Taxonomy, phylogeny and biogeography of francolins
(‘Francolinus’ spp.)
Aves: Order Galliformes
Family: Phasianidae
To
Thesis presented for the degree of
w
n
Tshifhiwa Gift Mandiwana-Neudani
DOCTOR OF PHILOSOPHY
ap
e
Faculty of Science
C
DST/NRF Centre of Excellence at the
of
Percy FitzPatrick Institute of African Ornithology
ty
Department of Biological Sciences
Private Bag X3
Rondebosch, 7701
South Africa
U
ni
ve
r
si
UNIVERSITY OF CAPE TOWN
Principal supervisor: Professor Timothy M. Crowe, DST/NRF Centre of Excellence at
the Percy FitzPatrick Institute of African Ornithology,
Department of Biological Sciences, University of Cape Town,
South Africa.
E-mail: timothy.crowe@uct.ac.za
Co-supervisor: Professor Rauri C.K. Bowie, Museum of Vertebrate Zoology and
Department of Integrative Biology,
University of California, Berkeley, USA.
E-mail: bowie@berkeley.edu
October 2013
n
of
C
ap
e
To
w
The copyright of this thesis vests in the author. No
quotation from it or information derived from it is to be
published without full acknowledgement of the source.
The thesis is to be used for private study or noncommercial research purposes only.
U
ni
ve
rs
ity
Published by the University of Cape Town (UCT) in terms
of the non-exclusive license granted to UCT by the author.
TABLE OF CONTENTS
Page(s)
Dedication
i
Declaration
ii
Acknowledgements
iii-viii
Abstract
ix-xi
Chapter 1
General introduction:
A review of the taxonomy, phylogeny and biogeography
of francolins (‘Francolinus’ spp.)…………………………………..1-36
Chapter 2
Phylogenetics of evolutionarily enigmatic terrestrial
gamebirds (Aves: Galliformes) with special regard to
‘Francolinus’ spp...............................................................................37-80
Chapter 3
A study of the gross morphological and histological
syringeal features of francolins and spurfowls in a
phylogenetic context………………………………………………81-103
Chapter 4
Taxonomic and phylogenetic utility of variation in
advertising calls of francolins and spurfowls (Galliformes:
Phasianidae)……………………………………………………...104-151
Chapter 5
Taxonomy and phylogeny of ‘true’ francolins………..................152-214
Chapter 6
Taxonomy and phylogeny of spurfowls…………………………215-273
Chapter 7
Historical biogeography of francolins and spurfowls……………274-319
Chapter 8
Summary and synthesis…………………………………………..320-327
Chapter 9
References………………………………………………………..328-363
DEDICATION
I dedicate this thesis to my parents who shared the same belief with Robert M.
Maciver who once said “…When you educate a man you educate an individual; when
you educate a woman you educate a whole family…” Without them, this work would
never have begun. To Enos, whose love and support made this journey possible, and
to my beloved Uhone whose flexibility and adaptability afforded the completion of this
work…
i
DECLARATION
I hereby declare that the work presented in this thesis is my own, unless otherwise
stated. Apart from the guidance received from my supervisors, assistance from all
institutions and individuals in this thesis is acknowledged. This thesis has not been
previously submitted for the degree at this or any other university and I therefore
present it for examination for the degree of PhD.
Signature Removed
10/03/2014
……………………………………….
………………………..............
Tshifhiwa G. Mandiwana-Neudani
Date
ii
ACKNOWLEDGEMENTS
Many people and institutions contributed to the success of this work in one way or the
other. First, there is of course Mrs B.P. Hall whose pioneering monograph on the
evolution of francolins published in 1963 laid key foundations for this research.
Second, I thank my supervisors, Professors Timothy M. Crowe and Rauri C.K.
Bowie, for helping to develop and extend her research on these fascinating birds. Their
financial support, mentorship and intellectual contributions were essential to the
completion of this academic journey.
I am indebted to the directors of institutions and heads and curators of collections
housed at the: American Museum of Natural History (Professor Joel L. Cracraft, Mr
Paul R. Sweet, Mr Peter Capainolo - Division of Vertebrate Zoology, Ornithology); the
Natural History Museum at Tring (Dr Robert Prys-Jones and Mark Adams –
Department of Zoology); Ditsong National Museum of Natural History (formerly
Transvaal Museum - Northern Flagship Institution) (former director Professor Francis
Thackeray, Ms Tamar Cassidy); Iziko Museums of Cape Town (Natural History) (Dr
Hamish Robertson, Ms Denise Hamerton – Terrestrial Vertebrates, Birds); and the
Field Museum of Natural History (Drs Shannon Hackett, John Bates and David E.
Willard - Division of Birds) for opening the doors of their collections allowing me to
personally examine the material under their care and their willingness to allow me to subsample toe pads for DNA analysis. I am also indebted to the University of Uganda, the
United States National Museum of Natural History, the Zoological Museum of the
University of Copenhagen, the Museum of Comparative Zoology (Harvard University),
the Los Angeles County Natural History Museum and the Louisiana State University
Natural History Museum for providing the specimens. Professor Rebecca Kimball
iii
provided key additional DNA sequences for a range of galliforms and wingshooters
provided key tissues.
The first sequences of DNA data were generated at the Molecular Ecology and
Evolution Programme (MEEP) laboratory at the University of Pretoria with the help of
the students who were using the laboratory at the time. Professor Paulette Bloomer,
the Director of the MEEP laboratory is thanked for generously affording me a space in
the laboratory and teaching me the basic techniques required for DNA sequencing.
Approximately 75% of the DNA samples sequenced in this project were subsampled from long-dead museum bird study skins, and this had implications with regard
to the technical expertise required to successfully sequence these samples. Dr Robert
G. Moyle and Ms Julie Feinstein (Division of Vertebrate Zoology, American Museum
of Natural History) played an important role in passing on their skills of sequencing
ancient DNA samples to me. I am very thankful to my friend and former classmate
Jack Henning, Professor Lorenzo Prendini (Division of Invertebrate Zoology,
American Museum of Natural History) and his wife Dr Elizabeth Scott who
generously accommodated me during my visits to the AMNH. The former manager of
the University of Cape Town (UCT) Department of Biological Sciences Systematics
laboratory Dr Tracey Nowell and Dr Pia Eldenäs, laboratory engineer of the
molecular laboratory (Swedish Museum of Natural History) are thanked for their help
with laboratory techniques. Professor Mari Källersjö and Dr J. Steven Farris
(molecular laboratory, Swedish Museum of Natural History) are thanked for making the
collaboration possible through the South African – Swedish Bilateral Partnership.
Recordings of vocalizations were sourced from various institutions including the
Transvaal Museum Sound Library, British Library Sound Archive and Macaulay
iv
Library of Natural Sounds. Michael Mills of Birding Africa is thanked for providing us
with the Gibbon’s (1995) and Chappuis’ (2000) collection and blood sample of
Swierstra’s Spurfowl. Callan Cohen (Birding Africa & Percy FitzPatrick Institute) is
thanked for passing on a recording of Jackson’s Spurfowl from Brian Finch of WINGS
birding. The analyses of calls would not have been possible without the help of Mr
Jean-pierre Richard of the Éthologie Animale et Humaine’ research group, Université
de Rennes, who taught me to use his software package ‘ANA’ (Jean-Pierre Richard,
Unpublished version) and to interpret sonograms. Dr Martine Hausberger, the director
of Éthologie Animale et Humaine’ research group (Université de Rennes) and Dr
Laurence Henry (Université de Rennes) are thanked for their useful comments on the
vocalization chapter of this thesis and also for handling all the logistics relating to my
trips to France and making my stay in Rennes enjoyable.
I am indebted to the wonderful contribution made by Dr Cecilia Kopuchian
(División Ornitología, Museo Argentino de Ciencias Naturales, Argentina) on the
chapter involving the anatomy of syringes. I owe this to Dr Pablo A. Goloboff
(Instituto Miguel Lillo, San Miguel de Tucumán, Argentina) who linked me with Dr
Kopuchian. She provided me with protocols to stain gross morphology of galliform
syringes and useful literature on syringeal-systematics work. She helped with the
interpretation of the observations made and also by making invaluable comments on the
syringeal draft manuscript. I am very grateful for the contribution made by Ms Morea
Peterson (Department of Human Biology, University of Cape Town) for her superb
technical assistance during the preparation of histological sections. Ms Barbara Moore
and Barbara Young (Department of Human Biology, UCT) also provided assistance
with histological procedures. I also would like to thank Professor Graham Louw
v
(Department of Human Biology, UCT) who helped with the interpretation of
histological sections and provided very useful comments on the histological component
of the chapter which focused on syringes of francolins and spurfowls.
Various gamebird hunters and Owen Davies (Percy FitzPatrick Institute of
African Ornithology, University of Cape Town) are also thanked for help with the
collection of whole specimens from which I extracted syringes and for providing hearts
for DNA-based research. I also would like to thank Ms Tamar Cassidy of Ditsong
National Museum of Natural History for providing me with the whole alcohol-preserved
specimens from which I extracted the syringes. Ms Anna Crowe, my fellow colleagues
at the FitzPatrick Institute, Graeme Oatley and Lisa Nupen are thanked for preparing
sub-samples of hearts of francolins and spurfowls on various farms in the Eastern,
Western Cape and Mpumalanga Province, and Owen Davies is thanked for his
technical involvement in the syringeal work.
J. Salvador Arias (Instituto Miguel Lillo, S.M. de Tucumán, Argentina) played
an important role in introducing me to the historical biogeographical analyses using his
spatial analysis of vicariance approach which was implemented in the software
Vicariance Inference Program (VIP). Professor Terry Hedderson (Department of
Biological Sciences, UCT) also helped with ancestral trait reconstruction analyses
which were performed on Mesquite and the interpretation of results thereof. Dr Fenton
Cotterill (AEON, UCT) provided illuminating discussions about the natural history of
key francolin species and the effects of topography and vegetation on francolin
biogeography. Dr Jérôme Fuchs, a post-doctoral fellow at the Percy FitzPatrick
Institute, UCT and Museum of Vertebrate Zoology, University of California, Berkeley
is thanked for the help he provided with some phylogenetic inference methods.
vi
Professor William Bond (Department of Biological Sciences, UCT) advised
me on aspects that relate to the vegetation map. I am also thankful to all my fellow
colleagues in the Department of Biological Sciences for consistently giving me support
through this journey. Ms Sandy Smuts (Department of Biological Sciences Senior
Administrator) is thanked for her administrative support and Mr Gonzalo Aquilar
(Department of Biological Sciences) for the technical support.
The administrative staff at the Percy FitzPatrick Institute is thanked for their
help in various ways. Ms Margaret Koopman (Niven librarian) is thanked for her fast
response and dedication in helping with literature. Mr Chris Tobler (Principal
Technical officer) is thanked for helping with computational technicalities. Ms Hilary
Buchanan and Tania Jansen are thanked for their quick response and help with all the
trips undertaken during this study and Mr Lionel Mansfield (former administrative
staff member) is thanked for ordering laboratory consumables when needed.
The financial support for this project was provided by the South African
Department of Science and Technology (DST)/National Research Foundation (NRF)
Centre of Excellence at the Percy FitzPatrick Institute of African Ornithology. My work
on vocalizations including visits to Rennes, France was funded through the South
African (NRF) - France (CNRS) Bilateral Programme while the syringeal component
was partly supported by CONICET PIP 112-200801-0074 through my collaboration
with Dr Kopuchian. The use of illustrations of francolins and spurfowls from
Handbook of the birds of the world: Vol. II. (del Hoyo et al. 1994) was allowed by
Lynx edicions, Barcelona.
I cannot forget to thank my friend Dr Tshifhiwa Nangammbi (Department of
Zoology, University of Venda) for always making sure that I stayed positive and
vii
acknowledging my effort at times when I could not see the light. Enos, your love and
support of all the time kept me alive. Uhone, you surprised me, you did not act like a
10-year old when you bowed to our earnest request that you needed to be away from me
so that I can complete this thesis. My sincere thanks to my sister-in-law Melta Neudane
who heartfully made sure that my house was always in order in my absence and
presence.
My parents played an immense role in shaping my future as I advanced in my
career. Their parenting role generated a sense of work and effort without which this
body of work would never have been completed.
viii
ABSTRACT
MANDIWANA-NEUDANI, T.G. 2013. Taxonomy, phylogeny and biogeography of
francolins (‘Francolinus’ spp.) Aves: Order Galliformes, Family Phasianidae. PhD
thesis, DST/NRF Centre of Excellence at the Percy FitzPatrick Institute of African
Ornithology, Department of Biological Sciences, University of Cape Town, Private Bag
X3, Rondebosch, 7701, South Africa.
Francolins (Francolinus spp.) are small to medium-sized, sedentary, Old World,
partridge/quail-like, terrestrial gamebirds (Order Galliformes) that occupy diverse
habitats ranging from dry/open/scrubby lowland and montane grasslands, bushveld and
savanna/woodland to mesic montane/lowland forests and forest edges. Some francolins
have complex distribution patterns and also are morphologically, ecologically and
behaviourally diverse. At the start of this research, Francolinus Stephens, 1819 was
considered a monophyletic galliform genus comprising 41 species (36 African and five
Asiatic) divided among eight putatively monophyletic species groups and four
taxonomically enigmatic species. However, different taxonomic revisions, especially
post Hall’s (1963) classic monograph, challenged the monophyletic status of the genus
and that of some of its designated species groups differed markedly in the number of
recognized subspecies. Furthermore, there was debate concerning the geographical
origin of the genus: Asia versus Africa. Some of the early molecular research on a few
exemplar francolin species based on partial mitochondrial Cytochrome-b DNA
sequences and Restriction Fragment Length Polymorphisms (RFLPs) also challenged
the monophyly of the genus and that of some of Hall’s (1963) species groups. These
findings suggested that francolins may form at least two distantly related lineages called
‘patryse’ (partridges) and ‘fisante’ (pheasants) by Afrikaans-speaking people. Patryse,
or ‘true’ francolins, had been divided into as many as five genera (Francolinus,
Ortygornis, Dendroperdix, Peliperdix, Scleroptila) and fisante, or spurfowls, all
grouped into a single genus, Pternistis. Research in this thesis is based on:
mitochondrial and nuclear DNA sequences (5554 base pairs), organismal and vocal
characters of francolins and spurfowls. Galliform terminal taxa rooted variably on
Anseriformes and Megapodes (Chapter 2), Gallus gallus and Bambusicola thoracica
(francolins) and Coturnix coturnix, Alectoris chukar (spurfowls) (Chapters 3, 4, 5, 6, 7).
The galliform phylogeny (Chapter 2), morphology of syringes (Chapter 3) and the
ix
vocalizations of many taxa (Chapter 4) confirm this phylogenetic dichotomy between
francolins and spurfowls and the Bayesian reconstruction of molecular divergence date
reveal they last shared a common ancestor at c. 33.6 mya. In addition, the taxonomic
and phylogenetic outcome of francolins (Chapter 5) and spurfowls (Chapter 6) suggest
the need to recognize two additional genera of francolins: Ortygornis for the Asian Grey
Francolin ‘Francolinus’ pondicerianus, Swamp Francolin, ‘Francolinus’ gularis (two
of four of Hall’s taxonomically enigmatic species), and the African Crested Francolin
‘Dendroperdix’ sephaena; and Afrocolinus gen. nov., for the African Latham’s
Francolin ‘Francolinus’ lathami (another of Hall’s taxonomically enigmatic species),
which links the basal francolins, Francolinus and Ortygornis spp. with relatively
terminal genera, Peliperdix and Scleroptila. The ‘true’ francolins, those related to the
nominate species, Black Francolin Francolinus francolinus, are monophyletic (having
originated at c. 7.6 mya), and sister to junglefowls (Gallus spp.) and Bamboo Partridges
(Bambusicola spp.), and are relatively small and have at most one (generally short)
tarsal spur positioned about half-way down the tarsus, give more musical and whistling
calls, generally roost on the ground, and have striped and barred rufous dorsal plumage
resembling that of quail. Spurfowls which originated at c. 8.7 mya are monophyletic,
but are distantly related to francolins. They are sister to the Bush Quail (Perdicula
asiatica) and the Sand Partridge (Ammoperdix heyi) and a range of Palaearctic and
African partridges and quails and are relatively large, often have two (generally long)
tarsal spurs (the lower of which is positioned about two-thirds down the tarsus), emit
raucous grating or cackling advertisement calls (given mainly at dawn/dusk). They have
been observed perching in trees and have dark dorsal plumage vermiculated with white
or buff. Thus, both francolins and spurfowls appear to have Asian, not African origins.
Four of eight of Hall’s (1963) groups were recovered and the phylogenetic hypotheses
for all four of Hall’s taxonomically enigmatic species are offered. The putative
taxonomic ‘link’ species between francolins and spurfowls, the Crested Francolin
Dendroperdix (now Ortygornis) sephaena, is shown to consistently group with two of
Hall’s unplaced Asian species O. pondicerianus and O. gularis.
Another important discovery was that one of the unplaced species, the African
Nahan’s Francolin Francolinus (now Ptilopachus) nahani is not a francolin, but is sister
x
to the African Stone Partridge Ptilopachus petrosus (Chapter 2), which in turn, are sister
to the New World quails (Odontophoridae).
xi
CHAPTER 1
A review of the taxonomy, phylogeny and biogeography of
francolins
Quail-like – ‘True’ Francolins – “patryse”
Francolinus francolinus
Ortygornis ‘Dendroperdix’ sephaena
Peliperdix coqui
Scleroptila shelleyi
1
Partridge-like Francolins – Spurfowls – “fisante”
Pternistis hartlaubi
Pternistis afer
Pternistis erckelii
Pternistis capensis
Pternistis squamatus
2
Nahan’s ‘Francolin’ ‘Acentrortyx’ ‘Francolinus’ nahani
3
Introduction – the status quo
Taxonomy - What are francolins?
Francolins (Table 1.1), at the initiation of this research, were classified as small to
medium-sized, sedentary, Old World, partridge/quail-like gamebirds adapted to varied,
but primarily tropical/sub-tropical, habitats ranging from dry, lowland grassland to
montane forests (Hall 1963, Johnsgard 1988, del Hoyo et al. 1994, Madge and
McGowan 2002). At the outset of this research, taxonomically, francolins were
invariably placed in the Order Galliformes and family Phasianidae. In some more finely
partitioned classifications, within the Phasianidae, they were placed in the sub-family
Phasianinae (including Phasianus and related species) and, with other Old World
partridge- and quail-like gamebirds (e.g. Perdix and Coturnix spp.), in the tribe
Perdicini (Chapin 1932, Peters 1934, Wolters 1975-82, Crowe et al. 1986, Johnsgard
1988, Sibley and Monroe 1990, del Hoyo et al. 1994, Madge and McGowan 2002).
Although relatively stable at higher taxonomic levels, there remains controversy with
respect to the number and limits of subspecies, species and genera of francolins.
A brief respite
The distributions of 36 francolin species are confined to Africa. Those of five further
species extend outside of Africa north to the Caspian Sea and east to central Asia and
southern China (Johnsgard 1988). Based on the findings of B.P. Hall’s (1963)
monograph, francolins were (and generally still are) lumped into a single perdicine
genus, Francolinus, comprising 41 species (del Hoyo et al. 1994). Francolinus
Stephens, 1819 is currently the most specious genus in the Galliformes (Morony et al.
4
1975, del Hoyo et al. 1994) and one of the largest genera in the class Aves (Bock and
Farrand 1980). However, other taxonomic treatments have partitioned francolins into as
many as eight genera (Francolinus, Pternistis, Peliperdix, Scleroptila, Dendroperdix,
Ortygornis, Acentrortyx, and Chaetopus) (Roberts 1924, Chapin 1932, Peters 1934,
Mackworth-Praed and Grant 1952, 1962, 1970, Hall 1963, Wolters 1975-82, Hockey et
al. 2005). Nevertheless, virtually immediately after lumping francolins into a single
genus, Hall (1963) hypothesized that all but four Francolinus spp. could be assigned to
eight putatively monophyletic groups (Table 1.1).
Morphology and behaviour
Francolins, like many perdicine gamebirds, are characterized by having 14 tail feathers
that are presumably moulted centrifugally (Johnsgard 1988). Some more partridge-like
species of francolins are: relatively large, often have two (generally long) tarsal spurs
(positioned about two-thirds down the tarsus), emit raucous grating or cackling
advertisement calls (given mainly at dawn/dusk). They have been observed perching in
trees or other elevated structures and have dark dorsal plumage vermiculated and/or
striped with white or buff (Crowe et al. 1986, Milstein and Wolff 1987, Crowe et al.
1992). Other more quail-like francolins are relatively small, have at most one (generally
short) tarsal spur (positioned about half-way down the tarsus), have relatively musical
and whistling calls, generally roost on the ground, and have striped and barred rufous
dorsal plumage resembling that of quails (Coturnix spp.) (Crowe et al. 1986, Milstein
and Wolff 1987). This has led the Afrikaans-speaking people to group francolins into
two supra-specific groupings; fisante (or pheasants) for the former and patryse (or
partridges) for the latter (Milstein and Wolff 1987).
5
However, the above distinction was difficult to apply to the Crested Francolin,
Dendroperdix sephaena, since it has attributes of both fisante and patryse. It is
relatively small, has quail-like back plumage like patryse, but emits a raucous call, has
well-spurred, red tarsi and roosts in trees like many fisante. Thus, it is a ‘linking form’
that persuaded Hall to lump all 41 species into a single genus Francolinus.
The tortuous taxonomic history of francolins
The genus Francolinus was first described by Stephens in 1819. Since that time, various
taxonomic treatments and revisions have attempted to partition the francolins into two
or more genera, a variable numbers of species and a plethora of subspecies (Table 1.2).
However, Hall’s (1963) monograph still remains the standard reference as far as the
classification of francolins is concerned (Johnsgard 1988, Sibley and Monroe 1990, del
Hoyo et al. 1994). Table 1.3 attempts to summarize the taxonomic chaos while Table
1.4 outlines the types of characters that have been used in various treatments.
Alternative taxonomic treatments
Roberts (1924), in his checklist of birds of South Africa, recognized six genera of
francolins (Table 1.2):
1. Pternistis Wagler, 1832 -- the species swainsonii, afer, humboldtii and
castaneiventer were recognized, the last of which was divided into three
subspecies castaneiventer, notatus and krebsi.
2. Chaetopus Swainson, 1837 -- capensis, adspersus and natalensis.
3. Peliperdix Bonaparte, 1856 -- hartlaubi.
6
4. Dendroperdix Roberts, 1922 -- rovuma (recognized by most avian taxonomists
as a subspecies) and sephaena.
5. Scleroptila Blyth, 1849 -- afra (= afra), levaillantii, shelleyi and gariepensis (=
levaillantoides).
6. Ortygornis Reichenbach, 1853 -- coqui.
Roberts, in his first edition (1940) of ‘The Birds of South Africa’ followed the
above-mentioned checklist, recognizing the same genera as in Roberts (1924), with just
one change in which he erected the genus Chapinortyx Roberts, 1928 to distinguish the
species hartlaubi. The same species as those in the 1924 checklist were retained, but
additional subspecies were delineated. Within Ortygornis coqui, three subspecies were
recognized (coqui, campbelli, vernayi); six within Dendroperdix sephaena (the
nominate sephaena, zuluensis, zambesiae, chobiensis, mababiensis, thompsoni); and six
within Scleroptila levaillantoides (levaillantoides, gariepensis, ludwigi, pallidior, langi,
kalaharica). In addition to the above species, Scleroptila jugularis was delineated
within which he recognized one subspecies cunenensis; four subspecies were
recognized for Pternistis swainsonii (swainsonii, damarensis, chobiensis, gilli) and two
subspecies for Pternistis afer (afer, cunenensis). A further subspecies lehmanni, was
recognized for Pternistis castaneiventer.
In the 2nd edition of ‘Roberts Birds of Southern Africa’ (McLachlan and
Liversidge 1957), the number of genera recognized was reduced to two (Francolinus
and Pternistis). Francolinus was assigned to all partridges and fisante (including
hartlaubi plus the three subspecies hartlaubi, bradfieldi, crypticus) that do not belong to
7
Hall’s Bare-throated Group, whereas Pternistis was assigned strictly to the Barethroated taxa (Table 1.2).
Even though the 1st and the 2nd editions covered the same geographic area, the
delineation of taxa continued to vary with the lumping and splitting of previously
recognized taxa and the erection of new taxa. This resulted in great inconsistency and
instability in the number of taxa recognized. The following taxa were typically
recognized: subspecies coqui, vernayi, with the addition of hoeschianus and the removal
of campbelli within F. coqui; subspecies mababiensis and chobiensis were synonymized
with F. sephaena; only five subspecies were recognized within F. levaillantoides
(levaillantoides, pallidior, langi, wattii and cunenensis which was considered a
subspecies of S. jugularis) leaving out gariepensis, ludwigi, kalaharica; subspecies gilli
was synonymized with swainsonii; and notatus, swynnertoni and humboldti were
atributed to P. afer. The subspecies lehmanni was removed from P. castaneiventer.
The 3rd and 4th (McLachlan and Liversidge 1970, 1978 respectively) editions
were similar to the 2nd edition with regard to the taxa which were recognized.
Subspecies lundazi was recognized as an additional subspecies of P. swainsonii, and
gilli was re-considered, whereas chobiensis was synonymized with swainsonii.
Subspecies lehmanni was re-considered as a valid subspecies of P. afer.
The 5th and the 6th editions (Maclean 1985, 1993 respectively) covered the
southern African region and recognized the same set of species, the only major
difference being that the 5th edition used the genus Francolinus for both patryse and
fisante and that there was no mention of subspecies with the exception that it is
8
indicated that rovuma was considered a subspecies of D. sephaena. The 6th edition
made use of the genera Francolinus and Pternistis as in the previous editions and had
the same composition of species as in the 4th edition.
The most recent 7th edition of Roberts’ Birds of Southern Africa (Hockey et al.
2005), recognized four genera:
1. Peliperdix -- coqui within which they recognized the subspecies coqui, which
included campbelii, hoeschianus, lynesi, kasaicus, ruahdae, stuhlmanni, vernayi
and angolensis.
2. Dendroperdix -- sephaena comprised of sephaena which included the
subspecies zuluensis, rovuma and zambesiaie (synonyms of which are
thompsoni, chobiensis and mababiensis).
3. Scleroptila -- afra, levaillantii, shelleyi within which they recognized subspecies
shelleyi (synonym sequestris) and canidorsalis; levaillantoides within which
they recognized the subspecies levaillantoides (synonyms ludwigi, langi and
kalaharica), jugularis (synonyms cunenensis and stresemanni) and pallidior
(synonym wattii).
4. Pternistis -- hartlaubi, adspersus (subspecies - kalahari and mesicus), capensis,
natalensis, afer (subspecies - afer, swynnertoni, castaneiventer) and swainsonii
(subspecies - swainsonii and lundazi).
Chapin (1932) divided francolins between two genera Pternistis and
Francolinus:
1. Pternistis -- included the three species afer, rufopictus and swainsonii
9
2. Francolinus -- this genus was assigned to all the other francolins with the
exception of nahani. Francolinus nahani was described by (Dubois, 1905) and
moved to Peliperdix by van Someren (1926). During the same year, Chapin
(1926) suggested nahani be placed in a monotypic genus, Acentrortyx, as he
argued that this perdicine was not congeneric with Francolinus.
Peters (1934) recognized 39 species and split the francolins between two genera:
Francolinus and Pternistis. However, he retained the use of the genus Francolinus in
the broader sense, pending a detailed revision of all the species. Peters’ deliniation of
Francolinus included members of Hall’s (1963) Spotted, Montane, Scaly,
Vermiculated, Red-winged, Striated and Red-tailed Groups. The genus Pternistis was
restricted to members of Hall’s Bare-throated Group. He recognized ‘schlegelii’ as a
subspecies of Francolinus coqui. Another Peters’ ‘peculiarity’ is that he elevated
Francolinus castaneicollis ‘atrifrons’ to full species status, whereas other authors
considered this taxon a subspecies of castaneicollis.
Mackworth-Praed and Grant (1952, 1962, 1970) also recognized two genera:
1. Pternistis -- rufopictus; leucoscepus (subspecies leucoscepus, infuscatus, muhamedben-abdullah, kilimensis); swainsonii (subspecies swainsonii, damarensis, chobiensis,
gilli), and afer (subspecies afer, cranchii, humboldtii, benguellensis, intercedens,
castaneiventer, krebsi, swynnertoni, notatus, lehmanni, cunenensis, loangwae, harterti,
nyanzae, leucoparaeus, bohmi, melanogaster, itigi).
2. Francolinus -- sephaena (subspecies sephaena, grantii, zambesiae); streptophora;
and rovuma (which they elevated to species, subspecies rovuma and spilogaster);
10
squamatus (subspecies squamatus, schuetti, doni, maranensis, usambarae, uzungwensis,
chyuluensis); griseostriatus and ahantensis (subspecies ahantensis, hopkinsoni); nobilis
(subspecies nobilis, chapini); castaneicollis (subspecies castaneicollis, gofanus,
ogoensis, atrifrons, kaffanus); jacksoni (subspecies jacksoni, gurae, pollenorum);
erckelii
(subspecies erckelii, pentoni); swierstrai; camerunensis; hildebrandti
(subspecies hildebrandti, altumi, johnstoni); harwoodi; bicalcaratus (subspecies
bicalcaratus, thornei, adamauae, ogilvie-granti, ayesha); clappertoni (subspecies
clappertoni,
gedgii,
heuglini,
cavei,
sharpii,
konigseggi,
nigrosquamatus);
icterorhynchus (subspecies icterorhynchus, emini); adspersus; capensis; natalensis
(subspecies natalensis, neavei); hartlaubi (subspecies hartlaubi, bradfieldi, crypticus);
coqui (subspecies coqui, angolensis, vernayi, hoeschianus, hubbardi, ruahdae, thikae,
maharao, kasaicus, spinetorum); schlegelii; albogularis (subspecies albogularis,
buckleyi, dewittei); finschi; levaillantoides (subspecies levaillantoides, gariepensis,
jugularis, pallidior, ludwigi, langi, cunenensis, kalaharica, wattii); levaillantii
(subspecies levaillantii, kikuyuensis, crawshayi, benguellensis, clayi); shelleyi
(subspecies shelleyi, whytei, theresae, elgonensis, uluensis, macarthuri); afer (=
africanus = afra) and the unplaced nahani and lathami (subspecies lathami, schubotzi).
Mackworth-Praed and Grant (1952, 1962, 1970) demonstrate the taxonomic
problems created by their variable use of the name ‘afer’ at both the specific and
subspecific level: ‘afer’ was used in their 1952 publication to refer to Grey-winged
Francolin, currently deliniated as Scleroptila afra (they described this species’
distribution as ranging from the Cape to Ethiopia, but with interruptions) and ‘cranchii’
for Red-necked Spurfowl currently Pternistis afer. The Red-necked Spurfowl is widely
11
distributed in the lowlands of central and east Africa, from Gabon, western Angola,
northwest Namibia, and Zimbabwe to South Africa. However, in 1962 they used ‘afer’
for both the Grey-winged Francolin and Red-necked Spurfowl. Currently, the specific
name ‘afer’ is used for Red-necked Spurfowl Pternistis afer whereas ‘africanus/afra’ is
used for the Grey-winged Francolin Scleroptila afra.
Furthermore, within the Grey-winged Francolin (presently considered a South
African and Lesotho endemic – Clancey 1986, Sinclair and Ryan 2003), MackworthPraed and Grant (1952, 1962, 1970) included several taxa which occur in sub-Saharan
Africa. These were uluensis, lorti, gutturalis, psilolaema, ellenbecki, archeri,
friedmanni, marcathuri and stantoni. Other authors (Table 1.2) assigned uluensis and
macarthuri to Shelley’s Francolin F. shelleyi, and gutturalis, archeri and friedmanni to
the Orange River francolin F. levaillantoides. While they considered psilolaema a
subspecies of Grey-winged Francolin, other authors (Table 1.2) recognized this taxon to
be a full species of the Red-winged Group with ellenbecki being one of its subspecies.
Wolters (1975-82) proposed the first cladistic system of genera and subgenera
for francolins in which he recognized Hall’s (1963) 41 species. He recognized six
genera:
1. Francolinus -- this genus included Hall’s Spotted taxa francolinus, pintadeanus,
pictus and her unplaced gularis, with francolinus, pintadeanus, pictus (except
gularis) lumped into a subgenus, Francolinus gularis was not assigned to any
subgenus.
2. Pternistis -- Hall’s Bare-throated leucoscepus, rufopictus, swainsonii and afer
(all placed in the subgenus Pternistis) as well as the Montane species erckelii,
12
ochropectus, castaneicollis, jacksoni, camerunensis and swierstrai (not placed in
any subgenus), the Scaly squamatus, griseostriatus and ahantensis (no
subgenus), adspersus and capensis (without subgenus). Wolters also assigned
nahani to genus Pternistis, placing it in subgenus Acentrortyx following Chapin
(1926). Also, placed in the genus Pternistis were the Vermiculated natalensis,
hildebrandti, harwoodi, hartlaubi, bicalcaratus, icterorhynchus and clappertoni,
all within the subgenus Chaetopus.
3. Scleroptila -- Hall’s Red-winged afra (= africanus), finschi, shelleyi,
psilolaema, levaillantii and levaillantoides.
4. Dendroperdix -- Hall’s Striated sephaena and streptophora.
5. Peliperdix -- Hall’s Red-tailed albogularis, schlegelii, coqui and the unplaced
lathami.
6. Ortygornis -- this genus was assigned to the unplaced pondicerianus.
Crowe and Crowe (1985) questioned, but did not refute, the monophyletic status
of the genus Francolinus as defined in Hall (1963). However, they retained this genus
for all the species and suggested a system of subgenera as follows: subgenus
‘Francolinus’ included the Spotted species francolinus, pictus, pintadeanus;
‘Ortygornis’ was assigned to the unplaced pondicerianus; ‘Scleroptila’ for all the Redwinged species finschi, levaillantii, afra, psilolaema, shelleyi, and levaillantoides that
also incorporated one species from Hall’s Striated Group streptophora; ‘Peliperdix’ for
the unplaced lathami; ‘Dendroperdix’ for the Striated sephaena; ‘Pternistis’ for the
Bare-throated afer, swainsonii, leucoscepus and rufopictus; and the subgenus
‘Acentrortyx’ for nahani. Within the subgenus Peliperdix, they assigned only lathami.
13
No subgenera were assigned to coqui, albogularis, schlegelii, gularis, griseostriatus,
ahantensis, squamatus, adspersus, hildebrandti, natalensis, hartlaubi, capensis,
castaneicollis, jacksoni, erckelii, ochropectus, nobilis, camerunensis, swierstrai,
harwoodi, icterorhynchus, clappertoni and bicalcaratus.
Crowe and Crowe (1985) also modified Hall’s (1963) classification, by
suggesting that F. streptophora should be moved from the Striated to the Red-winged
Group based on the cladistic analysis of organismal characters. They further maintained
that the two Montane Group species F. camerunensis and F. swierstrai form a second
superspecies (= a set of species with ranges that do not overlap) within the Montane
Group. However, they supported Hall’s suspicion that lathami has affinities with
members of the Red-tailed Group whereas nahani was thought to have affinities with
members of the Scaly Group.
Milstein and Wolff (1987) adopted Wolters’ (1975-82) classification system in
their work on the southern African francolins; they formally recognized the genus
‘Pternistis’ which included the Bare-throated species swainsonii and afer, and the
Vermiculated adspersus, capensis and natalensis. They assigned the genus ‘Scleroptila’
to the Red-winged afra, shelleyi, levaillantoides, and levaillantii whereas ‘Peliperdix’
and ‘Dendroperdix’ were assigned to a Red-tailed species coqui and the Striated species
sephaena, respectively.
Despite all this taxonomic debate, Crowe et al. (1986) and Sibley and Monroe
(1990) recognized only one genus ‘Francolinus’ with 41 species within it. They
14
recognized subspecies francolinus, arabistanicus, henrici, asiae, melanonotus within F.
francolinus; pallidus, pictus, watsoni within F. pictus; phayrei, pintadeanus within F.
pintadeanus; pondicerianus, mecranensis, interpositus within F. pondicerianus.
Francolinus gularis was recognized as a monotypic species. Francolinus lathami
consisted of two subspecies, lathami and schubotzi. Four subspecies were recognized
within F. coqui (spinetorum, maharao, hubbardi, coqui) with thikae being included in
maharao; buckleyi in spinetorum; angolensis, campbelli, ruahdae, lynesi, vernayi,
hoeschianus and kasaicus were all included in the nominate coqui.
Three subspecies albogularis, buckleyi and dewittei were recognized within F.
albogularis. Francolinus schlegelii, F. streptophora, F. afra, F. finschi and F. nahani
were recognized as monotypic species. Two subspecies, kikuyuensis and levaillantii
were recognized within F. levaillantii; subspecies shelleyi (which included uluensis,
canidorsalis, sequestris) and whytei were recognized within F. shelleyi; psilolaema
(which included ellenbecki) and elgonensis (included theresae) within F. psilolaema;
gutturalis, lorti (which included archeri), jugularis (which included cunenensis),
levaillantoides (which included pallidior, kalaharica) within F. levaillantoides;
sephaena, rovuma, spilogaster, zambesiae and grantii were recognized within F.
sephaena. Francolinus squamatus, F. ahantensis, F. griseostriatus, F. hartlaubi, F,
icterorhynchus, F. clappertoni, F. harwoodi, F. adspersus, F. capensis, F. natalensis, F.
hildebrandti, F. leucoscepus, F. rufopictus were all recognized as monotypic species.
The two subspecies ayesha and bicalcaratus were recognized for F. bicalcaratus.
Francolinus swierstrai, F. camerunensis, F. nobilis, F. jacksoni, F. ochropectus,
F. erckelii were recognized as monotypic species. Two subspecies were recognized
within F. castaneicollis, castaneicollis (which included bottegi, ogoensis, kaffanus,
15
gofanus) and atrifrons; lundazi and swainsonii (included damarensis, chobiensis, gilli)
within F. swainsonii; harterti, afer, cranchii (intercedens), leucoparaeus, melanogaster
(included loangwae), swynnertoni and castaneiventer (included notatus, lehmanni) were
recognized within F. afer.
Crowe et al. (1992) again challenged the status of monophyly of Francolinus
and proposed a system of genera and subgenera. Genus ‘Francolinus’ included the
Spotted pictus, francolinus, pintadeanus and were all placed in the subgenus
Francolinus; the unplaced pondicerianus was then placed in the subgenus Ortygornis
and gularis in the subgenus Limnocolinus (nov.). Genus ‘Peliperdix’ included the Redtailed coqui, albogularis and schlegelii, and the unplaced lathami, were all placed in the
subgenus Peliperdix; the Striated species sephaena was assigned to subgenus
Dendroperdix, but placed in genus Peliperdix. Genus ‘Scleroptila’ included the Striated
species streptophora and the Red-winged finschi, levaillantii, afra, psilolaema, shelleyi,
and levaillantoides.
The genus ‘Pternistis’ included nahani which was placed in the subgenus
Acentrortyx; the Vermiculated hartlaubi was placed in subgenus Chapinortyx; the other
Vermiculated species such as bicalcaratus, icterorhynchus, clappertoni and harwoodi
were placed in subgenus Chaetopus; the balance of the Vermiculated Group, adspersus,
capensis, natalensis, hildebrandti were placed in subgenus Notocolinus (nov.); the
Scaly
squamatus,
ahantensis,
and
griseostriatus
were
placed
in
subgenus
Squamatocolinus; the Bare-throated leucoscepus, rufopictus, afer, swainsonii in
subgenus Pternistis; the Montane jacksoni, nobilis, camerunensis, swierstrai,
castaneicollis, erckelii, ochropectus were all placed in subgenus Oreocolinus (nov.).
16
Despite the newly proposed taxonomic classification, del Hoyo et al. (1994)
followed Hall’s (1963) classification system and adopted the genus Francolinus (sensu
lato) for all the species, but did not comment on its monophyly. del Hoyo’s species and
subspecies delineations were similar to those in Crowe et al. (1986) with the exception
that no subspecies were attributed to castaneiventer.
Dickinson (2003) did not deviate much from Crowe et al. (1986) with regard to
genera, species and subspecies that were recognized. He attributed 41 species to the
genus genus Francolinus. Within F. francolinus, subspecies francolinus (which
included billypayni), arabistanicus, bogdanovi (included festinus), henrici, asiae
(including parkerae), melanonotus were recognized; pallidus, pictus, watsoni were
delineated within F. pictus; phayrei, pintadeanus within F. pintadeanus; pondicerianus
(included ceylonensis), mecranensis, interpositus (included prepositus) within F.
pondicerianus. Francolinus gularis (which included ridibundus) and F. lathami were
recognized as monotypic species. Four subspecies were recognized within F. coqui
(spinetorum, maharao, hubbardi, coqui) with thikae being included in maharao;
angolensis, campbelli, ruahdae, lynesi, vernayi, bourqoii, hoeschianus and kasaicus
were all included in the nominate coqui. The taxa albogularis, buckleyi, dewittei
(included meinertzhageni) were recognized within F. albogularis; F. schlegelii
(included confusus); F. streptophora; F. afra (which included proximus); F. finsch; F.
griseostriatus and F. nahani were recognized as monotypic species.
Three subspecies kikuyuensis (which included benguellensis, clayi), crawshayi,
and levaillantii were recognized within F. levaillantii; subspecies shelleyi (which
included uluensis, canidorsalis, sequestris) and whytei within F. shelleyi; psilolaema
17
(which included ellenbecki) and elgonensis (included theresae) within F. psilolaema;
gutturalis (included eritreae), archeri (included friedmanni, stantoni), lorti, jugularis
(included cunenensis, stresemanni), pallidior (included wattii) and levaillantoides
(included kalaharicus) within F. levaillantoides; sephaena (included zuluensis),
rovuma, spilogaster (included somaliensis), zambesiae (included thompsoni) and grantii
were recognized within F. sephaena; squamatus (included whytei), schuetti (included
tetraoninus), zappeyi, maranensis (included chyuluensis), usambarae, uzungwensis and
doni within F. squamatus; hopkinsoni and ahantensis were recognized within F.
ahantensis; F. hartlaubi was represented by three subspecies (hartlaubi, bradfieldi,
crypticus); altumi, hildebrandti (included helleri, fischeri), johnstoni (included grotei)
within F. hildebrandti; natalensis (included thamnobium) and neavei within F.
natalensis; ayesha, bicalcaratus (which included adamauae), thornei and ogilviegranti
(included molunduensis) within F. bicalcaratus; clappertoni, sharpii, heuglini,
konigseggi, gedgii (included cavei) and nigrosquamatus (included testis) within F.
clappertoni; F. harwoodi; F. capensis; F. swierstrai; F. camerunensi; F. nobilis
(included chapini); F. jacksoni (included pollenorum); F. ochropectus; F. erckelii; F.
rufopictus and F. leucoscepus were recognized as monotypic species. Two subspecies
were recognized within F. adspersus, adspersus (included kalahari) and mesicus;
castaneicollis (included bottegi, gofanus), ogoensis, kaffanus (included patrizii) and
atrifrons within F. castaneicollis; lundazi and swainsonii (included damarensis, gilli)
within F. swainsonii; harterti, afer (included chio, palliditectus), cranchii (boehmi,
punctulatus, benguellensis, intercedens, nyanzae, itigi), leucoparaeus, melanogaster
(included loangwae, aylwinae, tertius) swynnertoni and castaneiventer (included
notatus) were recognized within F. afer.
18
Phylogenetics
Hall (1963) presented the first phylogenetic hypothesis for francolins (Fig. 1.1).
Bloomer and Crowe (1998) provided the first DNA-based evidence (Fig. 1.2) to suggest
a lack of monophyly for Francolinus as circumscribed by Hall, but offered no
taxonomic recommendations. However, their molecular phylogeny provided support for
splitting the francolins into what is today referred as two distinct clades, the ‘francolins’
and ‘spurfowls’.
Highlights of some key taxonomic disagreements:
It is clear that taxonomic confusion exists, not just among taxonomists, but also
between taxonomists and gamebird enthusiasts and hunters. Mackworth-Praed
and Grant (1952, 1962, 1970) in all their reviews, split francolins into what they
called ‘francolins’ and ‘spurfowls’. They assigned the genus Pternistis to the
spurfowl taxa belonging to Hall’s (1963) Bare-throated Group, and the genus
Francolinus to all African francolins including the balance of the spurfowl
groups identified in Hall’s monograph (Table 1.1). The results of Bloomer and
Crowe (1998) led to the split of francolins into two major lineages (Fig. 1.2)
which they called the quail-francolin (or partridges)’ and partridge-francolin (or
pheasants) clades. The common names francolin and spurfowl will be used
hereafter, since Bloomer and Crowe (1998) found them to be evolutionarily
distantly-related from each other.
The evolutionary affinities of the Crested Francolin Ortygornis sephaena have
been debated given its possession of features that are found in both francolins
19
and spurfowls. Hall (1963) considered this species to be a 'linking form' and as a
consequence placed all 41 species into a single genus Francolinus.
Confusion has surrounded the number of genera recognized among francolins
and spurfowls. The earlier revisions (Roberts 1924, 1940) included seven genera
(Pternistis, Chaetopus, Peliperdix, Chapinortyx, Dendroperdix, Scleroptila,
Ortygornis), which were reduced to either two (Francolinus and Pternistis) in
Chapin (1932), Peters (1934), Mackworth-Praed and Grant (1952), McLachlan
and Liversidge (1957), Mackworth-Praed and Grant (1962), Mackworth-Praed
and Grant (1970), McLachlan and Liversidge (1970) or just one genus
Francolinus (Crowe and Crowe 1985, Maclean 1985, Crowe et al. 1986,
Maclean 1993) in several later revisions (Table 1.3). More recently, Crowe et al.
(1992) suggested that some earlier genera were indeed valid (Francolinus,
Dendroperdix, Peliperdix, Scleroptila and Pternistis). It should be noted that
based on Crowe et al. (2006) and the results emanating from this study, the
number of genera recognized within ‘francolins’ might see an increase, but are
expected to remain stable beyond the completion of this study.
The lack of stability and consistency in the subspecies recognized and several
revisions of the species complexes, even when similar evidence and geographic
areas were covered, has clearly demonstrated the profound difficulties with
respect to using morphological characters to delineate taxa.
Francolinus
nahani
has
remained
something of
an
enigma
among
francolins/spurfowl with a great deal of uncertainty pertaining to its
phylogenetic placement. Van Someren (1926) proposed the genus Peliperdix for
20
this species, and Chapin (1926) had during the same year proposed to move
nahani to the genus Acentrortyx. Chapin argued that this partridge was not
congeneric with the type species of the genus of francolins Francolinus
francolinus. However, subsequent revisions placed this species back in the
genus Francolinus (Peters 1934, Hall 1963, Mackworth-Praed and Grant 1970).
Wolters (1975-82) and Crowe et al. (1992) placed nahani within the genus
Pternistis in subgenus Acentrortyx. Most recently nahani has been moved from
Francolinus to ‘Ptilopachus’, a clade of two taxa, the other being the Stone
Partridge Ptilopachus petrosus which intriguingly forms the sister clade to New
World quails Odontophoridae (Crowe et al. 2006, Cohen et al. 2012).
The genus Francolinus - African or Asian in origin?
The geographical origin of the genus ‘Francolinus’ has remained controversial (Hall
1963, Crowe and Crowe 1985). Two hypotheses are postulated. The first based on the
suggestion that the genus Francolinus shares its closest affinities with other Palaearctic
and Asian genera. Hall (1963) strongly argued for the genus to be of Asian origin with
its age being traced back to the Oligocene +-25-35. She further alluded to reduced
competition as being the driving factor for the speciation of francolins in Africa.
Although Crowe and Crowe (1985) concurred with Hall that the ancestor was quail-like,
they hypothesized an African origin for the genus, with Asia being colonized by a
nomadic or migratory ancestor that diversified and became sedentary. The reported age
of the divergence francolins from Old World quail Coturnix spp. was traced to the
Pliocene (approx. 4.5-5 mya) using Shields and Wilson’s (1987) calibration of 2%
divergence per million years (Crowe 1992). Bloomer and Crowe (1998) recovered a
21
divergence date between the quail versus the partridge-like lineages at 5 and 8 mya,
whereas the transversion rate from suggested a divergence time of 3-6 mya using the
using Shields and Wilson’s (1987) conventional calibration of 2% per million years
implemented on mitochondrial Cytochrome-b data. They believe that the minimum
divergence estimates appear to be corroborated by the availability of well-differentiated
francolin fossils (Crested Francolin Ortygornis sephaena; Crowe 1992). However, as
hinted at (although not actually calculated) in Crowe et al. (2006), the divergence of
francolins and spurfowl may be considerably older.
Conclusions
The taxonomic review detailed above demonstrates that the confusion over the number
of genera and terminal taxa of francolins and spurfowls is profound and that what is
needed are characters from a range of sources that can provide solid and objective
explanations for every taxonomic decision that is taken. This is especially important
given the enormous morphological variation observed within certain species complexes,
which has generated considerable debate with respect to the delimitation of species and
subspecies. Defensible and robust analytical methods and approaches also need to be
adopted to enable objective conclusions to be drawn.
The major aims of this study
On the basis of the confusion/disagreement/hypotheses outlined above regarding: (1)
debate on the monophyly of the genus ‘Francolinus’, (2) the monophyly and
evolutionary relationships of, and within the different species-groups suggested by Hall
(1963), (3) the instability/inconsistency of the number of terminal taxa (species,
22
subspecies and genera) recognized, and (4) the dispute on where the Francolinus clade
first diversified – Africa or Asia - this study aimed at achieving the following:
1. To undertake a general review of the taxonomy of ‘francolins’ (taking views
largely – but not exclusively – of Mackworth-Praed and Grant (1952, 1962,
1970) and Hall 1963 as alternative hypotheses) [Chapter 1].
2. To contrast alternative hypotheses concerning the monophyly of the genus
‘Francolinus’ (e.g. Hall 1963 versus Bloomer and Crowe 1998) and to further
explore DNA-based, vocalization and behavioural evidence in order to test the
phylogenetic affinities of the Stone Partridge Ptilopachus petrosus and Nahan’s
Francolin Francolinus nahani as suggested in Crowe et al. (2006) [Chapter 2].
3. To investigate if there are any syringeal features (anatomical) that can be used to
further our understanding of the phylogenetic relationships among francolins
and spurfowls [Chapter 3].
4. To test the validity of the francolin-spurfowl taxonomic dichotomy based on
vocal characters [Chapter 4].
5. To test Hall’s (1963) hypotheses on the monophyly of the suggested eight
species groups within her circumscription of the the genus Francolinus, and
thereby provide a modern systematic revision of the terminal taxa and genera of
francolins and spurfowls [Chapter 5 & 6].
6. To describe and explore patterns of distribution of francolins and spurfowls,
using Hall’s (1963) monograph as a null hypothesis [Chapter 7].
Tables and Figures
23
Table 1.1. List of francolin and spurfowl species and their designated species groups
recognized by Hall (1963). English and specific names are as proposed in the IOC list
(Gill and Donsker 2013) and the generic classification follows Crowe et al. (2006) and
Mandiwana-Neudani, this thesis. Genus Francolinus Stephens, 1819, Peliperdix
Bonaparte, 1856, Scleroptila Blyth, 1849, Dendroperdix Roberts, 1922, Ortygornis
Reichenbach, 1852, Pternistis Wagler, 1832, Chaetopus Swainson, 1837, Afrocolinus
gen. nov. Mandiwana-Neudani, this thesis. Taxon authority citation with a year in
parentheses indicates that either the specific or the subspecific epithet was originally
published in another genus by the first author, but moved to the present genus by the
second revising author.
Species group
English name
Scientific name
1. Spotted
Black Francolin
Painted Francolin
Chinese Francolin
Francolinus francolinus (Linnaeus,
1766)
F. pictus (Jardine & Selby, 1828)
F. pintadeanus (Scopoli, 1786)
2. Bare-throated
Red-necked Spurfowl
Swainson’s Spurfowl
Yellow-necked Spurfowl
Grey-breasted Spurfowl
Pternistis afer (Müller, 1776)
P. swainsonii (Smith, 1836)
P. leucoscepus (Gray, 1867)
P. rufopictus Reichenow, 1887
3. Montane
Erckel’s Spurfowl
Djibouti Spurfowl
Chestnut-naped Spurfowl
Jackson’s Spurfowl
Handsome Spurfowl
Mount Cameroon Spurfowl
Swierstra’s Spurfowl
P. erckelii (Rüppell, 1835)
P. ochropectus (Dorst & Jouanin,
1952)
P. castaneicollis Salvadori, 1888
P. jacksoni O. Grant, 1891
P. nobilis Reichenow, 1908
P. camerunensis Alexander, 1909
P. swierstrai (Roberts, 1929)
4. Scaly
Ahanta Spurfowl
Scaly Spurfowl
Grey-striped Spurfowl
P. ahantensis Temminck, 1854
P. squamatus Cassin, 1857
P. griseostriatus O. Grant, 1890
5. Vermiculated
Double-spurred Spurfowl
Heuglin’s Spurfowl
Clapperton’s Spurfowl
P. bicalcaratus (Linnaeus, 1766)
P. icterorhynchus Heuglin, 1863
P. clappertoni (Children & Vigors,
1826)
P. hildebrandti Cabanis, 1878
P. natalensis Smith, 1833
Hildebrandt’s Spurfowl
Natal Spurfowl
24
Species group
English name
Scientific name
Hartlaub’s Spurfowl
Harwood’s Spurfowl
Red-billed Spurfowl
Cape Spurfowl
P. hartlaubi Bocage, 1869
P. harwoodi Blundell & Lovat, 1899
P. adspersus Waterhouse, 1838
P. capensis (Gmelin, 1789)
6. Striated
Crested Francolin
Ring-necked Francolin
Ortygornis sephaena (Smith, 1836)
Scleroptila streptophora O. Grant,
1891
7. Red-winged
Shelley’s Francolin
Grey-winged Francolin
Orange River Francolin
Red-winged Francolin
Finsch’s Francolin
Moorland Francolin
Scleroptila shelleyi O. Grant, 1890
S. afra (Latham, 1790)
S. levaillantoides Smith, 1836
S. levaillantii (Valenciennes, 1825)
S. finschi Bocage, 1881
S. psilolaema Gray, 1867
8. Red-tailed
Coqui Francolin
White-throated Francolin
Schlegel’s Francolin
Peliperdix coqui (Smith, 1836)
P. albogularis Hartlaub, 1854
P. schlegelii Heuglin, 1863
Latham’s Francolin
Swamp Francolin
Grey Francolin
Nahan’s Francolin
Afrocolinus lathami Hartlaub, 1854
Ortygornis gularis (Temminck, 1815)
O. pondicerianus (Gmelin, 1789)
Francolinus ‘Ptilopachus’ nahani
(Dubois, 1905)
Unplaced species
25
Table 1.2. Summary of the taxonomic designations from selected revisions pertaining to
francolins, detailed according to putative super-species group, genera, species and subspecies.
Fran - Genus Francolinus Stephens, 1819, Peli - Peliperdix Bonaparte, 1856, Scler Scleroptila Blyth, 1849, Dend - Dendroperdix Roberts, 1922, Orty - Ortygornis Reichenbach,
1852, Pter - Pternistis Wagler 1832, Chae - Chaetopus Swainson, 1837, Afrocolinus gen. nov.
(this thesis).
n/a denotes taxa not investigated (i.e. either falling outside the geographic region covered or
taxa not being of interest), + denotes taxa recognized, - denotes taxa not recognized, ~ denotes
taxa lumped/synonymized with others.
Putative genera - ~8, Putative species - ~41, Putative subspecies - ~164.
Taxon authority citation with a year in parentheses indicates that either the specific or the
subspecific epithet was originally published in another genus by the first author, but moved to
the present genus by the second revising author.
Putative
species
and subspecies
Roberts
(1924)
Hockey
et al.
(2005)
Peters
(1934)
MackHall
Wolters
worth(1963) (1975Praed
82)
& Grant
(1952,62,70)
MandiwanaNeudani
(this study)
Genera
Spotted
Group
n/a
n/a
Fran
n/a
Fran
Fran
Fran
Species
francolinus
subspecies
francolinus (Linnaeus, 1766)
arabistanicus
Zarudny & Härms, 1913
bogdanovi Zarudny, 1906
henrici Bonaparte, 1856
asiae Bonaparte, 1856
melanotus Hume, 1888
n/a
n/a
+
n/a
+
+
+
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
+
+
n/a
n/a
n/a
n/a
n/a
+
+
+
+
+
+
n/a
n/a
n/a
n/a
n/a
pictus
pictus (Jardine & Selby, 1828)
pallidus (Gray, 1831)
watsoni Legge, 1880
n/a
n/a
n/a
n/a
n/a
n/a
+
+
-
n/a
n/a
n/a
+
+
+
+
-
+
n/a
n/a
pintadeanus
pintadeanus (Scopoli, 1786)
phayrei (Blyth, 1843)
n/a
n/a
n/a
n/a
+
+
n/a
n/a
+
+
+
+
+
n/a
26
Putative
species
and subspecies
Roberts
(1924)
Hockey Peters
et al.
(1934)
(2005)
MackHall
Wolters
worth(1963) (1975Praed
82)
& Grant
(1952,62,70)
MandiwanaNeudani
(this study)
Genera
Striated
Group
Dend
Dend
Fran
Fran
Fran
Dend
Orty/Scler
sephaena
sephaena (Smith, 1836)
spilogaster Salvadori, 1888
somaliensis Grant & Praed, 1934
schoanus Heuglin, 1873
jubaensis Zedlitz, 1913
grantii Hartlaub, 1866
rovuma Gray, 1867
zambesiae Praed, 1920
chobiensis Roberts, 1932
thompsoni Roberts, 1924
zuluensis Roberts, 1924
mababiensis Roberts, 1932
+
n/a
n/a
n/a
n/a
n/a
+
+
+
+
+
-
+
n/a
n/a
n/a
n/a
n/a
+
+
~
~
~
~
+
+
+
+
+
+
+
+
~
~
~
+
+
+
+
+
-
+
+
+
+
+
-
+
+
+
+
-
+
+
~
~
~
+
+
~
~
~
~
~
streptophora
streptophora O. Grant, 1891
n/a
n/a
+
+
+
+
+
Orty
Peli
Fran
Fran
Fran Peli
Peli
coqui
coqui (Smith, 1836)
spinetorum Bates, 1928
buckleyi Peters, 1934
maharao Sclater, 1927
ruahdae Someren, 1926
hubbardi O. Grant, 1895
angolensis Rothschild, 1902
lynesi Sclater, 1932
vernayi (Roberts, 1932)
campbelli (Roberts, 1928)
thikae Grant & Praed, 1934
kasaicus White, 1945
hoeschianus Stresemann, 1937
stuhlmanni Reichenow, 1889
+
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
n/a
n/a
n/a
~
~
~
~
~
n/a
~
~
~
+
+
+
+
+
+
+
+
+
+
~
+
+
+
+
+
+
+
+
+
+
+
-
+
+
+
+
+
+
+
+
+
+
-
+
+
-
+
+
~
+
+
+
~
~
+
~
~
+
~
+
schlegelii
schlegelii Heuglin, 1863
n/a
n/a
+
+
+
+
+
albogularis
albogularis Hartlaub, 1854
buckleyi O. Grant, 1892
dewittei Chapin, 1937
meinertzhageni White, 1944
gambagae (Praed, 1920)
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
+
+
+
-
+
+
+
+
-
+
+
+
+
-
+
+
+
~
~
Red-tailed
Group
27
Putative
species
and subspecies
Roberts
(1924)
Hockey Peters
et al.
(1934)
(2005)
MackHall
Wolters
worth(1963) (1975Praed
82)
& Grant
(1952,62,70)
MandiwanaNeudani
(this study)
Genera
Red-winged
Group
Scler
Scler
Fran
psilolaema
psilolaema Gray, 1867
ellenbecki Erlanger, 1905
elgonensis O. Grant, 1891
theresae Meinertzhagen, 1937
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
-
shelleyi
shelleyi O. Grant, 1890
uluensis O. Grant, 1892
whytei Neumann, 1908
macarthuri Someren, 1938
trothae Reichenow, 1901
sequestris Clancey, 1960
canidorsalis (Lawson, 1963)
+
n/a
n/a
n/a
n/a
-
+
n/a
n/a
n/a
~
~
afra
afra (Latham, 1790)
+
Fran
Fran Scler
Scler
+
+
+
+
+
~
~
~
+
+
+
+
+
+
~
+
+
+
+
+
-
+
+
+
+
-
+
+
+
~
-
+
-
+
+
+
~
~
~
~
+
+
+
+
+
+
levaillantoides
levaillantoides Smith, 1836
kalaharica Roberts, 1932
pallidior Neumann, 1908
langi Roberts, 1932
wattii Macdonald, 1953
jugularis Büttikorfer, 1889
cunenensis Roberts, 1932
stresemanni
Hoesch & Niethammer, 1940
gutturalis (Rüppell, 1835)
lorti Sharpe, 1897
archeri Sclater, 1927
ludwigi Neumann, 1920
+
n/a
n/a
n/a
n/a
n/a
n/a
+
~
+
~
~
+
~
+
~
+
+
+
~
+
+
+
+
+
+
+
+
+
+
+
-
+
+
-
+
~
+
~
~
+
~
n/a
n/a
n/a
n/a
+
~
n/a
n/a
n/a
~
+
+
+
+
+
+
+
-
+
+
+
-
+
+
+
-
~
+
~
~
~
levaillantii
levaillantii (Valenciennes, 1825)
kikuyuensis O. Grant, 1897
crawshayi O. Grant, 1896
benguellensis Neumann, 1908
clayi White, 1944
+
n/a
n/a
n/a
n/a
+
n/a
-
+
+
+
+
-
+
+
+
+
+
+
+
+
-
+
-
+
+
+
~
~
finschi
finschi Bocage, 1881
n/a
n/a
+
+
+
+
+
28
Putative
species
and subspecies
Roberts
(1924)
Hockey Peters
et al.
(1934)
(2005)
MackHall
Wolters
worth(1963) (1975Praed
82)
& Grant
(1952,62,70)
MandiwanaNeudani
(this study)
Genera
Vermiculated
Group
Chae/Peli
Pter
Fran
Fran
Fran
Pter
Pter
hartlaubi
hartlaubi Bocage, 1869
crypticus Stresemann, 1939
bradfieldi (Roberts, 1928)
ovambensis (Roberts, 1928)
+
n/a
n/a
n/a
+
-
+
+
+
+
+
+
-
+
~
~
-
+
-
+
~
~
~
adspersus
adspersus Waterhouse, 1838
kalahari de Schauensee, 1931
mesicus Clancey, 1996
+
-
+
~
+
+
+
-
+
-
+
-
+
-
+
~
~
capensis
capensis (Gmelin, 1789)
+
+
+
+
+
+
+
natalensis
natalensis Smith, 1833
neavei Praed, 1920
+
-
+
-
+
+
+
+
+
+
+
-
+
~
hildebrandti
hildebrandti Cabanis, 1878
altumi
Fischer & Reichenow, 1884
helleri Mearns, 1915
fischeri Reichenow, 1887
johnstoni Shelley, 1894
grotei Reichenow, 1919
n/a
n/a
+
+
+
+
+
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
+
+
+
+
-
+
+
-
+
+
-
~
~
+
~
~
bicalcaratus
bicalcaratus (Linnaeus, 1766)
ogilvie-grantii Bannerman, 1922
ayesha Hartert, 1917
adamauae Neumann, 1915
thornei Ogilvie-Grant, 1902
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+
~
~
+
~
icterorhynchus
icterorhynchus Heuglin, 1863
dybowski Oustalet, 1892
ugandensis Neumann, 1907
emini Neumann, 1907
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
~
+
+
+
+
-
+
+
-
+
~
~
~
clappertoni
clappertoni
(Children & Vigors, 1826)
sharpii Ogilvie-Grant, 1892
n/a
n/a
n/a
n/a
+
+
+
+
+
+
+
+
+
+
29
Putative
species
and subspecies
Roberts
(1924)
Hockey Peters
et al.
(1934)
(2005)
MackHall
Wolters
worth(1963) (1975Praed
82)
& Grant
(1952,62,70)
MandiwanaNeudani
(this study)
Genera
Vermiculated
Group
Chae/Peli
heuglini Neumann, 1907
gedgii Ogilvie-Grant, 1891
nigrosquamatus Neumann, 1902
konigseggi Madarasz, 1914
testis Neumann, 1928
cavei Macdonald, 1940
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
harwoodi
harwoodi Blundell & Lovat, 1899 n/a
Fran
Fran
Pter
Pter
+
+
+
~
+
-
+
+
+
+
+
+
+
+
+
-
-
~
~
~
~
~
~
n/a
+
+
+
+
+
n/a
n/a
Fran
Fran
Fran
Pter
Pter
erckelii
erckelii (Rüppell, 1835)
pentoni Praed,1920
n/a
n/a
n/a
n/a
+
+
+
+
+
+
+
-
+
~
nobilis
nobilis Reichenow, 1908
chapini Grant & Praed, 1934
n/a
n/a
n/a
n/a
+
-
+
+
+
+
+
-
+
~
camerunensis
camerunensis Alexander, 1909
n/a
n/a
+
+
+
+
+
swierstrai
swierstrai (Roberts, 1929)
n/a
n/a
+
+
+
+
+
castaneicollis
castaneicollis Salvadori, 1888
bottegi Salvadori, 1898
gofanus Neumann, 1904
ogoensis Praed, 1920
kaffanus Grant & Praed 1934
atrifrons (Conover, 1930)
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
~
~
~
~
ochropectus
ochropectus
(Dorst & Jouanin, 1952)
n/a
n/a
-
-
+
+
+
jacksoni
jacksoni O. Grant, 1891
n/a
pollenorum Meinertzhagen, 1937n/a
gurae Bowen, 1931
n/a
n/a
n/a
n/a
+
+
+
+
-
+
+
-
+
-
+
~
~
Montane
Group
Pter
Fran
30
+
Putative
species
and subspecies
Roberts
(1924)
Hockey Peters
et al.
(1934)
(2005)
MackHall
Wolters
worth(1963) (1975Praed
82)
& Grant
(1952,62,70)
MandiwanaNeudani
(this study)
Genera
Scaly
Group
Fran
Fran
Pter
Pter
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
~
+
~
n/a
n/a
n/a
+
+
+
+
-
+
+
-
-
~
~
~
n/a
n/a
n/a
n/a
+
-
+
-
-
~
~
ahantensis
ahantensis Temminck, 1854
hopkinsoni Bannerman, 1934
n/a
n/a
n/a
n/a
+
-
+
+
+
+
+
-
+
~
griseostriatus
griseostriatus O. Grant, 1890
n/a
n/a
+
+
+
+
+
Pter
Pter
Pter
Pter
Fran
Pter
Pter
+
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+
+
-
+
~
~
~
~
~
+
~
~
n/a
~
+
+
+
+
+
+
+
+
+
+
-
~
+
~
+
n/a
n/a
~
~
~
~
~
+
~
~
~
~
-
+
+
+
+
+
+
+
-
+
+
+
+
+
-
+
+
+
-
~
~
~
~
~
~
~
~
squamatus
squamatus Cassin, 1857
maranensis Mearns, 1910
schuetti Cabanis, 1880
usambarae Conover, 1928
uzungwensis
Bangs & Loveridge, 1931
doni Benson, 1939
zappeyi Mearns, 1911
tetraoninus
Blundell & Lovat, 1899
chyuluensis Someren, 1939
Bare-throated
Group
n/a
n/a
Fran
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
afer
afer (Müller, 1776)
+
harterti Reichenow, 1909
n/a
nyanzae Conover, 1929
n/a
böhmi Reichenow, 1885
n/a
intercedens Reichenow, 1909
n/a
itigi (Bowen, 1930)
n/a
cranchii (Leach & Koenig, 1818) n/a
punctulatus (Gray, 1834)
n/a
benguellensis Bocage, 1893
n/a
leucoparaeus
(Fischer & Reichenow, 1884)
n/a
humboldtii (Peters, 1854
+
swynnertoni Sclater, 1921
n/a
castaneiventer
Gunning & Roberts, 1911
+
melanogaster Neumann, 1898
n/a
loangwae Grant & Praed, 1934 n/a
lehmanni Roberts, 1931
notatus Roberts, 1924
+
krebsi Neumann, 1920
+
cunenensis Roberts, 1932
n/a
cooperi Roberts, 1947
-
31
Putative
species
and subspecies
Roberts
(1924)
Hockey Peters
et al.
(1934)
(2005)
MackHall
Wolters
worth(1963) (1975Praed
82)
& Grant
(1952,62,70)
MandiwanaNeudani
(this study)
Genera
Bare-throated
Group
Pter
Pter
swainsonii
swainsonii (Smith, 1836)
lundazi White, 1947
gilli Roberts, 1932
damarensis Roberts, 1932
chobiensis Roberts, 1932
+
n/a
n/a
n/a
n/a
+
+
~
~
~
rufopictus
rufopictus Reichenow, 1887
n/a
leucoscepus
leucoscepus (Gray, 1867)
infuscatus Cabanis, 1868
holtemülleri Erlanger, 1904
keniensis Mearns, 1911
kilimensis Mearns, 1911
tokora Stoneham, 1930
muh.-ben-abdul. Erlanger, 1904
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Pter
Pter
Fran
Pter
Pter
+
~
~
~
+
+
+
+
+
~
~
~
-
+
-
+
~
~
~
~
n/a
+
+
+
+
+
n/a
n/a
n/a
n/a
n/a
n/a
n/a
+
+
+
+
~
+
+
+
+
+
+
+
~
-
+
-
+
+
~
~
~
~
~
Unplaced taxa
pondicerianus Hall (1963:167) speculates affinities with sephaena and/or coqui
pondicerianus (Gmelin, 1789)
mecranensis
Zarudny & Härms, 1913
interpositus Hartert, 1917
ceylonensis Whistler, 1941
n/a
n/a
+
n/a
+
+
+
n/a
n/a
n/a
n/a
n/a
n/a
+
+
-
n/a
n/a
n/a
+
+
+
-
n/a
n/a
n/a
gularis Hall (1963:167-168) “comparable in size and general proportions to the largest member of F.
francolinus” isolated from other francolins” …. “divergence over a long period”
gularis (Temminck, 1815)
n/a
n/a
+
n/a
+
+
+
nahani Hall (1963:166-167) speculates affinities with squamatus and other ‘Scaly’ francolins
nahani (Dubois, 1905)
n/a
n/a
+
+
+
+
-
lathami Hall (1963:165-166) speculates affinities with coqui and other ‘Red-tailed’ francolins
lathami Hartlaub, 1854
schubotzi Reichenow, 1912
n/a
n/a
n/a
n/a
+
+
+
+
+
+
+
-
A new genus Afrocolinus gen. nov. is recognized for lathami Hartlaub, 1854 in this study.
32
+
+
Table 1.3. Summary of the number of putative genera, species and subspecies outlined in
Table 1.2.
Taxonomic category
Author/s
Roberts (1924)
Peters (1934)
Mackworth-Praed & Grant (1952,
62, 70)
Hall (1963)
Wolters (1975-82
Crowe et al. (1986)
del Hoyo et al. (1994)
Dickinson (2003)
Hockey et al. (2005)
Mandiwana-Neudani (this thesis)
Geographic
area covered
Putative
genera
Putative
species
Putative
subspecies
South Africa
Asia & Africa
6
2
11
39
20
119
Africa
Asia & Africa
Asia & Africa
Asia & Africa
Asia & Africa
Asia & Africa
Southern Africa
Asia & Africa
2
1
5
1
1
1
4
8
34
41
41
41
41
41
11
40
113
101
56
50
50
85
17
54
33
Table 1.4. Summary of characters used by authors to justify their positions on the
classification of francolins.
Authors
Characters used
Roberts (1924)/Hockey et
al. (2005)
Plumage, calls, size, bare skin on throat, distribution range†, breeding
information
Chapin (1932)
Habitat type, distribution range†, morphology
Peters (1934)
Breeding information, distribution range†
Mackworth-Praed & Grant
(1952, 1962, 1970)
Morphology, geographic distribution†, habits, nest and eggs, breeding
information, calls
Hall (1963)
Plumage, size, sexual dimorphism, extent of bare skin, colour of bill
and legs, spurs, field habits, vocalizations, chicks, eggs
Wolters (1975-82)
Distribution range†
Crowe and Crowe (1985)
Skeletal and integument characters, vocalizations, ethological and
ecological information
Crowe et al. (1986)
Morphology, field characters, voice, habitat and food preferences,
breeding habits
Milstein and Wolff (1987)
Calls of adults and chicks, incubation periods, spur development, natal
down, hybridization, general behaviour (whether a particular species is
a squatter or runner)
Sibley and Monroe (1990)
Habitat preferences, hybridization information, distribution range†
Crowe et al. (1992)
Mitochondrial DNA Restriction Fragment Length Polymorphisms,
morpho-behavioural characters
del Hoyo et al. (1994)
Feeding habit, habitat preferences, breeding, distribution range†,
morphology
Bloomer and Crowe (1998)
Mitochondrial Cytochrome b sequence characters, morphobehavioural characters
Dickinson (2003)
Primarily results from published molecular DNA phylogenies, also
follow Peters (1934)
† indicates that the distribution ranges of taxa were not presented as point localities
instead as roughly defined ranges.
34
Figure 1.1. The re-drawn hypothetical cladogram of francolins according to Hall (1963). The
partridge-quail francolin dichotomy follows Milstein and Wolff (1987). The acronyms used on
the tree abbreviate the following: SPG stands for Spotted Group, BTG - Bare-throated Group,
MTG - Montane Group, SCG - Scaly Group, VMG - Vermiculated Group, RWG - Redwinged Group, STG - Striated Group and RTG abbreviates Red-tailed Group. AS stands for
Asia and AF - Africa.
35
Figure 1.2. The francolin phylogeny based on combined analysis of 200 Cytochrome-b
and 25 morphobehavioral characters (adopted from Bloomer and Crowe 1998).
Bootstrap support values are indicated at nodes.
36
CHAPTER 2
Phylogenetics of evolutionarily enigmatic terrestrial
gamebirds (Aves: Galliformes) with special regard to
‘Francolinus’ spp.
Part of the information presented in this chapter is derived from Crowe et al. (2006) and
Cohen et al. (2012) (both of which I am a co-author), where I contributed in a primary
capacity with regard to content and molecular DNA sequence data on francolins
(Francolinus, Ortygornis, Afrocolinus, Peliperdix and Scleroptila spp.) and spurfowls
(Pternistis). I was also involved at all levels with respect to the preparation of these
manuscripts.
Abstract
The ultimate goal of the cladistic analysis of any putative taxonomic group is to
demonstrate (or refute) its monophyly decisively. If the former proves to be the case, the
next task is to identify well-resolved and supported monophyletic assemblages within it.
Until Crowe et al. (2006), these goals had not been achieved for the terrestrial
gamebirds (Aves: Galliformes), a geographically widespread assemblage that harbours
some of the most economically important bird species and a spectrum of biologically
and geographically diverse taxa that have been used as models for studying broad-scale
physiological, ecological, evolutionary and biogeographical phenomena. However,
some authors, e.g., Eo et al. (2009), Shen et al. (2010), still maintain that this goal
37
remains unattained. In particular, the lack of a well-resolved hypothesis concerning
phylogenetic relationships within the crown Galliformes, and especially phasianine
galliforms (pheasants, partridges, ‘francolins’, quails, grouse, turkeys and peafowls),
hinders interpretation of their morphological and ecological evolution and their utility as
biogeographical indicators needed for conservation and management initiatives. In this
chapter, four mitochondrial and three nuclear markers were sequenced for a selection of
galliform species (and not genera, are terminals as in previous studies). Further evidence
in terms of the large number of francolin and spurfowl species analyzed, mitochondrial
and nuclear DNA characters (and new evidence particularly from vocalizations and
behaviour of particularly F. nahani) was explored in the context of the phylogenetic
affinities suggested by Crowe et al. (2006) with the goal of producing a robust
phylogenetic hypothesis for Africa's diverse Galliform lineages, and in particular
francolins sensu lato.
Results from this study rejects the traditional classification of galliform families,
and instead supports the recognition of only five families, which are the: Megapodiidae,
Cracidae, Numididae, Odontophoridae and Phasianidae. Members of the two
traditionally recognized families, the Meleagrididae (turkeys) and Tetraonidae (grouse
and allies), were nested within the Phasianidae. Regarding the focal taxa, Nahan’s
Francolin F. nahani is not a francolin but a partridge sister to the Stone Partridge P.
petrosus, and they diverged c. 9.6 mya. Their relationship is strongly supported and
contradicting all other published treatments of the galliformes. The two form a basal
clade relative to ‘true’ francolins and spurfowls suggesting that they represent a relictual
group sister to the New World quails (Odontophoridae), and are only distantly related to
the other Old World galliforms. This study also confirms earlier studies in
38
demonstrating that francolins and spurfowls are not each other’s closest relatives as
suggested by Hall (1963). The genus Francolinus comprises at least two distantly
related lineages (excluding F. nahani), the partridge-like spurfowls, which are related to
quails (e.g. Coturnix, Excalfactoria, Margaroperdix, Perdicula and Ammoperdix spp.)
and certain Old World partridges (e.g. Tetraogallus and Alectoris spp.), and the quaillike francolins (Dendroperdix, Peliperdix and Scleroptila spp.), which are related to the
‘true’ Asian francolins (Francolinus spp.), chickens (Gallus spp.), and other Old World
partridges (Bambusicola spp.). The estimated divergence date between the
Gallus/Bambusicola/francolins and the Coturnix/Alectoris/spurfowls clade is recovered
at c. 33.6 mya.
39
Introduction
Prior to 2006, there was no consensus on the phylogenetic relationships of, or within,
the Galliformes, except that the Galliformes were monophyletic (a view initially
challenged by Prager and Wilson 1976, Jolles et al. 1976, 1979). The sister-group
relationship between Galliformes and Anseriformes (ducks, geese and screamers) in the
avian tree of life is very strongly supported (Hackett et al. 2008), although relationships
among the lineages within Galliformes still remain controversial (Wang et al. 2013).
The number of supra-generic monophyletic assemblages within Galliformes has
varied greatly. Adopting an extreme ‘splitters’ viewpoint (Verheyen 1956, Johnsgard
1973, 1986, 1988, 1999), eight groupings with associated core biogeographical
affinities have been identified: Megapodiidae (Australasia - megapodes, scrubfowl,
brush-turkeys), Cracidae (Neotropics - curassows, guans and chachalacas), Numididae
(Afrotropics
–
guineafowls),
Phasiani(n/d)ae
(Afro/Asiotropical
-
pheasants,
junglefowls - chickens, peafowls, and peacock- and argus-pheasants), Perdicinae
(Palaearctic and Afro/Asiotropical - partridges, francolins and Old World quails),
Meleagridi(n/d)ae (Nearctic – turkeys), Tetraoni(n/d)ae (Holarctic – grouse) and
Odontophori(n/d)ae (Neotropical and Nearctic - New World quails).
With respect to these groupings (reviewed by Johnsgard 1973, 1986, 1988, 1999
and Sibley and Ahlquist 1985, 1990), there is general agreement that the megapodes and
cracids are cladistically basal (and probably sister-taxa – Wetmore 1960, Sibley and
Ahlquist 1990), followed by the guineafowls (Cracraft 1981, Crowe 1988) or New
World quails (Sibley and Ahlquist 1990, Kornegay et al. 1993) as sister to the
remaining phasianine assemblages. Within phasianine galliforms, it is traditionally
accepted that the turkeys and grouse form monophyletic clades, with Sibley and
40
Ahlquist (1990) speculating that they are each other's closest relatives. This leaves the
Phasiani(n/d)ae and Perdicinae for which several authors have suggested that their
constituent taxa are polyphyletic (e.g. Bloomer and Crowe 1998).
The Afrotropics harbour 49 species of galliform gamebirds, occurring in
virtually all habitats across the continent largely south of the Sahara (Crowe et al. 1986,
del Hoyo et al. 1994). Crowe et al. (2006) demonstrated decisively (see also Milstein
and Wolf 1987, Crowe et al. 1992, Bloomer and Crowe 1998), that Africa’s largest
gamebird genus (currently 36 spp.), Francolinus Stephens, 1819 (sensu Hall 1963)
comprises at least two distantly related African radiations. The partridge-like spurfowls
(Pternistis, 24 spp.) are related to quail (e.g. Coturnix, Excalfactoria, Margaroperdix,
Perdicula and Ammoperdix spp.) and certain Old World partridges (e.g. Tetraogallus
and Alectoris spp.). The quail-like francolins (Dendroperdix, Peliperdix and Scleroptila,
12 spp.) are related to ‘true’ Asian francolins (Francolinus, five spp.), junglefowls
(Gallus) and other Old World partridges (Bambusicola). The remaining African
galliforms comprise the Old World quails (Coturnix and Excalfactoria, three spp.), the
endemic guineafowls (Numididae, six spp.) and four species with putative IndoMalaysian affinities: the Congo Peafowl Afropavo congensis, the Stone Partridge
Ptilopachus petrosus, and the Udzungwa and Rubeho Forest Partridges Xenoperdix
udzungwensis and X. obscurata (Dinesen et al. 1994, Johnsgard 1988, Madge and
McGowan 2002, Bowie and Fjeldså 2005, Crowe et al. 2006). All of these African taxa
were thought to have their nearest phylogenetic relatives elsewhere in the Old World
(Sibley and Ahlquist 1990). Thus, it was surprising when Crowe et al. (2006) suggested
that the Stone Partridge Ptilopachus petrosus and Nahan’s Francolin Francolinus
41
nahani were sister-species, and that this clade (the Ptilopachinae, Bowie et al. 2013)
was sister to the New World Quails (Odontophoridae).
Based on an analysis of 158 ingroup taxa representing 65 genera and all putative
suprageneric galliform taxa rooted on representatives of the Anseriformes, Crowe et al.
(2006) (including myself as a junior author) offered a novel, generally well-resolved
and well-supported, phylogenetic hypothesis for the Galliformes (Fig. 2.1). Characters
analyzed included 102 morpho-behavioural attributes (Dyke et al. 2003) and 4452
nucleic acid base pairs (bp) from four mitochondrial markers (Cytochrome-b, ND2, 12S
and control region) and a single nuclear marker, ovomucoid intron G. At the generic
level, parsimony-based cladistic analysis of the concatenated character data set yielded a
single, completely resolved, supra-generic cladogram, often with high values of nodal
jackknife support (Fig. 2.1), which suggested the need for a revised classification for the
phasianine galliforms.
A brief review of the major galliform results of Crowe et al. (2006)
The megapodes are monophyletic (clade 1 in Fig. 3.1) and cladistically basal within a
monophyletic Galliformes. The next clade comprises the monophyletic cracids (clade
2), followed by the relatively basal guineafowls (clade 3), and not the New World quails
(clade 4), both of which are monophyletic and in turn sister to the balance of the
phasianids. The New World quails are basal within the remaining phasianids, but
include an African sister clade formed by the monotypic stone partridge Ptilopachus
petrosus and Hall’s (1963) previously phylogenetically enigmatic Nahan’s ‘Francolin’
(‘Francolinus/Acentrortyx’ nahani) (clade 4). Within the remainder of phasianids, the
grouse and ‘true’ (i.e. wattled) pheasants were the only demonstrably monophyletic
42
supra-generic traditionally recognized clades; Phasianids including the remainder of the
Phasianinae and Perdicinae (especially the latter) are polyphyletic; basal among these
phasianids is a novel clade comprised of Afro/Asian ‘partridges’ (Xenoperdix, Rollulus
and Arborophila spp. - clade 5), with the remaining Phasianinae and Perdicinae taxa
sundered phylogenetically into five clades which are: the Old World quails (Coturnix
and Excalfactoria spp.) including the monotypic Madagascar ‘partridge’ Margaroperdix
madagarensis), some Old World partridges (e.g. Alectoris, Tetraogallus, Ammoperdix
and Perdicula spp.), and some ‘francolins’ (Pternistis spp. which are commonly known
as spurfowls) (clade 6); the junglefowls (Gallus spp.), Bambusicola partridges and
‘true’ (= quail-like) francolins (Francolinus, Dendroperdix, Peliperdix and Scleroptila
spp.) (clade 7); the Afro-Asian peafowls sensu lato including the Argus Pheasant
(Argusianus and Rheinardia spp.) the Congo Peacock (Afropavo congensis), and the
peacock-pheasants (Polyplectron spp.) (clade 8); the turkeys (Meleagris spp.) and ‘the’
partridges (Perdix spp.) as sister-species (with relatively weak jackknife support) and
sister to the grouse (e.g. Tetrao, Bonasa and Tympanuchus spp.) (clade 9); and the ’true’
pheasants (Phasianus, Catreus, Syrmaticus, Chrysolophus, Lophura and Crossoptilon
spp.) that exclude the peacock-pheasants and junglefowls, but that include (with no
jackknife nodal support) the somewhat partridge-like taxa Ithaginis, Lophophorus,
Pucrasia and Tragopan spp. (clade 10).
A brief review of Galliform studies published since Crowe et al. (2006)
Since the publication of Crowe et al. (2006) several further manuscripts have been
published that have bearing on its conclusions. They, and their key findings, are as
follows: Cox et al. (2007) analyzed eight nuclear and three mitochondrial markers for
43
16 ingroup galliform taxa (rooted on megapodes), and produced a maximum likelihood
cladogram that was consistent with the finding that the guineafowls and not the New
World Quails are basal phasianoids, i.e. confirming Crowe et al. (2006).
Analysis of 20 ingroup galliforms (rooted on Anseriformes) for 25 retroposed
elements (large insertions of nuclear DNA commonly called ‘jumping genes’) by Kriegs
et al. (2007), recovered a cladogram that was consistent with that in Crowe (2006)
except in the placement of ‘the’ partridge, Perdix perdix. Rather than being sister to the
turkey (Meleagris gallopavo), it was placed as sister to the Chrysolophus pheasants,
leaving the turkey as sister to grouse. However, the methods and criteria used to
generate this cladogram were not stated. Subsequent correspondence with the authors
revealed that parsimony as implemented in PAUP*10b (Swofford 2002) and that the
IRREV.UP option had been used to generate their preferred tree. This strategy assumes
an ‘all-plesiomorphic’ outgroup and prevented the retroposed elements from being lost
(reversing to absent) once they appear in a clade. Although this may be a reasonable
assumption for retroposed elements, an unconstrained analysis with TNT (Goloboff et
al. 2008) using an all-zero outgroup did not resolve the placement of either the partridge
or turkey beyond grouping them in a polytomy with grouse and pheasants.
Kimball and Braun’s (2008) model-based analysis of four nuclear introns and
two mitochondrial coding regions for 41 ingroup taxa rooted on megapodes also
recovered a cladogram largely congruent with that in Crowe et al. (2006) except that the
turkey was placed as sister to grouse, and the partridges were placed as sister to the true
pheasants, while the pavonines emerged paraphyletic. Eo et al. (2009) conducted a
supertree analysis (Gatesy et al. 2002, Bininda-Edmonds et al. 2004) of available
GenBank data with the resulting topology also being highly congruent with that of
44
Crowe et al. (2006). Model-based phylogenetic analysis (Shen et al. 2010) of complete
mitochondrial genomes of 34 galliform taxa placed turkeys with grouse, the partridges
with true pheasants and the pavonines as polyphyletic. This dataset also placed the New
World quails as basal to guineafowls. Finally, a paper recently published by Wang et al.
(2013) represents the most comphrehensively sampled (in terms of number of
characters) molecular phylogeny to date. Their results refuted the traditional
consideration of seven families and strongly supported the recognition of five major
families within galliformes in the evolutionary sequence: Megapodiidae, Cracidae,
Numididae, Odontophoridae and Phasianidae. This is similar to the findings in Crowe et
al. (2006). The study also strongly supported the hypothesis that the deepest divergence
within extant galliforms is between the Megapodes and all the other galliforms species,
with the next divergence corresponding to that between Cracids and Phasianoidea.
Similarly, as with Crowe et al. (2006), the turkeys (traditionally classified in
Meleagrididae) and the grouse and ptarmigan (Tetraonidae) were nested within
Phasianidae, hence this finding rejects the hypothesis that the turkeys and grouse form
independent families.
Ksepka (2009) and a range of other fossil-related publications
Ksepka’s (2009) parsimony analysis of an expanded and revised matrix of 120
organismal characters combined with sequences from four mitochondrial markers
(control region, 12S rDNA, CYTB, and ND2) and a nuclear ovomucoid intron G for 56
ingroup species rooted on Anseriforms and a Tinamus sp. (Tinamiformes), produced a
cladogram largely congruent with the one presented in Crowe et al. (2006). However,
the major point of criticism presented by Ksepka (2009) is the phylogenetic placement
45
of a key Eocene fossil, Gallinuloides wyomingensis (+/-54 mya old) from North
America. He maintains that, contra Crowe et al. (2006), it is better placed at the stem
(and not in the crown) of the galliform cladogram. Furthermore, there are several other
fossil-based studies (Mlikovsky 1989, Stidham 2008, Elanowski and Stidham 2011,
Mourer-Chauvire et al. 2011) that have challenged the divergence times ascribed to
clades in Crowe et al. (2006).
This chapter
In this chapter, the focal taxa are francolins sensu lato. They are small to
medium-sized, sedentary, Old World, partridge/quail-like gamebirds which occur in
varied habitats, from dry, lowland grassland to montane forests (Hall 1963, Johnsgard
1988, del Hoyo et al. 1994, Madge and McGowan 2002). Taxonomically, francolins
were invariably placed in the family Phasianidae. In some more finely partitioned
classifications, within the Phasianidae, they are placed in the sub-family Phasianinae
(including Phasianus and related species), and together with other Old World partridgeand quail-like gamebirds (e.g. Perdix and Coturnix spp.), in the tribe Perdicini (Chapin
1932, Peters 1934, Wolters 1975-82, Crowe et al. 1986, Johnsgard 1988, Sibley and
Monroe 1990, del Hoyo et al. 1994, Madge and McGowan 2002). Controversy has
centered on the status of monophyly of the genus Francolinus sensu Hall (1963) and
debate over the geographic origin of the genus ‘Francolinus’, for which two contrasting
hypotheses have been postulated (Hall 1963, Crowe and Crowe 1985). Based on the
notion that the genus Francolinus shares its closest affinities with other Palaearctic and
Asian genera, Hall (1963) strongly argued for the genus to be of Asian origin with its
age being traced back to the Oligocene +/- 25-35. Although Crowe and Crowe (1985)
46
concurred with Hall that the ancestor was quail-like, they hypothesized an African
origin for the genus, with Asia being colonized by a nomadic or migratory ancestor that
diversified and became sedentary.
The enigmatic Stone Partridge occurs on rocky outcrops in the arid habitats of
the northern savanna belt including the Sahel south of the Sahara, from Gambia to
Ethiopia, and south to Cameroon and northern Kenya (Crowe et al. 1986). It was
described initially as a Tetrao by Gmelin (1789), but was subsequently placed into a
monotypic genus Ptilopachus by Swainson (1837). Nahan’s Francolin in contrast, is a
highly-localized species associated with core areas of primary forests of the eastern
equatorial lowlands of the Democratic Republic of the Congo and Uganda (Crowe et al.
1986, Sande et al. 2009). This taxon was first placed by Dubois (1905) in the genus
Francolinus, but subsequently moved by Chapin (1926) into a monotypic genus,
Acentrortyx. Hall (1963) placed it back into Francolinus because she doubted the value
of characters used to split it from Francolinus. Furthermore, she linked it tentatively to
members of her putatively monophyletic ‘Scaly Group’ of spurfowls (Ahanta Francolin
Pternistis ahantensis, Scaly Francolin P. squamatus and Grey-striped Francolin P.
griseostriatus; all previously placed into Francolinus by Hall) on the basis of bare-part
colouration and plumage characteristics. In a morphometric analysis based on
osteological features, Crowe and Crowe (1985) also placed F. nahani near members of
Hall’s Scaly Group, closest to P. ahantensis. In a further reworking of the francolins,
Crowe et al. (1992) placed F. nahani in a resurrected monotypic subgenus, Acentrortyx,
within the African spurfowl genus Pternistis, although speculating that it might
represent a phylogenetically relictual taxon, unrelated to other African galliforms.
47
In this chapter, further evidence in terms of the large number of francolin and
spurfowl species analyzed, mitochondrial and nuclear DNA (and new evidence from
vocalizations and behaviour of particularly F. nahani) is explored in the context of the
phylogenetic affinities suggested by Crowe et al. (2006) with the goal of producing a
robust phylogenetic hypothesis for Africa's diverse Galliform lineages, and in particular
francolins sensu lato.
Materials and methods
Collection of data
Taxon sampling
The taxon sampling was based on that of Crowe et al. (2006), with a number of
important changes. In order to increase the confidence that no taxa had been
overlooked, all additional African ‘francolin’ species (e.g. several additional Pternistis
spp.) as well as additional species of Asian and New World galliforms were sequenced
for this study and further sequences were obtained from GenBank (Table 2.1). A sample
of Ptilopachus petrosus was obtained from Ghana, and three samples of F. nahani were
obtained from Budongo Forest, Uganda.
Molecular approach
Four mitochondrial (mtDNA) markers and three nuclear (nucDNA) markers, which
occur on distinct chromosomes and thus provide independent estimates of phylogeny,
were used in this study. The mitochondrial markers (Cytochrome-b – CYTB, NADH
Dehydrogenase Subunit 2 - ND2, 12S Ribosomal DNA - 12S, and Control Region CR), and nuclear markers (Ovomucoid intron G – OVOG, Transforming Growth Factor
48
Beta 2 intron 5 – TGFB, and GAPDH intron 11 - GAPDH) were investigated since
these markers have helped to resolve the phylogenetic status of other galliform genera
and species (Armstrong et al. 2001, Dimcheff et al. 2000, 2002, Crowe et al. 2006,
Hackett et al. 2008).
Laboratory techniques
Total genomic DNA was extracted from blood, heart and liver tissue using the DNeasy
animal tissue protocol provided with the DNeasy tissue kit (Qiagen). The initial CYTB
primers amplified 1337 base pairs (Table 3.2). Due to the length of this region, an
internal primer (Table 2.2) was also used in sequencing this region. The initial CYTB
primer pair did not amplify Grey-striped Spurfowl Pternistis griseostriatus and Yellownecked Spurfowl P. leucoscepus, thus further galliform specific primers were also used
(Table 2.2).
Double stranded DNA templates were amplified by polymerase chain reaction
(PCR) using 0.75 units of BIOTAQTM DNA polymerase (Bioline) in 30 µl reactions.
Reactions also contained 1 x NH4 buffer, 2.5 mM MgCl2, each dNTP at 0.1 mM, each
primer at 0.3 µM, and 3 µl of extracted DNA was used as template. The thermal profile
used comprised an initial denaturation step at 94°C for two minutes, followed by 30
cycles of 94°C for one minute, 52°C for one minute and 72°C for two minutes, with a
final extension step of 72°C for seven minutes.
PCR-amplified products were cleaned from solution or gel using the GFX TM
PCR DNA and gel band purification kit (Amersham Biosciences) prior to cyclesequencing with the ABI PRISM Big DyeTM Terminator v3.1 cycle-sequencing Ready
Reaction Kit (Applied Biosystems). Sequencing products were resolved on an ABI
49
PRISM 3100 Genetic Analyser. Sequences were assembled and checked for incorrect
base calling and the presence of stop codons using SeqMan II (LaserGene systems
software, DNAstar, Inc.). Consensus sequences were aligned using Clustal and adjusted
manually in MegAlign (LaserGene systems software, DNAstar, Inc.).
Analyses of data
Phylogenetic analyses
Three methods of phylogenetic analysis with different optimality criteria were
employed to generate phylogenetic hypotheses: Bayesian inference (BI), maximum
likelihood (ML) and parsimony. In all analyses, indels and ambiguous character states
were considered as 'missing' and all characters were treated as non-additive.
Parsimony-based phylogenetic analyses were conducted using TNT (Tree analysis using
New Technology - Goloboff et al. 2008). In TNT, the search strategy employed was the
‘traditional’ search option. When multiple, equally parsimonious cladograms persisted,
a strict consensus cladogram was constructed. The extent to which each non-terminal
node is supported by different character partitions was determined by using the
‘jackknife’ resampling strategy with: 1000 replicates, TBR branch-swapping, five
random additions of taxa per replicate with the deletion of 36% of the characters per
jackknife replicate (Farris et al. 1996, Källersjö et al. 1998).
Since gene regions can evolve under different models of evolution, it has been
argued that a partioned, mixed-model approach should be used when concatenating
these different datasets in a model-based phylogenetic analysis (Ronquist and
Huelsenbeck 2003, Nylander et al. 2004). Mixed-model Bayesian analyses were
undertaken in MrBayes v3.1.2 (Huelsenbeck and Ronquist 2001, Ronquist and
50
Huelsenbeck 2003). Substitution models for each locus were determined in PAUP*4b10
(Swofford 2002) with Modeltest 3.06 (Posada and Crandall 1998), using the Akaike
Information Criterion (Akaike 1973, Posada and Buckley 2004). Mixed-model analyses
allowed different parameters (base frequencies, rate matrix or transition/transversion
ratio, shape parameter, proportion of invariable sites) to vary between the partitions
(gene regions and codon positions) (Nylander et al. 2004). Four Metropolis-coupled
MCMC chains (one cold and three heated) were run for 10 million generations with
trees sampled every 100 generations. A Dirichlet distribution was assumed for
estimation of the base frequency parameters and an uninformative (flat) prior was used
for the topology. The ‘burn-in’ period (discarded cycles before the chains had reached
stationarity) varied per analysis but was typically 500 000 generations (5000 trees);
posterior probabilities (PP) were estimated from the remaining generations. Each
Bayesian analysis was run twice (random starting point for each run). The loglikelihood values and posterior probabilities were checked using Tracer v1.4.1
(Rambaut and Drummond 2007) to confirm that the chains had reached stationarity. The
potential scale reduction factor was confirmed to approach 1.0 (for all parameters) and
the average deviation of split frequencies converged towards zero.
Mixed-model maximum likelihood analyses were performed using the
Randomised Axelerated Maximum Likelihood algorithm for High Performance
Computing (RAxML) v7.0.4 (Stamatakis 2006, Stamatakis et al. 2008) as implemented
on the CIPRES portal. Mixed-model RAxML analyses make use of a GTR++ model
partitioned by gene or codon postion. The following analyses were run: mixed-model
mtDNA (one model for each codon position, and also as a single data partition); a
mixed-model analysis of the nuclear DNA genes, partitioned by each of the four gene
51
regions, and a mixed-model analysis of the combined mtDNA and nuclear DNA
datasets. Support at nodes was assessed with 100 non-parametric bootstrap (BS)
pseudoreplicates. The use of different methodological approaches (optimality criteria)
facilitated the identification of method-based incongruence.
Divergence date estimation
The previous and most comprehensive dating analysis of Galliformes was
conducted by Crowe et al. (2006), which made use of the Eocene fossils Gallinuloides
wyomingensis (Green River Formation) and Amitabha urbsinterdictensis (Bridger
Formation) as calibration points. Further preparation and re-examination of Amitabha
resulted in this fossil being removed from Galliformes and it is now placed in the
Rallidae (Ksepka 2009). Further, Ksepka (2009) argues that Gallinuloides is best placed
at the stem and not within the crown of the galliform phylogeny, as previously
suggested by Crowe et al. (2006). If Ksepka (2009) is correct, this would suggest that
previous estimates of divergence dates among galliform lineages have been
overestimated. As a consequence, a new dating analysis was conducted using BEAST v.
1.6.2 (Drummond and Rambaut 2007), omitting both Gallinuloides and Amitabha and
instead calibrated via the use of three additional fossil calibration points: (1) a fossil of a
Crested Francolin Dendroperdix sephaena at 4.5-5.0 mya as a minimum date for the age
of the true Francolins (Crowe 1992), (2) a basal date for the Tetraoninae (Grouse and
allies) of 27-29 mya (Crowe and Short 1992), and (3) a basal date for Polyplectron
(Peacock-Pheasants) of 34-36 mya (Olson 1974, modified by T.M. Crowe unpubl.
data). An uncorrelated lognormal clock was used with the same data partitions and
nucleotide substitution models as described for the Bayesian analyses above. The
52
analysis was run for 80 million generations with tree sampling taking place every 2000
generations. Convergence was determined as described above for the Bayesian
phylogenetic analyses.
Field observations of behaviour and vocalizations of P. petrosus and F. nahani
Behavioural observations and vocalizations were recorded in the field: F. nahani was
observed in the Budongo (1.714°N, 31.543°E) and Mabira (0.399°N, 33.049°E) forests,
Uganda, in 1999, 2002, 2008 and 2009 by Callan Cohen; P. petrosus was observed near
Mora (11.083°N, 14.114°E) and Benoue National Park (8.116°N, 13.679°E) in Northern
Cameroon in 2002, 2004 and 2010, and near Bandiagara (14.359°N, 3.584°E), central
Mali, in 2006, also by Callan Cohen. Sound recordings were made using a stronglydirectional Sennheiser ME-67 microphone with a K6 power module. The recordings
were made onto various media including a Fostex FR-LE-2 solid-state recorder, a Sony
RH1 minidisc recorder in uncompressed format, and an Edirol R-09HR. These were
supplemented by further vocalizations of F. nahani from Brian Finch (unpubl. data) and
from Chappuis (2000).
Vocal analyses
Calls of P. petrosus and F. nahani were compared aurally to all available African
galliform species on Gibbon (1995) and Chappuis (2000), supplemented by additional
calls from the British Library Sound Archive and Macaulay Library of Natural Sounds.
In addition, sonograms were made from typical advertisement calls (heard most often at
dawn and dusk) for P. petrosus and F. nahani and compared with those of putative
sister taxa (spurfowls - Pternistis spp. and francolins - Scleroptila spp.) and other
53
African galliforms. Sonograms were generated in Raven Lite (Version 1.0, Cornell
Laboratory of Ornithology).
Results
Phylogenetic analyses
The parsimony tree adopted from Crowe et al. (2006) is presented in this chapter (Fig.
2.1) since it yielded a single, well-resolved, supra-generic cladogram, often with high
values of nodal jackknife (JK) support (Fig. 2.1). A combination of all seven sequenced
markers resulted in 5554 base pairs, involving 84 taxa, from which the Bayesian
inference (BI) and maximum likelihood (ML) analyses were generated (Fig. 2.2 and
2.3).
In comparing the BI (Fig. 2.2) and ML (Fig. 2.3) topologies generated with that
of Crowe et al.'s (2006) parsimony tree (Fig. 2.1: based on combined mtDNA, nucDNA
and organismal characters) all analyses recovered five major lineages of galliforms and
not seven families as traditionally circumscribed. The sequence of divergence of these
families is such that the Megapodiidae diverged first, followed in sequence by the
Cracidae, Numididae, Odontophoridae and Phasianidae. Members of the two
traditionally recognized families, the Meleagrididae (turkeys) and Tetraonidae (grouse
and allies), are nested within the Phasianidae with 1.0 posterior probability (PP) and
86% ML bootstrap (BS) support, contrary to the findings in parsimony analysis where
the turkeys (Meleagris spp.) and ‘the’ partridges (Perdix spp.) are sister (with relatively
weak 71% JK), and in turn sister to the grouse (e.g. Tetrao, Bonasa and Tympanuchus
spp.) (clade 9) with a lack of JK support.
54
The nodal support among the five prominent families (Megapodiidae, Cracidae,
Numididae, Odontophoridae and Phasianidae) is 1.0 PP, whereas ML analyses
recovered 100% BS support for the node between Cracids and Guineafowls,
Guineafowls and Odontophorids and 91% between Odontophorids and Phasianids with
no support between the basal Megapodes and Cracids. The parsimony topology
recovered 100% JK support between Megapodes and Cracids, Guineafowls and
Odontophorids, 98% between Cracids and Guineafowls and 91% between
Odontophorids and Phasianids. The phylogenetically enigmatic phasianines remain
Ithaginis, Tragopan, Pucrasia, Meleagris, and Perdix spp. (Table 2.3).
With regard to the francolins and spurfowls, the parsimony, BI and ML
topologies reject monophyly of the genus Francolinus decisively (Fig. 2.1, 2.2, 2.3) in
support of francolins sensu lato comprising two distantly related species assemblages.
Spurfowls (Pternistis spp.) form a strongly supported monophyletic lineage with 1.0 PP
and 100% BS but interestingly a lack of JK support in parsimony. Lack of support in
parsimony could be attributed to among other things, fewer spurfowl species having
been included in the analysis, specific differences underlying the analytical principles
for the three phylogenetic inference methods, and analyses based on genetic markers
representing loci evolving at different rates. On the other hand, francolins (Francolinus,
Dendroperdix, Peliperdix, Scleroptila spp.) are monophyletic with a PP of 0.99 and BS
of 78% respectively, but also with a lack of JK support in parsimony. This could be for
the same reasons as highlighted above. However, there is consensus among all three
phylogenetic methods in revealing that the closest evolutionary relatives of spurfowls
are Old World quails (Coturnix spp.), the monotypic Madagascar ‘partridge’
Margaroperdix sp., Old World partridges such as Alectoris spp., Tetraogallus spp.,
55
Ammoperdix spp. and Perdicula spp. In contrast, francolins are most closely related to
the chickens (Gallus spp.) and Bamboo partridges (Bambusicola spp.)
One other major revelation regards what was traditionally considered a francolin
that is, Nahan’s Francolin Francolinus nahani. All three inference methods strongly
refute this in support of Nahan’s Francolin being sister to the Stone Partridge
Ptilopachus petrosus with 1.0 PP and 100% BS (Fig. 2.4, see also Cohen et al. 2012).
Also these results demonstrate strong support that the ‘duo’ form a sister relationship
with representatives of the New World Odontophorid quails, Callipepla, Colinus,
Oreortyx and Cyrtonyx spp. with 1.0 PP and 100% ML BS (Fig. 2.2, 2.3, respectively),
100% JK (Fig. 2.4) and 98 JK (Fig. 2.1).
Analyses of data partitions
Mitochondrial versus nuclear DNA analyses
The combined mt- and nucDNA BI topologies are similar to that of the mt- and
nucDNA ML topologies, and as such only the mt- and nucDNA ML trees with branchlengths are presented. The combined mtDNA ML (Fig. 2.5) and BI analyses recovered
similar topologies to that of the combined total evidence BI (Fig. 2.2) and ML (Fig. 2.2)
DNA analyses, supporting the five major galliform families in the same sequence of
evolution, although with slighly less support at several nodes. In both mtDNA BI and
ML (Fig. 2.5) analyses, there is further support for the split of francolins into francolins
and spurfowls, and strong support for the phylogenetic placement of Nahan’s Francolin
and P. petrosus as sister species. Further, both spurfowl and francolin taxa were
recovered as distinct monophyletic assemblages in the BI and ML mtDNA analyses;
these results are consistent among each of the data partitions (Table 2.4).
56
Both the combined nucDNA BI and ML (Fig. 2.6) analyses appear to have been
influenced by missing data for some taxa. For instance, one Cracid species (Crax spp.)
is nested in the Megapodes and the placement of Rollulus rouloul is uncertain. Another
difference is that even though the PPs are generally high at the deeper nodes in the BI
analyses, there is poor nodal support (less than 50% BP) in the ML tree coupled with
very short branch-lengths. The branch between Numididae and Odontophoridade is
unresolved.
However, despite some uncertanty at the base of the phylogeny for nucDNA in
the BI and ML analyses, the phylogenetic resolution of the focal taxa, that is francolins
and spurfowls and also Nahan’s Francolin largely remain the same with the exception
that the francolins are recovered as monophyletic in both the nucDNA BI and ML (Fig.
2.6) trees. Otherwise, the francolin-spurfowl dichotomy is maintained with spurfowls
being monophyletic (1.0 PP, 100% BS) and the relationship between Nahan’s Francolin
and P. petrosus is strongly supported (1.0 PP, 100% BS). Further, the sister relationship
between Nahan’s Francolin and P. petrosus and the New World quails is maintained
with 1.0 PP and 99% BS.
Divergence times
The molecular divergence dates (Fig. 2.7) of the deeper nodes point to the oldest
split being between the Cracids/Megapodes and the rest of the galliform species and
could have happened at around 65.0 mya (HPD 53.5-79.8). The Megapodes and Cracids
diverged from each other at c. 56.8 mya (HPD 42.9-72.4), the Numididae at 46.5 mya
(HPD 42.1-51.8) and the Odontophorids and Phasianids diverged at c. 44.5 mya (HPD
40.6-49.2).
57
The estimated date for the timing of the split between P. petrosus and F. nahani
was 9.6 mya (95% HPD 5.8-14.1), and 37.4 mya (95% HPD 31.8-43.1) for the
divergence between this clade and the New World Quails. The estimated divergence
date between the Gallus/Bambusicola/francolins and the Coturnix/Alectoris/spurfowls
clade is recovered at c. 33.6 mya (HPD 29.8-37.1) with the split between
Gallus/Bambusicola and francolins at around 10.7 mya (HPD 8.4-13.7) and
Coturnix/Alectoris and spurfowls at c. 23.1 mya (HPD 20.7-30.0). Interestingly, the
time to most recent common ancestor of extant spurfowls is around 8.7 mya (HPD 7.410.4) and that for francolins slightly younger at c. 7.6 mya (HPD 7.0-8.3). Whether this
could be correlated to habitats in which spurfowls and francolins thrive is a difficult
question to answer.
Vocal and behavioural comparison between F. nahani and P. petrosus
The calls of P. petrosus (Fig. 2.8a) and F. nahani (Fig. 2.8b) are strikingly similar and
differ from those of other francolins and spurfowls both sonographically and aurally:
the exemplars presented here are from the widely available Chappuis (2000). They
consist of a long series of whistles that increase in volume and are often joined by
additional birds calling near the end of the sequence. The structure of the whistle begins
with a short lead in tone between 1-1.5 kHz, followed by a double-peaked whistle with
high and low frequency values of 1.5 and 2.5 kHz respectively, and associated
harmonics (Fig. 2.8c-e). Interspecific variation based on our additional recordings is
limited and influenced largely by the number of group members calling simultaneously.
These calls differ qualitatively to a large degree from any other African galliform. Thus,
it is not possible to identify homologous call units to enable direct comparison among
58
Galliform lineages. No other African galliform examined has a similar whistle structure.
In particular, these calls strongly contrast with typical spurfowl calls of the putative
relatives of F. nahani, which consist of slurred, almost grating, raucous calls that do not
very much in frequency (Fig. 2.8c-e; from Chappuis 2000).
Behavioural observations (substantiated by photographs and extensive field
observations) indicate that both P. petrosus (Fig. 2.9a) and F. nahani (Fig. 2.9b) hold
their tails in a distinctive, bantam-like cocked position.
Discussion
Phylogenetic relationships
In general the phylogenetic results for this study, including the family status within
galliformes, are similar to those reported by Crowe et al. (2006), with the exception that
here species, and not genera, are terminals. With respect to the focal taxa, the supported
sister relationship between F. nahani and P. petrosus contradicts all other published
treatments of the Galliformes (e.g. Hall 1963, Crowe et al. 1985, 1986, 1992, Johnsgard
1988, del Hoyo et al. 1994, Madge and McGowan 2002). Furthermore, the basal
position of this clade relative to ‘true’ francolins and spurfowls suggests that they
represent a relictual group sister to the New World quails (Odontophoridae), and are
only distantly related to the other Old World galliforms. Intriguingly, both species
occupy habitats - dense primary forest understorey and rocky outcrops - that have been
suggested by Kingdon (1989) as having a higher than expected proportion of relictual
species.
Further, the present study confirms the results of Crowe and Bloomer (1998) and
Crowe et al. (2006) in suggesting that francolins and spurfowls are not each other’s
59
closest relatives as suggested by Hall (1963). The genus Francolinus comprises at least
two distantly related lineages (excluding F. nahani, see above). The partridge-like
spurfowls, which are related to quails (e.g. Coturnix, Excalfactoria, Margaroperdix,
Perdicula and Ammoperdix spp.) and certain Old World partridges (e.g. Tetraogallus
and Alectoris spp.), and the quail-like francolins (Dendroperdix, Peliperdix and
Scleroptila spp.), which are related to the ‘true’ Asian francolins (Francolinus spp.),
chickens (Gallus spp.), and other Old World partridges (Bambusicola spp.).
Morphological, behavioural and vocal similarities between F. nahani
and P. petrosus
Morphological similarities shared by F. nahani and P. petrosus, include: small size, red
bare skin around the eye, lack of spurs and the lack of sexual dimorphism (Hall 1963,
Johnsgard 1988, Madge and McGowan 2002). Although it is well known that
Ptilopachus has a long, vaulted and regularly cocked tail (Johnsgard 1988, del Hoyo et
al. 1994, Madge and McGowan 2002), the same condition in F. nahani is less well
known, because of its rarity and dense forest habitat (Stevenson and Fanshawe 2002).
Hence, most bird artists have depicted the shape of the bird as that of a typical francolin
or spurfowl (see illustrations in Crowe et al. 1986, del Hoyo et al. 1994, Sinclair and
Ryan 2003). Only one relatively recent publication (Stevenson and Fanshawe 2002) has
depicted the posture of this species correctly. This posture is illustrated (Fig. 2.9) based
on two photographs of F. nahani in natural habitat (in Budongo and Mabira Forests,
Uganda) and one P. petrosus taken in Cameroon. Dendroperdix sephaena is the only
other African galliform known to cock its tail (Madge and McGowan 2002), but it is not
closely related to these species (Crowe et al. 2006).
60
The biology of F. nahani is very poorly known (Crowe et al. 1986, Sande et al.
2009), and its voice has only been described relatively recently (Chappuis 2000,
Stevenson and Fanshawe 2002), thus hampering the correct taxonomic placement of
this species. The calls of both F. nahani and P. petrosus are a series of whistles
increasing in volume, and are strikingly similar (see Fig. 2.8). Chappuis (2000), in a
booklet accompanying his CD set, noted this similarity, as did Brian Finch, who worked
on the voice section of Stevenson and Fanshawe (2002). Furthermore, our field
observations attest that both species live in small, family groups and have interactive
calling.
Given the long divergence time between these two species (5-13 mya), it is
interesting that the nature of the calls have been so well conserved. The group duetting
may indicate a strong social cohesiveness function and the calls could be subject to
stabilizing selection in this regard (Payne 1971). Another matter to consider is the exact
nature of the habitat of these species. Whereas Ptilopachus is found in the arid zone, it
does inhabit dense bush growth among large boulders, a challenging environment for
the broadcast of sounds, with many obstacles, similar to the dense forest understorey
inhabited by F. nahani. Indeed, given the likely Miocene divergence between these
species, it is most likely that their common ancestor inhabited forest habitats (Fjeldså
and Bowie 2008). The open savannas and arid land lineages of mammals only seem to
have radiated later, in the Plio-Pleistocene when dry habitat became much more
widespread in Africa (e.g. deMenocal 2004). The plumage of these similar birds seems
to have been very well conserved, and besides aspects of colouration that presumably
relate to camouflage (F. nahani is darker above, whereas P. petrosus is somewhat
paler), there has been remarkably little divergence.
61
Historical biogeography
As expected the omission of the fossil Gallinuloides wyomingensis did result in the
recovery of younger divergence times. For example, Crowe et al. (2006) estimated that
the stem Ptilopachus plus Odontophoridae clade to have diverged at 55.5 mya (95%
HPD 50.1-65.9) whereas our analyses based on three ingroup fossils recovered this
node at 37.4 (95% HPD 31.7-43.0). Overall, these results are in agreement with the
view of Ksepka (2009), that although stem galliformes likely existed in the Cretaceous
(i.e. pre 65 mya), the divergence of crown-group lineages remains inconclusive.
Our estimated divergence between Ptilopachus and the New World quails
occurred around the middle of the Eocene period (55.8-33.9 mya). The Eocene was a
remarkable period in earth history, with high temperatures and precipitation, in
essentially an ice-free world (Eberle and Greenwood 2012, Harrington et al. 2012).
Connections existed between Africa and Europe, and Europe and North America via
Greenland, although by about 40 mya, the time of our inferred Ptilopachus-New World
quail split, it seems unlikely that this landbridge was still open (Scotese 2001).
However, Eocene and Oligocene fossils have been discovered from France that are most
similar to New World quails (Crowe and Short 1992, Crowe et al. 2006, see MourerChauvire 1992 for another view) suggesting that Europe likely played an important part
in the biotic exchange between African and North American lineages. Should the
‘Greenland landbridge’ have been closed, an alternate connection may have been via
Asia to North America along the Bering Strait. Given the relatively sedentary habitats
of both Ptilopachus and New World quails, it seems highly unlikely that direct dispersal
between African and the New World occurred, as for example inferred for some
lineages of birds such as Thrushes (Turdus spp., Voelker et al. 2009). In summary,
62
although difficult to infer without additional fossil evidence, it seems likely that one of
the above mentioned landbridges played an important role in shaping the biogeographic
origins of both the African and New World members of the Odontophoridae. At 5-13
mya the super-African rainforest was likely still expansive (Fjeldså and Bowie 2008)
which may suggest that P. petrosus either secondary invaded its present arid and rocky
habitat, or that it was more geographically restricted in the past, and that F. nahani
likely occupies the ancestral habitat of these taxa. Interestingly many of the extant New
World Quails are more open habitat associated (de Hoyo et al. 1994), with closer habitat
affinities to P. petrosus than F. nahani.
The relative divergence time between Gallus/Bambusicola/francolins and the
Coturnix/Alectoris/spurfowls clade is recovered at c. 33.6 mya whereas the split
between Gallus/Bambusicola and francolins and Coturnix/Alectoris and spurfowls was
recovered at around 10.7 and 23.1 mya respectively. Thus the diversification of,
francolins in Africa happened much more recently. Further, the divergence times among
the lineages in each of the major clades seem to have occurred consistently over roughly
the same period and given that the spurfowls evolved at around 8.7 mya and the
francolins c. 7.6 mya, this indicates that spurfowls are a little older than the francolins.
Most spurfowl species occur in forests and along forest edges, with only a few species
occurring in distinctive habitats such as the rocky hills preferred by P. hartlaubi and
semi-wooded grasslands preferred by members of the Bare-throated species complex.
Francolins on the other hand occur in different types of grasslands including, montane,
wooded and scrubby grassland, with the exception of one species, Francolinus lathami,
which is a forest specialist.
63
Taxonomic recommendations
On the basis of the close genetic relationship between F. nahani and P. petrosus, as well
as their shared behavioural and vocal characters, we recommend that F. nahani be
moved to the genus Ptilopachus Swainson (on the basis of priority). We recommend the
placement of Ptilopachus in the Odontophoridae to emphasize its sister relationship to
this New World family of galliform birds. Ptilopachus should be placed first in the
sequence of genera of the Odontophoridae.
With regard to the spurfowls and francolins, this study supports assignment of
the genus Pternistis to all spurfowl taxa. However, there seems to be a dire need to
revise the taxonomic status of francolins in particular, the distinct forest species F.
lathami, one of Hall’s phylogenetically enigmatic species, as well as the other two
enigmatic species, the Asian F. gularis and F. pondicerianus, which emerged as sister
species to the enigmatic African Dendroperdix sephaena. This will form the focus of a
later chapter in this thesis.
Conclusions
After several studies focusing on improving our understanding of the galliform
phylogeny, it is frustrating that relationships within the Phasianidae, in particular, the
position of the turkeys and grouse and allies, remains unresolved. However, this study
clearly demonstrates the support for the five families considered in modern galliform
classification thereby rejecting the recognition of the seven traditionally recognized
families (including Meleagrididae and Tetraonidae) in the order outlined in Crowe et al.
(2006) and Wang et al. (2013). As Wang et al. 2013 concluded “additional data
collection will be necessary to resolve the remaining uncertainties within Galliformes”,
64
they also consider the challenges that remain to be resolve within the galliform
phylogeny, to be aggravated by for example, the use of group names such as
”pheasants” and “partridges” which do not imply common descent. It would be safe to
conclude that the traditionally known francolins represent at least two distantly related
lineages with F. nahani being pulled to the base of the galliform tree to go with P.
petrosus. Thus, F. nahani is not a francolin as traditionally considered but a partridge
related to the New World quails.
65
Tables and Figures
Table 2.1. GenBank accession numbers of the samples analyzed in this study.
Genus
Species
Sample no.
Locality
CYTB
ND2
CR
12S
OVOG
TGFB
GAPDH
Acryllium
vulturinum
–
–
AF536742
AF536745
–
AF536739
DQ832070
–
–
Afropavo
congensis
–
–
AF013760
DQ768253
DQ834507
–
AF170991
–
–
Alectoris
chukar
–
–
L083781
DQ768273
DQ834525
–
AF170987
FR694121
FR694070
Alectoris
Alectoris
Alectura
Arborophila
graeca
rufa
lathami
javanica
–
–
Z487724
–
DQ834524
–
–
–
–
–
–
–
–
Z487754
NC007227
–
AY274051
DQ834523
DQ834465
–
AY274004
AF170988
DQ832069
–
EU737326
–
–
–
–
AM236890
DG093804
–
DQ832097
DQ832074
–
–
Arborophila
torqueola
–
–
AM236889
–
DQ834475
–
–
–
–
Bambusicola
thoracica
–
–
EU165706
AF222538
DQ834513
EU165706
AF170978
–
–
Bonasa
Callipepla
Callipepla
umbellus
californica
gambelii
–
–
AF230167
AF222541
DQ834476
U83740
–
–
–
–
–
AB120131
AF028773
DQ834473
–
–
Submitted
Submitted
–
–
L083821
AF028761
DQ834472
–
–
–
–
Catreus
wallichii
–
–
AF028792
DQ768254
DQ834499
–
AF170980
–
–
Chrysolophus
amherstiae
–
–
AB120130
–
–
DQ832102
–
–
–
Chrysolophus
pictus
–
–
AF028793
DQ768255
DQ834497
–
–
–
–
Colinus
cristatus
–
–
–
–
–
–
–
EU737357
–
Colinus
virginianus
–
–
EU372675
AF222545
DQ834469
AF222576
–
–
Submitted
Coturnix
coturnix
–
–
L083771
X57246
DQ834529
X57245
–
Submitted
EU737363
Coturnix
japonica
–
–
NC003408
NC003408
–
NC003408
–
–
–
Crax
alector
–
–
AY141921
–
–
–
–
Submitted
–
Crax
rubra
–
–
AY956378
AY274050
AY145307
AY274003
–
–
–
Crossoptilon
crossoptilon
–
–
AF028794
DQ768256
DQ834500
–
AF170981
–
Cyrtonyx
montezumae
–
–
AF068192
AF028779
DQ834467
–
AF170976
–
–
Dendroperdix
sephaena
TMC9
Marico, SA
FR694140
DQ768274
DQ834515
FR691559
DQ832083
FR694111
FR694102
Falcipennis
canadensis
–
–
AF170992
AF222548
DQ834478
AF222577
AF170986
–
–
Francolinus
francolinus
FR691376
–
–
–
–
Francolinus
gularis
Francolinus
lathami
Francolinus
pictus
AMNH 776813
India
Francolinus
pondicerianus
Gallus
gallus
Gallus
AF013762
AMNH DOT8023
India
U90649
India
U906497
–
–
–
–
–
–
Cameroon
AM236893
DQ768257
FR691377
FR691546
DQ832082
FR694113
FR694080
FR694142
–
–
–
–
–
–
AMNH DOT8050
India
FR691632
DQ768279
–
FR691547
DQ832081
FR694114
FR694081
–
–
L083761
AB086102
DQ834510
NC001323
AF170979
FR694110
FR694078
varius
–
–
AB044988
AF222551
–
–
–
–
–
Guttera
pucherani
–
–
AM236882
–
–
–
–
–
–
Ithaginis
cruentus
–
–
AF068193
DQ768258
DQ834487
–
DQ832076
–
–
Leipoa
ocellata
–
–
AM236879
AF394619
–
AF222586
–
–
–
Lophophorus
impejanus
–
–
AF028796
DQ768259
DQ834486
DQ832098
DQ832075
–
–
Lophura
nycthemera
–
–
L083801
DQ768261
DQ834498
–
–
–
–
Margaroperdix
madagarensis
–
–
U906407
–
DQ834528
–
–
–
–
Megapodius
eremita
–
–
AF082065
AY274052
–
AY274005
–
–
–
Meleagris
gallopavo
–
–
L083811
AF222556
DQ834485
U83741
AF170984
–
–
Numida
meleagris
–
–
L083831
NC006382
DQ834466
AF222587
AF170975
EU737410
FR694071
Oreortyx
pictus
–
–
AF252860
AF028782
DQ834468
–
AF170977
Submitted
Submitted
Ortalis
vetula
–
–
L083841
AF394614
–
–
AF170974
–
–
Pauxi
pauxi
–
–
AF068190
AY140750
AF165439
AF165449
AF170973
–
–
Pavo
cristatus
–
–
L083791
AF394612
DQ834508
AY722396
AF170990
–
–
Peliperdix
coqui
PFIAO 45
Settlers, SA
AM236895
DQ768278
FR691379
FR691549
DQ832084
FR694115
FR694082
Perdix
perdix
–
–
AF028791
AF222560
DQ834484
AF222590
AF170982
–
–
Phasianus
colchicus
–
–
AY368060
AF222561
DQ834495
U837426
–
–
–
Polyplectron
bicalcaratum
–
–
AF534564
DQ768263
DQ834503
–
AF331959
–
–
Polyplectron
emphanum
–
–
AF330062
DQ768265
DQ834504
–
AF331955
–
AM236893
66
–
Genus
Species
Sample no.
Locality
CYTB
ND2
CR
12S
OVOG
TGFB
GAPDH
Pternistis
adspersus
PFIAO 206A
–
AM236910
DQ768276
DQ834535
DQ832113
DQ832095
FR694122
FR694087
Pternistis
afer
AM236908
DQ768281
DQ834533
DQ832111
DQ832092
FR694123
FR694088
Pternistis
bicalcaratus
U906377
FR691578
FR691370
FR691551
FR691690
FR694103
FR694089
Pternistis
camerunensis
TMC 42
Mount Cameroon
FR694142
FR691577
FR691382
FR691552
FR691694
FR694124
FR694090
Pternistis
capensis
PFIAO 229
Kakamas, SA
AM236909
DQ768282
DQ834534
DQ832112
DQ832093
FR694125
FR694091
Pternistis
castaneicollis
GB
–
AM236903
–
–
–
–
–
–
Pternistis
clappertoni
AMNH 541305
Takoukout,
Cameroon
FR691602
FR691576
FR691383
FR716655
FR691693
FR694126
FR694092
Pternistis
erckelii
AMNH DOT11039
Ethiopia
U906387
–
–
–
–
–
–
Pternistis
griseostriatus
Ndalla Tanda
AM236905
DQ768284
FR691384
FR691554
DQ832089
FR694128
FR694094
Pternistis
hartlaubi
Namibia
U906397
FR691572
FR691555
FR691692
FR694129
FR694095
Pternistis
hildebrandti
–
U906317
–
–
–
–
–
–
Pternistis
icterorhynchus
Fanadji
FR691601
–
–
–
–
–
–
Pternistis
jacksoni
East slope, Mt
Kenya
FR691594
–
–
–
–
–
–
Kenya
AM236906
FR691387
FR691556
DQ832090
FR694131
FR694097
FR694097
Marico River, SA
AM236911
DQ834536
FR691557
DQ832094
FR694132
FR694098
FR694098
West Ruwenzori
FR691592
–
–
–
–
–
–
Djibouti
FR691590
–
–
–
–
–
–
PFIAO 108
TM 14682
AMNH 541411
TMC 121
GB
AMNH 156922
AMNH26192
Watervalboven,
SA
Gold Coast,
Hinterland
Pternistis
leucoscepus
PFIAO 109
Pternistis
natalensis
TMC 120
Pternistis
nobilis
AMNH1759
Pternistis
ochropectus
FNHM 1971-1072
Pternistis
rufopictus
AMNH 202503
Gagayo, Muranza
FR691588
–
–
–
–
–
–
Pternistis
squamatus
AMNH 541409
Nr York Pass,
Sierra Leone
AM236904
DQ768286
DQ834531
DQ832109
DQ832088
FR694133
FR694099
Pternistis
swainsonii
TMC 40
Marico River, SA
AM236907
DQ768287
DQ834532
DQ832110
DQ832091
FR694134
FR694100
Pternistis
swierstrai
TMC 67
Angola
FR691593
–
–
–
–
–
–
Ptilopachus
nahani
–
Budongo forest,
Uganda
AM236885
DQ768288
FR691374
FR691545
DQ832071
FR694107
FR694075
Ptilopachus
petrosus
–
Ghana
AM236886
DQ768289
FR691375
FR691544
DQ832072
FR694108
FR694076
Pucrasia
macrolopha
–
–
AF028800
DQ768269
DQ834490
–
AF170983
–
–
Rollulus
rouloul
–
–
AM236888
–
–
–
–
Submitted
–
Scleroptila
afra
PFIAO 59
Eastern Cape, SA
AM236897
AF222550
DQ834517
AF222581
DQ832086
FR694116
FR694083
Scleroptila
finschi
AMNH 308887
Angola
AM236896
DQ768290
–
–
–
–
–
Scleroptila
levaillantii
TM 78622
Sterkspruit, SA
AM236913
DQ768291
DQ834516
DQ832106
DQ832085
FR694117
FR694084
Scleroptila
levaillantoides
TMC 12
Petrus steyn, SA
AM236900
DQ768292
DQ834519
DQ832108
–
FR694118
FR694085
Scleroptila
psilolaema
Kenya
FR691614
–
–
–
–
–
–
Scleroptila
shelleyi
PFIAO 47
Ayton farm, SA
AM236898
DQ768295
DQ834518
DQ832107
DQ832087
FR694119
FR694101
Scleroptila
streptophora
TMC11
Cameroon
FR691617
FR691573
FR691380
FR691550
–
FR694120
FR694086
Syrmaticus
ellioti
–
–
–
DQ768270
–
–
–
–
–
Syrmaticus
humiae
–
–
AF534706
–
DQ834491
DQ832099
DQ832077
–
–
Tetrao
urogallus
–
–
AB120132
AF222565
DQ834480
AF222594
–
–
–
Tragopan
temminckii
–
–
AF229838
AF222566
DQ834488
AF222595
–
–
–
Tympanuchus
Xenoperdix
phasianellus
udzungwensis
–
–
AF068191
AF222569
DQ834483
AF222598
AF170985
–
–
–
–
AM236887
DG093800
DQ834474
DQ832096
DQ832073
–
–
BM 80 1 1 1066
67
Table 2.2. DNA markers sequenced and primers used for PCR amplifications
and sequencing of preserved tissue.
Primer name
Primer sequence (5’to 3’)
Reference
All Galliformes (General primers)
Cytochrome b
L14578
MH15364
ML15347
H15915
cta gga atc atc cta gcc cta ga
act cta cta ggg ttt ggc c
atc aca aac cta ttc tc
aac gca gtc atc tcc ggt tta caa gac
J.G. Groth (pers. comm.)
P. Beresford (pers. comm.)
P. Beresford (pers. comm.)
Edwards & Wilson (1990)
Control region
PHDL
PH-H521
PH-L400
PHDH
agg act acg gct tga aaa gc
tta tgt gct tga ccg agg aac cag
att tat tga tcg tcc acc tca cg
cat ctt ggc atc ttc agt gcc
Fumihito et al. (1995)
E.A. Scott (pers. comm.)
E.A. Scott (pers. comm.)
Fumihito et al. (1995)
12S rRNA
L1267
H2294
aaa gca tgg cac tga ag(atc) tg
gtg cac ctt ccg gta cac ttac c
Moum et al. (1994)
O. Haddrath (S. Pereira
pers. comm.)
NADH dehydrogenase subunit 2 (ND2)
L5216
gcc cat acc ccr aaa atg
H6313
ctc tta ttt aag gct ttg aag gc
Sorenson et al. (1999)
Sorenson et al. (1999)
Ovomucoid G
OVO-G Forward
OVO-G Reverse
caa gac ata cgg caa caa rtg
ggc tta aag tga gag tcc crt t
Armstrong et al. (2001)
Armstrong et al. (2001)
GAPDH intron-11
GapdL890
GapdH950
acc ttt aat gcg ggt gct ggc att gc
cat caa gtc cac aac acg gtt gct gta
Friesen et al. (1997)
Friesen et al. (1997)
Transforming Growth Factor Beta2 intron-5
TGFb2-5F
ttg tta ccc tcc tac aga ctt gag tc
TGFb2-6R
gac gca ggc agc aat tat cc
Primmer et al. (2002)
Primmer et al. (2002)
Cytochrome b
Spurfowl-specific primers
L14851 (General)
Pt-H195
Pt-H194
MH15145
cct act tag gat cat tcg ccc t
ttt cgr cat gtg tgg gta cgg ag
cat gtr tgg gct acg gag g
aag aat gag gcg cca ttt gc
Kornegay et al. (1993)
R. Moyle & T. Mandiwana-Neudani
R. Bowie
P. Beresford
Pt-L143
Pt-H361
gcc tca tta ccc aaa tcc tca c
gtg gct att agt gtg agg ag
R. Moyle & T. Mandiwana-Neudani
R. Moyle & T. Mandiwana-Neudani
Pt-L330
Pt-H645
tat act atg gct cct acc tgt ac
ggg tgg aat ggg att ttg tca gag
R. Bowie
R. Moyle & T. Mandiwana-Neudani
Pt-L633
Pt-H901
ggc tca aac aac cca cta ggc
agg aag ggg att agg agt agg at
R. Moyle & T. Mandiwana-Neudani
R. Moyle & T. Mandiwana-Neudani
68
Primer name
Primer sequence (5’to 3’)
Reference
Cytochrome b
Spurfowl-specific primers
L2-2312
H15696
cat tcc acg aat cag gct c
aat agg aag tat cat tcg ggt ttg atg
R. Bowie
Edwards et al. (1991)
Pt-L851alt
Pt-H1050
cct att tgc cta cgc cat cct ac
gat gct gtt tgg ccg atg
R. Bowie
R. Bowie
Pt-L961
Pt-L961alt
HB20 (General)
cga acc ata aca ttc cca c
ctc atc cta ctc cta atc ccc
ttg gtt cac aag acc aat gtt
R. Moyle & T. Mandiwana-Neudani
R. Bowie
J. Feinstein (pers. comm.)
69
Table 2.3. Cladistic placement of phylogenetically enigmatic phasianine galliforms in
concatenated analyses.
Source of
data
Parsimony
Concatenated
organismal,
F. nahani
Meleagris
Ithaginis
Tragopan
Pucrasia
Perdix
sister to NWQ
sister to
sister to
sister to
sister to
sister to
Tetraonini
Pucrasia
Tetraophasis +
Ithaginis
true pheasants
mt & nucDNA
Lophophorus
Bayesian
inference
mt & nucDNA
sister to P. petrosus
sister to
sister to
sister to
sister to
sister to
Tetraonini
large Phasianine
Lophophorus
Perdix +
true pheasants
assemblage
mtDNA
sister to P. petrosus
true pheasants
sister to
sister to
sister to
sister to
sister to
Tetraonini
large Phasianine
Lophophorus
Perdix +
true pheasants
assemblage
nucDNA
Maximum
likelihood
mt & nucDNA
mtDNA
nucDNA
sister to P. petrosus
sister to P. petrosus
sister to P. petrosus
sister to P. petrosus
sister to
sister to
Tetraonini
true pheasants
_
polytomous with
polytomous with
large Phasianine
assemblage of
assemblage of
assemblage
Phasianine spp.
Phasianine spp.
_
sister to
sister to
sister to
sister to
Tetraonini
large Phasianine
true
true
assemblage
pheaseants
pheasants
sister to
sister to
sister to
sister to
sister to
Tetraonini
large Phasianine
true
true
true pheasants
pheasants
pheasants
sister to
assemblage
paraphyletic
with
_
paraphyletic with
paraphyletic with
large Phasianine
Assemblage
large Phasianine
assemblage
Lophophorus
70
large Phasianine
assemblage
Table 2.4. Support for the relationship of Ptilopachus petrosus and Francolinus nahani from
different data partitions, + indicates supported branch; U - unresolved (Adopted from Cohen et
al. 2012).
Bayesian
Inference
Parsimony
Maximum
Likelihood
Clade
mtDNA
nucDNA
mtDNA
nucDNA
mtDNA
nucDNA
CYTB
CR
ND2
12S
OVOG
TGFB2
GAPDH
Odontophoridae
sister to P.
petrosus & F.
nahani
1.0
1.0
100
100
+
+
+
+
+
+
+
+
+
P. petrosus sister
to F. nahani
1.0
1.0
100
100
+
+
+
+
+
+
+
+
U
71
Figure 2.1. Summary of the single strict consensus parsimony cladogram recovered for
the Galliformes by Crowe et al. (2006). Anseriforms, Megapodiidae, Cracidae,
Guineafowls Numididae, New World quails sensu stricto Odontophorines, Ptilopachus
spp., XEN/ROLL/ARB (Xenoperdix + Rollulus + Arborophila spp.),
COT/MAD/ALEC/TETG (Coturnix + Excalfactoria + Margaroperdix + Alectoris +
Tetraogallus spp.), Perdicula + Ammoperdix spp., Spurfowls Pternistis spp., Gallus +
Bambusicola spp., Francolins Francolinus, Ortygornis, Afrocolinus, Peliperdix,
Scleroptila spp., Rhenardia + Argusianus spp., Pavo + Afropavo spp., Polyplectron
spp., Perdix spp., Turkey Meleagris gallopavo, Grouse Tetraonini spp., Lophophorus
spp., Pucrasia macrolopha, Tragopan spp., Ithaginis cruentus, Wattled – Phesants
(Phasianus, Lophura, Syrmaticus, Catreus, Crossoptilon Chrysolophus spp. Numbers at
nodes are jackknife support values. See text for meaning of circled numbers (1-10) on
clades.
72
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
.96
.97
1.0
.96
.99
1.0
1.0
.97
1.0
.99
.99
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
.99
.98
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
.98
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Figure 2.2. The 50% Bayesian majority-rule consensus tree generated from the combined four
mitochondrial (Cytochrome-b – CYTB, NADH Dehydrogenase Subunit 2 - ND2, 12S Ribosomal DNA 12S, Control Region - CR) and three nuclear (Ovomucoid intron G – OVOG, Transforming Growth Factor
Beta 2 intron 5 – TGFB, GAPDH intron 11 - GAPDH) DNA markers. Numbers associated with nodes
represent posterior probability values (≥ 0.95 are shown).
73
100
100
Alectura lathami
Leipoa accelata
Megapodiidae
Megapodius sp.
Ortalis
sp.
100
Crax sp.
Cracidae
100
Pauxi pauxi
Acryllium vulturinum
100
Numididae
Guttera pucherani
Numida meleagris
Afropavo congensis
100
Pavo cristatus
Bambusicola thoracica
95
Gallus gallus
99
Gallus varius
Peliperdix coqui
F
Scleroptila streptophora
97
99
r
Scleroptila psilolaema
100
Scleroptila
finschi
a
100
Scleroptila levaillantoides
n
Scleroptila shelleyi
74
78
Scleroptila afra
74
c
Scleroptila levaillantii
o
Francolinus pictus
89
Francolinus francolinus
l
Francolinus lathami
i
Francolinus pondicerianus
100
99
Francolinus gularis
n
Dendroperdix sephaena
Margaroperdix
madagarensis
s
100
100 Coturnix japonica
Coturnix coturnix
Pternistis leucoscepus
89
Pternistis swainsonii
94
Pternistis
rufopictus
77
Pternistis afer
S
P
Pternistis hartlaubi p
h
Pternistis
squamatus
100
Pternistis icterorhynchus u
100
99
a
100
Pternistis bicalcaratus
r
s
Pternistis griseostriatus
100
Pternistis camerunensis
100
f
i
Pternistis nobilis
o
a
Pternistis castaneicollis
90
Pternistis ochropectus
79
w
n
Pternistis erckelii
100
74
l
Pternistis swierstrai
i
Pternistis clappertoni
s
d
Pternistis jacksoni
Pternistis adspersus
a
97
Pternistis natalensis
99
e
Pternistis hildebrandti
Pternistis capensis
Alectoris rufa
100
95
Alectoris chukar
Alectoris graeca
Polyplectron
emphanum
100
Polyplectron bicalcaratum
Bonasa umbellus
100
Tympanuchus phasianellus
100
Falcipennis canadensis
86
100
96
Tetrao urogallus
Meleagris gallopavo
Pucrasia macrolopha
Chrysolophus sp.
Crossoptilon crossoptilon
77
88 80
Lophura nycthemera
Catreus wallichii
100
85
Phasianus colchicus
Syrmaticus sp.
100
Perdix perdix
91
Lophophorus impejanus
Tragopan temminckii
Ithaginis cruentus
Xenoperdix udzungwensis
99
Rollulus rouloul
81
Arborophila
javanica
100
Arborophila torqueola
Cyrtonyx montezumae
Colinus sp.
100
100
Callipepla
californica
100
95
Callipepla gambelii
100
Odontophoridae
Oreortyx pictus
Ptilopachus
nahani
100
Ptilopachus petrosus
Figure 2.3. Maximum likelihood tree obtained from combined four mitochondrial (Cytochrome-b
– CYTB, NADH Dehydrogenase Subunit 2 - ND2, 12S Ribosomal DNA - 12S, Control Region CR) and three nuclear (Ovomucoid intron G – OVOG, Transforming Growth Factor Beta 2 intron
5 – TGFB, GAPDH intron 11 - GAPDH) DNA markers. Numbers associated with nodes are
bootstrap support values (only ≥ 70% are shown).
74
Ptilopachus petrosus
Francolinus nahani
Odontophoridae
Other Phasianines
Numididae
Megapodiidae
Figure 2.4. Parsimony cladogram for Ptilopachus petrosus and Francolinus nahani.
Numbers at nodes are parsimony jackknife/maximum likelihood bootstrap/Bayesian
probability values.
75
Pauxi pauxi
Cracidae
Crax rubra
Ortalis sp.
∞
Acryllium vulturinum
Numididae
Numida meleagris
*
∞
Guttera pucherani
Cyrtonyx montezumae
Oreortyx pictus
*∞
Colinus virgianianus
* *
∞
Odontophcallfornica
*
∞ * Callipepla
*
Callipepla callfornica
∞
oridae
∞ Ptilopachus
∞
nahani
*∞
Ptilopachus petrosus
Ithaginis cruentus
Pucrasia macrolopha
Syrmaticus sp.
Phasianus colchicus
Lophura nycthemera
*
∞
Catreus wallichii
*
**
*∞
∞
∞
Crossoptilon crossoptilon
Chrysolophus sp.
Perdix perdix
Tragopan temminckii
Lophophorus impejanus
*
Bonasa umbellus
∞
Tympanuchus phasianellus
* *
∞
Tetrao urogallus
∞*
*∞
Falcipennis
canadensis
∞
Meleagris gallopavo
Pavo cristatus
*
Afropavo congensis
∞
*∞
Gallus varius
*
gallus
*
∞ Gallus
*
∞
Bambusicola thoracica
∞
Francolinus lathami
F
*
Francolinus pictus
∞
*∞ Francolinus
r
francolinus
Francolinus pondicerianus
a
* * ∞*
Francolinus gularis
n
Dendroperdix sephaena
∞
Peliperdix coqui
c
Scleroptila streptophora
*
o
∞
Scleroptila
levaillantii
*
*
Scleroptila shelleyi
∞
l
*
∞
Scleroptila
levaillantoides
∞ *∞ Scleroptila afra
i
∞
*
Scleroptila psilolaema
∞
n
Scleroptila finschi
∞
s
Polyplectron bicalcaratum
*
Polyplectron emphanum
∞
Pternistis capensis
adspersus
*∞ Pternistis
Pternistis hildebrandti
*∞ Pternistis
natalensis
S
Pternistis griseostriatus
Pternistis hatlaubi p
*
∞
∞
Pternistis squamatus
∞
Pternistis icterorhynchus u
∞
* Pternistis
bicalcaratus
r
∞
Pternistis swierstrai
f
Pternistis ochropectus
o
erckelii
*∞∞* Pternistis
castaneicollis
w
*∞ ∞* Pternistis
Pternistis nobilis
*
Pternistis camerunensis l
∞
Pternistis clappertoni
s
Pternistis jacksoni
Pternistis
swainsonii
* Pternistis rufopictus
∞
* Pternistis afer
∞∞
Pternistis leucoscepus
Margaroperdix madagarensis
*∞
*
Coturnix coturnix
∞
* Coturnix
japonica
*∞ ∞Alectoris
graeca
*
Alectoris
chukar
∞
*
∞ Alectoris rufa
Arborophila javanica
*
Arborophila torqueola
∞
*
Rollulus
rouloul
∞
*
∞
Xenoperdix udzungwensis
Leipoa ocellata
Alectura lathami Megapodiidae
*
∞
∞
*
∞
Megapodius sp.
Figure 2.5. Maximum likelihood tree obtained from combined four mitochondrial (Cytochrome-b – CYTB,
NADH Dehydrogenase Subunit 2 - ND2, 12S Ribosomal DNA - 12S, Control Region - CR) DNA markers.
Symbols above branches (*) represent bootstrap support values (only ≥ 70% are presented) and symbols
below branches (∞) represent the Bayesian posterior probabilities (only ≥ 0.95 are presented) extracted from
the 50% majority-rule consensus tree.
76
P
h
a
s
i
a
n
i
d
a
e
Crax sp.
*
*
∞
∞
Al. lathami
Megapodius sp.
Megapodiidae
Pauxi pauxi
Cracidae
Ortalis sp.
Numida meleagris
Numididae
*∞
Acryllium vulturinum
Ptilopachus petrosus
*
Ptilopachus nahani
∞
Colinus sp.
Odontophoridae
*
*
Callipepla
callfornica
∞
∞
Oreortyx pictus
∞
Cyrtonyx montezumae
Coturnix coturnix
Pternistis hatlaubi
Pternistis capensis
S
*∞ Pternistis adspersus
p
Pternistis natalensis
u
* *
Pternistis griseostriatus
∞
r
∞
Pternistis squamatus
f
∞* Pternistis camerunensis
o
Pternistis bicalcaratus
w
∞
Pternistis leucoscepus
l
*
* ∞ * Pternistis swainsonii
s
*
∞
∞ Pternistis clappertoni
Pternistis afer
Alectoris chukar
*∞
*
∞ Alectoris rufa
Xenoperdix udzungwensis
F
Scleroptila shelleyi
r
*∞ * Scleroptila levaillantoides
a
afra
* ∞ Scleroptila
n
∞
Scleroptila streptophora
c
Scleroptila levaillantii
Phasianidae
o
Peliperdix coqui
l
Dendroperdix sephaena
*
i
∞ *
Francolinus pondicerianus
*∞
∞
n
Francolinus lathami
s
Bambusicola thoracica
* Gallus gallus
∞
Ithaginis cruentus
Perdix perdix
Pucrasia macrolopha
Syrmaticus sp.
*∞ Crossoptilon crossoptilon
Catreus wallichii
∞*
Tympanuchus phasianellus
*
∞ Falcipennis canadensis
Lophophorus impejanus
*
Meleagris gallopavo
∞
Polyplectron bicalcaratum
*
Polyplectron emphanum
∞
Pavo cristatus
*
Afropavo congensis
∞
Rollulus rouloul
*
∞
*
∞
*
∞
Figure 2.6. Maximum likelihood tree obtained from combined three nuclear (Ovomucoid intron G – OVOG,
Transforming Growth Factor Beta 2 intron 5 – TGFB, GAPDH intron 11 - GAPDH) DNA markers. Symbols
above branches (*) represent boostrap support values (only ≥ 70% are presented) and symbols below branches
(∞) represent the Bayesian posterior probabilities (only ≥ 0.95 are presented as extracted from the 50%
majority-rule consensus tree.
77
37.4
9.3
46.5
34.2
B
44.5
34.2
30.1
32.1
40.9
37.6
23.1
65.0
35.2
8.7
25.2
33.6
36.3
7.2
7.6
A
31.6
C
19.5
13.7
56.8
11.9
5.4
Numida meleagris
Numididae
Guttera pucherani
Acryllium vulturinum
Cyrtonyx montezumae
Callipepla californica
Callipepla gambelii
OdontopColinus sp.
Oreortyx pictus
horidae
Ptilopachus nahani
Ptilopachus petrosus
Arborophila javanica
Arborophila torqueola
Rollulus rouloul
Xenoperdix udzungwensis
Ithaginis cruentus
Tympanuchus phasianellus
Tetrao urogallus
Falcipennis canadensis
Bonasa umbellus
Meleagris gallopavo
Tragopan temminckii
Pucrasia macrolopha
Perdix perdix
Syrmaticus sp.
Phasianus colchicus
Crossoptilon crossoptilon
Catreus wallichii
Lophura nycthemera
Chrysolophus sp.
Lophophorus impejanus
Afropavo congensis
Pavo cristatus
Alectoris chukar
Alectoris graeca
Alectoris rufa
Pternistis capensis
Pternistis hildebrandti
Pternistis natalensis
Pternistis adspersus
P
Pternistis clappertoni
S
Pternistis jacksoni
h
p
Pternistis griseostriatus
a
Pternistis hartlaubi
u
Pternistis bicalcaratus
s
Pternistis icterorhynchus
r
i
Pternistis squamatus
f
a
Pternistis camerunensis
o
Pternistis nobilis
n
Pternistis erkelii
w
Pternistis ochropectus
i
l
Pternistis castaneicollis
d
Pternistis swierstrai
s
a
Pternistis afer
Pternistis rufopictus
e
Pternistis leucoscepus
Pternistis swainsonii
Coturnix coturnix
Coturnix japonica
Margaroperdix madagarensis
Francolinus lathami
F
Francolinus francolinus
r
Francolinus pictus
Francolinus gularis
a
Francolinus pondicerianus
n
Dendroperdix sephaena
Peliperdix coqui
c
Scleroptila finschi
o
Scleroptila psilolaema
Scleroptila afra
l
Scleroptila levaillantoides
i
Scleroptila shelleyi
Scleroptila levaillantii
n
Scleroptila streptophora
s
Gallus gallus
Gallus varius
Bambusicola thoracica
Polyplectron bicalcaratum
Polyplectron emphanum
Alectura lathami
Leipoa ocellata
Megapodiidae
Megapodius sp.
Megapodiidae
Crax sp.
Pauxi pauxi Cracidae
Ortalis sp.
Figure 2.7. The Bayesian reconstruction of divergence dates with uncorrelated lognormal molecular clock.
Numbers (only at critical nodes) at nodes are divergence dates estimated with three calibration points, (A) a
fossil of Crested Francolin Dendroperdix sephaena at 4.5-5.0 mya (Crowe 1992), (B) a basal date for the
Tetraoninae (Grouse and allies) of 27-29 mya (Crowe and Short 1992), and (C) a basal date for Polyplectron
(Peacock-Pheasants) of 34-36 mya (Olson 1974, modified by T. M. Crowe unpubl. data).
78
a
b
c
d
e
Figure 2.8. Sonograms of the call of (a) Stone Partridge Ptilopachus petrosus, (b) Nahan’s Francolin
Francolinus nahani, (c) Scaly Francolin Pternistis squamatus, (d) Ahanta Francolin Pternistis ahantensis
and (e) Red-necked Spurfowl Pternistis afer. Frequency (kHz) on the vertical axis with time (seconds) on the
horizontal axis.
79
,,)
(b)
Figure 2.9. Line drawing to show the posture of (a) Ptilopachus petrosus and (b) Francolinus
nahani, after photographs by Callan Cohen, Ron Hoff and Nik Borro.
80
CHAPTER 3
A study of gross morphological and histological
syringeal features of true francolins (Galliformes:
Francolinus, Ortygornis, Afrocolinus, Peliperdix and
Scleroptila spp.) and spurfowls (Pternistis spp.) in a
phylogenetic context
Abstract
Modern taxonomies of francolins recognize 41 congeneric species, forming the largest
genus of terrestrial gamebirds (Galliformes). Recent molecular, ecological and
behavioural studies challenge this view, suggesting that they comprise two unrelated,
monophyletic groups. There are 'true' francolins (Francolinus, Ortygornis, Afrocolinus,
Peliperdix and Scleroptila spp.) that are relatively small, ground-roosting birds (with the
exception of Ortygornis sephaena which roosts in trees), and spurfowls (Pternistis spp.)
that are large birds that can roost in trees.
This study explores gross morphological and histological syringeal anatomy of
francolins, spurfowls and their respective sister taxa to test whether differences are
concordant with a molecular-based hypothesis. Differences found were the presence of
a shield- versus diamond-shaped tympanum among francolins and spurfowls,
respectively. The first bronchial half rings are mineralized among francolins except in
O. sephaena, whereas almost no mineral deposition was observed among spurfowls.
81
Histologically, francolins have a small, rounded pessulus (except in O. sephaena, which
has a rounded, larger pessulus) in contrast to the larger pessulus observed among
spurfowls, which is rounded and triangular in Pternistis capensis and P. natalensis.
Both gross and histological similarities within, and differences between, francolin and
spurfowl syringes support this division. However, O. sephaena shows intermediate
features between francolins and spurfowls.
82
Introduction
Syringeal characters in taxonomy and systematic studies
The syrinx or avian voice box is located in the base of the neck, at the junction of the
trachea and bronchi (Ames 1971, Lewis 1983, Seller 1983, Gill 1990). It can be formed
from either the tracheal or bronchial tissues, or both. Myers (1917) categorized three
types of syrinx based on its location relative to the trachea and the bronchi: ‘syrinx
trachealis’ if it is found at the lower end of the trachea, ‘syrinx bronchialis’ in the case
where the syrinx is located below the bifurcation, and ‘syrinx trachea-bronchialis’ when
the syrinx is located at a position that includes both the lower end of the trachea and the
upper parts of the bronchi.
The syringeal structure of songbirds (Passeriformes) has been widely compared
with that of non-songbirds (Frank et al. 2006), with passerine birds known to produce
complex vocalizations as opposed to the relatively simple vocalizations given by many
non-passerine birds. Whether the complexity of the vocalizations is dependent on the
number or the complexity of the syringeal components is a question that requires
detailed functional investigation of each part. This is because certain non-passerine
species, e.g. cockatiels Nymphicus hollandicus, which are known to have only three
pairs of syringeal muscles and two pairs of tracheal muscles (Larsen and Goller 2002),
can mimic many types of sound (Tsukahara et al. 2008). The presence or absence of
certain components of the syringes or their musculature, contribute largely to voice
production, and has been found to play a significant role in the classification of birds.
This is supported by evidence in studies of the syringes of the Red jungle-fowl
(Chicken) Gallus gallus, the male Mallard Duck Anas platyrhynchos and the Greater
83
Sage-Grouse Centrocercus urophasianus (Myers 1917, Frank et al. 2006, Krakauer et
al. 2009, respectively).
The syrinx is anatomically complex and interspecifically diverse even in species
that lack special structures. Syringeal morphology has been found to be informative in
many systematic studies on passerine birds, e.g. Tyrannidae (Lanyon 1986, Prum and
Lanyon 1989, Mobley and Prum 1995), Pipridae (Prum 1992) and Furnariidae (Zimmer
et al. 2008), as well as in non-passerines, e.g. Anatidae (Delacour and Mayr 1945,
Humphrey 1955, Johnsgard 1961, Livezey 1986), Charadriiformes (Brown and Ward
1990), Falconidae (Griffiths 1994a, 1994b) and Psitacidae (Gaban-Lima and Höfling
2006). As an example of the utility of syringeal characters in non-passerines, Livezey
(1986) in his phylogenetic analysis of Anseriformes studied tracheal characters and
found several important synapomorphies with which to deliniate clades.
Because the demonstrated utility of syringeal characters in studies of
phylogenetic relationships among non-passeriform taxa and in the spirit of using
multiple lines of evidence to test hypotheses of monophyly (Templeton 1989, Farais et
al. 2000, Pruett and Winker 2010), this study explored the gross anatomy and histology
of francolin and spurfowl syringes, and the concordance of this feature with
phylogenetic relationships supported in previous studies on this group (Crowe and
Crowe 1985, Milstein and Wolff 1987, Crowe et al. 1992, Bloomer and Crowe 1998,
Crowe et al. 2006, Cohen et al. 2012, Chapter 2).
This is, to my knowledge, the first truly comparative paper of syringeal
morphology in francolins and spurfowls. Data on a number of lesser-known species in a
group that tends to be dominated by data from a few domesticated or managed northtemperate species is presented
84
Taxonomy, distribution, ecology and morphology
Based on traditional morphological research, the 41 currently recognized species of
francolins are placed within a single genus Francolinus Stephens, 1819 within the order
Galliformes and family Phasianidae (Hall 1963). They are distributed throughout subSaharan Africa (with an isolated population of one species Pternistis bicalcaratus
occurring in Morocco), the Middle East and Asia (Johnsgard 1988, Madge and
McGowan 2002), and are adapted to a variety of habitats, comprised primarily of
tropical and unforested vegetation types (McGowan 1994). All francolins have 14 tail
feathers and most species are sexually monomorphic in plumage, with males having
single- or double-spurred tarsi (Hall 1963, Johnsgard 1988).
However, relatively recent morphological, eco-ethological and molecular studies
of francolins (Crowe and Crowe 1985, Milstein and Wolff 1987, Crowe et al. 1992,
Bloomer and Crowe 1998, Crowe et al. 2006, Chapter 2) have suggested that they form
two distantly related lineages: the ‘true’ francolins and spurfowls (Fig. 3.1). ‘True’
francolins (allocated to five genera: Francolinus, Ortygornis (details for taxonomic
designation are in Chapter 5), Afrocolinus gen. nov. (details for taxonomic designation
are in Chapter 5), Peliperdix and Scleroptila) are relatively small, ground-roosting birds
with striped and barred rufous dorsal plumage resembling that of quails Coturnix spp.,
with only males possessing relatively small tarsal spurs. Spurfowls are placed within a
single genus Pternistis and are generally larger, often roost in trees, and have dark
dorsal plumage usually vermiculated with white or buff, and both sexes usually have
spurred (often two) tarsi that are much longer in males. Furthermore, spurfowls
generally emit atonal, raucous, grating advertisement calls (given at dawn and dusk),
whereas francolins have more tonal, often whistling, calls (Milstein and Wolff 1987).
85
This distinction between the two assemblages becomes blurred owing to O. sephaena,
the Creasted Francolin which, like other francolins, has quail-like plumage but, like
spurfowls has long tarsal spurs, roosts in trees and has an advertisement call with both
grating and tonal elements (Milstein and Wolff 1987). It was on the basis of this
‘linking form’ that Hall (1963) decided to place all 41 species into a single genus,
Francolinus, which DNA-based data no longer supports (Fig. 3.1). In this context, the
aim of this study was to examine the gross morphological and histological anatomical
structure of the syringes of selected francolins, spurfowls and their putative sister taxa,
as well as to determine whether there are any syringeal characters that could be
taxonomically and phylogenetically informative in investigating the proposed diphyletic
status of the ‘true’ francolins and spurfowls.
Materials and methods
Syringeal sampling (gross morphology and anatomy)
Syringes were examined from species that are representative of francolins and
spurfowls and those that are from their closest extant relatives (Table 3.1): Common
Quail Coturnix coturnix and Chukar Partridge Alectoris chukar (sister to the spurfowls),
G. gallus (sister to the francolins), and Helmeted Guineafowl Numida meleagris, a
distant relative to chickens, quails, partridges, francolins and spurfowls (Chapter 2).
It is appropriate at this point to highlight some of the differences between the
syringes of male and female birds. As explained by Frank et al. (2007), the syrinx of
male and female birds may have morphological differences that in turn could alter the
properties of the voices of each sex. However, Appel (1929) did not find differences
between the syringes of males and females in his research on the chicken Gallus gallus,
86
a galliform he studied intensively, even in ovariectomised females. Syringes from
immature birds were avoided in order to ensure that differences found between syringes
were not because of age differences between individuals. As Hogg (1982) described for
Galliformes, mineralisation begins well before maturity is reached and virtually
achieves its final extent during the growing period, so all adult birds are considered to
have a fully developed syrinx. Of course, it would be ideal to confirm our observations
by analysing more individuals of each of the species and in particular a developmental
sequence of one or more taxa, but this comparative material is at present not available.
Thus, the difficulty of acquiring multiple syringeal samples (represented by male and
female individuals of a particular species) and the inability to physically sex some of the
specimens were major constraints.
Analyses
All syringes examined for gross morphological and anatomical purposes were dissected
from frozen whole bird specimens with the exception of three syringes that were
dissected from alcohol-preserved whole specimens provided by the Ditsong National
Museum of Natural History, South Africa. The syringes were immersed in 70% ethanol
until such time that they were processed.
Gross syrynx morphology
The double-staining protocol in Cannell (1988) for the clearing and staining procedures
of the syringes was followed. Ossified tissues stained red with Alizarin red,
cartilaginous tissues stained blue with Alcian blue, and muscles stained brownish in
Lugol solution. Cleared and stained syringes were examined using a Leica S8APO
87
stereomicroscope and the LAS EZ version 1.7.0 software. Syringes of large birds such
as Gallus gallus, Numida meleagris, Pternistis capensis, P. swainsonii and P. natalensis
were stained for a few additional days relative to those excised from the smaller
francolins.
Histology
For the tissue-sectioning procedure, the syringes were transferred from 70% ethanol to
10% buffered formalin overnight and then taken through a series of 70%, 90% and
absolute ethanol washes before being cleared in xylene (1 h in each solution). They
were then embedded in paraffin wax and longitudinal sections of 5 μm thickness were
cut. Two stains were used, haematoxylin and eosin (Bancroft and Gamble 2002) for
more general biological examination, and a more specialised multiple stain, orceinpicroindigocarmine (Steven et al. 2000). The orcein-picroindigocarmine stain is
appropriate for staining bird syringes since it has the potential to differentiate tissues
and their components.
Tissue sections were stained with orcein, indigocarmine and picric acid following
the Steven et al. (2000) protocol. The sections were deparaffinised and placed in distilled
water, stained with orcein solution for 45 min at room temperature and rinsed in distilled
water. Sections were then differentiated in 96% ethyl alcohol (two times) for 30 s each,
rinsed in distilled water and stained with picroindigocarmine solution for 30 min. Slides
were drained and differentiated in 70% ethyl alcohol for 2 min and finally mounted with
Entellan and examined using a Nikon Stereoscopic Zoom microscope (NIS-Elements
version 2.10). The sectioning of tissues was more difficult for the larger birds examined
and this was a contributing factor in the distortion of those sections. Finally, even though
88
the staining of whole syringes and the sectioning of samples of larger syringes was
challenging, most of the stained syringes and sections were intact and clear with the
exceptions of Figure 3.5d and 3.6f.
Results
Syringeal gross morphology
The basic structure of the syringeal morphology (Fig. 3.2) of a typical
francolin/spurfowl is characterised by cartilaginous rings (which stained blue with
Alcian blue), a very distinctive mineralized tympanum (which stained red with Alizarin
red) and muscles (which stained brownish in Lugol solution). All the species analysed
have a typical ‘tracheobronchialis’ type of syrinx (Fig. 3.3-3.5), which means that both
the tracheal and the bronchial tissues are involved in the formation of the syrinx and
thus possibly in shaping the structure of their vocalizations.
Two significant gross morphological differences between true francolins and
spurfowls were observed. Firstly, unlike their putative sister taxon G. gallus, which has
a mineralized triangular-shaped tympanum (Fig 3.4a), francolins (Fig. 3.4b–f) and N.
meleagris (Fig. 3.3) have a shield-like calcified tympanum clearly visible from the
ventral perspective (Fig. 3.3). Numida meleagris, G. gallus and francolins show
extensive mineral deposition in the tympanum in both ventral and dorsal views but
Coqui Francolin Peliperdix coqui and Red-winged Francolin Scleroptila levaillantii
(Fig. 3.4c and d, respectively) show reduced mineral deposition on the dorsal side
compared with the ventral side.
In spurfowls, as in their putative sister taxa A. chukar and C. coturnix (Fig. 3.5a
and b, respectively), the tympanum is diamond-shaped with the degree of calcification
89
not as extensive as in francolins and relatives (Fig. 3.5c–e). Secondly, on the ventral
side, the first bronchial half rings are well mineralized in G. gallus (Fig. 3.4a) and
francolins (Fig. 3.4c, d and f), but not so in O. sephaena (Fig. 3.4b), spurfowls (Fig.
3.5c–e) and their relatives, A. chukar and C. coturnix (Fig. 3.5a and b). A further
observation is that some species such as S. levaillantoides, G. gallus and N. meleagris
show mineral deposition in some tracheal rings.
Histology
The histological structure of the syringes of francolins, spurfowls and their near
relatives are generally similar, with only a few small differences observed (Fig. 3.6).
The pessulus, which is present in all the species examined (Fig. 3.6a–h), varies
considerably in size and shape. It is generally small and rounded in the two true
francolins, i.e. S. levaillantii and Grey-winged Francolin S. afra, and N. meleagris (Fig.
3.6a, b and h, respectively) (Table 3.3). The pessulus is large and also rounded in O.
sephaena (Fig. 3.6c) and in spurfowls even though it is almost triangular in Natal
Spurfowl Pternistis natalensis (Fig. 3.6f) and C. coturnix (Fig. 3.6g). It is markedly
large and triangular in G. gallus (Fig. 3.6d, see also Myers 1917).
A distinctive interclavicular air sac, which is bound by the internal tympaniform
membrane, is found in the two francolins (S. levaillantii and S. afra), Cape Spurfowl
Pternistis capensis, C. coturnix and N. meleagris (Fig. 3.6a, b, e, g and h, respectively)
and is absent in O. sephaena and G. gallus (Fig. 3.6c and d, respectively), while its
presence could not be determined in P. natalensis (Fig. 3.6f). It was observed that the
inner wall of the interclavicular air sac and sometimes the medial bronchial wall are
lined with a layer of connective tissue that differs in thickness. Connective tissue fills
90
spaces and provides support to organs. The francolins S. levaillantii and S. afra (Fig.
3.6a and b) and their relative G. gallus (Fig 3.6d) have thin connective tissue lining the
walls of the membranes that tends to be moderate in P. natalensis, C. coturnix and N.
meleagris (Fig 3.6f, g and h, respectively). Numida meleagris has connective lining that
runs along the medial bronchial walls. Pternistis capensis (Fig 3.6e), as in O. sephaena
(Fig 3.6c), has a large amount of connective tissue that is restricted to the far corners of
the internal tympaniform membrane. This tissue pushes against the internal
tympaniform membrane forcing it to expand and hence results in the narrowing of the
surrounding airway. The external tympaniform membrane, which is found in all the
species examined, differs remarkably in length. This structure is extremely long in G.
gallus (Fig. 3.6d) and much shorter in the other species (Fig. 3.6a–c and e–h).
Discussion
Although a limited number of species were sampled, there are some marked structural
differences in the syrinx that distinguish francolins from spurfowls. The gross
morphology of the syrinx is generally consistent and supports the split of francolins and
spurfowls into two independent clades. The shape of the tympanum places O. sephaena
decisively with other francolins. However, O. sephaena, as in the two spurfowls P.
capensis and P. natalensis and their evolutionary relatives C. coturnix and A. chukar,
shows no mineral deposition in their first bronchial half rings with P. swainsonii
showing short first bronchial half rings with very little mineralization. Thus, the aspect
of mineral deposition is not fully consistent among francolins and spurfowls. Our
observation on the degree of mineralisation in tracheal rings of S. levaillantoides, G.
Gallus and N. meleagris could not at this stage translate to any coherent conclusion
91
apart from the fact that a similar observation was made by Hogg (1982), thought this
could have to do with conferring rigidity in the tracheal rings, which is an adaptation to
vocalization.
With regard to the histology of the syringes, the features that separate francolins
from spurfowls are based on the size of the pessulus. There is variation in the shape of
the pessulus, the presence/absence of the defined interclavicular air sac and the amount
of connective tissue such that differences are observed even between the two spurfowl
species. However, the amount of connective tissue puts species with tonal and whistling
calls together (S. levaillantii and S. afra), separated from those that have atonal, raucous
and grating calls (P. capensis, P. natalensis and O. sephaena). This feature could be
related to differences in sounds between them, given that whistle-like sounds could be
generated by shearing forces as a column of air is forced through a narrow aperture (see
Gaunt et al. 1982), and the connective tissue could exert pressure to modify the lumen
of the syrinx, which could be considered the column through which air passes,
generating the sound. At first glance, this explanation would be contrary to what we
would expect, given that spurfowls have thick connective tissue that would narrow their
syrinx, but they have raucous calls. However, we know the syrinx is only one
component in a vocal system (Gaunt et al. 1982) and other factors may be modulating
the final sound we hear.
Conclusions
A number of gross morphological and histological features were identified as
differentiating francolins from spurfowls (though with O. sephaena being a possible
exception). What emerged as an area for future research was investigation into the role
92
of each part of the syrinx in shaping francolin and spurfowl vocalizations, as well as the
need to include other galliform species in addition to the sampled francolin and
spurfowl taxa. This will require a reasonable number of individuals to be sampled per
species, such that any presence or absence of intra- or inter-specific variation can be
determined. Finally, it could be concluded that the outcome of this work points to the
distinction between francolin and spurfowl assemblages with O. sephaena (as indicated
in Hall 1963) still presenting some difficulties owing to its possession of characteristic
features that are typical of both francolins and spurfowls.
93
Tables and Figures
Table 3.1. List of species for which syrinxes were analyzed.
Gross morphology
Histology
Common name
Scientific name
Sample
name1
Sex2 Age
Scientific
name
Sample
Name1 Sex2 Age
Grey-winged Francolin
Orange River Francolin
Red-winged Francolin
Crested Francolin
Coqui Francolin
Cape Spurfowl
Swainson's Spurfowl
Natal Spurfowl
Common Quail
Chukar Partridge
Helmeted Guineafowl
Chicken/Red Jungle
Fowl
Scleroptila afra
S. levaillantoides
S. levaillantii
Ortygornis sephaena
Peliperdix coqui
Pternistis capensis
P. swainsonii
P. natalensis
Coturnix coturnix
Alectoris chukar
Numida meleagris
TMC67
TMC65
TMC60
TM78245
TM75627
TMC70
TMC48
TM60042
TMC50
TM74489
TMC45
M
F
F
M
F
M
M
?
F
?
M
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
?
Adult
S. afra
TMC01
F
Adult
S. levaillantii
O. sephaena
TMC15
TMC13
F
M
Adult
Adult
P. capensis
TMC07
M
Adult
P. natalensis
C. coturnix
TMC20
TMC14
M
F
Adult
Adult
N. meleagris
TMC17
F
Adult
Gallus gallus
TMC30
?
Adult
G. gallus
TMC02
?
Adult
1
TM, Transvaal Museum; TMC, Timothy M Crowe (Percy FitzPatrick Institute of African
Ornithology, (University of Cape Town).
2
? = Unknown.
Table 3.2. Comparison of gross morphological features of the syringes.
Taxon
Tympanum
shape
Bronchial half ring 1
mineralization
Tracheal ring
mineralization
S. afra
S. levaillantoides
S. levaillantii
O. sephaena
P. coqui
P. capensis
P. swainsonii
P. natalensis
C. coturnix
A. chukar
N. meleagris
G. gallus
Shield-like
Shield-like
Shield-like
Shield-like
Shield-like
Diamond
Diamond
Diamond
Diamond
Diamond
Shield-like
Triangular
Mineralized
Mineralized
Mineralized
Non-mineralized
Mineralized
Non-mineralized
Mineralized
Non-mineralized
Non-mineralized
Non-mineralized
Mineralized
Mineralized
Non-mineralized
Mineralized
Non-mineralized
Non-mineralized
Non-mineralized
Non-mineralized
Non-mineralized
Non-mineralized
Non-mineralized
Non-mineralized
Mineralized
Mineralized
94
Table 3.3. Comparison of the histological features of syringes.
Taxon
Pessulus shape
& size
Interclavicular
air sac
Amount of
connective
tissue
S. afra
S. levaillantii
O. sephaena
P. capensis
P. natalensis
C. coturnix
N. meleagris
G. gallus
Rounded, small
Rounded, small
Rounded, larger
Rounded, larger
Triangular, larger
Triangular, larger
Rounded, small
Triangular, larger
Bound by internal tympaniform membrane
Bound by internal tympaniform membrane
Absent
Bound by internal tympaniform membrane
?
Bound by internal tympaniform membrane
Bound by internal tympaniform membrane
Absent
Thin
Thin
Thick
Thick
Moderate
Thin
Thin
Thin
External
tympaniform
membrane
Shorter
Shorter
Shorter
Shorter
Shorter
Shorter
Shorter
Long
? = Presence or absence of feature cannot be determined from distorted tissue section.
95
Figure 3.1. Strict consensus cladogram for some 'true' francolins (Ortygornis,
Peliperdix and Scleroptila spp.) and spurfowls (Pternistis spp.) modified from Crowe et
al. (2006). Numbers at nodes are jackknife support values.
96
Figure 3.2. Example of a ventral view of the syrinx of a typical Coqui Francolin
Peliperdix coqui with illustrations of the various parts. Labels: tracheo-lateralis muscle
(tlm), tracheal cartilaginous ring (tcr), membrana-trachealis (mt), sterno-trachealis
muscle (stm), tympanum (t), external tympaniform membrane (etm), bronchial half ring
1 (bhr1), membrane-interanularis (mi), bronchus cartilaginous ring (bcr). Red colour mineralized cartilage, blue - cartilage, transparent light brown - muscle.
Figure 3.3. Ventral (V) and dorsal (D) views o f the syrinx of Helmeted Guineafowl
Numida meleagris. Red colour - mineralized cartilage, blue - cartilage, transparent light
brown - muscle.
97
Figure 3.4. Ventral (V) and dorsal (D) views of the syringes of: a. Gallus gallus, b.
Ortygornis sephaena, c. Peliperdix coqui, d. Scleroptila levaillantii, e. Orange River
Francolin S. levaillantoides and f. S. afra. Red colour - mineralized cartilage, blue cartilage, transparent light brown - muscle.
98
Figure 3.4 (concl.). Ventral (V) and dorsal (D) views of the syringes of: a. Gallus
domesticus, b. Ortygornis sephaena, c. Peliperdix coqui, d. Scleroptila levaillantii, e.
Orange River Francolin S. levaillantoides and f. S. afra. Red colour - mineralized
cartilage, blue - cartilage, transparent light brown - muscle.
99
Figure 3.5. Ventral (V) and dorsal (D) views of the syringes of: a. Alectoris chukar, b.
Coturnix coturnix, c. Pternistis natalensis, d. P. capensis, e. Swainson’s Spurfowl P.
swainsonii. Red colour - mineralized cartilage, blue - cartilage, transparent light brown muscle.
100
Figure 3.5 (concl.). Ventral (V) and dorsal (D) views of the syringes of: a. Alectoris
chukar, b. Coturnix coturnix, c. Pternistis natalensis, d. P. capensis, e. Swainson’s
Spurfowl P. swainsonii. Red colour - mineralized cartilage, blue - cartilage, transparent
light brown - muscle.
101
Figure 3.6. Histological structure of the syringes of francolins (a. S. levaillantii, b. S.
afra and c. O. sephaena), d. Chicken G. gallus, spurfowls (e. P. capensis and f. P.
natalensis), g. Common Quail C. coturnix and h. Helmeted Guineafowl N. meleagris.
References: tracheo-lateralis muscle (tlm), lumen of trachea (lt), interclavicular air sac
(ias), connective tissue (ct), pessulus (p), internal tympaniform membrane (itm),
external tympaniform membrane (etm), medial bronchial wall (mbw), bronchiodesmus
(bd), bronchial lumen (bl), sterno-trachealis muscle (stm).
102
Figure 3.6 (concl.). Histological structure of the syringes of francolins (a. S.
levaillantii, b. S. afra and c. O. sephaena), d. Chicken G. gallus, spurfowls (e. P.
capensis and f. P. natalensis), g. Common Quail C. coturnix and h. helmeted
guineafowl N. meleagris. Labels: tracheo-lateralis muscle (tlm), lumen of trachea (lt),
interclavicular air sac (ias), connective tissue (ct), pessulus (p), internal tympaniform
membrane (itm), external tympaniform membrane (etm), medial bronchial wall (mbw),
bronchiodesmus (bd), bronchial lumen (bl), sterno-trachealis muscle (stm).
103
CHAPTER 4
Taxonomic and phylogenetic utility of variation in
advertising calls of francolins and spurfowls
(Galliformes: Phasianidae)
Abstract
Systematists have not often made use of avian vocalizations to assess the taxonomic
rank of birds, or to infer their phylogenetic relationships. The likely reasons for this
stem from the perceived inability to distinguish genetic and ecological components of
variation in vocalizations, the difficulty in detecting homology across taxa, as well as
the diverse selection pressures acting on vocal characters which may make such
characters particularly prone to convergent evolution. In this study, we scored and
analysed DNA and vocal characters of two delineated assemblages of gamebirds,
francolins and spurfowls. Our phylogenetic results suggest that short strophes evolved
from longer strophes among taxa within the genera Scleroptila and Peliperdix. More
generally, our results corroborate the francolin-spurfowl dichotomy, with francolin calls
generally being long and tonal, containing a series of discrete elements that have
detectable harmonics. In contrast, most spurfowls render short, atonal calls with
elements that generally have no harmonics, although they may contain discrete
elements. Phylogenetically, Ortygornis sephaena is placed decisively with ‘true’
francolins and its closest relatives are the two phylogenetically enigmatic Asian
104
francolins, the Grey Francolin Ortygornis pondicerianus and Swamp Francolin O.
gularis.
105
Introduction
Vocalizations are a means of communication and are often of special
significance in the social behaviour of animals, and birds in particular. Many birds rely
on acoustic signals to convey information to each other over long distances, in dense
cover, and at night (Marler 1969, Gill 1990). Systematists have not often made use of
vocalizations to assess the taxonomic rank of birds, or to infer their phylogenetic
relationships. Possible reasons for this may stem from the perceived inability to
distinguish genetic and ecological components of vocalizations (McCracken and
Sheldon 1997), as well as the difficulty in detecting homology across taxa (Lanyon
1969). Further, McCracken and Sheldon (1997) maintain that the diverse selection
pressures acting on vocal characters may make such characters particularly prone to
convergent evolution. For example, species that live in dense vegetation tend to have
vocalizations with lower frequencies and narrower frequency ranges than those that
inhabit open habitats (Morton 1975). This is because longer wavelengths propagate
energy more efficiently through vegetation than shorter wavelengths, which attenuate
due to the scattering effects of the vegetation. In addition, several bird species from
diverse clades have repertoires of many distinct song types, whereas others have just
one simple and stereotyped song or call (Price and Lanyon 2002, Lei et al. 2005), and
some species learn the songs of other species, often making it difficult to distinguish
homologous components from learned songs (Lei et al. 2005). Other confounding
factors include the possibility that parts of the song may change in response to different
seasons (Lei et al. 2005, Aubrecht and Holzer 2000, Frank et al. 2007), as well as
morphological constraints of the syrinx (McCracken and Sheldon 1997). Finally, bird
song usually plays a strong role in mate recognition, and is thus under strong sexual
106
selection, and thereby adapted to the environment in which signalling is taking place
(Price and Lanyon 2002). Despite the above caveats, over the past two decades, authors
have started to place greater emphasis on exploring the systematic utility of vocal
character variation (Enggist-Düblin and Birkhead 1992, Alström 2001, Price and
Lanyon 2002, Päckert et al. 2003, Lei et al. 2005, Farnsworth and Lovette 2008).
Non-singing gamebirds such as francolins belonging to the genera Francolinus
Stephens, 1819, Ortygornis Reichenbach, 1852 (see Chapter 5 for details in taxonomic
designations), Afrocolinus gen. nov. (Mandiwana-Neudani, this thesis) (see Chapter 5
for details in taxonomic designations), Peliperdix Bonaparte, 1856 and Scleroptila
Blyth, 1849, as well as spurfowls Pternistis Wagler, 1832, are known to have a few
stereotyped call types making up their repertoires (van Niekerk 2010). Furthermore,
francolins and spurfowls appear to not be able to learn the calls of other related species
(Milstein and Wolff 1987). Here we explore the taxonomic and phylogenetic utility of
francolin and spurfowl ‘advertisement’ calls, that is those calls usually heard at dawn
and dusk, primarily during the breeding season.
Systematics
Genus Francolinus Stephens, 1819
Hall’s (1963) monograph of ‘francolins’ sensu lato placed 41 traditionally recognized
species into a single genus Francolinus, with 37 species partitioned into eight putative
monophyletic groups of related species and four unplaced species (Hall 1963, Table
1.1). Seven groups have representatives in Africa south of the Sahara, and one group is
confined to the Middle East and Asia. The current phylogenetic treatment divides
francolins into two distantly related groups of phasianine Galliforms: ‘francolins’
107
(shared among genera Francolinus, Ortygornis, Afrocolinus, Scleroptila, Peliperdix)
and ‘spurfowls’, all classified in one genus Pternistis (Milstein and Wolff 1987, Crowe
et al. 1992, Crowe et al. 2006, Mandiwana-Neudani, this thesis). Among attributes that
distinguish francolins from spurfowls, is the difference in the acoustic properties of their
advertising calls. Aurally, most spurfowl species have raucous grating or cackling calls
whereas francolins tend to have more musical and whistling calls (Crowe et al. 1986,
Milstein and Wolff 1987, Crowe et al. 1992). Furthermore, spurfowls can perch in trees
or other elevated structures and have scaly or striped underpart plumage, whereas
francolins roost on the ground and generally have barred, blotched or uniform
underparts and quail-like back plumage. However, the above distinction does not apply
well to the Crested Francolin O. sephaena because it has attributes of both francolins
and spurfowls. It is relatively small, but emits a somewhat raucous call and roosts in
trees like many spurfowls. Thus, it is the ‘linking form’ that persuaded Hall (1963) to
include all 41 species into a single genus Francolinus.
In this study, auditory, visual and quantitative descriptions of the advertising
calls of francolins and spurfowls are analysed in order to assess their phylogenetic
utility, investigate the validity of the francolin-spurfowl dichotomy and to assess
variation between putative related species.
Materials and methods
Collection of data
Call sampling
The calls were sourced from various institutional sound libraries and published
collections of recordings (Table 4.1), although many were unfortunately of poor quality.
108
Sonograms posted online at Xeno-canto [http://www.xeno-canto.org/] were also
examined, but were also mostly of poor quality. Frank et al. (2007) highlighted that the
syrinx of male and female birds may differ morphologically, which could in turn
determine the properties of their voices. However, given the difficulty in gathering
multiple vocalizations (for male and female individuals of a particular species), lack of
precise knowledge about the habitat from which the sourced call recordings were
collected, and often lack of knowledge as to whether the recorder actually saw the bird
in close proximity or not, we decided to analyse only the best quality calls and as a
consequence we were not always able to account for sex and/or the specific habitat the
bird was recorded in for all taxa.
Phylogenetic sampling
Molecular characters
Twenty terminal taxa of francolins (including two outgroups - Gallus gallus and
Bambusicola thoracica) and 24 spurfowl taxa (with two outgroups - Coturnix coturnix
and Alectoris chukar) were sequenced for the mitochondrial Cytochrome-b gene (CYTB
- 1143 base pairs- bp; Table 4.2). In contrast to the earlier work of Crowe et al. (1992)
and Bloomer and Crowe (1998), which focused on relatively few species, almost all
traditionally recognized species were included in our sampling. Total genomic DNA
was extracted from tissue using the animal tissue protocol provided with the DNeasy
tissue kit (Qiagen). Primers used in sequencing CYTB gene for fresh tissues and from
museum toe-pads are detailed in Table 4.3. Double stranded DNA templates were
amplified by polymerase chain reaction (PCR) using 0.75 units of BIOTAQTM DNA
polymerase (Bioline) in 30 µl reactions. Reactions also contained 1 x NH4 buffer, 2.5
109
mM MgCl2, each dNTP at 0.1 mM and each primer at 0.3 µM, as well as 3 µl of
template DNA. The thermal profile used comprised an initial denaturation step at 94°C
for two minutes, followed by 30 cycles of 94°C for one minute, 52°C for one minute
and 72°C for two minutes, with a final extension step of 72°C for seven minutes.
PCR-amplified products were cleaned from solution or gel using the GFXTM
PCR DNA and gel band purification kit (Amersham Biosciences) prior to cyclesequencing with the ABI PRISM Big DyeTM Terminator v3.1 cycle-sequencing Ready
Reaction Kit (Applied Biosystems). Sequencing products were resolved on an ABI
PRISM 3100 Genetic Analyser.
Vocal characters
Twenty terminal taxa of francolins (including two outgroups - Gallus gallus and
Bambusicola thoracica) and 24 spurfowl taxa (with two outgroups - Coturnix coturnix
and Alectoris chukar) were sampled and character matrices of vocal characters for the
francolins (Table 4.4.) and the spurfowls (Table 4.5) were generated. All calls (in wave
format) were analysed using a Linux-based software package ‘ANA’ (Jean-Pierre
Richard, Unpublished version). Multiple recordings of the calls of most species were
analysed to examine intraspecific variation and also to assess the quality of the calls for
each species. The terminology used hereafter is: strophe, element and pause (following
Alström et al. 2008) (Fig. 4.1). A ‘strophe’ is a call that may be repeated multiple times
in sequence and is comprised of elements that are defined as parts within a strophe
separated by a pause of duration of at least 0.01s. A ‘pause’ is an interval between the
elements of a strophe. Strophe duration was measured from the start of the first element
110
to the end of the last element on oscillograms and not on sonograms, in order to
maximise accuracy.
Sonograms with different acoustic properties were produced. The strophe
variables: strophe length; number of elements in a strophe; number of introductory
elements in a strophe; nature of the strophe in terms of whether it contains a harmonic
or trill; raucousness of advertisement; presence and absence of a warbling ending to the
strophe; whether the strophe is a simple musical advertisement, short complex musical
advertisement call, or a long complex musical call; whether the inter-element pause
exists and is distinct or indistinct; the presence or absence of a cackle-trill ending; the
presence or absence of ‘ka-waak/ko-rak' component; whether the pitch of the strophe
ascends or descends as it progresses; and whether the strophe is antiphonal or not; were
scored from the francolins and spurfowls vocal recordings of suitable quality (Table 4.4,
4.5).
Analyses of data
The generated nucleotide sequences were edited and assembled in the Staden Package
(Staden et al. 2003) and aligned using MAFFT (Katoh et al. 2009). All CYTB gene
sequences were checked for stop codons and insertions or deletions. Two phylogenetic
inference methods with different optimality criteria were employed to generate
phylogenetic hypotheses: Parsimony was used to construct phylogenetic trees for both
CYTB and the vocal data matrices, and a maximum likelihood (ML) approach was used
in analysing the CYTB data set. The francolin data matrices were rooted on Gallus
gallus and Bambusicola thoracica and the spurfowl matrices were rooted with Coturnix
coturnix and Alectoris chukar (see Crowe et al. 2006). Parsimony-based phylogenetic
111
analyses were conducted using PAUP* ver. 4.0b10 (Swofford 2002) under a full
heuristic search with all characters unordered and with equal weight, starting tree(s)
obtained via stepwise addition; tree-bisection-reconnection branch-swapping, and 1000
random addition replicates (Maddison 1991). When multiple, equally parsimonious
cladograms were recovered, a strict consensus cladogram was constructed. The extent to
which each non-terminal node is supported by different character partitions was
determined by using the bootstrap (BS) (Felsenstein 1985) resampling with 1000
pseudoreplicates, and five random addition replicates of taxa per bootstrap
pseudoreplicate.
Mixed-model maximum likelihood analyses were performed using the
Randomised Axelerated Maximum Likelihood algorithm for High Performance
Computing (RAxML) v7.0.4 (Stamatakis 2006, Stamatakis et al. 2008) implemented on
the CIPRES portal. Mixed-model RAxML analyses make use of a GTR++ model
partitioned by gene or codon position. The following analyses were run: mixed-model
mtDNA (one model for each codon position, and also as a single data partition).
Support
at
nodes
was
assessed
with
100
non-parametric
bootstrap
(BS)
pseudoreplicates. PAUP* was used to determine three measures of phylogenetic signal
by estimating: the consistency index, the retention index and the scaled retention index.
This enabled us to directly compare the phylogenetic signal inherent to the mtDNA and
vocal character matrices.
Results
Calls
112
Both structural and temporal characteristics distinguish the calls of francolins and
spurfowls, as well as among assemblages of taxa within them. Structural characteristics
involve: strophe duration, the number of elements per strophe, and use of harmonics,
trills and cackles. Temporal attributes include the sequence of elements and the relative
length of the pauses between them.
Francolins
Francolinus: The three members of the Spotted Francolin complex, F.
francolinus,
F. pictus and F. pintadeanus are characterised by having similarly structured strophes
which are 2.0 s, 2.11 s and 1.62 s long, respectively. These strophes are comprised of
distinctive elements: seven elements in F. francolinus, and five in both F. pintadeanus
and F. pictus (Appendix 4.1). The strophe of F. francolinus is relatively tonal with the
second element being higher-pitched. The first three elements of the strophe of F.
pintadeanus are tonal with the second and third elements ending in trills. The third
element is remarkably protracted whereas the fourth and fifth elements are fully trilled.
Francolinus pictus has a less tonal strophe with very fuzzy trilling elements.
These taxa also differ in that F. francolinus and F. pictus both have one
introductory element followed by a series of further elements, whereas F. pintadeanus
has two introductory elements. The pause between the introductory element/s and the
subsequent elements is relatively long in F. pictus (0.58 s), 0.43 s in F. francolinus and
0.24 s in F. pintadeanus. Another difference is seen in the shape of the harmonics,
which is rising in F. pintadeanus, but comprises a mainly overslurred fundamental
113
frequency in F. francolinus. The tonal part in the first element of F. pictus is also
overslurred.
Ortygornis: Ortygornis sephaena has a complicated strophe, which sounds like
a ‘hybrid’ between a francolin and spurfowl strophe to the ear (Appendix 4.1). The
advertisement strophe of O. sephaena is a high-pitched, squealing cooperative duet led
by a male and followed by a female (Little and Crowe 2000, van Niekerk 2010), and is
thought to be used in claiming a territory (van Niekerk 2010). Its strophe sounds similar
to that of the two Asian francolins, O. pondicerianus and O. gularis. The global
structure of the sonograms of these species looks similar despite the difference in the
duration of the strophes. Ortygornis pondicerianus and O. gularis are among the four
species that Hall (1963) did not place decisively in one of her eight species groups.
The duration of the strophes of O. sephaena is much shorter (0.55 s) compared
to that of the Kenyan O. grantii (0.74 s). The strophes of O. sephaena and O. grantii
have four elements, one introductory element followed by three complex elements.
There is interspecific variation that relates to the second element, which has two parts
that are fused in O. grantii but encompasses a distinctive pause in O. sephaena. The
first component of the second element has harmonics descending in shape in O.
sephaena and almost linear in O. grantii, and trilling parts. The third element has
complex variation, with the two parts either fused or separated. The fourth element is
quite faint and has an overslurred harmonic-like structure.
Thus, O. sephaena, O. grantii, O. gularis and O. pondicerianus have one
introductory element followed by a pause, and thereafter two to three additional
complex elements. The third element is almost broken into two components with
descending harmonics. Ortygornis gularis has strophe duration of 0.70 s, which,
114
approaches that of O. grantii (0.74 s), with O. pondicerianus having the shortest
duration (0.46 s) compared to O. sephaena and O. gularis. There is a much longer pause
(0.32 s) between the first and the last sets of elements in O. gularis when compared to
the other taxa: 0.14 s in O. grantii, 0.12 in O. sephaena, and 0.09 s in O. pondicerianus.
Peliperdix: All the Peliperdix spp. (also including A. lathami) have strophes
with an introductory element followed by a pause and then another five to seven
elements that trail away in volume, and have tonal and trilling parts. The global
structure of the sonograms remains similar despite differences in strophe duration, and
the number and structure of elements. The strophes of Peliperdix spp. (Appendix 4.1)
are relatively high pitched, harsh and tinny (Crowe et al. 1986) to the ear. Afrocolinus
lathami being an exclusively forest species, has a low frequency ‘cooing’ strophe which
suits the type of habitat it thrives in, but has the basic elemental structure similar to that
of the Peliperdix spp.
The major difference among the strophes of Peliperdix spp. is in their duration.
The strophe of Peliperdix albogularis is short (1.0 s) consisting of six tonal elements
with relatively stable harmonics with some parts of the elements trilled. The strophe of
P. schlegelii is a little longer (1.89 s) and consists of seven elements, with element E1
more tonal and has clearly defined rising harmonics, element E2 is fully trilled, and the
last four element (E 3-7) start with an overslurred component followed by trilling parts.
The strophe of P. coqui (1.95 s) has eight elements, which can be split into two parts.
The first part has stable harmonics followed by the second component which comprises
trilling and descending harmonics. The strophe of A. lathami (1.89 s) is made up of
eight very distinctive elements with E1 being more tonal followed by E 2-8 which begin
with tonal parts and end in trills. There is also a remarkable difference in the duration of
115
the pause between E1 and the rest of the elements in the strophes. The inter-element
pause (which narrows as the strophe progresses) is 0.10 s long in P. albogularis, 0.37 s
in P. schlegelii, 0.11 s in A. lathami, and somewhat longer than that in P. coqui (0.25 s).
There is a general decreasing trend in the duration of the inter-element pause in strophes
within the group. Thus, the strophes of the Peliperdix species represent a stepped cline
moving from the South African P. coqui to the Central African P. schlegelii, to the
West African P. albogularis.
Scleroptila: Unlike the Red-tailed Peliperdix francolins, which have harsh short
duration strophes, members of the Red-winged Group have whistling strophes. Four
species (S. levaillantoides, S. finschi, S. shelleyi and S. gutturalis) have strophes similar
in structure (four elements) that may be rendered “I’ll drink YER-BEER” (Newman
2002) and differ primarily in strophe and pause duration (Appendix 4.1). The strophe of
S. levaillantoides (0.53 s) is the shortest, with those of S. finschi (0.71 s), S. gutturalis
(0.79 s), and S. shelleyi, (0.89 s), respectively increasing in duration. The remaining
three species have strophes of much longer duration, (S. afra - 2.0 s, S. levaillantii 2.11 s, and S. streptophora - 2.22 s) and seven to ten elements (Appendix 4.1).
The typical advertisement strophe of S. levaillantoides, S. gutturalis, S. finschi
and S. shelleyi (also reported by Madge and McGowan 2002, Hockey et al. 2005
excluding S. gutturalis) are similar in that they have four complex elements with the
first two elements introducing the strophes. Strophes of Scleroptila finschi have
harmonics of complex shapes. The element following the two introductory elements is
very complex and does not have markedly definable third and fourth elements as in S.
levaillantoides and S. shelleyi. The first two elements of S. levaillantoides have rising
harmonics, which descend in the second component of E2, E3 is trilled, whereas E4
116
begins with trilling parts and ends with descending harmonics. The overall strophe
structure of S. shelleyi is similar with some minor variation in the structure of E3 and
E4.
Scleroptila streptophora, S. levaillantii and S. afra have longer strophes with
several elements. Scleroptila streptophora has a dove-like cooing strophe, which is very
tonal and has one introductory element followed by six additional elements. All the
elements have well-defined harmonics, which are stable in the first three elements and
descend in the fourth and fifth elements. The fourth through to the seventh elements
make a warbling sound. Scleroptila levaillantii has a strophe (with overslurred
harmonics) which is introduced by seven elements followed by three elements (E 8-10)
which make a warbling sound, whereas the strophe of S. afra (with overslurred
harmonics) is introduced by six elements followed by two elements (E 7 and 8), but
includes no warbles. There is a difference in the duration of the pause between the
introductory elements and the subsequent elements, which is quite long in S.
streptophora (0.46 s), and shorter in the other taxa: 0.27 s in S. shelleyi, 0.21 s in S.
afra, 0.18 s in S. finschi, 0.12 s in S. levaillantii, and 0.10 s in S. levaillantoides.
Spurfowls - Pternistis species
All four Bare-throated species have a basic atonal ‘ka-waak’ advertisement
strophe (0.75 s for P. swainsonii, 0.36 s – P. afer, 0.57 s – P. rufopictus and P.
leucoscepus) (Appendix 4.1) sounding higher-pitched and more guttural in P.
swainsonii and lower-pitched and more nasal to the ear in P. afer. Furthermore, the
strophes of P. leucoscepus and P. rufopictus are both high-pitched, with an element of
screeching and more protracted trilling second elements.
117
The strophes of P. swainsonii, P. afer, P. leucoscepus and P. rufopictus are
differentiated into two elements, which can be characterised as trills and are separated
by a fuzzy pause (except in P. swainsonii). The first element is more tonal in P.
swainsonii and trills in the other species, whereas E2 is completely trilled in P.
swainsonii, P. leucoscepus and P. rufopictus (exceptionally starting with a slower trill
and ending with a much faster trill). The inter-element pause is well-differentiated in P.
swainsonii (0.16 s long) in contrast to the fuzzy pause in P. afer (0.04 s), P. rufopictus
(0.08 s) and P. leucoscepus (0.09 s). There is a decreasing trend in the duration of
elements and the pause between them from the northeast African P. leucoscepus
through to the east African P. rufopictus to the southern African P. afer. Pternistis
swainsonii has a much longer strophe.
Like the Bare-throated spurfowls, the northern Vermiculated species P.
icterorhynchus, P. bicalcaratus, P. clappertoni and P. harwoodi, all give basic ‘kawaak’ strophes that differ mainly in the duration over which they are rendered (0.28 s,
0.33 s, 0.64 s and 0.61 s, respectively; Appendix 4.1). The strophes of these species
have two trilling elements, which are separated by a fuzzy pause that is 0.06 s, 0.04 s,
0.14 s and 0.08 s long in P. icterorhynchus, P. bicalcaratus, P. clappertoni, and P.
harwoodi, respectively. Among the four species, P. clappertoni is the only species in
which its strophe shows a much faster and loud trilling part at the beginning of the
second element. The southern Vermiculated species, P. adspersus, P. capensis, P.
hartlaubi, P. hildebrandti and P. natalensis give strophes which are extremely divergent
from one another, with the only similarity being that P. capensis and P. adspersus have
strophes ending in ‘cackle-trills’. The overall structure of the sonograms, the duration of
the strophes (2.10 s, 4.43 s, 1.90 s, 0.34 s and 0.62 s, respectively), number of elements
118
(some with components) and the duration of the inter-element pause differ remarkably.
Pternistis adspersus gives a strophe that is introduced by four elements followed by two
‘cackle-trill’ elements. Pternistis capensis renders the longest strophe within the group
which is more tonal consisting of seven elements (the first five elements with increasing
volume followed by a low-pitched ‘cackle-trill’). The first two elements have two
components, followed by two three-component elements, one four-component element
and ends in two five-component ‘cackle-trilled’ elements.
The most complicated strophe is that of Hartlaub’s Spurfowl P. hartlaubi, which
has a complex high-pitched antiphonal strophe often given by the male (Komen 1987;
Maclean 1993). This strophe has tonal and less well-defined trilling parts. Pternistis
hildebrandti in contrast, gives a trilling strophe with six elements that have no clearly
defined pause. The advertisement strophe of P. natalensis is a high-pitched, fourelement strophe characterized by some tonal, slow and fast trilling parts.
Among the members of the Scaly Group, the advertisement strophe of P.
squamatus is a high-pitched nasal ‘ke-rak’ which is repeated several times and increases
in volume as it progresses (Appendix 4.1). Pternistis ahantensis has a much shorter
strophe (0.44 s) compared to P. squamatus and P. griseostriatus (1.09 s and 0.81 s,
respectively).
The Montane species, P. camerunensis has a relatively tonal strophe with two
elements which are almost fused Appendix 4.1), with the first element being higherpitched relative to the second element. The same bird also gives a slow two-element
strophe with both elements descending in pitch and having a clearly defined pause in
between. Interestingly, the strophes from two recorded individuals of P. camerunensis
differed in their duration (0.37 s and 0.67 s, respectively), as well as the duration of the
119
pause between the two elements (0.03 s and 0.18 s, respectively). A second montane
species, P. nobilis, has an atonal, two-element strophe (0.60 s, which is reminiscent of
the ‘ka-waak’ strophes of members of the Bare-throated species group. The first
element has trilling and tonal parts followed by a pause (0.10 s) before a second
protracted trilling element. Pternistis ochropectus also gives a two-element tonal
strophe (0.48 s) with elements descending in pitch as it ends. The two elements were
separated by a distinctive pause (0.21 s). What is remarkable about P. ochropectus is
that based on the recordings analysed, the same individual sometimes switches the
sequence of the elements around starting with the lower-pitched second element
followed by a higher-pitched first element.
Three species, P. erckelii, P. swierstrai and P. castaneicollis, render strophes,
which are much longer (3.80 s, 4.24 s and 2.67 s long, respectively) with a series of
elements that build up in intensity and complexity as the strophes progress. The strophe
of P. erckelii is made up of 17 elements that increase in intensity from the start of E 1-4,
with E 5-14 beginning to subside and the last three elements (E 15-17) forming a
‘cackle-trill’ that further subsides in intensity as it ends. The duration of the interelement pause ranges from 0.05 s to 0.21 s. The strophe of P. swierstrai consists of
seven two-component elements that increase in intensity (from E 1-4) and decrease in
volume from E 5-7. The last three elements (E 5-7) form a ‘cackle-trill’, which starts to
break away in the fourth element becoming more apparent in the fifth element and so
forth. The duration of the inter-element pause increases from 0.14 s to 0.33 s.
The strophe of P. castaneicollis is characterized by six elements with the first
element having no components and followed by three two-component elements, one
three-component element, and ends in one four-component element. The last two
120
elements (E 5-6) again form a ‘cackle-trill’ as in P. erckelii and P. swierstrai. The
strophe of P. castaneicollis has variable intensity from the beginning to the end. The
first element starts at a low volume and then builds up in volume (E 2-4) and ends
fainter with E 5-6 comprising a 'cackle-trill'. There is an increasing trend in the duration
of the pause between elements (0.18 s - 0.21 s). It was reported in Madge and
McGowan (2002) that P. jacksoni gives very loud, high-pitched series of cackles that
sound similar to those of P. squamatus. The available recording of P. jacksoni was
unfortunately of poor quality such that no sonogram could be produced. However, to the
ear, it appears to give a loud series of cackles.
Phylogeny
Molecular and vocal characters - francolins
The parsimony analysis of the CYTB data set with 1143 characters, contained 248
parsimony informative characters (21.7%); a further 110 (9.6%) variable characters
were parsimony uninformative. Five trees of 965 steps were recovered. The vocal
character matrix, with nine characters, contained nine parsimony informative characters,
and recovered 28 equally parsimonious trees of 225 steps.
The CYTB parsimony and ML (not shown) trees (Fig. 4.2) recovered three of
Hall’s hypothesised monophyletic groups (Spotted - Francolinus, Red-tailed Peliperdix, Red-winged - Scleroptila). Further, the inclusion of Dendroperdix
streptophora into the Red-winged Group of francolins (Scleroptila spp.) in the CYTB
trees confirms the findings of Crowe and Crowe (1985) and Crowe et al. (1992). In the
parsimony derived vocal tree (Fig. 4.3), the Red-tailed species group together, but with
121
the inclusion of A. lathami; however, overall the resulting phylogeny is largely
unresolved.
Molecular and vocal characters - spurfowls
The parsimony analysis for the CYTB data set with 1143 characters, contained 231
parsimony informative characters (20.2%); a further 155 (13.6%) variable characters
were parsimony uninformative. Four trees of 926 steps were recovered. The vocal
character matrix, with eight characters, contained seven parsimony informative
characters, one parsimony uninformative character, and recovered 17 equally
parsimonious trees of 16 steps. The CYTB parsimony and ML (not shown) trees (Fig.
4.4), recovered some phylogenetic structure compared to the vocal tree (Fig. 4.5), which
largely forms a polytomy. The CYTB tree recovered at least one of Hall’s hypothesised
monophyletic groups, the Bare-throated Group (82% BS value – Fig. 4.4), though P.
hartlaubi is recovered within the group. Within the Bare-throated Group, P. rufopictus
and P. afer are recovered as sister taxa. The Vermiculated Group is split geographically
into a monophyletic clade comprising the southern Vermiculated taxa, whereas the
northern Vermiculated taxa are polyphyletic (P. bicalcaratus, P. clappertoni, P.
icterorhynchus, P. harwoodi). Montane species are primarily distributed towards the
base of the Pternistis tree, with the exception of P. jacksoni which is sister to the
montane Angolan endemic P. swierstrai, although with poor support. The East African
species (P. erckelii, P. ochropectus, P. castaneicollis) form a clade, with the montane
West and East African taxa P. camerunensis and P. nobilis forming a sister group basal
to the three East African montane species. The Namibian arid savanna endemic P.
122
hartlaubi is recovered as the basal taxon to all spurfowl species in the CYTB parsimony
tree.
Phylogenetic signal
In comparing three common parsimony metrics for the francolin CYTB and vocal data
sets, the vocal matrix has less homoplasy but considerably fewer characters: CYTB
consistency index (CI) of 0.516 compared to 0.818 in the vocal matrix, retention index
(RI) of 0.458 compared to 0.897 in vocal matrix, and a rescaled retention index (RC) of
0.236 compared to 0.734 for the vocal matrix. Similarly for the spurfowl CYTB data
matrix, the CI is 0.521 compared to 0.625 for the vocal matrix, with a RI of 0.457
versus 0.870, and a RC of 0.238 versus 0.543. In summary, there is less homoplasy in
the vocal matrices for both francolins and spurfowls than in the CYTB data sets, but
because of considerably fewer characters there is less phylogenetic resolution.
Discussion
The quantitative and visual structural descriptions of the advertisement calls support the
phylogenetic distinction between francolins and spurfowls. Francolins are found to
generally give long, tonal strophes (either whistling, or harsh and tinny calls) that
comprise a series of distinctive elements that usually have harmonics. Spurfowls are
characterised by having generally short (with a few exceptions), atonal and grating
strophes, with very few elements and no harmonics.
Within the Red-winged Group (Scleroptila spp.), both the CYTB and the vocal
trees recover the separation of the core-Scleroptila clade from S. streptophora and S.
levaillantii. This is generally in accordance with the sonographic presentations (except
123
S. afra), which break Scleroptila into two quite distinct subgroups: (1) species with
short, four-element strophes (S. finschi, S. levaillantoides S. gutturalis and S. shelleyi)
and those with long strophes consisting of many more elements (S. streptophora, S.
levaillantii and S. afra), often ending with a warble. It could be that the similarities in
the advertisement calls among Scleroptila species may be a consequence of sharing
common descent, as in contrast to the parsimony CYTB phylogeny, the ML-based
topology (not shown) places S. afra at the base of a clade comprising S. finschi, S.
levaillantoides, S. gutturalis and S. shelleyi. Even though the Red-winged Scleroptila
species occur in varied habitats at different latitudes, their primary preference is
grassland (Johnsgard 1988, Madge and McGowan 2002), which collectively comprises
the open lowland grasslands in which S. levaillantoides, S. gutturalis, S. finschi thrive,
and the open hilly (S. shelleyi) and highland grasslands preferred by S. afra, S.
psilolaema, S. levaillantii, and S. streptophora.
Within the Red-tailed group, Peliperdix species, the CYTB topology places P.
albogularis and P. schlegelii together, both taxa with fast, short strophes. Basal to these
two taxa is P. coqui that has a slower and longer strophe. Although the topology based
on vocal characters is poorly resolved, it interestingly recovers the enigmatic
Afrocolinus lathami to be nested inside Peliperdix, rather than as a highly distinct
lineage, sister to Francolinus revealed in the CYTB topology (although with limited
support). The grouping together of A. lathami and Peliperdix taxa in the vocal tree
could be attributed to it sharing a series of elements (almost eight) forming its strophe
that fade as the strophe progresses, as in other Peliperdix species, particularly P. coqui.
The three Red-tailed Peliperdix species inhabit open grassland and woodland savannas.
Afrocolinus lathami remains the only African francolin to be restricted to dense forested
124
habitats (del Hoyo et al. 1994, Madge and McGowan 2002), hence it has a low
frequency ‘cooing’ strophe (versus the high-pitched, broadband strophe in Peliperdix
spp.), which suits the forested habitat it thrives in. Frequency was not a character we
could reliably score, and hence it was not included in the phylogenetic analyses. If it
were included, it may well place A. lathami outside the Peliperdix clade.
The CYTB topology suggests that shorter strophes evolved from longer strophes
among both Scleroptila and Peliperdix species. The placement of Ortygornis
(Dendroperdix) sephaena has long been a puzzle because of its aberrant life-history and
vocal characteristics, that is, roosting in trees and emitting a somewhat raucous call
more similar to spurfowls than francolins. The CYTB and vocal topologies suggest O.
sephaena to be sister to the African O. grantii, with the DNA data suggesting this clade
to be sister to the Asian O. gularis and O. pondicerianus. Ortygornis sephaena may thus
represent a ‘linking form’ between Asian and African francolins. The global structure of
the sonograms of O. sephaena, O. grantii, O. gularis and O. pondicerianus look similar
despite the difference in the duration of the strophes. These species all have one
introductory element followed by a pause, and thereafter two to three additional
complex elements. The third element is almost broken into two components in O.
gularis and O. pondicerianus. Aurally, the strophe of O. sephaena and O. grantii sound
similar to that of the two Asian francolins, O. pondicerianus and O. gularis.
Despite the Spotted Francolinus species emerging as monophyletic, they have
remarkably variable strophes but share the feature of having similar introductory
elements. Differences between species of spurfowls are mainly in strophe duration, the
number of elements per strophe, the duration of the pause between the introductory
element and the subsequent elements. The Bare-throated species (P. swainsonii, P. afer,
125
P. rufopictus and P. leucoscepus) and perhaps the northern Vermiculated taxa (P.
bicalcaratus, P. icterorhynchus, P. clappertoni and P. harwoodi) have similar
sonographic features despite being well-separated in the CYTB topology. The Barethroated species also inhabit similar habitats to members of the Vermiculated Group (P.
hartlaubi, P. capensis, P. natalensis, P. adspersus, P. hildebrandti, P. clappertoni, P.
icterorhynchus, P. bicalcaratus, P. harwoodi), which generally occupy dense bushy
thickets and shrubby grasslands (del Hoyo et al. 1994, Madge and McGowan 2002).
The vocal topology recovered for spurfowls is almost completely unresolved and hence
is of limited phylogenetic utility.
There is strong divergence among the strophes of members of the Montane
species complex with just the two northeast African species (P. erckelii and P.
castaneicollis) and perhaps P. swierstrai (Angolan endemic) being brought together by
having multiple elements (with components) and a ‘cackle-trill’ that ends their strophes.
The relationship between the strophes of P. camerunensis, P. nobilis and P. ochropectus
is uncertain except that they all have two elements that descend in pitch. Pternistis
nobilis, however, gives a two-element strophe that is reminiscent of those of the Barethroated species, suggesting that the ‘ka-waak’ strophe might be the plesiomorphic state
within spurfowls. The vocal relationship within the Scaly Group (P. squamatus, P.
griseostriatus, P. ahantensis) is uncertain and their strophes are highly divergent.
Conclusions
Comparing the outcome of DNA and vocal characters in a phylogenetic context was not
an easy task due to the vocal phylogenetic topology being largely unresolved relative to
that recovered from the DNA characters, particularly among spurfowls. However, on
126
the basis of the level of phylogenetic signal within each data set (DNA and vocal
characters), vocal characters exhibit less homoplasy. Hence, if a sufficient number of
characters can be scored relative to the number of taxa being studied, such characters
should be of considerable use in reconstructing phylogenies. Spurfowls, unlike
francolins are very divergent vocally, morphologically, as well as with respect to habitat
preference. There is however, some phylogenetic concordance between DNA and vocal
characters of francolins. Perhaps the most useful phylogenetic outcome of this research
are the marked differences between the calls of francolins and spurfowls and the
decisive placement of the ‘linking form’ Ortygornis sephaena with francolins, in
particular with two of Hall’s (1963) unplaced taxa, O. pondicerianus and O. gularis.
127
Tables and Figures
Table 4.1. List of species for which calls were analysed. BLSA abbreviates
British Library Sound Archive, MLNS - Macaulay Library of Natural Sounds.
_____________________________________________________________________________
Taxon name
ID. No.
Supplier
Recorder
Locality
AFRICAN SPURFOWLS
Bare-throated Group
Pternistis afer
P. swainsonii
P. leucoscepus
P. rufopictus
CD2.5 - 198
CD2.5 - 199
100205
03052 R1C1
Michael Mills
M. Mills
MLNS
BLSA
G. Gibbon
G. Gibbon
R. Linda
Cornell
Natal
Namibia
Ethiopia
Tanzania
Montane Group
P. camerunensis
P. ochropectus
P. nobilis
P. erckelii
P. swierstrai
P. castaneicollis
P. jacksoni
CD5 - 95
cc26744
CD5 - 96
68756 BD10
Unknown
100242
Unknown
Michael Mills
BLSA
Michael Mills
BLSA
Michael Mills
MLNS
C. Cohen
C. Chappuis
G. Welch
C. Chappuis
S. Smith
M. Mills
R. Linda
B. Finch
Mt. Cameroon
Djibouti
Uganda
Unknown
Angola
Ethiopia
Kenya
Vermiculated Group
P. bicalcaratus
P. icterorhynchus
P. clappertoni
P. harwoodi
P. hildebrandti
P. capensis
P. natalensis
P. adspersus
P. hartlaubi
216-231
cc1238 BD20
cc1239 BD21
24718 BD1
CD5 – 92
CD2 - 195
CD2 - 196
61790 BD24
CD2 - 197
Michael Mills
BLSA
BLSA
BLSA
Michael Mills
Michael Mills
Michael Mills
BLSA
Michael Mills
C. Chappuis
C. Bourguignon
C. Chappuis
H. Shirihai
C. Chappuis
G. Gibbon
G. Gibbon
D. Watts
G. Gibbon
Ivory Coast
Zaire
Nigeria
Ethiopia
Malawi
South Africa
South Africa
Namibia
Namibia
Scaly Group
P. ahantensis
P. griseostriatus
P. squamatus
cc24693BD32
103772BD49
CD5 - 89
BLSA
BLSA
Michael Mills
C. Chappuis
I. Sinclair
C. Chappuis
Senegal
Angola
Gabon
MLNS
MLNS
BLSA
B. King
P. Holt
C. Carter
Sri Lanka
India
Burma
Michael Mills
Michael Mills
Michael Mills
C. Chappuis
G. Gibbon
G. Gibbon
Gabon
South Africa
South Africa
ASIATIC FRANCOLINS
Spotted Group
Francolinus pictus
F. francolinus
F. pintadeanus
551
86293
113300 BD42
AFRICAN FRANCOLINS
Red-winged Group
Scleroptila finschi
S. shelleyi
S. afra
CD5 - 86
CD2.5 - 191
CD2.4 - 190
128
___________________________________________________________________________
Taxon name
ID. No.
Supplier
Recorder
Locality
S. levaillantoides
S. gutturalis
S. levaillantii
CD2.5 - 193
48414 R1C1
CD2.5 - 192
Michael Mills
BLSA
Michael Mills
G. Gibbon
D. Pearson
G. Gibbon
Namibia
Somalia
South Africa
Red-tailed Group
Peliperdix coqui
P. schlegelii
P. albogularis
CD2 - 188
cc1214 BD22
cc1212 BD20
M. Mills
BLSA
BLSA
G. Gibbon
J. Brunel
C. Chappuis
South Africa
Chad
Senegal
Striated Group
Ortygornis sephaena
O. grantii
S. streptophora
1095
CD5 - 88
CD5 - 84
M. Hausberger
Michael Mills
Michael Mills
M. Hausberger
C. Chappuis
C. Chappuis
South Africa
Kenya
Cameroon
C. Chappuis
P. Holt
P. Holt
Ivory Coast
Nepal
Pakistan
Unplaced taxa (excluding Ptilopachus nahani)
Afrocolinus lathami
CD5 - 80
Michael Mills
Ortygornis gularis
65642 BD20
BLSA
O. pondicerianus
41433 R1C10
BLSA
129
4.2. Francolin and spurfowl taxa for which DNA sequences were generated.
Acronyms. AMNH abbreviates American Museum of Natural History, TM Transvaal Museum, BM - British Museum - Natural History Museum at Tring,
FMNH – Field Museum of Natural History, PFIAO - Percy FitzPatrick Institute of
African Ornithology, TMC - Timothy M. Crowe, University of Cape Town, South
Africa, ‘-’ – Unknown. Genera are as recorded on specimen label.
Taxa name
Sample no.
Origin
Date collected
GenBank no.
AMNH DOT8023 India
AMNH 776813
India
GenBank
China
-
AF013762
FR694142
NC011817
TMC 9
Marico River, South Africa
BM 1902 1 20 300 Hulul, Ethiopia
TMC 11
Cameroon
2004
1902
2005
FR694140
FR694144
PFIAO 47
PFIAO 59
TMC 12
AMNH 541174
TM 78622
AMNH 308887
Ayton Farm, South Africa
Eastern Cape, South Africa
Petrus Steyn, South Africa
Ethiopia
Sterkspruit, South Africa
Angola
2002
2002
2005
1941
AM236898
AM236897
FR691612
FR691613
U90642
FR691607
PFIAO 45
BM 1949 30 19
BM 1929 3 13 1
Settlers, South Africa
Mboro, Bahr-El-Ghazel, Chad1949
Farafeni, Gambia
1929
AM236895
FR694149
FR694145
Francolins
Spotted Group
F. francolinus
F. pictus
F. pintadeanus
Striated Group
Francolinus sephaena
F. sephaena grantii
F. streptophora
FR691617
Red-winged Group
F. shelleyi
F. afra
F. levaillantoides levaillantoides
F. l. gutturalis
F. levaillantii levaillantii
F. finschi
Red-tailed Group
F. coqui
F. schlegelii
F. albogularis
Ungrouped species
F. lathami
F. pondicerianus
F. gularis
GenBank
Cameroon
AMNH DOT8050 India
GenBank
India
-
AM236893
FR691632
U90649
PFIAO 108,
TMC 40
AMNH 202503
PFIAO 109
2004
2004
2004
AM236908
AM236907
FR691588
AM236906
Spurfowls
Bare-throated Group
Francolinus afer
F. swainsonii
F. rufopictus
F. leucoscepus
Tudor East, Watervalboven
Marico River
Gagayo, Muranza
Kenya
130
Taxa name
Sample no.
Origin
Date collected
GenBank no.
AMNH 541471
FMNH 1971-1072
GenBank
AMNH261929
AMNH1759
TMC 42
AMNH 419126
Badaltino, Shoa
Djbouti
East slope, Mt. Kenya
West Ruwenzori
Mount Cameroon
Angola
-
FR691589
FR691590
AM236903
FR691594
FR691592
FR691591
FR691593
PFIAO 117
AMNH 541411
Ndalla Tanda
-
AM236905
F. bicalcaratus
F. clappertoni
F. icterorhynchus
F. hildebrandti
F. natalensis
F. hartlaubi
F. capensis
F. adspersus
F. harwoodi
TM 14682
TMC 68
AMNH 156922
GenBank
TMC 120
TMC 121
PFIAO 229
PFIAO 206A
BM 1927.11.5.18
Gold Coast, Hinterland
Cameroon
Fanadji
Marico River, South Africa
Namibia
Kakamas, South Africa
-
1901
2005
2004
2006
1927
FR691624
FR691602
FR691601
FR691595
AM236911
FR691618
AM236909
FR691623
FR691600
Outgroups
Gallus gallus
Bambusicola thoracica
Alectoris chukar
Coturnix coturnix
-
-
-
L083761
EU165706
L083781
L083771
Montane Group
F. erckelii
F. ochropectus
F. castaneicollis
F. jacksoni
F. nobilis
F. camerunensis
F. swierstrai
Scaly Group
F. squamatus
F. griseostriatus
AM236904
Vermiculated Group
131
Table 4.3. DNA markers sequenced and primers used for PCR amplifications
and sequencing of preserved tissues.
Primer name
Primer sequence (5’to 3’)
Reference
Fresh tissues
Francolins & Spurfowls (General primers)
Cytochrome b
L14578
MH15364
ML15347
H15915
cta gga atc atc cta gcc cta ga
act cta cta ggg ttt ggc c
atc aca aac cta ttc tc
aac gca gtc atc tcc ggt tta caa gac
J.G. Groth (pers. comm.)
P. Beresford (pers. comm.)
P. Beresford (pers. comm.)
Edwards & Wilson (1990)
Toe-pads
Cytochrome b
Spurfowl-specific primers
L14851 (General)
cct act tag gat cat tcg ccc t
Pt-H195
ttt cgr cat gtg tgg gta cgg ag
Pt-H194
cat gtr tgg gct acg gag g
MH15145
aag aat gag gcg cca ttt gc
Kornegay et al. (1993)
R. Moyle & T. Mandiwana-Neudani
R. Bowie
P. Beresford
Pt-L143
Pt-H361
gcc tca tta ccc aaa tcc tca c
gtg gct att agt gtg agg ag
R. Moyle & T. Mandiwana-Neudani
R. Moyle & T. Mandiwana-Neudani
Pt-L330
Pt-H645
tat act atg gct cct acc tgt ac
ggg tgg aat ggg att ttg tca gag
R. Bowie
R. Moyle & T. Mandiwana-Neudani
Pt-L633
Pt-H901
ggc tca aac aac cca cta ggc
agg aag ggg att agg agt agg at
R. Moyle & T. Mandiwana-Neudani
R. Moyle & T. Mandiwana-Neudani
L2-2312
H15696
cat tcc acg aat cag gct c
R. Bowie
aat agg aag tat cat tcg ggt ttg atg Edwards et al. (1991)
Pt-L851alt
Pt-H1050
cct att tgc cta cgc cat cct ac
gat gct gtt tgg ccg atg
R. Bowie
R. Bowie
Pt-L961
Pt-L961alt
HB20 (General)
cga acc ata aca ttc cca c
ctc atc cta ctc cta atc ccc
ttg gtt cac aag acc aat gtt
R. Moyle & T. Mandiwana-Neudani
R. Bowie
J. Feinstein (pers. comm.)
132
Primer name
Primer sequence (5’to 3’)
Reference
Cytochrome b
Francolin-specific primers
L14851 (General)
cct act tag gat cat tcg ccc t
Franc-H1
cag cag aca cyt cyc tyg cct tc
MH15145
aag aat gag gcg cca ttt gc
Kornegay et al. (1993)
R. Bowie
P. Beresford
Franc-L1
Franc-H2
Franc-L2
Franc-H3
tgc ctc aca acc caa atc ctc ac
agg agr agr att act cct gtg ttt cag g
gcc tca ttc tty ttc aty tgy atc ttc c
ggr tgg aat ggg att ttg tca gag
R. Bowie
R. Bowie
R. Bowie
R. Bowie
Franc-L3
Franc-H4
tcatcyractcygacaaaatccc
gar rgg gat tag rag gag gat
R. Bowie
R. Bowie
Franc-L4
Franc-H5
tat tcg cct ayg cya tcc twc gct c
gta gga rag kga tgc tat ttg gcc
R. Bowie
R. Bowie
Franc-L5
HB20 (General)
ctc atc ctc ctc cta atc cc
ttg gtt cac aag acc aat gtt
R. Bowie
J. Feinstein (pers. comm.)
133
Table 4.4. Vocal character states scored and used for the phylogenetic analysis of
francolins.
1. Strophe length: <1.0 s = 1; =1.0 s <2 s = 2; ≥2.0 s = 3
2. No. of elements in strophe: <5 = 1; ≥5 = 2
3. Harmonics: Absent = 0; Indistinctive = 1; Predominately distinct = 2;
Distinct mixed with distinctive trills = 3
4. Raucous advertisement: Absent = 0; Type 1 = 1; Type 2 = 2
5. No. of introductory elements: one = 1; two = 2; More than two = 3
6. Strophe warbling ending: Absent = 0; Present = 1
7. Simple musical advertisement: Absent = 0; Slow = 1; Fast = 2; Very
fast = 3
8. Short complex musical advertisement call: Absent = 0; Slow I'll
drink-yer-beer = 1; Fast I'll drink-yer-beer = 2
9. Long complex musical call: Absent = 0; Type 1 = 1; Type 2 = 2
A matrix of vocal state scores used for phylogenetic analyses of the
francolins.
Characters
Taxon
1
2
3
4
5
6
7
8
9
Gallus gallus
1
1
2
1
1
0
0
0
0
Bambusicola thoracica
Francolinus francolinus
1
1
2
1
1
0
0
0
0
3
2
2
1
1
0
0
0
0
Francolinus pintadeanus
2
2
3
1
2
0
0
0
0
Francolinus pictus
3
2
1
1
1
0
0
0
0
Afrocolinus lathami
2
2
0
0
1
0
1
0
1
Ortygornis pondicerianus
1
1
2
1
1
0
0
0
0
Ortygornis gularis
1
1
2
1
1
0
0
0
0
Ortygornis grantii
1
1
3
2
1
0
0
0
0
Ortygornis sephaena
1
1
3
2
1
0
0
0
0
Peliperdix albogularis
1
2
2
0
1
0
3
0
1
Peliperdix coqui
2
2
2
0
1
0
1
0
1
Peliperdix schlegelii
2
2
2
0
1
0
2
0
1
Scleroptila afra
3
2
2
0
3
1
0
0
2
Scleroptila finschi
1
1
2
0
2
0
0
1
0
Scleroptila levaillantii
3
2
2
0
3
1
0
0
2
Scleroptila levaillantoides
1
1
2
0
2
0
0
2
0
Scleroptila gutturalis
1
1
2
0
2
0
0
1
0
Scleroptila shelleyi
1
1
2
0
2
0
0
1
0
Scleroptila streptophora
3
2
2
0
1
0
0
0
2
134
Table 4.5. Vocal characters and states scored and used for phylogenetic analysis of
spurfowls.
1. Strophe length: <1 s = 1; ˃1 s =2
2. No. of elements in strophe: 2-3 = 1; ≥4 = 2
3. Strophe type: Less to no trill = 1; Predominately trilled = 2; Predominately
harmonic = 3
4. Inter-element pause: Absent/indistinct = 1; Distinctive = 2
5. Cackle-trill ending: Absent = 0; Present = 1
6. Ka-waak/Ko-rak component: Absent = 0; Present = 1
7. Strophe pitch: Stable = 1; Descends as strophe ends = 2; Ascends as strophe ends =
3
8. Strophe antiphonal?: No = 0; Yes = 1
A matrix of vocal character used to generate the vocal character phylogeny.
Characters
Taxon
1
2
3
4
5
6
7
8
Coturnix coturnix
1
1
1
1
0
0
?
0
Alectoris chukar
1
1
1
1
0
0
?
0
Pternistis hartlaubi
2
2
1
2
0
0
1
1
Pternistis camerunensis
1
1
3
2
0
0
1
0
Pternistis nobilis
1
1
2
2
0
1
1
0
Pternistis erckelii
2
2
2
2
1
0
2
0
Pternistis swierstrai
2
2
2
2
1
0
2
0
Pternistis castaneicollis
2
2
2
2
1
0
2
0
Pternistis jacksoni
?
?
?
?
1
0
?
?
Pternistis ochropectus
1
1
3
2
0
0
1
0
Pternistis squamatus
Pternistis ahantensis
2
1
1
1
3
1
2
2
0
0
0
0
3
1
0
0
Pternistis griseostriatus
1
?
1
1
0
0
1
0
Pternistis bicalcaratus
1
1
2
2
0
1
1
0
Pternistis icterorhynchus
1
1
2
1
0
1
1
0
Pternistis clappertoni
1
1
2
1
0
1
1
0
Pternistis harwoodi
1
1
2
1
0
1
1
0
Pternistis hildebrandti
1
1
2
1
0
0
1
0
Pternistis natalensis
1
2
2
2
0
0
1
0
Pternistis adspersus
2
2
2
2
1
0
2
0
Pternistis capensis
2
2
3
2
1
0
3
0
Pternistis leucoscepus
1
1
2
1
0
1
1
0
Pternistis rufopictus
1
1
2
1
0
1
1
0
Pternistis afer
1
1
2
1
0
1
1
0
Pternistis swainsonii
1
1
2
2
0
1
1
0
135
Figure 4.1. Example of a sonogram of a typical francolin, Coqui Francolin Peliperdix
coqui, illustrating and defining the variables studied: E - element, E1 - element number
1, P - pause, H - harmonic, H1 - harmonic number 1 and FF - fundamental frequency.
136
Gallus gallus
Bambusicola thoracica
Scleroptila streptophora
Scleroptila shelleyi
Scleroptila levaillantoides
98
82
77
R
W
G
Scleroptila afra
Scleroptila guturalis
99
100
Scleroptila finschi
Scleroptila levaillantii
83
92
Peliperdix schlegelii
95
Peliperdix albogularis
86
99
R
T
G
Peliperdix coqui
Ortygornis sephaena
100
100
80
98
Ortygornis grantii
Ortygornis spp.
Ortygornis pondicerianus
100
100
Ortygornis gularis
Francolinus pintadeanus
Francolinus pictus
82
S
P
G
Francolinus francolinus
Afrocolinus lathami
Figure 4.2. A parsimony tree (1 of 5 most parsimonious trees) of francolins obtained from
mitochondrial Cytochrome-b characters. Numbers above branches represent parsimony boostrap
support values and those below branches are maximum likelihood bootstrap support values (only
≥ 70% are presented). RWG stands for Red-winged Group, RTG – Red-tailed Group, SPG –
Spotted Group.
137
Gallus gallus
Bambusicola thoracica
Francolinus francolinus
S
P
G
Francolinus pintadeanus
Francolinus pictus
Afrocolinus lathami
Peliperdix coqui
Peliperdix schlegelii
R
T
G
Peliperdix albogularis
Ortygornis pondicerianus
Ortygornis gularis
Ortygornis Spp.
Ortygornis grantii
81
Ortygornis sephaena
Scleroptila afra
90
Scleroptila levaillantii
Scleroptila streptophora
Scleroptila finschi
R
W
G
Scleroptila levaillantoides
78
Scleroptila guturalis
Scleroptila shelleyi
Figure 4.3. The strict consensus parsimony tree of francolins obtained from vocal characters.
Numbers above branches represent boostrap support values (only ≥ 70% are presented). RWG
stands for Red-winged Group, RTG – Red-tailed Group, SPG – Spotted Group.
138
Alectoris chukar
Coturnix coturnix
Pternistis adspersus
Pternistis capensis
87
95
Pternistis hildebrandti
99
98
S
V
t
Pternistis natalensis
Pternistis afer
81
81
B
T
G
Pternistis rufopictus
Pternistis leucoscepus
82
Pternistis swainsonii
100
97
Pternistis clappertoni
Pternistis harwoodi
Pternistis jacksoni
Pternistis swierstrai
Pternistis squamatus
Pternistis griseostriatus
M
T
t
S
C
t
Pternistis bicalcaratus
100
Pternistis icterorhynchus
Pternistis castaneicollis
89
92
88
Pternistis erckelii
93
90
89
90
100
Pternistis ochropectus
M
T
t
Pternistis camerunensis
Pternistis nobilis
Pternistis hartlaubi
Figure 4.4. A parsimony tree (1 of 4 most parsimonious trees) of spurfowls obtained from
mitochondrial Cytochrome-b characters. Numbers above branches represent parsimony boostrap
support values and numbers below branches represent maximum likelihood boostrap support
values (only ≥ 70% are presented). MTt stands for Montane taxa, SCt – Scaly taxa, SVt – Southern
Vermiculated taxa and BTG – Bare-throated Group.
139
Coturnix coturnix
Alectoris chukar
Pternistis hartlaubi
Pternistis camerunensis
Pternistis erckelii
Pternistis swierstrai
Pternistis castaneicollis
Pternistis jacksoni
Pternistis adspersus
Pternistis capensis
Pternistis squamatus
Pternistis ochropectus
Pternistis ahantensis
Pternistis natalensis
Pternistis nobilis
Pternistis bicalcaratus
Pternistis swainsonii
Pternistis icterorhynchus
Pternistis clappertoni
Pternistis harwoodi
Pternistis leucoscepus
Pternistis rufopictus
Pternistis afer
Pternistis hildebrandti
Pternistis ochropectus
Figure 4.5. The strict consensus parsimony tree of spurfowls obtained from vocal characters. No
nodes received bootstrap (i.e. > 50%) support.
140
Appendix 1
Francolinus francolinus
F. pictus
F. pintadeanus
i
Sonograms of the Francolinus species.
141
Ortygornis sephaena
O. grantii
O. pondicerianus
O. gularis
Sonograms of Ortygornis species.
142
Peliperdix albogularis
P. schlegelii
P. coqui
Afrocolinus lathami
Sonograms of Peliperdix species and Afrocolinus lathami.
143
Scleroptila finschi
S. levaillantoides
S. gutturalis
S. shelleyi
Sonograms of Scleroptila species.
144
Scleroptila streptophora
S. levaillantii
S. afra
Sonograms of Scleroptila species.
145
Pternistis swainsonii
P. afer
P. rufopictus
P. leucoscepus
i Sonograms of the Bare-throated spurfowls.
146
Pternistis icterorhynchus
P. bicalcaratus
P. clappertoni
P. harwoodi
Fi
Sonograms of the Vermiculated spurfowls.
147
Pternistis hildebrandti
P. hartlaubi
P. adspersus
P. natalensis
P. capensis
Fi
Sonograms of the Vermiculated spurfowls.
148
Pternistis squamatus
P. ahantensis
P. griseostriatus
Fi
Sonograms of the Scaly spurfowls.
149
Pternistis camerunensis (1)
P. camerunensis (2)
P. nobilis
P. ochropectus
Fi
Sonograms of the Montane spurfowls.
150
Pternistis erckelii
P. swierstrai
P. castaneicollis
Fi
Sonograms of the Montane spurfowls.
151
CHAPTER 5
Taxonomy and phylogeny of ‘true’ francolins
Abstract
The development of a plethora of species and subspecies concepts has had major effects
on the delineation of terminal taxa across the Class Aves. The taxonomy and phylogeny
of small, quail-like, Afro-Asian phasianine birds now known as ‘true’ francolins, but
traditionally placed in a single genus Francolinus Stephens, 1819 with a range of other
taxa are revised. The number of taxa that have been recognized as species and, to a
larger extent as subspecies among ‘true francolins’ (Francolinus, Dendroperdix,
Peliperdix and Scleroptila spp.) (sensu Crowe et al. 2006) has never been stable. This is
due to a lack of objective, evolutionarily relevant species, subspecies and generic
circumscription.
This study aimed to establish a classification system of francolins which takes
into account the evolutionary relationships among taxa, and which, in turn, could
generally bring stability to the number of taxa recognized as valid species, subspecies
and genera based on congruent multiple lines of evidence. The model-based Maximum
likelihood and parsimony analyses of separate and combined DNA and organismal
characters resulted in some putative subspecies being elevated to the species level,
others placed into more inclusive entities, and two new genera being established. Most
of the phylogenetic hypotheses presented by Hall (1963) were rejected. The genus
Dendroperdix is replaced by Ortygornis. The ‘true’ francolins are divided among five
genera in the following pectinate phylogenetic sequence: Francolinus / Ortygornis /
152
Afrocolinus gen. nov. / Peliperdix / Scleroptila. A multi-faceted character approach
seems to be a fitting strategy with which to delineate intra-generic relationships in
francolins.
153
Introduction
Despite the challenges and uncertainties that the galliform phylogeny still presents (see
Chapter 2), we do know that ‘francolins’ sensu lato do not share common ancestry as
traditionally circumscribed. The genus Francolinus Stephens, 1819 (sensu Hall 1963), is
split into two distantly related assemblages: ‘francolins’, represented by taxa classified in
the genera Francolinus Stephens, 1819, Dendroperdix Roberts, 1922, Peliperdix
Bonaparte, 1856 and Scleroptila Blyth, 1849, and ‘spurfowls’, with all members being
assigned to the genus Pternistis Wagler, 1832.
Traditionally, 41 species (Table 1.1.) are assigned to the genus Francolinus, 36
occur in sub-Saharan Africa and five in Asia, thus making it the most specious genus in
the Galliformes (Morony et al. 1975, del Hoyo et al. 1994) and one of the largest genera
in the class Aves (Bock and Farrand 1980). Francolins are placed in the sub-family
Phasianinae, and together with other Old World partridge- and quail-like gamebirds (e.g.
Perdix and Coturnix spp.) in the tribe Perdicini (Chapin 1932, Peters 1934, Wolters 197582, Crowe et al. 1986, Johnsgard 1988, Sibley and Monroe 1990, del Hoyo et al. 1994,
Madge and McGowan 2002). The traditional composition of the genus Francolinus has
been disputed for over 50 years, with continuing uncertainty regarding the delineation and
relationships of taxa within and between constituent species.
Species of francolins and their distribution
The focus in this chapter centres on the ‘true’ francolins (Francolinus, Dendroperdix,
Peliperdix and Scleroptila spp.), that is those taxa assigned to Hall’s (1963) species
groups (see Table 1.1) with the addition of four species that Hall failed to assign to any of
her groups. The species groups are: the Spotted (represented by three Asian species),
154
Striated, Red-tailed and the Red-winged Group (all restricted to Africa). The four species
she was unable to assign to a group are two African species, Latham’s Francolin
Francolinus lathami and Nahan’s Francolin Ptilopachus ‘Francolinus’ nahani, and two
Asian species, Swamp Francolin Francolinus gularis and Grey Francolin Francolinus
pondicerianus (Table 1.1).
Generally, francolins are small to medium sized, quail-like resident birds that can
run or fly for short distances when faced with a threat. They are mainly diagnosed by
having 14 tail feathers that moult centrifugally. Most species are sexually monomorphic
in plumage, with males of most species having single spurred tarsi, with females in only a
few cases having relatively smaller spurs (Johnsgard 1988). They represent a
morphologically, ecologically and behaviourally diverse group, and have complex
distribution patterns (Snow 1978). They occur in diverse habitats of a tropical to subtropical nature, with one African species F. lathami occurring in forested habitat
(excluding P. nahani) (Hall 1963, Johnsgard 1988, del Hoyo et al. 1994, Madge and
McGowan 2002). Francolins occur at varying altitudes and most of the species
assemblages show allopatric distributional patterns among their constituent taxa (Hall
1963).
Species group diversity, distribution and morphology
Spotted Group – Genus Francolinus Stephens, 1819
This Asian group is represented by three recognized species (Table 1.1), the Black
Francolin Francolinus francolinus (Linnaeus, 1766), Painted Francolin F. pictus
(Jardine & Selby, 1828) and Chinese Francolin F. pintadeanus (Scopoli, 1786). The
nominate species of the genus, F. francolinus, with its putative subspecies (Table 1.2)
155
occurs from Cyprus to Manipur in northeastern India (Forcina et al. 2012). This species
inhabits lowland cultivations, river deltas and lake edges with scrub and reeds (Forcina
et al. 2012). Francolinus pictus represents the southern counterpart of F. francolinus, is
distributed from Sri Lanka to north India, where it favours semi-dry undulating
grasslands with scrub or cultivation (Forcina et al. 2012). Francolinus pintadeanus
represents the eastern counterpart of F. francolinus (Forcina et al. 2012), inhabiting
subtropical or tropical lowland forest from northeastern India, across Myanmar to south
eastern China, extending across western and southern Thailand (Forcina et al. 2012).
Striated Group – Genus Dendroperdix Roberts, 1922
This group comprises two species, the polytypic Crested Francolin Dendroperdix
sephaena (Smith, 1836), and the monotypic Ring-necked Francolin Dendroperdix
streptophora O. Grant, 1891. The distributional range of D. sephaena extends from
Somalia through eastern Africa to KwaZulu-Natal in South Africa, and extends
westwards across Namibia and southern Angola, where the species inhabits mostly
acacia savanna and steppe habitats (Fig. 5.1; Mackworth-Praed and Grant 1952, 1962,
1970, Hall 1963). Several subspecies (Table 1.2) are recognized: in D. sephaena there
are sephaena (Smith, 1836), spilogaster Salvadori, 1888, somaliensis Grant & Praed,
1934, schoanus Heuglin, 1873, jubaensis Zedlitz, 1913, grantii Hartlaub, 1866, rovuma
Gray, 1867, zambesiae Praed, 1920, chobiensis Roberts ,1932, thompsoni Roberts,
1924, zuluensis Roberts, 1924 and mababiensis Roberts, 1932.
Generally, D. sephaena is a rich reddish-brown bird with broad white shaft
streaks. It has a chestnut collar that is interrupted with white. The patterned feathers of
the lower neck are rich dark brown or blackish with broad white centres. This species is
156
slightly dimorphic in that the females unlike the males are slightly vermiculated. The
rest of the belly is buff with triangular chestnut markings on the breast, narrow greyish
barring on the lower breast and it has an unpatterned belly. This species can be
categorized into two distinct types, the "coastal type", which has fine drop-shaped light
chestnut streaks on the belly and the "inland type" with no streaks on the belly (Fig.
5.1). Dendroperdix streptophora has a markedly disjunct distribution occurring in
savanna-grassland in Cameroon in the west, and northern Uganda and northwestern
Kenya and Tanganyika in the east (Fig. 5.2). It differs from most of the other Redwinged francolins in having an unpatterned crown, throat colour buff rimmed with
rufous, the absence of a gorget, barred breast, blotched and mottled belly, brown (not
red) primaries, unpatterned side of the head, and a tiny spur bump.
Red-winged Group – Genus Scleroptila Blyth, 1849
Members of this quail-like plumaged group, Shelley’s Francolin Scleroptila shelleyi O.
Grant, 1890, Grey-winged Francolin S. afra (Latham, 1790), Orange River Francolin S.
levaillantoides Smith, 1836, Red-winged Francolin S. levaillantii (Valenciennes, 1825),
Finsch’s Francolin S. finschi Bocage, 1881, Moorland Francolin S. psilolaema Gray,
1867 are relatively homogenous in their overall morphology (Hall 1963, Snow 1978)
and are distinguished by having red or rufous on their primaries, which is reduced in the
Grey-winged Francolin S. afra. The distributional range of the group encompasses most
of eastern and southern Africa extending from Ethiopia and Eritrea to the Cape, and
westwards to Angola and the Congo (Fig. 5.2, 5.3, 5.4). Some species are allopatric
(Snow 1978), and others have complex distributions, with species occupying diverse
habitats at different altitudes.
157
Taxonomically, this is Hall’s (1963) most contentious group. Several of the
species in this complex are polytypic with multiple subspecies as follows: four in S.
psilolaema (psilolaema Gray, 1867, ellenbecki Erlanger, 1905, elgonensis O. Grant,
1891, theresae Meinertzhagen, 1937); seven within S. shelleyi (shelleyi O. Grant, 1890,
uluensis O. Grant, 1892, whytei Neumann, 1908, macarthuri Someren, 1938, trothae
Reichenow, 1901, sequestris Clancey, 1960, canidorsalis (Lawson, 1963)); 12 in S.
levaillantoides (levaillantoides Smith, 1836, kalaharica Roberts, 1932, pallidior
Neumann, 1908, langi Roberts, 1932, wattii Macdonald, 1953, jugularis Büttikorfer,
1889, cunenensis Roberts, 1932, stresemanni Hoesch & Niethammer, 1940, gutturalis
(Rüppell, 1835), lorti Sharpe, 1897, archeri Sclater, 1927, ludwigi Neumann, 1920);
five in S. levaillantii (levaillantii (Valenciennes, 1825), kikuyuensis O. Grant, 1897,
crawshayi O. Grant, 1896, benguellensis Neumann, 1908, clayi White, 1944).
Scleroptila finschi Bocage, 1881 and S. afra (Latham, 1790) are monotypic species.
Red-tailed Group – Genus Peliperdix Bonaparte, 1856
The Red-tailed Group is represented by small francolins generally with a wing length of
less than 150 mm (Mackworth-Praed and Grant 1952, 1962, 1970). Their distribution
range stretches from Senegal to Sudan and from central Kenya west to the central
Congo and Angola, and south to the Transvaal and Natal in South Africa (Fig. 5.5).
There is an isolated subspecies in the Ethiopian Rift. The three traditionally recognized
species are Coqui Francolin Peliperdix coqui (Smith, 1836), White-throated Francolin
P. albogularis Hartlaub, 1854, and Schlegel’s Francolin P. schlegelii Heuglin, 1863
(Table 1.1), and all are allopatric (Hall 1963, Snow 1978). Members of this group
occupy wooded grasslands. All the above taxa generally have a quail-like patterning on
158
the back which is consistent and well-defined, though the basic colour and the colour of
the crown varies from brown to rufous in different taxa.
Peliperdix coqui is one of the two African francolins (including F. lathami) that
exhibits truly marked sexual dimorphism in plumage (Hall 1963). The male birds differ
from the females in having the sides of the head and throat ochre or light buff without a
black eye stripe or necklace. Females have the sides of the head similar in colour to the
males, but the throat is whiter and they have a black necklace and a black eye stripe that
continues as a black line down the sides of the head. The breast and lower neck of the
male is barred with black, whereas the barring on the breast of the female is overlaid by
a pinkish wash and there is no barring on the lower neck.
Peliperdix coqui and P. albogularis are the two species in this group that have
recognized subspecies (Table 1.2): 14 in P. coqui (coqui (Smith, 1836), spinetorum
Bates, 1928, buckleyi Peters, 1934, maharao Sclater, 1927, ruahdae Someren, 1926,
hubbardi O. Grant, 1895, angolensis Rothschild, 1902, lynesi Sclater, 1932, vernayi
(Roberts, 1932), campbelli (Roberts, 1928), thikae Grant & Praed, 1934, kasaicus
White, 1945, hoeschianus Stresemann, 1937, stuhlmanni Reichenow, 1889), and five
for P. albogularis (albogularis Hartlaub, 1854, buckleyi O. Grant, 1892, dewittei
Chapin, 1937, meinertzhageni White, 1944, gambagae (Praed, 1920)).
The unplaced species of Hall (1963)
The species Ptilopachus ’Francolinus’ nahani, one of Hall’s enigmatic species
(others F. lathami, F. gularis and F. pondicerianus) (Table 1.1), is excluded from this
chapter on francolins since this species was found not to be a ‘francolin’, but a
‘partridge’, with close phylogenetic affinity to the African Stone Partridge Ptilopachus
159
petrosus, and both these species in turn are the closest extant relatives of the New World
quails (see Chapter 2, Cohen et al. 2012, Bowie et al. 2013). Latham’s Francolin F.
lathami is a small francolin with a disjunct distribution and is restricted to forest habitat.
Its distribution extends from Sierra Leone to western Uganda and southern Sudan, with
two further outlying pockets of occupancy in the Congo (Fig. 5.6). Two subspecies,
lathami Hartlaub, 1854 and schubotzi Reichenow, 1912, are recognized (Table 1.2).
Francolinus pondicerianus lives on dry plains in semi-desert areas (Forcina et al. 2013) in
the Gulf of Oman and on the plains of India. Francolinus gularis is patchily distributed
between south-western Nepal and extreme north-eastern India (del Hoyo et al. 1994,
Forcina et al. 2012), where it is confined to marshes and reeds (Hall 1963) on the plains
of the Ganges River.
What do we know about their taxonomy and phylogeny?
Chapter 1 outlines the taxonomic disagreement and confusion that various authors have
had with delineating taxa at the subspecific level (Table 1.2), as well as with attempts to
assign francolin species to various genera. The classic monograph of Hall (1963) on
francolins remains one of the most significant works on any group of African birds. It has
significantly enhanced our knowledge and understanding of the taxonomic status of
francolins, their evolutionary relationships, current distribution ranges including how
species might have spread to their current habitats.
What ignited further investigation of systematic relationships among francolins
was Hall’s (1963) conclusion that the genus Francolinus was monophyletic, and that the
37 species she recognized could be divided among eight putatively monophyletic groups
comprising ecologically similar, but largely allo- or parapatric species (Bloomer and
160
Crowe 1998). The reason Hall (1963) did not recognize distinct genera was because of the
difficulties presented by the Crested Francolin Dendroperdix sephaena and an atypical
Ptilopachus nahani. However, she indicated that should generic subdivision be desirable
she would have two major groups, one consisting of members that she referred to as
belonging to the Spotted Group (Asian francolin group), Bare-throated (African
spurfowls), Montane (African spurfowls), Scaly (African spurfowls), Vermiculated Group
(African spurfowls) and one of the unplaced Asian francolin F. gularis. The second group
would be comprised of the Red-winged (francolins), Striated (francolins) and Red-tailed
groups (francolins), and two unplaced species F. pondicerianus (Asian francolin) and F.
lathami (African francolin).
Wolters (1975-82) like Roberts (1924) (Table 1.2) recognized the genera:
Francolinus, Scleroptila, Dendroperdix, Peliperdix and Ortygornis. Contrary to Roberts,
Peters (1934), and Mackworth-Praed and Grant (1952, 1963, 1970), Wolters assigned the
genus Pternistis (Table 1.2) to all taxa which are known today as ‘spurfowls’. In short,
Wolters’s system of genera represents what is used today (Table 1.2) with the exception
of the genus Ortygonis, which he assigned to F. pondicerianus. The number of
traditionally recognized francolin species ranged widely over time with 17 excluding P.
nahani (in Peters 1924 - who covered Africa and Asia), 11 excluding Scleroptila
psilolaema and P. nahani (Mackworth-Praed and Grant 1952, 1962, 1970 - Africa only),
17 excluding P. nahani (Wolters 1975-82 - Africa and Asia), and 17 excluding P. nahani
(Hall 1963 - Africa and Asia).
Phylogenetic difficulties have been articulated by Crowe and Crowe (1985),
Crowe et al. (1986), Crowe et al. (1992), Bloomer and Crowe (1998), Crowe et al. (2006)
and most recently by Forcina et al. (2012), with regard to the monophyletic status of the
161
genus Francolinus, and in particular, the status of the different putative monophyletic
groups recognized by Hall (1963). It is important to note that different studies used
different types of characters (Table 1.4), focussed on different geographic areas of
francolin distribution (Table 1.3), and made use of different methods of phylogenetic
analysis. Groups that various authors have recovered as monophyletic are the Spotted
Group (Crowe and Crowe et al. 1985, Crowe et al. 1992, Forcina et al. 2012), Red-tailed.
Group (Crowe and Crowe et al. 1985, Crowe et al. 1992), and the Red-winged Group
(Crowe and Crowe et al. 1985, Crowe et al. 1992, Bloomer and Crowe 1998). No study
has ever recovered the monophyly of the Striated Group, with Crowe et al. (1986) placing
the Ring-necked Francolin Dendroperdix streptophora in the Red-winged Group
(Scleroptila). In most analyses (Crowe and Crowe 1985, Crowe et al. 1992, Bloomer and
Crowe 1998, Chapter 2), F. pondicerianus and F. gularis group with D. sephaena and as
such it would have been ideal for Forcina et al. (2012) to have included some African
francolins to strengthen their findings on the Asian francolins and purported delineation
of the Spotted Group as being monophyletic.
On the basis of the chaotic taxonomic status and the phylogenetic uncertainty of
francolin taxa (Tables 1.1, 1.2), a formal revision of all francolin taxa is urgently needed.
What is taxonomy and how can we possibly study it?
A review of species and subspecies concepts
Delineating species and/or subspecies boundaries correctly is crucial to the discovery of
life’s diversity because it determines whether or not we can recognize when different
specimens are members of the same cohesive lineage (Dayrat 2005). Many different types
of taxonomic characters (morphological, physiological, chemical, behavioural, ecological,
162
molecular, among others) have been used to delineate taxa. In taxonomy, the fundamental
components of biodiversity, species and subspecies, are discovered, described, named,
and classified. The naming of taxa, in this case francolins, is to some degree of less
concern in the present study. However, the classification of these taxa at the species-level
and below, presently lacks consensus and therefore there is a need to establish a sound
classification system that can best reflect the evolutionary history of francolins.
A number of francolin species are polytypic, being complexes of taxa that exhibit
considerable geographical variation in their biological traits and this variation poses
serious challenges to classification (Pough 1990). The concept of ‘polytypic’ species is
said to have emerged as a way of simplifying classification by lumping several diverse
but difficult to delineate taxa into a single species (Clancey 1957, Maclean 1993). The
subspecies concept is one that has been debated for over 100 years with various authors
presenting different perspectives on the meaning of ‘subspecies’. Many authors interpret
subspecies as geographically partitioned variation, which may or may not exhibit
intergradation (Winker 2010). Subspecies are often construed as biological entities that
provide evidence of adaptation and the early stages of speciation; they are also considered
to be important in improving our understanding by alerting us to interesting geographic
and behavioural patterns (Cicero 2010).
Cicero (2010), Remsen (2010) and Winker (2010) see subspecies as a meaningful
rank, others consider subspecies to be a tool of convenience (Mayr 1982a, FitzPatrick
2010). Zink (2004) calls for the classification and rank names to reflect diagnosable units
and hence considers subspecies to be of limited utility. Phillimore et al. (2010) contends
that the concept has not been applied appropriately and objectively in that there has to be
a sound understanding of the processes that govern the origination and extinction of
163
subspecies. Pruett and Winker (2010) highlighted the importance of using multiple lines
of evidence in assessing subspecific status even though they acknowledged the challenge
that can arise in cases of discord caused by varying evolutionary rates of character
change.
Another complicated issue that continues to challenge biologists is that of
developing a species definition of universal application. As Darwin (1859) articulated
there is vast vagueness in what naturalists mean by ‘species’, as a consequence a suite of
species concepts have been postulated (reviewed in Mayden 1997, de Queiroz 2007) to
date and a consensus is hardly met. de Queiroz (2007) postulated a “Unified Species
Concept” which was meant to be a reconciliation of the various species concepts available
to date. He defined species as separately evolving metasubspecies lineages or segments
(Knowles and Carstens 2007). The properties that were previously treated as necessary
properties for species, for example intrinsic reproductive isolation in the case of the
Biological Species Concept; occupation of a distinct niche or adaptive zone in the case of
the Ecological Species Concept; fixed character state differences in the case of the
diagnosable version of the Phylogenetic Species Concept are no longer considered the
defining properties of the species category, rather there is a continuum that extends
between these alternate definitions of species that may relate to their relative evolutionary
age.
Taxonomic determinations made in this chapter
The polytypic nature of several species of francolins is highlighted above and tabulated
in Table 1.2. At the end of this chapter, it is expected that some of these taxa will be
elevated to species rank, provided they represent distinct phylogenetic and
164
biogeographical eco-evolutionary entities diagnosable in terms of organismal and
molecular characters. Other taxa will likely be subsumed into the same terminal taxon if
they show relatively little phylogenetic distinctiveness and geographical and/or
ecological partitioning. The primary grouping criterion followed in this study is
multifaceted consilience in character variation (sensu Pruett and Winker 2010), because
ultimately one has to draw the taxonomic line somewhere with the goal of finding
meaningful evolutionary entities.
Hybridization is known to occur even long after speciation (Mallet 2005, Mallet
2008) and hence is not over-emphasised as a criterion for taxonomic delineation. Thus,
in this study, the decision to rank a taxon as a species, is considered if it is
morphologically and/or behaviourally diagnosable (as defined above), ≥ 1-2% divergent
in unweighted molecular characters, and confined to a specific eco-region. For a
particular taxon to be considered a subspecies, it should exhibit relatively little
phylogenetic distinctiveness and geographical and/or ecological partitioning, be < 1%
divergent in unweighted molecular characters, and be confined to a specific eco-region.
Objectives of the study:
To review the taxonomic status of the ‘true’ francolins, with a particular focus
on African francolins.
To re-assess the monophyletic status of the various species groups proposed by
Hall (1963).
To investigate the phylogenetic relationship between Hall’s (1963) enigmatic
unplaced Asian species F. pondicerianus and F. gularis , and African species (F.
lathami) in light of the findings of Cohen et al. (2012) and Forcina et al. (2012).
165
To produce a revised classification of African francolins that takes into account
the evolutionary relationships among taxa.
Materials and methods
Data collection
Morpho-behavioural characters of francolins
In total, 24 organismal characters reflecting assessment of plumage/integument and
colour/pattern (Fig. 5.7), as well as measurements of certain qualitative and quantitative
structures (Table 5.1), and several vocal characters extracted from Crowe at al. (1992)
and chapter 4, were scored (Table 5.2). Morphometric characters representing billlength, tarsus- and spur-length were obtained using a Vernier Calliper. A stopped wingrule and a normal ruler were used to measure wing- and tail-length, respectively. Winglength was measured with the wing chord flattened and straightened for enhanced
accuracy.
Molecular characters
For within-group molecular analyses of ‘true’ francolins, 41 terminal taxa (including
two outgroup species) were sampled (Tables 5.3) for the entire mitochondrial
Cytochrome-b gene (CYTB - 1143 base pairs- bp). GenBank accession numbers of taxa
sequenced are detailed in Appendix 5.1. Primers used in sequencing molecular markers
are listed in Tables 5.4 and 5.5; 67% of DNA samples of francolins sequenced were
derived from toe-pads sub-sampled from museum skins.
For the toe-pad samples, instead of sequencing the CYTB gene in a single
reaction using the available universal primers, multiple sequence fragments (six for each
166
sample) were generated (using primers specifically designed for francolins, Table 5.5)
due to the degraded state of the DNA. This meant that 1143 bp for each toe-pad sample
would only be recovered by assembling six different PCR-amplified and sequenced
fragments.
Maps and mapping of distribution records of investigated taxa
The maps showing the distribution ranges of francolins were produced (Fig. 5.1 – 5.6).
This was a challenging task given that the actual point locality data for most taxa were
not available. Ranges of all the traditionally recognized species presented in Snow’s
(1978) atlas were used as the basis for generating range maps (Chapter 5 and 6) because
these ranges were produced from the point locality data associated with skins housed in
museums. The distribution data in Harrison et al. (1997) were then overlaid on Snow’s
ranges and this was very helpful in filling in the gaps in the ranges at least of the
southern African species. The 2nd step used Hall’s (1963) distributions to cover the
ranges of both species and subspecies that she recognized.
Step 3 involved the superimposition of Mackworth-Praed and Grant’s (1952,
1962, 1970) ranges on Snow’s ranges to cover both species and subspecies that these
authors recognized. Mackworth-Praed and Grant (1952, 1962, 1970) and Hall (1963)
are the two main taxonomic revisions that covered the Africa and Africa and Asia
species respectively, contrary to revisions in Roberts (1924), Clancey (1967) and
Hockey et al. (2005) that covered smaller geographic areas in which francolins occur. In
Step 4 Clancey (1967) and Hockey et al. (2005) were used to account for the
distributions ranges of other populations as they recognized them. These two revisions
covered the southern African region only. Distribution ranges in steps 2-4 were
167
generally superimposed on Snow’s ranges and gaps, which may either translate to the
real gaps in distribution or are indicative of under-sampling in the intervening area. It is
only with the exception of Peliperdix schlegelii where the range is not broken according
to Snow since the species is confirmed to have a continuous distribution from eastern
Cameroon through to Central African Republic, to the western Bahr-el-Ghazal in Sudan
(Mackworth-Praed and Grant 1970, Hall 1963). With regard to the names of studied
localities of taxa, old rural names were used to a certain extent when translating them
posed difficulty. In order to prepare distribution ranges of taxa to be used in the spatial
analysis of vicariance, distribution records were mapped on a continuous basis using a
2x2
degree
grid
on
the
world
map
which
was
accessed
on
[http://earthobservatory.nasa.gov/Features/BlueMarble/].
Data analyses
Phylogenetic analyses
The generated nucleotide sequences were edited and assembled in the Staden Package
(Staden et al. 2003) and aligned in MAFFT (Katoh et al. 2009). All CYTB (coding
genes) sequences generated in this study were checked for stop codons before they were
analyzed by translating them into amino acids and this was done online at EMBL
[http://www.ebi.ac.uk/emboss/transeq/]. Two phylogenetic inference methods with
different optimality criteria were employed to generate phylogenetic hypotheses:
maximum likelihood (CYTB) and parsimony (CYTB, organismal, combined
CYTB/organismal characters). As suggested by the results of the much larger data set in
Chapter 2, all data matrices were rooted on Gallus gallus and Bambusicola thoracica.
For all analyses, characters were treated as non-additive. Parsimony-based phylogenetic
168
analyses were conducted using PAUP ver. 4.0b10 (Swofford 2002). The following
settings were effected: full heuristic search with all characters unordered and with equal
weight, starting tree(s) obtained via stepwise addition; tree-bisection-reconnection
branch-swapping, 1000 random additions of taxa (Maddison 1991), one tree being held
at each step during stepwise addition , branches were collapsed (creating polytomies) if
the maximum branch-length was zero. When multiple, equally parsimonious
cladograms were recovered, a strict consensus cladogram was constructed. The extent to
which each non-terminal node is supported by different character partitions was
determined by using the bootsrap (BS) (Felsenstein 1985) resampling strategy with
1000 pseudoreplicates, with 5 random additions of taxa per bootstrap pseudoreplicate.
Since different codon positions evolve under different models of evolution, it
has been argued that a partitioned, mixed-model approach should be adopted (Ronquist
and Huelsenbeck 2003, Nylander et al. 2004). Mixed-model analyses allowed different
parameters (base frequencies, rate matrix or transition/transversion ratio, shape
parameter, proportion of invariable sites) to vary among the three codon positions.
Mixed-model maximum likelihood analyses were performed using the
Randomised Axelerated Maximum Likelihood algorithm for High Performance
Computing (RAxML) v7.0.4 (Stamatakis 2006, Stamatakis et al. 2008) as implemented
on the CIPRES portal. Mixed-model RAxML analyses make use of a GTR++ model
partitioned by gene or codon postion. Support at nodes was assessed with 100 nonparametric bootstrap (BS) pseudoreplicates. The use of different methodological
approaches (optimality criteria) facilitated the identification of method-based
incongruence.
169
Genetic distances
Uncorrected pairwise distances (Table 5.6) were calculated in PAUP ver. 4.0b10
(Swofford 2002) for the CYTB data matrix and were converted to percentage sequence
divergence.
Results and Discussion
Separate versus combined phylogenetic analyses
The parsimony analysis for the CYTB data set (with 1143 bp characters, yielded 315
parsimony informative characters and 110 variable characters that were parsimony
uninformative. Four trees of 1356 steps were recovered. The organismal data set
comprised 24 characters, 23 of which were parsimony informative, and one that was
parsimony uninformative. In total, 390 trees of 104 steps were recovered. The combined
CYTB/organismal data set comprised 1167 characters of which 334 were parsimony
informative and 115 that were variable but parsimony uninformative. In total, 24 trees
of 1491 steps were recovered.
The systematics of Hall’s species groups and unplaced species
Striated Group and the unplaced Asian species
Traditionally, the Striated Group encompasses two recognized species, a polytypic
Dendroperdix sephaena and a monotypic species Dendroperdix streptophora, and 12
subspecies assigned to D. sephaena: sephaena, spilogaster, somaliensis, schoanus,
jubaensis, grantii, rovuma, zambesiae, chobiensis, thompsoni, zuluensis and
mababiensis.
170
In comparing the phylogenetic trees (Fig. 5.8, 5.9, 5.10, 5.11), none of these
analyses ever recovered the monophyly of the Striated Group as hypothesized by Hall
(1963). In all the trees, D. streptophora joins the Red-winged Group whereas D.
sephaena (labelled Ortygornis in the trees) is well separated from the Red-Winged
Group and remains at the basal parts of the trees. These results confirm earlier
suggestions of the distinction of these two taxa (Crowe and Crowe 1985 and Crowe et
al. 1992). Dendroperdix sephaena (together with rovuma, grantii, see below) form a
sister relationship in most analyses (Fig. 5.8, 5.9, 5.11, respectively) with the two
unplaced Asian species F. pondicerianus and F. gularis, which in turn emerge as sister
species with high support. The organismal tree (Fig. 5.10) does not support the
phylogenetic association of Dendroperdix taxa with F. pondicerianus and F. gularis,
but instead places D. sephaena with other members of the genus Francolinus.
Francolinus gularis differs from F. pondicerianus with 5% sequence divergence (Table
5.6), from sephaena with 8%, and there is 9% divergence between pondicerianus and
sephaena. Francolinus gularis and F. pondicerianus diverged from each other 3.1 mya
and both species diverged from D. sephaena at 5.5 mya – Fig. 2.7).
Francolinus gularis is quite different in colouration and patterning (with barred
back) from the other Asian francolins. The white streaks on the belly strongly contrast
with the rufous throat and wings. Francolinus pondicerianus is finely patterned with the
nominate subspecies pondicerianus being the darkest and the least grey and has the
greatest amount of chestnut markings. The consistent grouping of F. pondicerianus and
F. gularis with the Dendroperdix taxa (Fig. 5.8, 5.9, 5.11, 2.2, 2.3, 2.5, Bloomer and
Crowe 1998) refutes the finding in Forcina et al. 2012 that the five Asian species (F,
francolinus, F. pictus, F. pintadeanus, F. pondicerianus, F. gularis) are monophyletic.
171
Therefore, it is recommended that F. pondicerianus and F. gularis be considered
congeneric and be placed in the genus Ortygornis Reichenbach, 1853 over
Dendroperdix Roberts, 1922 based on priority, and because Ortygornis was once used
for pondicerianus (Roberts, 1940). Therefore, there will be Ortygornis pondicerianus
(Gmelin, 1789) and O. gularis (Temminck, 1815) (Appendix 5.2).
The Dendroperdix taxa form a monophyletic assemblage with sephaena being
sister to rovuma and grantii (Fig. 5.8, 5.9, 5.11) with moderate to high support (see
rationale below for splitting these taxa from D. sephaena). Rovuma and grantii, are in
turn, sister to each with high support in most analyses (Fig. 5.8, 5.9, 5.11, respectively)
with the exception of the organismal only dataset (Fig. 5.10), where relationships
among these taxa are unresolved. With respect to uncorrected pairwise genetic
divergence, rovuma differs from sephaena by 4%, and sephaena differs from grantii by
5%, and rovuma differs from grantii by 4%.
Dendroperdix sephaena can be categorized into two distinct types, the coastal
types that have fine drop-shaped light chestnut streaks on the belly and the inland type
with no streaks on the belly. Grantii has a reduced band of chestnut triangular marks on
the breast, narrow greyish somewhat U-shaped streaks on breast and belly with the
distal part of the belly being unpatterned. Rovuma has a band of chestnut triangular
marks being reduced without covering much of the breast, no barring on the belly, has
drop-shaped chestnut streaks on the belly and not much distally. In the north, the
Ethiopian subspecies spilogaster is barred and streaked on the breast with barring being
extensive as in sephaena. The Somalian subspecies somaliensis is like spilogaster, but
the streaking is not extensive. There is confirmed evidence of hybridization taking place
in the north between the southern subspecies rovuma and the northern spilogaster (Hall
172
1963, Snow 1978). The distribution range of rovuma and spilogaster is intercepted by
that of grantii such that the two streaked subspecies are isolated geographically. It was
not possible to fully account for the taxonomic status of spilogaster in the absence of
genetic information. Given the difference in morphology between spilogaster and
rovuma and the geographic isolation between the two subspecies, spilogaster (in the
north) and rovuma (in the south) both separated by grantii, the recommendation is that
rovuma (once recognized as a species in Roberts 1924) be considered a separate species
within which two subspecies rovuma and spilogaster are recognized. Furthermore, since
grantii, jubaensis and schoanus are morphologically inseparable and genetic evidence
showed that little genetic distance exists between grantii and jubaensis and schoanus,
they are synonymized with grantii which is recognized as a full species.
The taxa, sephaena, zambesiae, chobiensis, zuluensis, thompsoni and
mababiensis all have buff bellies (with very narrow greyish barring) and a broad band
of chestnut triangular markings on the breast and an unpatterned belly. Based on similar
morphological attributes and the availability of genetic evidence for some subspecies,
zambesiae, chobiensis, zuluensis, thompsoni and mababiensis should be synonymized
with sephaena.
Based on the phylogenetic association of the Dendroperdix taxa with the Asian
species Ortygornis gularis and O. pondicerianus, it is recommended that the genus
Ortygornis Reichenbach, 1853 replace Dendroperdix Roberts, 1922 based on priority.
Therefore, the taxonomic status of the African members of this genus would be:
Ortygornis sephaena (Smith, 1836), Ortygornis rovuma with subspecies Ortygornis
rovuma rovuma Gray, 1867 and Ortygornis rovuma spilogaster Salvadori, 1888 and
Ortygornis grantii Hartlaubi, 1866 (Appendix 5.2).
173
The unplaced Francolinus lathami
This is one of Hall’s (1963) enigmatic species, that she could not place in any of her
groups even though she speculated that it is closest to the Red-tailed francolins
(Peliperdix spp.) and that morphological similarities between it and the spotted Asian
Black Francolin Francolinus francolinus are superficial or plesiomorphic. Two putative
subspecies, lathami and schubotzi are recognized.
The underparts of nominate lathami are largely black with white spots in the
male as opposed to a brown belly colour with white spots in female. Both sexes have a
black throat and patterned face (greyish in male and brownish in female). The
upperparts are mottled rufous and brown with pronounced white streaks (margined
mostly with black) restricted to the lower neck. There is geographical variation in
plumage between the nominate subspecies and schubotzi. The males of schubotzi have
the black and white pattern extending down the bulk of the belly continuing to the
undertail coverts (Chapin 1932, Hall 1963), where the ground colour still predominates.
The females cheeks are more rufous than grey. The species exhibits striking plumage
dichromatism and is markedly distinct from any of the African francolins.
The phylogenetic inferences place F. lathami near the base of the tree in nearly
all the analyses (Fig. 5.8, Fig 5.9, Fig 5.11). The two putative subspecies lathami and
schubotzi are supported as sister taxa in all the analyses with high support. Wolters
(1975-82), Crowe and Crowe (1985) and Crowe et al. (1992) supported Hall’s (1963)
suspicion that F. lathami has affinities with members of the Red-tailed Group, with
Crowe and Crowe (1985) and Crowe et al. (1992) placing this species in the subgenus
Peliperdix. In our analyses, only the organismal data matrix places F. lathami close to
members of the Red-tailed Group (Fig. 4.10). Dating of the molecular characters
174
suggests Francolinus lathami and the Peliperdix taxa split a long time ago, c. 7.6 mya
(Fig. 2.7). The two recognized subspecies, lathami and schubotzi differ by 1% in
sequence divergence (Table 5.6).
Francolinus lathami links the relatively basal Asio-Afrotropical francolins and
the quail-like Red-tailed Peliperdix and Red-winged Scleroptila taxa. On the basis of its
distinct phylogenetic placement, morphology, unique forest habitat, geography and
large genetic divergence, lathami should be recognized in a genus of its own and it
should continue to comprise two subspecies lathami and schubotzi. Therefore, the genus
Francolinus Stephens, 1819 will be replaced with a newly erected genus Afrocolinus
gen. nov. Afrocolinus is formed from two words, afro- meaning African and -colinus for
‘quail’. The subspecies should be Afrocolinus lathami lathami Hartlaub, 1854 and
Afrocolinus lathami schubotzi Reichenow, 1912 (Appendix 5.2).
Red-tailed group
This group is at represent represented by three species, Peliperdix coqui, P. albogularis
and P. schlegelii. Within P. coqui there are 14 recognized subspecies: coqui,
spinetorum, buckleyi, maharao, ruahdae, hubbardi, angolensis, lynesi, vernayi,
campbelli, thikae, kasaicus, hoeschianus and stuhlmanni, and five subspecies are
recognized within P. albogularis: albogularis, buckleyi, dewittei, meinertzhageni and
gambagae. Peliperdix schlegelii is at present considered a monotypic species, but was
once considered a subspecies of Peliperdix ‘Francolinus’ coqui (Peters 1934).
The Red-tailed Group is recovered as monophyletic in all analyses (Fig. 5.8, 5.9,
5.10, 5.11) with moderate to high support, with the exception of the organismal analyses
where support is lacking. This finding corroborates those of Crowe and Crowe (1985)
175
and Crowe et al. (1992). What stands out within the Red-tailed Group is the split
between the P. schlegelii/P. albogularis complex and the rest of the Peliperdix taxa in
all the analyses. Peliperdix coqui consists of a number of geographical variants and
exhibits sexual dichromatism in plumage (Hall 1963). The various subspecies all have
bellies mostly narrowly barred (except ruahdae with broader, blacker and more widelyspaced barring), though the degree of barring is variable and the females have a pink
wash on their upper belly. Both sexes generally have grey wings, with the exception of
kasaicus and vernayi that have redder, rufous and pinkish wash in the wings,
respectively. The colour of the belly varies from being buffish white in coqui, vernayi,
ruahdae, stuhlmanni, to tawny in kasaicus. The degree of barring in stuhlmanni is
intermediate between that in the nominate coqui and hubbardi in that the barring is not
pronounced especially on the belly. Within P. coqui, the various subspecies form a cline
in which both sexes of the west African spinetorum have unbarred bellies, the east
African hubbardi has barring on the flanks leaving the centre of the belly unbarred, the
east African maharao is like the southern African P. coqui subspecies in having wholly
barred bellies.
Among the recognized subspecies of P. coqui, there is a geographical split
between the east/west African subspecies (maharao, hubbardi, spinetorum) and
central/southern African subspecies (coqui, vernayi, stuhlmanni, kasaicus) (Fig. 5.8,
5.9, 5.11). However, nodal support for the east/west African clade is only recovered in
the ML CYTB tree (Fig. 5.8 - 69% BS). What is also supported in the CYTB parsimony
tree (Fig. 5.10, 4.11) is the sister relationship between maharao and hubbardi (Fig. 5.9 57%, Fig. 5.11, 5% sequence divergence). In the CYTB ML tree maharao is sister to
spinetorum with poor BS support (Fig. 5.9), whereas hubbardi and spinetorum are sister
176
taxa in the organismal tree (55% BS, 6.4% sequence divergence; Fig. 5.10). Coqui pairs
with vernayi in the CYTB ML (93% BS, 3% sequence divergence) and CYTB
parsimony (70% BS) tree, whereas coqui pairs with ruahdae in both the organismal and
CYTB/organismal trees (all 64%). Stuhlmanni and kasaicus are sister taxa in the CYTB
parsimony tree but with no BS support; they differ by 5% sequence divergence. Their
systematic relationship is unresolved in the ML CYTB and organismal tree. Overall, the
nearest non-coqui CYTB phylogenetically close and with the least genetic distance to
coqui is hubbardi, differing at 5% sequence divergence.
It is recommended that lynesi, campbelli and angolensis be included in coqui
based on morphological similarity, phylogenetic affinities and that there exists limited
genetic divergence (<1%), and that vernayi (Roberts, 1932) be considered a subspecies
within P. coqui based on their morphological differences (the red wings as opposed to
grey wings in nominate coqui, as well as the much redder crown as opposed to the less
red crown of coqui). their phylogenetic affinity and small genetic divergence from coqui
(3%), It should therefore be recognized as a subspecies Peliperdix coqui vernayi
(Roberts, 1932). Hoeschianus Stresemann, 1937 should be synonymized with vernayi
based on similar overall morphological appearance and geographic proximity.
Specimens of ruahdae were not examined directly and as such the morphological
description in Mackworth-Praed and Grant, and Hall, was used to classify the Ugandan
ruahdae and hoeschianus. The genetic distance between ruahdae (tissue from Rwanda)
and the South African nominate coqui is 4%. This is not surprising since the two
subspecies are morphologically different (ruahdae having broad, much black and
spaced barring on the belly) and they are geographically isolated. This taxon should
177
therefore be considered a subspecies within P. coqui to be named Peliperdix coqui
ruahdae van Someren, 1926.
Stuhlmanni could be a cryptic taxon that morphologically looks like an
intermediate between coqui (strongly barred) and thikae and hubbardi (barring
restricted to flanks). The rest of the belly is not strongly barred as in coqui, but the
barring covers the rest of belly and not just the flanks as in thikae and hubbardi.
Stuhlmanni is supported as sister to kasaicus in the parsimony CYTB and parsimony
CYTB/organismal tree even though with no BS support and it is very divergent
genetically (~5-8%) compared to any other P. coqui subspecies. The two subspecies
should be considered species and hence Peliperdix stuhlmanni Reichenow, 1889.
Perhaps kasaicus, like stuhlmanni, is a cryptic taxon based on the genetic distance,
geographic isolation, and phylogenetic placement despite the fact that it has similar
morphology to coqui. Kasaicus should be considered a species based on the
geographical isolation with vernayi and be named Peliperdix kasaicus White, 1945. The
western spinetorum and east African maharao and hubbardi form a clade. Spinetorum is
unarguably evolutionarily distinct, geographically isolated from the rest of P. coqui
subspecies, genetically different from other subspecies by 6-9%, and both sexes have
plain unbarred bellies, but have rufous wings as in maharao. It should therefore be
considered a species and be named Peliperdix spinetorum Bates, 1928.
The isolated Kenyan subspecies maharao has its closest phylogenetic
association with hubbardi. On the other hand, there is considerable genetic divergence
(5%). Unlike hubbardi, the rest of the belly of maharao is barred. On the basis of the
remarkable genetic distance, differences on the belly and the isolated distribution,
178
maharao should be considered a species and be named Peliperdix maharao Sclater
1927.
Thikae and hubbardi have a genetic difference of 3%, occur side by side
geographically and have the similar belly morphology despite the difference in wing
colour (rufous in thikae and grey in hubbardi). Therefore, hubbardi Ogilvie-Grant, 1895
should be recognized as a species and thikae Grant and Mackworth-Praed, 1934 must be
synonymized with it and be named Peliperdix hubbardi Ogilvie-Grant, 1895 (Appendix
5.2).
Peliperdix schlegelii is the sister species to the P. albogularis complex and it
shares the least sequence difference (1%) with buckleyi (Table 5.6). There is no nodal
support for the P. schlegelii/P. albogularis complex clade in the organismal tree. Within
the P. albogularis complex, albogularis and buckleyi are sister taxa (Fig. 5.8, 5.9, 5.11
– 99%, 100%, 100% BS respectively) whereas albogularis, buckleyi and dewittei are
similar in the organismal tree (Fig. 5.10). The taxon dewittei/ ‘meinertzhageni’ forms a
sister relationship with albogularis (with 5% sequence divergence) and buckleyi with
high support in the molecular analyses (Fig. 5.8, Fig. 5.9, 5.11). The nearest relative of
buckleyi is albogularis with 1% sequence divergence.
The Peliperdix albogularis complex has a disjunct distribution across west and
west-central Africa with the nominate subspecies being restricted to Guinean savanna in
West Africa between Senegal and Cameroon (Cotterill 2006). Peliperdix albogularis
resembles P. c. coqui, but the quail-type patterning is less well-defined in the females
and the shaft streaks and barring are much narrower. The wings are more rufous as in
the east Kenyan subspecies of coqui. The females (but not the males) also have the
black facial pattern and necklace, but both are poorly-defined. Both sexes have a buff
179
throat contrasting with the ochre cheeks. The rest of the underparts of male albogularis
are quite distinct from any other forms of coqui, being chestnut on the upper belly with
ochre shaft streaks and richer ochre on the lower belly, lacking any dark barring. The
females are barred (to varying degrees) with a faint pinkish wash or rufous on the upper
belly similar to the Ethiopian subspecies of P. c. maharao and the east Kenyan P. c.
thikae. The subspecies in which the females are less patterned with narrow barring
restricted to the upper belly and flank is in Gambia (albogularis – Hall 1963). There is a
subspecies that Peters (1934) recognized as gambagae (Mackworth-Praed, 1920) from
the type locality Gambaga, Gold Coast Colony) which Hall and Serle (1957)
synonymized with buckleyi. The taxon dewittei (including meinertzhageni) occurs south
of the Congo forest in moist montane grasslands and dambo floodplains (Cotterill 2006)
and it is considered common in some parts, even though it was said to not have been
collected on the Marungu plateau since 1931 (Dowsett and Prigogine 1974). Birds of
both sexes are larger than albogularis (wing length ranging from 142 to 147 mm –
Mackworth-Praed and Grant, 1970) and more richly coloured than both albogularis and
buckleyi with the female dewittei being more heavily barred although this is not as
extensive as in buckleyi. The eastern Angolan and possibly the northwestern Zambian
taxon meinertzhageni comprise the darkest birds with heavily barred females. This
subspecies is found in montane grassland and it is restricted to the Upper Zambezi
floodplains (Cotterill 2006).
The taxon dewittei is indisputably morphologically different from the other
subspecies recognized within P. albogularis, and is divergent genetically. The
morphology of dewittei (southeastern Congo) resembles that of the Angolan and
northwestern Zambian ‘meinertzhageni’ (both sexes are large birds with a wing length
180
of 140-147 mm, rich dark-coloured birds, heavily patterned). Unfortunately, dewittei
could not be sequenced. Geographically, the two subspecies occur in different parts (but
occupy similar habitat). Hall (1963) speculated that the northwest Zambian
meinertzhageni subspecies may be more closely related to the Congo dewittei than the
Angolan meinertzhageni subspecies, this speculation may have been based on the
geographical proximity.
Based on morphology, geographical and molecular evidence for the Angolan
meinertzhageni (in the absence of molecular evidence from northwest meinertzhageni
subspecies), meinertzhageni White 1944 should be moved to species level. It is not
possible at this stage to comment on Hall’s speculation about the relationship of the
north-west Zambia meinertzhageni and the Congo dewittei subspecies. Despite the
absence of molecular evidence for dewittei Chapin, 1937, its name has to be given the
priority and meinertzhageni will be synonymized with it. The name must be Peliperdix
dewittei Chapin, 1937.
The West African buckleyi and gambagae are inseparable morphologically and
they are almost identical genetically with 0.4% difference. The two subspecies should
be merged and form a subspecies with the name buckleyi Ogilvie-Grant, 1892 taking
priority over gambagae (Mackworth-Praed, 1920). The recognized name must be
Peliperdix albogularis buckleyi Grant, 1892. So, Peliperdix albogularis Hartlaub, 1854
will be represented by two subspecies, Peliperdix albogularis albogularis Hartlaubi,
1854 and Peliperdix albogularis buckleyi Ogilvie-Grant, 1892 (Appendix 5.2).
Another controversial taxon within the Red-tailed Group is Peliperdix schlegelii.
This taxon was considered a subspecies of P. coqui in Chapin (1932) and Peters (1934).
Contrary to this, Hall (1963) and Mackworth-Praed and Grant (1970) recognized this
181
taxon as a species. This is a rare bird; its distribution stretches from eastern Cameroon
through the Central African Republic to the western Bahr-el-Ghazal in Sudan. The back
plumage colouration and patterning of P. schlegelii is closer to that of P. albogularis
than P. coqui, though the quail-type patterning is much reduced and the extent of sexual
dimorphism is also pronounced. The differences are seen in the male P. schlegelii,
which has broad buff shaft streaks and some streaks with few transverse blackish brown
bars whereas the female is almost unpatterned. The back in both sexes is vinous
chestnut with the belly being buff to off-white. The male P. schlegelii is similar to
southern P. coqui subspecies with ochre sides of the head and throat and narrow black
and white barring on the upper and lower belly. The female resembles the male on the
crown and throat but has the upper belly feathers edged with black, giving a more
mottled appearance of simply triangular marks. The flanks are sparsely barred.
The findings demonstrated that P. schlegelii is an evolutionarily distinct taxon
both morphologically and genetically. It is more closely related to P. albogularis than it
is to P. coqui. Therefore, P. schlegelii must be given full specific status and be named
Peliperdix schlegelii Heuglin, 1863. These findings suggest that Chapin (1932) and
Peters (1934) were conservative in considering this taxon a subspecies of P. coqu
(Appendix 4.2).
Red-winged Group
The Red-winged Group is represented by six species, Scleroptila shelleyi, S. afra, S.
levaillantoides, S. levaillantii, S. finschi and S. psilolaema. Scleroptila afra and S.
finschi are monotypic species whereas the balance of the species are polytypic.
Scleroptila shelleyi comprises seven subspecies (shelleyi, uluensis, whytei, macarthuri,
182
trothae, sequestris, canidorsalis); S. psilolaema four subspecies (psilolaema, ellenbecki,
elgonensis, theresae); S. levaillantii five subspecies (levaillantii, kikuyuensis,
crawshayi, benguellensis, clayi); S. levaillantoides 12 subspecies (levaillantoides,
kalaharica, pallidior, langi, wattii, jugularis, cunenensis, stresemanni, gutturalis, lorti,
archeri, ludwigi).
The Red-winged Group is monophyletic in all the analyses (Fig. 5.8, 5.9, 5.10,
5.11) with varying levels of support. One further common aspect, is the incorporation of
S. streptophora as a basal taxon to the group in all the analyses (Fig. 5.8, 5.9, 5.10,
5.11). This finding is consistent with that of Crowe and Crowe (1985) and Crowe et al.
(1992). This group presents a very complicated and variable phylogenetic structure with
few nodes are consistently supported in most of the analyses: (1) kikuyuensis and
crawshayi and (2) psilolaema and theresae.
The major taxonomic change relevant to this group has been the shift of the
monotypic Scleroptila streptophora from the Striated to the Red-winged Group (see
above). It has a markedly disjunct distribution occurring in savanna-grassland in
Cameroon in the west, and northern Uganda and northwestern Kenya and Tanganyika in
the east. It differs from most of the other Red-winged francolins in having an
unpatterned crown, throat colour buff rimmed with rufous, the absence of a gorget,
barred breast, blotched and mottled belly, brown (not red) primaries, unpatterned side of
the head, and tiny spur bump. It differs from uluensis and shelleyi by 8% sequence
divergence.
The monotypic specific status of S. finschi has never been disputed. This species
is found disjunctly in western Angola and parts of western Congo (Mackworth-Praed
and Grant 1962, Hall 1963). It differs from the other Red-winged francolins in lacking
183
black and white patterning on the face and neck and by having the upper belly and the
lower end of the lower belly grey. The centre of the belly is buff with chestnut blotching
and it is essentially unbarred. The sides of the face and border of the throat are ochre as
in S. levaillantii but this is less extensive on the hind neck. This species is long-billed
like S. levaillantii. It shares the least genetic distance with shelleyi (5% sequence
divergence – Table 4.6).
Scleroptila shelleyi has a buff to white throat, and a distinct black necklace, a
moderately developed gorget, blotched and mottled upper belly and a broadly barred
lower belly. It occurs in moist, rocky hillsides below 2000 m throughout southeastern
Africa from Mozambique in the north to South Africa, and shares the least genetic
distance with gutturalis (4% sequence divergence). Gutturalis was considered by Hall
as one of the northeastern subspecies (lorti and friedmanni, stantoni, eritreae, archeri)
attributed to S. levaillantoides. It occurs in arid savanna steppe in eastern Ethiopia,
Eritrea and northeastern Kenya, and differs from other northern Red-winged francolins
in having the lower belly largely unpatterned with bars and throat base freckled, with
dark above a well-defined necklace and gorget. It shares the least genetic distance with
uluensis (3%). Uluensis occurs in central Kenya, southern Uganda and northeastern
Malawi. It is similar in form and habitat to S. shelleyi, but the necklace and gorget are
somewhat less distinct and the bill markedly shorter. Some specimens have the throat
moderately freckled with black indicating possible hybridization with gutturalis. The
Malawian taxon whytei is very different from shelleyi and uluensis in having a rufous
buff throat and an indistinct necklace. It has a moderately-developed gorget blotched
with reddish chestnut distinguishing it from the other Scleroptila spp. The belly is buffy
184
white with broad black bars on a background of rufous buff. Its nearest CYTB relatives
are finschi, gutturalis and psilolaema at 5% sequence divergence.
Scleroptila afra is endemic to southern Africa generally occurring in montane
grasslands above 1800 m. It differs from uluensis, shelleyi and whytei in having a buff
to white throat flecked with black. The reddish colour which is apparent in the wings of
other scleroptilids is extremely reduced to absent. The degree of black barring on the
buffy belly is very narrow whereas the barring on the underparts of the others is broad.
The gorget in S. afra is well-defined and eventually grades to the upper belly which
forms a band mottled with grey, chestnut and tawny. The former eastern Transvaal
province of South Africa subspecies proximus (which does not differ genetically from
those to the south) is according to Clancey (1967) “similar to afra, but darker and more
saturated”. “On the underside, rather deeper tawny colour to the ground of the lower
throat and upper breast, the grey tipping to the feathers more extensive and darker; rest
of the underside more yellowish tinged, and chestnut spotting darker and heavier”. The
nearest relative is ellenbecki differing at 2% sequence divergence. Ellenbecki is very
similar to afra in having a buff throat freckled with black, a poorly developed necklace,
but differs in having broad black and white barring on the lower belly. It inhabits
montane grassland in southern Ethiopia above 2500 m.
Scleroptila psilolaema and theresae are typical Red-winged francolins and are
similar in appearance in having the most red in their primaries, but psilolaema is
significantly smaller and shorter billed. Both inhabit montane grassland above 2500 m,
the former from southern Ethiopia and the latter from Mt. Kenya and the Aberdares.
They differ from ellenbecki in having little or no freckling on the throat and have a
much more poorly developed necklace and gorget which is replaced by rich chestnut
185
with black indistinct spots. Scleroptila psilolaema differs from theresae with 2.6% and
from pallidior at 5% sequence divergence.
Scleroptila levaillantii is endemic to rank highland (1600-2000 m) grasslands of
southeastern Africa. It is the largest of the southern African francolins and differs from
the other Red-winged francolins in having an ochre collar on the sides of the face and
edges of the throat and also inside the black and white facial pattern. Its face is like that
of S. levaillantoides, but this species has a long bill, it is darker and richer in colour. The
black and white patterning of the well-defined necklace (forming a well-developed
gorget) extends in a complete collar round the hind neck below the ochre collar and the
black and white stripes from above the eye (which in S. levaillantoides and S. shelleyi
run down the side of the face) in the South African S. levaillantii subspecies behind the
head to join at the back. It also differs from its congeners in having more red on the
wings, a longer bill and shorter spurs. It differs from ellenbecki at 6.6% sequence
divergence.
On the other hand, S. levaillantoides is a species endemic to the arid grasslands
of northern Namibia, Botswana and central South Africa, occurring at lower altitudes
than shelleyi, levaillantii and afra. It is also smaller than these taxa. It has a moderately
developed gorget with the upper belly blotched with buff and reddish brown. The
generally paler subspecies pallidior is still tentatively placed with levaillantoides
despite its genetic affinity with ellenbecki (9% sequence divergence) and that it is
otherwise identical to the nominate subspecies in its overall biology. It shares the least
genetic distance with afra at 3% sequence divergence. The taxonomic treatment renders
S. levaillantoides paraphyletic in molecular tree. Jugularis is according to Hall (1963)
endemic to southeastern Angola and northwestern Namibia and classified as a
186
subspecies of S. levaillantoides. It differs from this taxon in being paler overall and
having a well-developed necklace which grades to the breast to form a broad gorget. It
differs from crawshayi at 4% sequence divergence.
Crawshayi is endemic to northeastern Malawi and western Tanzania and is
richer in colour than levaillantii, with still more red in the wing, and with more black
marking on the belly, whereas kikuyuensis (which occurs in Angola and southern Congo
– Fig. 5.6) is without the stripe on the hind neck, but otherwise similar to crawshayi.
Both have similar habitats to levaillantii. The nearest relative of crawshayi is afra
which differ at 3% sequence divergence. The nearest relative of kikuyuensis is
crawshayi with 2% sequence divergence.
The taxon theresae was considered a subspecies of S. shelleyi by Mackworth
-Praed and Grant and a synonym of psilolaema by Hall (1953). Since it is sister to
psilolaema here and is 3% divergent from it (as opposed to 6% to shelleyi), Hall appears
to have been correct. On the other hand, ellenbecki was recognized by MackworthPraed and Grant as a subspecies of S. afra, but synonymized with psilolaema by Hall.
The genetic difference between ellenbecki and S. afra is 2%, and 5% between ellenbecki
and psilolaema. Moreover, S. afra and ellenbecki indeed have morphological affinities,
i.e. they have similar underparts with freckled throat and barred bellies. So, in this
instance Hall appears to have erred in her decision. One other controversial aspect is the
taxonomic status of psilolaema. Hall considered this a species, whereas MackworthPraed and Grant considered this taxon a subspecies within S. afra. Genetically,
psilolaema differs from S. afra by 6% and the two have no apparent phylogenetic
affinity.
187
Within the Scleroptila levaillantoides, the genetic distance observed between the
three northern subspecies that Hall (1963) attributed to levaillantoides (lorti, archeri,
gutturalis) is in the range of ~3.1-3.4% to S. levaillantoides and ~3-4% to S. shelleyi.
Among themselves they differ by <2%. These three subspecies are phylogenetically
related with archeri and lorti being sisters (results not shown on the cladogram) and
they are in turn sister to gutturalis. Since uluensis is genetically closest to guttaralis,
both seem to have been incorrectly placed by Hall and warrant species status.
The southern jugularis shares sister relationship with whytei in all the analyses
(with 77% BS support only in the combined CYTB/organismal tree) except in the
organismal tree and it differs from the other southern subspecies (levaillantoides,
pallidior), the northern (gutturalis) subspecies, as well as shelleyi by ~7% sequence
divergence. It is closest to the nominate levaillantoides, only in the organismal tree but
differs by 7% of genetic distance. Despite having a well-developed gorget, jugularis has
some similarities with pallidior in that it has a white to buffy white belly with dark
chestnut blotches. Pallidior differs by ~3-4% from levaillantoides, gutturalis and S.
shelleyi. It has the closest phylogenetic affinities with S. afra and ellenbecki (though not
supported) and the two taxa differ genetically by 2% and 3% respectively. Since
pallidior has little morphological resemblance to S. afra and ellenbecki it is still
tentatively placed with levaillantoides, subject to further phylogeographic study.
Scleroptila levaillantii differs genetically from kikuyuensis and crawshayi by 9
and 8% respectively, and a 3% difference exists between kikuyuensis and crawshayi.
Scleroptila shelleyi remains a valid species including the Natal subspecies sequestris
Clancey, 1960 and the Inhambane, southern Mozambique subspecies canidorsalis
(Lawson, 1963). The Malawi subspecies whytei should be recognized as a separate
188
species based on the remarkable difference observed in morphology, genetic distance
and distinct phylogenetic placement and geographic isolation, and hence be Scleroptila
whytei Neumann, 1908. The phylogenetic position of uluensis does not favour
Mackworth-Praed and Grant’s classification system, that is, uluensis is not
phylogenetically related to S. afra even though the two are in the same clade in the
parsimony CYTB/organismal tree. They have a genetic distance of 4%. Despite that the
morphological divergence and the geographic isolation of the distribution ranges of the
uluensis and S. afra are difficult to reconcile. Therefore, uluensis has to be recognized
as an evolutionarily distinct taxon and is upgraded to species Scleroptila uluensis
(Grant, 1892) with macarthuri (van Someren, 1938) being included as a subspecies
(Appendix 5.2).
Mackworth-Praed and Grant’s classification of psilolaema as a subspecies of S.
afra is probably not correct. The two are phylogenetically and geographically apart,
have a genetic difference of 6% and differ morphologically except for the presence of
the grey flecks on the throat of S. afra and indistinct black spots on the throat of
psilolaema. Thus, Hall (1963) was probably correct in recognizing S. psilolaema as a
valid species, Scleroptila psilolaema Gray 1867.
The taxonomic status of theresae is also uncertain. Mackworth-Praed and Grant
considered it a subspecies of S. shelleyi whereas Hall included it in S. psilolaema.
Theresae is sister to psilolaema and the two differ genetically by 3% (and 6% when
compared to S. shelleyi). Also, theresae is morphologically similar to psilolaema and
not to S. shelleyi. It has rich chestnut in wings and differs from S. shelleyi in having
some barring on the tips of the primaries and the underparts are rich buff mottled with
chestnut and with some less defined brown to black bars. The gorget is poorly-defined
189
and replaced by rich chestnut with black spots and has lighter chestnut throughout the
belly. Therefore, theresae should be considered a subspecies, Scleroptila psilolaema
theresae (Meinertzhagen, 1936) with elgonensis (Grant, 1891) being synonymized with
it. The phylogenetic position of ellenbecki and the small genetic divergence between
ellenbecki and S. afra (2%) certainly confirm Mackworth-Praed and Grant’s
classification system. Hall included ellenbecki in S. psilolaema (which differ genetically
by 5%). Careful examination of the specimens revealed that S. afra and ellenbecki both
have similar underparts with freckled throats and barred bellies. The major difference is
that the width of barring on breast and belly is finer in S. afra and broader in ellenbecki.
Furthermore, ellenbecki has rich darker chestnut in wings (opposed to grey in S. afra)
and some barring at the tips of the primaries. It has a poorly-defined gorget (as opposed
to a well-developed gorget in S. afra) replaced by rich chestnut and has dark chestnut
throughout the belly different from the narrow black barring in S. afra. In consideration
of the similarity in morphology, the small genetic distance and the phylogenetic
affinities between ellenbecki and S. afra, and geographical isolation, it is recommended
that ellenbecki be considered a species closely related to S. afra and hence Scleroptila
ellenbecki Erlanger, 1905.
Gutturalis should be recognized as a separate taxon, i.e. as a species Scleroptila
gutturalis (Rüppell, 1835), and gutturalis takes priority over archeri Sclater, 1927 and
lorti Sharpe, 1897. Friedmanni Grant and Mackworth-Praed, 1934 and stantoni (Cave,
1940) should be synonymized with archeri and eritreae Zedlitz, 1910 with gutturalis.
Scleroptila levaillantoides levaillantoides Smith, 1836 remains the valid
nominate subspecies with ludwigi Neumann, 1920 and gariepensis Smith, 1843 being
included in it. This subspecies is strongly marked on the upper and lower belly and
190
there exists large genetic difference when compared to pallidior (4.2%). Pallidior
Neumann, 1908, wattii Macdonald, 1953, langi Roberts, 1932 and kalaharica Roberts,
1932 should be considered a single taxon, i.e. subspecies within S. levaillantoides based
on its morphological attributes (strongly marked on upper belly with a relatively
unmarked lower belly) irrespective of its phylogenetic position and the least genetic
difference when compared to S. afra. Scleroptila levaillantoides pallidior Neumann
1908 takes priority over the other names.
From the south, jugularis Büttikorfer, 1889 generally looks like the nominate
levaillantoides, but it is relatively unmarked on the lower belly, posses a well-developed
gorget, there is remarkable genetic difference between the two and they are wellsupported sister taxa (99%). So, jugularis should be recognized as a species, Scleroptila
jugularis Büttikorfer 1889. This name receives priority over cunenensis (Roberts, 1932)
and stresemanni Hoesch & Niethammer, 1940.
There are marked morphological difference between levaillantii and the other
subspecies, kikuyuensis, crawshayi, benguellensis, mulemae and clayi. Generally,
differences concern whether a taxon has either rich reddish chestnut colour in the belly
feathers, a broken or unbroken ochre collar, rufous band across nape (in clayi) and a
streak below the eye with less black spotting and some irregular black patches on the
belly in benguellensis. Among the subspecies for which there are DNA sequence data,
the genetic divergence is remarkable between levaillantii and the two sister subspecies
kikuyuensis (9%) and crawshayi (8%) even though they appear similar morphologically.
The recommendation is that kikuyuensis and crawshayi be considered as separate taxa
taking the status of a species i.e. with crawshayi - Ogilvie-Grant 1896 receiving priority
191
over kikuyuensis Ogilvie-Grant, 1897 and becoming the nominate subspecies and hence
the name Scleroptila crawshayi crawshayi Ogilvie-Grant, 1896.
Morphologically crawshayi and kikuyuensis are similar except that kikuyuensis
is without the stripe on the hind neck. Thus, benguellensis and mulemae are similar to
kikuyuensis. There is no genetic evidence for the recognition of clayi, benguellensis and
mulemae the decision is based on morphological characters and the locality of the taxa.
Since benguellensis Neumann, 1908, clayi White, 1944 and mulemae Ogilvie-Grant,
1903 are said to be similar to kikuyuensis these subspecies should be synonymized with
crawshayi due to the geographic proximity and not with kikuyuensis. Kikuyuensis is
isolated from crawshayi and has some morphological differences and therefore, it
should be considered a subspecies within S. crawshayi O. Grant, 1896, and named
Scleroptila crawshayi kikuyuensis Ogilvie-Grant, 1897.
Scleroptila finschi and S. streptophora are the two distinct species with no
geographic variation and their species status is not disputed (Appendix 5.2).
Conclusions
This study has attempted to comprehensively revise the systematics of francolins. There
is congruence in the topologies resulting from the various analyses, but some
discordance in certain areas of the trees, especially within the species groups, remains.
What is apparent is that the combined analyses yielded better phylogenetic results. One
of the major finding is that the Striated Group is not monophyletic as hypothesised by
Hall (1963) whereas the Red-tailed and Red-winged Groups were largely diagnosable.
The consistent phylogenetic placement of the two previously unplaced Asian species
Ortygornis pondicerianus and O. gularis as the closest relatives of the African
192
Ortygornis taxa strongly rejects the monophyletic status of the five Asian francolin
species as hypothesized by Forcina et al. (2012). Francolinus lathami has distinct
evolutionary trajectory, hence it is recommended that it is assigned it own genus
Afrocolinus gen. nov.
193
Tables and Figures
Table 5.1. Morpho-behavioural characters and character states scored for phylogenetic
analysis of francolins.
1. Crown patterning: unpatterned = 0; streaked = 1; mottled = 2
2. Chestnut collar: none = 0; chestnut = 1
3. Upper back patterning: spotted = 1; streaked = 2; barred or barred and streaked = 3
4. Back plumage patterning: streaked = 1; barred = 2; quail-like = 3
5. Throat base colour: white-buff = 1; rufous-chestnut = 2; black = 3; buff-rimmed with
rufous = 4
6. Throat flecking/freckling: none = 0; present = 1
7. Gorget: absent = 0; present, but poorly developed = 1; moderately developed = 2; well
developed = 3
8. Breast patterning: unpatterned = 0; barred = 1; streaked = 2; spotted = 3;
blotched/mottled = 4
9. Belly patterning: unpatterned = 0; spotted = 1; barred = 2; streaked = 3;
blotched/mottled = 4; streaked and barred = 5
10. Under tail patterning: unpatterned = 0; barred or barred and blotched = 1
11. Wing patterning: unpatterned = 0; blotched = 1
12. Wing base colour: greyish brown = 1; greyish brown with some red = 2; moderate
to well-developed red = 3
13. Side of the head patterning: unpatterned = 0; streaked = 1; striped
(eye and/or jaw – progresses to gorget) = 2
14. Leg colour: yellow = 1; red, reddish range = 2
15. Spur number: none/poorly developed = 0; one = 1
16. Spur length: < 1mm = 0; 1-10mm = 1; 11-20mm = 2
17. Wing length (mm): < 140 = 1; < 160 = 2; < 180 = 3
18. Bill length (mm) / Wing length: < .15 = 1; < .17 = 2; < .18 = 3; > .18 = 4
19. Tail length (mm) / Wing length: < .50 = 1; <.55 = 2; < .60 = 3; > .60 = 4
20. Sexual plumage dimorphism: 0 = none; 1 = slight; 2 = marked
21. Raucous advertisement call: absent = 0; type 1 = 1; type 2 = 2
22. Simple musical advertisement call: 0 = absent; slow = 1; fast = 2; very fast =3
23. Short complex musical advertisement call: absent = 0; slow ki-bee-tilli = 1; fast
ki-bee-tilli = 2
24. Long complex musical call: absent = 0; type 1 = 1; type 2 = 2
194
Table 5.2. A matrix of organismal character and character state scores used in phylogenetic
analysis of francolins.
Character no.
Taxon
Gallus gallus
Bambusicola thoracica
Francolinus francolinus
Francolinus pictus
Francolinus pintadeanus
Ortygornis gularis
Ortygornis pondicerianus
Ortygornis sephaena
Ortygornis grantii
Ortygornis r. rovuma
Ortygornis r. spilogaster
Afrocolinus l. lathami
Afrocolinus l. schubotzi
Peliperdix coqui
Peliperdix kasaicus
Peliperdix maharao
Peliperdix ruandae
Peliperdix stuhlmanni
Peliperdix c. vernayi
Peliperdix hubbardi
Peliperdix spinetorum
Peliperdix albogularis
Peliperdix a. buckleyi
Peliperdix dewittei
Peliperdix schlegelii
Scleroptila streptophora
Scleroptila finschi
Scleroptila levaillantii
Scleroptila c. crawshayi
Scleroptila c. kikuyuensis
Scleroptila levaillantoides
Scleroptila l. pallidior
Scleroptila jugularis
Scleroptila p. psilolaema
Scleroptila p. theresae
Scleroptila afra
Scleroptila ellenbecki
Scleroptila shelleyi
Scleroptila uluensis
Scleroptila gutturalis
Scleroptila whytei
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0
2
1
1
1
0
0
1
1
1
1
0
0
2
2
2
2
2
2
2
2
2
2
2
2
0
2
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
2
3
1
1
1
3
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
1
3
1
1
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
2
3
2
1
2
1
1
1
1
1
3
3
1
1
1
1
1
1
1
1
1
1
1
1
4
1
4
4
4
1
1
1
1
1
1
1
1
1
1
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
2
0
0
0
0
0
0
3
3
3
1
1
3
2
2
3
2
2
2
2
3
0
2
1
3
1
2
1
2
2
2
2
3
3
0
1
1
0
1
1
1
1
1
1
1
1
1
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
0
0
2
1
2
3
2
2
2
3
3
1
1
2
2
2
2
2
2
0
0
2
2
2
2
4
2
2
2
2
0
0
0
2
2
2
2
2
2
5
2
0
0
1
0
1
0
1
0
0
0
0
1
1
1
1
1
1
1
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
2
1
1
1
1
1
1
1
1
2
1
1
1
2
1
1
2
2
2
1
1
2
3
3
3
3
3
2
3
3
1
3
2
2
3
3
0
0
0
0
1
0
1
1
1
1
1
0
0
2
2
2
2
2
2
2
0
2
2
2
0
0
0
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
0
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
195
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
0
1
1
1
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
3
2
2
1
1
3
1
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
1
1
2
2
4
2
2
2
2
2
2
1
1
2
2
2
2
2
2
2
2
2
2
2
2
3
3
4
4
4
3
3
3
3
3
3
3
3
3
3
3
4
3
4
3
3
4
4
4
4
4
4
2
2
3
3
3
3
3
3
3
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
1
2
0
0
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
1
1
2
2
2
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
3
3
3
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
2
1
1
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
2
2
2
0
0
0
0
0
1
0
2
2
0
2
Table 5.3. Francolin taxa for which DNA sequences were generated. Acronyms. AMNH American Museum of Natural History, TM - Transvaal Museum, BM - British Museum,
Natural History Museum at Tring, SAM - Iziko Museums of Cape Town (Natural History),
PFIAO - Percy FitzPatrick Institute of African Ornithology, TMC - Timothy M. Crowe,
University of Cape Town, South Africa, GB - GenBank, ‘-’ - Unknown, Br. muscle - Breast
muscle, Pect. muscle - Pectoral muscle. The scientific names below are as recorded on
specimen labels.
Taxa name
Sample no.
Origin
Date coll.
Sample type
AMNH DOT8023
AMNH 776813
GB EU165707
India
India
China
-
Br. muscle
Toe-pad
-
TMC 9
TMC 10
TM 6410
BM 1902 1 20 300
TMC 11
Marico River, South Africa
South Africa
Zimbiti, Beira, Mozambique
Hulul, Ethiopia
Cameroon
2004
1910
1902
2005
Heart
Br. muscle
Toe-pad
Toe-pad
Br. muscle
BM 80 1 1 1066
BM 1965 M 2073
AMNH 68972
PFIAO 47
TMC 13
TM 27245
PFIAO 59
TMC 12
TM 13217
SAM 55386
SAM 57819
AMNH 541174
AMNH 541185
AMNH 192573
AMNH 542359
TM 78622
AMNH 406156
TM 23166
AMNH 308887
Kenya
Mount Kenya, Kenya
Kenya
Ayton Farm, South Africa
Kenya
Mzimba, Malawi
Eastern Cape, South Africa
Petrus Steyn, South Africa
Quickborn, Namibia
Botswana
Namibia
Ethiopia
Somalia
Ethiopia
Angola
Sterkspruit, South Africa
Kenya
Nyika Plateau, Malawi
Angola
1965
1949
2002
2002
1923
2005
1940
1941
Toe-pad
Toe-pad
Liver
Br. muscle
Toe-pad
Liver
Liver
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Br. muscle
Toe-pad
Toe-pad
Toe-pad
PFIAO 45
TM 25325
BM 1933 7 14 105
TM 17263
BM 1953 54 52
TM 52676
BM 1956 5 14 85
AMNH 261916
TM 23158
BM 1963 6 2
BM 1953 54 49
BM 19339 5 2
BM 1949 30 19
BM 1929 3 13 1
BM 1999 9 20 2
BM 1934 3 16 28
Settlers, South Africa
Luluabourg, Dem. Rep. Congo
Richmond, Natal, South Africa
Botswana
Luluaborg, Dem. Rep. Congo
Rwinkwaku, Rwanda
Mega, Ethiopia
Kidong Valley, Kenya
Mzimba, Malawi
Arusha, Tanzania
Luluaborg, Dem. Rep. Congo
Azare banchi, Nigeria
Mboro, Bahr-El-Ghazel, Chad
Farafeni, Gambia
Gambaga, Ghana
Ejura Ashanti Ghana
1939
1933
1953
1956
1934
1963
1953
1933
1949
1929
1999
1934
Pect. Muscle
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Spotted Group
F. francolinus
F. pictus
F. pintadeanus
Striated Group
Francolinus sephaena
F. sephaena
F. s. rovuma
F. s. grantii
F. streptophora
Red-winged Group
F. psilolaema psilolaema
F. p. theresae
F. p. ellenbecki
F. shelleyi
F. s. uluensis
F. s. whytei
F. afra
F. levaillantoides levaillantoides
F. l. pallidior
F. l. kalaharica
F. l. jugularis
F. l. gutturalis
F. l. lorti
F. l. archeri
F. l. cunenensis
F. levaillantii levaillantii
F. l. kikuyuensis
F. l. crawshayi
F. finschi
Red-tailed Group
F. coqui coqui
F. c. angolensis
F. c. campbelli
F. c. vernayi
F. c. lynesi
F. c. ruahdae
F. c. maharao
F. c. hubbardi
F. c. stuhlmanni
F. c. thikae
F. c. kasaicus
F. c. spinetorum
F. schlegelii
F. albogularis albogularis
F. a. gambagae
F. a. buckleyi
196
Taxa name
Sample no.
Origin
Date coll.
Sample type
F. a. meinertzhageni
BM 1957 35 13
Luacano, Angola
1957
Toe-pad
GB AM236893
BM 1949 30 17
AMNH DOT8050
GB U90649
Cameroon
Benengal Zandi, Sudan
India
India
1949
-
Blood
Toe-pad
Br. muscle
-
Ungrouped taxa
F. lathami lathami
F. l. schubotzi
F. pondicerianus
F. gularis
197
Table 5.4. DNA markers sequenced and primers used for PCR-amplification and sequencing
of preserved francolin tissues.
Primer name
Primer sequence (5’to 3’)
Reference
Cytochrome b
L14578
MH15364
ML15347
H15915
cta gga atc atc cta gcc cta ga
act cta cta ggg ttt ggc c
atc aca aac cta ttc tc
aac gca gtc atc tcc ggt tta caa gac
J.G. Groth (pers. comm.)
P. Beresford (pers. comm.)
P. Beresford (pers. comm.)
Edwards & Wilson (1990)
Table 5.5. DNA marker sequenced and primers used for PCR-amplification and sequencing
of museum toe-pads for francolins.
Primer name
Primer sequence (5’ to 3’)
Reference
Cytochrome-b
Francolin-specific primers
L14851 (General)
cct act tag gat cat tcg ccc t
Franc-H1
cag cag aca cyt cyc tyg cct tc
MH15145
aag aat gag gcg cca ttt gc
Kornegay et al. (1993)
R. Bowie
P. Beresford
Franc-L1
Franc-H2
tgcctcacaacccaaatcctcac
R. Bowie
agg agr agr att act cct gtg ttt cag g R. Bowie
Franc-L2
Franc-H3
gcc tca ttc tty ttc aty tgy atc ttc c
ggr tgg aat ggg att ttg tca gag
R. Bowie
R. Bowie
Franc-L3
Franc-H4
tcatcyractcygacaaaatccc
gar rgg gat tag rag gag gat
R. Bowie
R. Bowie
Franc-L4
Franc-H5
tat tcg cct ayg cya tcc twc gct c
gta gga rag kga tgc tat ttg gcc
R. Bowie
R. Bowie
Franc-L5
HB20 (General)
ctc atc ctc ctc cta atc cc
ttg gtt cac aag acc aat gtt
R. Bowie
J. Feinstein (pers. comm.)
198
Table 5.6. The inter- and intraspecific uncorrected ‘P’ distances calculated from CYTB.
Taxo n name
1 Franco linus franco linus
2 Franco linus pintadeanus
3 Ortygo rnis gularis
4 A fro co linus latham i
5 A fro co linus l. s chubo tzi
6 Franco linus pictus
7 Ortygo rnis po ndicerianus
8 Ortygo rnis s ephaena grantii
9 Ortygo rnis s ephaena (Kenya)
10 Ortygo rnis s . ro vum a
11 Ortygo rnis s ephaena (SA)
12 Ortygo rnis s . zam bes iae
13 P eliperdix albo gularis
14 P eliperdix a. buckleyi
15 P eliperdix a. gam bagae
16 P eliperdix a. m einertzhageni
17 P eliperdix co qui
18 P eliperdix c. ruahdae (Rwanda)
19 P eliperdix co qui (Zambia)
2 0 P eliperdix c. ango lens is
2 1 P eliperdix c. cam pbelii
2 2 P eliperdix hubbardi
2 3 P eliperdix c. kas aicus
2 4 P eliperdix c. lynes i
2 5 P eliperdix c. m aharao
2 6 P eliperdix c. s pineto rum
2 7 P eliperdix c. s tuhlm anni
2 8 P eliperdix c. thikae
2 9 P eliperdix c. vernayi
3 0 P eliperdix s chlegelii
3 1 S clero ptila afra
3 2 S clero ptila fins chi
3 3 S clero ptila levaillantii
3 4 S clero ptila l. craws hayi
3 5 S clero ptila l. kikuyuens is
3 6 S clero ptila levaillanto ides
3 7 S clero ptila l. kalaharica (Bo ts wana)
3 8 S clero ptila l. jugularis
3 9 S clero ptila l. archeri
4 0 S clero ptila l. gutturalis
4 1 S clero ptila l. lo rti
4 2 S clero ptila l. pallidio r
4 3 S clero ptila ps ilo laem a
4 4 S clero ptila p. theres ae
4 5 S clero ptila p. ellenbecki
4 6 S clero ptila s helleyi uluens is
4 7 S clero ptila s . whytei
4 8 S clero ptila s helleyi (SA)
4 9 S clero ptila s trepto pho ra
1
2
3
4
5
6
7
8
9 10
11 12 13 14
15 16
17 18 19 2 0 2 1 2 2 2 3 2 4
25 26 27 28 29 30 31 32 33 34 35 36 37 38 #
# 41 42 # 44 45 46
47 48 49
11%
9% 12%
9% 12% 9%
10% 12% 9% 1%
7% 10% 9% 9% 9%
9% 13% 5% 10% 10% 8%
10% 14% 9% 10% 11% 9% 9%
10% 14% 9% 10% 11% 9% 9% 1%
10% 14% 9% 11% 12% 10% 10% 4% 4%
9% 13% 8% 9% 10% 11% 9% 5% 5% 6%
10% 12% 8% 9% 9% 11% 10% 5% 6% 4% 0%
10% 14% 12% 11% 11% 14% 12% 12% 12% 12% 11% 13%
11% 14% 12% 11% 11% 14% 12% 12% 12% 12% 11% 12% 1%
11% 13% 12% 11% 11% 14% 12% 12% 12% 12% 11% 12% 2% 0%
11% 14% 11% 10% 11% 10% 11% 11% 12% 12% 11% 10% 5% 5% 6%
10% 13% 10% 10% 11% 10% 11% 11% 11% 11% 10% 8% 8% 8% 9% 8%
11% 17% 15% 13% 14% 9% 13% 15% 14% 15% 12% 9% 8% 7% 7% 10% 4%
11% 14% 11% 11% 11% 11% 11% 11% 12% 12% 11% 7% 8% 8% 9% 8% 0% 4%
11% 14% 11% 11% 11% 11% 12% 12% 12% 12% 11% 8% 8% 8% 9% 8% 0% 4% 0%
10% 14% 10% 10% 10% 10% 11% 11% 11% 11% 10% 8% 8% 7% 8% 8% 2% 4% 2% 2%
10% 13% 11% 10% 10% 9% 11% 11% 11% 12% 11% 10% 8% 8% 9% 8% 5% 5% 5% 5% 6%
10% 11% 10% 10% 10% 10% 11% 9% 10% 10% 11% 10% 9% 9% 7% 8% 7% 2% 8% 8% 6% 9%
11% 14% 11% 11% 11% 11% 11% 12% 12% 12% 11% 7% 9% 8% 9% 8% 1% 4% 0% 0% 2% 5% 8%
10% 13% 11% 10% 11% 9% 11% 12% 12% 13% 12% 11% 10% 10% 10% 9% 6% 11% 6% 6% 8% 5% 9% 6%
11% 14% 11% 11% 11% 10% 11% 11% 12% 12% 10% 11% 9% 9% 9% 10% 8% 7% 8% 8% 8% 6% 10% 8%
7%
10% 11% 10% 10% 10% 10% 10% 10% 10% 10% 10% 9% 8% 8% 7% 8% 5% 9% 6% 5% 7% 6% 5% 6%
8% 8%
10% 12% 10% 10% 11% 9% 11% 10% 11% 12% 10% 10% 9% 8% 9% 8% 6% 9% 6% 6% 6% 3% 7% 6%
5%
7% 7%
11% 17% 13% 12% 13% 9% 12% 14% 14% 15% 11% 9% 7% 6% 7% 9% 3% 1% 3% 3% 3% 4% 0% 3% 10% 6% 7% 8%
10% 13% 10% 10% 10% 8% 11% 10% 11% 12% 10% 9% 8% 8% 7% 8% 9% 8% 9% 9% 8% 8% 7% 9% 9% 8% 8% 9% 8%
9% 12% 10% 11% 11% 9% 10% 10% 11% 10% 10% 9% 9% 10% 8% 10% 9% 13% 10% 10% 9% 10% 5% 10% 10% 10% 5% 10% 12% 9%
10% 13% 10% 11% 11% 9% 10% 11% 11% 11% 11% 10% 11% 11% 10% 10% 10% 13% 10% 11% 10% 10% 8% 10% 11% 10% 8% 10% 12% 11% 6%
9% 13% 10% 10% 10% 9% 11% 10% 10% 10% 10% 10% 10% 10% 10% 10% 9% 13% 9% 9% 9% 9% 9% 9% 10% 11% 8% 10% 12% 11% 7% 8%
9% 12% 10% 11% 11% 7% 11% 11% 12% 11% 11% 11% 10% 11% 9% 10% 10% 12% 11% 11% 11% 10% 5% 11% 11% 12% 6% 10% 11% 9% 2% 6% 8%
10% 14% 11% 11% 11% 8% 11% 11% 12% 11% 11% 10% 11% 11% 11% 11% 9% 12% 10% 10% 10% 9% 8% 9% 12% 11% 8% 9% 11% 10% 5% 7% 9% 2%
10% 12% 11% 11% 10% 8% 11% 11% 11% 11% 11% 8% 10% 10% 9% 10% 9% 11% 10% 10% 9% 9% 7% 10% 10% 10% 6% 10% 10% 9% 3% 5% 8% 5% 5%
11% 15% 14% 15% 17% 9% 12% 15% 15% 16% 14% 9% 13% 12% 10% 12% 12% 9% 12% 12% 12% 11% 5% 12% 14% 11% 12% 12% 9% 14% 7% 10% 13% 7% 8% 4%
10% 13% 11% 11% 11% 9% 10% 9% 8% 11% 6% 9% 10% 11% 12% 12% 12% 11% 12% 12% 11% 11% 12% 12% 12% 12% 11% 10% 11% 12% 8% 10% 10% 8% 4% 7% 5%
9% 11% 9% 10% 9% 4% 9% 10% 11% 11% 10% 9% 10% 9% 8% 9% 9% 11% 9% 10% 9% 9% 5% 10% 10% 9% 6% 9% 10% 8% 4% 6% 7% 5% 5% 3% 6% 7%
10% 11% 9% 10% 10% 8% 9% 11% 11% 11% 11% 8% 11% 10% 9% 10% 10% 11% 10% 11% 10% 10% 6% 10% 10% 10% 6% 10% 11% 9% 4% 6% 7% 5% 6% 3% 6% 8% 1%
10% 11% 9% 10% 10% 8% 10% 11% 11% 11% 11% 9% 10% 10% 9% 9% 10% 12% 10% 10% 9% 10% 6% 10% 10% 10% 6% 9% 10% 9% 4% 6% 7% 4% 6% 3% 7% 8% 1% 2%
9% 12% 10% 11% 11% 8% 11% 11% 11% 11% 10% 9% 9% 9% 9% 10% 9% 12% 9% 9% 9% 10% 5% 9% 11% 10% 6% 10% 11% 10% 2% 7% 7% 3% 4% 4% 7% 8% 4% 4% 4%
10% 12% 10% 11% 10% 10% 11% 11% 11% 11% 11% 10% 11% 11% 11% 11% 11% 14% 12% 12% 12% 11% 9% 12% 11% 10% 9% 11% 13% 11% 6% 6% 8% 6% 7% 6% 9% 9% 5% 5% 6% 5%
11% 12% 10% 10% 10% 10% 10% 11% 11% 11% 11% 10% 11% 11% 11% 11% 11% 13% 11% 11% 11% 10% 9% 11% 10% 10% 8% 11% 12% 11% 6% 6% 8% 7% 8% 6% 9% 9% 6% 6% 6% 7% 3%
10% 13% 10% 12% 12% 9% 11% 10% 11% 10% 10% 9% 10% 10% 10% 11% 10% 12% 10% 10% 10% 10% 8% 10% 10% 11% 7% 9% 12% 11% 2% 5% 7% 3% 5% 3% 5% 7% 4% 4% 3% 3% 5% 6%
10% 12% 9% 10% 11% 9% 10% 10% 11% 10% 10% 9% 10% 11% 9% 10% 10% 14% 10% 10% 10% 10% 6% 10% 11% 11% 6% 10% 13% 9% 4% 6% 8% 6% 7% 4% 9% 8% 3% 3% 3% 5% 6% 6% 4%
9% 12% 9% 10% 9% 5% 10% 10% 10% 11% 10% 10% 10% 10% 10% 10% 11% 14% 11% 11% 10% 10% 8% 11% 10% 10% 8% 10% 13% 9% 6% 5% 8% 7% 6% 6% 9% 6% 5% 6% 6% 7% 5% 6% 6% 6%
10% 12% 10% 10% 10% 9% 10% 10% 11% 11% 11% 10% 10% 10% 9% 9% 9% 11% 10% 10% 9% 9% 7% 10% 10% 10% 7% 10% 10% 9% 3% 5% 7% 5% 6% 2% 7% 7% 4% 4% 3% 4% 6% 6% 4% 4%
6%
10% 13% 9% 10% 10% 9% 10% 11% 11% 11% 10% 9% 10% 11% 11% 10% 9% 12% 10% 10% 10% 11% 9% 10% 11% 10% 9% 11% 11% 10% 8% 8% 8% 9% 8% 8% 12% 10% 8% 8% 8% 8% 9% 8% 8% 8%
9% 8%
199
3
4
3
Striated Group
1. Ortygornis sephaena
2. O. rovuma rovuma
3. O. grantii
4. O. rovuma spilogaster
2
1
1
2
1
Figure 5.1. Distribution ranges of Hall’s Striated Group taxa.
200
1
1
1
2
2
3
Red-winged Group
1.
2.
3.
4.
5.
6.
Scleroptila gutturalis
S. streptophora
S. finschi
S. jugularis
S. levaillantoides pallidior
S. levaillantoides levaillantoides
4
3
5
6
Figure 5.2. Distribution ranges of Hall’s Red-winged Group taxa.
201
3
2
Red-winged Group (cont.)
1. Scleroptila levaillantii
3
2. S. crawshayi crawshayi
3. S. c. kikuyuensis
1
Figure 5.3. Distribution ranges of Hall’s Red-winged Group taxa.
202
6
7
5
4
Red-winged Group (concl.)
1.
2.
3.
4.
5.
6.
7.
Scleroptila afra
S. shelleyi
S. whytei
S. uluensis
S. psilolaema theresae
S. p. psilolaema
S. ellenbecki
2
3
2
2
1
Figure 5.4. Distribution ranges of Hall’s Red-winged Group taxa.
203
8
8
8
9
11
10
7
5
6
4
4
Red-tailed Group
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Peliperdix coqui coqui
P. c. vernayi
P. stuhlmanni
P. kasaicus
P. c. ruahdae
P. hubbardi
P. maharao
P. spinetorum
P. albogularis albogularis
P. a. buckleyi
P. schlegelii
P. dewittei
12
1
3
1
2
1
Figure 5.5. Distribution ranges of Hall’s Red-tailed Group taxa.
204
1
1
2
2
1
2
1. Afrocolinus lathami lathami
2. A. l. schubotzi
Figure 5.6. Distribution of Afrocolinus lathami.
205
Figure 5.7. Illustration of francolin body partitions scored to generate the plumage
character matrix. C - Crown, HN - Hindneck, M - Mantle, B - Back, TF - Tail feathers,
T - Throat, G - Gorget, BR - Breast, BE - Belly, UNTC - Undertail coverts, SOH - Side
of the head, W - Wing (primaries).
206
Francolinus pintadeanus
86
Francolinus pictus
Francolinus francolinus
S
P
G
Bambusicola thoracica
Afrocolinus lathami schubotzi
100
Afrocolinus lathami lathami
Ortygornis sephaena
82
Ortygornis rovuma rovuma
99
African
Ortygornis grantii
Ortygornis pondicerianus
Ortygornis gularis
100
Asian
Peliperdix hubbardi
Peliperdix maharao
69
Peliperdix spinetorum
88
93
Peliperdix coqui coqui
Peliperdix coqui vernayi
Peliperdix stuhlmanni
Peliperdix schlegelii
78
94
Peliperdix albogularis buckleyi
99
R
T
G
Peliperdix albogularis albogularis
Peliperdix dewittei
99
Peliperdix kasaicus
Scleroptila levaillantii
75
Scleroptila crawshayi kikuyuensis
Scleroptila crawshayi crawshayi
58
Scleroptila levaillantoides levaillantoides
Scleroptila shelleyi
Scleroptila finschi
52
Scleroptila jugularis
98
Scleroptila whytei
Scleroptila psilolaema theresae
Scleroptila psilolaema psilolaema
R
W
G
Scleroptila gutturalis
Scleroptila afra
Scleroptila levaillantoides pallidior
Scleroptila uluensis
Scleroptila ellenbecki
Scleroptila streptophora
Gallus gallus
Figure 5.8. Maximum likelihood tree obtained from mitochondrial Cytochrome-b characters.
Numbers above branches represent boostrap support values (only ≥ 70% are shown). SPG stands
for Spotted Group, RTG – Red-tailed Group, RWG – Red-winged Group.
207
Scleroptila streptophora
Scleroptila whytei
Scleroptila jugularis
Scleroptila finschi
Scleroptila psilolaema theresae
100
Scleroptila psilolaema psilolaema
Scleroptila uluensis
Scleroptila ellenbecki
Scleroptila crawshayi kikuyuensis
98
Scleroptila crawshayi crawshayi
Scleroptila levaillantoides pallidior
Scleroptila afra
Scleroptila shelleyi
72
Scleroptila levaillantoides levaillantoides
Scleroptila gutturalis
Scleroptila levaillantii
Peliperdix spinetorum
Peliperdix hubbardi
Peliperdix maharao
70
Peliperdix coqui vernayi
Peliperdix coqui coqui
R
T
G
Peliperdix stuhlmanni
Peliperdix kasaicus
Peliperdix schlegelii
Peliperdix dewittei
Peliperdix albogularis buckleyi
100
Peliperdix albogularis albogularis
Francolinus pintadeanus
S
Francolinus pictus
P
G
Francolinus francolinus
97
Ortygornis sephaena
99
99
100
Ortygornis rovuma rovuma
Ortygornis grantii
Ortygornis pondicerianus
Ortygornis gularis
African
Asian
Afrocolinus lathami lathami
Afrocolinus lathami schubotzi
Bambusicola thoracica
Gallus gallus
Figure 5.9. A parsimony tree (1 of 4 most parsimonious trees) obtained from mitochondrial
Cytochrome-b characters. Numbers above branches represent boostrap support values (only ≥
70% are presented). SPG stands for Spotted Group, RTG – Red-tailed Group, RWG – Redwinged Group.
208
R
W
G
Gallus gallus
Bambusicola thoracica
Francolinus francolinus
Francolinus pintadeanus
Francolinus pictus
S
P
G
Ortygornis sephaena
96
Ortygornis grantii
Ortygornis rovuma rovuma
Ortygornis rovuma spilogaster
Ortygornis pondicerianus
African
Asian
95
Afrocolinus lathami lathami
Afrocolinus lathami schubotzi
Peliperdix coqui coqui
Peliperdix ruahdae
Peliperdix kasaicus
Peliperdix coqui vernayi
Peliperdix maharao
Peliperdix stuhlmanni
Peliperdix hubbardi
Peliperdix spinetorum
R
T
G
Peliperdix albogularis albogularis
Peliperdix albogularis buckleyi
Peliperdix dewittei
Peliperdix schlegelii
Scleroptila streptophora
Scleroptila finschi
Scleroptila levaillantii
Scleroptila crawsha. crawshayi
Scleroptila craw. kikuyuensis
Scleroptila afra
94
74
70
Scleroptila levaillantoi. levaillantoides
Scleroptila levaillantoides pallidior
Scleroptila jugularis
Scleroptila psilolaema psilolaema
Scleroptila ellenbecki
Scleroptila psilolaema theresae
Scleroptila shelleyi
Scleroptila uluensis
Scleroptila gutturalis
Scleroptila whytei
Ortygornis gularis
Figure 5.10. A parsimony tree (1 of 397 most parsimonious trees) obtained from organismal
characters. Numbers above branches represent bootstrap support values (only ≥ 70% are presented).
SPG stands for Spotted Group, RTG – Red-tailed Group, RWG – Red-winged Group.
209
R
W
G
Gallus gallus
Bambusicola thoracica
Francolinus francolinus
Francolinus pictus
Francolinus pintadeanus
Peliperdix coqui coqui
Peliperdix coqui ruahdae
S
P
G
Peliperdix coqui vernayi
Peliperdix maharao
Peliperdix hubbardi
Peliperdix spinetorum
Peliperdix kasaicus
Peliperdix stuhlmanni
Peliperdix albogularis albogularis
100
Peliperdix albogularis buckleyi
Peliperdix dewittei
Peliperdix schlegelii
93
71
71
83
77
87
100
87
Scleroptila streptophora
Scleroptila levaillantii
Scleroptila finschi
Scleroptila jugularis
Scleroptila whytei
Scleroptila psilolaema psilolaema
Scleroptila psilolaema theresae
Scleroptila crawshayi crawshayi
100
Scleroptila crawshayi kikuyuensis
Scleroptila afra
Scleroptila uluensis
Scleroptila levaillantoides pallidior
Scleroptila ellenbecki
85
R
T
G
R
W
G
Scleroptila levaillantoides levaillantoides
Scleroptila shelleyi
Scleroptila gutturalis
Ortygornis gularis
Ortygornis pondicerianus
Ortygornis sephaena
Ortygornis grantii
100
73
97
83
Ortygornis rovuma rovuma
Ortygornis rovuma spilogaster
Afrocolinus lathami lathami
Asian
African
71
100
Afrocolinus lathami schubotzi
Figure 5.11. A parsimony tree (1 of 24 most parsimonious trees) obtained from combined
mitochondrial Cytochrome-b and organismal characters. Numbers above branches represent bootstrap
support values (only ≥ 70% are presented). SPG stands for Spotted Group, RTG – Red-tailed Group,
RWG – Red-winged Group.
210
Appendix 5.1. List of francolin taxa included in the analysis and the GenBank accession
number for CYTB.
CYTB
GenBank
no.
Taxon
Francolinus francolinus
AF013762
Francolinus pictus
Francolinus pintadeanus
FR694142
NC011817
Ortygornis pondicerianus
Ortygornis gularis
FR691632
U90649
Afrocolinus lathami lathami
Afrocolinus lathami schubotzi
AM236893
FR694139
Ortygornis grantii
Ortygornis sephaena Kenya
FR694144
FR694141
Ortygornis rovuma rovuma
Ortygornis sephaena South Africa
FR694135
FR694140
Ortygornis sephaena zambesiae
Peliperdix albogularis albogularis
FR694143
FR694145
Peliperdix a. buckleyi
Peliperdix a. gambagae
FR694147
FR694146
Peliperdix a. meinertzhageni
FR694148
Peliperdix coqui coqui
Peliperdix coqui angolensis
AM236895
FR694153
Peliperdix coqui campbelii
Peliperdix coqui hubbardi
FR694156
FR694151
Peliperdix coqui kasaicus
Peliperdix coqui lynesi
FR694150
FR694155
Peliperdix coqui maharao
Peliperdix coqui spinetorum
FR691635
FR694154
Peliperdix coqui stuhlmanni
Peliperdix coqui thikae
FR694152
FR691634
Peliperdix coqui vernayi
Peliperdix schlegelii
FR694157
FR694149
Scleroptila afra
Scleroptila finschi
AM236897
FR691607
Scleroptila levaillantii
Scleroptila crawshayi crawshayi
U90642
FR691605
Scleroptila c. kikuyuensis
Scleroptila levaillantoides
FR691606
FR691612
Scleroptila levaillantoides archeri
Scleroptila levaillantoides gutturalis
FR691610
FR691613
Scleroptila jugularis
Scleroptila levaillantoides lorti
FR691608
FR691611
Scleroptila levaillantoides pallidior
Scleroptila psilolaema
FR691609
FR691614
Scleroptila psilolaema ellenbecki
Scleroptila psilolaema theresae
FR691616
FR691615
211
Scleroptila uluensis Kenya
FR691620
Scleroptila shelleyi South Africa
Scleroptila uluensis
AM236898
FR691622
Scleroptila whytei
Scleroptila streptophora
FR691621
FR691617
212
Appendix 5.2. A revised classification of African francolins that is based on evidence
presented in this chapter. An asterisk (*) indicates that the taxonomic status of that
particular taxon may change when DNA characters become available and are included
in the analyses. Species and subspecies authority appear in Table 1.2.
Family: Phasianidae
Sub-family: Phasianinae
Genus: Francolinus Stephens, 1819
Francolinus francolinus
Francolinus pintadeanus
Francolinus pictus
Genus: Ortygornis Reichenbach, 1853
Ortygornis pondicerianus
Ortygornis gularis
Ortygornis sephaena
Ortygornis rovuma rovuma
Ortygornis rovuma spilogaster*
Ortygornis grantii
Genus: Afrocolinus gen. nov.
Afrocolinus lathami
Afrocolinus lathami schubotzi
Genus: Peliperdix Bonaparte, 1856
Peliperdix coqui coqui
Peliperdix coqui vernayi
Peliperdix stuhlmanni
Peliperdix kasaicus
Peliperdix coqui ruahdae
Peliperdix hubbardi
Peliperdix maharao
Peliperdix spinetorum
Peliperdix albogularis albogularis
Peliperdix albogularis buckleyi
Peliperdix schlegelii
Peliperdix dewittei
Genus: Scleroptila Blyth, 1849
Scleroptila levaillantoides levaillantoides
Scleroptila levaillantoides pallidior
Scleroptila jugularis
Scleroptila gutturalis
Scleroptila finschi
Scleroptila streptophora
Scleroptila levaillantii
Scleroptila crawshayi crawshayi
Scleroptila crawshayi kikuyuensis
213
Scleroptila afra
Scleroptila shelleyi
Scleroptila whytei
Scleroptila uluensis
Scleroptila psilolaema psilolaema
Scleroptila psilolaema theresae
Scleroptila ellenbecki
214
CHAPTER 6
Taxonomy and phylogeny of spurfowls
Abstract
Delimitation of taxa continues to be a challenging exercise in the absence of a universal
definition of species and subspecies that can be applied consistently across different
groups of study. The spurfowls, members of which share a common evolutionary path
(genus Pternistis Wagler, 1832, sensu Crowe et al. 2006), which are the group of focus
in this chapter, are no exception. Over and above the challenges resulting from
suspected high levels of hybridization, discordance between organismal and molecular
characters, difficulty in determining the genetic threshold when delimiting taxa, and
unavailability of fresh samples of many taxa for DNA analyses, this study sought to
establish a classification system of spurfowls which takes into account the evolutionary
relationships among taxa, with the goal of bringing stability to the taxa recognized as
valid species, subspecies and genera, based on congruent multiple lines of evidence.
The phylogenetic inference methods, Maximum likelihood and parsimony were
performed on separate and combined DNA and organismal characters. The outcome
resulted in some subspecies being elevated to the species level with others being placed
into more inclusive entities; no generic recommendations are made. Most of the
phylogenetic hypotheses presented by Hall (1963) were rejected. Spurfowls are
represented by extremely morphologically and vocally divergent taxa. Thus, it is
difficult to identify homologuous characters that can be used to investigate their
evolutionary relationships. However, a multi-faceted character approach seems to be a
215
fitting strategy for working out intra-generic relationships within spurfowls.
Phylogeographic studies of various species complexes are recommended.
216
Introduction
What are spurfowls?
The hypothesised monophyly of the genus Francolinus Stephens 1819 (sensu Hall 1963),
was rejected resulting in the split of birds traditionally known as ‘francolins’ into two
lineages, ‘true’ francolins and spurfowls, that are evolutionary distant (Crowe et al. 2006).
Spurfowls are represented by members currently assigned to a single genus Pternistis
Wagler 1832 (Table 1.1, 1.2). Spurfowls, like francolins are placed in the sub-family
Phasianinae, and together with other Old World partridge- and quail-like gamebirds (e.g.
Perdix and Coturnix spp.) are placed in the tribe Perdicini (Chapin 1932, Peters 1934,
Wolters 1975-82, Crowe et al. 1986, Johnsgard 1988, Sibley and Monroe 1990, del Hoyo
et al. 1994, Madge and McGowan 2002). This assemblage is comprised of 23
traditionally recognized species with their core distribution mainly in sub-Saharan Africa
and no species occur in Asia (Hall 1963, Johnsgard 1988). The West African Pternistis
bicalcaratus is the only species that extends its distribution into Morocco, in North
Africa.
Spurfowl species and their distribution
This chapter focuses on spurfowls (Pternistis spp.), which Hall (1963) categorized into
four putative monophyletic species groups (see Table 1.1), all represented by African
species. These are the Bare-throated Group, Montane, Vermiculated, and the Scaly
Group. Generally, spurfowls like francolins are (partridge-like) resident birds that can run
or fly over short distances when facing a threat. Most species are sexually monochromatic
in plumage, but males of most species have a single spur on their tarsus while females in
some instances also have relatively smaller spurs (Johnsgard 1988). Members of the
217
different species groups have complex morphology, ecology, behaviour and distribution
patterns (Snow 1978) and occur in different habitats primarily of a tropical or sub-tropical
nature (Hall 1963, Johnsgard 1988, del Hoyo et al. 1994, Madge and McGowan 2002).
Species group diversity, distribution and morphology
Bare-throated Group
The Bare-throated Group consists of four traditionally recognized species, the Rednecked Spurfowl Pternistis afer (Müller, 1776), Swainson’s Spurfowl P. swainsonii
(Smith, 1836), Yellow-necked Spurfowl P. leucoscepus (Gray, 1867) and Grey-breasted
Spurfowl P. rufopictus Reichenow, 1887 (Table 1.1). A number of subspecies are
traditionally recognized, 20 in P. afer (afer (Müller, 1776), harterti Reichenow, 1909,
nyanzae Conover, 1929, böhmi Reichenow, 1885, intercedens Reichenow, 1909, itigi
(Bowen, 1930), cranchii
(Leach & Koenig, 1818), punctulatus (Gray, 1834),
benguellensis Bocage, 1893, leucoparaeus (Fischer & Reichenow, 1884), humboldtii
(Peters, 1854), swynnertoni Sclater, 1921, castaneiventer Gunning & Roberts, 1911,
melanogaster Neumann, 1898, loangwae Grant & Praed, 1934, lehmanni Roberts, 1931,
notatus Roberts 1924, krebsi Neumann, 1920, cunenensis Roberts, 1932, cooperi
Roberts, 1947); five in P. swainsonii (swainsonii (Smith, 1836)), lundazi White, 1947,
gilli Roberts, 1932, damarensis Roberts, 1932, chobiensis Roberts, 1932); seven in P.
leucoscepus (leucoscepus (Gray, 1867)), infuscatus Cabanis, 1868, holtemülleri
Erlanger, 1904, keniensis Mearns, 1911, kilimensis Mearns, 1911, tokora Stoneham,
1930, muhamed-ben-abdullah Erlanger, 1904) and a monotypic P. rufopictus (Table
1.2).
218
These taxa have widespread distribution in eastern Africa from Eritrea to the
Cape Province and extend westward south of the Congo forest to Gabon, Angola and
northern Namibia (Hall 1963) (Fig. 6.1, 6.2). They inhabit grasslands with cover of
trees and scrubs and they are found near water courses. Members of the Bare-throated
Group are large with a body mass ranging from 480 – 960 g (Johnsgard 1988). These
birds are distinguished from the other spurfowls by having a patch of bare skin on the
throat and round the eye, with the underparts being diagnostic in all four species. The
presence of fairly well-developed spurs is a remarkable feature (probably linked to the
origin of the group name ‘spurfowls’). The males have a long and strong lower spur
with a short blunt upper spur being common in P. leucoscepus and P. rufopictus, and
less common among individuals of P. afer and P. swainsonii.
The “afer” type subspecies have no vermiculation and they are strongly
patterned in black, white and grey on the belly and face. The belly of P. afer is broadly
streaked in white and black, with feathers having black centres and white edges. In
short, what characterizes what Hall (1963) refers to as the cranchii-type subspecies is
that the belly feathers have buff central streak vermiculated with blackish grey and
margined with broad chestnut (degree of chestnut colour varies from one subspecies to
the other) whereas the afer-type subspecies have broad greyish black central streak with
buff margins (particularly in the nominate subspecies afer) or with thin greyish black
central streak separating the long buff parallel streaks margined with black or
sometimes maroon. The feathers on the belly of P. swainsonii have drop-shaped dark
brown shaft streaks which increase in size over the lower belly and flanks. Pternistis
leucoscepus is the most distinctive member of the group with a yellow bare throat skin
and it presents slight variation compared to P. afer. The back is diagnosed by white
219
shaft streaks and the underparts are streaked with white and chestnut in a way that each
feather has chestnut with narrow white edges and a triangular white patch at the tip,
tapering up the shaft. The primaries are said to have a conspicuous patch which
according to Hall (1963) is visible during flight. Unlike the other Bare-throated species,
P. rufopictus has orange pink throat skin, grey with black shaft streaks breast with the
belly belly having the narrow central black streak separated from rufous chestnut
margins by broad buff to white streaks.
Montane Group
This group consists of seven traditionally species: Erckel’s Spurfowl P. erckelii
(Rüppell, 1835), Djibouti Spurfowl P. ochropectus (Dorst & Jouanin, 1952), Chestnutnaped Spurfowl P. castaneicollis Salvadori, 1888, Jackson’s Spurfowl P. jacksoni O.
Grant, 1891, Handsome Spurfowl P. nobilis Reichenow, 1908, Mount Cameroon
Spurfowl P. camerunensis Alexander, 1909, Swierstra’s Spurfowl P. swierstrai
(Roberts, 1929) (Table 1.1). The putative subspecies are: two in P. erckelii (erckelii
(Rüppell, 1835)), pentoni Praed, 1920, two in P. nobilis (nobilis Reichenow, 1908,
chapini Grant & Praed, 1934); six within P. castaneicollis (castaneicollis Salvadori,
1888, bottegi Salvadori, 1898, gofanus Neumann, 1904, ogoensis Praed, 1920, kaffanus
Grant & Praed, 1934, atrifrons (Conover, 1930)); three in P. jacksoni (jacksoni O.
Grant, 1891, pollenorum Meinertzhagen, 1937, gurae Bowen, 1931) (Table 1.2).
In terms of its distribution, the Montane Group is formed of scattered
communities in the mountains of north eastern Africa from Eritrea to Mount Kenya, on
the eastern Congo border, in the highlands of Angola, and on mount Cameroon (Hall
1963) (Fig. 6.3). Its members are birds which require some trees for cover and roosting
220
and the majority of species are in or near montane evergreen forest. Pternistis erckelii is
the only member which sometimes wanders into grasslands of the high plateau of
northern Ethiopia. The Montane species are said to be the least homogeneous (Snow
1978) of the spurfowl groups morphologically, so much so that, it is impossible to
designate any group character other than that the males have the crown, lower back,
primaries and tail plain brown or red-brown. In case where the sexes are dichromatic,
the lower back and tail are vermiculated.
Hall (1963) acknowledged that variation in some characters follows
geographical trends (Meiri and Dayan 2003, James 1970), with the birds of the extreme
north eastern parts being the largest and most heavily spurred with dark bills, yellowish
legs, no bare skin round the eyes, and with the sexes alike. Birds of the two isolated
western subspecies and the one central subspecies are the smallest with a mass that
according to Johnsgard (1988) ranges from ~500-600 g in P. camerunensis and P.
swierstrai and 600-890 g in P. nobilis. They are the least heavily spurred with red bills
and legs, and with the sexes quite unlike (Snow 1978) except for P. nobilis that exhibits
no sexual dimorphism. The Cameroon species, P. camerunensis has an extensive area of
red bare skin round the eye.
Generally, the diagnostic characters of the Montane spurfowls are found mainly
on the back and to a larger extent on the underparts particularly on the belly in the midventral feathers. The patterning of the belly feathers of P. castaneicollis, which has an
extensive double-patterning on the back with wing coverts and breast being clearly
defined in black and white, with some ochre and chestnut) is generally similar to that in
P. erckelii and P. swierstrai with the feathers being made up of a broad buff central
streak which is constricted in the middle and expanded distally into a tear drop distally,
221
margined with rufous. The central buff streak in P. ochropectus is margined with
greyish black U-shaped streak. The feathers on the belly of P. camerunensis, P. nobilis
and P. jacksoni are uniformly coloured with buff margins in P. jacksoni and P.
camerunensis, and grey margins in P. nobilis.
Scaly Group
The Scaly Group comprises three traditionally recognized allopatric species, Ahanta
Spurfowl P. ahantensis Temminck, 1854, Scaly Spurfowl P. squamatus Cassin, 1857
and Grey-striped Spurfowl P. griseostriatus O. Grant, 1890 (Table 1.1), with nine
putative subspecies recognized within P. squamatus (squamatus Cassin, 1857,
maranensis Mearns, 1910, schuetti Cabanis, 1880, usambarae Conover, 1928,
uzungwensis Bangs & Loveridge, 1931, doni Benson, 1939, zappeyi Mearns, 1911,
tetraoninus Blundell & Lovat, 1899, chyuluensis Someren, 1939); two subspecies in P.
ahantensis (ahantensis Temminck, 1854, hopkinsoni Bannerman, 1934) being
recognized while Pternistis griseostriatus is a monotypic species (Table 1.2).
Members of this group are characterized by having scaly underparts and they
inhabit the lowland forest in Upper and Lower Guinea, and cultivations and clearings
(Hall 1963) (Fig. 6.4). Compared to other spurfowls, these species have very little
defining plumage pattern and strong colouration except for P. ahantensis which shows
some distinct patterning with feathers having paler edges and darker centres. The
underparts have some parts that are creamy-buff or brown with the colour and pattern
varying in different subspecies, but all have narrow darker edges to breast feathers
giving a scaly appearance. There is no marked plumage dimorphism except that the
females tend to be more vermiculated than the males.
222
Pternistis squamatus is not sexually dimorphic but females tend to be less
strikingly vermiculated. This species is characterized by having indistinctly
vermiculated upperparts with faint U-patterning on the lower neck and the feathers have
blackish centres tinged with red-brown. The belly is plain brown with scaly pattern and
ill-defined dark shaft streaks margined buff. In P. ahantensis, the rest of belly feathers
are richly coloured and streaked with dark brown chestnut edged buff while the
upperparts are vermiculated with some white U-patterning (indistinct on the back but
very distinct on the lower neck). On the other hand, P. griseostriatus is the most distinct
member of the group in that the lower neck feathers and wing coverts are chestnut and
broadly vermiculated. It has plain belly, the breast and flank feathers are chestnut and
edged with greyish or creamy buff.
Vermiculated Group
This is the most widespread of all African spurfowl species groups (Fig. 6.5, 6.6, 6.7)
consisting of nine traditionally recognized species (Hall 1963). The species are the
Double-spurred Spurfowl P. bicalcaratus (Linnaeus, 1766), Heuglin’s Spurfowl P.
icterorhynchus Heuglin, 1863, Clapperton’s spurfowl P. clappertoni (Children &
Vigors, 1826), Hildebrandt’s Spurfowl P. hildebrandti Cabanis, 1878, Natal Spurfowl
P. natalensis Smith, 1833, Hartlaub’s Spurfowl P. hartlaubi Bocage, 1869, Harwood’s
Spurfowl P. harwoodi Blundell & Lovat, 1899, Red-billed Spurfowl P. adspersus
Waterhouse, 1838 and Cape Spurfowl P. capensis (Gmelin, 1789) (Table 1.1). The
putative subspecies are four within P. hartlaubi (hartlaubi Bocage, 1869, crypticus
Stresemann, 1939, bradfieldi (Roberts, 1928), ovambensis (Roberts, 1928)); three in P.
adspersus (adspersus Waterhouse, 1838, kalahari de Schauensee, 1931, mesicus
223
Clancey, 1996); two in P. natalensis (natalensis Smith 1833, neavei Praed, 1920); six
subspecies in P. hildebrandti (hildebrandti Cabanis, 1878, altumi Fischer & Reichenow,
1884, helleri Mearns, 1915, fischeri Reichenow, 1887, johnstoni Shelley, 1894, grotei
Reichenow, 1919); five in P. bicalcaratus (bicalcaratus (Linnaeus, 1766), ogilviegrantii Bannerman, 1922, ayesha Hartert, 1917, adamauae Neumann, 1915, thornei
Ogilvie-Grant, 1902); four in P. icterorhynchus (icterorhynchus Heuglin, 1863,
dybowski Oustalet, 1892, ugandensis Neumann, 1907, emini Neumann, 1907); eight
subspecies in P. clappertoni (clappertoni (Children & Vigors, 1826), sharpii OgilvieGrant, 1892, heuglini Neumann, 1907, gedgii Ogilvie-Grant, 1891, nigrosquamatus
Neumann, 1902, konigseggi Madarasz, 1914, testis Neumann, 1928, cavei Macdonald,
1940) (Table 1.2). Pternistis capensis and P. harwoodi are monotypic species.
Members of the group are known to exhibit variation in their ecology, with the
northern vermiculated species (P. bicalcaratus, P. icterorhynchus, P. clappertoni and P.
harwoodi) forming a homogenous assemblage occupying grasslands and savanna
habitats (Hall 1963). Within the Vermiculated member species the underparts are
diagnostic. Pternistis bicalcaratus has buff underparts distinctly and heavily streaked
with blackish and chestnut with small arrow-shaped buff marks on most belly feathers.
In P. clappertoni, the upperparts are greyish brown above with barring in flight feathers.
It is buff below with black to brownish marks while the hind and lower neck have Upatterning as in P. icterorhynchus whose underparts are buff with dark brown markings.
The male P. harwoodi that was examined is grey speckled and finely barred with
blackish and buff above. The hind and lower neck, sides of face, and throat are speckled
with black and white. It has irregular double-V patterning on the belly which tends to be
scattered just on the lower extreme of the belly on the buff background. Pternistis
224
hartlaubi is one of the spurfowls that exhibit striking plumage dimorphism with the
difference in the known populations being mainly in the degree of colouration. The
male P. hartlaubi has buff throat with black streaks which continues to the sides of neck
through to the hind neck and to the breast. The belly of male is like the underparts of
female which are pale tawny. The underparts of the female are tawny chestnut while
males are darker and have broader black streaks on the belly. Pternistis hildebrandti
shows marked plumage dimorphism with the males of the nominate form hildebrandti
being greyish brown above with vermiculation, the hind and lower neck streaked black
with white margins, the underparts are marked and mottled with black and white almost
like in P. natalensis. The females have similar backs to the males but they have
different underparts which are rusty. In P. natalensis, the hind neck mottled black and
white, the back is highly vermiculated and it is greyish brown with variable black,
whitish and buffish markings. The rest of the belly is buff with the upper belly to midbelly being heavily patterned in black and buff. The patterning is concentrated on the
upper part of the belly with the extreme lower belly having no or few marks. Pternistis
adspersus is a different looking species with the nominate form adspersus consisting of
minute vermiculation on the upperparts and with distinct black and white barring on the
underparts and variably on the lower neck. It is characterized by having distinctive
uniform brown and white double V- or U-patterning on the back and on the belly while
the pattern on the throat is reduced to form some irregular black fleckings. The belly
patterning also has distinct white shaft streaks.
What do we know about their taxonomy and phylogeny?
225
Profound taxonomic disagreements and confusion as reflected in Table 1.2 also affect the
spurfowls and this disagreement is largely at the subspecies level and to a limited extent
at the species level. Hall’ (1963) monograph has made a significant contribution
(Johnsgard 1988, Sibley and Monroe 1990, del Hoyo et al. 1994) towards the
understanding and enhancement of our knowledge about the evolutionary relationships
and distribution ranges of these birds. The monophyletic status of genus Francolinus as
hypothesized in Hall (1963), comprises 37 traditionally recognized species which were
divided among the eight putatively monophyletic groups comprising ecologically similar
but largely allo- or parapatric species. Even though Hall (1963) did not partition genus
Francolinus, she suggested that if generic division were necessary, she would have two
major groups, one consisting of members that she referred to as belonging to among
others, the spurfowls, encompassing members of the Bare-throated, Montane, Scaly,
Vermiculated Group. Another group would be represented by the francolins (see Chapter
5).
Contrary to Roberts (1924), Peters (1934) and Mackworth-Praed and Grant (1952,
1962, 1970) and Hall (1963), Wolters (1975-82) is the one who assigned the genus
Pternistis (Table 1.2) to all taxa which are known as ‘spurfowls’ today. Peters (1934)
recognized the genera Francolinus and Pternistis, and like Roberts he recognized
Pternistis for the Bare-throated species only (Table 1.2). Mackworth-Praed and Grant
(1952, 1962, 1970) also recognized two genera, Francolinus and Pternistis, with
Pternistis being assigned to members of the Bare-throated Group only (Table 1.2).
Phylogenetic difficulties have been articulated by Crowe and Crowe (1985),
Crowe et al. (1986), Crowe et al. (1992), Bloomer and Crowe (1998), Crowe et al. (2006)
and most recently by Forcina et al. (2012), with regard to the monophyletic status of the
226
genus Francolinus, and in particular, the status of the different putative monophyletic
groups recognized by Hall (1963). It is important to note that different studies used
different types of characters (Table 1.4), focussed on different geographic areas of
francolin distribution (Table 1.3), and made use of different methods of phylogenetic
analysis. Furthermore, only the Montane Group was once recovered as monophyletic
(Crowe and Crowe 1985, Crowe et al. 1992) and the Bare-throated, Vermiculated and the
Scaly Groups were either found to be paraphyletic or polyphyletic (Crowe and Crowe
1985, Crowe et al. 1992, Bloomer and Crowe 1998). Based on these results, there is a
dear need to test the monophyly of the species groups of spurfowls as delineated by Hall
(1963) hypotheses.
What is taxonomy and how can we possibly study it?
Species and subspecies concept review
The species and subspecies concept review is as outlined as it pertains the spurfowls in
Chapter 5.
Taxonomic determinations made in this chapter
Most of the spurfowl species are polytypic, encompassing more than one putative
subspecies (see Table 1.2). The same criteria as outline in Chapter 5 are used here.
Objectives of the study:
To review the taxonomic status of spurfowls.
To re-assess the monophyletic status of the species groups as delineated by Hall
(1963).
227
To investigate the phylogenetic relationship between taxa within Hall’s (1963)
putative monophyletic groups.
To produce a revised classification of spurfowls that takes into account the
evolutionary relationships among delineated taxa.
Materials and methods
Data collection
Morpho-behavioural characters of spurfowls
Spurfowls were divided into discrete sections and scored for variation in plumage
patterning and colouration (Fig. 5.7). In total, 24 organismal characters reflecting
assessment of plumage/integument, colour/pattern, as well as measurements of certain
qualitative and quantitative structures (Table 6.1), and several vocal characters extracted
from Crowe at al. (1992) and Chapter 4 were scored (Table 6.2). Morphometric
characters representing bill-length, tarsus- and spur-length were obtained using a
Vernier Calliper. A stopped wing-rule and a normal ruler were used to measure wingand tail-length, respectively. Wing-length was measured with the wing chord flattened
and straightened for maximum accuracy.
Molecular characters
For within-group molecular analyses of spurfowls, 34 terminal taxa (including
the two outgroup taxa) were sampled (Tables 6.3) with respect to mitochondrial
Cytochrome-b (CYTB - 1143 base pairs- bp) characters. Primers used in sequencing
molecular markers are listed in Table 6.4 and 6.5. The 1143 bp long CYTB gene was
sequenced for all taxa and Genbank accession numbers are detailed in Appendix 6.1.
228
72% of DNA samples were toe-pads sub-sampled from museum skins. Due to the
degraded state of the DNA, the CYTB gene for the toe-pads was sequenced in multiple
fragments (six for each sample) using spurfowl-specific primers – Table 6.5).
Maps and mapping of distribution records of investigated taxa
The maps showing the distribution ranges of spurfowls (Fig. 6.1 – 6.7) were produced
following the same procedure as outlined in Chapter 5.
Data analyses
Phylogenetic analyses
Two phylogenetic inference methods, parsimony and maximum likelihood with
different optimality criteria were employed to generate phylogenetic hypotheses as
outlined for francolins in Chapter 5. As suggested by the results in Chapter 2 (Fig. 2.1,
2.2, 2.3, 2.5, 2.6), all the data matrices were rooted on Alectoris chukar and Coturnix
coturnix.
Genetic distances
Uncorrected pairwise distances (Table 6.6) were calculated in PAUP ver. 4.0b10
(Swofford 2002) and were transformed into percentages.
Results and Discussion
Separate versus combined phylogenetic analyses
229
The parsimony analysis for the CYTB data set (with 1143 bp characters, yielded 252
parsimony informative characters, 150 variable characters which are parsimony
uninformative and two trees with 995 steps), for the organismal data set (with 33
characters, yielded 33 parsimony informative characters and 397 trees with 192 steps),
for combined CYTB and organismal data set (with 1176 characters, yielded 283
parsimony informative characters, 148 variable characters which are parsimony
uninformative and two trees with 1191 steps).
The systematics of Hall’s species groups of spurfowls
Bare-throated Group
The Bare-throated Group consists of four traditionally recognized species, polytypic
Pternistis afer, P. swainsonii, P. leucoscepus, and a monotypic species P. rufopictus.
Pternistis afer is comprised of the following recognized subspecies: (afer, harterti,
nyanzae, böhmi, intercedens, itigi, cranchii, punctulatus, benguellensis, leucoparaeus,
humboldtii, swynnertoni, castaneiventer, melanogaster, loangwae, lehmanni, notatus,
krebsi, cunenensis, cooperi); P. swainsonii (swainsonii, lundazi, gilli, damarensis,
chobiensis); and P. leucoscepus (leucoscepus, infuscatus, holtemülleri, keniensis,
kilimensis, tokora, muhamed-ben-abdullah) (Table 1.2).
All four of the analyses recovered monophyly of the Bare-throated Group,
although with varying levels of support (Fig. 6.8, 6.10, 6.11). Pternistis nobilis joins the
Bare-throated Group in the organismal tree (Fig. 6.9). These results are consistent with
those of Bloomer and Crowe (1998), but contradict those of Crowe and Crowe (1985),
Crowe et al. (1992), who suggested the Bare-throated group to be paraphyletic.
230
Within the Bare-throated Group, there is a split resulting in P. afer, P. rufopictus
and P. cranchii grouping together in most analyses with moderate support (Fig. 6.8,
6.10, 6.11). The putative subspecies afer and humboldtii are sister to each other and
they are, in turn sister to P. rufopictus, with P. cranchii being basal in this clade.
Pternistis rufopictus is embedded within the afer and cranchii clade in the combined
CYTB/organismal, organismal and ML trees, where it is sister to P. afer. The southern
P. swainsonii is sister to the eastern P. leucoscepus of which the two subspecies, P. l.
leucoscepus and P. l. infuscatus share a sister relationship.
Pternistis afer complex
This is the most widespread and variable of all the species within the Bare-throated
Group and it has an extremely complex distribution (Fig. 6.1). Pternistis afer is mainly
diagnosed by having red bare throat skin, bill and legs. Hall (1963) maintained that this
species forms a complex which could be split into two types with intervening blocks of
hybridization, the “cranchii” and the “afer” type subspecies. The “cranchii” subspecies
includes all the subspecies of the southern Congo, northern Angola, Northern Zambia,
western Tanzania, Uganda and Lake Victoria shores. These are the subspecies
characterized by having vermiculated underparts, with sparse chestnut streaks on the
belly. Cranchii has a black and grey facial pattern and it shares the least genetic distance
with afer differing at 1% of sequence divergence (Table 6.1).
The “afer” subspecies forming the complex have no vermiculation and are
strongly patterned in black, white and some grey on the belly and face. In the nominate
afer subspecies the face is white and the belly is streaked in white and black, with
feathers having black centres and white edges. The black and white subspecies have a
231
broken distribution with the nominate subspecies afer being restricted to south western
Angola along the escarpment and the Cunene basin. The underparts of the Benguella,
southern Angola subspecies benguellensis has a mix of afer- and cranchii-type features
(black, white and chestnut streaks) due to hybridization between the afer- and cranchiitype subspecies. Among the South African subspecies, the face is black and the
underparts are streaked with black, white and maroon. The bird from north-central
Africa (humboldtii) (Fig. 6.1) is diagnosed by having white jaw feathers and black belly
patch. The feathers on the breast are mainly grey with black shaft streaks that contrast
with those of the lower belly which form a black patch and those of the flanks that are
streaked black and white. The least genetic distance exists between humboldtii and afer
differing at 2% of sequence divergence. In short, what characterizes Hall’s cranchiitype subspecies is that the belly feathers have a buff central streak vermiculated with
blackish grey and margined with broad chestnut (degree of chestnut colour varies from
one subspecies to the other). The belly feather of the afer-type subspecies have a broad
greyish black central streak with buff margins (particularly in the nominate subspecies
afer) or with a thin greyish black central streak separating the long buff parallel streaks
margined with black or sometimes maroon. In Mozambique, Tanzania and south Kenya
the afer-type subspecies have very broad black margins and a much narrower central
buff to white streak. A range of additional subspecies of intermediate phenotype have
been described where these three subspecies are para/sympatric, but they lack the
morphological cohesion necessary for recognition.
Generally, the nucleotide sequence divergences do not reveal clear cut
differences among the P. afer subspecies. Those from the Cunene area of western
Angola are 1% sequence divergent from cranchii, suggesting close genetic ties between
232
afer and cranchii, presumably due to hybridization. Indeed, the <1% genetic differences
between the subspecies, nyanzae, böhmi, itigi, intercedens and harterti and cranchii
indicate that these may be recently diverged lineages. The furthest distance that exists
between the cranchii-type is that with humboldtii at 3%, subsequently followed by 3%
to the South African afer.
The results point to a split of P. afer into what Hall (1963) referred to as the
“afer-type” and the “cranchii-type”. The afer-type lineage which include the nominate
subspecies afer, castaneiventer/nudicollis, lehmanni, notatus and krebsi and also a
group consisting of black and white subspecies that have a black belly patch as in
humboldtii (including swynnertoni, melanogaster, leucoparaeus and loangwae). The
nominate afer is disjunctly distributed from the South African afer-type allies. The
black and white subspecies have a broken distribution with the nominate afer being
restricted to the south-western part of Angola along the escarpment and the Cunene
basin. The core cranchii-type includes cranchii, nyanzae, harterti, intercedens, itigi and
böhmi, cunenensis, benguellensis and punctulatus. These subspecies are diagnosed by
having vermiculated underparts and sparse chestnut streaks on the belly (chestnut being
replaced with maroon in harterti, and chestnut and a tinge of black in punctulatus). The
three subspecies, böhmi, itigi and intercedens have in addition a mixture of white and
black tinge on the belly. The subspecies, cunenensis and benguellensis are included in
the cranchii-type because they seem to be intermediate between the cranchii- and the
afer-type subspecies and this is revealed in their morphological attributes, genetic
affinities and in how they relate with the other forms. It is recommended that Pternisti
afer (Müller, 1776) be broken into two subspecies, that is, Pternistis afer afer (Müller,
1776) which includes castaneiventer, nudicollis, lehmanni, notatus and krebsi and
233
another subspecies Pternistis afer humboldtii (Peters, 1854) which include swynnertoni,
melanogaster, leucoparaeus and loangwae (Appendix 6.2). There is a geographical
break between P. a. humboldtii (the black belly patched subspecies) and the south
distributed P. a. afer which extends to the north of the Limpopo River basin.
Pternistis rufopictus
Pternistis rufopictus is a monotypic species that occurs from the south eastern shores of
Lake Victoria to the Wembere in Tanzania along the Acacia belt (Fig. 6.2). Unlike the
other Bare-throated species, P. rufopictus has orange pink throat skin, grey brown back
plumage with dark vermiculation and shaft streaks, and has brown legs. The wing
coverts and the feathers on the back are edged with rufous chestnut; the breast is grey
with black shaft streaks with the lower belly being richly patterned. The belly feathers
have narrow central black streaks separated from rufous chestnut margins by broad buff
to white streaks. It is similar to the cranchii-type subspecies in western Tanzania except
that it has a partial buff to white jaw feathers (as in humboldtii), and has no
vermiculation. This taxon’s distribution range overlaps with that of P. leucoscepus in
the southern parts of its distribution. The least genetic distance exists between P.
rufopictus and cranchii differing at 2% of sequence divergence. It is an intriguing
species that is genetically very different from P. swainsonii and P. leucoscepus
(differing from both at 4%). Its phylogenetic placement (embedded within the
cranchii/afer clade) points to the possibility that it could be a hybrid between the
various taxa belonging to the Bare-throated Group. It is the largest species among the
Bare-throated members with a mass of 779-964 g (Johnsgard 1988), has orange pink
throat skin which could possibly be an indication of a mix between the red throat skin in
234
P. afer and a yellow throat skin in P. leucoscepus. Its morphologically (and the least
genetic distance) places it close to the cranchii-type subspecies but it is
phylogenetically close to the afer-type subspecies.
Our results provide new evidence that reveals the possibility of P. rufopictus
being of hybrid origin. This taxon shows morphological, phylogenetic, genetic affinities
and a sympatric distribution with the cranchii-type subspecies. Vocally, it sounds very
similar to P. leucoscepus (details in Chapter 4) except that its call is much faster.
Despite the potential of a hybrid origin, morphologically, Pternistis rufopictus
Reichenow, 1887 still maintains its distinctiveness and has diverged sufficiently from P.
leucoscepus, P. swainsonii and from some of the cranchii-type subspecies to be
considered a distinct species.
Pternistis swainsonii complex
Pternistis swainsonii is a polytypic south western African (Fig. 6.2) species diagnosed
by having red bare throat skin and black legs. Within P. swainsonii the underparts are
less diagnostic and the subspecies (swainsonii, lundazi, gilli, damarensis, chobiensis)
are characterized by having a narrow central greyish black streak separated from greyish
chestnut margins by broad buff grey vermiculated streaks. The general colour
swainsonii is grey brown and lacks the black and white streaking found on the
underparts of the nominate subspecies afer. The feathers on the belly have drop-shaped
dark brown shaft streaks which increase in size over the lower belly and flanks. The
differences among P. swainsonii subspecies are quantitative rather than qualitative such
that there are no remarkable plumage differences. Genetically, the smallest divergence
is between P. swainsonii and P. rufopictus at 4% sequence divergence, although
235
phylogenetic analyses suggest P. swainsonii is either the basal member of the Barethroated Group or sister to P. leucoscepus. Therefore, Pternistis swainsonii (Smith,
1836) maintains its species status while all the subspecies, lundazi White 1947, gilli
Roberts, 1932, damarensis Roberts, 1932 and chobiensis Roberts, 1932 are
synonymized with it based on priority.
Pternistis leucoscepus complex
This is the most distinctive member of the Bare-throated group comprising the north
eastern African (Fig. 6.2) subspecies, leucoscepus, infuscatus, holtemülleri, keniensis,
kilimensis, tokora, muhamed-ben-abdullah with yellow bare throat skin. These
subspecies have similar morphological patterns but differ in the amount of white and
chestnut in the feathers of the belly. The back plumage among the different subspecies
of P leucoscepus is diagnosed by white shaft streaks and the underparts are streaked
with white and chestnut in a way that each feather has chestnut with narrow white edges
and a triangular white patch at the tip, tapering up the shaft. The primaries have a
conspicuous white patch which is visible during flight (Hall 1963). The northern
subspecies infuscatus (differing at 1% sequence divergence from the nominate
subspecies) differs from leucoscepus in having more chestnut than white on the
underparts contrary to the dominant white over chestnut in leucoscepus. Genetically, P.
leucoscepus is close to P. swainsonii, differing by 5% sequence divergence.
The morphology and genetic distance point to two subspecies. It is
recommended that two subspecies should be recognized: Pternistis leucoscepus
leucoscepus (Gray, 1867) and Pternistis l. infuscatus Cabanis, 1868. The rest of the
236
subspecies (holtemülleri, keniensis, kilimensis, tokora, muhamed-ben-abdullah) should
be synonymized with P. l. infuscatus Cabanis, 1868 (Appendix 6.2).
Vermiculated Group
The Vermiculated Group comprises nine traditionally recognized species which are
Pternistis hartlaubi, P. bicalcaratus, P. icterorhynchus, P. clappertoni, P. hildebrandti,
P. natalensis, P. harwoodi, P. adspersus and P. capensis. The putative subspecies are as
follows: for P. hartlaubi (hartlaubi, crypticus, bradfieldi, ovambensis); P. adspersus
(adspersus, kalahari, mesicus); P. natalensis (natalensis, neavei); P. hildebrandti
(hildebrandti, altumi, helleri, fischeri, johnstoni, grotei); P. bicalcaratus (bicalcaratus),
ogilvie-grantii, ayesha, adamauae, thornei); P. icterorhynchus (icterorhynchus,
dybowski, ugandensis, emini) and P. clappertoni (clappertoni, sharpii, heuglini, gedgii,
nigrosquamatus, konigseggi, testis, cavei) (Table 1.2).
In comparing all the phylogenetic trees (Fig. 6.8, 6.9, 6.10, 6.11), the position of
P. hartlaubi is consistently basal to all the other spurfowl taxa in the parsimony trees
except in the organismal parsimony tree where it forms a sister relationship with the
Montane Group member P. camerunensis (Fig. 6.9). In the ML (Fig. 6.11) tree P.
hartlaubi is embedded within the Montane group with an unresolved branch. These
results are in agreement with those in Bloomer and Crowe (1998). Also Crowe et al.
(1992) hinted at the uncertain genetic relationships between P. hartlaubi and the other
spurfowls and found that it was morphometrically the most distinct spurfowl and
speculated that this taxon shared morphological similarities to the Old World quails.
Bloomer and Crowe (1998) concluded that it represents a very early offshoot from the
partridge-francolin lineage.
237
The other eight species split themselves geographically into the northern (P.
bicalcaratus, P. icterorhynchus, P. clappertoni, P. harwoodi) and southern (P. capensis,
P. natalensis, P, hildebrandti, P. adspersus) Vermiculated species in all the trees (Fig.
5.8, 5.9, 5.10, 5.11). The northern Vermiculated clade emerges to be sister to the Barethroated Group in the combined CYTB/organismal (Fig. 6.8) and organismal (Fig. 6.9)
tree except in the CYTB (Fig. 6.10) tree where it is only members of the P. clappertoni
complex and P. harwoodi that group with the Bare-throated Group while members of
the P. bicalcaratus complex and P. icterorhynchus form a paraphyletic clade at the
basal parts of the tree. The northern Vermiculated taxa are paraphyletic in the ML tree.
Surprisingly P. harwoodi is embedded within P. clappertoni complex forming a sister
relationship with P. c. sharpii with 90% and 99 BS in the combined CYTB/organismal
(Fig. 6.8), organismal (Fig. 6.9) and CYTB (Fig. 6.10) trees. Pternistis harwoodi and P.
clappertoni complex are outliers in the spurfowl clade in the ML tree, The sister
relationship between the northern Vemiculated and the Bare-throated clades as
consistently recovered in this study, contrasts the finding in Bloomer and Crowe (1998)
who found that the southern Vermiculated taxa and not the northern Vermiculated taxa
were consistently grouping with each other. This kind of relationship is supported in
only one analysis (CYTB – Fig. 6.10) in this study even though the Bare-throated clade
is part of a large clade which incorporates two species from the northern Vermiculated
clade and also the two Montane sister species (P. jacksoni and P. swierstrai) (CYTB –
Fig. 6.10). The southern Vermiculated taxa (P. capensis, P. natalensis, P, hildebrandti,
P. adspersus) form a monophyletic clade only in the combined CYTB/organismal (Fig.
5.8), CYTB (Fig. 6.10) and Fig. 6.11 tree (64%, 88%, 95% BS respectively). Pternistis
capensis is an outlier in the southern Vermiculated clade forming a paraphyletic
238
relationship with P. adspersus, P. natalensis and P. hildebrandti with P. natalensis
being closely related to the two sister subspecies, hildebrandti and fischeri (98% BS CYTB/organismal (Fig. 6.8 tree). Pternistis capensis emerges as a sister taxon to P.
adspersus in CYTB (Fig. 6.10) and ML (Fig. 6.11) tree, but this relationship received no
support. In the organismal (Fig. 6.9) tree, P. natalensis is sister to P. adspersus but with
no nodal support. It is very clear that the Vermiculated Group is polyphyletic just as
found in Crowe and Crowe (1985), Crowe et al. (1992) and Bloomer and Crowe (1998).
However, the relationships among taxa within the northern Vermiculated clade appear
consistent unlike among taxa within the southern Vermiculated clade; the position of P.
hartlaubi as a basal taxon seems consistent.
Pternistis bicalcaratus complex
Pternistis bicalcaratus is distributed from Senegambia to the Central African Republic,
except for one taxon, ayesha, that occurs in isolation in Morocco (Fig. 6.5). All the
recognized subspecies of P. bicalcaratus are similarly patterned above and below
differing in the degree of colouration, patterning, vermiculation and the size of the
arrow-shaped buff marks in the centre of the belly feathers. The nominate subspecies
bicalcaratus is paler with a more rufous crown, vermiculated with V-patterning and
slightly rufous on the lower neck. It has buff underparts distinctly and heavily streaked
with blackish and chestnut with small arrow-shaped buff marks on most belly feathers.
The heavily patterned ayesha (from Morocco – not mapped) looks very similar to
bicalcaratus, but is faintly vermiculated and slightly more rufous on the lower neck,
with small arrow-shaped buff marks on belly feathers. The darkest subspecies is
adamauae (with 2% sequence divergence from bicalcaratus) with very little rufous on
239
lower neck, and the underparts are more buff with extremely reduced chestnut and have
larger arrow-shaped buff marks in centre of belly feathers. The least genetic distance
exists between bicalcaratus and P. icterorhynchus differing at 3% of sequence
divergence. It is apparent that P. bicalcaratus represents a cline with regard to the
overall colouration of back plumage of the various subspecies and the patterning of
plumage on the belly even though these subspecies are not strikingly different. The
subspecies bicalcaratus is heavily patterned like thornei but the two differ in that
bicalcaratus is less dark compared to thornei. Adamauae and Ogilvie-granti are the
darkest subspecies but with less patterning.
The morphological and genetic evidence reveal that adamauae is distinct from
bicalcaratus suggesting that there may be two subspecies within the P. bicalcaratus
complex. This should be Pternistis bicalcaratus bicalcaratus (Linnaeus, 1766) which
should include ayesha, thornei and ogilvie-granti and Pternistis bicalcaratus adamauae
Neumann, 1915. It is possible that ogilvie-granti could well be made a synonym of P. b.
adamauae but this decision could not be taken in the absence of genetic evidence.
Pternistis clappertoni complex
This species has its distribution range covering Cameroon, Central African Republic,
Chad, Eritrea, Ethiopia, Mali, Mauritania, Niger, Nigeria, Sudan, and Uganda (Fig. 6.5).
Within P. clappertoni, the nominate subspecies clappertoni is greyish brown above with
barring on the flight feathers. This subspecies is buff below with black to brownish
marks. The hind and lower neck have U-patterning as in P. icterorhynchus. The
subspecies sharpii (which differs at 1% sequence divergence clappertoni) has marks on
the belly which look streakier than those in clappertoni in a buffy white background
240
with the breast being similarly V-patterned extending on to the back. The primaries are
barred as in the nominate subspecies. It is genetically close to P. harwoodi differing at
1% of sequence divergence.
The major plumage differences lie in the colour of the background on the belly
and the shape of the marks on the belly. There is very little genetic divergence between
the nominate clappertoni and sharpii and nigrosquamatus. Therefore, two subspecies
are recognized which are Pternistis clappertoni clappertoni (Children & Vigors, 1826)
and P. c. sharpii Ogilvie-Grant, 1892. Pternistis c. clappertoni should include heuglini,
gedgii and cavei while konigseggi, nigrosquamatus and testis should be synonymized
with P. c. sharpii.
Pternistis icterorhynchus complex
Pternistis icterorhynchus is distributed in the Central African Republic, Democratic
Republic of the Congo, Sudan, and Uganda (Fig. 6.5). It consists of four putative
subspecies, icterorhynchus, dybowski, ugandensis and emini mainly diagnosed by
having distinct U-patterning on the lower neck and more vermiculation on the back than
other vermiculated taxa. The underparts are buff with dark brown markings. It shares
the least genetic distance with bicalcaratus differing at 3% of sequence divergence.
There was no success in sequencing all the subspecies within P. icterorhynchus.
Morphologically, there are no striking differences between icterorhynchus and
emini except that emini is less patterned on the belly as opposed to icterorhynchus
which is extensively patterned whereas ugandensis has some chestnut on the flanks. In
the absence of genetic data, no subspecies are recognized.
241
Pternistis harwoodi
This is a poorly known species from the Blue Nile gorges (Fig. 6.5), the male bird
which was examined is grey speckled and finely barred with blackish and buff above.
The hind and lower neck, sides of face, and throat are speckled with black and white. It
has irregular double-V patterning on the belly which tends to be scattered just on the
lower extreme of the belly on the buff background. Pternistis harwoodi shares the least
genetic distance with sharpii differing at 1% of sequence divergence.
Phylogenetically, P. harwoodi and sharpii are sister taxa and also share a
sympatric distribution. The two are not similar morphologically except that they are
both barely patterned distally on the belly. This is a case where possibily with time
morphology and DNA will align. While Pternistis harwoodi Blundell & Lovat, 1899 is
recognized as a valid species, it could be concluded from the above evidence that P.
harwoodi could tentatively be a well-marked subspecies of P. clappertoni or a linking
form through sharpii.
Pternistis hildebrandti complex
Pternistis hildebrandti, a species distributed in central and western Kenya, extending
south through Tanzania to south eastern Zaire, northeastern Zambia and Malawi (Fig.
6.6), shows marked plumage dimorphism with the males of the nominate subspecies
hildebrandti being greyish brown above with vermiculation, the hind and lower neck
streaked black with white margins, the underparts are marked and mottled with black
and white almost like in P. natalensis. The females have similar backs to the males but
they have different underparts which are rusty. The subspecies johnstoni (differing at
1% of sequence divergence) is like hildebrandti, but differs in that the female has the
242
nape, hind neck and breast uniform with the rest of the plumage. It shares the least
genetic distance with P. natalensis differing at 1% of sequence divergence.
The genetic and morphological evidence support a split of P. hildebrandti into
two subspecies, that is Pternistis hildebrandti hildebrandti Cabanis 1878 and Pternistis
hildebrandti fischeri Reichenow, 1887. The subspecies, altumi, 1884 and helleri are
synonymized with P. h. hildebrandti, whereas johnstoni and grotei are synonymized
with P. h. fischeri.
Pternistis natalensis complex
Pternistis natalensis is a species complex with the putative subspecies natalensis and
neavei from south eastern Africa (Fig. 6.6). The hind neck is mottled black and white,
the back is highly vermiculated and it is greyish brown with variable black, whitish and
buffish markings. The rest of the belly is buff with the upper belly to mid-belly being
heavily patterned in black and buff. The patterning is concentrated on the upper part of
the belly with the extreme lower belly having no or few marks. Pternistis natalensis
shares the least genetic distance with P. hildebrandti differing at 1% of sequence
divergence. However, it is particularly close to altumi, fischeri and johnstoni differing at
1%. The morphological similarities between P. natalensis and male P. hildebrandti
could have either been a possibility for P. natalensis and P. hildebrandti interbreeding
in the southeast region or that P. natalensis and P. hildebrandti could have been the
same taxon that got broken up with some changing in appearance but not genetically.
Based on the above-mentioned evidence and the absence of genetic data for neavei and
thamnobium, the decision would be to recognize Pternistis natalensis Smith 1833 as a
species and synonymize neavei with it.
243
Pternistis adspersus complex
Pternistis adspersus consists of three putative subspecies (adspersus, kalahari, mesicus)
and it is distributed in Namibia and Botswana (Fig. 6.6). It is a very different looking
species with the overall plumage dominated by minute vermiculation on the upperparts
and with distinct black and white barring on the underparts and variably on the lower
neck. This species shares the least genetic distance with P. capensis differing at 4% of
sequence divergence. The subspecies, adspersus and kalahari are inseparable
morphologically. The specimens of mesicus were not examined and no description was
provided in the literature in which it was recognized. This study acknowledges that
Pternistis adspersus Waterhouse 1838 is a valid species, and make no taxonomic
decision on the status of other subspecies.
Pternistis capensis
Pternistis capensis is an undisputed monotypic species which is endemic to the
southwestern Cape of South Africa (Fig. 6.6). It is the largest vermiculated species
(males weighing ~900 g with females being ~600 g – Johnsgard 1988) and is endemic
to the south western part of South Africa (the Cape) (Fig. 6.6). It is characterized by
having distinctive uniform brown and white double V- or U-patterning on the back and
on the belly while the pattern on the throat is reduced to form some irregular black
flecking. The belly patterning also has distinct white shaft streaks. This species shares
the least genetic distance with P. adspersus at 4% sequence divergence.
Pternistis capensis (Gmelin, 1789) is acknowledged as a valid species due to its
morphological and genetic differentiation (4-5%) compared to P. natalensis, P.
adspersus and P. hildebrandti and also its phylogenetic position.
244
Pternistis hartlaubi complex
Pternistis hartlaubi occupies granite and sandstone outcrops surrounded by semi-desert
steppe and is confined to northern Namibia and extreme south western Angola (Fig.
6.7). It is comprised of the subspecies, hartlaubi, bradfieldi, crypticus and ovambensis.
Is one of the spurfowl species that exhibits striking plumage dimorphism with the
difference in the subspecies being mainly in the degree of colouration. The male
hartlaubi has a buff throat with black streaks that continues to the sides of neck through
to the hind neck and to the breast. The belly of male birds is like the underparts of the
females which is pale tawny. The sides of face and chin and belly of females is pale
tawny. It shares the least genetic distance with P. ochropectus differing at 9% of
sequence divergence.
There is limited genetic and morphological evidence for the separation of
hartlaubi and crypticus and therefore, only one taxon is recognized, Pternistis hartlaubi
Bocage 1869 and bradfieldi, crypticus and ovambensis are synonymized with it.
Nonetheless, P. hartlaubi is clearly an ancient African paleoendemic.
The Montane Group
The Montane Group encompasses seven species, P. erckelii, P. ochropectus, P.
castaneicollis, P. jacksoni, P. nobilis, P. camerunensis and P. swierstrai. The putative
subspecies are as follows: within Pternistis erckelii (erckelii, pentoni); P. nobilisis
(nobilis, chapini); P. castaneicollis (castaneicollis, bottegi, gofanus, ogoensis, kaffanus,
atrifrons; P. jacksoni (jacksoni, pollenorum, gurae) (Table 1.2).
The Montane Group is polyphyletic in all the analyses (Fig. 6.8, 6.9, 6.10, 6.11)
with all the Montane taxa appearing at the base of the tree. This is in contrast with
245
Crowe and Crowe (1985) and Crowe et al. (1992) wherein the monophyletic status was
certain. Despite the Montane Group being polyphyletic, the north eastern species (P.
erckelii, P. ochropectus, P. castaneicollis) form a monophyletic clade in all the analyses
(Fig. 6.8, 6.9, 6.10, 6.11). Pternistis ochropectus is consistently sister to P. erckelii, and
P. nobilis is consistently sister to P. camerunensis. The phylogenetic position of
Pternistis swierstrai and P. jacksoni is uncertain in all the trees. The evolutionary
history of the Montane spurfowls still needs further scrutiny in line with the challenge
that this study rejects Hall’s (1963) speculation that they form a monophyletic
assemblage. Further, she also postulated that P. swierstrai and P. camerunensis were
the first species to be isolated from the ancestral stocks of the other species. This study
consistently reveals that P. camerunensis and P. nobilis were isolated first.
Pternistis ochropectus
This is a monotypic species endemic to the evergreen forest of Djibouti (Fig. 6.3). A
closer analysis of the plumage patterning on the belly feathers of P. ochropectus, P.
erckelii and P. castaneicollis renders them almost similar. They have a broad buff
central streak constricted in the middle and expanded distally into a tear drop, margined
by a greyish black U-shaped streak. This species shares the least genetic distance with
P. erckelii differing at 3% of sequence divergence. This study acknowledges Pternistis
ochropectus (Dorst & Jouanin, 1952) to be a valid species.
Pternistis jacksoni complex
This is a large species endemic to Kenya (Fig. 6.3) with a body mass ranging from
~1130-1160 g (Johnsgard 1988). It encompasses three subspecies, jacksoni, pollenorum
246
and gurae. It has a buff throat and greyish lower neck with the proximal part of the
lower neck being similarly patterned to the rest of belly. The lower neck feathers are
chestnut-coloured edged with buffish to white, but the degree of chestnut and buff and
white varies. It shares the least genetic distance with P. griseostriatus with which it
differs at 5% of sequence divergence.
There is a possibility that there could be two entities (possibly subspecies)
within P. jacksoni i.e., jacksoni and pollenorum being one entity and gurae being
another entity based only on morphology. However, in the absence of molecular
evidence for pollenorum and gurae, our decision is speculation only and hence only
Pternistis jacksoni O. Grant, 1891 can be acknowledged as a valid species.
Pternistis nobilis complex
This species comprises two putative subspecies, nobilis and chapini and is distributed in
the mountain forests of the eastern Democratic Republic of the Congo, southwest
Uganda and borders between Burundi and Rwanda (Fig. 6.3). The head, uppertail
coverts and primaries are grey brown. The wing coverts and the lower neck are deep
maroon, with light grey scalloping on the lower neck. The throat colour is buff white
with the rest of the belly being chestnut with grey edges replacing the broad buff to
white edges found in the nominate jacksoni. It is genetically close to P. camerunensis
differing at 7% of sequence divergence.
The morphological difference observed between nobilis and chapini are
quantitative with regard to colouration and hence only one species is recognized in the
absence of molecular evidence even though the two subspecies are found on different
247
mountains. Pternistis nobilis Reichenow, 1908 remains a valid species and chapini
should be synonymized with it pending further investigation.
Pternistis erckelii complex
Pternistis erckelii occurs in the Ethiopian massif (Fig. 6.3) and it is the largest spurfowl
(males have a maximum mass of ~1590 g, with females having a maximum mass of
~1136 g) (Johnsgard 1988), and is characterized by its black forehead, eye-stripe and
chestnut crown. The lower neck is grey like the upper belly, but has greyish brown
margins and a thin central buff streak whereas the upper belly feathers have central
greyish black streaks. The lower belly feathers consist of a broad buff central streak
constricted in the middle and expanded distally into a tear drop, margined with rufous.
It shares the least genetic distance with P. ochropectus at 3% of sequence divergence.
The specimens of pentoni were not available for examination. Irrespective of the
quantitative differences presented in literature, the recommendation is that pentoni
should in the meantime be included in Pternistis erckelii (Rüppell, 1835).
Pternistis castaneicollis complex
Pternistis castaneicollis is distributed in Ethiopia, Somalia, and possibly Kenya (Fig.
6.3). The subspecies, castaneicollis, bottegi, gofanus, ogoensis, kaffanus and atrifrons
are recognized within P. castaneicollis (Table 1.2). Pternistis castaneicollis is generally
similar to P. erckelii with the belly feathers being made up of a broad buff central streak
which is constricted in the middle and expanded distally into a tear drop, margined with
rufous. The eastern Ethiopian subspecies have an extensive double-patterning on the
back with wing coverts and the upper belly being clearly defined in black and white,
248
with some ochre and chestnut. The subspecies atrifrons from southern Ethiopia is quite
different from the other subspecies with a lesser degree of colouration and patterning on
the belly. The less defined belly patterning is in brown and buff, and the throat and the
belly cream instead of white as in other subspecies. Despite its distinct morphology,
atrifrons was found to have similar voice, habits and environment (Benson 1945) to
other subspecies of P. castaneicollis. It shares the least genetic distance with P.
camerunensis differing at 8% of sequence divergence.
The subspecies, ogoensis, bottegi, gofanus and kaffanus are morphologically
similar to the nominate castaneicollis. The southern Ethiopian subspecies, atrifrons is
the one that is quite distinct in being less coloured and patterned on the belly as well as
having the well-defined patterning and rich colour on the back. Despite the similar
voice, habit and habitat that atrifrons shares with other P. castaneicollis subspecies
(Hall 1963), it would therefore be crucial to include its DNA in the analysis in order to
gain insight into its phylogenetic placement given its distinct morphology. It could be
that atrifrons is a valid species sister to P. castaneicollis Salvadori, 1888 which remains
a valid species.
Pternistis camerunensis
This is a monotypic, sexually dimorphic species which is endemic to Mount Cameroon
(Fig. 6.3). Its taxonomic status as a monotypic species has never been disputed. The
male P. camerunensis has the back (strictly excluding the lower neck) and the wings
coloured rich dark brown, crown grey brown and the rest of the belly and the lower
neck are plain grey with some black feather centres (Hall 1963). The upper- and
underparts of the female are mottled and vermiculated with black, dark brown and buff
249
and some off-white U- to V-patterning on belly and U-patterning on the lower neck. It
shares the least genetic distance with P. nobilis differing at 7% of sequence divergence.
Pternistis camerunensis Alexander, 1909 is acknowledged as a valid species.
Pternistis swierstrai
This monotypic species is endemic to the patches of evergreen forest in the highlands
and along the escarpment of Angola (Fig. 6.3). It is slightly sexually dimorphic in its
plumage and it is very different and isolated from the other members of the group. Male
and female birds have white eye stripes not found in any other member of the Montane
Group. The male breast is black contrasting with the buff throat while the belly consists
of broad buff central streaks with blackish margins. The females were not examined and
their underparts are according to Hall (1963), buff with irregular black or brown
blotches or bars which are concentrated on the upper belly to form a mottled band and
are sparse in the centre of the belly. It shares the least genetic distance with afer
differing at 6% of sequence divergence.
Scaly Group
The Scaly Group comprises three allopatric species, P. ahantensis, P. squamatus, P.
griseostriatus, with subspecies, P. squamatus (squamatus, maranensis, schuetti,
usambarae, uzungwensis, doni, zappeyi, tetraoninus, chyuluensis) and P. ahantensis
(ahantensis, hopkinsoni) being recognized (Table 1.2). The Scaly taxa are paraphyletic,
in which case P. squamatus and P. griseostriatus have close phylogenetic association
(Fig. 6.8, 6.10, 6.11). Squamatus and schuetti appear as sister taxa in some analyses.
250
(Fig. 6.8, 6.9), and schuetti is sister to P. griseostriatus others (Fig. 6.10, 6.11).
Generally, the phylogenetic relationship of the Scaly Group members is unresolved.
The general finding is that there are no morphological characters that link up the
three species within the group except that they all have some degree of scaliness. There
is no qualitative break in characters. There is a cline from the west African P.
ahantensis which has clearly defined patterning through to north, east and central
African P. squamatus (moderately patterned) and down to southwestern African P.
griseostriatus which has less defined patterning. Pternistis griseostriatus is likely to
represent the end of the cline. There is a large genetic jump of 4% between P.
squamatus and P. ahantensis and 5% between P. squamatus and P. griseostriatus.
Pternistis squamatus complex
Pternistis squamatus, a species distributed in south central Nigeria, south eastern to
western Zaire and east to Uganda, south Sudan and south western Ethiopia, then south
to east Zaire, north eastern Tanzania and extreme northern Malawi (Fig. 6.4), consists of
the following subspecies, squamatus, maranensis, schuetti, usambarae, uzungwensis,
doni, zappeyi, tetraoninus, chyuluensis (Table 1.2). Pternistis squamatus is not sexually
dimorphic, although female birds tend to be less strikingly vermiculated. The nominate
subspecies squamatus is characterized by having indistinctly vermiculated upperparts
with faint U-patterning on the lower neck and the feathers have blackish centres tinged
with red-brown. The belly is plain brown with a scaly pattern and ill-defined dark shaft
streaks margined buff. Schuetti (with 1% sequence divergence) differs from squamatus
in being less vermiculated with the pattern on the lower neck being more clearly defined
and with more red-brown at the centre of feathers. The edges of the darker belly feathers
251
have buff edges giving the streaky effect. It shares the least genetic distance with P.
griseostriatus differing at 3% of sequence divergence. The two subspecies, schuetti and
maranensis have slightly diverged genetically and morphologically from squamatus and
warrants only subspecies status for the former. The genetic divergence between schuetti
and maranensis is just above 1% and the two subspecies differ in the degree of
colouration. Therefore, on the basis of the marked genetic distances and different
morphology, two subspecies are recognized within P. squamatus, that is, Pternistis
squamatus squamatus Cassin, 1857 and Pternistis squamatus schuetti Cabanis, 1880.
Pternistis s. schuetti will include maranensis, chyuluensis, doni, usambarae,
uzungwensis, zappeyi and tetraoninus.
Pternistis ahantensis complex
This species occurs in lowlands of Senegambia to south western Nigeria (Fig. 6.4) and
it comprises two putative subspecies, ahantensis and hopkinsoni. The belly feathers are
richly coloured and streaked with dark brown chestnut edged buff while the upperparts
are vermiculated with some white U-patterning (indistinct on the back but very distinct
on the lower neck). It shares the least genetic distance with P. squamatus from which it
differs with 4% of sequence divergence. The two subspecies, ahantensis and hopkinsoni
show little morphological differentiation. Pternistis ahantensis Temminck, 1854
remains a valid species.
Pternistis griseostriatus
Pternistis griseostriatus is an indistinctive, monotypic species endemic to Angola (Fig.
6.4). The lower neck feathers and wing coverts are chestnut broadly vermiculated as in
252
P. s. squamatus and P. a. ahantensis, but the belly is plain and the upper belly and flank
feathers are chestnut and edged with greyish or creamy buff. It shares the least genetic
distance with squamatus differing at 3% of sequence divergence. This study
acknowledges Pternistis griseostriatus O. Grant 1890 as a valid species due to its
marked genetic distinctiveness (4%) compared to P. squamatus and P. ahantensis.
Conclusions
The definition of species and subspecies will continue to be contested as more and more
data are accumulated, analytical methods continue to be improved and people have
freedom to think and present their own views. The good side of this will be that there
will be great improvement on how species and subspecies are defined and delimited.
The conflict exists between characters from different sources possibly because of
hybridization and incomplete lineage sorting and these conditions make taxonomic
delineation of spurfowls challenging. Despite the striking diversity within the group, all
spurfowls share a single evolutionary path. However, among all of Hall’s putative
monophyletic species groups, the Bare-throated Group is the only one to be recovered
as monophyletic while others remain largely paraphyletic.
The number of species among spurfowls did not increase significantly; P. afer is
the only species that was split into two P. afer and P. cranchii. The number of
subspecies is reduced (Appendix 6.2). The recommendation is that phylogeographic
studies should be conducted for the various species complexes and fresh DNA samples
for all the other known subspecies need to be collected and sequenced so that sound
taxonomic decisions can be made.
253
Table and Figures
Table 6.1. Characters and character states scored and used for phylogenetic analysis of
spurfowls.
1. Crown margins: unmargined = 0, grey = 1, buff = 2, grey brown = 3
2. Nares: black = 1, chestnut = 2, grey brown = 3, buff or white =4 buff = 4, white = 5
3. Hindneck patterning: unpatterned = 0, mottled = 1, streaked = 2
4. Hindneck base colour: grey brown = 1, grey black = 2, grey chestnut = 3
rufous brown = 4, black = 5
5. Hindneck margins: unmargined = 0, grey =1, buff = 2, grey brown = 3
6. Lower neck patterning: streaked = 1, mottled = 2, barred = 3
7-10. Back plumage: absent = 0, streaked = 1 (7), mottled = 1 (8),
vermiculated = 1 (9), barred =1 (10)
11-13. multistate Uppertail coverts: absent = 0, barred = 1 (11), vermiculated = 1 (12),
streaked = 1 (13)
14. Throat: feathered = 1, yellow skin = 2, orange = 3, red = 4
15-18. multistate Undertail coverts: absent = 0, barred = 1 (15), streaked = 1 (16),
vermiculated = 1 (17), mottled = 1 (18)
19. Bare skin around eye: none = 0, red = 1, yellow = 2
20. Leg colour: yellow = 1, red = 2, orange red = 3, orange = 4, olive green
= 5, orange yellow = 6, black = 7
21. Number of spurs: one = 1, two = 2
22. Wing length (males): <160 mm = 1, <180 mm =2, <200 = 3, >200 =4
23. Culmen length/wing length: <.16 = 1, <.20 = 2, >= .20 =3
24. Tail length/ wing length: <.54 = 1, >= .54 = 2
25. Sexual dimorphism (plumage): absent = 0, present = 1
26. Sexual dimorphism (wing length): female >= .9 of male = 0, < .9 = 1
27. Vocalization strophe duration: <= 0.3 sec = 1, > 0.3 < 0.6 = 2, > 0.6 = 3
28. Number of elements: 1 = 1, 2 = 2, >2 = 3
29. Inter-element interval: absent or indistinct = 1, distinct = 2
30. Cackle-trill: absent = 0, present = 1
31-32. multistate Strophe character: tonal = 0 (31), trill = 1 (32)
33. ‘KO-WAAARK’ advertisement call: absent = 0, present = 1
254
Table 6.2. Organismal character matrix used in phylogenetic analysis. ? indicates information
which could not be sourced.
Character no.
Taxon
1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3 3 3
123456789 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3
Pternistis hartlaubi
Pternistis camerunensis
Pternistis nobilis
Pternistis erckelii
Pternistis swierstrai
Pternistis castaneicollis
Pternistis c. atrifrons
Pternistis ochropectus
Pternistis jacksoni
Pternistis squamatus
Pternistis s. schuetti
Pternistis ahantensis
Pternistis griseostriatus
Pternistis b. bicalcaratus
Pternistis b. ayesha
Pternistis b. adamauae
Pternistis icterorhynchus
Pternistis clappertoni
Pternistis c. sharpii
Pternistis harwoodi
Pternistis h. hildebrandti
Pternistis h. fischeri
Pternistis natalensis
Pternistis adspersus
Pternistis capensis
Pternistis leucoscepus
Pternistis l. infuscatus
Pternistis rufopictus
Pternistis afer afer
Pternistis cranchii
Pternistis afer humboldtii
Pternistis swainsonii
011101110 0 1 1 0 1 1 0 0 0 0 1 2 1 3 2 1 0 3 3 2 0 1 1 0
011202011 0 1 1 0 1 1 0 0 0 1 2 2 2 1 1 1 0 2 2 2 0 1 0 0
011201100 0 0 0 0 1 0 0 0 0 1 2 2 3 1 1 0 0 2 2 2 0 1 1 1
021301001 0 0 0 0 1 0 1 1 0 0 2 2 4 1 2 0 1 3 3 2 1 0 1 0
010501100 0 0 0 0 1 0 0 0 0 0 2 2 3 2 2 1 0 3 3 2 1 0 1 0
010301101 0 0 0 0 1 0 1 0 0 1 2 2 4 1 2 0 1 3 3 2 1 0 1 0
012121101 0 0 0 0 1 0 1 0 0 1 2 2 4 1 2 0 1 3 3 2 1 0 1 0
021301001 0 0 0 0 1 0 1 1 0 0 2 2 4 1 2 0 1 2 2 2 0 1 0 0
010301101 0 0 0 0 1 0 1 0 0 0 2 2 4 1 2 0 0 ? ? ? ? ? ? 0
030102011 0 1 1 0 1 0 0 0 1 0 3 2 2 1 2 0 0 3 3 2 0 1 1 0
031102011 0 1 1 0 1 0 1 0 0 0 3 2 2 1 2 0 0 3 3 2 0 1 1 0
040101101 0 0 1 01 0 1 1 0 0 4 2 2 1 2 0 0 2 2 2 0 1 1 0
030101101 1 1 0 0 1 0 0 0 0 0 3 1 1 1 2 0 0 3 1 1 0 0 1 0
011401101 0 1 1 0 1 0 1 0 0 0 5 2 2 2 1 0 0 1 2 1 0 01 0
011401101 0 1 1 0 1 0 1 0 0 0 5 2 2 2 1 0 0 1 2 1 0 0 1 0
011202101 0 1 1 0 1 0 1 0 0 0 5 2 2 2 1 0 0 1 2 1 0 0 1 0
011201001 1 1 1 0 1 1 0 0 0 2 6 2 2 2 1 0 0 1 2 1 0 0 1 0
011401100 0 1 1 0 1 0 1 0 0 1 2 2 2 2 1 0 0 2 2 1 0 0 1 0
010101100 0 1 1 0 1 0 1 0 0 1 2 2 2 2 1 0 0 2 2 1 0 0 1 0
011401100 0 1 1 0 1 0 1 0 0 1 2 2 2 2 2 ? 1 2 2 1 0 0 1 0
112211001 1 1 1 0 1 1 0 1 0 0 2 2 2 1 2 1 0 1 3 2 0 0 1 0
112211001 1 1 1 0 1 1 0 1 0 0 2 2 2 1 2 1 0 1 3 2 0 0 1 0
030101001 1 1 1 0 1 1 0 0 0 0 2 1 2 1 2 0 02 3 2 0 1 1 0
010203000 1 1 0 0 1 1 0 0 0 2 2 1 2 1 2 0 0 3 3 2 1 1 1 0
232121100 0 1 0 0 1 0 1 0 0 0 2 2 4 1 2 0 0 3 3 2 0 1 0 0
342131101 0 1 1 0 2 0 0 1 0 1 7 2 3 2 1 0 0 3 2 2 0 0 1 1
012101001 0 1 1 0 2 0 0 1 0 1 7 2 3 2 1 0 0 3 2 2 0 0 1 1
030101101 0 0 1 1 3 0 1 0 0 1 7 2 4 2 1 0 0 2 2 2 0 0 1 1
352101100 0 0 0 0 4 0 1 0 0 1 2 2 3 2 1 0 0 2 2 2 0 1 1 1
010101101 0 0 1 0 4 0 1 1 0 1 2 2 3 2 1 0 0 2 2 2 0 1 1 1
342131100 0 0 0 0 4 0 0 0 0 1 2 2 3 2 1 0 0 2 2 2 0 1 1 1
332131101 0 0 1 1 4 1 0 1 0 1 7 1 3 2 1 0 0 3 2 2 0 1 1 1
255
Table 6.3. Taxa for which DNA sequences were generated. AMNH stands for American
Museum of Natural History, FHHM = Franch Natural Histiry Museum, TM = Transvaal
Museum, BM = British Museum - Natural History Museum at Tring, SAM = Iziko Museums
of Cape Town (Natural History), PFIAO = Percy FitzPatrick Institute of African
Ornithology, TMC = Timothy M. Crowe, University of Cape Town, South Africa, GB =
GenBank, ‘-’ = Unknown, Br. muscle = Breast muscle, Pect. muscle = Pectoral muscle. The
specific and susubpecific epithets are as recorded on specimen label.
Taxa name
Sample No.
Origin
Date coll.
Sample type
PFIAO 108,
AMNH 267682
AMNH 541485
BM 1903.10.14.91
BM 1932.5.10.214
TM 28584
TM 20341
BM 1953.54.56
AMNH 202502
AMNH 416180
AMNH211906
TMC 40
SAM 2055756a
SAM 2003501
AMNH 202503
PFIAO 109
AMNH 419169
AMNH 541581
Tudor East, Watervalboven
Mombola
Russisi River
E. Transvaal
S. Tanganyika
Cunene River
Selindu, Mabsettler
Mwinilunga, N. Rhodesia
Poona Singida
Tukuyu
Buhumbiro
Marico River
Deka
Victoria falls
Gagayo, Muranza
KenyaTana River, Kenya colony
-
2004
1903
1932
1957
1935
1953
2004
1969
1904
2004
-
Liver
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Liver
Toe-pad
Toe-pad
Toe-pad
Heart
Toe-pad
Toe-pad
AMNH 541471
Badaltino, Shoa
AMNH DOT11039 Ethiopia
FMNH 1971-1072 Djbouti
GB
AMNH541435
Rafissa, Abyssinia
AMNH541426
Lower Sheikh
AMNH261929
East slope, Mt. Kenya
AMNH1759
West Ruwenzori
TMC 42
Mount Cameroon
AMNH 419126
Angola
TMC 67
Angola, 14.49 S 13.23 E
-
30/06/2010
Toe-pad
Liver
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Liver
Toe-pad
Blood
AMNH 541409
PFIAO 117
AMNH 541407
AMNH 763912
AMNH 541411
Nr York Pass, Sierra Leone
Kilimanjaro district
Tshibati, Dem. Rep. Congo
Ndalla Tanda
-
Toe-pad
DNA
Toe-pad
Toe-pad
Toe-pad
TM 14682
AMNH 541250
AMNH 541280
AMNH 704359
Gold Coast, Hinterland
1901
Forest of Mamora
Kavene District, Sierra Leone Cameroon
-
Bare-throated Group
P. afer
P. afer benguellensis
P. afer harterti
P. afer nudicollis
P. afer böhmi
P. afer cunenensis
P. humboldtii swynnertoni
P. cranchii cranchii
P. cranchii itigi
P. cranchii intercedens
P. cranchii nyanzae
P. swainsonii
P. s. lundazi
P. s. chobiensis
P. rufopictus
P. leucoscepus
P. l. infuscatus
P. l. muhamed-ben-abdullah
Montane Group
P. erckelii
P. erckelii
P. ochropectus
P. castaneicollis
P. c. bottegi
P. c. ogoensis
P. jacksoni
P. nobilis
P. camerunensis
P. swierstrai
P. swierstrai
-
Scaly Group
P. ahantensis
P. squamatus
P. s. maranensis
P. s. schuetti
P. griseostriatus
Vermiculated Group
P. b. bicalcaratus
P. b. ayesha
P. b. thornei
P. b. adamauae
256
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Taxa name
Sample No.
Origin
Date coll.
Sample type
AMNH 541305
TMC 68
AMNH 541324
AMNH 541341
AMNH 156922
GB
AMNH 551345
AMNH 261945
AMNH 347277
AMNH 207771
TMC 120
TMC 121
AMNH 703654
PFIAO 229
PFIAO 206A
BM 1927.11.5.18
Takoukout, Camergon
Cameroon
Adarte
S. Ethiopia
Fanadji
Gilgil River
N. Tanganyika Territory
Mafinga Mt., N. Rhodesia
Neng
Marico River, South Africa
Namibia
Erungo Plateau
Kakamas, South Africa
-
2005
2004
2006
1927
Toe-pad
Br. muscle
Toe-pad
Toe-pad
Toe-pad
Blood
Toe-pad
Toe-pad
Toe-pad
Toe-pad
Liver
Br. muscle
Toe-pad
Heart
Liver
Toe-pad
Vermiculated Group
P. clappertoni
P. clappertoni
P. c. sharpii
P. c. nigrosquamatus
P. icterorhynchus
P. hildebrandti
P. h. altumi
P. h. fischeri
P. h. johnstoni
P. h. helleri
P. natalensis
P. hartlaubi
P. h. crypticus
P. capensis
P. adspersus
P. harwoodi
257
Table 6.4. DNA markers sequenced and primers used for PCR amplification and sequencing
of preserved tissues.
Primer name
Primer sequence (5’to 3’)
Reference
cta gga atc atc cta gcc cta ga
act cta cta ggg ttt ggc c
atc aca aac cta ttc tc
aac gca gtc atc tcc ggt tta caa gac
J.G. Groth (pers. comm.)
P. Beresford (pers. comm.)
P. Beresford (pers. comm.)
Edwards & Wilson (1990)
Spurfowls
Cytochrome b
L14578
MH15364
ML15347
H15915
Table 6.5. DNA marker sequenced and primers used for DNA amplification and sequencing
of museum toe-pads.
Primer name
Primer sequence (5’ to 3’)
Reference
Cytochrome b
Spurfowl-specific primers
L14851 (General)
cct act tag gat cat tcg ccc t
Pt-H195
ttt cgr cat gtg tgg gta cgg ag
Pt-H194
cat gtr tgg gct acg gag g
MH15145
aag aat gag gcg cca ttt gc
Kornegay et al. (1993)
R. Moyle & T. Mandiwana-Neudani
R. Bowie
P. Beresford
Pt-L143
Pt-H361
gcc tca tta ccc aaa tcc tca c
gtg gct att agt gtg agg ag
R. Moyle & T. Mandiwana-Neudani
R. Moyle & T. Mandiwana-Neudani
Pt-L330
Pt-H645
tat act atg gct cct acc tgt ac
ggg tgg aat ggg att ttg tca gag
R. Bowie
R. Moyle & T. Mandiwana-Neudani
Pt-L633
Pt-H901
ggc tca aac aac cca cta ggc
agg aag ggg att agg agt agg at
R. Moyle & T. Mandiwana-Neudani
R. Moyle & T. Mandiwana-Neudani
L2-2312
H15696
cat tcc acg aat cag gct c
aat agg aag tat cat tcg ggt ttg atg
R. Bowie
Edwards et al. (1991)
Pt-L851alt
Pt-H1050
cct att tgc cta cgc cat cct ac
gat gct gtt tgg ccg atg
R. Bowie
R. Bowie
Pt-L961
Pt-L961alt
HB20 (General)
cga acc ata aca ttc cca c
ctc atc cta ctc cta atc ccc
ttg gtt cac aag acc aat gtt
R. Moyle & T. Mandiwana-Neudani
R. Bowie
J. Feinstein (pers. comm.)
258
Table 6.6. The inter- and intraspecific uncorrected ‘P’ distances calculated from CYTB data
partition only. Taxon names are as known prior to this study.
Taxo n name
1 P ternis tis hartlaubi
2 P ternis tis h. crypticus
3 P ternis tis cam erunens is
4 P ternis tis no bilis
5 P ternis tis cas taneico llis
6 P ternis tis c. bo ttegi
7 P ternis tis c. o go ens is
8 P ternis tis erckelii
9 P ternis tis o chro pectus
10 P ternis tis s wiers trai
11 P ternis tis jacks o ni
12 P ternis tis ahantens is
13 P ternis tis s quam atus
14 P ternis tis s . m aranens is
15 P ternis tis s . s chuetti
16 P ternis tis gris eo s triatus
17 P ternis tis b. bicalcaratus
18 P ternis tis b. adam auae
19 P ternis tis b. ayes ha
2 0 P ternis tis b. tho rnei
2 1P ternis tis ictero rhynchus
2 2 P ternis tis clapperto ni
2 3 P ternis tis c. nigro s quam atus
2 4 P ternis tis c. s harpii
2 5 P ternis tis harwo o di
2 6 P ternis tis capens is
2 7 P ternis tis ads pers us
2 8 P ternis tis natalens is
2 9 P ternis tis hildebrandti
3 0 P ternis tis h. altum i
3 1 P ternis tis h. fis cheri
3 2 P ternis tis h. helleri
3 3 P ternis tis h. jo hns to ni
3 4 P ternis tis afer (Ango la)
3 5 P ternis tis afer (SA)
3 6 P ternis tis a. nudico llis
3 7 P ternis tis a. benguellens is
3 8 P ternis tis a. bo hm i
3 9 P ternis tis a. cranchii
4 0 P ternis tis a. cunenens is
4 1 P ternis tis a. harterti
4 2 P ternis tis a. intercedens
4 3 P ternis tis a. itigi
4 4 P ternis tis a. nyanzae
4 5 P ternis tis a. s wynnerto ni
4 6 P ternis tis rufo pictus
4 7 P ternis tis leuco s cepus
4 8 P ternis tis l. infus catus
4 9 P ternis tis l. m uh-ben-abdullah
5 0 P ternis tis s wains o nii
5 1 P ternis tis s . cho biens is
5 2 P ternis tis s . lundazi
1
2
3
4
5
6
7
8
9 10 11 12 13
14
15 16 17 18 19 2 0 2 1 2 2 2 3 2 4
2 5 2 6 2 7 2 8 2 9 3 0 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 3 9 4 0 4 1 4 2 4 3 4 4 4 5 4 6 4 7 4 8 4 9 50
51 52
0%
11% 10%
11% 11% 7%
10% 10% 8% 8%
10% 10% 8% 8% 0%
10% 10% 8% 8% 1% 1%
9% 9% 7% 8% 4% 4% 4%
8% 8% 8% 8% 4% 4% 4% 3%
11% 11% 8% 9% 7% 7% 7% 6% 8%
11% 11% 9% 10% 8% 8% 7% 7% 7% 7%
9% 7% 8% 8% 7% 5% 7% 6% 8% 7% 8%
10% 10% 8% 9% 7% 7% 6% 7% 7% 7% 6% 4%
9% 9% 8% 8% 6% 7% 6% 6% 7% 7% 6% 4% 3%
8% 8% 8% 8% 6% 7% 6% 5% 6% 5% 7% 4% 4%
1%
10% 10% 8% 9% 6% 6% 6% 6% 6% 7% 5% 5% 4% 4% 3%
10% 10% 8% 9% 7% 7% 7% 6% 7% 7% 7% 6% 6%
5% 6% 5%
10% 10% 8% 8% 7% 7% 7% 6% 7% 7% 7% 6% 6% 6% 6% 5% 2%
11% 10% 8% 9% 7% 7% 7% 7% 7% 7% 7% 6% 6% 6% 7% 5% 1% 1%
11% 10% 8% 9% 7% 7% 7% 7% 7% 7% 7% 6% 6% 6% 7% 5% 1% 1% 1%
11% 11% 9% 9% 8% 8% 8% 7% 7% 8% 8% 6% 7%
7% 8% 6% 3% 3% 3% 3%
9% 9% 8% 9% 6% 6% 6% 5% 6% 6% 5% 6% 5%
5% 5% 4% 5% 6% 6% 6% 6%
9% 9% 8% 9% 6% 6% 6% 6% 7% 7% 6% 7% 5%
5% 5% 5% 6% 6% 6% 6% 7% 1%
9% 9% 8% 9% 6% 6% 6% 6% 7% 7% 6% 7% 5%
5% 5% 5% 6% 6% 6% 6% 7% 1% 1%
9% 9% 8% 9% 6% 6% 6% 6% 7% 7% 6% 7% 5%
5% 5% 5% 6% 6% 6% 6% 7% 1% 1% 1%
10% 10% 8% 9% 6% 7% 7% 7% 7% 6% 6% 8% 5%
5% 5% 4% 6% 6% 6% 6% 7% 5% 5% 5% 5%
10% 10% 9% 10% 7% 7% 7% 7% 7% 7% 6% 8% 6%
5% 6% 5% 6% 7% 7% 7% 7% 5% 5% 6% 6% 4%
10% 10% 8% 9% 7% 7% 7% 7% 7% 7% 6% 6% 5%
5% 5% 5% 6% 6% 6% 6% 7% 4% 5% 5% 5% 4% 4%
10% 10% 9% 9% 7% 7% 7% 8% 8% 7% 7% 5% 5%
5% 4% 6% 7% 6% 7% 7% 8% 6% 6% 6% 6% 5% 5% 3%
10% 10% 9% 10% 7% 7% 8% 8% 8% 7% 7% 7% 6%
5% 5% 6% 7% 6% 7% 7% 8% 6% 6% 6% 6% 5% 5% 3% 1%
10% 10% 8% 9% 6% 6% 7% 7% 7% 7% 6% 7% 5%
5% 6% 5% 6% 6% 6% 7% 7% 5% 5% 5% 5% 4% 4% 1% 3% 3%
9% 9% 9% 8% 7% 8% 7% 7% 7% 6% 7% 7% 6% 4% 4% 6% 6% 6% 6% 7% 7% 5% 6% 6% 6% 4% 4% 2% 2% 1% 3%
10% 10% 8% 9% 7% 7% 7% 7% 7% 7% 6% 7% 5%
5% 6% 5% 6% 6% 7% 7% 7% 5% 5% 5% 5% 4% 5% 1% 3% 3% 0% 3%
10% 10% 8% 8% 7% 8% 7% 7% 7% 7% 6% 7% 5% 6% 5% 6% 7% 7% 8% 8% 7% 5% 5% 5% 5% 6% 7% 6% 6% 6% 6% 6% 6%
10% 10% 8% 9% 7% 7% 7% 7% 7% 7% 6% 6% 6% 6% 6% 6% 7% 7% 8% 7% 8% 6% 6% 6% 6% 5% 6% 6% 6% 6% 6% 6% 6% 3%
9% 9% 8% 9% 6% 7% 6% 6% 7% 8% 7% 6% 6%
5% 5% 6% 7% 7% 7% 7% 8% 5% 5% 5% 5% 6% 6% 6% 6% 6% 6% 6% 6% 3% 3%
9% 9% 8% 8% 6% 6% 6% 6% 6% 7% 6% 7% 5%
5% 5% 6% 6% 7% 7% 7% 7% 5% 5% 4% 5% 5% 6% 5% 5% 5% 5% 5% 5% 1% 3% 2%
9% 9% 7% 8% 6% 6% 6% 5% 5% 7% 5% 6% 4%
5% 3% 5% 6% 6% 6% 6% 6% 4% 5% 4% 5% 5% 5% 5% 5% 5% 5% 5% 5% 2% 3% 2% 1%
9% 9% 8% 8% 6% 7% 6% 6% 6% 7% 6% 6% 5%
5% 5% 6% 7% 7% 7% 7% 7% 5% 5% 5% 5% 5% 6% 5% 5% 5% 5% 5% 5% 2% 4% 2% 0% 1%
8% 7% 7% 7% 6% 7% 6% 5% 6% 6% 5% 7% 5%
5% 5% 5% 6% 6% 7% 7% 6% 5% 5% 5% 5% 5% 5% 4% 5% 5% 5% 4% 5% 3% 3% 3% 1% 1% 1%
9% 9% 8% 8% 6% 6% 6% 6% 7% 7% 6% 6% 5%
5% 5% 6% 7% 7% 7% 7% 7% 5% 5% 4% 5% 5% 6% 5% 6% 6% 5% 5% 5% 2% 3% 2% 0% 1% 1% 1%
9% 9% 8% 8% 6% 6% 6% 6% 7% 7% 6% 6% 5%
5% 5% 6% 7% 7% 7% 7% 7% 5% 5% 5% 5% 5% 6% 5% 6% 6% 5% 5% 5% 2% 4% 2% 1% 1% 1% 1% 0%
10% 9% 8% 8% 7% 7% 6% 6% 7% 7% 6% 7% 5%
5% 6% 6% 7% 7% 7% 7% 8% 5% 5% 5% 5% 6% 6% 6% 6% 6% 6% 6% 6% 2% 4% 3% 1% 0% 1% 1% 1% 1%
9% 9% 8% 8% 6% 7% 6% 6% 6% 7% 6% 6% 5%
5% 5% 6% 7% 7% 7% 7% 7% 5% 5% 5% 5% 5% 6% 5% 5% 5% 5% 5% 5% 2% 4% 2% 0% 1% 0% 1% 1% 1% 1%
10% 9% 9% 9% 7% 7% 7% 7% 7% 8% 7% 7% 6% 6% 6% 7% 7% 7% 8% 7% 8% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 3% 4% 1% 2% 2% 3% 3% 2% 2% 3% 3%
9% 9% 8% 8% 6% 6% 6% 6% 7% 7% 6% 6% 5% 5% 4% 6% 7% 7% 7% 7% 8% 5% 5% 5% 5% 5% 6% 6% 6% 6% 6% 5% 6% 2% 3% 2% 2% 2% 2% 2% 2% 2% 2% 2% 2%
10% 10% 9% 9% 7% 7% 7% 7% 7% 8% 7% 7% 6% 6% 6% 6% 7% 7% 7% 7% 8% 5% 5% 5% 5% 6% 7% 6% 6% 7% 6% 6% 6% 4% 6% 5% 4% 4% 4% 4% 4% 4% 4% 4% 5% 4%
9% 9% 8% 9% 7% 7% 7% 6% 7% 7% 6% 6% 5% 6% 5% 6% 6% 7% 7% 7% 7% 5% 5% 5% 5% 6% 6% 6% 6% 6% 6% 5% 6% 5% 6% 5% 4% 4% 4% 4% 4% 4% 4% 4% 5% 4% 1%
9% 9% 8% 9% 6% 6% 6% 6% 6% 7% 6% 7% 5%
10% 9% 7% 9% 6% 6% 6% 6% 6% 7% 6% 8% 6%
5% 5% 5% 6% 7% 6% 7% 7% 5% 5% 4% 4% 6% 6% 6% 6% 6% 6% 5% 6% 4% 5% 5% 4% 4% 4% 4% 4% 4% 4% 4% 5% 4% 2% 2%
5% 5% 6% 7% 7% 7% 7% 7% 5% 5% 5% 5% 6% 6% 6% 6% 6% 6% 6% 6% 5% 5% 4% 4% 4% 4% 5% 4% 4% 4% 4% 5% 4% 5% 4% 4%
11% 11% 9% 11% 8% 8% 8% 8% 8% 9% 6% 7% 8% 8% 7% 7% 8% 8% 9% 8% 9% 7% 8% 7% 7% 7% 8% 7% 9% 8% 7% 9% 7% 8% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 7% 8% 7% 5%
10% 10% 9% 11% 8% 8% 8% 7% 8% 9% 8% 8% 7%
7% 6% 8% 8% 8% 8% 8% 9% 6% 6% 6% 6% 7% 8% 7% 8% 8% 7% 9% 7% 8% 7% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 6% 7% 7% 6% 3% 2%
259
1
2
1
1
Bare-throated Group
1. Pternistis cranchii
2. P. afer humboldtii
3. P. a. afer
3
2
1
1
2
3
3
Figure 6.1. Distribution ranges of Hall’s Bare-throated Group taxa.
260
1
2
3
Bare-throated Group (concl.)
1.
2.
3.
4.
P. leucoscepus leucoscepus
P. l. infuscatus
P. rufopictus
P. swainsonii
4
4
4
Figure 6.2. Distribution ranges Hall’s Bare-throated Group taxa.
261
2
3
2
4
5
1
6
7
Montane Group
1.
2.
3.
4.
5.
6.
7.
8.
Pternistis camerunensis
P. erckelii
P. ochropectus
P. castaneicollis castaneicollis
P. c. atrifrons
P. jacksoni
P. nobilis
P. swierstrai
8
Figure 6.3. Distribution ranges of Hall’s Montane Group of spurfowls.
262
2
1
3
1
2
2
2
Scaly Group
1.
2.
3.
4.
Pternistis ahantensis
P. squamatus squamatus
P. s. schuetti
P. griseostriatus
2
4
Figure 6.4. Distribution ranges of Hall’s Scaly Group taxa.
263
4
4
3
1
3
5
2
6
3
6
Vermiculated Group
1.
2.
3.
4.
5.
6.
P. bicalcaratus bicalcaratus
P. b. adamauae
P. clappertoni clappertoni
P. c. sharpii
P. harwoodi
P. icterorhynchus
Figure 6.5. Distributional ranges of Hall’s Vermiculated Group.
264
1
1
Vermiculated Group (cont.)
1.
2.
3.
4.
5.
P. hildebrandti hildebrandti
P. h. fischeri
P. natalensis
P. adspersus
P. capensis
2
3
4
2
4
3
4
5
Figure 6.6. Distributional ranges of Hall’s (1963) Vermiculated Group.
265
Vermiculated Group (concl.)
1. Pternistis hartlaubi
1
Figure 6.7. Distributional ranges of Hall’s (1963) Vermiculated Group.
266
Coturnix coturnix
Alectoris chukar
Pternistis hartlaubi
Pternistis camerunensis
93
Pternistis nobilis
Pternistis erckelii
95
M
T
t
Pternistis ochropectus
Pternistis castaneicollis castaneicollis
95
Pternistis castaneicollis atrifrons
Pternistis jacksoni
Pternistis squamatus squamatus
Pternistis squamatus schuettii
Pternistis griseostriatus
S
C
t
90
Pternistis hildebrandti hildebrandti
73
Pternistis hildebrandti fischeri
98
S
V
t
Pternistis natalensis
Pternistis adspersus
Pternistis capensis
Pternistis ahantensis
83
73
100
Pternistis bicalcaratus bicalcaratus
Pternistis bicalcaratus ayesha
Pternistis bicalcaratus adamauae
N
V
t
Pternistis icterorhynchus
Pternistis clappertoni clappertoni
100
Pternistis clappertoni sharpii
90
Pternistis harwoodi
100
Pternistis leucoscepus leucoscepus
Pternistis leucoscepus infuscatus
Pternistis swainsonii
Pternistis rufopictus
89
71
93
84
B
T
G
Pternistis afer afer
Pternistis afer humboldtii
Pternistis cranchii
Pternistis swierstrai
Figure 6.8. A parsimony tree (1 of 2 most parsimonious trees similar to Strict Concensus tree) obtained from
combined mitochondrial Cytochrome-b and organismal characters. Numbers above branches represent
boostrap support values (only ≥ 70% are presented). MTt stands for Montane taxa SCt – Scaly taxa, SVt –
Southern Vermiculated taxa, NVt – Northern Vermiculated taxa and BTG – Bare-throated Group.
267
Coturnix coturnix
Alectoris chukar
Pternistis hartlaubi
Pternistis camerunensis
Pternistis squamatus squamatus
70
Pternistis squamatus schuettii
Pternistis nobilis
Pternistis afer humboldtii
Pternistis afer afer
Pternistis leucoscepus leucoscepus
Pternistis leucoscepus infuscatus
B
T
G
Pternistis swainsonii
Pternistis cranchii
Pternistis rufopictus
Pternistis bicalcaratus bicalcaratus
Pternistis bicalcaratus ayesha
Pternistis bicalcaratus adamauae
Pternistis icterorhynchus
98
92
Pternistis hildebrandti hildebrandti
Pternistis hildebrandti fischeri
Pternistis clappertoni clappertoni
N
V
t
Pternistis harwoodi
Pternistis clappertoni sharpii
Pternistis ahantensis
65
Pternistis erckelii
Pternistis ochropectus
Pternistis castaneicollis castaneicollis
Pternistis castaneicollis atrifrons
M
T
t
Pternistis jacksoni
Pternistis swierstrai
Pternistis capensis
Pternistis natalensis
Pternistis adspersus
S
V
t
Pternistis griseostriatus
Figure 6.9. A parsimony tree (1 of 397 most parsimonious trees) obtained from organismal characters.
Numbers above branches represent boostrap support values (only ≥ 70% are presented). MTt stands
for Montane taxa, SCt – Scaly taxa, SVt – Southern Vermiculated taxa, NVt – Northern Vermiculated
taxa and BTG – Bare-throated Group.
268
Coturnix coturnix
Alectoris chukar
Pternistis adspersus
Pternistis capensis
88
Pternistis hildebrandti hildebrandti
98
Pternistis hildebrandti fischeri
S
V
t
Pternistis natalensis
Pternistis afer afer
76
Pternistis afer humboldtii
70
Pternistis rufopictus
87
B
T
G
Pternistis cranchii
100
89
Pternistis leucoscepus leucoscepus
Pternistis leucoscepus infuscatus
Pternistis swainsonii
Pternistis clappertoni clappertoni
100
Pternistis clappertoni sharpii
99
Pternistis harwoodi
Pternistis jacksoni
Pternistis swierstrai
Pternistis squamatus squamatus
S
C
t
Pternistis griseostriatus
Pternistis squamatus schuettii
Pernistis bicalcaratus bicalcaratus
80
N
V
t
Pternistis bicalcaratus ayesha
100
Pternistis bicalcaratus adamauae
Pternistis icterorhynchus
Pternistis castaneicollis castaneicollis
91
92
94
Pternistis erckelii
Pternistis ochropectus
79
89
Pternistis camerunensis
M
T
t
Pternistis nobilis
Pternistis hartlaubi
Figure 6.10. A parsimony tree (1 of 2 most parsimonious trees similar to Strict Consensus tree)
obtained from combined mitochondrial Cytochrome-b and organismal characters. Numbers above
branches represent boostrap support values (only ≥ 70% are presented). MTt stands for Montane
taxa, SCt – Scaly taxa, SVt – Southern Vermiculated taxa, NVt – Northern Vermiculated taxa and
BTG – Bare-throated Group.
269
Pternistis swainsonii
Pternistis cranchii
Pternistis afer humboldtii
B
T
89
Pternistis afer afer
G
Pternistis rufopictus
Pternistis leucoscepus infuscatus
100
Pternistis leucoscepus leucoscepus
Pternistis swierstrai
Pternistis nobilis
100
Pternistis camerunensis
Pternistis hartlaubi
Pternistis castaneicollis castaneicollis
88
M
T
t
Pternistis ochropectus
90
Pternistis erckelii
Pternistis jacksoni
99
98
Pternistis hildebrandti fischeri
Pternistis natalensis
Pternistis hildebrandti hildebrandti
95
S
V
t
Pternistis adspersus
Pternistis capensis
Pternistis squamatus schuettii
79
Pternistis griseostriatus
Pternistis squamatus squamatus
S
C
t
Pternistis bicalcaratus ayesha
Pternistis bicalcaratus adamauae
99
Pternistis icterorhynchus
Pernistis bicalcaratus bicalcaratus
91
Pternistis clappertoni clappertoni
N
V
t
Pternistis harwoodi
Pternistis clappertoni sharpii
Alectoris chukar
Coturnix coturnix
Figure 6.11. Maximum likelihood tree obtained from mitochondrial Cytochrome-b characters.
Numbers above branches represent boostrap support values (only ≥ 70% are presented). MTt stands for
Montane taxa, SCt – Scaly taxa, SVt – Southern Vermiculated taxa, NVt – Northern Vermiculated taxa
and BTG – Bare-throated Group.
270
Appendix 6.1. GenBank accession numbers for taxa sequenced for mitochondrial
Cytochrome-b.
CYTB
GenBank
no.
Taxon
Pternistis hartlaubi hartlaubi
FR691618
Pternistis hartlaubi crypticus
FR691619
Pternistis adspersus
FR691623
Pternistis afer Angola
FR694158
Pternistis afer benguellensis
FR694159
Pternistis afer böhmi
FR694162
Pternistis cranchii cranchii
FR694164
Pternistis afer cunenensis
FR694160
Pternistis afer harterti
FR694161
Pternistis afer intercedens
FR694165
Pternistis afer itigi
FR694166
Pternistis afer nudicollis
FR694163
Pternistis afer nyanzae
FR694167
Pternistis afer South Africa
AM236908
Pternistis afer swynnertoni
FR694168
Pternistis bicalcaratus bicalcaratus
FR691624
Pternistis bicalcaratus adamauae
FR691626
Pternistis bicalcaratus ayesha
FR691625
Pternistis bicalcaratus thornei
FR691627
Pternistis camerunensis
FR691591
Pternistis capensis
AM236909
Pternistis castaneicollis castaneicollis
AM236903
Pternistis castaneicollis bottegi
FR691629
Pternistis castaneicollis ogoensis
FR691628
Pternistis clappertoniclappertoni
FR691602
Pternistis clappertoni nigrosquamatus
FR691604
Pternistis clappertoni sharpii
FR691603
Pternistis erckelii
FR691589
Pternistis griseostriatus
AM236905
Pternistis harwoodi
FR691600
Pternistis hildebrandti hildebrandti
FR691595
Pternistis hildebrandti altumi
FR691597
Pternistis hildebrandti fischeri
FR691598
Pternistis hildebrandti helleri
FR691599
Pternistis hildebrandti johnstoni
FR691596
Pternistis icterorhynchus
FR691601
Pternistis jacksoni
FR691594
Pternistis leucoscepus
AM236906
Pternistis leucoscepus infuscatus
FR691587
Pternistis leucoscepus muhamed-ben-abdullah
FR691586
271
CYTB
GenBank
no.
Taxon
Pternistis natalensis
AM236911
Pternistis nobilis
FR691592
Pternistis ochropectus
FR691590
Pternistis rufopictus
FR691588
Pternistis squamatus squamatus
AM236904
Pternistis squamatus maranensis
FR691630
Pternistis squamatus schuettii
FR691631
Pternistis swainsonii swainsonii
AM236907
Pternistis swainsonii chobiensis
FR694170
Pternistis swainsonii lundazi
FR694169
Pternistis swierstrai
FR691593
272
Appendix 6.2. A revised classification of spurfowls which was based on multiple lines of
evidence presented in this chapter. The status of taxa with an asterisk (*) might change when
DNA sequence data (or with increased number of nucleotide bases) are included in the
analyses. Species and subspecies authority appear in Table 1.2.
Family: Phasianidae
Sub-family: Phasianinae
Genus: Pternistis Wagler, 1832
Pternistis hartlaubi
Pternistis camerunensis
Pternistis erckelii
Pternistis ochropectus
Pternistis castaneicollis castaneicollis
Pternistis castaneicollis atrifrons*
Pternistis jacksoni
Pternistis nobilis
Pternistis swierstrai
Pternistis bicalcaratus bicalcaratus
Pternistis bicalcaratus adamauae
Pternistis clappertoni clappertoni
Pternistis clappertoni sharpii
Pternistis harwoodi
Pternistis icterorhynchus
Pternistis hildebrandti hildebrandti
Pternistis hildebrandti fischeri
Pternistis natalensis
Pternistis adspersus
Pternistis capensis
Pternistis ahantensis ahantensis*
Pternistis squamatus squamatus
Pternistis squamatus schuetti
Pternistis griseostriatus
Pternistis cranchii
Pternistis afer afer
Pternistis afer humboldtii
Pternistis leucoscepus leucoscepus
Pternistis leucoscepus infuscatus
Pternistis rufopictus
Pternistis swainsonii
273
CHAPTER 7
Historical biogeography of francolins and spurfowls
(Galliformes: Phasianidae)
Abstract
The biogeography of francolins has not received much attention since the monograph of
Hall (1963). Apart from the dispute about their monophyly, uncertainties in the
taxonomic designations and their phylogenetic relationships have generated opposing
hypotheses ("out of Asia" or "out of Africa") pertinent to the origin of the genus
Francolinus. Francolins (sensu lato) are galliform birds that occur in Africa, and are
mainly restricted to sub-Saharan and the Indian Sub-continent; inhabiting primarily
tropical/sub-tropical areas with some species occurring in forested habitats. They were
recently split into two distantly related assemblages, francolins and spurfowls
(divergence date going back to c. 33.6 mya) which are split among five francolin genera
Francolinus, Ortygornis, Afrocolinus, Peliperdix, Scleroptila as opposed to the single
genus Pternistis assigned to all spurfowl taxa.
This chapter seeks to understand the history of the current geographic
distribution patterns of francolins and spurfowls in light of their phylogeny and to test
the two opposing hypotheses on their origin. A range of organismal and DNA
characters were analyzed using parsimony and Bayesian methods. A new
biogeographical reconstruction method, spatial analysis of vicariance, was used to
detect disjunctions and to infer barriers while the ancestral area and habitat were
reconstructed in Mesquite. The colonization of Africa by ancestral species may have
274
been through dispersal from Asia, which resulted in the formation of disjunct
distributions and the somewhat rapid diversification of francolins and spurfowls within
Africa.
The Rift Valley system, Lake Chad, Upper Guinea and the Congolian forest,
major rivers such as Limpopo, Zambezi, Rovuma, Volta and Rufiji, the Okavango
Swamp and Sahara Desert emerged as the major physical breaks that may have created
and maintained the distributions observed today and promoted speciation among the
African francolins and spurfowls. Sharp diversity gradients in habitat appear to also
have played an important ecological role in facilitating diversification among these taxa.
275
Introduction
General background
Cracraft (1994) argued that a scientific problem that is central to systematists and
ecology has been to explain the spatio-temporal patterns of species diversity and
attempt to understand why taxa occur in particular areas and not others, as well as how
they assembled or spread to those regions. This becomes a difficult exercise to do given
that historical biogeography is at best poorly known. Some authors, e.g. de Quieroz
(2005), Donoghue and Moore (2003), Lamm and Redelings (2009) argue that the
challenge of historical biogeography is to distinguish dispersal and vicariance whereas
(Hovenkamp 1997, 2001) argued for the recognition of barriers being an important task
in a framework that can examine both inter and intra-continental distributional patterns.
In this chapter the focus is on francolins and spurfowls (Table 1.1), here
considered two independent radiations of taxa (Crowe et al. 2006, Chapter 2). These
assemblages are francolins, divided among the genera Francolinus, Ortygornis,
Afrocolinus, Peliperdix and Scleroptila (details in Chapter 5) and spurfowls, which are
attributed to a single genus Pternistis (Crowe et al. 2006, Chapter 6). The closest
relatives of each of these independent radiations occur in Asia and Europe: the genera
Gallus, Bambusicola spp., which are the close relatives of the francolins, and Coturnix,
Margaroperdix, Alectoris, Perdicula, Tetraogallus, Excalfactoria and Ammoperdix
spp., which are closely related to the spurfowls.
Habitat preferences of francolins and spurfowls
Francolins and spurfowls have complex distribution patterns (Snow 1978),
occurring in different habitats primarily of a tropical/sub-tropical nature, with some
276
species occurring in forested habitats (Hall 1963, Johnsgard 1988, del Hoyo et al. 1994,
Madge and McGowan 2002). The three Spotted Middle Eastern-Asian Francolinus
species, F. francolinus, F. pictus and F. pintadeanus, thrive in thick bush and patches of
scrub jungle (Johnsgard 1988). The Ortygornis taxa, O. sephaena and O. grantii occur in
scrubby African woodlands with grass habitats. The Asian francolin O. gularis is a
swamp-dwelling species, whereas Asian O. pondicerianus thrives in woodland and
bushes. The Red-tailed Peliperdix spp. inhabit open grassland and woodland savannas.
Afrocolinus lathami remains the only African francolin to be restricted to deeply forested
habitats (del Hoyo et al. 1994, Madge and McGowan 2002). Although the Red-winged
Scleroptila spp. occur in varied habitats at different latitudes, they primarily occupy
grassland (Johnsgard 1988, Madge and McGowan 2002), which could be open lowland
grasslands in which S. levaillantoides, S. gutturalis, S. finschi thrive, or the open hilly (S.
shelleyi) and highland grasslands preferred by S. afra, S. psilolaema, S. levaillantii, and S.
streptophora.
With regard to the spurfowls, the Bare-throated Group includes four relatively
large species, P. swainsonii which inhabits thickets and dry savannas, P. afer which
thrives in thickets, riparian scrub and adjoining grassland and forest edges, P.
leucoscepus occupies thicket and open savanna scrub areas and adjacent fields, and P.
rufopictus which occurs in thickets and in Acacia scrub and woodland. Members of the
Vermiculated Group (P. hartlaubi, P. capensis, P. natalensis, P. adspersus, P.
hildebrandti, P. clappertoni, P. icterorhynchus, P. bicalcaratus, P. harwoodi) generally
occupy dense bushy thickets and shrubby grasslands (del Hoyo et al. 1994, Madge and
McGowan 2002). In contrast to the habitat preferences of the Bare-throated and the
Vermiculated spp., members of the Montane Group (P. camerunensis, P. swierstrai, P.
277
nobilis, P. ochropectus, P. jacksoni, P. castaneicollis) inhabit largely montane forest
patches and scrubby slopes (Madge and McGowan 2002) with the exception of P.
erckelii being the only member that ventures into adjacent grasslands. The Scaly Group
spp., P. squamatus, P. griseostriatus and P. ahantensis inhabit forest edges with dense
undergrowth (Madge and McGowan 2002).
The origin of the genus Francolinus Stephens, 1819
The relationships between francolins with Palaearctic and Asian genera influenced
Hall’s (1963) hypothesis on the origin of the genus Francolinus. Even though Hall
(1963) produced commendable work in describing the geographical distribution patterns
of francolins, the origin of the genus ‘Francolinus’; however, continued to be highly
disputed (Hall 1963, Crowe et al. 1992). Based on the fact that francolins and spurfowls
share the closest relationships with Palearctic and Asian genera, Hall (1963) strongly
argued for this genus to be of Asian origin and she speculated that its age extended back
to the Oligocene 25-35 mya (Hall 1963, Sibley and Alquist 1985). She further
maintained that speciation of francolins in Africa was likely promoted by other factors
such as reduced competition. In contrast, Crowe and Crowe (1985) and Bloomer and
Crowe (1998) argued that the ancestor of francolins was African, inferring that
francolins dispersed from Africa to Asia. Despite the difference in opinion about the
origin of this lineage, Hall (1963), Crowe and Crowe (1985) and Bloomer and Crowe
(1998), all hypothesized that the ancestral francolin (traditionally including spurfowls)
was a small, quail-like phasianid based on the observation that the plumage and other
integumentary features of immature francolins resemble those of quails (Crowe et al.
1986).
278
A brief overview of historical biogeographical approaches
One way to investigate the biogeography of taxa is by tracing ancestral patterns i.e., the
history of character evolution and also by reconstructing the evolution of geographic
ranges of taxa on phylogenetic trees. In addressing the evolution of the distribution
patterns of francolins (sensu lato), Hall (1963) made a direct comparison of the ranges
of species with vegetation types and “tentatively tried to correlate some of the climatic
eras postulated with those known” as outlined in the monograph in Appendix 1, on page
173-174. Her attempts to explain the evolution of species involved two assumptions.
The first one being that the degree of divergence shown by two isolated populations can
be correlated with the length of time spent in isolation and secondly, that the species
assemblage present originated in some part or parts of their current range.
Until recently a commonly utilized approach to investigate historical
biogeography was that of dispersal-vicariance analysis (Ronquist 1997), which is
implemented in the software DIVA (Ronquist 1996). This approach requires one fullyresolved phylogenetic tree and minimizes the number of dispersal and extinction events
required to explain the species’ distributions by way of optimizing the ancestral areas
onto internal nodes of a phylogeny. Jønsson et al. (2010) used what they call a newly
developed Bayesian approach to dispersal-vicariance analysis, ‘Bayes-diva’ (Nylander
et al. 2008), which unlike Ronquist’s DIVA approach, accounts for phylogenetic
uncertainty. Ree and Smith (2008) developed a dispersal-extinction-cladogenesis (DEC)
model that reconstructs the evolution of geographic ranges. The DEC model has been
criticized for "ignoring vicariance” and assuming that the rate of evolution of
geographic ranges occurs independently, an assumption that may not be applicable to
some biogeographic inferences.
279
Spatial analysis of vicariance by Arias et al. (2011) is a new method which is
based on the ideas of Hovenkamp (1997, 2001) as implemented in VIP (Arias et al.
2011),
which
instead
of
looking
at
pre-defined
areas,
uses
direct
geographic/georeferenced data to detect disjunctions between sister groups and infers
barriers associated with them. It uses an optimality criterion framework which
maximizes the number of possible pairs of disjunct sister nodes while minimizing the
number of eliminated distributions (Arias et al. 2011). This method requires a
phylogeny.
In this chapter, the aim was to understand the history of the current geographic
distributional patterns of francolins and spurfowls in light of the species groups
suggested by Hall (1963), with modifications as outlined in Chapters 5 and 6, in order to
detect disjunctions and infer barriers, as well as to resolve the contrasting hypotheses on
the origin of the genus Francolinus sensu lato outlined by Hall (1963; our of Asia) and
Crowe and Crowe (1985; our of Africa), respectively.
Materials and methods
Data collection
Sampling of taxa and characters
Only francolin and spurfowl taxa that were recognized as valid terminal taxa in Chapter
5 (francolins) and Chapter 6 (spurfowls) were analyzed in this chapter.
Maps and mapping of distribution records of investigated taxa
The maps depicting the distribution ranges of francolins and spurfowls were produced
as outlined in Chapter 5 and 6.
280
Ancestral state reconstructions
The multistate matrices for francolins and spurfowls were generated for two traits: area
and habitat. The francolin area states were scored as follows: Southern Asia (0), West
Africa (1), Southern Africa (2), Central Africa (3) and East Africa (4), whereas those for
habitat were: open grasslands (0), scrubby grasslands (1), wooded grasslands (2),
Montane grasslands (3) and Forest (4). The states for area and habitat of the spurfowls
were coded as Southern Africa (0), West Africa (1), North Africa (2), Central Africa (3)
and East Africa (4) and those for habitat were: savanna and arid woodlands (0), mesic
woodlands (1), forest/forest edge (FFE) (2), lowland fynbos (3) and macchia (a
scrubland vegetation of the Mediterranean region which is composed primarily of
leathery, broad-leaved evergreen shrubs).
Spatial analysis of vicariance
In the absence of direct georeferenced distribution data, the distribution ranges of taxa
were used to detect disjunctions and infer barriers between sister groups.
Data analyses
Chapter 2 demonstrated that instead of francolins being considered a single evolutionary
lineage, they represent two lineages and as a result francolin and spurfowl analyses were
conducted separately
Ancestral state reconstructions
To reconstruct the ancestral patterns, a maximum likelihood approach (based on a
Bayesian topology) which seeks to find the ancestral states that maximize the
281
probability that the observed states would evolve under a stochastic model of evolution
(Schluter et al. 1997, Pagel 1999) was used. What differentiates this approach and
makes it a better method in reconstructing ancestral states over the parsimony approach
is that it incorporates both topology and branch-lengths. Another advantage of a
Bayesian reconstruction is that the marginal probabilities are assigned to nodes and thus
the level of confidence can be gained for any assignment. Ancestral area and habitat
reconstructions were implemented in Mesquite (Maddison and Maddison 2006) with the
idea of tracing the evolution of area and habitat among francolins and spurfowls.
Bayesian and parsimony phylogenetic trees were appended to the input files and the
analyses were performed under the mk1 model (Lewis 2001).
Spatial analysis of vicariance
Distributional ranges of taxa were analysed and an overlap of up to 25% was accepted
as a disjunct distribution. The cost of overlapped (i.e. non-disjunct) distributions was set
to the default of 1.00 and the cost for removal of distribution was 2.00, which means
that a removal was only accepted if it was found at least in one additional pair of
disjunct sister groups. To detect barriers, a simple hill climbing heuristic search was
performed starting with a sector of 20 nodes over 1000 iterations during which a
“flipping nodes” strategy was used. This procedure is quite effective when the
distributions are close together, but it becomes less accurate for distributions placed far
away, as it is placed in the line that is equidistant to each point.
282
Phylogenetic analyses
Bayesian and parsimony methods were used in analysing combined mitochondrial and
nuclear DNA data partitions from Chapter 2, 4 to 6, and the combined organismal
(Chapter 4, 5 and 6) and DNA data partitions, respectively. The francolin and spurfowl
trees were each rooted with the most basal taxon among members of these assemblages,
that is, F. francolinus for the francolins and P. hartlaubi for the spurfowls. Combined
organismal and DNA parsimony analyses consisting of 5141 characters of francolins
and 5203 characters of spurfowls were performed and these characters were used to
build the strict consensus trees (the boldest hypotheses of relationship) implemented in
TNT (Tree analysis using New Technology - Goloboff et al. 2008). All characters were
equally weighted and treated as non-additive. It must be noted that, in order to reduce
the denseness of taxa on the tree, the phylogenetic trees presented only included taxa
that are recognized in Chapter 5 and 6. The settings for the analyses followed the option
of a full heuristic search with the starting tree(s) being obtained via stepwise addition
and random addition of sequences (Maddison 1991) over 1000 replicates. One tree was
held at each step during stepwise addition. The tree-bisection-reconnection option was
implemented and Farris et al.’s (1996) jackknife measure was used to assess branch
support over 1000 pseudoreplicates.
Bayesian analyses of the combined mitochondrial and nuclear DNA characters
for francolins (5116 bp) and spurfowls (5170 bp) were implemented in MrBayes 3.1.2
(Ronquist and Heulsenbeck 2003). The francolin and spurfowl Bayesian analyses
excluded one taxon each, O. r. spilogaster and P. castaneicollis atrifrons, respectively
due to unavailability of DNA characters. Over and above parsimony analyses, the use of
a Bayesian approach was to investigate the influence of mixed models (parameter-rich)
283
on the phylogeny. Modeltest 3.7 (Posada and Crandall 1998) was used to search for the
best-fit model for each molecular marker and each codon data partition for CYTB and
ND2. The parameters for the model considered were those under the Akaike
Information Criterion (AIC). The best-fit model of nucleotide substitution suggested for
all the data partitions was the GTR+I+G model. These analyses involved two
independent runs (with random starting trees) and they proceeded without a molecular
clock being imposed upon the rate of sequence evolution.
The Markov Chain Monte Carlo (MCMC) for the spurfowl and francolin data
set was run for 3 x 106 and 1.5 x 106 generations, respectively, sampling every 100th
generation using four chains (one cold and three heated). Tracer v 1.5 (Rambaut and
Drummond 2007) was used to check the ‘burn-in’ phase and the state of convergence of
the two parallel runs: 300 000 and 150 000 trees from the burn-in phase for the
spurfowls and francolins, respectively, were discarded and convergence was attained
when the average deviation of split frequencies converged towards zero. Finally, the
50% majority rule consensus of sampled trees (excluding the burn-in) was obtained
with posterior probabilities (PPs) appearing above nodes.
Results
Ancestral state reconstructions
Even though the parsimony ancestral area and habitat reconstructions of francolins and
spurfowls were performed, the results were used for comparison with those for the
Bayesian reconstructions irrespective of the differences inherent in the methodology of
these two approaches. The results discussed in subsequent sections are based solely on
Bayesian reconstructions.
284
Francolins – area and habitat
Figure 7.1 and 7.2 respectively show the Bayesian reconstructions through which the
ancestral area and habitat among Asian and African francolins were traced. The
marginal probabilities are assigned to each node and these are estimated based on the
regional and habitat data that were incorporated into the input file and they indicate the
confidence that each state can be assigned at a particular node. Figure 7.1 showed mixed
results, e.g. some nodes such as 1, 7, 9, 10, 11, 12 have probabilities above 0.90; node 2,
6, 14 were assigned with probabilities between 0.80 and 0.90 and the rest of the nodes
showed great uncertainties. The root, node 1, points to southern Asia being the ancestral
area that gave rise to all other francolins. Node 3 links the Asian francolins and the rest
of Africa, but with very low probabilities (0.52 and 0.36) being assigned to the rest of
Africa and Asia, respectively. The immediate link of Asia and Africa is at node 4, which
links the Asian O. gularis/O. pondicerianus and the African Ortygornis clade (Fig. 7.1).
The probabilities are very low with the proportions 0.48 and 0.40 pointing to southern
Africa and southern Asia respectively as being the ancestral areas. These species
possibly evolved in scrubby grassland habitats (probabilities of 0.99).
Afrocolinus lathami is the only West African (and forest endemic) species
supported with a probability of 0.99. This species probably made its way to the forest
habitat from a scrubby grassland ancestor. The Red-tailed and the Red-winged clade
(node 6) are truly southern African clades supported with 0.88 probability of which the
probability assigned to their ancestral habitat (Fig. 7.2) is uncertain. Within the Redtailed Group clade (node 7), members share a most recent common ancestor in southern
African (0.99 probability) that likely thrived in wooded grasslands from which these
clades diversified in other parts of Africa i.e., two species spread to West Africa (P.
285
albogularis and P. spinetorum), one to Central (P. schlegelii) and two to East Africa (P.
maharao and P. hubbardi). When these species diversified in different parts of Africa
they retained the ancestral preference of wooded habitats with a slight difference being
that P. albogularis occurred in mesic woodlands, whereas the other three species
occurred in arid woodlands.
Seemingly, the most recent common ancestral area of the Red-winged taxa
(node 10) was in southern Africa (0.91 probability), although this clade underwent its
greatest diversification in East Africa. Most species inhabit open grasslands, three
species, with two taxa (S. psilolaema psilolaema and S. ellenbecki) that went into
montane grasslands in East Africa and one species S. afra made its way to montane
grasslands in southern Africa. Scleroptila streptophora spread to the scrubby mesic
habitat in Cameroon and around Lake Victoria that differs from that of its former group
member O. sephaena that mainly occupies arid scrubby grasslands and woodlands.
Thus, the Red-winged Group diversified into mesic, arid, and montane grasslands. The
Striated members spread from the south to East Africa and the probability assigned to
this node is 0.80. O. grantii and O. r. spilogaster occupy arid scrubby habitat. The
nominate O. r. rovuma occupies mesic scrub habitat. It seems that, in Africa, francolins
diversified out of southern Africa, spreading mainly to the East, then West and P.
schlegelii (inhabitant of savanna and arid woodlands) evolved in central Africa where it
remained isolated. The Red-tailed diverged to occupy both mesic and arid woodland
habitats.
286
Spurfowls – area and habitat
The likelihood reconstruction for area (Fig. 7.3), unlike that for habitat (Fig. 7.4)
exhibited ambiguity at several nodes with respect to the ancestral distribution of
spurfowl. Although with a low probability of assignment the root node suggests
southern Africa as the potential ancestral area. Habitat reconstruction (Fig. 7.4) showed
some support (0.64) for tracing all spurfowls back to forest (node 1).
The Bare-throated Group/P. clappertoni/P. leucoscepus/P. swainsonii (node 7
which combines node 8 and 9) is assigned with a low probability values with east Africa
(0.47) being favoured to be an ancestral area over southern Africa (0.40). With regard to
habitat, taxa that thrive in mesic woodlands (P. rufopictus, P. cranchii, P. a. afer, P. a.
humboldtii and P. l. leucoscepus, P. l. infuscatus, P. swainsonii) are reconstructed as
occupying savanna and arid woodlands with high probability (0.95; Fig. 9.12). Node 4
joins the Montane Group (forest inhabitants) to P. clappertoni (savanna and arid
woodland), P. a. afer (mesic woodlands), P. c. sharpii, P. harwoodi and the Barethroated Group, and is reconstructed to likely have had a forest ancestor with a
probability of 0.70. The most recent common ancestor of P. b. bicalcaratus, P. b.
adamauae, P. icterorhynchus, ‘ayesha’ has a savanna and arid woodland ancestor (0.87
probability).
In Fig. 7.4, node 10 connects the Scaly Group and the southern Vermiculated
clade has the highest probability of 0.72 and this is in support of a forest ancestor, while
the most recent common ancestor of the Scaly Group is supported with 0.99 probability
to have lived in forest in southern Africa (0.68). The probability of 0.89 supports the
southern African origin of the Vermiculated species P. adspersus, P. capensis, P.
hildebrandti, and P. natalensis. Pternistis hildebrandti has its distribution in southern
287
and eastern Africa. Most of the nodes on the ancestral area reconstruction showed great
uncertainty in the assignment of marginal probabilities. It is, however, clear that
colonization of areas happened more than once with the exception of North Africa
where P. b. ayesha has an isolated distribution in Morocco in the macchia type of
habitat. Areas too were colonized more than once except for the lowland fynbos, a home
for P. capensis.
Spatial analysis of vicariance
The spatial analysis of vicariance of francolins yielded two reconstructions from which
a consensus reconstruction analysis was performed yielding the best score associated
with a cost of 9.0, and 320 hits of 1000 iterations. Thirty optimum disjunct sister groups
were recovered and these disjunctions are mapped at nodes on the parsimony tree (Fig.
7.5). The spurfowl distribution ranges resulted in one reconstruction with 28 optimum
disjunctions (Fig. 7.6) accompanied by a best score with a cost of 6.0 for 471 hits of
1000 iterations.
The history of evolution of distribution ranges
Francolins - The phylogenetic results confirmed that the Asian francolins are the most
basal of all francolins and the main geographic barriers exist around the Indian subcontinent (Fig. 7.7a-b). The first break being near the Windhya Range of Mountains
(Fig. 7.7a), that are located in Rajasthan, south of Delhi, between F. francolinus and F.
pictus (node 1, Fig. 7.5). According to Hall (1963), Rajasthan has no clear natural
barrier. Thus, it remains uncertain whether the Windhya Range and possibly the Ganges
tributaries should be considered effective physical barriers. The Ganges River in
288
northeast India separates Francolinus pictus from F. pintadeanus (Fig. 7.7b, node 2 in
Fig. 7.5) as well as the Asian/African francolin clade represented by all taxa restricted to
the sub-Saharan region.
The Naga Hills in Burma in addition to the Chin Hills and the Manipur Range as
alluded to by Hall (1963) act as the possible barriers between F. pintadeanus and the
Asian+African francolin clade (Fig. 7.7c, node 3 in Fig. 7.5). The disjunction (node 4 in
Fig. 7.5) between the clade that includes members of Ortygornis and the African clade
circumscribed by A. lathami/Peliperdix/Scleroptila is associated with a barrier that runs
along the Rift Valley system down to the south and in the west meets the Congolian
forest, setting itself as the main barrier in the south separating the African Ortygornis
taxa on the east, from the rest of the African francolins in the west (Fig. 7.7d). Another
barrier is set by the Limpopo River in the south separating the African Ortygornis clade
from specifically S. levaillantii and S. afra in the area where they overlap (Fig. 7.7d).
The barrier between the O. gularis/O. pondicerianus clade and the O.
sephaena/grantii/rovuma clade (node 5 in Fig. 7.5) is quite ambiguous, because in this
case the barrier could be any intervening feature, possibilities include the: Arabian Sea,
and the large rivers and mountain ranges in Pakistan and Iran (Persia) (Fig. 7.7e). The
disjunction that exists within the clade that includes O. pondicerianus + O. gularis and
O. sephaena/grantii/rovuma clade might be an indication of colonization of Africa by
the clade ancestor from Asia or a return of the clade ancestor to Asia. In colonizing
Africa, the clade ancestor landed itself on the east side of the Rift Valley system thereby
inhabiting arid habitats. The O. pondicerianus/O. gularis/O. sephaena clade diverged c.
7.2 mya (Chapter 2). The estimated age for the formation of the Rift Valley is c. 6 mya.
It could be speculated that the clade ancestor possibly crossed the Rift system
289
diversifying in central and west Africa occurring along the mesic Congolian and Upper
Guinea forests.
Within the O. sephaena/grantii/rovuma clade, the Zambezi River acts as the
major barrier that separates O. sephaena (in the south) and the O. grantii clade in the
north (Fig. 7.7f). The Rufiji River separates the southern African O. r. rovuma from O.
grantii (Fig. 7.7h) whereas the northeastern O. r. spilogaster is separated from O.
grantii by the Ganane Basin and possibly the Ethiopian Rift (Fig. 7.7g). In addition, the
difference in habitat, xeric versus mesic scrub, seems to have an influence within the O.
sephaena/grantii/rovuma clade. The xeric habitats differ in the amount of mean annual
rainfall received, ~400-800 mm where O. sephaena and O. grantii occurs, ~200-400
mm where O. r. spilogaster occurs (Clark 1967). The mesic scrubs receive the mean
annual rainfall of ~800-1600 mm and this is occupied by O. r. rovuma. The mesic
versus xeric habitat preference seems to set itself as the possible barrier between the
southern O. r. rovuma and its northern counterpart O. r. spilogaster (Fig. 7.7h). There is
a disjunction (node 9 in Fig. 7.5) between the A. lathami/Peliperdix and Scleroptila
clades (Fig. 7.7i). The Congolian and the Upper Guinea forest emerged to be effective
breaks between the two clades. While there is no detected barrier to separate A. lathami
and the Peliperdix clade, the west African forest population A. l. lathami is set apart
from the central A. l. schubotzi by the Congo Basin (Fig. 7.7j). Both taxa occur in high
rainfall areas of above 3200-4000 mm annually (Clark 1967).
Due to the presence of a disjunction in just one of the reconstructions, there was
no barrier detected between node 11 and node 14 in Figure 7.5 which links the P.
schlegelii/P. albogularis and P. coqui clade. Therefore, there are no definitive barriers
to infer between P. schlegelii and P. albogularis (Fig. 7.7k, node 11 in Fig. 7.5),
290
between P. a. dewittei and P. albogularis albogularis/P. a. buckleyi clade (Fig. 7.7l),
node 12 in Fig. 7.5) and also between P. a. albogularis and P. a. buckleyi (Fig. 7.7m).
Both clades occur in habitats that receive similar amounts of annual rainfall (800-1600
mm).
The occurrence of P. a. dewittei on montane wooded grasslands south of the
Congo forest (Cotterill 2006) could be attributed to the fact that during the time when
Africa was wet the forest may have formed a continuous belt from west Africa through
to central and down to south central Africa (Clark 1967, Crowe and Crowe 1982).
During this time the proto-albogularis may have had a continuous distribution along the
forest belt. During a dry period (possibly glaciation), forest contracted and formed
pockets leaving P. a. dewittei being isolated south of the Congo. Morphologically,
there are subtle differences between P. a. albogularis and P. a. dewittei which relate to
the size and extent of barring on the underparts. The predicted mean annual rainfall in
areas where P. a. albogularis and P. a. dewittei occur is 100-1400 and 1400-1800 mm,
respectively.
Within the clade that includes P. stuhlmanni/P. kasaicus and P. c. coqui/P. c.
vernayi/P. c. ruahdae/P. hubbardi/P. maharao/P. spinetorum (node 14 in Fig. 7.5), the
Rovuma River was detected as an effective barrier that exists on the east at least
between P. stuhlmanni/P. hubbardi and P. maharao in the south the Limpopo River for
a barrier separating P. stuhlmanni from P. c. coqui/P. c. vernayi (Fig. 7.7n). P. kasaicus
is separated from the other taxa in this clade, but there is no definite barrier attributable
to this disjunction. Peliperdix kasaicus and P. stuhlmanni both occur in mesic
conditions which differ in areas where the mean annual rainfall is 1000-1400 mm. The
mesic woodland P. c. ruahdae (800-1600 mm of mean annual rainfall) on the west is
291
separated from P. hubbardi, which occurs in much more xeric habitat to the east around
Lake Victoria and Lake Tanganyika (400-800 mm of mean annual rainfall) (Fig. 7.7o).
The southern African P. c. coqui and P. c. vernayi are separated by the
Okavango swamp (Fig. 7.7p) with P. c. ruahdae and P. c. vernayi being separated by
the mesic and arid habitat respectively (Fig. 7.7q). The area between the Ethiopian and
the East African Rift might possibly be the barrier separating P. hubbardi from P.
maharao in Malawi (400-800 mm of mean annual rainfall), whereas all the remaining
taxa occur in arid savanna (including P. spinetorum 200-400 mm of mean annual
rainfall) (Fig. 7.7r). Habitat separation definitely has an influential role among these
disjunctions.
The first divergence within the Scleroptila clade led to the separation of S.
streptophora/S. levaillantii from the rest of the Scleroptila taxa (Fig. 7.7s, node 21 in
Fig. 7.5). The Scleroptila taxa mainly occur in xeric environments. In the south the
Limpopo and Zambezi Rivers act as barriers between S. levaillantii and the other
Scleroptila taxa (Fig. 7.7t), whereas northern Lake Tanganyika and the Rufiji River
separate the east African S. streptophora population from the other Scleroptila taxa
(Fig. 7.7u). Mesic habitat links the two west and east S. streptophora populations that
occur on either side of the Congo Basin. It is quite strange that not even habitat links the
disjunct west and east African S. streptophora populations with those of southern
African S. levaillantii. Scleroptila streptophora has a puzzling distribution being found
in grasslands in Cameroon where the annual rainfall reaches ~2600-3200 mm and in
east Africa (800-1600 mm) and it was included in the Striated Group together with O.
sephaena by Hall (1963). From examination of museum skins of the two populations of
S. streptophora, there were no differences observed. The two species S. streptophora
292
and O. sephaena are distinct morphologically and at least ecologically (Hall 1963), but
the east African population shares almost the same type of habitat with that of O.
sephaena. Scleroptila streptophora occurs in mesic habitat (at an altitude ranging from
1050–1200 m in Cameroon and 600-1800 m in scrub-covered hillsides in east Africa –
del Hoyo et al. 1994), whereas S. levaillantii is found in xeric to mesic habitats (Fig.
7.7t) at an altitude of 600-2 250 m (Hall 1963, Snow 1978) and the mean annual rainfall
of 600-800 mm.
The disjunction between the S. psilolaema and S. uluensis/S. shelleyi is not
associated with any detectable barrier. However, the African Great lakes (Lake Victoria,
Lake Tanganyika, Lake Malawi) form a barrier that separate S. psilolaema from S.
finschi (Fig. 7.7v), with S. finschi being separated from S. whytei/S. jugularis (Fig.
7.7w, node 25 in Fig. 7.5). The African Great lakes appear to be effective barriers
between S. uluesis and S. shelleyi (Fig. 7.7x), whereas S. gutturalis and S. uluensis are
separated by the Ethiopian Rift, with but taxa occurring in xeric grassland habitats (Fig.
7.7y). Among members of the S. shelleyi/S. levaillantoides and the S. ellenbecki clade,
the distribution ranges of S. shelleyi and S. levaillantoides are maintained by two
breaks, Lake Malawi in the north and the Limpopo River in the south (Fig. 7.7z).
Scleroptila shelleyi and S. levaillantoides overlap in the southern part of their ranges
and there was no detectable barrier. The mean annual rainfall is the area according to
Clark (1967) is 600-800 mm, i.e. in areas where S. levaillantoides, S. shelleyi, S.
levaillantii occur and 400-600 mm in areas where S. afra occurs. The impact of the
African Great Lakes is also observed between S. c. crawshayi and S. c. kikuyuensis (Fig.
7.7aa, node 30 in Fig. 7.5). The ranges of the taxa that belong to S. ellenbecki have no
barriers detected among them and they are completely disjunct. The Zambezi and
293
Limpopo valleys appear to be effective barriers among some of the species and this
confirms the findings in Benson et al. (1962).
Spurfowls - the results in Figure 7.8a reveal that the most basal disjunct node (1 in Fig.
7.6) is associated with a barrier that separates the clade comprising Ammoperdix heyi
from the rest of the African spurfowls. The Red Sea emerges to be the main barrier with
some overlap along the Bab al Mandab Mountains that lie between the Red Sea and the
Gulf of Aden. The barrier may have resulted from the opening of the Red Sea. The
barrier between A. heyi and P. asiatica (Fig 9.8b, node 2 in Fig. 7.5) could be anything
between Saudi Arabia and India with the possible barriers being the Arabian Sea, and
the rivers and mountain ranges in Pakistan and Iran.
It is possible that dispersal route may have started in Asia, into Saudi Arabia and
subsequently into Africa or vice versa. It is somewhat difficult to explain clearly how
Africa was colonized by the ancestral-spurfowl, but to some extent it could have been
that the ancestral-spurfowl dispersed from Asia along the desert of Saudi Arabian coast
and settled in the desert in the south western part of Africa i.e., in Namibia and south
western Angola where P. hartlaubi occurs (Fig. 7.8c). The ancestral-hartlaubi
diversified to other parts of Africa from the desert habitat. Whereas P. hartlaubi is said
to occur in savanna and arid woodlands where the mean annual rainfall is below 100
mm (Clark 1967), it has been observed in boulder-strewn slopes and rocky outcrops
(Komen 1987, Sinclair and Ryan 2005). The subsequent Africa spurfowl radiation
involved the evolution of the montane lineage that includes species that were confined
to the mountain forests. The first break, the African Great Lakes is detected between the
clades comprising P. nobilis and P. swierstrai (node 4 in Fig. 7.6) with habitat acting as
294
another barrier (Fig. 7.8d). Pternistis nobilis and P. camerunensis inhabit the mesic
montane forest occurring on the extreme side of the Congo Basin, whereas the other
taxa that they are being compared with in this context are found in arid savannas and
they are both separated by the Congo forest (Fig. 7.8e). Pternistis nobilis occurs in an
area where the mean annual rainfall is ~ 800-1600 mm, whereas P. camerunensis is
endemic to an area which receives an average of more than 3200 mm of rain annually.
The disjunction exists between the Angolan highlands endemic species P. swierstrai
and the north eastern P. c. castaneicollis clade, and this separation is mainly promoted
by habitat differences, mesic versus xeric environment, even though they all occur in
forests (Fig. 7.8f). The separation seems large such that vegetation alone may not
explain the present distribution.
The main barrier between the P. c. castaneicollis clade and the other spurfowls
(especially P. jacksoni as other ranges are removed) is the area that forms part of the
Rift Valley system between the Ethiopian and the East African Rift as they are generally
found in similar habitat (Fig. 7.8g, node 7 in Fig. 7.6). There is a slight overlap between
the clades comprising P. c. castaneicollis and P. ochropectus along the Ethiopian Rift
(Fig. 7.8h). The distribution ranges of P. c. castaneicollis and P. c. atrifrons are
possibly maintained by part of the Ethiopian Rift (Fig. 7.8i). Pternistis erckelii and P.
ochropectus prefer similar habitat, they are montane forest species, with the exception
of P. erckelii, which sometimes wanders into montane grasslands. Their distribution
ranges are also maintained by part of the Ethiopian Rift (Fig. 7.8j).
The disjunction that exists between P. ahantensis and P. icterorhynchus cannot be
explained geographically because the sister groups were removed (Fig. 7.6). Figure 7.8k
shows the detected barrier between P. ahantensis and P. griseostriatus clade (node 12 in
295
Fig. 7.6) which cuts through east of Lake Victoria, along Lake Tanganyika (where they
slightly overlap) separating the two clades into mesic versus somewhat xeric habitat
taxa with the Congolian and Upper Guinea forest being effective barrier between the
northern and the southern taxa. The Niger River strongly separates P. ahantensis which
occurs in the far west in the Upper Guinea forest from P. squamatus (Fig. 7.8l) which is
mainly confined to the Congolian forest while P. s. squamatus is separated from P. s.
schuetti by the area between the Ethiopian and the East African Rift (Fig. 7.8m).
Pternistis griseostriatus is endemic to Angola and occurs in riverine forest (Sinclair and
Ryan 2005). The P. griseostriatus clade also include the southern Vermiculated species
P. capensis, P. natalensis, P. hildebrandti and P. adspersus that occur along the arid
belt which acts as a barrier between them and P. griseostriatus (Fig. 7.8n). In Figure
7.8o, it is apparent that P. capensis and P. natalensis clades (node 16 in Fig. 7.6) are
dissected by the Okavango swamps along the area of slight overlap whereas they are all
found in the intervening arid habitats. Between P. capensis and P. adspersus the Orange
River emerges as an effective barrier (Fig. 7.8p) whereas the Zambezi River and Lake
Malawi separate P. hildebradti from P. natalensis (Fig. 7.8q). Lake Malawi also acts as
a barrier between P. h. hildebrandti and P. h. fischeri along the area where they overlap
slightly (Fig. 7.8r).
Although some parts of the mesic and arid savannas are covered, the Upper
Guinea and the Congolian forest act as a barrier between the clades that include P.
icterorhynchus and P. clappertoni/P. leucoscepus along the area where they overlap
(Fig. 7.8s, node 20 in Fig. 7.6) with P. icterorhynchus and P. bicalcaratus being
separated from each other by the Niger River and Lake Chad (Fig. 7.8t). The Volta and
Niger River likely both play a role in breaking P. b. adamauae from P. b. bicalcaratus
296
clade (Fig. 7.8t) with the isolated Moroccan P. b. ayesha being strongly separated from
P. b. bicalcaratus by the Sahara Desert (Fig. 7.8t). The spurfowls occur in almost the
rest of the sub-Saharan region with the exception of P. bicalcaratus that expanded its
range to Morocco in north Africa where ayesha occurs (Fig. 7.8a).
Spurfowl typically occur in various montane/lowland habitats as well as
xeric/mesic habitats that vary in mean annual rainfall. The Ethiopian Rift may have
been effective in maintaining the break between the P. clappertoni and P. leucoscepus
clades (Fig. 7.8u). The eastern branch of the East African Rift and possibly the Rufiji
River separate the distribution range into P. leucoscepus clade in the north and P.
swainsonii clade in the south (Fig. 7.8v, node 25 in Fig. 7.6), with P. l. leucoscepus and
P. l. infuscatus being separated by the Ethiopian Rift (Fig. 7.8w). No barrier was
detected between the P. swainsonii and the P. rufopictus clade, whereas P. cranchii is
separated from P. rufopictus by the Rift Valley system in the east and the Congolian
forest in the west (Fig. 7.8x).
Pternistis cranchii primarily occurs in mesic woodlands. Habitat seems to be
playing a role in maintaining the break between P. rufopictus and P. a. afer (Fig. 7.8y).
According to Hall (1963), they all occur along the acacia belt that is differentiated by
different Acacia species. Pternistis rufopictus occupied largely the mesic part with P. a.
afer and P. a. humboldtii getting into the dry and mesic parts. The Zambezi River
strongly separates the two subspecies P. a. afer and P. a. humboldtii (Fig. 7.8z).
297
Discussion
Reconstruction of ancestral area and habitat
Francolins and spurfowls
On the basis of the results, it was difficult to determine where the ancestor of francolins
could have lived based on the evolution of area, whereas in terms of habitat, it could be
concluded that the ancestor of francolins likely evolved in scrubby grassland. Even
though the spurfowl area reconstruction had the poorest probabilities assigned to nodes,
habitat reconstruction suggests that the ancestral distribution was restricted to forest.
However, it was difficult to draw definitive conclusions on the basis of the area and
habitat maximum likelihood reconstructions due to uncertainties in nodal probabilities
assigned to states among francolins and spurfowls.
The two biogeographic methods, one focusing on ‘detecting barriers’ and the
other one focusing on ‘mapping of ancestral areas and habitat’ both seek to explain the
history of distribution ranges of taxa, but differ in their underlying theoretical
applications. Even though both applications made use recovered phylogeny, the
ancestral reconstructions seem uncertain when there is poor phylogenetic resolution. In
the context of francolins and spurfowls, the spatial analysis of vicariance results
provided a much better picture on the history of their distribution patterns.
Spatial analysis of vicariance
Francolins
Some barriers that were detected among the Asian species are geologically and
geographically associated with the Indian sub-continent. While it may be possible that
F. francolinus and F. pictus are ecologically separated, the Wandhya Range seems
298
unlikely to be a significant barrier. The Asian species generally have different habitat
preferences. Ortygornis gularis is confined to marshes and reeds in the plains of the
Ganges River, O. pondicerianus occurs in the Gulf of Oman in the plains of India, F.
pictus inhabits the plains of India, whereas F. pintadeanus occurs in marshes and hills
such as the Naga and Chin and Manipur range.
Francolinus francolinus prefers mountainous areas for example, the Himalayan
Mountains in Nepal, and extending to the west reaching Iran and Afghanistan. The
results here suggest that habitat type could have been an effective ecological break
between these F. francolinus and F. pictus. The Naga and Chin Hills as well as the
higher ranges of Manipur are effective barriers along most of the boundary between F.
francolinus and F. pintadeanus even though the two species might be expected to meet
along the coastal strip.
It could be possible that speciation may have started around the India–Eurasia
border even though the actual age of this event has always been uncertain. The latest
evidence points to c. 40 mya (Bouilhol 2013) while the split between
Gallus/Bambusicola/francolins and Coturnix/Alectoris/spurfowls is at around 33.6 mya
which is Oligocene. Plate tectonics may have led to the formation of different
topographical features such as the Himalayan Mountains and the Kūhha-ye-zāgros in
Iran. At the same time during collision the ancestral species may have been able to
disperse into Africa possibly cutting through south eastern Saudi Arabia into Africa.
The unfavourable era may have pushed the range of the ancestral species southwards,
isolating the ancestral stock of the three species in south western Asia that is where F.
francolinus currently occurs, southern India (F. pictus) and south eastern Asia where F.
pintadeanus occurs.
299
The Oligocene to Miocene is said to have marked the start of generalized
cooling, with glaciers forming in Antarctica for the first time. The increase in ice sheets
led to a fall in sea level while the tropics diminished, giving way to cooler woodlands
and grasslands. Diversification of habitat has occurred with A. lathami being the only
francolin species that is endemic to primary forest. Among the Scleroptila taxa, there is
a somewhat weaker, but possible retreat of some of the species to montane areas
possibly seeking the favourable habitat. Examples of these are the southern African S.
afra, the east African S. p. psilolaema, S. p. theresae and P. ellenbecki.
Scleroptila afra and S. levaillantii have sympatric distributions, but they are
divided altitudinally (del Hoyo et al. 1994, Little and Crowe 2000, Sinclair and Ryan
2005). The Peliperdix taxa are all found in arid or mesic wooded savannas. Habitat
preference seems to play an influential role among members of the P. coqui clade.
They all occur along the dry belt differing in the amount of annual rainfall that is
received in those areas with the exception of P. c. kasaicas and P. stuhlmanni which
occur in mesic habitats with mean annual rainfall of 800-1600 mm. The Rift Valley
system which includes the Great African lakes (Lake Victoria, Lake Tanganyika, Lake
Malawi), the Upper Guinea and the Congolian forest, major rivers such as the Limpopo,
Zambezi and Rovuma may have played a crucial role in maintaining the distribution
ranges of francolins.
The two streaked subspecies, the northern O. r. spilogaster (xeric scrubs) and
the southern O. r. rovuma (mesic scrubs) are separated by habitat types which are
characterized by difference in the amount of rainfall received in a year.
300
Spurfowls
A number of nodes have disjunctions that can be explained by geography and had
detectable barriers associated with them. It is possible that the tectonic effect that
resulted from the collision of Africa with Asia may have caused the clade ancestor to
disperse along the Saudi Arabian coast into Africa moving to the desert in southwest
Africa. Pternistis hartlaubi is the most basal spurfowl which probably is a paleoendemic species that occurs in the desert of Namibia and southwestern Angola where
the mean annual rainfall is below 100 mm. The India-Saudi Arabia-Africa route and
vice versa is quite possible with difficulty arising with regard to understanding how
different parts of Africa including different habitats were colonized as the spurfowls
diversified within various bushy habitats. However, it seems possible that, in Africa,
diversification may have been initiated in southwestern Africa where P. hartlaubi
occurs and subsequently followed by the radiation into forests of montane areas (Clark
1967). During this dry period the range of forest was reduced leading to montane forest
pockets being occupied by the few species that preferred these kinds of habitat. Some
species such as P. squamatus, P. ahantensis and P. griseostriatus remained in lowland
forests, while others remained in other lowland habitats such as mesic and arid bushy
and wooded grassland habitats and diversified there. Habitat type is clearly linked to the
rainfall amount that each habitat type receives and this clearly separate out parts even
within species.
The Rift Valley system is acting as an effective barrier that is maintaining the
ranges of most of the spurfowl distributions, for example among the north eastern and
east African montane spurfowls, among the Bare-throated taxa P. leucoscepus/P.
swainsonii/P. rufopictus/P. afer, between P. clappertoni/P. harwoodi and the Bare-
301
throated clade, as well as between the P. squamatus/P. ahantensis and P.
griseostriatus/southern Vermiculated species clade. Major rivers such as the Niger,
Volta, Zambezi, Orange and Rufiji River have played significant roles as barriers. The
Upper Guinea and Congolian forest, Okavango swamp and also the mesic versus arid
habitats were found to be effective barriers.
Conclusions
The history of taxa can never be explained with ease because history is generally
unknown. The colonization of Africa by the ancestral species may have been through
dispersal which resulted in the formation of disjunct distributions and the somewhat
rapid diversification of francolins and spurfowls in Africa. Even though francolins and
spurfowls evolved at almost the same time that is, during the Miocene francolins (c. 7.6
mya) are slightly younger than spurfowls (c. 8.7 mya). However, the main divergence
between these two lineages occurred at around 33.6 mya in the Oligocene and this is
probably the initial speciation of genus Francolinus around India. The major physical
breaks that may have maintained the distributions and at the same time promoted
speciation among the African francolins were the Rift Valley system which includes the
Great African lakes (Lake Victoria, Lake Tanganyika, Lake Malawi), Lake Chad, the
Upper Guinea and the Congolian forest, and major rivers such as Limpopo, Zambezi
and Rovuma River. The diversity in habitat for example, xeric/mesic savanna versus
mesic savanna, montane and lowland grasslands and forest habitat has also played an
important ecological role in facilitating diversification among francolins. Among
spurfowls, similar barriers as those among francolins were inferred with the addition of
302
the Niger and Rufiji River, Okavango Swamp and Sahara Desert acting as barriers
between some taxa.
In conclusion, the overall possible biogeographic explanation among francolins
is that the ancestral species may have dispersed from Asia to Africa and later possibly
returned to Asia. If Asian and African species are monophyletic, it could be fitting that
during the collision of Asia and Africa some individuals remained in Asia while others
remained in Africa resulting in the two groups differentiating with subsequent
speciation occurring on both continents. In this case, the Asian Ortygornis gularis clade
is sister to the O. sephaena complex clade and this fits the former explanation.
The direction of dispersal is difficult to determine for the spurfowls. One
possible explanation on the origin of the spurfowls is that the ancestor may have
dispersed from Asia, into the desert of Saudi Arabia along its coast and then into Africa
where it eventually arrived in the desert in south western Africa. Generally, there is
support for Hall’s (1963) hypothesis on the evolution of the genus Francolinus having
dispersed from Asia where its related genera occur, as opposed to Crowe and Crowe
(1985) and Bloomer and Crowe’s (1998) hypothesized African origin of the genus
Francolinus. Despite that the ancestral area and habitat from the maximum likelihood
reconstructions were highly undecided, the spatial analysis of vicariance results at least
clearly demonstrated that the ancestral francolin was Asian and they lived in dry and
mesic scrubby and wooded grassland habitat.
It is important to note that since the two approaches (spatial analysis of variance,
ancestral character reconstructions) are dependent on the phylogeny, poor phylogenetic
resolution influences the pattern. In other words, phylogenetic signal plays a core role in
the outcome of the analyses and in the absence of well-resolved phylogenies, informed
303
ancestral patterns cannot be traced and the colonization route can hardly be determined.
The maximum likelihood ancestral area and habitat reconstruction provided little and
decisive insight into the history of the distribution of ranges of francolins and spurfowls,
and a genomic approach to data collection may be required to fully resolve spurfowl
relationships.
304
Francolinus francolinus
F. pictus
F. pintadeanus
Ortygornis gularis
O. pondicerianus
4
14
O. grantii
O. r. rovuma
O. sephaena
Afrocolinus lathami lathami
A. l. schubotzi
Peliperdix a. albogularis
P. a. buckleyi
8
P. dewittei
P. schlegelii
P. c. coqui
7
P. c. ruahdae
9
P. c. vernayi
P. hubbardi
1
P. maharao
2
P. spinetorum
3
P. stuhlmanni
P. kasaicus
5
Scleroptila afra
6
S. ellenbecki
S. crawshayi crawshayi
13
12
11
S. jugularis
S. levaillantoides pallidior
Scleroptila l. levaillantoides
S. shelleyi
10
Character: Area
S. gutturalis
S. uluensis
Southern Asia
S. finschi
West Africa
Scleroptila whytei
Southern Africa
S. p. psilolaema
Central Africa
S. p. theresae
S. crawshayi kikuyuensis
East Africa
S. levaillantii
S. streptophora
Figure 7.1. The Bayesian reconstruction of ancestral area of francolins inferred from Bayesian
topology and branch-lengths. Numbers 1-14 identifies the nodes they are associated with. The
marginal probabilities are not mapped on the tree but references to them are in the main text.
305
Francolinus francolinus
F. pictus
F. pintadeanus
Ortygornis gularis
O. pondicerianus
4
14
O. grantii
O. sephaena
O. r. rovuma
Afrocolinus lathami lathami
A. l. schubotzi
Peliperdix a. albogularis
P. a. buckleyi
8
P. dewittei
P. schlegelii
P. c. coqui
7
P. c. ruahdae
P. hubbardi
P. c. vernayi
I
1
P. maharao
2
P. spinetorum
3
P. stuhlmanni
P. kasaicus
5
Scleroptila afra
6
S. ellenbecki
13
12
S. crawshayi crawshayi
S. l. levaillantoides
S. l. pallidior
11
Character: Habitat
Open grasslands
S. jugularis
S. shelleyi
10
S. gutturalis
S. uluensis
Scrubby grasslands
S. finschi
Wooded grasslands
S. whytei
Montane grasslands
S. crawshayi kikuyuensis
Forest
S. p. theresae
S. p. psilolaema
S. levaillantii
S. streptophora
Figure 7.2. The Bayesian reconstruction of ancestral habitat of francolins inferred from Bayesian
topology and branch-lengths. Numbers 1-14 identifies the nodes they are associated with. The
marginal probabilities are not mapped on the tree but references to them are in the main text.
306
Pternistis hartlaubi
P. bicalcaratus bicalcaratus
2
P. b. adamauae
P. icterorhynchus
P. b. ayesha
P. camerunensis
P. nobilis
P. erckelii
5
6
P. ochropectus
P. castaneicollis
P. swierstrai
P. jacksoni
P. clappertoni clappertoni
4
P. c. sharpii
8
P. harwoodi
1
P. leucoscepus leucoscepus
7
P. l. infuscatus
3
P. swainsonii
P. rufopictus
9
P. cranchii
P. afer humboldtii
P. a. afer
P. squamatus squamatus
Character: Area
11
Southern Africa
P. s. schuetti
P. griseostriatus
West Africa
North Africa
P. ahantensis
10
Central Africa
P. hildebrandti hildebrandti
East Africa
12
Pternistis h. fischeri
P. natalensis
P. capensis
P. adspersus
Figure 7.3. The Bayesian reconstruction of ancestral area of spurfowls inferred from Bayesian
topology and branch-lengths. Numbers 1-12 identifies the nodes they are associated with. The
marginal probabilities are not mapped on the tree but references to them are in the main text.
307
Pternistis hartlaubi
P. bicalcaratus bicalcaratus
2
P. b. adamauae
P. icterorhynchus
Pternistis b. ayesha
P. camerunensis
P. nobilis
5
Pternistis erckelii
6
P. ochropectus
P. castaneicollis
P. swierstrai
Pternistis jacksoni
P. clappertoni clappertoni
4
P. c. sharpii
8
1
P. harwoodi
Pternistis leucoscepus leucoscepu
7
P. l. infuscatus
3
P. swainsonii
P. rufopictus
9
P. cranchii
P. afer humboldtii
P. a. afer
P. squamatus squamatus
11
Character: Habitat
P. s. schuetti
P. griseostriatus
Savanna & arid woodlands
P. ahantensis
10
Mesic woodlands
P. hildebrandti hildebrandti
Forest/forest edge
12
Lowland fynbos
P. h. fischeri
P. natalensis
Macchia
P. capensis
Pternistis adspersus
Figure 7.4. The Bayesian reconstruction of ancestral habitat of francolins inferred from
Bayesian topology and branch-lengths. Numbers 1-12 identify the nodes they are associated
with. The marginal probabilities are not mapped on the tree but references to them are in the
main text.
308
Gallus sp.
Francolinus francolinus
Francolinus pictus
Francolinus pintadeanus
Ortygornis pondicerianus
Ortygornis gularis
Ortygornis sephaena
Ortygornis grantii
Ortygornis rovuma spilogaster
Ortygornis rovuma rovuma
Afrocolinus lathami schubotzi
Afrocolinus lathami lathami
Peliperdix schlegelii
Peliperdix dewittei
Peliperdix albogularis buckleyi
Peliperdi albogularis albogularis
Peliperdix stuhlmanni
Peliperdix kasaicus
Peliperdix coqui coqui
Peliperdix coqui vernayi
Peliperdix coqui ruahdae
Peliperdi hurbbadi
Peliperdix spinetorum
Peliperdix maharao
Scleroptila streptophora
Scleroptila levaillantii
Scleroptila psilolaema theresae
Scleroptila psilolaema psilolaema
Scleroptila finschi
Scleroptila whytei
Scleroptila jugularis
Scleroptila uluensis
Scleroptila gutturalis
Scleroptila shelleyi
Scleroptila levaillantoides levaillantoides
Scleroptila ellenbecki
Scleroptila levaillantoides pallidior
Scleroptila afra
Scleroptila crawshayi kikuyuensis
Scleroptila crawshayi crawshayi
Figure 7.5. The francolin parsimony tree showing disjunct nodes which are associated with barriers
(filled black squares) and disjunct nodes indicative of removed distribution (unfilled squares). The
309
unmarked nodes could not be explained geographically due to large overlap in distributions.
Ammoperdix heyi
Perdicula asiatica
Pternistis hartlaubi
Pternistis nobilis
Pternistis camerunensis
Pternistis swierstrai
Pternistis castaneicollis atrifrons
Pternistis castaneicollis castaneicollis
Pternistis ochropectus
Pternistis erckelii
Pternistis jacksoni
Pternistis ahantensis
Pternistis squamatus schuetti
Pternistis squamatus squamatus
Pternistis griseostriatus
Pternistis capensis
Pternistis adspersus
Pternistis natalensis
Pternistis hildebrandti fischeri
Pternistis hildebrandti hildebrandti
Pternistis icterorhynchus
Pternistis bicalcaratus adamauae
Pternistis b. bicalcaratus ‘ayesha’
Pternistis bicalcaratus bicalcaratus
Pternistis clappertoni clappertoni
Pternistis harwoodi
Pternistis clappertoni sharpii
Pternistis leucoscepus infuscatus
Pternistis leucoscepus leucoscepus
Pternistis swainsonii
Pternistis cranchii
Pternistis rufopictus
Pternistis afer humboldtii
Pternistis afer afer
Figure 7.6. The spurfowl parsimony tree showing showing disjunct nodes which are
associated with barriers (filled black squares) and disjunct nodes indicative of removed
distribution (unfilled squares). The unmarked nodes could not be explained geographically due
to large overlap in distributions.
310
(a)
(b)
(c)
Figure 7.7 (a-aa). The range maps of francolins indicating where barriers lie between each
the two circumscribed clades that are being compared. The range in blue is a reference
taxon/clade which is compared to the range of the taxon/clade in red. The yellow lines are
barriers detected, whereas the red and dark blue lines are connecting lines. Green grids
indicate area of overlap of distribution ranges.
311
(d)
(e)
Figure 7.7. (cont.)
312
(f)
(g)
(h)
(i)
(j)
(k)
(l)
(m)
Figure 7.7. (cont.)
313
(n)
(o)
(p)
(q)
(r)
(s)
(t)
(u)
Figure 7.7 (cont.)
314
(v)
(w)
(x)
(y)
(z)
(aa)
Figure 7.7 (concl.).
315
(a)
(b)
Figure 7.8 (a-z). The range maps of spurfowls indicating where barriers lie between two
clades that are being compared. The range in blue is a reference taxon/clade which is
compared to the range of the taxon/clade in red. The yellow lines are barriers detected
whereas the red and dark blue lines are connecting lines. Green grids indicate areas of
overlap of distribution ranges.
316
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(j)
Figure 7.8. cont.
317
(k)
(l)
(m)
(n)
(o)
(q)
(p)
(r)
Figure 7.8. cont.
318
(s)
(t)
(u)
(v)
(w)
(y)
(x)
(z)
Figure 7.8. concl.
319
CHAPTER 8
Summary and Synthesis
Aims of this thesis:
1. To undertake a general review of the taxonomy of ‘francolins’ [Chapter 1].
2. To contrast alternative hypotheses concerning the monophyly of the genus
‘Francolinus’ (e.g. Hall 1963 versus Bloomer and Crowe 1998) and to further explore
DNA-based, vocalization and behavioural evidence in order to test the phylogenetic
affinities of the Stone Partridge Ptilopachus petrosus and Nahan’s Francolin
Francolinus nahani as suggested in Crowe et al. (2006) [Chapter 2].
3. To investigate if there are any syringeal features (anatomical) that can be used to
further our understanding of phylogenetic relationships among francolin and spurfowl
taxa [Chapter 3].
4. To test the validity of the francolin-spurfowl taxonomic dichotomy based on vocal
characters [Chapter 4].
5. To test Hall’s (1963) hypotheses on the monophyly of the suggested eight species
groups within her circumscription of the genus Francolinus, and thereby provide a
modern systematic revision of the terminal taxa and genera of francolins and
spurfowls [Chapter 5 & 6].
6. To describe and explore the patterns of distribution of francolins and spurfowls using
Hall’s (1963) monograph as a null hypothesis [Chapter 7].
320
The key findings of this thesis
A general review of the taxonomy of Francolinus (sensu lato)
There has been persistent confusion regarding the taxonomic status of francolins sensu lato
since the description of the genus Francolinus in the mid-1830s. This is readily demonstrated
in the outline of the various taxonomic reviews carried out on the genus (Chapter 1). The
confusion centred on the taxonomic and phylogenetic status of the genus, the taxonomic
status and the number of valid genera, terminal species and subspecies of francolins, the
phylogenetic position of Hall’s (1963) putative monophyletic species groups, as well as the
phylogenetic position of Ortygornis sephaena and Hall’s (1963) four unplaced enigmatic
species (now Ptilopachus nahani, Afrocolinus lathami, Ortygornis pondicerianus, Ortygornis
gularis), and finally the origin of the genus Francolinus which involves two conflicting
hypotheses postulated by Hall (1963) and Crowe and Crowe (1985), respectively.
One of the major motivations for conducting this study was to establish a classification
system that accounts for the evolutionary history of francolin and spurfowl taxa. Most species
of francolins and spurfowls vary, often strikingly, in morphology, and they form complexes of
species that make delimitation of taxa and the endeavour to define species complicated and
difficult to achieve. The conclusions drawn from the taxonomic reviews were that taxon and
character sampling procedures, types of characters (organismal versus molecules) and
analytical methods (quantitative statistical methods versus model-/non-model-based
phylogenetic methods) have significantly improved our understanding of the systematics of
francolins and spurfowls, that have important implications for designating taxonomic rank:
e.g. whether a taxon should be considered a species, subspecies or a synonym. These results
in turn have improved our understanding of the biogeography of francolins and spurfowls.
321
Contrasting ideas:
Monophyly of the genus Francolinus (sensu Hall 1963) and the species groups
The advent of molecular data has revolutionized the understanding of evolutionary aspects of
francolins and spurfowls. One of Hall’s major challenges was the difficulty of determining the
phylogenetic position of Ortygornis sephaena and hence she was forced to place 41
traditionally recognized species in a single genus Francolinus. Without the use of discrete
molecular data and modern phylogenetic inference methods, as well as the need for excellent
taxon sampling, early studies could not decisively reject monophyly of Francolinus and had
to simply allude to the possibility of francolins forming two major lineages, and hence often
opted to retain all species in a single genus Francolinus.
This study decisively demonstrates that the genus Francolinus is not monophyletic,
instead it can be split into two distantly related assemblages (Chapter 2): ‘francolins’
(partitioned among the genera Francolinus, Ortygornis, Afrocolinus, Peliperdix, Scleroptila)
and ‘spurfowls’ (all placed in the genus Pternistis). The Bayesian reconstruction for the time
of divergence between these two lineages, that is, the Coturnix/Alectoris/spurfowls and
Gallus/Bambusicola/francolins, was recovered at c. 33.6 mya. The estimated evolutionary
date for the time-to-most-recent-common-ancestor for francolins and spurfowls was recovered
at c. 7.6 and 8.7 mya, respectively.
Monophyly of four (the Red-tailed, Red-winged, Spotted and the Bare-throated
Groups) of eight species groups, as delimited by Hall (1963), was recovered. Significantly,
this study was able to decisively place the problematic species O. sephaena within the ‘true’
francolins, despite its possession of mixed francolin and spurfowl features. This study also
revealed that its closest evolutionary relatives are the two unplaced Asian species O.
322
pondicerianus and O. gularis. Ortygornis sephaena may well be a derived francolin that
emerged early in the history of the genus Francolinus.
Another major revelation was that Hall’s (1963) difficulty in assigning Nahan’s
Francolin Francolinus ‘Ptilopachus’ nahani to any specific species group made sense since
this study unequivocally revealed that Nahan’s Francolin is not a francolin but a partridge,
and sister to the African Stone Partridge Ptilopachus petrosus (having diverged c. 9.6 mya details in Chapter 2), and the duo are in turn, sister to the New World quails. The remaining
enigmatic morphological taxon of Hall (1963), Afrocolinus lathami was in most analyses,
found to represent its own evolutionary trajectory and a new genus name Afrocolinus gen.
nov. was assigned in Chapter 5.
Further investigation into the phylogenetic affinities of the Stone Partridge Ptilopachus
petrosus and Nahan’s Francolin Francolinus nahani (sensu Crowe et al. 2006)
Francolinus nahani is a taxonomically enigmatic African galliform restricted to the
interior of remnant primary forests of the eastern equatorial lowlands of the Democratic
Republic of the Congo and Uganda. This thesis extended the earlier result of Crowe et al.
(2006) by confirming that the closest extant relative of Francolinus nahani is Ptilopachus
petrosus with an estimated divergence time at c. 9.6 mya. The two species are in turn most
closely related to the New World quails (Odontophoridae) than any other Old World
galliform. On the basis of the close phylogenetic relationship between F. nahani and P.
petrosus, and their shared vocal and behavioural characters, this study recommended that F.
nahani be moved to the genus Ptilopachus Swainson, based on name priority, subfamily
Ptilopachinae subfam. nov. (Bowie et al. 2013), and family Odontophoridae.
323
Francolin-spurfowl dichotomy – syryngeal and vocalization perspective
With ever increasing data suggesting a deep divergence between the francolin and
spurfowl assemblages, the syringes (Chapter 3) of selected francolins and spurfowls and the
calls (Chapter 4) of the various species were analyzed in an effort to further validate the
divergence between these two lineages of primarily African galliforms. It was found that there
were clear-cut vocal differences between (and among) francolins and spurfowls with
francolins generally rendering longer tonal strophes with distinct elements that have
detectable harmonics. Spurfowls, in contrast, generally give shorter atonal strophes with
elements that generally have no harmonics. The structure of the syringes also supports this
distinction, with the shape of the tympanum placing O. sephaena decisively with other
francolins despite the lack of the mineralized bronchial half rings that are present in other
francolins, and the size of the pessulus corroborates the francolin-spurfowl dichotomy. The
amount of connective tissue places species with tonal and whistling calls together (Scleroptila
levaillantii and S. afra) to the exclusion of those taxa that have atonal, raucous and grating
calls (Pternistis capensis, P. natalensis and O. sephaena).
The validity of terminal taxa - francolins and spurfowls
In Chapter 5 and 6, the focus was to establish an evolutionary-based classification system to
enable the identification of those francolin and spurfowl taxa that represent valid species and
subspecies, as well as to determine the validity of various proposed generic names. The major
disagreement recorded was at the subspecies level, followed by the generic and to some
degree at the species level. As outlined in Chapters 5 and 6, this study proposes a tremendous
decrease in the number of subspecies traditionally recognized in the earlier reviews (see
Chapter 1) with most of these taxa being synonymized (with those deemed valid subspecies)
324
whenever they were found to not represent distinct evolutionary units and fail to meet the
criteria as outlined in Chapter 5. In summary, the number of francolin and spurfowl species
increased while additional genera were recognized for francolins. Five francolin genera
Francolinus, Ortygornis, Afrocolinus gen. nov., Peliperdix and Scleroptila are recognized and
all spurfowl species are placed in a single genus Pternistis.
A number of traditionally recognized subspecies were elevated to species in this
revision, resulting in 31 and 24 species of francolins and spurfowls, respectively. Although
some subspecies are morphologically distinct, combined phylogenetic evidence from
organismal and DNA characters, together with sequence divergence estimates based on
Cytochrome-b characters, suggests that these taxa have yet to reach an independent
evolutionary trajectory that warrants species status, hence they are synonymized with other
taxa. What differentiates the taxonomic approach adopted in this study from those in the
previous revisions is that the evolutionary relationships among taxa played a central role.
The history of distributional patterns of francolins and spurfowls and the origin of the
genus Francolinus
The focus in Chapter 7 was to interpret the current distributional patterns of francolins
and spurfowls using historical information. Further, in light of the two alternate hypotheses
for the origin of francolins and spurfowl, this study sought to determine the origin of the
ancestral francolin and spurfowl. Between the two approaches that were used, the disjunctiondetecting and barrier-inferring spatial analysis of vicariance or the ancestral reconstruction of
area and habitat, it could be concluded that the spatial analysis of vicariance provided a better
picture of the history of distribution patterns of francolins and spurfowls. So, the colonization
of Africa by the ancestral species may have been through dispersal which resulted in the
325
formation of disjunct distributions and the somewhat rapid diversification of francolins and
spurfowls in Africa. The direction of dispersal might have gone either way, i.e. from Asia to
Africa, and vise versa. The sister relationship between the Asian O. gularis clade and the O.
sephaena complex could fit the explanation that, during the collision of Asia and Africa, some
individuals may have remained in Asia while others remained in Africa resulting in the two
groups differentiating with subsequent speciation occurring on both continents. On the other
hand, one other possible explanation on the origin of the spurfowls is that the ancestor may
have dispersed from Asia, into the desert of Saudi Arabia along its coast and then into Africa
where it eventually arrived in the desert in south western Africa. Generally, there is strong
support for Hall’s (1963) hypothesis for the evolution of the genus Francolinus being of
Asian origin.
The Rift Valley system, Lake Chad, Upper Guinea and the Congolian forest, major
rivers such as Limpopo, Zambezi, Rovuma and the Volta River were found to be maintaining
the distribution patterns and possibly promoting speciation among African francolins. In
addition, the Niger and Rufiji River, Okavango swamp and the Sahara desert have maintained
the distribution ranges of the spurfowl taxa. Habitat heterogeneity e.g. xeric/mesic savanna
versus mesic savanna, montane and lowland grasslands and forest habitat also play an
important ecological role in facilitating diversification among francolins. Among spurfowls,
similar habitat barriers to those detected among francolins were inferred.
Future prospects
Taxonomy remains a challenging discipline especially in the absence of a universal definition,
which can be applied at both the species and subspecies level. The long-standing debate on
the definition of species (as well as subspecies) is one that will remain difficult to resolve. The
326
conflict exists between character partitions from different sources possibly due to the high
level of hybridization and incomplete lineage sorting and these conditions make taxonomic
practice complicated. Thus, there has been profound taxonomic confusion.
Future prospects to specifically focus on the phylogeographic studies of selected
species complexes of francolins and spurfowls and also to investigate the nature of hybrid
zones, are brought forth. The role of various syringeal features in voice production among
francolins and spurfowls should be investigated.
327
CHAPTER 9
References
Akaike H 1973 Information theory as an extension of the maximum likelihood
principle. Pp. 276-281. In: Second International Symposium on Information
Theory. (Petrov BN, Csaki F, eds). Akademiai Kiado, Budapest.
Alexander, 1909 Francolinus camerunensis. Bulletin of the British Ornithologists´ Club
25, p. 12.
Alström P 2001 The use of sounds in bird systematics. Introductory Research Essay
No. 2. Uppsala University, Uppsala, Sweden.
Alström P, Rasmussen PC, Olsson B, Sundberg P 2008 Species delimitation based
on multiple criteria: the Spotted Bush Warbler Bradypterus thoracicus complex
(Aves: Megaluridae). Zoological Journal of the Linnean Society 154, pp. 291307.
Ames PJ 1971 The morphology of the syrinx in passerine birds. Bulletin of the Peabody
Museum of Natural History 37, pp 1-94.
Appel FW 1929 Sex dimorphism in the syrinx of the fowl. Journal of Morphology
47(2), pp. 497-517.
Arias JS, Szumik CA, Goloboff PA 2011 Spatial analysis of vicariance: a method for
using direct geographical information in historical biogeography. Cladistics 27,
pp. 1-12.
Armstrong MH, Braun EL, Kimball RT 2001 Phylogenetic utility of avian
Ovomucoid intron G: a comparison of nuclear and mitochondrial phylogenetics
in Galliformes. The Auk 118, pp. 799-804.
328
Aubrecht G, Holzer G 2000 Die Stockenten: Biologie, O¨ kologie, Verhalten. 1.
Östechischer, Agrarverlag, Vienna, Austria.
Bangs & Loveridge, 1931 Pternistis squamatus uzungwensis. Proceedings of the New
England Zoological Club 12, p. 93.
Bannerman, 1934 Francolinus ahantensis hopkinsoni. Bulletin of the British
Ornithologists´ Club 55, p. 5.
Bannerman D, 1934 Francolinus ahantensis hopkinsoni. [A new race of the
Ahanta Francolin from Gambia and Portuguese Guinea]. Bulletin of the British
Ornithologists´ Club 55, pp. 5-6.
Bannerman, 1922 Francolinus bicalcaratus ogilviegranti. Bulletin of the British
Ornithologists´ Club 42, p. 132.
Bancroft JD, Gamble M 2002 Theory and practice of histological techniques (5th ed).
Churchill Livingstone, NY.
Bates, 1928 Francolinus coqui spinetorum. Bulletin of the British Ornithologists´ Club
49, p. 33.
Benson CW, 1939 Francolinus squamatus doni Benson. A new Francolin from
Nyasaland. Bulletin of the British Ornithologists´ Club 59, pp. 42-43.
Benson, 1939 Pternistis squamatus doni. Bulletin of the British Ornithologists´ Club 59,
p. 42.
Benson CW 1945 Notes on the birds of sourthern Abyssinia (part). Ibis 87, pp. 366400.
Benson CW, Irwin MPS, White CM 1962 The significance of valleys as avian
zoogeographical barriers. Symposium on causes and problems of animal
distribution with special reference to southern Africa. Annals of the Cape
329
Provincial Museum 2, pp. 155-189.
Bininda-Edmonds ORP, Jones KE, Price SA, Cardillo M, Grenyer R, Purvis A
2004 Garbage in garbage out: data issues in supertree constructions. Pp. 267-280. In:
Phylogenetic supertrees: combining information to reveal the Tree of
Life (Binida-Edmonds ORP, ed). Kluwer, Dordrecht.
Bloomer P, Crowe TM 1998 Francolin phylogenetics: molecular, morphobehavioral,
and combined evidence. Molecular Phylogenetics and Evolution 9(2), pp. 236-254.
Blundell & Lovat, 1899 Francolinus squamatus tetraoninus. Bulletin of the British
Ornithologists' Club 10, p. 22.
Blundell & Lovat 1899 Francolinus harwoodi. Bulletin of the British Ornithologists´
Club 10, p. 22.
(Blyth, 1843) Francolinus pintadeanus phayrei. Journal of the Asiatic Society of Bengal
12, p. 1011.
Blyth, 1852 Scleroptila. Catalogue of the birds in the museum [of the] Asiatic, p. 250.
Bocage, 1869 Francolinus hartlaubi hartlaubi. Jornal de Sciencias mathematicas,
physicas e naturas, publicado sob os auspicos da Academia real das sciencias
da Lisboa 2(8), p. 350.
Bocage, 1881 Francolinus finschi. Ornithologie d'Angola pt2, p. 406.
Bocage, 1893 Pternistis afer benguellensis. Jornal de Sciencias mathematicas,
physicase naturas, publicado sob os auspicos da Academia real das sciencias da
Lisboa 3(2), p. 154.
Bock WJ, Farrand J 1980 The number of species and genera of Recent birds.
American Museum Novitates 2703, pp. 1-29.
Bonaparte, 1856 Francolinus francolinus asiae. Comptes Rendus 42, p. 882.
330
Bonaparte, 1856 Francolinus francolinus henrici. Comptes Rendus 42, p. 882.
Bonaparte, 1856 Peliperdix. Comptes Rendus 42, p. 882.
Bouilhol P, Jagoutz O, Hanchar JM, Dudas FO 2013 Dating the India–Eurasia
collision through arc magmatic records. Earth and Planetary Science Letters
366 (2013), pp. 163-175.
Bowen, 1930 Pternistis afer itigi. Proceedings of the Academy of Natural Sciences of
Philadelphia 82, p. 86.
Bowen, 1931Francolinus jacksoni gurae. Proceedings of the Academy of Natural
Sciences of Philadelphia 83, p. 302.
Bowie RCK, Fjeldså J 2005 Genetic and morphological evidence for two species in
the Udzungwa Forest Partridge Xenoperdix udzungwensis. Journal of East
African Natural History 94, pp. 191-201.
Bowie RCK, Cohen C, Crowe TM 2013 Ptilopachinae: a new subfamily of the
Odontophoridae (Aves: Galliformes). Zootaxa 3670(1), pp. 097-098.
Brown C, Ward D 1990 The morphology of the syrinx in the Charadriiformes (Aves):
possible phylogenetic implications. Bonner Zoologische Beiträge 41, pp. 95-107.
Büttikofer, 1889 Francolinus levaillantoides jugularis. Notes from the Leyden Museum
11, p. 76.
Cabanis, 1868 Pternistes leucoscepus infuscatus. Journal of Ornithology 16, p. 413.
Cabanis, 1878 Pternistis hildebrandti hildebrandti. Journal of Ornithology 26(142), p.
206.
Cabanis, 1880 Pternistis squamatus schuetti. Journal of Ornithology 28(152), p. 351.
Cannell PF 1988 Techniques for the study of avian syrinxes. Wilson Bulletin 10, pp.
289-293.
331
Chapin JP 1926 A new genus Acentrortyx proposed for Francolinus nahani Dubois.
The Auk 43, pp. 235.
Chapin JP 1932 The birds of the Belgian Congo. Bulletin of the American Museum of
Natural History. Vol. LXV, New York.
Chapin JP, 1937 Francolinus albogularis dewittei. A new race of Francolinus
albogularis from Marungu. Revue de zoologie et de botanique africaines 29, pp.
395-396.
Chappuis, C 2000 African bird sounds. Birds of North, West and Central Africa.
Audio CD set. With the collaboration of the British Library National Sound
Archive (London). Société d Études Ornitholgiques de France, Paris.
Cassin, 1857 Pternistis squamatus squamatus. Proceedings of the Academy of Natural
Sciences of Philadelphia 8, p. 321.
(Children & Vigors, 1826) Francolinus clappertoni clappertoni. Narrative of Travels
and Discoveries in Northern and Central Africa App. 21, p. 198.
Cicero C 2010 The significance of subspecies: a case study of sage sparrows
(Emberizidae, Amphispiza belli). The American Ornithologists’ Union 67, pp.
103-113.
Clancey PA 1957 On the range and status of Certhilauda falcirostris Reichenow, 1916:
Port Nolloth, N.W. Cape. Bulletin of the British Ornithilogists’ Club 77, pp.
133-137.
Clancey PA, 1960 Francolinus shelleyi sequestris. Miscellaneous taxonomic notes on
African birds. XV. 1. A new race of Shelley’s Francolin Francolinus shelleyi
Ogilvie-Grant from Natal and Zululand. Durban Museum Novitates 6(2), pp. 1314.
332
Clancey, 1996 Francolinus adspersus mesicus. Bulletin of the British Ornithologists´
Club 116, p. 106.
Clancey PA 1967 Gamebirds of Southern Africa. Purnell & Sons, Cape Town.
Clancey PA 1986 Endemicity in the southern African avifauna. Durban Museum
Novitates 13(20), pp. 245-284.
Clark JD 1967 Atlas of African prehistory. The University of Chicago, London.
Cohen C, Wakeling JL, Mandiwana-Neudani TG, Sande E, Dranzoa C, Crowe
TM, Bowie RCK 2012 Phylogenetic affinities of evolutionarily enigmatic
African galliforms: the Stone Partridge Ptilopachus petrosus and Nahan’s
Francolin Francolinus nahani, and support for their sister relationship with New
World quails. Ibis 154, pp. 768-780.
Conover, 1929 Pternistes afer nyanzae. Auk 46, p. 345.
Conover, 1928 Pternistis squamatus usambarae. Auk 45, p. 356.
(Conover, 1930) Francolinus castaneicollis atrifrons. Proceedings of the Biological
Society of Washington 43, p. 3.
Cotterill FDP 2006 Taxonomic status and conservation importance of the avifauna of
Katanga (south-east Congo Basin) and its environs. Ostrich - Journal of African
Ornithology 77(1&2), pp. 1-2.
Cox WA, Kimball RT, Braun EL 2007 Phylogenetic position of the New World
quails: eight nuclear loci and three mitochondrial regions contradict morphology
and the Sibley/Ahlquist tapestry. The Auk 124(1), pp. 71-84.
Cracraft J 1994 Species diversity, biogeography, and the evolution of biotas.
American Zoologist 34, pp. 33-47.
Cracraft J 1981 Toward a phylogenetic classification of the recent birds of the world.
333
The Auk 98, pp. 681-714.
Crowe TM, Crowe AA 1985 The genus Francolinus as a model for avian evolution
and biogeography in Africa. Pp. 207-231. In: Proceedings of the International
Symposium on African vertebrates (Schuchmann KL, ed). Museum Alexander
Koenig, Bonn.
Crowe TM, Keith S, Brown LH 1986 Phasianidae, guineafowl, Congo peacock, quail,
partridges and francolins. Pp. 1-75. In: The Birds of Africa (Urban EK, Fry CH,
Keith S, eds). Vol II. Academic Press, London.
Crowe TM 1992 Morphometric research on Miocene and Pliocene fossil phasianids
from the west coast of southern Africa: some preliminary taxonomic and
phylogenetic conclusions. Pp. 261-267. In: Proceedings of the 7th Pan-African
Ornithological Congress (Bennun L, ed). PAOC Committee, Nairobi.
Crowe TM, Short LL 1992 A new gallinaceous bird from the Oligocene of Nebraska,
with comments on the phylogenetic position of the Gallinuloididae. Natural
History Museum of Los Angeles County Science Series 36, pp. 179-185.
Crowe TM, Harley EH, Jakutowicz M 1992a Phylogenetic relationships of southern
African francolins (Francolinus spp.) as suggested by the structure of their
mitochondrial DNA and their morphology, behaviour and ecology. Proceedings
of the 7th Pan-African Ornithological Congress, pp. 261-267.
Crowe TM, Harley EH, Jakutowicz MB, Komen J, Crowe AA 1992b Phylogenetic,
taxonomic and biogeographical implications of genetic, morphological, and
behavioural variation in francolins (Phasianidae. Francolinus). The Auk 109(1),
pp. 24-42.
Crowe TM, Bowie RCK, Bloomer P, Mandiwana TG, Hedderson TAJ, Randi E,
334
Pereira SL, Wakeling J 2006 Phylogenetics, biogeography and classification
of, and character evolution in, gamebirds (Aves: Galliformes): Effects of
character exclusion, data partitioning and missing data. Cladistics 22, pp. 1-38.
Darwin C 1859 On the origin of species. Facsmile of first edition (1964). Harvard
University Press, Cambridge, Massachusetts.
Dayrat B 2005 Towards integrative taxonomy. Biological Journal of the Linnean Society 85,
pp. 407-415.
Delacour J, Mayr E 1945 The Family Anatidae. Wilson Bulletin 57, pp. 3-55.
del Hoyo J, Elliott A, Sargatal J 1994 Handbook of the birds of the world: New
World Vultures to Guineafowl. Vol. II. Lynx edicions, Barcelona.
de Queiroz A 2005 The resurrection of oceanic dispersal in historical biogeography.
Trends in Ecology and Evolution 20, pp. 68–73.
de Queiroz K 2007 Species concept and species delimitation. Systematic Biology 56(6),
pp. 879-886.
De Schauensee, 1931 Francolinus adspersus kalahari. Proceedings of the Academy of
Natural Sciences of Philadelphia 83, p. 453.
Dickinson EC (ed) 2003 The Howard and Moore complete checklist of the birds of the
World. 3rd ed. Christopher Helm, London.
Dimcheff DE, Drovetski SV, Krishnan M, Mindell DP 2000 Cospeciation and
horizontal transmission of avian sarcoma and leucosis virus gag genes in
galliform birds. Journal of Virology 74, pp. 3984-3995.
Dimcheff DE, Drovetski SV, Krishnan M, Mindell DP 2002 Phylogeny of
Tetraoninae and other galliform birds using mitochondrial 12S and ND2 genes.
Mololecular Phylogenetics and Evolution. 24, pp. 203-215.
335
Dinesen L, Lehmberg TJ, Svendsen O, Hansen TA, Fjeldså J 1994 A new genus and
species of perdicine bird from Tanzania: a relict form with Indo-Malayan
affinities. Ibis 136, pp. 3-11.
Donoghue MJ, Moore BR 2003 Toward an integrative historical biogeography.
Integrative and Comparative Biology 43, pp. 261-270.
(Dorst & Jouanin, 1952) Pternistis ochropectus. L'Oiseau et la Revue Francaise
d'Ornithologie 22, p. [71].
Dowsett RJ, Prigogine A 1974 The avifauna of the Marungu Highlands. Exploration
Hydrobiologique du Bassin du lac Bangweolo et du Luapula 14, pp. 1-67.
Drummond AJ, Rambaut A 2007 BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evolutionary Biology 7, pp. 214.
(Dubois, 1905) Francolinus nahani. Annales du Musée du Congo 1(pl.10), p. 17.
Dyke
GJ
2003
The
phylogenetic
position
of
Gallinuloides
Eastman
(Aves:
Galliformes) from the Tertiary of North America. Zootaxa 199, pp. 1-10.
Dyke GJ, Gulas BE, Crowe TM 2003 Suprageneric relationships of galliform birds
(Aves, Galliformes): a cladistic analysis of morphological characters. Zoological
Journal of the Linnean Society 137, pp. 227-244.
Edwards SV, Wilson AC 1990 Phylogenetically informative length polymorphism and
sequence variability in mitochondrial DNA of Australian songbirds
(Pomatostomus). Genetics 126, pp. 695-711.
Edwards SV, Arctander P, Wilson AC 1991 Mitochondrial resolution of a deep branch in
the genealogical tree for perching birds. Proceedings of the Royal
Society 243, pp. 99-107.
336
Elzanowski A, Stidham TA 2011 A gallofanserine quadrate from the Late Cretaceous
Lance formation of Wyoming. The Auk 128(1), pp.138-145.
Enggist-Düblin P, Birkhead TR 1992 Differences in the calls of European and North
American Black-billed Magpie and the Yellow-billed Magpie. The International
Journal of Animal Sound and its Recording 4, pp. 185-194.
Eberle JJ Greenwood DR 2012 Life at the top of the greenhouse Eocene world – a
review of the Eocene flora and vertebrate fauna from Canada’s High Arctic.
Bulletin of the Geological Society of America 124, pp. 3-23.
Eo SH, Bininda-Evans ORP, Carroll JP 2009 A phylogenetic supertree of the fowls
(Galloanserae, Aves). Zoologica Scripta 38(5), pp. 465-481.
Farais IP, Ortí G, Meyer A 2000 Total evidence: molecules, and the phylogenetics of
cichlid fishes. Journal of Experimental Zoology (Molecular and Developmental
Evolution) 288, pp. 76-92.
Farnsworth A, Lovette IJ 2008 Phylogenetic and ecological effects on interspecific
variation in structurally simple avian vocalizations. Biological Journal of the
Linnean Society 94, pp. 155-173.
Farris JS, Albert VA, Källersjö M, Lipscomb D, Kluge AG 1996 Parsimony
jackknifing outperforms neighbor-joining. Cladistics 12, pp. 99-124.
Felsenstein J 1985 Confidence limits on phylogenies: an approach using the
Bootstrap. Evolution 39(4), pp. 783-791.
(Fischer & Reichenow, 1884) Pternistis afer leucoparaeus. Journal of Ornithology
32(166), p. 263.
Fischer & Reichenow, 1884 Francolinus hildebrandti altumi. Journal of Ornithology
32(165), p. 179.
337
FitzPatrick JW 2010 Subspecies are for convenience. The American Ornithologists’
Union 67, pp. 54-61.
Fjeldså J, Bowie RCK 2008 New perspectives on the origin and diversification
of Africa's forest avifauna. African Journal of Ecology 46, pp. 235-247.
Forcina G, Panayides P, Guerrini M, Nardi F, Gupta BK, Mori DE, Al-Sheikhly
EF, Mansoori J, Khaliq GI, Rank DN, Parasharya BM, Khan AA,
Hadjigerou P, Barbanera F 2012 Molecular evolution of the Asian francolins
(Francolinus, Galliformes): a modern reappraisal of a classic study in
speciation. Molecular Phylogenetics and Evolution 65(2012), pp. 523-534.
Frank T, Walter I, Probst A, König HE 2006 Histological aspects of the syrinx of the
male Mallard (Anas platyrhynchos). Anatomia, Histologia, Embryologia 35, pp.
396-401.
Frank T, Probst A, König HE, Walter I 2007 The syrinx of the male Mallard (Anas
platyrhynchos): special anatomical features. Anatomia, Histologia, Embryologia
36, pp. 121-126.
Friesen VL, Congdon BC, Walsh HE, Birt TP 1997 Intron variation in Marbled
Murrelets detected using analyses of single-stranded conformation
polymorphisms. Molecular Ecology 6, pp.1047-1058.
Fumihito A, Miyake T, Takada M, Ohno S, Kondo N 1995 The genetic link
between the Chinese bamboo partridge (Bambusicola thoracica) and the
chicken and junglefowls of the genus Gallus. Proceedings of the National
Academy of Sciences of the United States of America 92, pp. 11053–11056.
338
Gaban-Lima R, Höfling E 2006 Comparative anatomy of the syrinx in the tribe
Arini (Aves: Psittacidae). Brazilian Journal of Morphological Sciences 23,
pp. 501-512.
Gatesy J, Mathee C, DeSalle R, Hayashi C 2002 Resolution of a supertree/supermatrix
paradox. Systematic Biology 51, pp. 652-654.
Gaunt AS, Gaunt SLL, Casey RM 1982 Syringeal mechanics reassessed: evidence
from Streptopelia. The Auk 99, pp. 474-494.
Gibbon G 1995 Southern African Bird sounds. Southern, Durban, South Africa,
African Birding cc.
Gill FB 1990 Ornithology. 2nd ed. WH Freeman and Company, New York.
Gill F, Donsker D (eds) 2013 IOC World Bird List (version 3.4). Available at
http://www.worldbirdnames.org/ [Accessed "20/08/2013"].
(Gmelin, 1789) Francolinus capensis. System Naturae 1 (pt2), p. 759.
(Gmelin, 1789) Francolinus pondicerianus pondicerianus. Systema Naturae 1(pt2), p.
760.
Goloboff PA, Farris JS, Nixon KC 2008 TNT, a free program for phylogenetic
analysis. Cladistics 24, pp. 774-786.
Grant CHB & Mackworth-Praed CW, 1934 Francolinus coqui thikae. Descriptions of
two new races of Francolin]. Bulletin of the British Ornithologists’ Club 54, pp.
173-174.
Grant & Mackworth-Praed, 1934 Pternistis afer loangwæ. Bulletin of the British
Ornithologists' Club 55, p. 17.
Grant CHB & Mackworth-Praed CW, 1934 Francolinus castaneicollis kaffanus.
[Descriptions of two new races of Francolin.]. Bulletin of the British
339
Ornithologists´ Club 54, pp. 173-174.
Grant CHB & Mackworth-Praed CW, 1934 Francolinus nobilis chapini. On a new
subspecies of Francolin from eastern Africa. Bulletin of the British
Ornithologists´ Club 55, pp. 62-63.
Grant CHB & Mackworth-Praed CW, 1934 Francolinus sephaena somaliensis. [The
eastern African races of Francolinus sephæna (Smith).]. Bulletin of the British
Ornithologists´ Club 54, pp. 170-173.
(Gray JE, 1831) Francolinus pictus pallidus. Illustrations of Indian Zoology
[Hardwicke] 1(pt8, pl.55).
(Gray JE, 1834) Pternistis afer punctulatus. Illustrations of Indian Zoology[Gray &
Hardwicke] 2(pl.43, f.2).
Gray GR, 1867 Francolinus psilolaema psilolaema. List of the specimens of Birds in the
British Museum (pt5), p. 50.
(Gray GR, 1867) Pternistis leucoscepus leucoscepus. List of the specimens of Birds in the
British Museum (pt5), p. 48.
Gray GR, 1867 Francolinus sephaena rovuma. List of the specimens of Birds in the British
Museum (pt5), p. 52.
Griffiths CS 1994a Monophyly of the Falconiformes based on syringeal morphology.
The Auk 111(4), pp. 787-805.
Griffiths CS 1994b Syringeal morphology and the phylogeny of the Falconidae. The
Condor 96, pp. 127-140.
Gunning & Roberts, 1911 Pternistis afer castaneiventer. Annals of the Transvaal
museum 3(2), p.110.
Hackett SJ, Kimball RT, Reddy S, Bowie RCK, Braun EL, Braun MJ, Chojnowski
340
JL, Cox WA, Han K-L, Harshman J, Huddleston CJ, Marks BD, Miglia
KJ, Moore WS, Sheldon FH, Steadman DW, Witt CC, Yuri T 2008 A
phylogenomic study of birds reveals their evolutionary history. Science 320, pp.
1763-1768.
Hall BP 1963 The francolins, a study in speciation. Bulletin of the British Museum
(Natural History) 10, pp. 8-204.
Hall R 2001 Cenozoic geological and plate tectonic evolution of SE Asia and the SW
Pacific: computer-based reconstructions, model and animations. Journal of
Asian Earth Sciences 20, pp. 353-431.
Harrington GJ, Eberle J, Le-Page BA, Dawson M, Hutchison JH 2012 Arctic
plant diversity in the Early Eocene greenhouse. Proceedings of the royal society
B - Biological Sciences 279, pp. 1515-1521.
Harrison JA, Allan DG, Underhill LG, Herremans M, Tree AJ, Parker V, Brown,
CJ (eds) 1997 The Atlas of Southern African Birds: Non-passerines. Vol. 1.
BirdLife South Africa, Johannesburg.
Hartert, 1917 Francolinus pondicerianus interpositus. Novitates Zoologicae 24, p. 288.
Hartert, 1917 Francolinus bicalcaratus ayesha. Novitates Zoologicae 24, p. 291.
Hartlaub, 1854 Francolinus lathami lathami. Journal of Ornithology 2(9), p. 210.
Hartlaub, ["1865"]1866 Francolinus sephaena grantii. Proceedings of the Zoological
Society of London (pt3, pl.39), p. 665.
Heuglin, 1863 Francolinus schlegelii, Journal of Ornithology 11(64), p. 275.
Hockey PAR, Dean WRJ, Ryan PG (eds) 2005 Roberts - Birds of Southern Africa. 7th
ed. The Trustees of the John Voelcker Bird Book Fund, Cape Town.
Hoesch W & Niethammer, G 1940 Francolinus gariepensis stresemanni. Die
341
Vogelwelt Deutsch-Südwest-Afrikas namentlich des Damara- und Namalandes.
Journal of Ornithology 88, pp. 1-404.
Hogg DA 1982 Ossification of the laryngeal, trachea and syringeal cartilages in the
domestic fowl. Journal of Anatomy 134, pp 57-71.
Hovenkamp P 1997 Vicariance events, not areas, should be used in biogeographical
analysis. Cladistics 13, pp. 67-79.
Hovenkamp P 2001 A direct method for the analysis of vicariance patterns. Cladistics
17, pp. 260-265.
Huelsenbeck JP, Ronquist F 2001 MrBayes: Bayesian inference of phylogeny.
Bioinformatics 17, pp. 754-55.
Hume, 1888 Francolinus francolinus melanonotus. Stray Feathers 11, p. 305.
Humphrey PS 1955 The relationships of the seaducks (Tribe Mergini). Unpublished
Ph.D. thesis, University of Michigan-Ann Arbor.
James FC 1970 Geographic size variation in birds and its relationship to climate.
Ecology 51(3), pp. 365-390.
(Jardine & Selby, 1828) Francolinus pictus pictus. Illustrations of Ornithology
1(pl.50).
Johnsgard PA 1961 Tracheal anatomy of the Anatidae and its taxonomics significance.
Wildfowl 12, pp. 58-69.
Johnsgard PA 1973 The Grouse and Quails of North America. University of Nebraska
Press, Lincoln.
Johnsgard PA 1986 The Pheasants of the World. Oxford University Press, Oxford.
Johnsgard PA 1988 The Quails, Partridges, and Francolins of the World. Oxford
University Press, Oxford.
342
Johnsgard PA 1999 The Pheasants of the World. 2nd ed. Smithsonian Institution Press,
Washington, D.C.
Jolles J, Schoentgen F, Jolles P, Wilson AC 1976 Amino acid sequence and
immunological properties of chachalaca egg white lysozymes. Journal of Molecular
Evolution 8, pp. 59-78.
Jolles J, Ibrahimi IM, Prager EM, Schoentgen F, Jolles P, Wilson AC 1979 Amino
acid sequence of pheasant lysozyme. Evolutionary change affecting processing
of pre-lysozyme. Biochemistry 18, pp. 2744-2752.
Jønsson KA, Bowie RCK, Moyle RG, Les Christidis JAN, Benz BW, Fjeldså J
2010 Historical biogeography of an Indo-Pacific passerine bird family
(Pachycephalidae): different colonization patterns in the Indonesian and
Melanesian archipelagos. Journal of Biogeography 37, pp. 245-257.
Källersjö M, Farris JS, Chase MW, Bremer B, Fay MF, Humphries CJ, Petersen
G, Seberg O, Bremer K 1998 Simultaneous parsimony jackknife analysis of 2538
rbcL DNA sequences reveals support for major clades of green plants, land
plants, seed plants and flowering plants. Plant Systematics and Evolution 213, pp.
259-287.
Katoh K, Asimenos G, Toh H 2009 Multiple alignment of DNA sequences with
MAFFT. Methods in Molecular Biology 537, pp. 39-64.
Kimball RT, Braun EL 2008 A multigene phylogeny of Galliformes supports a single
origin of erectile ability in non-feathered facial traits. Journal of Avian Biology
39, pp.438-445.
Kingdon J 1989 Island Africa. Collins, London.
Knowles LL, Carstens BC 2007 Delimiting species without monophyletic gene trees.
343
Systematic Biology 56(6), pp. 887-895.
Komen J 1987 Preliminary observations of the social pattern, behaviour and
vocalization of Hartlaub’s francolin. South African Journal of Wildlife Research
Supplement 1, pp. 82-86.
Kornegay JR, Kocher TD, Williams LA, Wilson AC 1993 Pathways of lysozyme
evolution inferred from the sequences of cytochrome b in birds. Journal of
Molecular Evolution 37, pp. 367-379.
Krakauer AH, Tyrrell M, Lehmann K, Losin N, Goller F, Patricelli GL 2009 Vocal
and anatomical evidence for two-voiced sound production in the greater sagegrouse Centrocercus urophasianus. Journal of Experimental Biology 212, pp
3719-3727.
Kriegs JO, Matzke A, Churakov G, Kuritizin A, Mayr G, Brosius J, Schmitz J
2007 Waves of genomic hitchhikers shed light on the evolution of gamebirds
(Aves: Galliformes). BMC Evolutionary Biology 7, pp.190-212.
Ksepka DT 2009 Broken gears in the avian molecular clock: new phylogenetic
analyses support stem galliform status for Gallinuloides wyomingensis and rallid
affinities for Amitabha urbsinterdictensis. Cladistics 25, pp. 173-197.
Lamm
KS,
Redelings
BD
2009
Reconstructing
ancestral
ranges
in
historical
biogeography: properties and prospects. Journal of Systematics and Evolution
47, pp. 369-382.
Lanyon WE 1969 Vocal characters and avian systematics. Pp. 291-310. In: Bird
Vocalizations: their relation to current problems in Biology and Psychology
(Hinde RA). Cambridge University Press, London.
Lanyon WE 1986 A phylogeny of the thirty-three genera in the Empidonax assemblage
344
of tyrant flycatchers. American Museum Novitates 2846, pp. 1-64.
Larsen ON, Goller F 2002 Direct observation of syringeal muscle function in
songbirds and a parrot. Journal of Experimental Biology 205, pp. 25-35.
(Latham, 1790) Francolinus afra. Index ornithologicus, sive systema ornithologiae 2, p.
648.
(Lawson, 1963) Francolinus shelleyi canidorsalis. Durban Museum Novitates 7, p. 77.
(Lawson WJ, 1963) Francolinus shelleyi canidorsalis. A contribution to the
Ornithology of Sul Do Save, southerm Moçambique. Durban Museum Novitates
7, pp. 73-124.
(Leach & Koenig KD, 1818) Pternistis afer cranchii. Narrative of an expedition to
explore the river Zaire[Tuckey] App.4, p. 408.
Legge, 1880 Francolinus pictus watsoni. Birds Ceylon 3, p. 745.
Lei F, Wang A, Wang G, Yin Z 2005 Vocalizations of Red-necked Snow Finch,
Pyrgilauda ruficollis on the Tibetan Plateau, China – a syllable taxonomic
signal?. Folia Zoologica 54(1-2), pp. 135-146.
Lewis B 1983 Bioacoustics: a comparative approach. Academic Press, London.
Lewis PO 2001 A likelihood approach to estimating phylogeny from discrete
morphological character data. Sytematic Biology 50(6), pp. 913-925.
(Linnaeus, 1766) Francolinus francolinus francolinus. Systema Naturae, edition 12, p.
275.
(Linnaeus, 1766) Francolinus bicalcaratus bicalcaratus Systema Naturae, edition 12 p.
277.
Little R, Crowe T 2000 Gamebirds of southern Africa. Struik publishers, Cape Town.
Livezey BC 1986 A phylogenetical analysis of recent Anseriform genera using
345
morphological characters. The Auk 103, pp 737-754.
Macdonald JD, 1940 Francolinus clappertoni cavei. A new race of Francolin and a
new race of Lark from the Sudan. Bulletin of the British Ornithologists´ Club 60,
pp. 57-59.
Macdonald JD, 1953 Francolinus levaillantoides wattii. The races in South West
Africa of the Orange River Francolin. Bulletin of the British Ornithologists´
Club 73, pp. 34-36.
Mackworth-Praed CW, Grant CHB 1952 African Handbook of Birds. Series I: Birds
of Eastern and North Eastern Africa. Vol. I. Longmans, London.
Mackworth-Praed CW, Grant CHB 1962 African Handbook of Birds. Series II:
Birds of the Southern Third of Africa. Vol. II. Longmans, London.
Mackworth-Praed CW, Grant CHB 1970 African Handbook of Birds. Series III:
Birds of West Central and Western Africa. Vol. I. Longmans, London.
Mackworth-Praed, 1920 Francolinus castaneicollis ogoensis. Bulletin of the British
Ornithologists´ Club 40, p. 141.
Mackworth-Praed, 1920 Francolinus erckelii pentoni. Bulletin of the British
Ornithologists´ Club 40, p. 141.
Mackworth-Praed, 1920 Francolinus natalensis neavei. Bulletin of the British
Ornithologists´ Club 40, p.140.
Mackworth-Praed, 1920 Francolinus sephaena zambesiae. Bulletin of the British
Ornithologists´ Club 40, p.140.
Maclean GL 1985 Roberts’ Birds of Southern Africa. 5th ed. The Trustees of the John
Voelcker Bird Book Fund, Cape Town.
Maclean GL 1993 Roberts’ Birds of Southern Africa. 6th ed. The Trustees of the John
346
Voelcker Bird Book Fund, Central News Agency, Cape Town.
Madarasz 1914 Francolinus clappertoni koenigseggi, Annales Historico-Naturales
Musei Nationalis Hungarici 12, p. 560.
Maddison DR 1991 The discovery and importance of multiple islands of most
parsimonious trees. Systematic Biology 40, pp. 315-328.
Maddison WP, Maddison DR 2006 Mesquite: a modular system for evolutionary
analysis. Version 1.12. Available from: <http://mesquiteproject.org>.
Madge S, McGowan P 2002 Pheasants, Partridges, and Grouse: a guide to the
Pheasants, Partridges, Quails, Grouse, Guineafowl, Buttonquails, and
Sandgrouse of the world. Princeton Univ. Press, Princeton, N.J.
Mallet J 2005 Hybridization as an invasion of the genome. Trends in Ecology &
Evolution 20, pp. 229-237.
Mallet J 2008 Hybridization, ecological races and the nature of species: empirical
evidence for the ease of speciation. Philosophical Transactions of the Royal
Society B: Biological Sciences 363, pp. 2971-2986.
Marler P 1969 Tonal quality of bird sounds. Pp. 5-18. In: Bird Vocalizations: Their
Relation to Current Problems in Biology and Psychology (Hinde RA, ed).
Cambridge University Press, London.
Mayden RL 1997 A hierarchy of species concepts: the denouement in the saga of the
species problem. Pp. 381-424. In: Claridge MF, Dawah HA, Wilson MR.
Species: the units of biodiversity. Chapman & Hall, UK.
Mayr E 1982 The growth of biological thought. Belknap P. of Harvard U.P.:
Cambridge (Mass.)
Mccracken KG, Sheldon FH 1997 Avian vocalizations and phylogenetic
347
signal. Proceedings of the National Academy of Sciences of the United States of
America 94, pp. 3833-3836.
McGowan PJK 1994 Family Phasianidae (Pheasants and Partridges). In: del Hoyo J,
Elliott A, Sargatal J. Handbook of the birds of the world: New world Vultures to
Guineafowls, Vol. 2. Lynx edicions, Barcelona.
McLachlan GR, Liversidge R 1957 Roberts Birds of South Africa. 2nd ed. The
Trustees of the South African Bird Book Fund, Cape Town.
McLachlan GR, Liversidge R 1970 Roberts Birds of South Africa. 3rd ed. The Trustees
of the John Voelcker Bird Book Fund, Central News Agency, Cape Town.
McLachlan GR, Liversidge R 1978 Roberts Birds of South Africa. 4th ed. The Trustees
of the John Voelcker Bird Book Fund, Cape Town.
Mearns, 1910 Pternistis squamatus maranensis. Smithsonian miscellaneous collections
56(14), p. 1.
Mearns, 1911 Francolinus squamatus zappeyi. Smithsonian Miscellaneous Collections
66(20), p. 4.
Mearns, 1911 Pternistes leucoscepus keniensis. Smithsonian Miscellaneous Collection
66(20), p. 1.
Mearns, 1911 Pternistes leucoscepus kilimensis. Smithsonian Miscellaneous Collections
66(20), p. 2.
Mearns, 1915 Francolinus hildebrandti helleri. Proceedings of the U.S. National
Museum 48, p. 381.
Meinertzhagen, 1937 Francolinus shelleyi theresae. [Six new races from Mt. Kenya,
Kenya Colony.]. Bulletin of the British Ornithologists´ Club 57, pp. 67-70.
Meinertzhagen R, 1937 Francolinus jacksoni pollenorum. [Six new races from Mt.
348
Kenya, Kenya Colony.]. Bulletin of the British Ornithologists´ Club 57, pp. 6770.
Meiri S, Dayan T 2003 On the validity of Bergmann’s rule. Journal of Biogeography
30, pp. 331–351.
Milstein P le S, Wolff SW 1987 The over-simplification of our “francolins”. South
African Journal of Wildlife Research Supplement 1, pp. 58-65.
Mlikovsky J 1989 A new guineafowl (Aves: Phasianidae) form the late Eocene of
France. Annalen des Naturhistorischen Museums in Wien 90, pp.63-66.
Mobley JA, Prum RO 1995 Phylogenetic relationships of the Cinnamon Tyrant,
Neopipo cinnamomea, to the tyrant flycatchers (Tyrannidae). The Condor 97,
pp. 650-662.
Moore WS, Sheldon FH, Steadman DW, Witt CC, Yuri T 2008 A phylogenomic
study of birds reveals their evolutionary history. Science 320, pp. 1763-1768.
Morton ES 1975 Ecological sources of selection of selection on avian sounds.
American Naturalist 109, pp. 17-34.
Morony JJ, Bock WJ, Farrand J 1975 Reference List of the Birds of the World.
American Museum of Natural History, New York.
Moum T, Johansen S, Erikstad KE, Piatt JF 1994 Phylogeny and evolution of the
auks (subfamily Alcinae) based on mitochondrial DNA sequences. Proceedings
of the National Academy of Sciences of the United States of America 91, pp.
7912-7916.
Mourer-Chauvire C, Pickford M, Senut B 2011 The first Palaeogene galliform from
Africa. Journal of Ornithology 152, pp. 617-622.
(Müller PLS, 1776) Pternistis afer afer. Natursystems Supplements, p. 129.
Myers JA 1917 Studies on the syrinx of Gallus domesticus. Journal of Morphology
349
20(1), pp. 165-215.
Neumann, 1898 Pternistis afer melanogaster. Journal of Ornithology 46(2, pl.3, f.1.), p.
299.
Neumann, 1902 Francolinus clappertoni nigrosquamatus. Ornithologische
Monatsberichte 10, p. 8.
Neumann, 1904 Francolinus castaneicollis gofanus. Journal of Ornithology 52, p. 353.
Neumann, 1907 Francolinus clappertoni heuglini. Ornithologische Monatsberichte 15,
p. 199.
Neumann, 1907 Francolinus icterorhynchus emini. Ornithologische Monatsberichte 15,
p. 98.
Neumann, 1907 Francolinus icterorhynchus ugandensis. Ornithologische
Monatsberichte 15, p. 199.
Neumann, 1908 Scleroptila levaillantoides pallidior. Bulletin of the British
Ornithologists´ Club 21, p. 45.
Neumann, 1908 Francolinus levaillanti benguellensis. Bulletin of the British
Ornithologists' Club 21, p. 44.
Neumann, 1908 Scleroptila shelleyi whytei. Bulletin of the British Ornithologists´ Club
21, p. 76.
Neumann, 1915 Francolinus bicalcaratus adamauae. Ornithologische Monatsberichte
23, p. 73.
Neumann, 1920 Pternistis afer krebsi. Journal of Ornithology 68, p. 78.
Neumann, 1920 Francolinus gariepensis ludwigi. Journal of Ornithology 68, p. 79.
Neumann, 1928 Francolinus clappertoni testi. Journal of Ornithology 76, p. 784.
Newman K 2002 Newman’s Birds of Southern Africa. Struik Publishers, Cape Town.
350
Nylander JAA, Ronquist F, Huelsenbeck JP, Nieves-Aldrey J 2004. Bayesian
phylogenetic analysis of combined data. Journal of Systematic Biology 53(1),
pp. 47-67.
Nylander JAA, Olsson U, Alström P, Sanmartín I 2008 Accounting for Phylogenetic
Uncertainty in Biogeography: a Bayesian approach to Dispersal-Vicariance
analysis of the Thrushes (Aves: Turdus). Systematic Biology 57(2), pp. 257-268.
Ogilvie-Grant, 1890 Francolinus griseostriatus. Ibis (pl.10), p. 349.
Ogilvie-Grant, 1890 Francolinus shelleyi shelleyi. Ibis, p. 348.
Ogilvie-Grant, 1891 Francolinus clappertoni gedgii. Ibis, p. 124.
Ogilvie-Grant, 1891 Pternistis jacksoni jacksoni. Ibis, p. 123.
Ogilvie-Grant, 1891 Francolinus psilolaema elgonensis. Ibis, p. 126.
Ogilvie-Grant, 1891 Francolinus streptophora. Ibis, p. 126.
Ogilvie-Grant, 1892 Francolinus coqui buckleyi. Ibis, p. 41.
Ogilvie-Grant, 1892 Francolinus shelleyi uluensis Ibis, p. 44.
Ogilvie-Grant, 1892 Francolinus clappertoni sharpii. Ibis p.47.
Ogilvie-Grant, 1896 Francolinus levaillantii crawshayi. Ibis (pl. 12), p.482.
Ogilvie-Grant, 1897 Francolinus levaillantii kikuyuensis. Bulletin of the British
Ornithologists´ Club 6, p. 23.
Ogilvie-Grant, 1902 Francolinus bicalcaratus thornei. Bulletin of the British
Ornithologists´ Club 13, p. 22.
Olson SL 1974 Telecrex restudied: a small Eocence guineafowl. Wilson Bulletin 86, pp.
246-250.
Oustalet, 1892 Francolinus icterorhynchus dybowski. Le Naturaliste 6(2), p. 232.
Päckert M, Martens J, Kosuch J, Nazarenko A, Veith M 2003 Phylogenetic signal in
351
the song of crests and kinglets (Aves: Regulus). Evolution 57(3), pp. 616-629.
Pagel M 1999 The maximum likelihood approach to reconstructing ancestral character
states of discrete characters on phylogenies. Systematic Biology 48, pp. 612–
622.
Payne RB1971 Duetting and chorus singing in African birds. Ostrich Supplementary 9,
pp. 125-146.
(Peters, 1854) Pternistis afer humboldtii. Monatschrifte der Deutschen Akademie der
Wissenschaften zu Berlin, p. 134.
Peters JL 1934 Check-list of Birds of the World. Vol II. Harvard University Press,
Cambridge.
Phillimore AB 2010 Subspecies origination and extinction in birds. The American
Ornithologists’ Union 67, pp. 42-53.
Posada D, Crandall KA 1998 Modeltest: testing the model of DNA substitution.
Bioinformatics 14, pp. 817-818.
Posada D, Buckley TR 2004 Model selection and model averaging in phylogenetics:
advantages of Akaike information criterion and Bayesian approaches over
likelihood ratio tests. Systematic Biology 53, pp. 793-808.
Pough FH 1990 Vertebrate life. MacMillan, New York.
Prager EM, Wilson AC 1976 Congruency of phylogenies derived from different proteins.
A molecular analysis of the phylogenetic position of cracid birds. Journal of
Molecular Evolution 9, pp. 45-57.
Price JJ, Lanyon SM 2002 Reconstructing the evolution of complex bird song in the
oropendalas. Evolution 56(7), pp. 1514-1529.
Primmer CR, Borge T, Lindell J, Saetre GP 2002 Single nucleotide polymorphism
352
characterization in species with limited available sequence information: high
nucleotide diversity revealed in the avian genome. Molecular Ecology 11, pp.
603-612.
Pruett CL, Winker K 2010 Alaska Song Sparrows (Melospiza Melodia) demonstrate
that genetic marker and method of analysis matter in subspecies assessments.
The American Ornithologists’ Union 67, 162-171.
Prum RO 1992 Syringeal morphology, phylogeny, and evolution of the Neotropical
manakins (Aves: Pipridae). American Museum Novitates 3043.
Prum RO, Lanyon WE 1989 Monophyly and phylogeny of the Schiffornis group
(Tyrannoidea). The Condor 91, pp. 444-461.
Rambaut A, Drummond AJ 2007 Tracer v1.4.
[http:// tree.bio.ed.ac.uk/software/tracer/].
Ree RH, Smith SA 2008 Maximum likelihood inference of geographic range evolution
by dispersal, local extinction and cladogenesis. Systematic Biology
57(1), pp. 4-14.
Reichenbach, ["1852"]1853 Ortygornis. Avium systema natural das natürliche System
der Vogel mit hundert Tafeln grosstentheils Original-Abbildungen der bis jetzt
entdecken fast zwölfhundert typischen Formen, p. 28.
Reichenow, 1885 Pternistes afer böhmi. Journal of Ornithology 33, p. 465.
Reichenow, 1887 Francolinus hildebrandti fischeri. Journal of Ornithologist 67, p. 51.
Reichenow, 1887 Pternistis rufopictus. Journal of Ornithology 35(177), p. 52.
Reichenow, 1889 Francolinus stuhlmanni. Journal of Ornithology, p. 270.
Reichenow, 1908 Francolinus nobilis nobilis. Ornithologische Monatsberichte 16, p.
81.
353
Reichenow, 1909 Pternistis afer harterti. Ornithologische Monatsberichte 17, p. 41.
Reichenow, 1909 Pternistis afer harterti. Ornithologische Monatsberichte 17, p. 41.
Reichenow, 1909 Pternistes afer intercedens. Ornithologische Monatsberichte 17, p. 88.
Reichenow, 1912 Francolinus lathami schubotzi. Journal of Ornithology 60, p.320.
Reichenow, 1919 Francolinus hildebrandti grotei. Journal of Ornithology 67, p. 334.
Remsen Jr. JV 2010 Subspecies as a meaningful taxonomic rank in avian
classification. The American Ornithologists’ Union 67, pp. 62-78.
Roberts A 1924 Synoptic check list of the birds of South Africa. Annals of the
Transvaal Museum. Vol. X. South Africa.
(Roberts, 1929) Francolinus swierstrai. Annals of the Transvaal Museum 13, p. 72.
Roberts, 1922 Dendroperdix. Annals of the Transvaal Museum 8, p. 194.
Roberts, 1924 Francolinus sephaena zuluensis. Annals of the Transvaal Museum 10, p.
78.
Roberts, 1924 Francolinus sephaena thompsoni. Annals of the Transvaal Museum 10,
p. 78.
Roberts, 1924, Francolinus afer notatus, A. “Synoptic Check-list of Birds of South
Africa.” Annals of the Transvaal Museum 10, pp. 89-125.
(Roberts, 1928) Francolinus hartlaubi bradfieldi. Annals of the Transvaal Museum 12,
p. 292.
(Roberts, 1928) Francolinus hartlaubi ovambensis. Annals of the Transvaal Museum
12, p. 293.
Roberts, 1931 Pternistis afer lehmanni. Annals of the Transvaal Museum 14, p. 238.
Roberts, 1932 Francolinus sephaena chobiensis. Annals of the Transvaal Museum 16,
p. 21.
354
Roberts, 1932 Francolinus jugularis cunenensis. Annals of the Transvaal Museum 15,
p. 22.
Roberts, 1932 Pternistis swainsoni gilli. Annals of the Transvaal Museum 15, p. 23.
Roberts, 1932 Francolinus gariepensis kalaharica. Annals of the Transvaal Museum
16, p. 22.
Roberts, 1932 Francolinus gariepensis langi. Annals of the Transvaal Museum 16, p.
22.
Roberts, 1932 Francolinus sephaena mababiensis Annals of the Transvaal Museum 16,
p. 22.
Roberts, 1931 Pternistis swainsoni damarensis. Annals of the Transvaal Museum 14, p.
238.
Roberts, 1932 Pternistis swainsoni chobiensis. Annals of the Transvaal Museum 15, p.
23.
Roberts, 1932 Pternistis afer cunenensis. Annals of the Transvaal Museum 15, p. 22.
Roberts, A 1940 The Birds of South Africa. 1st ed. Riverside Press, Edinburgh.
Roberts A, 1947 Pternistis cooperi. A new species of Pternistis from Salisbury, S.
Rhodesia. The Ostrich 18, p. 197.
Ronquist F 1996 DIVA, version 1.1. Computer program and manual available by
anonymous FTP from Uppsala University, Uppsala, Sweden. Available at:
ftp.uu.se or ftp.systbot.uu.se.
Ronquist F 1997 Dispersal-vicariance analysis: a new approach to the quantification of
historical biogeography. Systematic Biology 45, pp.195-203.
Ronquist F, Huelsenbeck JP 2003 MRBAYES 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19, pp.1572-1574.
355
Rosen DE 1978 Vicariant patterns and historical explanation in biogeography.
Systematic Zoology 27, pp. 159-188.
Rothschild, 1902 Francolinus coqui angolensis. Bulletin of the British Ornithologists'
Club 12, p. 76.
(Rüppell, 1835) Francolinus levaillantoides gutturalis. Neue Wirbelthiere zu der Fauna
von Abyssinien gehörig, p. 13.
(Rüppell, 1835) Francolinus erckelii erckelii. Neue Wirbelthiere zu der Fauna von
Abyssinien gehörig (pl.6), p. 12.
Salvadori, 1888 Francolinus castaneicollis castaneicollis. Annali del Museo Civico di
Storia Naturale di Genova 26, p. 542.
Salvadori, 1888 Francolinus sephaena spilogaster. Annali del Museo Civico di Storia
Naturale di Genova 26, p. 541.
Salvadori, 1898 Francolinus castaneicollis bottegi. Annali del Museo Civico di Storia
Naturale di Genova 38, p. 652.
Sande E, Dranzoa C, Wegge P, Carroll JP 2009 Home ranges and survival of
Nahan’s Francolin Francolinus nahani in Budongo Forest, Uganda. African
Journal of Ecology 47, pp. 457-462.
Schluter, Price T, Mooers AØ, Ludwig D 1997 Likelihood of ancestor states in
adaptive radiation. Evolution 51, pp. 1699-1711.
Sclater WL, 1921 Pternistis afer swynnertoni. Bulletin of the British Ornithologists´
Club 41, p. 134.
Sclater WL 1927 Francolinus levaillantoides archeri. Bulletin of the British
Ornithologists´ Club 48, p. 51.
Sclater WL, 1927 Francolinus coqui maharao. Bulletin of the British Ornithologists´
356
Club 48, p. 51.
Sclater WL, 1932 Francolinus coqui lynesi. Bulletin of the British Ornithologists' Club
62, p. 143.
(Scopoli, 1786) Francolinus pintadeanus pintadeanus. Deliciae florae faunae insubricae
2, p. 93.
Scotese CR 2001 Atlas of Earth History. Vol. 1. Paleogeography. PALEOMAP
project, Arlington, Texas.
Seller TJ 1983 Control of sound production in birds. Pp. 93-124. In: Bioacoustics: a
Comparative Approach (Lewis B, ed). Academic Press, London.
Serle W 1957 A contribution to the ornithology of the eastern region of Nigeria. Ibis
99, pp. 371-418, 628-685.
Sharpe, 1897 Francolinus levaillantoides lorti. Bulletin of the British Ornithologists´
Club 6, p. 47.
Shelley, 1894 Pternistis hildebrandti johnstoni. Ibis, p. 24.
Shen Y-Y, Liang, L, Sun Y-B, Yue B-S, Yang Xiao-J, Murphy RW, Zhang Y-P
2010 A mitogenomic perspective on the ancient, rapid radiation in the Galliformes
with an emphasis on the Phasianidae. BMC Evolutionary Biology 10, pp 132
doi:10.1186/1471-2148-10-132.
Shields GF, Wilson AC 1987 Calibration of mitochondrial DNA evolution in geese.
Journal of Molecular Evolution 24, pp. 212-217.
Sibley CG, Ahlquist JE 1985 The relationships of some groups of African birds, based
comparisons of the genetic material, DNA. Pp. 115-161. In: Proceedings of the
International symposium on African vertebrates (KL Schuchmann, ed). Museum
Alexander Koenig, Bonn.
357
Sibley CG, Ahlquist JE 1990 Phylogeny and Classification of Birds. Yale University
Press, New Haven.
Sibley CG, Monroe Jr. BL 1990 Distribution and Taxonomy of Birds of the World.
Yale University Press, New Haven.
Sinclair I, Ryan P 2003 Birds of Africa South of the Sahara. Struik publishers, Cape
Town.
Smith A, 1833 Francolinus natalensis natalensis. South African Quarterly Journal 2(1),
p. 48.
(Smith A, 1836) Francolinus coqui coqui. Report of the expedition for exploring central
Africa, p. 55.
Smith A, 1836 Francolinus levaillantoides levaillantoides. Report of the expedition for
exploring central Africa, p. 55.
(Smith A, 1836) Francolinus sephaena sephaena. Report of the expedition for exploring
central Africa, p. 55.
(Smith A, 1836) Pternistis swainsonii swainsonii. Report of the expedition for
exploring central Africa, p. 54.
Sorenson MD, Ast JC, Dimcheff DE, Yuri T, Mindell DP 1999 Primers for a PCRbased approach to mitochondrial genome sequencing in birds and other
vertebrates. Molecular Phylogenetics and Evolution 12, pp. 105-114.
Snow DW 1978 An Atlas of Speciation in African Non-passerine birds. British
Museum, London.
Staden R, Judge DP, Bonfield JK 2003 Analysing sequences using the Staden
package and EMBOSS. Introduction to Bioinformatics. A Theoretical and
Practical Approach. (Stephen A, Krawetz SA, Womble DD, eds). Human
358
Pressinc., Totawa, NJ.
Stamatakis A 2006 RAxML-VI-HPC: maximum likelihood-based phylogenetic
analyses with thousands of taxa and mixed models. Bioinformatics 22, pp. 26882690.
Stamatakis A, Hoover P, Rougemont J 2008 A rapid bootstrap algorithm for the
RAxML web-servers. Systematic Biology 75, pp. 758-771.
Stephens, 1819 Francolinus. General Zoology [Shaw] 11(pt.2), p. 316.
Steven P, Paulsen F, Tillmann B 2000 Orcein-Picroindigocarmine - a new multiple
Stain. Archives of Histology and Cytology 63(4), pp. 397-400.
Stevenson T, Fanshawe J 2002 Field Guide to the Birds of East Africa. T & AD
Poyser, London
Stidham TA 2008 The first fossil of the Congo peafowl (Galliformes: Afropavo). South
African Journal of Science 104, pp. 511-512.
Stoneham, 1930 Pternistis leucoscepus tokora. Bateleur 2, p. 113.
Stresemann E, 1937 Francolinus coqui hoeschianus subsp. nova. Ornithologische
Monatsberichte 45(2), pp. 66-67.
Stresemann E, 1939 Francolinus hartlaubi crypticus. Zwei neue Vogelrassen
aus Südwest-Afrika. Ornithologische Monatsberichte 47, pp. 61-62.
Swainson, 1837 Chaetopus. On the natural history and classification of birds[D
Lardner]. The Cabinet Cyclopaedia 2(92), p. 344.
Swofford DL 2002 PAUP*: Analysis Using Parsimony, Version 4. Sinaeur Associates,
Sunderland, Massachusetts.
(Temminck, 1815) Francolinus gularis. Histoire naturelle generale des pigeons et des
gallinaces 3, pp. 401, 731.
359
Temminck, 1854 Francolinus ahantensis ahantensis. Bijdragen Tot De Dierkunde
1(pt6), p. 49.
Templeton AR 1989 The meaning of species and speciation: a genetic perspective. Pp
3-27. In: Speciation and its consequences (Otte D, Endler JA, eds). Sinauer
Associates, Sunderland, Massachusetts.
Tsukahara N, Yang Q, Sugita S 2008 Structure of the syringeal muscles in Jungle
Crow (Corvus macrorhynchus). Anatomical Science International 83, pp. 152158.
(Valenciennes, 1825) Francolinus levaillantii levaillantii. Dictionnaire des Sciences
Naturelles 38, p. 441.
Van Someren, 1926 Francolinus coqui ruahdae. Journal of the East Africa and Uganda
Natural History Society 25, p. 34.
Van Niekerk JH 2010 Vocal behaviour of Crested Francolin Ortygornis sephaena in
response to playback calls. Ostrich - Journal of African Ornithology 81(2), pp.
149-154.
Van Someren VGL, 1938 Francolinus africanus macarthuri. A new race of Grey-wing
Francolin from Kenya Colony. Bulletin of the British Ornithologists´ Club 59, p.
7.
Van Someren VGL, 1939 Francolinus squamatus chyuluensis. Reports on the Corydon
Museum Expedition to the Chyulu Hills. 2. The birds of the Chyulu Hills.
Journal of the East Africa and Uganda Natural History Society 14, pp. 15-129.
Verheyen R 1956 Contribution a l’anatomie et a la systematique des Galliformes.
Bulletin de l'Institut Rroyal des Sciences Naturelles de Belgique 32, pp. 1-24.
360
Voelker G, Outlaw RK, Bowie RCK 2010 Pliocene forest dynamics as a primary
driver of African bird speciation. Global Ecology and Biogeography 19, pp.
111-121.
Von Erlanger, 1904 Pternistes leucoscepus höltemulleri. Ornithologische Monatsberichte
12, p. 98.
Von Erlanger, 1904 Pternistes leucoscepus muhamed-ben-abdullah. Ornithologische
Monatsberichte 12, p. 98.
Von Erlanger, 1905 Francolinus psilolaema ellenbecki. Journal of Ornithology 63, p.
151.
Von Heuglin, 1873 Francolinus sephaena schoanus. Systematische Uebersicht der
Vögel Nord-Ost-Afrika's mit Einschluss der arabischen Kueste des Rothen
Meeres und der Nil-Quellen-Länder 2(pt.l, pl.29), p. 891.
Wagler, 1832 Pternistis. Isis von Oken 25, col. 1229.
Wang N, Kimball RT, Braun EL, Liang B, Zhang Z 2013 Assessing phylogenetic
relationships among Galliformes: a multigene phylogeny with expanded taxon
sampling in Phasianidae. Public Library of Science One 8(5), pp. e64312.
doi:10.1371/journal.pone.0064312.
Waterhouse, 1838 Francolinus adspersus adspersus. An Expedition of the discovery
into the Interior of Africa[Alexander] 2 App., p. 267.
Wetmore A 1960 A classification for the birds of the world. Smithson. Smithsonian
Miscellaneous Collections 139(1), pp.1-37.
Whistler H, 1941 Francolinus pondicerianus ceylonensis. Recognition of new
subspecies of Birds in Ceylon. Ibis 5(14), pp. 319-320.
White CMN, 1944 Francolinus albogularis meinertzhageni. A new race of
361
Francolin from Northern Rhodesia. Bulletin of the British Ornithologists' Club
65, pp. 7-8.
White CMN, 1944 Francolinus levaillantii clayi. A new race of Scrub Robin and a
New race of Red-winged Francolin from Northern Rhodesia. Bulletin of the
British Ornithologists´ Club 64, pp. 49-50.
White CMN, 1945 Francolinus coqui kasaicus. The ornithology of the KaondeLunda province, Northern Rhodesia. Part III. Systematic List. Ibis 87, pp. 309345.
White CMN, 1947 Pternistis swainsonii lundazi. Bulletin of the British Ornithologists´
Club 67, pp.72-73.
Winker K 2010 Subspecies represent geographically partitioned variation, a gold mine
of evolutionary biology, and a challenge for conservation. The American
Ornithologists’ Union 67, pp. 6-23.
Wolters HE 1975-82 Die Vogelarten der Erde. Paul Parey, Hamburg & Berlin.
Wortley AH, Scotland RW 2006 The effect of combining molecular and
morphological data in published phylogenetic analyses. Systematic
Biology 55 (4), pp. 677-685.
Zarudny, 1906 Francolinus francolinus bogdanovi. Ornithologische Monatsberichte
14, p. 151.
Zarudny & Härms, 1913 Francolinus pondicerianus mecranensis. Ornithologische
Monatsberichte 21, p. 53.
Zarudny & Härms, 1913 Francolinus francolinus arabistanicus. Ornithologische
Monatsberichte 21, p. 54.
Zedlitz, 1913 Francolinus sephaena jubaensis. Ornithologische Monatsberichte 21, p.
362
59.
Zimmer KJ, Robbins MB, Kopuchian C 2008 Taxonomy, vocalizations, syringeal
morphology and natural history of Automolus roraimae (Furnariidae). Bulletin
of the British Ornithologists' Club 128(3), pp 14-33.
Zink RM 2004 The role of subspecies in obscuring avian biological diversity and
misleading conservation policy. Proceedings of the Royal Society of London,
Series B 271, pp. 561-564.
363