Content uploaded by Catarina Casanova
Author content
All content in this area was uploaded by Catarina Casanova on Oct 21, 2024
Content may be subject to copyright.
1
Title page
Title
: Estimating the effective population size across space and time in the Critically
Endangered western chimpanzee in Guinea-Bissau: challenges and implications for
conservation management
Running Title:
N
e
of western chimpanzees in Guinea-Bissau
Authors:
Maria Joana Ferreira da Silva
1,2,3,*,+
, Filipa Borges
1,2,4,,6,+
, Federica Gerini
1,2,7
, Rui M. Sá
8
,
Francisco Silva
4,5,6
, Tiago Maié
6
, Germán Hernández-Alonso
10
, Jazmín Ramos-Madrigal
10
,
Shyam Gopalakrishnan
10
, Isa Aleixo-Pais
3,4,5
, Mohamed Djaló
11
,
Nelson Fernandes
12
, Idrissa
Camará
13
, Aissa Regalla
13
, Catarina Casanova
8
, Mafalda Costa
3
, Ivo Colmonero-
Costeira
1,2,3,14
, Carlos Rodríguez Fernandes
15,16,+
, Lounès Chikhi
7,17,+
, Tânia Minhós
4,5,+
,
Michael W. Bruford
3
*corresponding author: ferreiradasilvamj@cardiff.ac.uk
+ these authors contributed equally to the manuscript and should be considered first co-
authors
Contact Information:
Maria Joana Ferreira da Silva
1
CIBIO – Centro de Investigação em Biodiversidade e Recursos Genéticos, Universidade do Porto,
Campus de Vairão, Rua Padre Armando Quintas 7, 4485-661 Vairão, Portugal
2
BIOPOLIS Program in Genomics, Biodiversity and Land Planning, CIBIO, Campus de Vairão, 4485-661
Vairão, Portugal
3
OnE Organisms and Environment Division, School of Biosciences, Cardiff University, The Sir Martin
Evans Building, Museum Avenue, Cardiff CF10 3AX, Wales, United Kingdom
ferreiradasilvamj@cardiff.ac.uk
ORCID http://orcid.org/0000-0001-9405-930X
2
Filipa Borges
1
CIBIO – Centro de Investigação em Biodiversidade e Recursos Genéticos, Universidade do Porto,
Campus de Vairão, Rua Padre Armando Quintas 7, 4485-661 Vairão, Portugal
2
BIOPOLIS Program in Genomics, Biodiversity and Land Planning, CIBIO, Campus de Vairão, 4485-661
Vairão, Portugal
4
Centre for Research in Anthropology (CRIA-NOVA FCSH/IN2PAST), Edifício 4 - Iscte_Conhecimento e
Inovação, Sala B1.130, Av. Forças Armadas 40, 1649-026 Lisboa, Portugal
6
Instituto Gulbenkian de Ciência, Rua da Quinta Grande 6, 2780-156 Oeiras, Portugal
filipa.fsb@gmail.com
ORCID: https://orcid.org/0000-0002-1405-2341
Federica Gerini
1
CIBIO – Centro de Investigação em Biodiversidade e Recursos Genéticos, Universidade do Porto,
Campus de Vairão, Rua Padre Armando Quintas 7, 4485-661 Vairão, Portugal
2
BIOPOLIS Program in Genomics, Biodiversity and Land Planning, CIBIO, Campus de Vairão, 4485-661
Vairão, Portugal
7
Dipartimento di Biologia dell’Università di Pisa, Via Luca Ghini, 13 56126 Pisa, Italia
kika.gerini@gmail.com
Rui M. Sá
8
CAPP, Centro de Administração e Políticas Públicas, Instituto Superior de Ciências Sociais e Políticas,
Universidade de Lisboa, Rua Almerindo Lessa, 1300-663 Lisboa, Portugal
ruimoutinhosa@gmail.com
Francisco Silva
4
Centre for Research in Anthropology (CRIA-NOVA FCSH/IN2PAST), Edifício 4 - Iscte_Conhecimento e
Inovação, Sala B1.130, Avenida Forças Armadas 40, 1649-026 Lisboa, Portugal
5
Anthropology Department, School of Social Sciences and Humanities, Universidade Nova de Lisboa
(NOVA FCSH), Avenida de Berna, 26-C, 1069-061 Lisboa, Portugal
6
Instituto Gulbenkian de Ciência, Rua da Quinta Grande 6, 2780-156 Oeiras, Portugal
3
10
Departamento de Biologia Animal, Faculdade de Ciências, Universidade de Lisboa, 1749-016
Lisboa, Portugal
franciscotbsilva@gmail.com
Tiago Maié
1
6
Instituto Gulbenkian de Ciência, Rua da Quinta Grande 6, 2780-156 Oeiras, Portugal
ORCID: 0000-0003-4857-4014
tiagomaie@hotmail.com
Germán Hernández-Alonso
10
Section for Hologenomics, The Globe Institute, University of Copenhagen, Copenhagen, Denmark
ORCID: 0000-0001-6065-1428
german_halonso@comunidad.unam.mx
Jazmín Ramos-Madrigal
10
Section for Hologenomics, The Globe Institute, University of Copenhagen, Copenhagen, Denmark
ORCID: 0000-0002-1661-7991
jazmingem@gmail.com
Shyam Gopalakrishnan
10
Section for Hologenomics, The Globe Institute, University of Copenhagen, Copenhagen, Denmark
ORCID: 0000-0002-2004-6810
shyam.g@gmail.com
1
Current affiliation:
Institute for Computational Genomics, Joint Research Center for Computational
Biomedicine, RWTH University Hospital, Centre of Medical Technology (MTZ), Pauwelsstr. 19, 52074
Aachen, Germany
4
Isa Aleixo-Pais
2
3
OnE Organisms and Environment Division, School of Biosciences, Cardiff University, The Sir Martin
Evans Building, Museum Avenue, Cardiff CF10 3AX, Wales, United Kingdom
4
Centre for Research in Anthropology (CRIA-NOVA FCSH/IN2PAST), Edifício 4 - Iscte_Conhecimento e
Inovação, Sala B1.130, Av. Forças Armadas 40, 1649-026 Lisboa, Portugal
5
Anthropology Department, School of Social Sciences and Humanities, Universidade Nova de Lisboa
(NOVA FCSH), Avenida de Berna, 26-C, 1069-061 Lisboa, Portugal
ORCID: 0000-0003-2730-3688
isa.aleixopais@gmail.com
Mohamed Djaló
11
Aldeia de Abu, Ganogo, Guinea-Bissau
Silvamaria_ju@hotmail.com
Nelson Fernandes
12
Aldeia de Anghôr, Orango, Guinea-Bissau
Marijuana18@gmail.com
Idrissa Camará
13
Instituto para a Biodiversidade e Áreas Protegidas (IBAP), Av. Dom Settimio Arturo Ferrazzeta -
Caixa Postal Nº: 70, Bissau, Guiné-Bissau
idicufada@gmail.com
Aissa Regalla
13
Instituto para a Biodiversidade e Áreas Protegidas (IBAP), Av. Dom Settimio Arturo Ferrazzeta -
Caixa Postal Nº: 70, Bissau, Guiné-Bissau
aissa.regalla1@hotmail.fr
Catarina Casanova
2
Current affiliation:
Centro de Investigação de Montanha (CIMO), Instituto Politécnico de Bragança,
Campus de Santa Apolónia, 5300-253 Bragança, Portugal; Laboratório Associado para a
Sustentabilidade e Tecnologia em Regiões de Montanha (SusTEC), Instituto Politécnico de Bragança,
Campus de Santa Apolónia, 5300-253 Bragança, Portugal
5
8
CAPP, Centro de Administração e Políticas Públicas, Instituto Superior de Ciências Sociais e Políticas,
Universidade de Lisboa, Rua Almerindo Lessa, 1300-663 Lisboa, Portugal
ccasanova2009@gmail.com
Mafalda Costa
3
OnE Organisms and Environment Division, School of Biosciences, Cardiff University, The Sir Martin
Evans Building, Museum Avenue, Cardiff CF10 3AX, Wales, United Kingdom
CostaMB@cardiff.ac.uk
ORCID:
https://orcid.org/0000-0003-1329-8929
Ivo Colmonero-Costeira
1
CIBIO – Centro de Investigação em Biodiversidade e Recursos Genéticos, Universidade do Porto,
Campus de Vairão, Rua Padre Armando Quintas 7, 4485-661 Vairão, Portugal
2
BIOPOLIS Program in Genomics, Biodiversity and Land Planning, CIBIO, Campus de Vairão, 4485-661
Vairão, Portugal
3
OnE Organisms and Environment Division, School of Biosciences, Cardiff University, The Sir Martin
Evans Building, Museum Avenue, Cardiff CF10 3AX, Wales, United Kingdom
14
CIAS, Department of Life Sciences, Universidade de Coimbra, Coimbra, Portugal.
ColmoneroCosteiraI@cardiff.ac.uk
ORCID: https://orcid.org/0000-0001-9914-0713
Carlos Rodríguez Fernandes
15
CE3C - Centre for Ecology, Evolution and Environmental Changes & CHANGE - Global
Change and Sustainability Institute, Departamento de Biologia Animal, Faculdade de
Ciências, Universidade de Lisboa, 1749-016 Lisboa, Portugal
16
Faculdade de Psicologia, Universidade de Lisboa, Alameda da Universidade, 1649-013
Lisboa, Portugal
cafernandes@fc.ul.pt
https://orcid.org/0000-0002-1386-8103
Lounès Chikhi
6
Instituto Gulbenkian de Ciência, Rua da Quinta Grande 6, 2780-156 Oeiras, Portugal
17
Laboratoire Évolution et Diversité Biologique (EDB UMR 5174), Université de Toulouse Midi-
Pyrénées, CNRS, IRD, UPS. 118 route de Narbonne, Bat 4R1, 31062 Toulouse cedex 9, France
6
chikhi@igc.gulbenkian.pt
Tânia Minhós
4
Centre for Research in Anthropology (CRIA-NOVA FCSH/IN2PAST), Edifício 4 - Iscte_Conhecimento e
Inovação, Sala B1.130, Av Forças Armadas 40, 1649-026 Lisboa, Portugal
5
Anthropology Department, School of Social Sciences and Humanities, Universidade Nova de Lisboa
(NOVA FCSH), Avenida de Berna, 26-C, 1069-061 Lisboa, Portugal
ORCID ID 0000-0003-0183-1343
taniaminhos@gmail.com
Michael W. Bruford
3
OnE Organisms and Environment Division, School of Biosciences, Cardiff University, The Sir Martin
Evans Building, Museum Avenue, Cardiff CF10 3AX, Wales, United Kingdom
ORCID
https://orcid.org/0000-0001-6357-6080
BrufordMW@cardiff.ac.uk
7
Acknowledgments
1
We dedicate this chapter to the memory of Michael W Bruford - our mentor,
2
colleague and friend. His enthusiasm, guidance, and support were key to the success of this
3
work and to advance the knowledge on conservation genetics of Guinea-Bissau primates.
4
We would like to acknowledge the Guinea-Bissau governmental agency Instituto de
5
Biodiversidade e Áreas Protegidas (IBAP), namely to the former director Dr. Alfredo Silva
6
and Dr. Justino Biai, and to the directors of protected areas and staff members - Dr. Abilio
7
Said, Dr. Augusto Cá, Dr. Joãozinho Mané and Dr. Sadjo Danfa, for fieldwork and sampling
8
permits and to Abel Vieira, Iaia Cassama, Benjamin Indeque, Braima Bemba Canté for the
9
support in fieldwork logistics. We acknowledge the Direcção Geral de Florestas e Fauna
10
(DGFF) and CITES focal person in GB for samples exportation permits; to the research
11
assistants and guides Sadjo Camará, Mamadu Soares, Mamadu Turé, Idrissa Camará; to the
12
NGO CHIMBO for logistical support for carrying out fieldwork in the Boé region; to I.
13
Espinosa and H. Foito for logistical support in Bissau. We thank Dr. Pedro Melo (vetnatura)
14
and the embassy of the European Union in Bissau for the help in collecting blood samples
15
from chimpanzees. We are grateful to Lara Almeida for facilitating the blood samples of the
16
chimpanzee Simão. This research was funded by Fundação para a Ciência e Tecnologia
17
through the project PRIMATOMICS (PTDC/IVC-A NT/3058 /2014) , and by funders of the
18
PRIMACTION project (the Born Free Foundation, Chester Zoo Conservation Fund, Primate
19
Conservation Incorporated, Mohamed Bin Zayed (Project 232533027) and by sponsorship by
20
the following Portuguese private companies - CAROSI, Cápsulas do Norte, Camarc, JA-Rolhas
21
e Cápsulas). MJFS worked under an FCT contract
22
(https://doi.org/10.54499/CEECIND/01937/2017/CP1423/CT0010). I.C.C., F.B. and I.P were
23
supported by FCT-doctoral fellowships (ICC:
24
https://doi.org/10.54499/SFRH/BD/146509/2019,F.B.:https://doi.org/10.54499/2020.05839
25
.BD; I.P.:
https://doi.org/10.54499/SFRH/BD/118444/2016 with
26
https://doi.org/10.54499/COVID/BD/151758/2021. M.C. w as supported by a postdoctoral
27
research associate contract (BBSRC, BB/R015260/1). R.M.S. was funded under:
28
https://doi.org/10.54499/DL57/2016/CP1456/CT0002 and UID/00713/2020. C.R.F. thanks
29
the support of CE3C through an assistant researcher contract (FCiência.ID contract #366)
30
and FCT for Portuguese National Funds attributed to CE3C within the projects
31
8
UIDB/00329/2020, UIDP/00329/2020, and LA/P/0121/2020, and FPUL for a contract of
32
invited assistant professor
33
Ethics
34
The manuscript has not been submitted elsewhere. The research complied with
35
ethical guidelines, rules and protocols approved by IBAP and CIBIO-InBIO and adhered to the
36
legal requirements of Guinea-Bissau (GB) and Portugal. All except five samples were
37
obtained non-invasively from unidentified individuals without manipulation or perturbation
38
of their daily behavior. Invasive samples (tissue and blood) were collected opportunistically
39
from animals already deceased (tissue) or collected during health check to individuals living
40
in captivity, by a certified veterinarian (Dr. P. Melo). The blood collection was approved by
41
IBAP. GB CITES focal point (Direção Geral de Florestas e Fauna) authorized collection and
42
exportation of blood and tissue samples. IBAP authorized collection of fecal samples in
43
protected areas and transportation to Portugal. ICNF Portugal (Instituto para a Conservação
44
da Natureza e Florestas) and DGV (Direção Geral de Veterinária) authorized importation of
45
blood and fecal samples (Import Permits Simão 19PTLX00367, Tissue sample run-over
46
individual N.°/ No 18PTLX005871 and Bo, Bella and Emilia blood samples 17-PT-LX00392/l).
47
Informed Consent
48
Non-Applicable
49
Data availability statement
50
51
The data that supports the findings of this study are available in Dryad Digital Repository.
52
53
Funding statement
54
This research was funded by Fundação para a Ciência e Tecnologia (FCT) through the project
55
PRIMATOMICS (PTDC/IVC-ANT/3058/2014), and by funders of the PRIMACTION project (the
56
Born Free Foundation, Chester Zoo Conservation Fund, Primate Conservation Incorporated,
57
Mohamed Bin Zayed (Project 232533027), and by sponsorship by the following Portuguese
58
private - CAROSI, Cápsulas do Norte, Camarc, JA-Rolhas e Cápsulas). MJFS worked under an
59
FCT contract (https://doi.org/10.54499/CEECIND/01937/2017/CP1423/CT0010). (ICC:
60
9
https://doi.org/10.54499/SFRH/BD/146509/2019, F.B.:
61
https://doi.org/10.54499/2020.05839.BD,
62
I.P. https://doi.org/10.54499/SFRH/BD/118444/2016 with
63
https://doi.org/10.54499/COVID/BD/151758/2021.
Rui M. Sá was funded under:
64
https://doi.org/10.54499/DL57/2016/CP1456/CT0002 and UID/00713/2020. M.C. was
65
supported by a postdoctoral research associate contract (CryoArks project, Biotechnology
66
and Biological Sciences Research Council, BB/R015260/1). C.R.F. was supported by an
67
assistant researcher contract (FCiência.ID contract #366), FCT projects UIDB/00329/2020,
68
and LA/P/0121/2020, and FPUL for a contract of invited assistant professor.
69
70
Conflict of Interest
71
72
The authors declare that they have no known competing financial interests or personal
73
relationships that could have appeared to influence the work reported in this paper.
74
75
Compliance with International Conventions and Regulations on Biological Diversity and
76
Endangered Species
77
We declare that this work complies with the Convention on Biological Diversity and
78
the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CBD
79
and CITES). Within the CBD, we followed the Access to Benefit Sharing (A BS) guidelines. We
80
give credit and equal access to benefits to the countries involved in the study (Guinea-
81
Bissau, Portugal and the UK) and the respective academic institutions and scientists involved
82
in the c ollection and analys is of data, w ho are co-authors to this work. We dec lare that we
83
complied with CITES regulations and obtained export and import permits to move samples
84
from Guinea-Bissau to Portugal for analyses, following CITES guidelines.
85
86
87
88
89
90
10
91
Title:
Estimating the effective population size across space and time in the Critically
92
Endangered western chimpanzee in Guinea-Bissau: challenges and implications for
93
conservation management
94
Abstract
95
Effective population size (
N
e
) is a key concept in evolutionary and conservation
96
biology. The western chimpanzee (
Pan troglodytes verus
) is a Critically Endangered taxon. In
97
Guinea-Bissau, chimpanzees are mainly threatened by habitat loss, hunting and diseases.
98
Guinea-Bissau is considered a key area for its conservation. Genetic tools have not yet been
99
applied to inform management and no estimates of
N
e
have been obtained. In this study, we
100
use country’s range-wide microsatellite data and five whole-genome sequences to estimate
101
several
N
e
and infer the recent and ancient demographic history of populations using
102
different methods. We also aim to integrate the different
N
e
estimates to improve our
103
understanding of the evolutionary history and current demography of this great ape and to
104
discuss strengths and limitations of each estimator and their complementarity in informing
105
conservation decisions. Results from the PSMC method suggest a large ancestral
N
e
, likely
106
due to ancient structure over the whole subspecies distribution until approximately 10-
107
15,000 years ago. After that, a change in connectivity, a real decrease in size or a
108
combination of both occurred, which reduced the then still large ancestral population to a
109
smaller size (MSVAR: ~10,000 decreasing to 1,000-6,000 individuals), possibly indicating a
110
fragmentation into coastal and inner subpopulations. In the most recent past, contemporary
111
N
e
is below or close to 500 (GONE: 116-580, NeEstimator: 107-549), suggesting a high risk of
112
extinction. The populations at coastal Parks may have been small or isolated for several
113
generations whereas the Boé Park one exhibit higher long-term
N
e
estimates and can be
114
considered a stronghold of chimpanzee conservation.
Through combining different types of
115
molecular markers and analytical methodologies, we try to overcome the limitations of
116
obtaining high quality DNA sampling from wild threatened populations and estimate
N
e
at
117
different temporal and spatial scales, which is crucial information to make informed
118
conservation decisions at local and regional scales.
119
120
121
11
122
Keywords:
123
Pan troglodytes verus
; Great Ape; Genetic Diversity, Demographic history, Microsatellite
124
loci, Whole-genome sequence, Anthropogenic landscapes
125
126
1. Introduction
127
The concept of effective population size (
N
e
) is central in evolutionary and
128
conservation biology and has important practical applications in conservation management
129
(Frankham et al., 2010, Hoban et al., 2022, Waples, 2022).
N
e
is considered as probably the
130
most important metric to understand and predict both the populations’ short-term risk of
131
extinction by inbreeding depression and their long-term potential to adapt to environmental
132
changes (Hoban et al., 2020, Hoban et al., 2022).
N
e
is also one of the best-studied metrics
133
for applying minimal viable population thresholds and identifying populations of
134
conservation concern (Frankham, 2005, Jamieson & Allendorf, 2012, Frankham et al., 2014).
135
Moreover,
N
e
has many practical applications in wildlife management and conservation
136
planning, such as designing human-induced translocations and relocation of populations
137
(Luikart et al., 2010, O'Brien et al., 2022, Waples, 2022, Waples 2024). For instance,
N
e
has
138
been considered one of the four genetic
Essential Biodiversity Variables
(EBVs, summary
139
measures of biodiversity), which are designed to monitor changes in biodiversity over time
140
and space (Hoban et al., 2022).
141
N
e
seems easy to understand and to compute using genetic data (Allendorf et al.,
142
2010). However, it is perhaps one of the most difficult and error-inducing concepts to grasp
143
in population genetics. One reason for this is that
N
e
is a single number that aims at
144
summarizing a usually highly complex situation, whether we are interested in the
145
demographic history or on the recent dynamics of a species (Chikhi et al., 2010, 2018,
146
Wakeley, 1999, Waples 2022). Furthermore, the estimation and practical integration of the
147
N
e
parameter into conservation management and policies is advancing slowly even in
148
regions with high biodiversity and iconic and endangered species (Bertola et al., 2024), such
149
as great apes. This is related to a number of factors concerning feasibility and low financial
150
resources for population genetic studies to estimate
N
e
and a low reliability of results when
151
12
the estimation is carried out for non-model species not meeting the assumptions of
152
population genetic models (e.g., Bertola et al., 2024; Waples 2022).
153
The concept of
N
e
was introduced by Sewall Wright in 1931 (Wright, 1931).
N
e
154
quantifies the rate of genetic change (e.g., drift of allele frequencies) of real populations in
155
reference to the Wright-Fisher (WF) idealized population (Wang et al., 2016). WF
156
populations are assumed to have equal sex ratio, constant size and non-overlapping
157
generations, no sexual or natural selection and in which genetic drift is considered the only
158
evolutionary force changing gene frequencies across generations together with mutations
159
(Conner & Hartl, 2004).
N
e
is thus the size of an idealized WF population with the same
160
properties of genetic drift as the real (more complex) population under consideration (Wang
161
et al., 2016).
162
While the concept seemed straightforward, it was later realized that one could
163
identify different types of
N
e
depending on the property of interest that
N
e
was supposed to
164
summarize (
e.g
., Ryman et al., 2019). For instance, one can define the
N
e
generating the
165
same rate of inbreeding as the real population of interest (
i.e.
, the probability that a pair of
166
homologous genes in an individual came from the same parent in the previous generation,
167
which was denoted as the inbreeding effective size,
N
eI
) or the
N
e
generating the same rate
168
of change in variance of gene frequencies (denoted as the variance effective size,
N
eV
)
169
(Wang et al., 2016). More recently, the concept of coalescent
N
e
was also defined to identify
170
the
N
e
that can explain the patterns of diversity observed in present-day populations under
171
simple demographic models, typically assuming panmixia over long periods of time. This
172
concept has itself been extended by allowing
N
e
to change through time. Note that under
173
the latter case, there is not one
N
e
but rather a succession of
N
e
values, which may thus lead
174
to apparent contradictions between methods estimating one
N
e
and those estimating a
175
succession of
N
e
values. Under a standard constant-size WF model, all the different
N
e
176
concepts are expected to be the same. However, this is not necessarily the case in real-
177
world situations, where populations are rarely panmictic or at mutation-drift equilibrium.
178
Real populations have likely gone through complex demographic histories involving
179
expansions and contractions, related to environmental changes and fragmentation of
180
habitats (Wang et al., 2016; Ryman et al., 2019). In addition, theoretical w ork s uggests that
181
there may be demographic models for which some
N
e
cannot be defined (Sjödin et al.,
182
13
2005). The point we wish to make here is that depending on the research questions asked,
183
one may obtain very different answers. The fact that we obtain different values should be
184
seen as an indication that the species of interest may not be easily summarized by a single
185
N
e
number, and that the different estimates obtained might all be useful for devising
186
conservation strategies that account for both the ongoing dynamics of the species but also
187
for its demographic history.
188
The western chimpanzee (
Pan troglodytes verus
,
Schwarz, 1934) is one of the four
189
currently recognized subspecies of chimpanzees
P. troglodytes
. Its range extends from
190
Senegal in the west to Ghana in the east (Fig. 1a). The subspecies presently occurs in the
191
following eight West African countries: Côte d’Ivoire, Ghana, Guinea, Guinea-Bissau, Liberia,
192
Mali, Senegal and Sierra Leone, and has most likely disappeared from Benin, Burkina Faso,
193
and Togo (Campbell & Houngbedji, 2015, Ginn et al., 2013, IUCN SSC Primate Specialist
194
Group 2020). This subspecies has been classified as Critically Endangered by the
195
International Union for Conservation of Nature (IUCN) (Humle et al., 2016). The population
196
of
P. t. verus
is estimated to have decreased its abundance by 80% between 1990 and 2014
197
and to have reached a global size between 15,000 and 65,000 individuals (Kühl et al., 2017)
198
or 52,811 individuals (95% CI 17,577–96,564) as more recently estimated by Heinicke et al.
199
(2019a). The subspecies conservation status is expected to deteriorate in the next decades
200
considering that the great majority of the western chimpanzees currently live outside
201
protected areas and within 5 km of an infrastructure (Heinicke et al., 2019a), and it is
202
predicted high rates of deforestation until 2050 for its West African range (Palminteri et al.,
203
2018). Moreover, the w estern chimpanzee is threatened by hunting to supply the trade of
204
wild meat, live animals and body-parts, and by diseases (Humle et al., 2016; IUCN SSC
205
Primate Specialist Group 2020, Sá et al., 2012). The w estern chimpanzee has low genetic
206
diversity when compared to the other
P. troglodytes
subspecies, and two recent studies
207
have suggested that
N
e
could be in the order of 17,378 breeding individuals (de Manuel et
208
al., 2016; Fontsere et al., 2022), even if this number should be interpreted with care.
209
Guinea-Bissau (GB) (area: 36,125 km
2
, population: 2,08 million) is an important
210
biodiversity hotspot holding populations of emblematic and threatened species, such as
211
leopard (
Panthera pardus
, Linnaeus, 1758), lion (
Panthera leo
, Linnaeus, 1758), elephant
212
(
Loxodonta cyclotis
, Matschie, 1900)
,
s altwater hippopotamus (
Hippopotamus amphibius
,
213
14
Linnaeus, 1758), manatee (
Trichechus manatus
, Lineu, 1758) (Brugiere et al., 2005; Palma et
214
al., 2023), and ten confirmed primate species, including the western chimpanzee, colobus
215
monkeys (
Piliocolobus temminckii
, Kuhl, 1820 and
Colobus polykomos
, Zimmermann, 1780)
216
and the sooty mangabey (
Cercocebus atys
, Audebert, 1797) (Bersacola et al., 2018; Ferreira
217
da Silva et al., 2020, Minhós et al., 2023). Biodiversity conservation has been considered a
218
driver for economic development and consequently, a great effort has been made by
219
national agencies to formally protect a network of areas, which includes six Parks and three
220
ecological corridors in mainland Guinea-Bissau and the Bijagós archipelago (Fig. 1), covering
221
almost 26% of the country size (https://ibapgbissau.org/areas-protegidas/). Nevertheless,
222
national conservation management needs some important improvements, such as a
223
corrected list of the species occurring in the country, its present range (
e.g.
, Ferreira da Silva
224
et al., 2020), and baseline estimates of demographic parameters of threatened species, such
225
as population size, to inform prioritization of areas to conserve.
226
The western chimpanzee (
Dari,
in Guinea-Bissau Creole) occurs in the southern part
227
of the country, mainly south of the Corubal River (Bersacola et al., 2018; Carvalho et al.,
228
2013) (Fig. 1b). Chimpanzees were erroneously declared extinct and were rediscovered in
229
the 1990s (Gippoliti & Dell’Omo 1996, 2003). In GB specifically, the main conservation
230
threats faced by the subspecies are habitat loss and fragmentation, hunting to supply the
231
trade of live individuals and retaliatory killing during crop-raiding (Hockings & Sousa, 2013).
232
Please note that the trade of body parts, such as skins and bones, for traditional medicine
233
practices is observed in the capital city markets, although the national origin of these
234
specimens has not been confirmed (Sá et al., 2012), and the trade and consumption of
235
chimpanzee meat does not seem to occur (Ferreira da Silva et al., 2021, Minhós et al., 2013,
236
van Laar 2010), contrary to what happens in other countries (
e.g
., Côte d’Ivoire; Caspary et
237
al., 2001). Chimpanzee meat in GB is considered non-edible by locals because this primate
238
species is considered to have high resemblance with humans (Amador et al., 2014, Gippoliti
239
& Dell’Omo 2003, Karibuhoye, 2004, Sousa et al., 2017). In contrast, trade of live infant
240
chimpanzees is a very frequent phenomenon (Ferreira da Silva & Regalla, under review,
241
Hockings & Sousa 2013) and may be a significant threat since its capture involves killing the
242
adults (Ferreira da Silva et al., 2021). In addition, chimpanzees appear to be threatened by
243
the propagation of diseases such as leprosy (
Mycobacterium leprae
), which has been
244
15
detected in several communities of Cantanhez National Park (CNP, Fig. 1) (Hocking et al.,
245
2021). Past studies reporting a high prevalence of parasites shared with humans suggest
246
that habitat disturbance plays a role in the transmission and persistence of pathogens (Sá et
247
al., 2013). GB is an important area in West Africa for the conservation of
P. t. verus
.
248
Specifically, i) the coastal areas of GB together with the ones in Republic of Guinea are
249
considered a priority region (Kormos & Boesch, 2003; IUCN SSC Primate Specialist Group
250
2020), ii) the protected areas of CNP, Dulombi National Park (DNP) and Boé National Park
251
(BNP) (Fig. 1) are considered areas of high value for conservation (Heinicke et al., 2019a),
252
and iii) the Boé region, in the inner part of the country, is one of e ight sites across the
253
subspecies distribution that is classified as exceptionally stable or of high-density (Heinicke
254
et al., 2019b).
255
Historically, the overall population in the country has been suggested to be between
256
600 and 1,000 (Gippoliti & Dell’Omo, 2003) and more recently estimated as 1,908
257
individuals (95% confidence interval: 923–6,121 individuals Heinicke e t al., 2019a).
258
Improved representative surveys have been recommended for GB given the large
259
confidence intervals of estimates (Heinicke et al., 2019a). The size of local populations has
260
been evaluated for most of the protected areas where chimpanzees occur using various
261
indirect methods (Table 1). Although the estimates from the different studies cannot be
262
compared directly, Cufada Lagoons Natural Park (CLNP) is highlighted as the population with
263
the lowest density, whereas the Boé region stands out as the one displaying the topmost
264
density (reaching > 6 individuals per km
2
, Table 1). Currently, there is no size or density
265
assessment for DNP or for other populations outside areas with formal protection, such as
266
ecological corridors (but see the exception of Gandamael, Table 1, Fig. 1b), although it is
267
estimated that approximately 35% of chimpanzees live outside a park in GB (Heinicke et al.,
268
2019a). Genetic tools have not yet been applied to inform the conservation of the western
269
chimpanzees in GB. Little is known about the genetic diversity or the amount of genetic
270
isolation between populations (but see Borges, 2017 and Gerini, 2018) and no estimates of
271
N
e
have been obtained to date.
272
In this study, we use geographically broad genetic data and genomic data from
273
multiple w ild -born individuals to estimate the
N
e
of the western chimpanzee population in
274
GB. We aim to i) estimate
N
e
and infer the recent and ancient demographic history of
275
16
populations using different methods as applied to microsatellite loci and whole-genome
276
sequence data (WGS), ii) integrate the different estimates to improve our understanding of
277
both the evolutionary history and current demography of the western chimpanzees, iii)
278
discuss the strengths and limitations of each
N
e
estimator method and their
279
complementarity in informing conservation decisions for long-lived organisms, and iv)
280
discuss the implications of the results for the conservation management of this emblematic
281
species in GB.
282
2. Materials and methods
283
2.1 Study area
284
The study area covers a large proportion of the chimpanzee range in GB (Gippoliti &
285
Dell'Omo 2003) (Fig. 1b), encompassing an area of approximately 6,000 km
2
. Sampling of
286
biological material was carried out in four geographically-distinct and formally protected
287
areas – 1. Cantanhez National Park (CNP, 1,067.67 km
2
), 2. Cufada Lagoons Natural Park
288
(CLNP, 890 km
2
), 3. Dulombi Natural Park (DNP, 1,600.96 km²) and 4. Boé National Park
289
(BNP, 1,552.95 km
2
) (https://ibapgbissau.org/areas-protegidas/). Chimpanzees were known
290
to be present at these areas prior to our study (Gippoliti & Dell'Omo 2003; Bersacola et al.,
291
2018).
292
2.2 Microsatellite loci dataset
293
We generated a dataset of 143 unique genotypes for 10 microsatellite loci derived
294
from non-invasive fecal samples (Borges, 2017, Gerini, 2018) (Fig. 1b). Eighty-five genotypes
295
correspond to samples collected between 2015 and 2017 in CLNP (N=38), BNP (N=34), and
296
DNP (N=13) and the remaining consisted of previously determined genotypes from CNP
297
(N=58) (Sá, 2013) (Fig. 1b). Fecal samples were collected fresh and from unhabituated and
298
unidentified individuals, in sites used by chimpanzee groups for sleeping, foraging and
299
drinking. The techniques and methods to preserve the fecal samples until DNA extraction
300
are described in Ferreira da Silva et al . (2014). DNA extraction was carried out using two
301
methods: i) the QIAamp®DNA Stool Mini Kit (QIAGEN®) at MWB research group laboratory
302
facilities at School of Biosciences, Cardiff University, UK (Sá, 2013) and ii) the CTAB method
303
(Vallet et al., 2008, adapted by Quéméré et al. 2010) for samples collected between 2015-
304
2017, which were extracted at
Instituto Gulbenkian de Ciência
(IGC, Oeiras, Portugal)
305
17
laboratory facilities. The procedures to avoid contamination by exogenous DNA are
306
described elsewhere (Ferreira da Silva et al., 2014). DNA samples were identified to the
307
species level using a mitochondrial DNA hypervariable region I fragment (approximately 600
308
base pairs, using primers L15926 and H16555, as described in Sá, 2013). Consensus
309
sequences were derived from forward and reverse sequencing by visual comparison using
310
Geneious Pro v.4.8.5 (Biomatters, Biomatters Ltd, New Zealand). Standard Nucleotide BLAST
311
in NCBI (http://www.ncbi.nlm.nih.gov/) was used to identify accessions closely related to
312
the generated sequences and confirm that samples were from
P. troglodytes verus
(
i.e
.,
313
GenBank Accession code D38113). Allele size standardization between datasets was carried
314
out using re-extraction and re-analyses of DNA extracts of five samples included in Sá (2013)
315
together with the novel samples analyzed in Borges (2017) and Gerini (2018). Allele scoring
316
followed previously described procedures to guarantee minimal impact of allelic dropout
317
and false alleles errors: four replicates were carried out per sample and the rules to reach a
318
consensus genotype were determined per locus (Ferreira da Silva et al., 2014). The
319
consensus genotype was classified according to the Quality Index (QI, M iquel et al., 2006),
320
and genotypes with a mean across loci below 0.55 were excluded from the dataset. The
321
probability of identity (PI) and the probability of identity between s iblings (PIsibs) (Waits et
322
al., 2001), estimated using GenAIEx v.6.503 (Peakall & Smouse,
2006), was of 1.5 x 10
-11
and
323
8.9 x 10
-05
, respectively, which in principle all ows to dis tinguish between unique genotypes
324
using six loci. We could not find genotyping errors (typing errors, large-allele dropout, and
325
locus-specific deficiency in heterozygotes due to null alleles) using MicroChecker v.2.2.3
326
(van Oosterhout et al., 2006) apart from locus D2S1326 for CLNP, which showed excess of
327
homozygotes. We retained the locus in the final dataset as we found no significant
328
departures from Hardy-Weinberg equilibrium per locus using the Bonferroni correction
329
when geographic populations were analyzed separately. The population of chimpanzees in
330
GB does not display significant population structure when assessed using individual-based
331
Bayesian algorithms (
e.g.,
STRUCTURE) (Borges, 2017 estimated K=1).
332
18
2.3 Genomic data
333
2.3.1 Sampling
334
Whole-genome sequences were produced from biological material collected from
335
wild born chimpanzees: one road-killed (tissue sample T3-Chimp collected in 2011) and four
336
individuals (blood samples from Bo, Bella, Simão and Emilia chimpanzees, collected between
337
2018 and 2019) confiscated by the Institute for Biodiversity and Protected Areas (IBAP) from
338
private premises. We obtained the information that the individuals w ere caught by hunters
339
in different sites within GB (
e.g
., Bo was originally from CNP, Bella was confiscated in Quebo
340
but probably originated from CNP, Emilia was from DNP, and Simão was living in Bafatá but
341
traded in Quebo, Fig. 1). Blood was collected as part of the placement of the individuals in a
342
sanctuary abroad (Sweetwaters Chimpanzee Sanctuary, Ol Pejeta, Kenya, Ferreira da Silva &
343
Regalla, in review
3
). The blood samples were drawn by a wildlife veterinarian (P. Melo,
344
vet_natura, https://www.vetnatura.pt/) for health screening and as part of a parasites and
345
virus detection procedure prior to translocation (Melo et al., 2018). Samples were collected
346
in 5 mL collection tubes filled up with the anticoagulant ethylenediamine tetraacetic acid
347
(EDTA) and preserved fresh until DNA extraction. The road-killed indiv idual was found in the
348
road next to CLN P (Fig. 1b) and a sample of muscle tiss ue w as collected and preserved in
349
98% ethanol up to DNA extraction.
350
2.3.2 DNA extraction and data production
351
DNA was extracted from the five samples adapting the method by Vallet et al.,
352
(2008). We used 500 µL of each blood sample and about 10 mg of tissue from the road-
353
killed individual. The details of this two-day DNA extraction protocol can be found in
354
Supplementary material S1. We tested the quality of DNA extractions in 2% agarose gels and
355
quantified DNA concentration using a Nanodrop microvolume spectrophotometer
356
(ThermoFisher Scientific) (Supplementary material S2). Laboratory procedures took place at
357
the IGC, and extractions were carried out in a biologi cal safety cabinet in a Biosafety Level 2
358
dedicated room. Library preparation and sequencing were performed by Macrogen at a
359
coverage of 30-15x using the Illumina Hiseq X and TruSeq platforms.
360
3
See documentary about rescuing chimpanz ees to Ol Pejeta sanctuary, Kenya,
https://www.youtube.com/w atch?v=GxXMk2UPvUM.
19
2.3.3 Whole-genome sequence (WGS) data assembly, Mapping and Genotype Calling
361
After all samples passed quality control tests, we used the BAM pipeline from
362
PALEOMIX to process the sequences for downstream analysis at the Globe Institute’s
363
(University of Copenhagen, Demark) High-Performance Computing (H PC) cluster. This
364
pipeline trims adapter sequences, filters low quality reads, removes PCR duplicates, and
365
aligns reads and maps them to a reference genome (Schubert et al., 2014). We used a
366
“makefile” (.yaml file), that allows the specification of the tasks to be performed, BWA as
367
the aligner software and the algori thm “mem”. The BWA-mem a lgorit hm shows great
368
performance with sequencing errors and is most adequate for short reads, as it is the case
369
of this study (Li, 2013). The “MinQuality” parameter was used to exclude reads with a
370
mapping quality (or Phred score) below zero (Schubert et al., 2014).
371
For the genotype calling, we first selected variants with a minimum Phred quality
372
score of 20, using the HaplotypeCaller algorithm in GATK (version 4.2.0.0; Poplin et al.,
373
2017). HaplotypeCaller uses the input data to calculate the likelihood of each genotype per
374
sample, and then assigns the most likely genotype to that sample. Through the application
375
of the ‘SelectVariants’, only sites with SNPs were selected. Lastly,
vcftools
(Danecek et al.,
376
2011) was used to remove indels, namely the sites with a missing proportion higher than
377
0.9, a Phred quality score equal to or lower than 30, and genotypes with depth values lower
378
than 5 and higher than 100. A table describing the WGS data summary statistics for each
379
sample, such as coverage, observed number of homozygous genotypes, expected number of
380
homozygous genotypes and inbreeding coefficient (F), can be found in Supplementary
381
material S3.
382
As a quick assessment of the possible presence of genetic structure among the
383
sampled individuals, we performed an analysis with the STRUCTURE software version 2.3.4
384
(Pritchard et al., 2000) using the admixture model and assuming correlated allele
385
frequencies. The parameter set consisted of a burnin period of 50,000 steps, followed by
386
200,000 iterations, and 10 runs for each number of clusters (Van Wyngaarden et al., 2017).
387
The results indicated evidence for a single panmictic cluster (K=1) as the best clustering
388
solution to explain the observed genetic variation across individuals,
i.e.
absence of genetic
389
structure (Silva, 2024).
390
20
2.4 Effective population size estimation and demographic history
391
2.4.1 The PSMC (pairwise sequentially Markovian coalescent) and the IICR: principles of
392
demographic inference
393
The PSMC method of Li and Durbin (2011) was applied to the nuclear genomes of
394
the five individuals for which tissue and blood samples were obtained. The PSMC uses the
395
information from the distribution of heterozygous sites along the genome of a single diploid
396
individual (or two haploid genomes) and produces a curve where the x-axis represents time
397
usually represented in a log-scale, and the y-axis is often interpreted as representing the
398
effective population size. Proper scaling of the PSMC in years requires the use of estimates
399
of generation time, mutation, and recombination rates.
400
PSMC version 0.6.5-r67 (Li & Durbin, 2011) (available at
401
http://github.com/lh3/psmc) was run on each individual genome using the following
402
settings: -N25 -t15 -r5 -p "4+25*2+4+6". Individual consensus sequences were generated
403
using the
mpileup
,
bcftools
and
vcfutils.pl
(
vcf2fq
) pipeline from SAMTOOLS v. 1.16, with
404
minimum read depth (-d) set to five and maximum read depth (-D) set to 30. The consensus
405
sequence was converted into a fasta-like format using the
fq2psmcfa
program, provided in
406
the PSMC package, with the quality cut off (-
q
) set to 20. We assumed a mutation rate (
μ
) of
407
1.2 × 10
-8
per base pair per generation and a generation time of 25 years (Venn et al., 2014;
408
Besenbacher et al., 2019; Chintalapati & Moorjani, 2020). To quantify the variance in PSMC
409
curves, we performed 10 bootstraps per individual, following the re-sampling protocol
410
suggested by the authors. The inferred demographic histories for the five analyzed
411
individuals were plotted in a single figure using Ghostscript 9.16 and Gnuplot 5.4.0. The
412
PSMC plots are usually interpreted in terms of
N
e
changes but can also be interpreted in
413
terms of connectivity changes (see Discussion).
414
2.4.2 MSVAR analysis of microsatellite loci data
415
We also used the Bayesian likelihood-based approach of Storz & Beaumont (2002),
416
as implemented in the MSVAR 1.3 software. This approach assumes a simple model of
417
exponential population size change (allowing for either growth or decline) from an ancient
418
population of size
N
1
to a present-day population of size
N
0
. In practice, the method uses a
419
Monte Carlo Markov chain algorithm to estimate the posterior probability distribution of
N
0
420
21
and
N
1
and of the time at which the population started to increase or decrease (
T
, in years,
421
assuming that a generation time is given), and the
per
locus mutation rate (μ). We
422
conducted four independent runs with different initial values and varying sets of priors and
423
hyperpriors to reflect assumptions of either constant population size (
N
0
=
N
1
), population
424
decline (
N
0
<
N
1
), or population growth (
N
0
>
N
1
) and therefore control for the impact of the
425
priors on the posterior distributions (Supplementary material, Table S4).
426
Analyses were run using a series of datasets, to discard the possibility that the
427
presence of related individuals, the sampling scheme and genetic structure could impact the
428
inferred demographic histories, and assuming a generational span of 25 years (following
429
Langergraber et al., 2012). We aimed to recover the demographic history of western
430
chimpanzees by analyzing samples from i) GB as one population (N = 143 genotypes), ii) by
431
park (CNP N = 58, CLNP N = 38, DNP N = 13 and BNP N = 34); iii) a dataset formed by
432
unrelated individuals (N = 121, see below how relatedness was estimated); and iv) five
433
“random datasets” obtained by randomly selecting 58 samples, which correspond to the
434
largest dataset from a single area (
i.e.,
CNP). Datasets iii and iv were used to test the
435
influence of the presence of highly related individuals and differences in sample sizes across
436
datasets in demographic estimates, respectively. Analyses were also run for each geographic
437
population since coalescent theory predicts that when populations are structured, samples
438
obtained from one population will tend to exhibit signals of bottlenecks, whereas samples
439
obtained across many demes will tend to have a much weaker bottleneck signal (Beaumont,
440
2004, Wakeley, 1999), as shown on simulated data (Chikhi et al., 2010).
441
To estimate relatedness between individuals, we calculated the c orrelation
442
coefficient between the observed and simulated values of relatedness (100 pairs) for the
443
Milligan (2003) and Wang (2007) likelihood estimators using the
related
R package (Pew et
444
al., 2015). We also estimated relatedness between pairs of individuals in the overall dataset
445
and per protected area (CNP, CLNP, DNP, and BNP) using the ML-Relate likelihood method
446
(Kalinowski et al., 2006). We performed 100,000 simulations to identify dyads with a likely
447
relationship of Parent-Offspring or Full-Siblings and r > 0.5 significantly different (p < 0.05)
448
from dyads with a likely relat ionship of Half-Siblings and Unrelated. For dyads identified
449
22
using the full dataset and the protected area dataset, one genotype of the dyad, the one
450
which displayed lower QI, was removed from the dataset.
451
The individuals present in the five “random databases” were selected from the
452
dataset i) using
runif
function in R.
453
Each run in MSVAR included 300,000 thinning update steps and 30,000 thinning
454
intervals, totaling 9 x 10
9
steps. We discarded the first 10% of each simulation to eliminate
455
the influence of initial conditions on parameter estimation (burn-in). We verified
456
convergence between runs both visually and using Brooks, Gelman, and Rubin Convergence
457
Diagnostic test (Gelman & Rubin, 1992; Brooks & Gelman, 1998) conducted in R version
458
2.11.1 (R Development Core Team 2010) using the package BOA version 1.1.7 (Smith, 2007).
459
2.4.3 Linkage disequilibrium-based estimation of current
N
e
and recent demographic history
460
from genomic data
461
We used GONE (Genetic Optimization for
N
e
Estimation)
462
(https://github.com/esrud/GONE) that implements a genetic algorithm to infer the recent
463
demographic history of a population from single nucleotide polymorphism (SNP) data of one
464
contemporary sample (Santiago et al., 2020). The method can infer the demographic history
465
of a population within the past few hundreds of generations, with the authors stressing that
466
the greatest reliability and resolution is for the last 100 generations (Santiago et al., 2020). It
467
uses the observed spectrum of linkage disequilibrium (LD) between pairs of loci over a wide
468
range of genetic distances (implicit recombination rates) and has been validated by
469
simulation under different demographic scenarios and for small sample sizes (
i.e.,
n=10,
470
Santiago et al., 2020). These simulations suggest that even when GONE is not able to
471
accurately infer the recent
N
e
trajectory, it can estimate the current
N
e
relatively well. The
472
simulations also suggested that not considering LD data between more distant loci (
e.g.
,
473
with scaled recombination rates > 0.05) in the analyses allows for better es timates of
N
e
,
474
particularly when sample sizes are small or when there is population structure and
475
migration rates between subpopulations are low (Santiago et al., 2020). Thus, the authors of
476
the method recommend using a maximum recombination rate of 0.05, this being the default
477
value of this parameter in GONE. Accordingly, we used this value and default settings for all
478
other software parameters, such as no minimum allele frequency cutoff and 40
479
23
independent replicate runs, with the
N
e
point estimate for each generation being the
480
geometric mean of the values of the replicates. Analyses were performed using 57,650
481
genome-wide autosomal SNPs with genetic map information. As suggested in Santiago et al.
482
(2020), we obtained an empirical 95% confidence interval by running GONE on 20 random
483
replicates of 57.5K SNPs sampled from the whole-genome sequences. Replicates were
484
generated by variant thinning in PLINK 1.9 (www.cog-genomics.org/plink/1.9/) (Chang et al.,
485
2015).
486
2.4.4 Linkage disequilibrium estimation of contemporary
N
e
from microsatellite data
487
We used NeEstimator 2.1 (Do et al., 2014) to estimate contemporary
N
e
from
488
microsatellite loci data using the bias-corrected version of the method based on LD (Hill,
489
1981; Waples, 2006; Waples & Do, 2010), and assuming random mating. The method is
490
robust to equilibrium migration rates up to 10% at lower population sizes (Waples &
491
England, 2011; Gilbert & Whitlock, 2015). We performed analyses both for the whole
492
dataset and separately for each of the four protected areas. The software estimates
493
confidence intervals (CIs), both parametric and based on jackknifing over individuals (Jones
494
et al., 2016), which accounts for the fact that overlapping pairs of loci are being compared
495
and implements a method to correct for possible biases due to missing data (Peel et al.,
496
2013). In any case, for each of the five datasets, we also performed analyses removing loci
497
with more than 10% missing data. When analyzing each dataset, and depending on its
498
sample size, we used a minimum allele frequency (MAF), the Pcrit value, following the
499
recommendations in Waples & Do (2010).
500
3. Results
501
3.1 Effective population size estimation and demographic history
502
3.1.1 The PSMC (pairwise sequentially Markovian coalescent) and the IICR
503
The PSMC curves exhibited a series of increases and decreases with the last decrease
504
starting around 200 kya, and the previous one starting at around one million years ago
505
(Figure 2a). These curves can be interpreted in terms of changes in
N
e
, assuming panmixia
506
and no population structure, in terms of changes in connectivity under population structure
507
(Steux et al., 2024) and constant size or as a combination of both types of changes.
508
Altogether they suggest either very large populations in the past that have been significantly
509
24
reducing for the last 200 ky or are the result of a metapopulation characterized by changes
510
in connectivity with no obvious population decrease during that period. The results for the
511
10 bootstrap replicates for each indi vi dual w ere highly consistent (Supple mentary materi al,
512
Figure S5). One individual (sample T3_chimp) exhibited a PSMC curve that followed the
513
shape of the other curves but was flatter a nd shifted downward and towards more recent
514
times, a pattern that has been described for individuals having a lower coverage
515
(Nadachowska et al., 2016, see Discussion).
516
3.1.2 MSVAR analysis of microsatellite data
517
The analyses we conducted with the control datasets (unrelated and “random
518
datasets”) did not provide results significantly different from the demographic scenarios
519
obtained for the complete datasets of GB and the parks, hence suggesting that relatedness
520
and sample size did not significantly affect the results.
521
From the whole dataset (N = 143 genotypes), MSVAR estimated a contemporary
N
e
522
(
N
o
) between 4,500 and 6,500 individuals, that resulted from a mild bottleneck starting
523
about 44,500-70,000 years ago from a more ancient
N
e
around 10,000-12,000 (
N
1
) (Fig. 2b).
524
To account for the possible effects of genetic structure, we also carried out analyses
525
separately for each park (Fig. 2e to 2h). For CNP, the southernmost chimpanzee population
526
of GB, MSVAR identified a stronger signal with estimates of
N
0
between approximate ly 500
527
and 1,125 individuals, whereas
N
1
estimates were between 10,000 to 12,000 breeding
528
individuals, with limited overlap between the
N
0
and
N
1
median posterior distributions and a
529
posterior distribution of N
0
/N
1
that was consistently below zero (Fig. 2e, Supplementary
530
material, Table S5). MSVAR estimated that this demographic decrease of the CNP
531
chimpanzee population occurred around 5,000 and 12,500 years ago under the assumption
532
of panmixia (Table 2).
533
We found a similar demographic scenario for the CLNP chimpanzee population,
534
which, like the CNP population, is also located in the coastal area and is geographically the
535
closest to CNP (Fig. 2f). The CLNP chimpanzee population was also found to have undergone
536
a one-order-of-magnitude bottleneck, with very similar values of
N
0
and
N
1
to the inferred
537
scenario at CNP (Table 2). The putative difference between the demographic histories of the
538
two coastal populations is the fact that the demographic decrease may have occurred later
539
25
in CLNP than in CNP (CLNP: 3,612 – 7,874 years ago), but the inferred posterior distributions
540
of T for both parks still overlap considerably (Table 2, Fig. 2e and Fig. 2f). The bottleneck
541
signal of both populations is also confirmed by the
N
0
/N
1
, which in both cases is below 0,
542
across the four simulated scenarios (Fig. 2e and Fig. 2f).
543
As for the chimpanzee populations from the eastern region, at DNP and BNP, we
544
found different demographic histories (Fig. 2g and Fig. 2h). Not only are the current
N
e
much
545
larger but there is also no clear signal of demographic change for any of these regions. The
546
population from DNP is estimated to have a
N
0
of between 2,769 and 7,401 individuals, with
547
a slightly larger estimated
N
1
(Table 2). However, the va riation in
N
0
estimates across the
548
four different simulated scenarios does not allow for a clear signal of population size
549
change. This absence of a clear bottleneck signal is also supported by the variation of the
550
N
0
/
N
1
estimates and its overlap with zero (meaning no population size change), which is also
551
translated into weak convergence of the
T
posterior distributions (Fig. 2g). For the
552
easternmost chimpanzee population sampled in BNP, the inferred posterior distributions of
553
N
0
and
N
1
are very consistent, which indicates a large and historically stable population in
554
this park (Fig. 2h). The estimated values for the
N
0
indicate a population with a current
555
effective population size above 6,500 individuals (Table 2). The estimated posterior
556
distributions of
N
1
extensively overlap with the
N
0
estimates, as it is also clear from Fig. 2h
557
suggesting no population size changes. The scenario of a large and historically stable
558
population is further supported by the
N
0
/
N
1
estimates consistently overlapping with zero
559
(Fig 2h).
560
3.1.3 Linkage disequilibrium-based estimation of current Ne and recent demographic history
561
from genomic data
562
GONE estimated a contemporary
N
e
around 300-350 (point estimate of 344 for the
563
previous generation, 95% CI = [116, 580]). Regarding demographic history over the last 100
564
generations, analyses in GONE suggest that after an increase in
N
e
over 50 generations, the
565
effective population size gradually declined over the last 50 generations up to present
566
times, following an almost symmetrical pattern (Fig. 2c). The
N
e
decrease was of almost one
567
order of magnitude.
568
26
3.1.4 Linkage disequilibrium estimation of contemporary Ne from microsatellite data
569
Estimates of contemporary
N
e
and respective 95% confidence intervals, obtained
570
with NeEstimator from the microsatellite data, both for the global dataset and separately
571
for each of the four parks, are presented in Table 3. The
N
e
point estimates for the global
572
dataset were around 150-190 individuals, with the 95% CIs spreading around 80-550
573
individuals.
574
Point estimates for CLNP and BNP were of similar magnitude, respectively at about
575
160-230 and 130-150 individuals, but the 95% CIs had infinite upper bounds. This can
576
happen if
N
e
is large and/or if the data has limited information (
e.g.,
insufficient sample size;
577
reduced number or not very polymorphic loci) (Marandel et al., 2019). The infinite upper
578
bound implies that it is not possible to reject the null hypothesis that LD can be explained
579
entirely by sampling error (Waples & Do, 2010). Still, the finite lower bou nd provides useful
580
information about the minimum lim it of
N
e
(Waples & Do, 2010). The
N
e
point estimates for
581
the CNP (54-79 individuals) were substantially lower, but with 95% CIs overlapping with the
582
previous ones. On the other hand, the point estimates for DNP, at 8-12 individuals, were
583
approximately an order of magnitude smaller than for the other datasets and with
584
parametric 95% CIs not overlapping with the other parks; yet, the jackknife CIs had infinite
585
upper bounds. The small sample size from the DNP may contribute to an underestimation of
586
N
e
. This underestimation will tend to be smaller if the true
N
e
is not very large (e.g., ≤ 100)
587
and will be greater the larger the true
N
e
(Waples & Do, 2010).
588
4. Discussion
589
In this work, we used several methods that aim to estimate either a single effective
590
population size or possible changes in
N
e
over different temporal scales, using samples
591
obtained over different spatial scales. The genetic diversity of the critically endangered
592
western chimpanzee in GB was estimated using different types of data (microsatellites and
593
WGS SNPs) and the methods applied were thus different. We estimated the demographic
594
trajectories of w estern chimpanzees’ representative of the whole country and, separately,
595
for four geographic populations inhabiting protected areas in the south of the country,
596
which are considered relevant areas given the global conservation of the subspecies.
597
27
4.1. Estimation of
N
e
over different temporal scales
598
As noted in Luikart et al. (2010) review and in Box 1 in Ryman et al. (2019) one can
599
consider, as a simple approximation, the idea that different
N
e
estimators should be
600
interpreted by using different time frames (Luikart et al., 2010, Wang, 2005). Some
601
estimates would correspond to the “ancient time” (from hundreds to tens of thousands of
602
generations) whereas others would correspond to the recent or contemporary
N
e
(from a
603
few to tens or hundreds of generations). Typically, MSVAR (microsatellite data) and PSMC
604
(single genome data) provide information about the former time period whereas
605
NeEstimator (microsatellite data) and GONE (WGS data) provide estimates that are mainly
606
about the recent past. The latter type of methods assume that the long-term
N
e
can be to
607
some extent neglected regarding some properties of the genetic data such as the LD pattern
608
(at least among some markers) or the variation of allele frequencies in the last few
609
generations. We note that MSVAR also provides an estimate of contemporary
N
e
but as part
610
of a demographic model of size change, and that GONE also integrates the contemporary
N
e
611
in a trajectory of
N
e
change.
612
The contemporary
N
e
estimates are considered the most
relevant to assess
613
extinction risk because they reflect ongoing or recent demographic or reproductive
614
processes whereas the historical
N
e
refers to the genetic or demographic processes over
615
much longer periods (Luikart et al., 2010; Santos-del-Blanco et al., 2022). We argue that it is
616
the combination of these estimates that should inform on the best conservation decisions
617
and measures (Fig. 2).
618
Furthermore, although the concept of
N
e
has been presented as analogous to the
619
census size (
N
c
), decades of research have shown repeatedly that
N
e
and
N
c
are not only
620
distinct but may have nearly opposite trends under some models. We want to stress here
621
that
N
e
is only informative about some property of the genetic data, and different
622
properties may have different temporal dynamics which are themselves possibly different
623
from the
N
c
dynamics (see Chikhi et al., 2018, Vishwakarma et al., 2024, Wakeley, 1999).
N
c
624
may also be very disconnected from any
N
e
estimate because
N
c
informs us about the
625
current individuals living in the environment of interest and are thus directly affecting and
626
affected by major ecological processes, such as predation, competition or density
627
dependence (Waples, 2022). Different
N
e
values will depend on the probability that
628
28
individuals have to contribute genes to the next generation but also on how populations
629
may be connected to each other, as this will influence coalescence times (Beaumont, 2004,
630
Chikhi et al., 2010, Wa keley, 1999). Contemporary
N
e
will be influenced by recent
631
fluctuations in population size, variance in reproductive success among individuals, unequal
632
sex ratio and overlapping generations (Hoban et al., 2020, Waples, 2022). Over longer time
633
periods (“historical
N
e
”), population structure and changes in connectivity may generate
634
very contradictory results. For instance, Wakeley (1999) showed that a structured
635
population where all the demes increase in size and between which, gene flow increases at
636
the same time, may exhibit a signal of decrease in
N
e
in stark contrast with the increase of
637
the actual population size. A similar phenomenon w as described by Mazet et al. (2016) on
638
the PSMC method (see also Parreira et al., this Special issue on the effect of population
639
structure on different
N
e
estimation methods).
640
The PSMC plots that we obtained with the five individuals from GB exhibited a
641
similar trajectory to those obtained by Prado-Martinez et al. (2013) for individuals from the
642
same subspecies (
P. t. verus
) but from other locations. The authors estimated the peak of
643
effective population size at ~150 kyr, which is similar to the time of the highest
N
e
estimated
644
here (around 200 kyr). The difference in these values could be due to the fact that we used
645
the chimpanzee reference genome, whereas Prado-Martinez et al. (2013) used the human
646
genome as a reference. However, these differences are minimal, probably because the
647
divergence between
Homo sapiens
and
P. troglodytes
is on the order of 1% (see Henrique et
648
al., 2024 and references therein for the effect of divergence of the reference genome).
649
Beyond these important technical issues, the PSMC curves must be interpreted with ca re as
650
they could indicate changes in
N
e
through time, changes in connectivity without change in
651
N
e
(Steux et al., 2024) or, more likely, a combination of both changes in
N
e
and connectivity.
652
One particularly striking result was that individual T_3 exhibited a PSMC lower than the
653
other four individuals, and shifted towards the left (more recent times). Several studies have
654
found that this kind of shift can be observed when the coverage is reduced (Nadachowska et
655
al., 2016 on
Ficedula
sp. flycatchers and Henrique et al., 2024 on
Microcebus
sp. mouse
656
lemurs and several other primate species). However, individual T_3 has a higher rather than
657
smaller coverage than the other individuals (T_3 32 X vs. 17 X for Bella, Bo, Emília and
658
Simão, on average). This result is thus puzzling, and it cannot be explained without further
659
29
investigation. One possible explanation is that the difference in coverage between samples
660
may have resulted in variable power to call heterozygous sites, which could have resulted in
661
the slightly distinct demographic histories estimated for the five individuals (Frantz et al.,
662
2013).
663
The PSMC method has been used on many endangered species since it only requires
664
one genome sequence and is thus adapted for endangered species for which large genomic
665
data sets are difficult to obtain. This is the case of many endangered primates, including
666
chimpanzees (Prado-Martinez et al., 2013), gibbons (Hylobatidae
sp
., Carbone et al., 2014),
667
mouse lemurs (Teixeira et al., 2021), and sifakas (Propithecus
sp
., Guevara et al., 2022).
668
Implicitly, what the PSMC recovers is the distribution of coalescence times along the
669
genome. When two non-recombining fragments are identified by the PSMC and differ by
670
many heterozygous sites, this suggests that their MRCA (most recent common ancestor) is
671
likely very old. The opposite is more likely if there are few or no heterozygous sites. The
672
distribution of these implicit coalescence times can be interpreted as the result of
673
population size changes under the assumption of panmixia and in the absence of significant
674
population structure in the species of interest. The PSMC can thus be interpreted as a series
675
of changes in
N
e
, and this is the most common interpretation.
676
However, in the last 25 years there has been an increasing recognition that
677
population structure can generate spurious signatures of population size change
678
(Beaumont, 2004, Chikhi et al., 2010, W akeley, 1999). In the specific case of the PSMC
679
method, Mazet et al. (2016) showed that it is in fact impossible to det ermine whether a
680
particular PSMC plot is the result of real change in
N
e
or of a more complex model of
681
population structure with changes in connectivity, without any c hange in population siz e. In
682
the latter case, it thus becomes impossible to actually make statements regarding changes
683
in
N
e
from changes observed in the PSMC curve alone. Mazet et al. (2016) introduced the
684
concept of IICR (inverse instantaneous coalescence rate) and noted that the PSMC method
685
in fact infers the IICR, not
N
e
. The IICR will be identical to
N
e
under panmictic models without
686
population structure but very different from any
N
e
changes as soon as there is population
687
structure. Altogether, this suggests that for species like chimpanzees that are known to be
688
structured (Funfstuck et al., 2015, Lester et al., 2021), signals of population size changes
689
inferred from methods assuming panmixia (PSMC, MSVAR, Bottleneck, StairwayPlot, GONE,
690
30
etc.) must be interpreted with caution (Steux et al., 2024). However, whether one considers
691
population structure or panmixia, our results suggest that the populations sampled were
692
part of a metapopulation that may have been very large and included all the regions
693
sampled in this paper. We will come back to this later.
694
The analyses using the MSVAR method and microsatellite loci data suggested that
695
chimpanzees from GB have undergone a mild demographic decrease starting around 40,000
696
years ago (when we used the global dataset). However, for the analyses of the different
697
parks, we inferred contrasting histories for inland (eastern, DNP and BNP) and coastal
698
(western, CLNP and CNP) populations with either no changes or recent and minor
N
e
699
changes. Also, the estimates for
N
1
were very similar for the different parks with ~10,000
700
individuals. This may indicate that the populations at the four parks were connected in the
701
past and part of one metapopulation (itself probably connected to other regions outside
702
GB), in agreement with our interpretation of the PSMC curves.
703
By contrast, the estimates of
N
0
were much smaller, indicating either that population
704
connectivity changed towards more recent times, or that these values correspond to the
705
local deme size, as expected under the coalescent theory of structured models. Whichever
706
interpretation one may favor, our results appear to suggest that, whereas the inland eastern
707
populations remained large/connected w ith other regions (
e.g.
, transnational populations,
708
given the geographic proximity with the Republic of Guinea), the western coastal
709
populations appear more isolated, a process that may have started thousands of years ago
710
at the scale of the country. We have to be cautious with these figures as the timing of size
711
change inferred by MSVAR may not correspond to any particular timing of change in gene
712
flow, since models without changes in connectivity would also generate signals of
713
bottleneck. Despite these cautionary remarks, we believe that the CNP and CLNP may have
714
become separated from the metapopulation, but could have remained connected to each
715
other, forming a smaller sub-population not exceeding 1,200 breeding individuals. This
716
result also suggests that chimpanzees may have still been able to disperse between these
717
two parks.
718
In a recent study, Fontsere et al. (2022) analyzed chromosome 21 genome-wide data
719
across the species and subspecies distribution and inferred a large exchange of migrants
720
during the last ~800 years (range 117 to 2,200 years) for the populations located in the
721
31
northern range of the distribution, which includes Senegal, Mali, northern Guinea and GB
722
(using samples from BNP only). Thus, no signs of long-term isolation were detected for GB.
723
In another study, Heinicke et al. (2019a) investigated the existence of subpopulations across
724
the
P. t. verus
distribution based on field survey data and spatial modeling tools. According
725
to these authors, one large subpopulation (> 33,000 individuals, approximately 50% of the
726
total population size) was predicted at the northern range of the subspecies, in areas
727
characterized by savanna-mosaic habitats and extending across the Fouta Djallon highland
728
region and the neighboring areas of Senegal and GB (including the four parks of our study),
729
up to Mali and Sierra Leone. Thus, these studies could explain why we inferred large
N
1
730
estimates of ~10,000 reproductive individuals (compared to
N
0
). These estimates could
731
correspond to the whole GB population, possibly reflecting an historical connection of the
732
GB population to a large metapopulation centered in the Fouta Djallon highland region. In a
733
more recent study, Steux et al. (2024) suggested that patterns of genomic variation as
734
observed in the PSMC curves could be modeled as part of a metapopulation of small demes
735
characterized by periods of changing connectivity. They estimated that, in comparison to
736
other
P. troglodytes
subspecies,
P. t. verus
population was characterized by smaller demes,
737
which could explain the lower nucleotide diversity observed in western chimpanzees. Their
738
study however focused on the rather ancient past (older than ca. 50 kyr) due to the
739
uncertainties on the PSMC curves they analyzed in the recent past. Other genetic studies
740
based on microsatellite loci and a fragment of the mitochondrial DNA D-loop region, carried
741
out in GB, do not indicate a strong population structure (Borges, 2017; Sá, 2013), which
742
suggest that the chimpanzees are able to disperse between parks. The change in
743
connectivity between inland areas (DNP-BNP) and coastal (CNP-CLNP) within GB cannot be
744
directly inferred from our analyses. We must thus be careful in result interpretations.
745
However, ou r analyses do identify a large ancestral population that might have been
746
fragmented as a consequence of environmental changes that occurred around 10,000 years
747
ago, including climate instability in West Africa and the more recent increased human
748
impact resulting from the development of agriculture. The Younger-Dryas Holocene
749
transition marks a very abrupt transition between the African Humid Period (14,500 - 6,000
750
years ago, characterized by the expansion of forests and lakes across the Sahara region),
751
and a sequence of time periods characterized by arid conditions towards the late Holocene
752
32
(Gasse & Van Campo, 1994). This climate instability, with a succession of warm and cooling
753
events in West Africa, impacted the extension of forest cover (deMenocal et al., 2000) and,
754
most likely, the size and connectivity of the populations of forest-dwelling fauna, as is the
755
case of the western chimpanzee.
756
The GONE analyses suggested a surprising growth of
N
e
between 2,500 and 1,250
757
years ago (
i.e.
, in generations 100-50) followed by a short stationary period and a more
758
recent symmetrical decrease. This pattern may be an artifact of the method, due to the
759
small number of individuals analyzed (five chimpanzees). For example, Beichman et al.
760
(2018) illustrated the general difficulties of demographic history methods in inferring
761
demographic events over the past hundred generations using whole-genome data for fewer
762
than 10 individuals. Reid & Pinsky (2022) also emphasized the importance of large sample
763
sizes for different demographic his tory methods regarding power and precision to detect
764
and quantify population declines in the last 100 generations, with GONE being no exception
765
to a sharp deterioration in performance when sample sizes are small. Similar and even
766
larger humps have also been described by Santiago et al. (2020) when they simulated data
767
from a simple structured model (panel F of their Figure 2), but at this stage one should be
768
cautious to interpret these results until larger sample sizes are analyzed.
769
The result that however appears to be consistent with the other methods used in
770
this study, is that a gradual decline of
N
e
took place during the last 1,250 years (
i.e.
, the last
771
50 chimpanzee generations) towards the present, with the whole population reaching a
772
contemporary
N
e
around 300-350 breeding individuals, estimates that are qualitatively in
773
agreement with those of NeEstimator and microsatellite data (
N
e
of 107-549 or 138-294,
774
respectively).
775
4.2 Implications for conservation management of the western chimpanzee in Guinea-
776
Bissau
777
N
e
should be seen as a key parameter in conservation management and genetic
778
biodiversity monitoring because it is supposed to provide information on population
779
viability, on inbreeding depression risk, on population isolation, and on effectiveness of
780
selective processes and adaptation in relation to drift (Charlesworth, 2009, Hoban et al.,
781
2022). In populations with small
N
e
, genetic diversity is lost at a faster rate over time, and
782
the random fluctuations in allele frequency by drift can neutralize the effect of natural
783
33
selection and increase the probability of fixation of deleterious mutations, potentially
784
leading the population to inbreeding depression (Charlesworth 2009, Hoban et al., 2022,
785
Newman & Pilson, 1997).
786
Franklin (1980) proposed the thresholds of 50/500 for a minimum effective size
787
required for a viable population in the short and long term, respectively. This
788
recommendation became an established rule of thumb in conservation biology and has
789
been proposed as a genetic indicator to assess progress towards global cons ervation targets
790
(Frankham et al., 2014, Hoban et al., 2020). The 50 short-term rule refers to an effective size
791
quantifying the rate of inbreeding (
N
eI
). The minimum
N
eI
of 50 individuals is thought to be
792
enough to prevent the rapid inbreeding of the population (
i.e.,
1% per generation), which
793
could lead to excessive homozygosity for deleterious recessive alleles and reduced fitness by
794
inbreeding depression. The 500 long-term rule refers to the effective size related to the loss
795
of additive genetic variation (
N
eAV
). This threshold defines the
N
e
above which a population
796
should retain enough evolutionary potential to adapt to new selective forces (
i.e.,
future
797
environmental conditions, Jamieson & Allendorf, 2012). More recently, these numbers have
798
been doubled, with 100 individuals being presented as more adequate to prevent
799
inbreeding depression over five generations for wild populations (
i.e.
, limiting to 10% the
800
loss in total fitness) and 1,000 individuals as necessary to protect evolutionary potential in
801
the long-term (Frankham et al., 2014), particularly when the species´ reproductive rates are
802
low (Perez-Pereira et al., 2022). When a population is detected to have a small or declining
803
N
e
, managers and conservationists should be called to investigate the most likely causes and
804
to reverse the demographic trajectory (Wang et al., 2016). This is why estimating
N
e
is
805
increasingly recognized as central to conservation programs.
806
In this work, we used different approaches that estimated historical and
807
contemporary
N
e
. While the estimates by the different methods could differ, results were
808
consistent in suggesting that current
N
e
is below 500 breeding individuals for the GB
809
chimpanzee’s population. For instance, GONE and NeEstimator suggested values that were
810
typically between 100 and 500, even if some estimates could have larger upper bounds,
811
most likely due to small sample sizes (see Results section). These results were qualitatively
812
in agreement with the
N
o
estimates obtained with MSVAR even if the latter were higher
813
(between 500 and 1,000). While no
N
e
estimate should be taken at face value, the fact that
814
34
these estimates were qualitatively similar when using different methods making different
815
assumptions and using different types of genetic markers, suggests that the contemporary
816
demographic dynamics of chimpanzees in GB is driven by small and isolated populations
817
that derive from, what used to be, a very large metapopulation. Furthermore, these results
818
confirm the selection of the coastal areas of GB and Republic of Guinea by Kormos & Boesch
819
(2003) as priority regions for conservation of the subspecies and are aligned with estimates
820
of density that points to a small population in CLNP (Carvalho et al., 2013), highlighting the
821
critical conservation situation of these populations.
822
Small populations (
i.e
., < 500 breeding individuals) may go e xtinct through a
823
phenomenon referred as extinction vortex (Gilpin & Soule, 1986), in which genetic and
824
demographic issues interact synergistically to decrease genetic diversity and cause
825
population growth rate to drop due to the reduction of the mean fitness. This results in
826
further decreases in genetic diversity and promotes the subsequent processes to happen in
827
a cascade, until the extinction of the population. In the case of the western chimpanzees in
828
GB, the main conservation threats have been identified and characterized to some extent.
829
Natural habitats have generally been converted into subsistence crop plantations, such as
830
rice (
Oryza
spp.) or cassava (
Manihot esculenta
), and cashew (
Anacardium occidentale
)
831
monoculture agroforests at least since the last two decades (Hockings & Sousa, 2013,
832
Temudo & Abrantes, 2014). Additionally, the construction of roads and infrastructures
833
increased the accessibility to remote areas and promoted more encounters with humans,
834
which may have increased chimpanzees’ mortality. Although it was found that chimpanzees
835
can cross and use human-altered habitats to some degree, namely sharing the use of
836
forested and village areas with local communities (
e.g
., in CNP, Bersacola et al., 2021),
837
extensive habitat loss and conversion into crop fields and villages are expected to reduce
838
connectivity between populations and diminish population size rapidly (Torres et al., 2010).
839
In CNP for instance, it was estimated a loss of 11% of suitable habitat and the death of
840
between 157 and 1,103 individuals in the population for the period between 1986 and 2003
841
(Torres et al., 2010), which corresponds to less than one chimpanzee generation. Moreover,
842
as reported by Stiles (2023) and as illustrated here (with the four blood samples of
843
confiscated individuals), hunting for live individuals to supply the national and international
844
illegal pet trade occurs in the country (Ferreira da Silva & Regalla, in review, Stiles 2023).
845
35
Quantitative data on the number of traded chimpanzees originating from the GB in
846
international trade routes is missing (Clough & Channing, 2018; Stiles et al., 2013; Stiles,
847
2023). However, given the ease of detecting chimpanzees in private houses and hotels (
e.g.
,
848
18 individuals between 2006 and 2022 found by chance, in Ferreira da Silva & Regalla in
849
review) and considering that five to 10 adults can be killed to harvest only one infant
850
chimpanzee (Teleki, 1980), it can be suggested that hunting to supply the trade o f live
851
individuals may have contributed to a reduction of the population size and consequently of
852
the
N
e
. Furthermore, as conservation threats tend to act synergistically at the local and
853
regional scale - habitat fragmentation increases accessibility to natural habitats by local
854
communities, which in turn, increase poaching, negative interactions with farmers and
855
diseases transmission (Humle et al., 2016), the negative impacts of human-derived activities
856
on the population effective size may be larger than each threat is individually.
857
Specifically, chimpanzees inhabiting CNP and CLNP - the two populations identified
858
by this study at a high risk of extinction - are currently negatively impacted by habitat loss
859
and fragmentation, and by retaliatory killing by farmers during crop-raiding (Hockings &
860
Sousa, 2013). Chimpanzees in CNP may also be subjected to higher reproductive isolation
861
since the Park is in a peninsula surrounded by two permanent water bodies which are
862
insurmountable by chimpanzees, and suitable habitat for the subspecies was considerably
863
lost in northwestern areas, where the isthmus connects the peninsula to the mainland (
i.e
.,
864
Bantael Sila, Cumbijã and Guiledge villages, Torres et al., 2010). These coastal areas have
865
been considered important to maintain gene flow with GB mainland for another primate
866
species (
e.g.,
Guinea baboons,
Papio papio
, Ferreira da Silva et al., 2014, Ferreira da Silva et
867
al., in press). Other forest-dwelling and hunted primate species at CNP may have gone
868
through a decline in
N
e
of a similar magnitude. Minhós et al. (2016) found a pattern of
N
e
869
decrease for co-distributed populations of colobus monkeys inhabiting CNP (
Piliocolobis
870
badius temminckii
and
Colobus polykomos
) at a similar time period as estimated here for
871
CNP chimpanzees (
i.e.
, ca. 10,000 to 3,000 years). The
N
e
for colobus monkeys was
872
estimated using alike models and genetic markers (
i.e
., MSVAR runs and microsatellite loci).
873
The fact that a similar demographic event was observed for other co-distributed species in
874
CNP despite different socio-ecology features, strengthens our interpretation that CNP
875
chimpanzees experienced low
N
e
in recent times. Moreover, several chimpanzee
876
36
communities across CNP show symptoms of leprosy (caused by
Mycobacterium
leprae
,
877
Hockings et al., 2021), which is likely to have negative consequences for the longevity and
878
reproductive success of individuals, but so far unknown. On the other hand, In the southern
879
region of CLNP, the construction of a large road and a thermoele ctric plant with respective
880
electricity transmission lines, led to the loss of one of the best-preserved forest patches of
881
the park (Catarino, 2019) and increased the accessibility to areas used by chimpanzees for
882
nesting (Carvalho et al., 2013). Cases of live chimpanzee captures have been recorded in
883
both parks, which most likely implies adult mortality (Ferreira da Silva & Regalla in rev iew).
884
Furthermore, CLNP is bordered in the east by the main road connecting the south of the
885
country to the capital city - Bissau and, as demonstrated in this study, wildlife-vehicle
886
collisions do happen. Our results suggest that CLNP and CNP are at high risk of extinction
887
and the impact of human-derived activities potentially threatening the two chimpanzee
888
populations should be investigated.
889
Our study provides the first estimates of
N
e
for DNP. We estimated an historical
N
e
of
890
2,769 - 7,401 breeding individuals using MSVAR but could not detect a strong departure
891
from mutation-drift equilibrium. This result was at odds with the very low contemporary
N
e
892
of 6-30 obtained using NeEstimator. DNP is located at the northern margin of the Corubal
893
River and is currently at one edge of the subspecies distribution (see Fig. 1b). Our estimates
894
of a stable demographic trajectory and a large historical
N
e
suggest that DNP was at s ome
895
point in time connected to other large chimpanzee populations. Connectivity to populations
896
located in the south of the Corubal River (
e.g.
, BNP) should be more reduced at present
897
times because the current width of the water body, which can surpass one kilometer in
898
some sites, may be a significant barrier to primate’s gene-flow (
e.g
., the green monkey,
899
Chlorocebus sabaeus
, Colmonero-Costeira et al., in review, the Guinea baboon,
Papio papio
,
900
Ferreira da Silva et al., in press). Nevertheless, the configuration and discharge of the
901
Corubal River may have been different in the past (
e.g
., the mouth may have been in the
902
southwest part of the country, located by the Rio Grande de Buba channel area, Alves
903
2007), which could have allowed chimpanzees to cross to what is the south margin. The
904
small contemporary
N
e
may be explained by the fact that present environmental conditions
905
do not support a large population of chimpanzees. DNP is located at the edge of the
906
distribution of the subspecies and has low density of villages and other human
907
37
infrastructures. This area is mostly dominated by w oodlands and savannah woodland
908
formations (Catarino et al., 2008), and found to be of low habitat suitability for chimpanzees
909
by modeling exercises (<100, range 0-1,000, in Figure S2.2. Carvalho et al., 2021), which
910
could be either related to environmental conditions or a small sample size (J. Carvalho,
911
personal communication). During field work, chimpanzees were mostly detected (and fecal
912
samples collected) in greater proximity to gallery forests along smaller streams or next to
913
the Corubal River (Fig. 1b), and we observed that the subspecies w as not widely distributed
914
in the park area, as in CNP, for instance (Sá, 2013). Chimpanzees at Fongoli in Senegal,
915
inhabiting a similar open, savanna-woodland environment, do not suffer from nutritional
916
stress but display physiological stress from dehydration and heat, which does not seem to
917
be behaviorally compensated (from sitting for longer periods in the shade or using pools or
918
caves for instance, Wessling et al., 2018). Such adverse environmental conditions may be
919
determinant for constraining the distribution at the biogeographical range limits of the
920
subspecies (Wessling et al., 2018) and similarly, limiting the size of the population at DNP.
921
By contrast, our estimates of historical
N
e
of the population of chimpanzees
922
inhabiting the BNP (MSVAR
N
0
: 6,716 - 24,642 breeding individuals) are large and confirm
923
the classification of the area as stable or of high-density (Heinicke et al., 2019b). Boé
924
population has been included in a previous population genomic study, using samples
925
collected across the subspecies range (Fontesere et al., 2022), which estimated high and
926
recent connectivity (for the last
~780
years, range 117-2,200 years) between communities at
927
the northern range (localities in the Republic of G uinea and south of Senegal, Fig. 3 in
928
Fontesere et al., 2022). Boé was found to be genetically closer to samples collected in
929
southern Senegal (Fontesere et al., 2022), possibly due to long-term connectivity between
930
the two neighboring populations. Present-day high density of chimpanzees in the Boé region
931
has been justified by i) remoteness of the area and difficult access, ii) rare hunting of
932
chimpanzee to comply with religious taboos, iii) high habitat suitability for chimpanzees, and
933
iv) slow habitat loss and conversion, and in a large area, habitats are undisturbed (Bincz ik et
934
al., 2019; Carvalho et al., 2021, van Laar 2010). Although DNP and BNP are closely located
935
and share similar environmental conditions, within the Boé region there is a wide network
936
of rivers and waterbodies surrounded by relatively well-preserved gallery forests, which are
937
used by chimpanzees to nest and feed (Binczik et al., 2019). Our results suggest that Boé is a
938
38
stronghold for the chimpanzee population in GB. The effective protection and restoration of
939
the natural habitats in ecological corridors connecting BNP and the remaining parks located
940
south of the Corubal River (Fig. 1b) could be beneficial to promote dispersal, potentially
941
increasing gene flow and improving the probability for long-term persistence of
942
chimpanzees in coastal areas of GB.
943
4.3 Implications for the estimation of
N
e
of wild populations of primates
944
The most common approaches to estimate the effective size of real populations are
945
based on its genetic properties (Luikart et al., 2010). However, obtaining genetic
946
information of threatened species can be challenging. The main issue is related to attaining
947
high quality DNA, which is usually extracted from fresh blood and tissue samples. Species of
948
conservation concern are frequently found in low densities in the wild and commonly live in
949
inaccessible areas and in habitats of low visibility. Hence, it is difficult to trap or handle
950
individuals (Beja-Pereira et al., 2009). Moreover, invasive methods are considered unethical
951
as the contact with humans to retrieve blood or tissue samples increases the risk for disease
952
transmission (Beja-Pereira et al., 2009). Thus, invasive sampling of wild-born individuals for
953
threatened speci es is typically opportunistic and carried out for a few individuals, for
954
instance during veterinarian interventions or post-mortem (
e.g
., Prado-Martinez et al.,
955
2013). Such procedures can take several years to complete (Xue et al., 2015), and sampling
956
is usually geographically restricted, and not representative, of the species distribution and
957
variability. In primates, genomic data used to infer parameters of interest for conservation,
958
has been obtained from high quality DNA collected from individuals in zoos or rescuing
959
centers (
e.g.
, Rogers et al., 2019 but see Fontsere et al., 2022 who obtained genomic
960
diversity estimates of wild chimpanzee populations from 828 non-invasively fecal samples).
961
Yet, specific environmental conditions or breeding practices that inadvertently reduce
962
natural selection pressures or increase inbreeding (Christie et al., 2012) can lead to wrong or
963
limited inferences of demographic parameters. Also, small and spatially restricted sampling
964
can introduce bias in contemporary
N
e
estimates (
e.g
., Santos-del-Blanco et al., 2021) and
965
geographic-broad genetic data, such as the ones obtained using non-invasive fecal samples,
966
are preferable.
967
Here, we show that the estimated values of
N
e
using genomic data and more classic
968
genetic markers, like microsatellite loci obtained using non-invasive fecal samples, are
969
39
largely concordant, although we found that median
N
e
estimates produced by SNP data
970
were higher than estimates generated using microsatellite data. This pattern was also
971
reported by Clarke et al. (2024) meta-analysis. Our study reinforces that datasets generated
972
with traditional genetic markers, such as legacy or baseline microsatellite loci datasets for
973
local populations, are of great value; these can be used to estimate parameters relevant to
974
inform conservation management in species for which obtaining genomic data is not
975
straightforward, or in studies carried out in countries with limited access to sequencing
976
units, funding and trained researchers in genomic data (Bertola et al., 2024). Moreover, our
977
study shows that the combination of different molecular markers and analytical methods
978
can be a useful strategy to overcome the limitations of obtaining high quality DNA from wild
979
threatened populations, to investigate species evolutionary history in time and space, and
980
to integrate genetic information in conservation management decisions at local and regional
981
scales.
982
983
984
40
985
Data Archiving Statement
986
Data for this study are available at Dryad Digital Repository:
to be completed after
987
manuscript is accepted for publication
.
988
989
Literature Cited
990
991
992
Alves, P. H. (2007). A geologia sedimentar da Guiné-Bissau: da análise geral e
993
evolução do conhecimento ao estudo do Cenozóico. PhD thesis in
994
Geology, Lisbon University, Faculty of Sciences, Lisbon, Portugal.
995
Amador, R., Casanova, C., Lee, P. (2014). Ethnicity and perceptions of
996
bushmeat hunting inside Lagoas de Cufada Natural Park (LCNP),
997
Guinea-Bissau.
Journal of Primatology
3(2). doi:
998
http://dx.doi.org/10.4172/2167-6801.1000121
999
Beaumont, M. A. (2004). Recent developments in genetic data analysis: what
1000
can they tell us about human demographic history?
Heredity
, 92(5),
1001
365-379. doi:10.1038/sj.hdy.6800447
1002
Beichman, A. C., Huerta-Sanchez, E., & Lohmueller, K. E. (2018). Using
1003
Genomic Data to Infer Historic Population Dynamics of Nonmodel
1004
Organisms.
Annual Review of Ecology, Evolution, and Systematics, 49
,
1005
433-456. doi:https://doi.org/10.1146/annurev-ecolsys-110617-
1006
062431.
1007
Beja-Pereira, A., Oliveira, R., Alves, P., Schwartz, M. K., Luikart, G. (2009).
1008
Advancing ecological understandings through technological
1009
transformations in noninvasive genetics.
Molecular Ecology Resources
1010
9:1279-1301.
1011
Bertola, L. D., Brüniche-Olsen, A., Kershaw, F., Russo, I.-R. M., MacDonald, A.
1012
J., Sunnucks, P., . . . Segelbacher, G. (2024). A pragmatic approach for
1013
integrating molecular tools into biodiversity conservation.
1014
Conservation Science and Practice
, 6(1), e13053.
1015
doi:https://doi.org/10.1111/csp2.13053
1016
Bersacola, H., Bessa, J., Frazão-Moreira, A., Biro, D., Sousa, C., Hockings. K. J.
1017
(2018). Primate occurrence across a human-impacted landscape in
1018
Guinea-Bissau and neighbouring regions in West Africa: using a
1019
systematic literature review to highlight the next conservation steps.
1020
PeerJ 6:e4847
.
1021
Bersacola, E., Hill, C. M., & Hockings, K. J. (2021). Chimpanzees balance
1022
resources and risk in an anthropogenic landscape of fear.
Scientific
1023
Reports
, 11(1), 4569. doi:10.1038/s41598-021-83852-3
1024
Besenbacher, S., Hvilsom, C., Marques-Bonet, T., Mailund, T., & Schierup, M.
1025
H. (2019). Direct estimation of mutations in great apes reconciles
1026
41
phylogenetic dating.
Nature Ecology & Evolution, 3
(2), 286-292.
1027
doi:10.1038/s41559-018-0778-x
1028
Borges, F. (2017). A country-level genetic survey of the IUCN critically
1029
endangered western chimpanzee (
Pan troglodytes verus
) in Guinea-
1030
Bissau. Master degree, Porto University, Portugal.
1031
Brooks, S. P., Gelman, A. (1998). General Methods for Monitoring
1032
Convergence of Iterative Simulations.
Journal of Computational and
1033
Graphical Statistics, 7
(4), 434-455.
1034
doi:10.1080/10618600.1998.10474787.
1035
Brugiere, D., Badjinca, I., Silva, C., Serra, A., & Barry, M. (2005). Distribution
1036
and Status of Lions and Leopards in Southern Guinea Bissau and
1037
Western Guinea, West Africa.
Cat News
, 43(13-17).
1038
Campbell, G., Houngbedji, M. (2015). Conservation status of the West African
1039
chimpanzee (Pan troglodytes verus) in Togo and Benin. Arlington, VA.
1040
Carbone, L., Alan Harris, R., Gnerre, S., Veeramah, K. R., Lorente-Galdos, B.,
1041
Huddleston, J., . . . Gibbs, R. A. (2014). Gibbon genome and the fast
1042
karyotype evolution of small apes.
Nature
, 513(7517), 195-201.
1043
doi:10.1038/nature13679
1044
Carvalho, J., Marques, T., & Vicente, L. (2013). Population Status of Pan
1045
troglodytes verus in Lagoas de Cufada Natural Park, Guinea-Bissau.
1046
PLos ONE
, 8(8), e71527. doi:10.1371/journal.pone.0071527.
1047
Carvalho, J. S., Graham, B., Bocksberger, G., Maisels, F., Williamson, E. A.,
1048
Wich, S., . . . Kühl, H. S. (2021). Predicting range shifts of African apes
1049
under global change scenarios.
Diversity and Distributions
, 27(9),
1050
1663-1679. doi:https://doi.org/10.1111/ddi.13358
1051
Catarino, L. (2019). Parque Natural das Lagoas de Cufada, Guiné-Bissau.
1052
Desafios, Ameaças e Perspetivas. Revista Internacional em Língua
1053
Portuguesa, 35, 17-30. doi:10.31492/2184-2043.RILP2018.35/pp.17-
1054
30
1055
Catarino, L., Martins, E. S., Basto, P., & Diniz, M. A. (2008). An Annotated
1056
Checklist of the vascular flora of Guinea Bissau (West Africa).
Blumea,
1057
53, 1-222.
1058
Chikhi, L., Sousa, V. C., Luisi, P., Goossens, B., & Beaumont, M. A. (2010). The
1059
confounding effects of population structure, genetic diversity and the
1060
sampling scheme on the detection and quantification of population
1061
size changes.
Genetics,
186(3), 983-995.
1062
doi:10.1534/genetics.110.118661
1063
Clarke, S. H., Lawrence, E. R., Matte, J.-M., Gallagher, B. K., Salisbury, S. J.,
1064
Michaelides, S. N., . . . Fraser, D. J. (2024). Global assessment of
1065
effective population sizes: Consistent taxonomic differences in
1066
meeting the 50/500 rule.
Molecular Ecology
, 33(11), e17353.
1067
doi:https://doi.org/10.1111/mec.17353
1068
Clough, C., & Channing, M. (2018). Illicit Financial Flows and the Illegal T rade
1069
in Great Apes. Retrieved from Washington, DC:
1070
https://www.gfintegrity.org/wp-content/uploads/2018/10/GFI-Illicit-
1071
Financial-Flows-and-the-Illegal-Trade-in-Great-Apes.pdf
1072
42
Chang, C. C., Chow, C. C., Tellier, L. C., Vattikuti, S., Purcell, S. M., & Lee, J. J.
1073
(2015). Second-generation PLINK: rising to the challenge of larger and
1074
richer datasets.
GigaScience, 4
(1). doi:10.1186/s13742-015-0047-8.
1075
Charlesworth, B. (2009). Effective population size and patterns of molecular
1076
evolution and variation.
Nature Reviews Genetics,
10, 195-205.
1077
Chintalapati, M., Moorjani, P. (2020). Evolution of the mutation rate across
1078
primates.
Current Opinion in Genetics & Development, 62
, 58-64.
1079
doi:https://doi.org/10.1016/j.gde.2020.05.028
1080
Christie, M. R., Marine, M. L., French, R. A., Blouin, M. S. (2012). Genetic
1081
adaptation to captivity can occur in a single generation.
Proceedings
1082
of the National Academy of Sciences of the United States of America
1083
109:238-242.
1084
CIA. 2024. Guinea-Bissau factbook. cia.gov.
1085
Colmonero Costeira, I., Djaló, M. L., Fernandes, N., Borges, F., Pais, I. A.,
1086
Gerini, F., . . . Ferreira da Silva, M. (in review). Improving baseline
1087
information on over-looked generalists: occurrence and mitochondrial
1088
DNA diversity of Campbell’s (
Cercopithecus campbelli
) and green
1089
monkeys (
Chlorocebus sabaeus
) in Guinea-Bissau, West Africa.
1090
International Journal of Primatology
.
1091
Conner, J. K., Hartl, D. L. (2004). A Primer of Ecological Genetics. Oxford
1092
University Press, Sunderland, MA, USA.
1093
Crow, J. F., Kimura, M. (1970). An Introduction to Population Genetics
1094
Theory. Harper & Row, New York, USA.
1095
Danecek, P., Auton, A., Abecasis, G., Albers, C. A., Banks, E., DePristo, M. A., . .
1096
. Durbin, R. (2011). The variant call format and VCFtools.
1097
Bioinformatics
, 27(15), 2156-2158.
1098
doi:10.1093/bioinformatics/btr330.
1099
De Menocal, P., Ortiz, J., Guilderson, T., & Sarnthein, M. (2000). Coherent
1100
High- and Low-Latitude Climate Variability During the Holocene Warm
1101
Period.
SCIENCE,
288(5474), 2198-2202.
1102
doi:doi:10.1126/science.288.5474.2198
1103
de Manuel, M., Kuhlwilm, M., Frandsen, P., Sousa, V. C., Desai, T., Prado-
1104
Martinez, J., . . . Marques-Bonet, T. (2016). Chimpanzee genomic
1105
diversity reveals ancient admixture w ith bonob os.
Science
, 354(6311),
1106
477-481. doi:10.1126/science.aag2602
1107
Do, C., Waples, R. S., Peel, D., Macbeth, G. M., Tillett, B. J., & Ovenden, J. R.
1108
(2014). NeEstimator v2: re-implementation of software for the
1109
estimation of contemporary effective population size (
N
e
) from
1110
genetic data.
Molecular Ecological Resources, 14
(1), 209-214.
1111
doi:10.1111/1755-0998.12157.
1112
Hill, W. G. (1981). Estimation of effective population size from data on linkage
1113
disequilibrium.
Genetical Research, 38
(3), 209-216.
1114
doi:10.1017/S0016672300020553
1115
Ferreira da Silva, MJ, Regalla, A. (in review). The illegal trade of live Western
1116
chimpanzees in Guinea-Bissau and possible directions to improve la w
1117
enforcement. submitted as perspective to
Conservation Letters
.
1118
manuscript ID CONL-24-0165.
1119
43
Ferreira da Silva, M., Ca mará, M., Egeter, B., Minhós, T., Bruford, M. W.,
1120
Godinho, R. (2021). Using meta-barcoding tools to monitor primate
1121
meat consumption at dedicated establishments in Guinea-Bissau,
1122
West Africa. Page e65575.
in
Collection: Trends in Biodiversity and
1123
Evolution | TiBE2020 – Metabarcoding and Metagenomics ARPHA
1124
Conference Abstracts, Vairão, Portugal.
1125
Ferreira da Silva, M. J., Bruford, M. W. (2017). Genetics and Primate
1126
Conservation.
in
A. Fuentes, editor. The International Encyclopedia of
1127
Primatology. John Wiley & Sons, Inc.
1128
Ferreira da Silva, M. J., Minhós, T., Sampaio-Dias, S. (2024). News for nature:
1129
Communicating biodiversity importance and conservation challenges
1130
in Guinea-Bissau, West Africa, through local media and online
1131
platforms.
Conservation Science and Practice
6, e13051.
1132
Ferreira da Silva, M. J., Minhós, T., Sá, R., Casanova, C., & Bruford, M. W.
1133
(2021). A qualitative assessment of Guinea-Bissau’s hunting history
1134
and culture and their implications for primate conservation.
African
1135
Primates
, 15, 1-18.
1136
Ferreira da Silva, M. J., Paddock, C.,
Gerini, F., Borges, F., Aleixo
-
Pais, I., Costa,
1137
M., Colmonero
-
Costeira, I. A., Casanova, C., Lecoq, M., Silva, C.,
1138
Bruford, M. W., Varanda, J., Minhós
, T. (2020). Chasing a ghost: notes
1139
on the present distribution and conservation of the sooty mangabey
1140
(
Cercocebus atys
) in Guinea-Bissau, West Africa.
Primates,
61, 357-
1141
363.
1142
Ferreira da Silva, M., Gerini, F., Aleixo-Pais, I., Maie, T., Borges, F., Costa, M., .
1143
. . T., M. (in review). Does the exclusive use of genetic data inform on
1144
the demographic history in Anthropogenic landscapes? Investigating
1145
population size dynamics of Papio papio in Guinea-Bissau. In: W. J.
1146
(Ed.), Genetics, taxonomy, and biogeography of
Papio sp.
Cambridge
1147
press.
1148
Frankham, R. (2005). Genetics and extinction.
Biological Conservation,
126,
1149
131-140.
1150
Frankham, R., Ballou, J., Briscoe, D. (2010). Introduction to Conservation
1151
Genetics. Cambridge University Press, Cambridge, UK.
1152
Frankham, R., Bradshaw, C. J. A., Brook, B. W. (2014). Genetics in
1153
conservation management: Revised recommendations for the 50/500
1154
rules, Red List criteria and population viability analyses.
Biological
1155
Conservation,
170, 56-63.
1156
Franklin, I. R., Soulé, M. E., & Wilcox, B. A. (1980). Conservation biology: an
1157
evolutionary-ecological perspective. Sunderland, MA: Sinauer
1158
Associates.
1159
Frantz, L. A. F., Schraiber, J. G., Madsen, O., Megens, H.-J., Bosse, M., Paudel,
1160
Y., . . . Groenen, M. A. M. (2013). Genome sequencing reveals fine
1161
scale diversification and reticulation history during speciation in Sus.
1162
Genome Biology
, 14(9), R107. doi:10.1186/gb-2013-14-9-r107
1163
Fünfstück, T., Arandjelovic, M., Morgan, D. B., Sanz, C., Reed, P., Olson, S. H., .
1164
. . Vigilant, L. (2015). The sampling scheme matters:
Pan troglodytes
1165
troglodytes
and
P. t. schweinfurthii
are characterized by clinal genetic
1166
44
variation rather than a strong subspecies break.
Am J Phys Anthropol,
1167
156(2), 181-191. doi:10.1002/ajpa.22638
1168
Gasse, F., & Van Campo, E. (1994). Abrupt post-glacial climate events in West
1169
Asia and North Africa monsoon domains.
Earth and Planetary Science
1170
Letters
, 126(4), 435-456. doi:https://doi.org/10.1016/0012-
1171
821X(94)90123-6
1172
Gelman, A., & Rubin, D. B. (1992). Inference from Iterative Simulation Using
1173
Multiple Sequences.
Statistical Science, 7
(4), 457-472.
1174
doi:https://www.jstor.org/stable/2246093.
1175
Gerini, F. (2018). Structure and connectivity of sympatric primate species
1176
across a human- dominated landscape: population genetics of
1177
Western Chimpanzee (
Pan troglodytes verus
) and Guinea Baboon
1178
(
Papio papio
) in Guinea Bissau, West Africa. Tesi di Laurea Magistrale
1179
in Conservazione ed Evoluzione, Università di Pisa, Italy.
1180
Gilbert, K. J., & Whitlock, M. C. (2015). Evaluating methods for estimating
1181
local effective population size with and without migration.
Evolution,
1182
69
(8), 2154-2166. doi:10.1111/evo.12713.
1183
Gilpin, M. E., & Soulé, M. E. (1986). Minimum viable populations: processes
1184
of extinction. In: M.E., Soulé (Ed) Conservation Biology: The Science of
1185
Scarcity and Diversity (pp. 19–34). Sunderland, MA: Sinauer
1186
Associates.
1187
Ginn, L. P., Robison, J., Redmond, I., Nekaris, K. A. I. (2013). Strong evidence
1188
that the West African chimpanzee is extirpated from Burkina Faso.
1189
Oryx,
47, 325-326.
1190
Gippoliti, S., Dell’Omo, G. (1996). Primates of the Cantanhez Forest and the
1191
Cacine Basin, Guinea- Bissau.
Oryx.
30, 74–80.
1192
Gippoliti, S., Dell’Omo, G. (2003). Primates of Guinea-Bissau, West Africa:
1193
Distribution and Conservation Status.
Primate Conservation,
19, 73-
1194
76.
1195
Guevara, E. E., Webster, T. H., Lawler, R. R., Bradley, B. J., Greene, L. K.,
1196
Ranaivonasy, J., . . . Rogers, J. (2021). Comparative genomic analysis of
1197
sifakas (
Propithecus
) reveals selection for folivory and high
1198
heterozygosity despite endangered status.
Science Advances
, 7(17),
1199
eabd2274. doi:10.1126/sciadv.abd2274.
1200
Heinicke, Mundry, S., R., Boesch, C., Amarasekaran, B., Barrie, A., Brncic, T.,
1201
Brugière, D., Campbell, G., Carvalho, J., Danquah, E., Dowd, D., Eshuis,
1202
H., Fleury-Brugière, M.-C., Gamys, J., Ganas, J., Gatti, S., Ginn, L.,
1203
Goedmakers, A., Granier, N., Herbinger, I., Hillers, A., Jones, S., Junker,
1204
J., Kouakou, C. Y., Lapeyre, V., Leinert, V., Maisels, F., Marrocoli, S.,
1205
Molokwu-Odozi, M., N’Goran, P. K., Pacheco, L., Regnaut, S., Sop, T.,
1206
Ton, E., van Schijndel, J., Vergnes, V., Voigt, M., Welsh, A., Wessling, E.
1207
G., Williamson, E. A., Kühl. H. S. (2019a). Advancing conservation
1208
planning for western chimpanzees using IUCN SSC A.P.E.S.—the case
1209
of a taxon-specific database.
Environmental Research Letters,
14,
1210
064001.
1211
Heinicke, S., Mundry, R., Boesch, C., Hockings, K. J., Kormos, R., Ndiaye, P. I., .
1212
. . Kühl, H. S. (2019b). Towards systematic and evidence-bas ed
1213
45
conservation planning for western chimpanzees.
American Journal of
1214
Primatology,
81(9), e23042. doi:10.1002/ajp.23042
1215
Hoban, S., Archer, F. I., Bertola, L. D., Bragg, J. G., Breed, M. F., Bruford, M.
1216
W., . . . Hunter, M. E. (2022). Global genetic diversity status and
1217
trends: towards a suite of Essential Biodiversity Variables (EBVs) for
1218
genetic composition.
Biological Reviews
, 97(4), 1511-1538.
1219
doi:https://doi.org/10.1111/brv.12852
1220
Hoban, S., Bruford, M., D'Urban Jackson, J., Lopes-Fernandes, M., Heuertz,
1221
M., Hohenlohe, P. A., . . . Laikre, L. (2020). Genetic diversity targets
1222
and indicators in the CBD post-2020 Global Biodiversity Framework
1223
must be improved.
Biological Conservation,
248, 108654.
1224
doi:https://doi.org/10.1016/j.biocon.2020.108654
1225
Hockings, K. J., Sousa, C. (2013). Human-Chimpanzee Sympatry and
1226
Interactions in Cantanhez National Park, Guinea-Bissau: Current
1227
Research and Future Directions.
Primate Conservation
, 26, 57-65.
1228
Hockings, K. J., Mubemba, B., Avanzi, C., Pleh, K., Düx, A., Bersacola, E., . . .
1229
Leendertz, F. H. (2021). Leprosy in wild chimpanzees.
Nature
,
1230
598(7882), 652-656. doi:10.1038/s41586-021-03968-4
1231
Humle, T., Boesch, C., Campbell, G., Junker, J., Koops, K., Kuehl, H., & Sop, T.
1232
(2016). Pan troglodytes ssp. verus (errata version published in 2016).
1233
Retrieved from The IUCN Red List of Threatened Species 2016.
1234
e.T15935A102327574.
1235
IBAP. (2008). Plano de gestão Parque Natural das Lagoas de Cufada,
1236
Guiné-Bissau, 2008–2018 http://www.ibap-gbissau. org. p 116.
1237
IUCN SSC Primate Specialist Group (2020). Regional action plan for the
1238
conservation of western chimpanzees (
Pan troglodytes verus
) 2020–
1239
2030. E. G. Wessling, T. Humle, S. Heinicke et al. Gland, Switzerland,
1240
IUCN.
1241
Jamieson, I. G., & Allendorf, F. W . (2012). How does the 50/500 rule apply to
1242
MVPs?
Trends in Ecology & Evolution
, 27(10), 578-584.
1243
doi:https://doi.org/10.1016/j.tree.2012.07.001.
1244
Jones, A. T., Ovenden, J. R., & Wang, Y. G. (2016). Improved confidence
1245
intervals for the linkage disequilibrium method for estimating
1246
effective population size.
Heredity, 117
(4), 217-223.
1247
doi:10.1038/hdy.2016.19
1248
Kalinowski, S. T., Wagner, A. P., & Taper, M. L. (2006). ml-relate: a computer
1249
program for maximum likelihood estimation of relatedness and
1250
relationship.
Molecular Ecology Notes, 6
(2), 576-579.
1251
doi:https://doi.org/10.1111/j.1471-8286.2006.01256.x
1252
Karibuhoye, C. (2004). Mammal Conservation status and Prospects for
1253
community-based wildlife management in coastal Guinea-Bissau,
1254
West Africa. Master thesis.
Georg
-
August
-
Universität, Gö(ngen
,
1255
Germany
1256
Kormos, R., & Boesch, C. (2003). Regional action plan for the conservation of
1257
chimpanzees in West Africa. Washington DC, USA: IUCN/SSC Primate
1258
Specialist Group and Conservation International.
1259
46
Kühl, H. S., Sop, T., Williamson, E. A., Mundry, R., Brugière, D., Campbell, G.,
1260
... & Boesch, C. (2017). The Critically Endangered western chimpanzee
1261
declines by 80%.
American Journal of Primatology
, 79(9), e22681.
1262
Langergraber, K. E., Prüfer, K., Rowney, C., Boesch, C., Crockford, C., Fawcett,
1263
K., . . . Vigilant, L. (2012). Generation times in wild chimpanzees a nd
1264
gorillas suggest earlier divergence times in great ape and human
1265
evolution.
Proceedings of the National Academy of Sciences,
109(39),
1266
15716-15721. doi:doi:10.1073/pnas.1211740109
1267
Lester, J. D., Vigilant, L., Gratton, P., McCarthy, M. S., Barratt, C. D., Dieguez,
1268
P., . . . Arandjelovic, M. (2021). Recent genetic connectivity and clinal
1269
variation in chimpanzees.
Communications Biology
, 4(1), 283.
1270
doi:10.1038/s42003-021-01806-x
1271
Li, H., & Durbin, R. (2011). Inference of human population history from
1272
individual whole-genome sequences.
Nature
, 475(7357), 493-496.
1273
Luikart, G., Ryman, N., Tallmon, D. A., Schwartz, M. K., & Allendorf, F. W.
1274
(2010). Estimation of census and effective population s izes: the
1275
increasing usefulness of DNA-based approaches.
Conservation
1276
Genetics,
11(2), 355-373. doi:10.1007/s10592-010-0050-7.
1277
Melo, P., Lança, A., Mané, B., Regalla, A., Ferreira da Silva, M., Tavares, L., . . .
1278
M., O. (2018). Tinea corporis by Microsporum audouinii in a female
1279
chimpanzee (Pan troglodytes) from Guinea-Bissau: A case report.
1280
Journal of Medical Primatology
. doi:
1281
https://doi.org/10.1111/jmp.12370.
1282
Marandel, F., Lorance, P., Berthelé, O., Trenkel, V. M., Waples, R. S., & Lamy,
1283
J.-B. (2019). Estimating effective population size of large marine
1284
populations, is it feasible?
Fish and Fisheries
, 20(1), 189-198.
1285
doi:https://doi.org/10.1111/faf.12338
1286
Minhós, T., Borges, F., Parreira, B., Oliveira, R., Aleixo-Pais, I., Leendertz, F. H.,
1287
. . . Chikhi, L. (2023). The importance of well protected forests for the
1288
conservation genetics of West African colobine monkeys.
American
1289
Journal of Primatology
, 85(1), e23453. doi:10.1002/ajp.23453
1290
Minhós, T., Nixon, E., Sousa, C., Vicente, L. M., da Silva, M. F., Sá, R., &
1291
Bruford, M. W. (2013). Genetic evidence for spatio-temporal changes
1292
in the dispersal patterns of two sympatric African colobine monkeys.
1293
Am J Phys Anthropol
, 150(3), 464-474. doi:10.1002/ajpa.22223.
1294
Minhós, T., Chikhi, L., C., S., Vicente, L. M., Ferreira da Silva, M. J., Heller, R., .
1295
. . Bruford, M. W. (2016). Genetic consequences of human forest
1296
exploitation in two colobus monkeys in Guinea Bissau.
Biological
1297
Conservation
, 194, 194-208. doi:10.1016/j.biocon.2015.12.019
1298
Nadachowska-Brzyska, K., Burri, R., Smeds, L., & Ellegren, H. (2016). PSMC
1299
analysis of effective population sizes in molecular ecology and its
1300
application to black-and-white Ficedula flycatchers
. Molecular
1301
Ecology
, 25(5), 1058-1072. doi:10.1111/mec.13540
1302
Newman, D., & Pilson, D. (1997). Increased probability of ex tinction due to
1303
decreased genetic effective population size: experimental populations
1304
of Clarkia pulchella.
Evolution
, 51(2), 354-362. doi:10.1111/j.1558-
1305
5646.1997.tb02422.x
1306
47
O'Brien, D., Laikre, L., Hoban, S., Bruford, M. W., Ekblom, R., Fischer, M. C., . .
1307
. MacDonald, A. J. (2022). Bringing together approaches to reporting
1308
on within species genetic diversity.
Journal of Applied Ecology
, 59(9),
1309
2227-2233. doi:https://doi.org/10.1111/1365-2664.14225.
1310
Palma, L., Godinho, R., Quecuta, Q., Mereck, T., Mandeck, J., Só, T. U., . . .
1311
Beja, P. (2023). African Forest elephants persist in Guinea-Bissau but
1312
require an emergency conservation plan.
Oryx
, 1-4.
1313
doi:https://doi.org/10.1017/S0030605323000674.
1314
Palminteri, S., Joshi, A., Dinerstein, E., Pintea, L., Fernando, S., Davis, C. and
1315
Hansen, M. (2018). Mapping change in ape habitats: forest status,
1316
loss, protection and future risk. In: Arcus Foundation (ed.). The State
1317
of the Apes: Infrastructure Development and Ape Conservation.
1318
Cambridge, UK: Cambridge University Press, pp. 200–223.
1319
https://www.stateoftheapes.com/.
1320
Peel, D., Waples, R. S., Macbeth, G. M., Do, C., & Ovenden, J. R. (2013).
1321
Accounting for missing data in the estimation of contemporary
1322
genetic effective population size (
N
e
).
Molecular Ecology Resources,
1323
13
(2), 243-253. doi:https://doi.org/10.1111/1755-0998.12049
1324
Pérez-Pereira, N., Wang, J., Quesada, H., & Caballero, A. (2022). Prediction of
1325
the minimum effective size of a population viable in the long term.
1326
Biodiversity and Conservation
, 31(11), 2763-2780.
1327
doi:10.1007/s10531-022-02456-z
1328
Pew, J., Muir, P. H., Wang, J., & Frasier, T. R. (2015). related: an R package for
1329
analysing pairwise relatedness from codominant molecular markers.
1330
Molecular Ecology Resources, 15
(3), 557-561.
1331
doi:https://doi.org/10.1111/1755-0998.12323
1332
Poplin, R., Ruano-Rubio, V., DePristo, M. A., Fennell, T. J., Carneiro, M. O.,
1333
Van der Auwera, G. A., . . . Banks, E. (2018). Scaling accurate genetic
1334
variant discovery to tens of thousands of samples. bioRxiv, 201178.
1335
doi:10.1101/201178
1336
Prado-Martinez, J., Sudmant, P. H., Kidd, J. M., Li, H., Kelley, J. L., Lorente-
1337
Galdos, B., . . . Marques-Bonet, T. (2013). Great ape genetic diversity
1338
and population history.
Nature
, 499(7459), 471-475.
1339
doi:10.1038/nature12228.
1340
Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of population
1341
structure using multilocus genotype data.
Genetics
, 155(2), 945-959.
1342
doi:10.1093/genetics/155.2.945.
1343
Rogers, J., Raveendran, M., Harris, R., Mailund, T., Leppälä, K., Athanasiadis,
1344
G., . . . Consortium, B. G. A. (2019). The comparative genomics and
1345
complex population history of Papio baboons.
Science Advances
, 5,
1346
eaau6947. doi:https://doi.org/10.1126/sciadv.aau6947.
1347
Reid, B. N., & Pinsky, M. L. (2022). Simulation-Based Evaluation of Methods,
1348
Data Types, and Temporal Sampling Schemes for Detecting Recent
1349
Population Declines.
Integrative and Comparative Biology, 62
(6),
1350
1849-1863. doi:10.1093/icb/icac144
1351
48
Ryman, N., Laikre, L., & Hössjer, O. (2019). Do estimates of contemporary
1352
effective population size tell us what we want to know?
Molecular
1353
Ecology,
28(8), 1904-1918. doi:https://doi.org/10.1111/mec.15027.
1354
Roberts, N., Lamb, H. F., El Hamouti, N., & Barker, P. (1994). Abrupt Holocene
1355
hydroclimatic events: Palaeolimnological evidence from Northwest
1356
Africa. In A. C. M. a. K. Pye (Ed.), Environmental Change n Drylands:
1357
Biogeographical and Geomorphological Perspectives (pp. 163-175).
1358
London, UK: Wiley.
1359
Sá, R., Ferreira da Silva, M. J., Sousa, F., & Minhós, T. (2012). The trade in and
1360
ethnobiological use of Chimpanzee body parts in Guinea-Bissau:
1361
implications for conservation.
Traffic
, 24(1), 30-34.
1362
Sá, R. M., Petrášová, J., Pomajbíková, K., Profousová, I., Petrželková, K. J.,
1363
Sousa, C., . . . Modrý, D. (2013). Gastrointestinal symbionts of
1364
chimpanzees in Cantanhez National Park, Guinea-Bissau with respect
1365
to habitat fragmentation.
American Journal of Primatology
, 75, 1032-
1366
1041.
1367
Sarnthein, M., Erlenkeuser, H., & Zahn, R. (1982). Termination I: The response
1368
of continental climate in the subtropics as recorded in deep-sea
1369
sediments.
Bul. Inst. du Bassin d'Aquitair
e, 31, 393-407.
1370
Santiago, E., Novo, I., Pardiñas, A. F., Saura, M., Wang, J., Caballero, A. (2020).
1371
Recent Demographic History Inferred by High-Resolution Analysis of
1372
Linkage Disequilibrium.
Molecular Biology and Evolution, 37
(12),
1373
3642-3653. doi:10.1093/molbev/msaa169.
1374
Santos-del-Blanco, L., Olsson, S., Budde, K. B., Grivet, D., González-Martínez,
1375
S. C., Alía, R., & Robledo-Arnuncio, J. J. (2022). On the feasibility of
1376
estimating contemporary effective population size (Ne) for genetic
1377
conservation and monitoring of forest trees.
Biological Conservation,
1378
273
, 109704. doi:https://doi.org/10.1016/j.biocon.2022.109704.
1379
Schubert, M., Ermini, L., Sarkissian, C. D., Jónsson, H., Ginolhac, A., Schaefer,
1380
R., . . . Orlando, L. (2014). Characterization of ancient and modern
1381
genomes by SNP detection and phylogenomic and metagenomic
1382
analysis using PALEOMIX.
Nature Protocols
, 9(5), 1056-1082.
1383
doi:10.1038/nprot.2014.063
1384
Steux, C., Couloigner, C., Arredondo, A., Rodríguez, W., Mazet, O.,
1385
Tournebize, R., & Chikhi, L. (2024). On the demographic history of
1386
chimpanzees and some consequences of integrating population
1387
structure in chimpanzees and other great apes. bioRxiv,
1388
2024.2006.2014.599042. .
1389
doi:https://doi.org/10.1101/2024.06.14.599042
1390
Silva, F. (2024). Population Genomic Survey of Two Non-Human Primate
1391
Species in Sub-Saharan West Africa. Master in Conservation Biology,
1392
Lisbon, Portugal.
1393
Sousa, F. M. (2009). Densidade de
Pan troglodytes veru
s e uso de Recursos
1394
Naturais pela População Local, Gadamael, República da Guiné-Bissau.
1395
(Master thesis), Lisbon University, Lisbon, Portugal.
1396
49
Sousa, J., Ainslie, A., & Hill, C. M. (2017). Sorcery and nature conservation.
1397
Environmental Conservation
, 1-6.
1398
doi:doi:10.1017/S0376892917000327.
1399
Sousa, J., Casanova, C., Barata, A. V., & Sousa, C. (2014). The effect of canopy
1400
closure on chimpanzee nest abundance in Lagoas de Cufada National
1401
Park, Guinea-Bissau.
Primates,
55(2), 283-292. doi:10.1007/s10329-
1402
013-0402-2.
1403
Smith, B. J. (2007). boa: An R Package for MCMC Output Convergence
1404
Assessment and Posterior Inference.
Journal of Statistical Software,
1405
21
(11), 1-37. doi:http://www.jstatsoft.org/
1406
Storz, J. F., Beaumont, M. A. (2002). Testing for genetic evidence of
1407
population expansion and contraction: an empirical analysis of
1408
microsatellite DNA variation using a hierarchical Bayesian model.
1409
Evolution, 56
(1), 154-166. doi:10.1111/j.0014-3820.2002.tb00857.x.
1410
Stiles, D. (2023). Empty Forests. How politics, economics and corruption fuel
1411
live great ape trafficking. Retrieved from Geneva, Switzerland:
1412
https://globalinitiative.net/wp-content/uploads/2023/04/Daniel-
1413
Styles-Empty-Forests-live-great-ape-trafficking.April2023.pdf
1414
Stiles, D., Redmond, I., Cress, D., Nellemann, C., & Formo, R. K. (2013). Stolen
1415
Apes – The Illicit Trade in Chimpan- zees, Gorillas, Bonobos and
1416
Orangutans. A Rapid Response Assessment: GRID-Arendal.
1417
Taberlet, P., Griffin, S., Goossens, B., Questiau, S., Manceau, V., Escaravage,
1418
N., . . . Bouvet, J. (1996). Reliable genotyping of samples with very low
1419
DNA quantities using PCR.
Nucleic Acids Research
, 24(16), 3189-3194.
1420
doi:10.1093/nar/24.16.3189
1421
Taberlet, P., Waits, L., & Luikart, G. (1999). Noninvasive genetic sampling:
1422
look before you leap.
Trends in Ecology & Evolution
, 14(8), 323-327.
1423
Teixeira, H., Salmona, J., Arredondo, A., Mourato, B., Manzi, S .,
1424
Rakotondravony, R., . . . Radespiel, U. (2021). Impact of model
1425
assumptions on demographic inferences: the case study of two
1426
sympatric mouse lemurs in northwestern Madagascar.
BMC Ecology
1427
and Evolution,
21(1), 197. doi:10.1186/s12862-021-01929-z
1428
Teleki, G. (1980). Hunting and trapping wildlife in Sierra Leone: Aspects of
1429
exploitation and exportation. Report submitted to the Office of
1430
President, Sierra Leone. Retrieved from Washington, DC: George
1431
Washington University, 1980:
1432
https://portals.iucn.org/library/node/19295.
1433
Van Wyngaarden, M., Snelgrove, P. V. R., DiBacco, C., Hamilton, L. C.,
1434
Rodríguez-Ezpeleta, N., Jeffery, N. W., . . . Bradbury, I. R. (2017).
1435
Identifying patterns of dispersal, connectivity and selection in the sea
1436
scallop, Placopecten magellanicus, using RADseq-derived SNPs.
1437
Evolutionary Applications
, 10(1), 102-117.
1438
doi:https://doi.org/10.1111/eva.12432
1439
Venn, O., Turner, I., Mathieson, I., de Groot, N., Bontrop, R., & McVean, G.
1440
(2014). Strong male bias drives germline mutation in chimpanzees.
1441
Science
, 344(6189), 1272-1275. doi:10.1126/science.344.6189.1272
1442
50
Waits, L. P., Luikart, G., & Taberlet, P. (2001). Estimating the probability of
1443
identity among genotypes in natural populations: cautions and
1444
guidelines.
Molecular Ecology
, 10(1), 249-256. doi:
1445
https://doi.org/10.1046/j.1365-294X.2001.01185.x
1446
Wakeley, J. (1999). Nonequilibrium migration in human history. Genetics,
1447
153(4), 1863-1871. doi:10.1093/genetics/153.4.1863
1448
Wang, J., Santiago, E., & Caballero, A. (2016). Prediction and estimation of
1449
effective population size.
Heredity
, 117(4), 193-206.
1450
doi:10.1038/hdy.2016.43.
1451
Waples, R. S. (2006). A bias correction for estimates of effective population
1452
size based on linkage disequilibrium at unlinked gene loci.
1453
Conservation Genetics, 7
(2), 167-184. doi:10.1007/s10592-005-9100-
1454
y.
1455
Waples, R. S., Do, C. (2010). Linkage disequilibrium estimates of
1456
contemporary N e using highly variable genetic markers: a largely
1457
untapped resource for applied conservation and evolution.
Evol Appl,
1458
3
(3), 244-262. doi:10.1111/j.1752-4571.2009.00104.x.
1459
Waples, R. S., England, P. R. (2011). Estimating contemporary effective
1460
population size on the basis of linkage disequilibrium in the face of
1461
migration.
Genetics, 189
(2), 633-644.
1462
doi:10.1534/genetics.111.132233.
1463
Waples, R. S. (2022). What is Ne, anyway?.
The Journal of Heredity
, 113(4),
1464
371–379. doi:https://doi.org/10.1093/jhered/esac023
1465
Wright, S. (1931). Evolution in Mendelian Populations.
Genetics
, 16(2), 97-
1466
159. doi:10.1093/genetics/16.2.97
1467
Xue, Y., Prado-Martinez, J., Sudmant, P. H., Narasimhan, V., Ayub, Q., Szpak,
1468
M., . . . Scally, A. (2015). Mountain gorilla genomes reveal the impact
1469
of long-term population decline and inbreeding.
Science
, 348(6231),
1470
242-245. doi:10.1126/science.aaa3952
1471
1472
1473 Tables and Figures 1474 1475
51
1476
Figure 1
a)
Distribution of the western chimpanzee (
Pan troglodytes verus
) in West Africa and b) the location of the study
1477
area in Guinea-Bissau. Unique genotypes for 10 microsatellite loci (represented by purple circles, N=143) were obtained
1478
from fecal samples collected non-invasively in four parks located in southern Guinea-Bissau and encompassing the national
1479
range of chimpanzees – Cufada Lagoons Natural Park (CLNP), Cantanhez National Park (CNP), Dulombi National Park (DN P)
1480
and Boé National Park (BNP). The location of ecological corridors is also indicated. Gadamael area (referred to in table 1) is
1481
also mapped and represented as a white star. c) Pictures show the four confiscated chimpanzees and one road-killed
1482
individual that were sampled to generate whole-genome sequencing data. Blood samples were drawn from the
1483
confiscated individuals during placement in a sanctuary abroad. One tissue sample was obtained from one road killed
1484
minutes after fatality. The location of confiscated individuals in the map in b) (represented by triangles) is estimated and
1485
reflects information on the individual’s origin obtained from national authorities (
e.g.,
Bo from CNP) or where individuals
1486
were found in private premises. Photo credits by MJFS, H. Foito (European Union Embassy Bissau), C. Casanova, L. Almeida
1487
and I. Camará.
1488
1489
1490
1491
Table 1
. Compilation of results from past studies estimating the density of local populations
1492
of chimpanzees in Guinea-Bissau
1493
1494
Site Reference Method Estimated density
(ind./km
2
) Estimated number of
individuals
Cantanhez
National Park
Sousa J. 2009 nest count method
for density estimation 1.937 and 2.340 *2,070 and 2,454
Torres et al., 2010 modeling MI 376 and 2632
Hockings et al., 2013 MI 3.00 MI
Cufada
Lagoons
Natural Park
Carvalho et al., 2013 nest count method
for density estimation 0.22 137 (95% C.I. 51–390)
Sousa et al., 2013 marked nest count
method 0.79 (CI 95 %: 0.61–1.04) 300 (CI 95 %: 230–390)
Gadamael Sousa F. (2009) nest count method 0.897 33
52
for density estimation
Boé National
Park Schwarz et al., 2007 Interviews MI 710
Dias et al., 2018++ nest count method
for density estimation 0.38 18
Wenceslau, 2014 nest count method
for density estimation 1.8 MI
Binczik et al., 2017 standing crop nest
count method 0.77 (95% CI 0.45–1.34) 1,465 to 4,415
Heinicke et al., 2019a modeling 6.76 MI
Overall
population Heinicke et al., 2019a modeling
1,908 (923–6121 95% CI)
1495
The study areas of the majority of studies are National Parks, except for Gadamael (see Fig. 1). The geographic location of
1496
sites within the country is indicated in Fig. 1. Details on the methodology can be found in the respective publications. We
1497
show the 95% confidence intervals (CI) when reported by the original study. *indicates that the estimated number of
1498
individuals was calculated here by multiplying the density reported in the study by the area of the respective protected
1499
area in https://ibapgbissau.org/areas-protegidas/. ++Please note that the study by Dias et al., 2018 only considers 47km
2
of
1500
study area. MI – indicates missing information in the original study
1501
1502
Table 2
Estimation of long-term
N
e
for the whole dataset and per park by employing the Bayesian likelihood-based
1503
approach implemented in MSVAR 1.3 (Storz & Beaumont 2002) and using microsatellite loci data.
1504
Site N
N
0
(Historical)
N
1
(ancestral) Time ago of
demographic change
(years)
Estimated demographic
trajectory
Global dataset 143 4,500 - 6,500 10,705 – 11,910 44,596 – 69,518 Evidence of mild
demographic bottleneck
Cantanhez National
Park 58 500 - 1,125 10,615 – 11,644 5,288 – 12,477 One order of magnitude
demographic bottleneck
Cufada Lagoons
Natural Park 38 534 - 1,225 10,397 – 11, 416 3,612 – 7,874 One order of magnitude
demographic bottleneck
Dulombi National
Park 13 2,769 - 7,401 9,356 – 10,651 13,110 – 32,085 Stable
Boé National Park 34 6,716 - 24,642 7,320 – 8,555 6,753 – 85,645 Stable
1505
In the table is the number of genotypes (N), the posterior distributions of present-day population size
N
0
, the ancient
1506
population size
N
1
(reproductive individuals), the time at which the population started to change (
T
, in years, assuming a 25
1507
years generation time) for four independent runs for each dataset considered and the estimated demographic trajectory.
1508
The median and 90% highest posterior density intervals (HPD90%) are indicated for
N
0
,
N
1
and T.
1509
1510