Introduction
It is often claimed that captive populations are important for biodiversity conservation, especially given continuing threats to wild populations and their habitats (Conde et al., Reference Conde, Flesness, Colchero, Jones and Scheuerlein2011; Barongi et al., Reference Barongi, Fisken, Parker and Gusset2015). However, this claim is regularly challenged (Conway, Reference Conway2003; Leader-Williams et al., Reference Leader-Williams, Balmford, Linkie, Mace, Smith, Stevenson, Zimmermann, Hatchwell, Dickie and West2007; Balmford et al., Reference Balmford, Kroshko, Leader-Williams and Mason2011) because of issues prevalent within zoos or in captive breeding programmes (Lacy, Reference Lacy2013). Most captive populations have limited viability as a result of low population sizes and inbreeding (Lees & Wilcken, Reference Lees and Wilcken2009; Conway, Reference Conway2011; Conde et al., Reference Conde, Colchero, Gusset, Pearce-Kelly, Byers and Flesness2013), low genetic diversity (Muñoz-Fuentes et al., Reference Muñoz-Fuentes, Green and Sorenson2008; Shen et al., Reference Shen, Zhang, He, Yue, Zhang and Zhang2009; Atkinson et al., Reference Atkinson, Kitchener, Tobe and O'Donoghue2018) and limited or skewed breeding success (Roullet, Reference Roullet2012; Kaumanns et al., Reference Kaumanns, Singh and Silwa2013; Penfold et al., Reference Penfold, Powell, Traylor-Holzer and Asa2014; Edwards et al., Reference Edwards, Walker, Dunham, Pilgrim, Okita-Ouma and Shultz2015). There are also uncertainties regarding the taxonomy or geographical provenance of captive animals and issues related to subspecific or interspecific hybridization (Hvilsom et al., Reference Hvilsom, Frandsen, Børsting, Carlsen, Sallé, Simonsen and Siegismund2013). Additional constraints include diseases (Thompson et al., Reference Thompson, Hilliard, Kittel, Lipper, Giddens, Black and Eberle2000), behavioural or genetic adaptation to captivity (McPhee, Reference McPhee2004; Frankham, Reference Frankham2008) and the dominance of non-threatened species over threatened species (Conde et al., Reference Conde, Colchero, Gusset, Pearce-Kelly, Byers and Flesness2013). Therefore, rigorous and transparent criteria that are open to scrutiny are required to justify the maintenance of captive populations for conservation (Balmford et al., Reference Balmford, Mace and Leader-Williams1996; IUCN/SSC, 2014), and existing captive populations need to be assessed for their conservation value (Hvilsom et al., Reference Hvilsom, Frandsen, Børsting, Carlsen, Sallé, Simonsen and Siegismund2013; Gilbert et al., Reference Gilbert, Gardner, Kraaijeveld and Riordan2017; Johann et al., Reference Johann, Roullet, Herrmann and Fienieg2018).
Genetic studies are being used increasingly to help assess captive populations (Ogden et al., Reference Ogden, Chuven, Gilbert, Hosking, Gharbi and Craig2020). Taxonomic uncertainties can be evaluated (Hvilsom et al., Reference Hvilsom, Frandsen, Børsting, Carlsen, Sallé, Simonsen and Siegismund2013; Senn et al., Reference Senn, Banfield, Wacher, Newby, Rabeil and Kaden2014), genetic diversity can be compared between captive populations or with wild populations (El Alqamy et al., Reference El Alqamy, Senn, Roberts, McEwing and Ogden2012; Svengren et al., Reference Svengren, Prettejohn, Bunge, Fundi and Björklund2017), and levels of relatedness and inbreeding can be quantified (Svengren et al., Reference Svengren, Prettejohn, Bunge, Fundi and Björklund2017; Atkinson et al., Reference Atkinson, Kitchener, Tobe and O'Donoghue2018). Moreover, by using genetic information, the probable geographical provenance of founder animals can be ascertained. This can be particularly useful to help with decision-making in relation to the suitability of captive populations or of individual captive animals for prospective reintroduction or other release projects (Ogden et al., Reference Ogden, Ghazali, Hopper, Čulík and King2018), especially of threatened species that have been extirpated locally at some sites or that persist in isolated populations at risk of losing genetic diversity (Farré et al., Reference Farré, Johnstone, Hopper, Kitchener, Roos and King2022).
Lemurs are one of the most threatened groups of primates (Schwitzer et al., Reference Schwitzer, Mittermeier, Johnson, Donati, Irwin and Peacock2014). Ruffed lemurs (including the black-and-white ruffed lemur Varecia variegata and the red ruffed lemur Varecia rubra) occur only in the eastern rainforests of Madagascar (Fig. 1) and are particularly sensitive to habitat loss and disturbance (Vasey, Reference Vasey, Goodman and Benstead2003). Both species are categorized as Critically Endangered on the IUCN Red List (Borgerson et al., Reference Borgerson, Eppley, Patel, Johnson, Louis and Razafindramanana2020; Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020). As with most lemur species, ruffed lemurs are threatened by habitat loss because of deforestation and climate change (Morelli et al., Reference Morelli, Smith, Mancini, Balko, Borgerson and Dolch2020), and hunting for food and live-trapping for the illegal pet trade are additional threats (Golden, Reference Golden2009; Reuter & Schafer, Reference Reuter and Schaefer2017; Borgerson et al., Reference Borgerson, Johnson, Hall, Brown, Narváez-Torres and Rasolofoniaina2022). Habitat protection remains the priority for ensuring the survival of ruffed lemurs (King et al., Reference King, Rasolofoharivelo and Chamberlan2013a,Reference King, Rasolofoharivelo, Randrianasolo, Dolch, Randrianarimanana, Ratolojanahary, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramananab; Schwitzer et al., Reference Schwitzer, King, Robsomanitrandrasana, Chamberlan, Rasolofoharivelo, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013a), with 10 sites supporting V. variegata and two sites supporting V. rubra populations, all listed as priority lemur conservation sites in the IUCN lemur conservation strategy (Schwitzer et al., Reference Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013b). Nevertheless, large numbers of ruffed lemurs are held in captivity, with > 800 V. variegata and > 600 V. rubra reported globally (Schwitzer et al., Reference Schwitzer, King, Robsomanitrandrasana, Chamberlan, Rasolofoharivelo, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013a; Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020). Smaller numbers of both species are also held in Madagascar in recognized facilities (including 35 V. variegata according to a 2014 census) and illegally (Schwitzer et al., Reference Schwitzer, King, Robsomanitrandrasana, Chamberlan, Rasolofoharivelo, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013a; Reuter et al., Reference Reuter, Gilles, Wills and Sewall2016; Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020). Therefore, the two species have been identified as having high potential for integrating in situ and ex situ conservation planning (Schwitzer et al., Reference Schwitzer, King, Robsomanitrandrasana, Chamberlan, Rasolofoharivelo, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013a), with the North American captive population of V. variegata having already been used as a source of captive-born lemurs for reinforcing one small, isolated wild population (Britt et al., Reference Britt, Welch, Katz, Iambana, Porton and Junge2004). An assessment of the conservation value of the captive populations of these two species would aid conservation decision-making.
Fig. 1 Geographical distribution of ruffed lemurs. The red ruffed lemur Varecia rubra occurs in the north-east, whereas the black-and-white ruffed lemur Varecia variegata has a wider and scattered distribution from north to south. Although three subspecies of V. variegata are recognized (V. variegata variegata, V. variegata subcincta and V. variegata editorum), the taxonomic status and distribution of the subspecies are not fully resolved.
Unresolved subspecific taxonomy complicates the current assessment of the conservation value of the captive population of V. variegata (King et al., Reference King, Rasolofoharivelo and Chamberlan2013a; Baden et al., Reference Baden, Holmes, Johnson, Engberg, Louis and Bradley2014). Three subspecies are recognized (Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020): V. variegata subcincta, V. variegata editorum and V. variegata variegata. Varecia v. subcincta occurs in the north of the species range, V. v. editorum in the south and V. v. variegata in the area between V. v. editorum and V. v. subcincta (Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020). Yet the most comprehensive genetic study to date of wild V. variegata populations found a genetic distinction between V. variegata populations located to the north and to the south of the Mangoro River (Baden et al., Reference Baden, Holmes, Johnson, Engberg, Louis and Bradley2014). Although this is a major biogeographical barrier for many taxa in Madagascar (Ganzhorn et al., Reference Ganzhorn, Goodman, Nash, Thalmann, Lehman and Fleagle2006; Wilmé et al., Reference Wilmé, Goodman and Ganzhorn2006), it is not traditionally considered to represent the distributional limit between the two southern subspecies of V. variegata. Current texts consider V. v. editorum to occur on both sides of the Mangoro River, with the distribution limit and possible overlap with V. v. variegata located in the general region of Zahamena National Park (Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020), which is > 200 km north of the Mangoro River. Additionally, genetic evidence suggests that V. v. subcincta may not be a valid subspecies (Baden et al., Reference Baden, Holmes, Johnson, Engberg, Louis and Bradley2014; Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020). Further work is underway to resolve these taxonomic issues (Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020). However, previous genetic research (Baden et al., Reference Baden, Holmes, Johnson, Engberg, Louis and Bradley2014) provides a baseline for ascertaining the geographical provenance of captive ruffed lemurs. Work is also needed to determine the genetic diversity of captive ruffed lemurs in Europe and Madagascar and their relationships with ruffed lemurs in North America and in the wild.
Here we assess the mitochondrial genetic diversity of captive ruffed lemurs in Madagascar, Europe and North America, focusing primarily on V. variegata, using analyses of new samples from lemurs in Madagascar and Europe and data published previously on lemurs in North America. We compare the results from captive lemurs with published data from wild lemurs, with a particular emphasis on ascertaining the geographical provenance of the founders of the global captive population. Our findings will help inform decision-making regarding the potential conservation value and roles of captive populations and the integration of ex situ and in situ conservation activities for ruffed lemurs.
Methods
Molecular biology techniques
We obtained 51 new samples for this study, including nine muscle samples from V. variegata in the CryoArks Biobank at National Museums Scotland (derived from animals in UK zoos during 1989–2012) and 42 hair (with follicle), whole-blood and Whatman FTA card (Merk, Darmstadt, Germany) blood samples from captive ruffed lemurs collected by zoo and/or veterinary professionals in Madagascar and Europe, including samples from V. rubra, V. v. subcincta and V. variegata of undetermined subspecies (but phenotypically not of V. v. subcincta; Supplementary Table 1). In addition, we retrieved 229 mitochondrial DNA (mtDNA) D-loop sequences (accession numbers KJ700486–KJ700626, AF173507–AF173547, AF475865–AF475904, AF493668–AF493671 and AY584494) from GenBank (we later removed AF173519, AF173521, AF173522 and AF173530 from the analysis because of their short sequence length).
We obtained DNA from hair, whole-blood and muscle tissue samples using the GeneJET Genomic DNA Purification Kit (ThermoFisher Scientific, Waltham, USA). We obtained DNA from dried blood samples from Whatman FTA cards following a custom method for DNA extraction (Supplementary Material 1). We performed PCR to amplify the mtDNA D-loop region of ruffed lemurs using heavy strand dLp5 (5′-CCATCGWGATGTCTTATTTAAGRGGAA-3′; Baker et al., Reference Baker, Perry, Bannister, Weinrich, Abernethy and Calambokidis1993) and light strand dLp1.5 (5′-GCACCCAAAGCTGARRTTCTA-3′; Wyner et al., Reference Wyner, Amato and DeSalle1999) primers following standard protocols, resulting in 536 base-pair fragments (Supplementary Material 1).
Data analysis
We analysed 276 DNA sequences of ruffed lemurs (Supplementary Table 1). We aligned all sequences using BioEdit 7.2.6 (Hall, Reference Hall1999). We divided the sample into nine groups (Table 1) for downstream analysis according to the known species or subspecies classification, the geographical origin of wild V. v. editorum samples (north or south of the Mangoro River), the status of the individuals (wild or captive) and whether captive individuals were part of the captive population in Madagascar, the European Association of Zoos and Aquaria Ex-situ Programme (EEP; formerly known as European Endangered Species Programme), the North American Association of Zoos and Aquariums Species Survival Plan Programmes (SSP) or a non-EEP European population (samples from Fenn Bell Conservation Project; FBC).
Table 1 DNA polymorphism and population expansion test of ruffed lemurs (Varecia spp.). Bold indicates significant test of neutrality (P < 0.05).
1 W, wild; C, captive; S/N, south/north of the Mangoro River; EEP, European Endangered Species Programme; SSP, Species Survival Plan Programmes; FBC, Fenn Bell Conservation Project.
2n, number of sequences used. 3S, segregating sites. 4H, number of haplotypes. 5Hd, haplotype (gene) diversity. 6π, Jukes–Cantor nucleotide diversity. 7k, average number of nucleotide differences. 8R 2, Ramos-Onsins and Rozas’ value. 9Fs, Fu's Fs value. 10Values calculated based on the full dataset of DNA sequences.
We imported DNA sequences into DnaSP 6.12.03 (Rozas et al., Reference Rozas, Ferrer-Mata, Sánchez-DelBarrio, Guirao-Rico, Librado, Ramos-Onsins and Sánchez-Gracia2017). We obtained several DNA polymorphism indices in DnaSP, including number of segregating sites (S), number of haplotypes (H), haplotype diversity (Hd), nucleotide diversity (π) and average number of nucleotide differences (k). We evaluated changes in population size using Ramos-Onsins and Rozas' R 2 and Fu's Fs values calculated with DnaSP, with significance tested using the coalescent process for a neutral infinite-sites model and assuming a large constant population size (1,000 replications).
To estimate the genetic structure amongst the groups, we obtained pairwise genetic differentiation (F ST) values and the numbers of net nucleotide substitutions per site between groups (Da) using DnaSP. We carried out an analysis of molecular variance in Arlequin 3.5.2.2 (Excoffier et al., Reference Excoffier, Laval and Schneider2005). We generated a phylogenetic network from the haplotypes using PopART (Leigh & Bryant, Reference Leigh and Bryant2015), with the Median Joining algorithm selected.
Results
We obtained 51 new D-loop sequences belonging to the genus Varecia (GenBank accession numbers MZ615228–MZ615278; Supplementary Table 1). These included 23 from captive lemurs in Madagascar, of which three were V. v. subcincta and 20 were V. variegata (unidentified subspecies), four from the EEP V. rubra (from ZooParc de Beauval), 20 from the EEP V. variegata, of which three were V. v. subcincta (from Port Lympne Reserve) and 17 were V. variegata of undetermined subspecies but not V. v. subcincta phenotypically (six from Howletts Wild Animal Park, two from Marwell Zoo and nine from specimens at National Museums Scotland), and four from captive V. variegata of undetermined subspecies held in Europe but not included within the EEP (from FBC).
Genetic diversity
The DNA sequence alignment for the whole dataset (276 D-loop sequences) was 613 base pairs long and contained 455 sites (excluding gaps and missing data) from which 411 were invariable, 41 were informative and three were singletons. We found the highest Hd and π values in the sample from captive V. variegata in Madagascar, and the lowest values in the sample from wild populations of V. v. editorum south of the Mangoro River (Table 1). Genetic diversity in the sample from the EEP V. variegata was much lower than in the SSP sample, and the EEP sample had the second lowest diversity value across all groups (Table 1).
Consistent with the low genetic diversity values, we found small pairwise differences between sequences in the samples from wild populations of V. v. editorum south of the Mangoro River and from the EEP V. variegata, whereas other groups showed some larger pairwise differences, with frequencies of 5–16% (Supplementary Fig. 1). None of the R 2 and Fs values were significantly different from neutral expectations, with there being no indications of sudden population expansions or contractions except for the sample from wild populations of V. v. editorum south of the Mangoro River, which showed a significantly low and negative Fs value (Table 1).
Genetic structure
We found significant pairwise F ST values amongst all groups (Table 2). The values ranged from 20.0% between the samples from captive V. variegata in Madagascar and the samples from wild populations of V. v. variegata, to 97.7% between the samples from the EEP V. variegata and the samples from wild populations of V. v. editorum south of the Mangoro River. These values are even higher than those from pairwise comparisons of different species (e.g. V. rubra and V. variegata). We found the highest pairwise divergence values (Da) to be between V. rubra and all V. variegata groups followed by pairwise comparisons between V. v. subcincta and V. v. variegata and V. v. editorum, and we found the lowest Da values in pairwise comparisons between V. v. variegata and V. v. editorum (Table 3). The analysis of molecular variance showed that 29.1% of the variation was between wild vs captive animals, and 41.3% of the variation was amongst groups and 29.5% was within groups.
Table 2 Pairwise differentiation values (F ST) amongst groups of ruffed lemurs (Varecia spp.). Varecia variegata-Fenn Bell Conservation Project (C) is not included because of small sample size.
1 W, wild; C, captive; S/N = south/north of the Mangoro River; EEP, European Endangered Species Programme; SSP, Species Survival Plan Programmes.
Table 3 Pairwise divergence values (Da) amongst groups of ruffed lemurs (Varecia spp.). Varecia variegata-Fenn Bell Conservation Project (C) is not included because of small sample size.
1 W, wild; C, captive; S/N = south/north of the Mangoro River; EEP, European Endangered Species Programme; SSP, Species Survival Plan Programmes.
The phylogenetic network reflected the taxonomic classification of ruffed lemurs (Fig. 2). The V. rubra and V. v. subcincta haplotypes appeared separately from each other and from other ruffed lemur groups by several mutational steps, and V. v. editorum haplotypes appeared closer together than to any other V. variegata haplotypes. The remaining haplotypes showed a mix of small and large numbers of mutational steps obtained from samples from wild populations of V. v. variegata and from the four groups of captive V. variegata (in Madagascar, the EEP, the SSP and FBC). The most common haplotypes were Hap_5 and Hap_20 (with 84 and 29 sequences, respectively), all belonging to the sample from wild populations of V. v. editorum south of the Mangoro River, followed by Hap_4 (27 sequences) belonging to a mix of groups. The EEP and FBC samples were represented by only two haplotypes (Hap_2 and Hap_4), whereas the SSP sample was represented by seven haplotypes. The captive animals in Madagascar had the highest haplotype diversity, with 10 different haplotypes. Hap_12 was shared between the sample from wild populations of V. v. editorum north of the Mangoro River and the captive samples from Madagascar and the SSP, which indicates that those captive animals from Madagascar and the SSP not identified to subspecies level could have a V. v. editorum maternal origin. Furthermore, several haplotypes belonging to wild V. v. variegata and to captive V. variegata from Madagascar, the EEP, the SSP and FBC were distantly related to other V. v. variegata haplotypes. Specifically, Hap_9, Hap_10 and Hap_11 (all from wild V. v. variegata) were separated by eight mutational steps and distantly related to all other haplotypes.
Fig. 2 Phylogenetic haplotype network of ruffed lemurs (Varecia spp.) based on D-loop mitochondrial DNA sequence data. The node sizes are proportional to the number of sequences belonging to the haplotypes. Mutations between haplotypes are shown using hatch marks. Inferred internal nodes are shown in black. W, wild; C, captive; S/N, south/north of the Mangoro River; Mada, Madagascar; EEP, European Endangered Species Programme; SSP, Species Survival Plan Programmes; FBC, Fenn Bell Conservation Project. (Readers of the printed journal are referred to the online article for a colour version of this figure.)
Discussion
We assessed the mitochondrial genetic diversity of captive ruffed lemurs in Madagascar, Europe and North America and compared this to published data from wild lemurs to ascertain the geographical provenance of the founders of the global captive population. The results could inform decision-making regarding the potential conservation value and roles of the captive ruffed lemur populations and the potential integration of ex situ and in situ conservation practices for ruffed lemurs. We report 10 mtDNA haplotypes that have not yet been recorded from wild or captive Varecia populations. We found one of the novel haplotypes in the four samples from the European captive V. rubra population but there was only one published V. rubra haplotype from wild populations for comparison, with which the captive haplotype clustered. All three haplotypes found in captive V. v. subcincta (one from the European captive population and two from Madagascar) have not been reported previously but clustered with the six published haplotypes from wild V. v. subcincta populations. We found only two other mtDNA haplotypes in the European captive V. variegata population, both of which were identical to previously published haplotypes from wild populations. Conversely, out of 10 mtDNA haplotypes present in the non-V. v. subcincta captive population in Madagascar, only four had been reported previously from wild populations and six are newly reported here.
There are 18 recognized founders of the non-subcincta EEP population and four founders of the V. v. subcincta EEP population (Johann et al., Reference Johann, Roullet, Herrmann and Fienieg2018; Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020). Seven of the 18 founders of the non-subcincta EEP population were female, of which we were able to sample four indirectly in this study through their descendants along uninterrupted maternal lines. As several of the founders of the EEP population are also founders of the North American captive population (SSP), another seven of the EEP founders had been sampled previously in a study of the SSP population (Wyner et al., Reference Wyner, Amato and DeSalle1999), including one female founder sampled indirectly through a maternal-line descendant and six male founders sampled directly. Therefore, we have been able to identify the mtDNA haplotypes of 11 of the 18 founders of the EEP captive population. Studbook analysis illustrates that these 11 founders have contributed 72% of the genetic diversity of the EEP population, illustrating that despite sampling only a small proportion of the EEP population our results are representative of the majority of the living population.
We identified only four different mtDNA haplotypes within the historic non-subcincta EEP population (including only two within contemporary samples), despite having sampled 11 of the 18 founder animals directly or indirectly. Six of the sampled founders share Hap_4 (founders 4, 5, 11, 13, 21 and 25) and three share Hap_2 (founders 14, 15 and 35). Putative founders having the same haplotype suggests they could share common maternal ancestry, and if so, perhaps should not be considered as unrelated founders for the purposes of the analysis of theoretical summary genetic statistics for captive populations or for the purposes of calculating mean kinship and founder representation, and for making breeding recommendations. We recommend additional genetic analysis to ascertain the true levels of relatedness of the putative founders of the captive ruffed lemur populations, as has been done for other species in captivity (Svengren et al., Reference Svengren, Prettejohn, Bunge, Fundi and Björklund2017; Atkinson et al., Reference Atkinson, Kitchener, Tobe and O'Donoghue2018). This would allow a more accurate assessment of the genetic diversity of the original founder populations, with implications for our understanding of the conservation value of the existing captive populations and of the specific breeding recommendations to maximize their conservation potential.
Studbook analysis illustrates that the six founders sharing Hap_4 have contributed 46% of the genetic diversity of the non-subcincta EEP population and the three founders sharing Hap_2 have contributed 19% of this genetic diversity. Therefore, these nine founders sharing only two mtDNA haplotypes have contributed 65% of the genetic diversity of the EEP population. At over 300 individuals (Johann et al., Reference Johann, Roullet, Herrmann and Fienieg2018), the European captive population of non-subcincta V. variegata is relatively large but our results show that mitochondrial genetic diversity is relatively low and that the founders of the population represent only a small proportion of the genetic diversity of the species. Nevertheless, one haplotype in the non-subcincta EEP population is not represented in the captive population from Madagascar, and exchanges between the various regional captive populations could increase the genetic diversity of each and therefore increase their conservation value. The genetic diversity of the global V. rubra captive population is also considered to be low (Borgerson et al., Reference Borgerson, Eppley, Patel, Johnson, Louis and Razafindramanana2020).
Geographical origins of captive ruffed lemurs
All mtDNA haplotypes reported from captive ruffed lemurs were either identical to or clustered with published haplotypes originating from wild Varecia populations located north of the Mangoro River. There is currently no evidence for the presence of lemurs from south of the Mangoro River being incorporated into global captive populations. Moreover, the distantly related haplotypes within V. variegata and specifically those seen in wild V. v. variegata suggest a cryptic genetic structure in ruffed lemurs, warranting further genetic characterization of individuals in the wild.
Our results confirm the low haplotype diversity found in a previous genetic study (Wyner et al., Reference Wyner, Amato and DeSalle1999) but contradict the conclusion from that study that the captive-born V. variegata from North American zoos, which were released into Betampona Reserve from 1997 to 2001 to reinforce a small, isolated wild population (Britt et al., Reference Britt, Welch, Katz, Iambana, Porton and Junge2004), probably originated from the south of the species' range and therefore were not particularly appropriate for release in this Reserve because of its location in the northern part of the species range (Wyner et al., Reference Wyner, Amato and DeSalle1999). This previous study used population aggregation analysis to test for phylogenetic clusters based on diagnostic nucleotide positions and used a restricted baseline from wild populations for comparison. Instead, our results show that the mtDNA of the North American captive population originates from wild populations north of the Mangoro River and therefore that the lemurs released were more suitable genetically for the population reinforcement project than suggested previously (Wyner et al., Reference Wyner, Amato and DeSalle1999). Should further releases of captive V. variegata be considered appropriate in Madagascar within the context of integrating in situ and ex situ lemur conservation (King et al., Reference King, Rasolofoharivelo and Chamberlan2013a; Schwitzer et al., Reference Schwitzer, King, Robsomanitrandrasana, Chamberlan, Rasolofoharivelo, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013a), our results suggest that the current global captive population would be more suitable from a genetic perspective for releases in sites located north of the Mangoro River. However, there are many other issues that would need to be considered prior to any potential releases, including behavioural assessments such as potential naivety to predators, disease risk analyses and socioeconomic considerations, as detailed in international guidelines (IUCN/SSC, 2013).
The most prevalent mtDNA haplotype in the European captive EEP population (Hap_4) has been reported in the wild only from Zahamena National Park (Baden et al., Reference Baden, Holmes, Johnson, Engberg, Louis and Bradley2014), suggesting that much of the genetic diversity of the EEP probably originated from this part of the species' range. The second most prevalent haplotype (Hap_2) has been reported in the wild from Betampona Reserve only, but the wild samples were collected after the release of captive-bred lemurs from the North American captive SSP population, so this haplotype could be derived from the released lemurs rather than the original wild population of Betampona Reserve. This haplotype is similar to Hap_4 and so could have originated from closer to Zahamena National Park; a letter from the 1970s regarding one of the founders with Hap_2 (ISB35) claims that this individual was captured 50 miles north-east of Ambatondrazaka, which would be in or near Zahamena National Park. Betampona Reserve and Zahamena National Park are at similar latitudes in the species' range, so the distinction is unlikely to be significant from an evolutionary perspective. Of the two remaining haplotypes identified from the EEP founders, Hap_26 has not been reported from wild populations but is also similar to Hap_4 from Zahamena National Park, whereas Hap_12 is different and has been obtained from wild lemurs identified as V. v. editorum in the Mantadia, Andasibe and Torotorofotsy sample sites at the southern end of the Ankeniheny–Zahamena Corridor (Baden et al., Reference Baden, Holmes, Johnson, Engberg, Louis and Bradley2014). The three haplotypes that probably originated from in or around Zahamena National Park can be traced back to 10 founder individuals who have contributed at least 60% of the genetic diversity of the EEP population, providing strong evidence for the probable geographical origins of a large proportion of the captive EEP population.
Using mtDNA, this study has helped to ascertain the geographical provenance of several of the founders of the global captive ruffed lemur population, provided insights into the taxonomic classification of captive individuals and determined the genetic diversity of captive ruffed lemurs. Although the use of mtDNA has limitations in comparison with other molecular markers (Nielsen et al., Reference Nielsen, Beger, Henriques and von der Heyden2020), it is useful for species identification and for wildlife forensics (Alacs et al., Reference Alacs, Georges, FitzSimmons and Robertson2010). The use of mtDNA to identify the probable population of origin of captive animals in Madagascar could help us to understand where ruffed lemurs or other species are being captured illegally from wild populations (Reuter et al., Reference Reuter, Gilles, Wills and Sewall2016; Reuter & Schafer, Reference Reuter and Schaefer2017). Our results illustrate that several captive ruffed lemurs in Madagascar have mtDNA haplotypes that have not yet been identified from wild populations. Therefore, we recommend prioritizing the genetic analysis of wild populations that have not yet been sampled, including utilizing historical museum specimens of known origin if available and non-invasive samples from populations under community-based conservation initiatives such as the Andriantantely lowland forest and the western Ankeniheny–Zahamena Corridor (King et al., Reference King, Rasolofoharivelo, Randrianasolo, Dolch, Randrianarimanana, Ratolojanahary, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013b,Reference King, Ravaloharimanitra, Randrianarimanana, Rasolofoharivelo and Chamberlanc; Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020). This would provide us with a better understanding of the genetic diversity of wild ruffed lemur populations and could provide baseline genetic diversity for identifying where lemurs are being captured illegally, especially if there is a particular sampling focus on areas where illegal capture is most likely to be occurring (e.g. forests within relatively easy reach of the major markets in Toamasina). Areas that are identified as probable sites of illegal captures or of other threats to ruffed lemurs or their habitats should also be considered for urgent conservation interventions to mitigate these threats. Community-based conservation of forests and lemurs is a well-established model in Madagascar (King et al., Reference King, Rasolofoharivelo, Randrianasolo, Dolch, Randrianarimanana, Ratolojanahary, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013b,Reference King, Ravaloharimanitra, Randrianarimanana, Rasolofoharivelo and Chamberlanc; Rasolofoharivelo et al., Reference Rasolofoharivelo, Randrianarimanana, King, Randrianasolo, Dolch, Ratolojanahary, Schwitzer, Mittermeier, Davies, Johnson, Ratsimbazafy and Razafindramanana2013; Ravaloharimanitra et al., Reference Ravaloharimanitra, Randrianarimanana, Randriahaingo, Mihaminekena and King2015; Louis et al., Reference Louis, Sefczek, Raharivololona, King, Morelli and Baden2020) and could be implemented or increased for any remaining wild ruffed lemur populations through appropriate local community support.
Conclusion
Given the continuing crisis facing rainforests and ruffed lemurs in Madagascar (Jenkins et al., Reference Jenkins, Keane, Rakotoarivelo, Rakotomboavonjy, Randrianandrianina and Razafimanahaka2011; Seaman et al., Reference Seaman, Randriahaingo, Randrianarimanana, Ravaloharimanitra, Humle and King2018; Morelli et al., Reference Morelli, Smith, Mancini, Balko, Borgerson and Dolch2020), the use of all available tools to tackle these issues should be considered an urgent priority for lemur conservation. Genetic analyses, such as those presented here, can help to inform conservation decision-making. The results of this study indicate that the large global captive population of ruffed lemurs could have some value as a source of potential release stock for reintroduction or reinforcement projects, that the much smaller captive population in Madagascar has higher genetic diversity and greater potential for contributing suitable release candidates, and that effective conservation of wild populations should be considered the highest priority for the conservation of ruffed lemurs and their remaining genetic diversity.
Acknowledgements
We thank Lemurs Park, Lemurialand, Vakona Lodge, Parc Ivoloina, Tsimbazaza Zoo, Howletts Wild Animal Park, Port Lympne Reserve, ZooParc de Beauval, Marwell Zoo, National Museums of Scotland and the Fenn Bell Conservation Project for the provision of samples; Jamie Robertson, Matt Ford, Ellen Holding, Antoine Leclerc, Baptiste Mulot, Justine Shotton, Danielle Free and Natalie Terry for facilitating sample collection; Christiane Schwarz for laboratory work; Lisbeth Høgh, Gina Ferrie and Christie Eddie for providing unpublished data on the European and international studbooks for Varecia variegata; and two anonymous reviewers and the editor for their suggestions and comments. Tissue samples from National Museums Scotland were made available through CryoArks (supported by BBSRC grant BB/R015260/1). Funding was provided by Parc Zoologique de Paris, Muséum National d'Histoire Naturelle and The Aspinall Foundation, with in-kind contributions from Canterbury Christ Church University and the German Primate Center.
Author contributions
Study design: RV, TK; collection of samples: JH, ACK, JC, DR, ER, JDH; laboratory work: RV, JDH, CR; data analysis: RV, TK; writing and editing: all authors.
Conflicts of interest
None.
Ethical standards
This research abided by the Oryx guidelines on ethical standards. Ethical guidelines and considerations by Canterbury Christ Church University, The Aspinall Foundation and the Fenn Bell Conservation Project were followed for procuring hair and blood samples from captive lemurs.
Open access
