From mid-November to late December 2011, we surveyed 12 sites in nine communes in and around the south-western part of the range of the Endangered crowned sifaka Propithecus coronatus in western Madagascar. We observed sifaka appearing to be P. coronatus at four sites in the Menabe Region; at three of them several of the sifaka were melanistic. Decken's sifaka P. deckenii were recorded at three sites, and Verreaux's sifaka P. verreauxi at one site. We found no evidence of sifakas between the Tsiribihina and Manambolo rivers west of latitude about 45°E, and local people claimed they have never existed there. We therefore recommend that the forests in this area be excluded from the current distribution range of P. coronatus. We found no sifakas resembling P. deckenii in the melanistic P. coronatus groups, which appears to confirm the hypothesis that most previous reports of possible sympatry between these two species can be better explained by melanism in P. deckenii. The typical individuals of P. coronatus we report showed lightly or heavily washed rufous coloration on the forearms and upper back. The melanistic forms we describe can be categorized as either a “very dark” form, characterized by dark brown to blackish coloration on the forearms and upper back, or as an “intermediate” form, showing dull rufous or light brown forearms and upper back. In reality there appeared to be a continuum in chromatic variation from the typically colored individuals, through the intermediate melanistic form, to the very dark form. These melanistic sifakas appear similar to the darker melanistic forms of P. deckenii north of the Manombolo River, rather than to the melanistic forms of P. coronatus reported to the south-east of our survey sites, or to melanistic forms of P. verreauxi south of the Tsiribihina River. Ancient gene-flow between sifaka taxa may be one of the causes of these melanistic tendencies, but we recommend further research to clarify the situation. These newly reported populations are at a high risk of local extinction. Containing melanistic forms not found elsewhere, we propose that they should be considered of considerable conservation importance with regard to preserving unique chromatic variation, and probably also genetic diversity, in P. coronatus.
Introduction
The crowned sifaka Propithecus coronatus is an Endangered (IUCN 2012; Salmona et al. 2014) lemur endemic to Madagascar. Recent work has led to a greatly improved understanding of its distribution (King et al. 2012; Rakotonirina et al. 2014) and abundance (Salmona et al. 2014). However, while the distribution of the species is fairly clear in the northern part of its range, where it is limited to remaining dry deciduous forest fragments between the Betsiboka and Mahavavy rivers, it is less clear in the southern parts (Mittermeier et al. 2008, 2010; Rakotonirina et al. 2014). Rakotonirina et al. (2014) therefore recommended further surveys to refine the known limits of the species in the south-west of its range, between the Mahajilo, Manambolo and Tsiribihina rivers (Figs. 1 and 2). This same region is also important for investigating patterns of sifaka chromatic variation and melanism, with melanistic sifaka occurring to the north of the Manambolo River within populations of Decken's sifaka P. deckenii (Petter and Peyrieras 1972; Petter et al. 1977; Tattersall 1986; Curtis et al. 1998; Thalmann et al. 2002; Rakotonirina et al. 2014), to the south of the Mahajilo River in a group of crowned sifaka (Razafindramanana and Rasamimanana 2010; Rakotonirina et al. 2014), and to the south of the Tsiribihina River in populations of Verreaux's sifaka P. verreauxi (Mittermeier et al. 2010). Defining species distributions is important for conservation decision-making (Anderson and Martinez-Meyer 2004; Thorn et al. 2009; Rakotonirina et al. 2011), and understanding variation within a species distribution can help define what needs to be conserved (Blair et al. 2011).
With the aim of contributing to a species conservation project for the crowned sifaka (The Aspinall Foundation 2009, 2010; King et al. 2012; Rakotonirina et al. 2014), we therefore undertook a survey of the south-west part of the crowned sifaka range in late 2011. Here we present the results of this survey, including newly reported sites supporting sifakas, descriptions and photographs of chromatic variation amongst the sifaka populations, and preliminary assessments of the threats facing the sites.
Figure 1.
Map of the study site in western Madagascar showing major rivers (dark gray), approximate forest cover (light gray), and survey sites (numbered symbols) where we observed typical P. coronatus only (gray stars), P. coronatus occurring with melanistic forms (black stars), P. deckenii (circles), P. verreauxi (triangles), or an absence of sifakas (diamonds). The names of our numbered survey sites are: 1-Marolaka; 2-Ankaboka; 3-Itondy; 4-Bekopaka; 5-Bevinoa; 6-Antsakavirohazo; 7-Beanka; 8-Andrea; 9-Ankoadava; 10-Ankitapo; 11-Masoarivo; 12-Ambalakapoaky.
Table 1.
Sites surveyed during 2011, summarizing the main results for the presence of sifaka.
Methods
From mid November to late December 2011, two of us (L. Rakotonirina and A. Rakotoarisoa) surveyed 10 sites in seven communes of the Menabe Region of western Madagascar, and an extra two sites in the Melaky Region (Table 1; Fig. 1). Ten of the survey sites were at low elevation (12 to 261 m above sea level), the other two were at low to middle elevations (189 to 780 m above sea level). This area comprises large areas of wooded grassland-bushland mosaics, with restricted areas of fragmented western dry forest, sometimes as gallery forest along watercourses (Moat and Smith 2007).
Following similar methods to those used in previous surveys (Rakotonirina et al. 2014), we consulted local authorities and local populations to select the sites for surveying. We surveyed the sites for one to three days each (Table 1). Accompanied by local guides, we searched for sifakas following paths in the forest or by walking along the edge of forest patches or gallery forests. On each occasion that we observed sifakas we took a GPS point and noted the date and time of the observation, the size of the group, the sex and age-class of each individual wherever possible, and described their coloration. We also noted evidence of threats to the sites. Lemur nomenclature follows Mittermeier et al. (2010).
Results
We made direct observations of sifaka at eight of the 12 survey sites (Table 1; Fig. 1). We observed sifaka appearing to be P. coronatus at four sites, at three of which several of the sifaka were melanistic (Tables 1, 2; Fig. 1). We found P. deckenii at three sites, and P. verreauxi at one site (Tables 1, 2; Fig. 1). Sifaka appeared to be absent at the four remaining sites (Table 1), including at the three sites in the relatively forested part of the survey area between the Tsiribihina and Manambolo rivers (Fig. 1). We recorded no other species of lemur during the surveys.
The P. deckenii individuals we observed (Fig. 3) were of typical coloration, as described by Mittermeier et al. (2010) and Rakotonirina et al. (2014). The three P. verreauxi individuals we observed were also of generally typical coloration for the species, although one showed a quite dark brownish patch covering much of its back (Fig. 3).
At the four sites where we recorded P. coronatus, 56% of individuals we saw (nine of 16) had typical P. coronatus coloration (Table 2; Fig. 4). The other 44% (seven individuals, including two infants) showed melanistic tendencies (Table 2; Figs. 5, 6). Two of five groups contained only typical individuals, the three other groups contained both typical and melanistic individuals (Table 2). The two infants we observed were melanistic, and both were carried by melanistic adults (Table 2; Fig. 5).
The typical individuals showed lightly or heavily washed rufous coloration on the forearms and upper back (Table 3; Figs. 4–6). For the purposes of describing the chromatic variation we recognized two melanistic forms, a “very dark” form characterized by dark brown to blackish coloration on the forearms and upper back, and an “intermediate” form with dull rufous or light brown forearms and upper back (Table 3; Figs. 5 and 6). In reality there appeared to be a continuum in chromatic variation from the typically colored individuals, through the intermediate melanistic form to the very dark form.
The sizes of P. coronatus groups we observed ranged from one to five individuals (Table 2; mean = 3.2, SD = 1.48, n = 5), and of groups of P. deckenii from two to five individuals (Table 2; mean = 3.5, SD = 1.05, n = 6).
We noted the presence of various threats to the sifakas and their habitat at each of the eight sites where sifakas were seen (Table 4). With the notable exception of the Beanka site, where we noted significant disturbance only near the major road running through it, all the sites had several threats, particularly related to habitat disturbance and fragmentation (Table 4). There was logging at two sites, and hunting pressure was high at five sites (Table 4).
Table 2.
Details of groups of P. coronatus, P. deckenii and P. verreauxi observed during our survey.
Figure 2.
Map of west and western central Madagascar showing major rivers (dark gray), approximate forest cover (light gray), distributional records of P. coronatus (gray stars), P. deckenii (gray circles), P. coronatus occurring sympatrically with melanistic forms (black stars) and P. deckenii occurring sympatrically with melanistic forms (black circles). Note that at the site on the south bank of the Manambolo river, one P. deckenii was also recorded in addition to six P. coronatus (Thalmann and Rakotoarison 1994). Distribution records are taken from Wilmé et al. (2006) and references therein, Tattersall (1986), Thalmann and Rakotoarison (1994), Razafindramanana and Rasamimanana (2010), King et al. (2012), Rakotonirina et al. (2014), Salmona et al. (2014), and this paper.
Table 3.
Chromatic description of P. coronatus and melanistic forms observed at the sites of Marolaka, Ankaboka, Itondy and Bekopaka, in the Menabe Region of western Madagascar, December 2011.
Discussion
Geographic range of P. coronatus
Whilst our observations of Propithecus deckenii and P. verreauxi are unremarkable, occurring within the known ranges for the two species as given by Mittermeier et al. (2010), our observations of P. coronatus at four newly reported sites in the Menabe Region, and of an absence of sifakas at three sites in the western parts of this region, help to resolve the confusion over sifaka species limits in the south-west of the P. coronatus range, especially between the Mahajilo, Manambolo and Tsiribihina rivers (Rakotonirina et al. 2014; Fig. 2).
The three most northerly P. coronatus sites we report here, in the Ankavandra and Itondy communes, confirm that the Manambolo River is a boundary between the current distributions of P. coronatus and P. deckenii, even in its upper reaches (Fig. 2). More significantly, our results suggest an absence of sifakas between the Tsiribihina and Manambolo rivers west of latitude about 45°E (Figs. 1, 2). We found no evidence of sifakas in this region, and local people claimed they have never existed there. Indeed the Sakalava people native to the area consider the potential sighting of sifakas in this area as a malediction. With only one, non-georeferenced, published observation of P. coronatus in this general area (Thalmann and Rakotoarison 1994, which also included a single P. deckenii), and one puzzling observation of P. verreauxi lacking detail or explication (Zicoma 1998 in Wilmé et al. 2006), we recommend that the forests in this area be excluded from the current distribution range of P. coronatus. According to the forest cover data gathered from satellite images in 1999 and 2000 by the Madagascar Vegetation Mapping Project data (Moat and Smith 2007), this area includes approximately 1,300 km2 of forests, therefore representing over 30% of the forest cover included in the total range of P. coronatus as proposed by Salmona et al. (2014) when estimating potential total population size for the species. Total forest cover in the range of P. coronatus is therefore probably between 1,600 and 2,850 km2, rather than the 2,350 to 4,150 km2 suggested by Salmona et al. (2014) based on the differing vegetation cover data-sets of MEFT-USAID-CI (2009) and Moat and Smith (2007) respectively. This would consequently have an impact on the total population estimates for P. coronatus given by Salmona et al. (2014), but would not alter the IUCN Red List category they propose of Endangered B1ab.
Table 4.
Threats observed at each of the survey sites where we observed sifaka. (+++: severe; ++: frequent; +: occasional).
Melanism in P. coronatus
Our descriptions and photographs of melanistic individuals from three of the P. coronatus sites we report here add significantly to our knowledge of melanism in this species (King et al. 2012), which in most of its range exhibits fairly limited chromatic variation (Milne-Edwards and Grandidier 1875; Petter and Peyrieras 1972; Rakotonirina et al. 2014), with a general tendency of increasing rufous coloration on the back and limbs from the north to the south of its distribution (Rakotonirina et al. 2014). Melanism in P. coronatus had previously only been reported from a pair captured at an unspecified location “on the track from Tsiroanomandidy to Ankavandra” (Paulian 1953, cited by Petter and Peyrieras 1972), where the male was slightly more rufous dorsally than a typical P. coronatus, and the female was heavily black on the head, upper back, arms, hands, and ventrally, and gray on the base of the nape, the lower back and tail, and from the most southerly known P. coronatus site of Dabolava, where some individuals show varying degrees of dark blackish patches on their shoulders, mid-back, and upper aspects of their arms and thighs (Razafindramanana and Rasamimanana 2010; Rakotonirina et al. 2014).
The presence of a large proportion of melanistic sifakas in the P. coronatus groups at the two sites we report in the Ankavandra commune, south of the Manambolo river but only 30 km south of P. deckenii populations to the north of the Manambolo containing melanistic sifakas (Randrianarisoa et al. 2001; Thalmann et al. 2002; Rakotonirina et al. 2014; Fig. 2), suggests that the melanistic tendencies in both species may be related to former gene flow between them, perhaps during periods of increased habitat connectivity. Other potential causes of melanism in several sifaka species have been proposed, including intermediate environmental or climatic conditions leading to reduced selection pressures and increased expression of variation (Petter and Peyrieras 1972; Petter et al. 1977), but occasional gene flow seems the most likely explanation (Thalmann et al. 2002; King et al. 2012), perhaps accentuated by intermediate environmental or climatic conditions.
Figure 5.
Chromatic variation of P. coronatus at the Marolaka and Ankaboka survey sites, Ankavandra Commune, including typical P. coronatus individuals (top left and center), an individual with an unusual rufous cap (top right), and various melanistic forms (bottom line). Photographs by L. Rakotonirina.
Our observation of melanism in P. coronatus at the Bekopaka site in the Bemahatazana Commune, in gallery forest along a minor northern tributary of the Tsiribihina River, is intriguing as the site is closer to the range of P. verreauxi to the south of the Tsiribihina than to the range of P. deckenii. However, the melanistic sifakas at the site (Fig. 6) appear similar to those in the P. coronatus groups in the Ankavandra commune further north (Fig. 5), and to the darker melanistic forms of P. deckenii north of the Manombolo (Rakotonirina et al. 2014), rather than to melanistic forms of P. verreauxi (Mittermeier et al. 2010, pp. 526–527). Interestingly, some of the P. coronatus individuals at Dabolava, only 50 km east of Bekopaka but south of the Mahajilo River (a major tributary of the Tsiribihina) show melanistic tendencies similar to those expressed by P. verreauxi, of dark blackish patches on the shoulders and limbs (Fig. 7; Razafindramanana and Rasamimanana 2010; Rakotonirina et al. 2014), rather than those expressed at the P. coronatus sites we report here, or at P. deckenii sites. The implication is perhaps that melanism in P. coronatus at Dabolava may be linked to former gene flow with P. verreauxi, whilst at sites between the Tsiribihina and Manambolo rivers it may be linked to former gene flow with P. deckenii.
The precise mechanisms by which gene flow might lead to melanistic tendencies in sifakas remain unclear, especially as melanistic forms do not often resemble first-generation hybrids (Petter and Peyrieras 1972; Petter et al. 1977; Rakotonirina et al. 2014). However melanin-based pigmentation, the most frequent form of pigmentation in primate skin and hair, is known to be genetically controlled (Bradley and Mundy 2008). Two pigmentation genes important in melanin synthesis, the melanocortin-1 receptor (MC1R) and agouti signaling protein (ASIP), have been identified as having a major contribution to chromatic variation in a wide number of animals (Hubbard et al. 2010), although their role in primate coat color variation is complex (Mundy and Kelly 2003, 2006; Bradley and Mundy 2008). Ancestral polymorphism in such pigmentation genes across the Propithecus genus might be one plausible explanation for increased chromatic variation in sifaka populations subject to inter-taxa gene-flow. Again, we recommend further research, including genetic analysis, into the factors influencing chromatically variable sifaka populations throughout Madagascar (Rakotonirina et al. 2014; Fig. 8).
Figure 6.
Sifaka at the Bekopaka survey site, Bemahatazana Commune, including typical P. coronatus individuals (top left), and melanistic forms. Photographs by L. Rakotonirina.
Another interesting observation is that we found no sifakas resembling P. deckenii in the melanistic P. coronatus groups. This appears to confirm our previous hypothesis that most previous reports of possible sympatry between these two species can better be explained by melanism in P. deckenii, where some “intermediate” melanistic forms resemble P. coronatus (Rakotonirina et al. 2014). Any significant gene flow between the two species is therefore likely to be ancient rather than current. Confirmed observations of the species occurring together (Tattersall 1982, 1988; Thalmann and Rakotoarison 1994; Thalmann et al. 2002) are rare, involve only one or two individuals of one species amongst larger numbers of the other, and can be plausibly explained by local and infrequent (perhaps including human-assisted) crossing of biogeographical barriers (Rakotonirina et al. 2014), with minimal subsequent gene flow. Despite the distinctive cranial features of P. coronatus museum specimens (Groves and Helgen 2007), the lack of sympatry, probable former gene flow leading to melanistic tendencies, the potential to hybridize (Petter 1969), and the lack of genetic differentiation (Rumpler et al. 2011), all add weight to the argument that recognizing P. coronatus and P. deckenii as full species might be regarded as a case of taxonomic inflation (Tattersall 2007, 2013).
Threats and conservation
The sifaka group sizes we recorded during this survey were generally similar to those recorded for these species elsewhere (Pichon et al. 2010; King et al. 2012; Rakotonirina et al. 2014; Salmona et al. 2014), but six of the eight sites where we found sifakas contained limited and highly fragmented forest cover, including all four of the sites where we observed P. coronatus (Table 4). Coupled with severe hunting pressure at most sites, other anthropogenic threats (Table 4), and demographic factors influencing small isolated populations (Gilpin and Soulé 1986; Frankham 2005), these newly reported populations are at a high risk of local extinction. The number of sifakas at these isolated sites is probably only a small fraction of the total wild population of P. coronatus (Salmona et al. 2014). However, the populations contain melanistic forms not found elsewhere, and as such should be considered of considerable conservation importance in terms of preserving unique chromatic variation, and probably also genetic diversity, within P. coronatus. We therefore reiterate our previous recommendation (Rakotonirina et al. 2014) to facilitate the creation and management of conservation zones that ensure the maintenance of the full range of chromatic and genetic diversity of P. coronatus.
Figure 7.
Melanistic forms of P. coronatus at Dabolava, near Miandrivazo, in May 2010. Photographs by F.-G. Grandin / MNHN.
Figure 8.
A pale form of P. verreauxi with a white rather than brown cap at Berenty private reserve, southern Madagascar, in October 2010 (left); a melanistic form of P. deckenii at Orimbato, near Ambohijanahary Special Reserve, in March 2010 (center; see Rakotonirina et al. 2014), and two melanistic female sifaka currently in the American Museum of Natural History and collected by A. L. Rand and R. Archbold at “Ambararatabe; Soalala” on 30 March 1931 (right; see Tattersall 1986). Photographs by T. King (left), L. Rakotonirina (center) and Jen Crick / AMNH (right).
Acknowledgments
We thank the Environment and Forests Ministry of the government of Madagascar, represented by the Direction of the Protected Areas System and the Regional Directions of Menabe, Melaky and Bongolava, for their permission to carry out this research. We also thank all the traditional and administrative authorities of the zones we visited, the rural communes, fokontany, Gendarme, and local guides, porters and cooks, for their generous collaboration during this study. We thank Christelle Chamberlan, Mohamad Mbaraka, Rose Marie Randrianarison and Maholy Ravaloharimanitra for their help in organizing the study, François-Gilles Grandin for the use of his photos, and Matthias Markolf and Ian Tattersall for their excellent reviews of the paper. The survey was financed by The Aspinall Foundation, within the framework of their Tsibahaka project for the conservation of the crowned sifaka.
