According to the IUCN Red List, c. 94% of the > 100 lemur species of Madagascar are threatened, making them the most endangered mammalian group globally (Schwitzer et al., Reference Schwitzer, Mittermeier, Johnson, Donati, Irwin and Peacock2014). The Endangered ring-tailed lemur Lemur catta, a flagship species for the country, is believed to be suffering precipitous population declines (Gould & Sauther, Reference Gould and Sauther2016; LaFleur et al., Reference LaFleur, Gould, Sauther, Clarke and Reuter2018; LaFleur & Gould, Reference LaFleur and Gould2020). Once considered widespread throughout southern Madagascar (Sussman, Reference Sussman1977), L. catta is now considered locally extinct or nearly extinct at several locations where it was found historically (Gould & Sauther, Reference Gould and Sauther2016; LaFleur et al., Reference LaFleur, Clarke, Reuter and Schaeffer2016, Reference LaFleur, Gould, Sauther, Clarke and Reuter2018; LaFleur & Gould, Reference LaFleur and Gould2020; Fig. 1). Nevertheless, much of the historical range of the species remains under-surveyed, raising concerns that current population estimates potentially underrepresent L. catta numbers in the wild (Murphy et al., Reference Murphy, Ferguson and Gardner2017).

Fig. 1 Locations of the 83 surveys for the ring-tailed lemur Lemur catta conducted during June 2018–August 2019 in southern Madagascar. Supplementary Table 1 contains the corresponding site names and numbers. Extant and historical L. catta distributions are from the IUCN Red List (LaFleur & Gould, Reference LaFleur and Gould2020). ‘New survey’ refers to data from this study for sites for which there are no other published surveys. ‘Resurvey’ indicates sites for which other published surveys are also available (Supplementary Table 1). ‘Confirmed population’ indicates sites where species presence has been confirmed despite being previously identified as likely locally extinct (Gould & Sauther, Reference Gould and Sauther2016; LaFleur et al., Reference LaFleur, Clarke, Reuter and Schaeffer2016).

Accurate population assessments are essential for informing species-orientated conservation management (Rylands et al., Reference Rylands, Mittermeier and Williamson2020), including that of L. catta (LaFleur & Gould, Reference LaFleur and Gould2020). To aid these efforts, we present survey results from 83 sites to verify the species' occurrence, refine knowledge of its distribution and contribute to population estimates (Fig. 1).

During June–July 2019 we conducted surveys at five sites where L. catta occurred historically that were either excluded from recent population estimates or had not been systematically surveyed since 1995 (Hawkins, Reference Hawkins1999): (1) Zombitse Forest, (2) Vohibasia Forest, (3) Fiherenana Forest, (4) Ranomay Forest, and (5) two trail circuits in Isalo National Park (Namaza Circuit and Piscine Naturelle) (Fig. 1). We conducted week-long surveys using a standard line transect methodology along existing trails, to minimize forest disturbance (Whitesides et al., Reference Whitesides, Oates, Green and Kluberdanz1988; Hilário et al., Reference Hilário, Rodrigues, Chiarello and Mourthé2012; see Supplementary Material 1 for full details).

During 2-week periods in June, August, September and October 2018 and March, April and August 2019, we surveyed a further 78 sites across seven fokontany (an administrative unit comprising one or more villages) throughout the Mahafaly Plateau, a vast region of south-western Madagascar where surveys for this species have been insufficient (LaFleur et al., Reference LaFleur, Clarke, Reuter and Schaeffer2016; Fig. 1). The sites surveyed around the Mahafaly Plateau lack the infrastructure for conducting line transect sampling; we therefore used broad survey methods following Sussman et al. (Reference Sussman, Green, Porton, Andrianasolondraibe and Ratsirarson2003) to count individuals at sleeping sites (see Supplementary Material 1 for full details). We utilized broad survey methods because of their ability to generate comparative datasets at low cost over relatively short periods of time and their utility for confirming the presence and relative numbers of the population of a species within a proscribed area (Muckenhirn et al., Reference Muckenhirn, Mortensen, Vessey, Fraser and Singh1975).

The combined results from our surveys confirm the presence of L. catta at 65 (78.3%) of the 83 sites. We identified a minimum of 792 L. catta individuals in total (summing only the maximum group size across sites; Supplementary Table 1) and as many as 1,221 individuals (using estimated population counts) from 104 groups at 60 sites (Supplementary Table 1). Our estimates at four sites in the Mahafaly Plateau (Vintany, Andranoilovy, Grotte Maiky and Andranovao South) corroborate the findings presented in a recent report (Kasola et al., Reference Kasola, Atrefony, Louis, Odilon, Ralahinirina, Menjanahary and Ratovonamana2020). Furthermore, we confirmed the presence of L. catta at a further five sites on the Mahafaly Plateau (Antsono, Ankamena, Andramaniloke, Marohazo and Vohindambo), although individuals were obscured from view, making accurate counts impossible. Local guides indicated L. catta was present at the remaining 18 sites, although our surveys did not confirm this.

Among our surveys were three sites where L. catta populations were believed previously to be locally extinct or nearly extinct: Zombitse, Vohibasia and Fiheranana (Supplementary Table 1). As in a previous study (La Fleur et al., Reference LaFleur, Clarke, Reuter and Schaeffer2016), we did not detect L. catta in Zombitse Forest, but we did confirm presence at both Vohibasia and Fiheranana forests. At Vohibasia, which was last surveyed in 2007 (Siers, Reference Siers2007), we observed 10 individuals, thus indicating species persistence in this region, albeit at low densities. We observed a total of 40 individuals in five groups at Fiheranana; local guides suggest that additional L. catta populations occur along the Fiheranana River Valley where it extends beyond the range of our survey. Our survey efforts also confirmed that Ranomay, a historical L. catta stronghold (Murphy et al., Reference Murphy, Ferguson and Gardner2017), and multiple sites throughout the Mahafaly Plateau (e.g. Bemananga, Mananiho and Telomaly) still support a substantial presence of L. catta (Supplementary Table 1).

Although encouraging, our findings are preliminary; additional systematic surveys are needed to improve population estimates of L. catta. Nevertheless, our data are important for informing species-orientated conservation management (Kéry & Schmidt, Reference Kéry and Schmidt2008). Our findings are also a cautionary reminder that we must not confuse data deficiency with population deficiency; both survey extent and habitat heterogeneity (i.e. the ‘geographical template’ sensu Lomolino, Reference Lomolino, Lomolino and Heaney2004) must be considered when inferring species occupancy, as the absence of a species from one area might not be indicative of its absence from an entire region (Angermeier et al., Reference Angermeier, Krueger, Dolloff, Scott, Heglund, Morrison, Haufler, Raphael, Wall and Samson2002). Thus, it is important that additional surveys are conducted beyond the regions included in this study. Failure to do so could lead to so-called Wallacean shortfalls (i.e. gaps in understanding of the geographical distribution of a species; Lomolino, Reference Lomolino, Lomolino and Heaney2004). We therefore echo a recent call (LaFleur et al., Reference LaFleur, Gould, Sauther, Clarke and Reuter2018) for widespread collaboration between researchers, conservationists and park management officials to generate more comprehensive information on the density and distribution of L. catta throughout its geographical range, to inform management of the species appropriately.