Predicting potential habitat and population size for reintroduction of the Far Eastern leopards in the Russian Far East
Highlights
► The Far Eastern Leopard is perhaps the world’s most endangered large cat subspecies. ► We identified habitat for reintroductions and potential population size. ► Leopards required high ungulate densities and low human activity. ► We identified >10,000 km2 of potential leopard habitat for 105 adult leopards. ► Reintroduction into high quality habitat is required for recovery.
Introduction
Restoration of large carnivores is recognized as an important strategy for biodiversity conservation (Ray et al., 2005), as an indicator of landscape-scale conservation success (Linnell et al., 2000, Carroll et al., 2001), and because of their important trophic role (Berger et al., 2001, Hebblewhite et al., 2005). As a result, policies towards large carnivores in many areas of the world have shifted away from active persecution (Musiani and Paquet, 2004). This paradigm shift has enabled carnivores to recolonize parts of their historic ranges, e.g., wolves (Canis lupus) in the European Alps (Breitenmoser, 1998, Marucco and McIntire, 2010), and the northwestern USA (Pletscher et al., 1997). However, for many endangered carnivores, favorable policies alone have not been enough, and active reintroductions have been required (Breitenmoser et al., 2001). Reintroductions have established successful populations where significant dispersal barriers or human persecution still exist, such as Lynx (Lynx canadensis) to Colorado (Devineau et al., 2010) and the Swiss Alps (L. lynx)(Breitenmoser, 1998), and lions (Panthera leo) and other carnivores in Africa (Hayward et al., 2007a, Hunter et al., 2007). Given declining trends for many carnivores, more reintroductions will likely be required in the 21st century (Weber and Rabinowitz, 1996, Dalerum et al., 2009).
The Far Eastern, or Amur leopard (Panthera pardus orientalis; Schlegel 1857) is probably the most endangered large felid subspecies in the world, and has been listed as critically endangered on the IUCN red list since 1996 (Jackson and Nowell, 2008). Originally distributed broadly in the southern-most portions of the Russian Far East, Northeast China and much of the Korean peninsula (Pocock, 1930, Nowell and Jackson, 1996), fewer than 10 leopards likely exist in China today (Yang et al., 1998), and the main population occurs in southwestern Primorski Krai (Province) in the Russian Far East (Fig. 1). In 1970 there were no more than 46 individuals in three isolated populations in southern Primorski Krai (Fig. 1) (Abramov and Pikunov, 1974). By 1985, leopards had disappeared from the two northern sites (Pikunov and Korkishko, 1985), and since then, a small population of 25–35 leopards has persisted in Southwest Primorski Krai (Fig. 1, Pikunov et al., 2009). This isolated population of leopards is faced with imminent extinction from continued poaching, illegal harvest of prey, habitat fragmentation, and inbreeding depression (Miththapala et al., 1996, Uphyrkina et al., 2002). It is critical that numbers of leopards increase to reduce extinction risk. However, given that natural expansion of this population seems unlikely because of high levels of agricultural and infrastructure development surrounding the population in Russia and China, reintroduction of individuals to form additional populations appears to be the best option for increasing numbers in the wild.
Quantitative assessments of reintroduction programs for mammals and birds (Griffith et al., 1989) indicate that one of the primary determinants of success is habitat quality of the release area. Various modeling approaches have been employed to identify potential habitat and colonization patterns of carnivores (Schadt et al., 2002a, Schadt et al., 2002b, Cianfrani et al., 2010). For many endangered carnivores, vegetation-based landscape covariates used in many habitat models are insufficient for two reasons. First, human-caused mortality – a parameter not easily tied to landscape covariates – is a primary influence on population persistence for many large carnivore populations (Chapron et al., 2008, Goodrich et al., 2008). Second, in the absence of high human-caused mortality, carnivore densities are correlated with preferred prey density (Karanth et al., 2004; Hayward et al., 2007a, Hayward et al., 2007b, Miquelle et al., 2010a). Therefore, to accurately identify potential habitat for reintroductions, it will be necessary to include anthropogenic influences as well as measures of prey abundance (Marucco and McIntire, 2010).
Our goal was to identify and prioritize potential habitat for reintroduction of Far Eastern leopards to their historical range in Russia, as requested by Russian authorities as part of the recovery planning process. We used extensive winter track surveys in the current range of leopards to develop a resource selection function (RSF, Boyce and McDonald, 1999) that defined the relative probability of leopard occurrence as a function of key resources. We compared models that included estimates of the probability of prey occurrence versus those based on landscape covariates, and also included anthropogenic covariates that relate to human-caused mortality of carnivores in the Russian Far East (Goodrich et al., 2008, Kerley et al., 2002). Next, we extrapolated this RSF to the historic range of Far Eastern leopards to identify best potential reintroduction sites. Using a habitat-based ratio estimator (Boyce and Waller, 2003) and estimates of the current population size, we then predicted leopard population sizes for potential habitat patches. Finally, we used the RSF models to assess connectivity of these patches with a least-cost path analysis (Chetkiewicz and Boyce, 2009). Habitat patch size, potential population size, and connectivity of patches were used as criteria for identifying priority areas for initiating a reintroduction program.
Section snippets
Study area
RSF models were first derived for the current range of leopards in Southwest Primorski Krai (current range), and then applied across their historic range in Southern Primorski Krai (historic range, Fig. 1). Their current range encompasses ∼7000 km2 along the eastern slopes of the East Manchurian Mountains on the border with China. Human activity is concentrated in low elevation agricultural river valleys, along main roads and railways, and near the main cities of Vladivostok (>700,000 people)
Leopard and ungulate track surveys
Six snow track surveys of leopards have been conducted in Southwest Primorski Krai between 1997 and 2005 (Pikunov et al., 1997; Aramilev et al. 1998; Pikunov et al., 2000, Pikunov et al., 2009, Pikunov et al., 2003, Aramilev and Fomenko, 2000). Snow track surveys were conducted shortly after new snowfall to eliminate double counting (e.g., Jedrzejewska and Jedrzejewski, 1998). A stratification process eliminated areas and vegetation types where the probability of leopards was very low (e.g.
Ungulate prey models
In winter all three ungulate species consistently selected oak forests, avoided areas with high road density, and showed weaker selection for areas with low snow cover and lower elevations (Table 2). Sika deer selected birch, Korean pine and regenerating forests, whereas roe deer selected shrub fields, and wild boar selected Korean pine while avoiding agricultural areas (Table 2). Sika deer were more common in Southeast Sikhote-Alin (especially coastal areas) and Southwest Primorski (current
Discussion
Based on the habitat relationships of the sole remaining population of Far Eastern leopards, we identified >10,000 km2 of potential leopard habitat divided into seven large habitat patches within their historic range in the southern Sikkhote-Alin in the Russian Far East. These areas could harbor a potential population size of 105.3 (57.9–147.2) additional adult Far Eastern leopards, significantly increasing the population size and geographic distribution of this critically endangered species.
Conclusion
Our results paint a potentially optimistic future for recovery of Far Eastern leopards, with reintroduction efforts capable of creating a population in the Russian Far East that could be 3–4 times greater than the existent population, with over 10,000 km2 of potential habitat in their historic range in the southeastern Sikhote-Alin Mountains. The potential habitat for an additional 105 adult leopards, one in a single population of ∼70 individuals centered on Lazovski Zapovednik, dramatically
Acknowledgements
This study was funded by NASA through a grant from the University of Virginia to the Wildlife Conservation Society and the University of Montana. Thanks to Dr. H. Shugart and N. Sherman for their cooperative effort. Support for surveys was provided by The Wildlife Conservation Society, World wide Fund for Nature, Tigris, Amur, and Zoological Society of London. Many people assisted in data collection during leopard surveys, including the following coordinators: V.K. Abramov, A.I. Belov, P.V.
References (78)
- et al.
Relating populations to habitats using resource selection functions
Trends in ecology and evolution
(1999) - et al.
Evaluating resource selection functions
Ecological Modelling
(2002) Large predators in the alps: the fall and rise of man’s competitors
Biological Conservation
(1998)- et al.
A habitat-based framework for grizzly bear conservation in Alberta
Biological Conservation
(2006) - et al.
Leopards in the far east of the USSR and their protection
Bull. Moscow Soc. for Nature – Amateur. Sec. Biol.
(1974) - Aramilev, S.V., 2010. Distribution and Other Aspects of the Ecology of Sika Deer (Cervus nippon hortulorum [Swinhoe,...
- Aramilev, V.V., Fomenko, P.V., 2000. Simultaneous Survey of Far Eastern leopards and Amur tigers in Southwest Primorski...
- et al.
A 1998 survey of leopards
Zov Taiga
(1999) - et al.
Feasibility of translocating panthers into northern Florida
Journal of Wildlife Management
(1993) - et al.
A mammalian predator–prey imbalance. Grizzly bear and wolf extinction affect avian neotropical migrants
Ecological Applications
(2001)