Integrative Zoology 2015; 10: 389–402
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
doi: 10.1111/1749-4877.12146
ORIGINAL ARTICLE
Identifying ecological corridors for Amur tigers (Panthera tigris altaica) and Amur leopards (Panthera pardus orientalis) Dale G. MIQUELLE,1 Vyachaslav V. ROZHNOV,2 Victor ERMOSHIN,3 Andre A. MURZIN,3 Igor G. NIKOLAEV,4 Jose A. HERNANDEZ-BLANCO2 and Sergie V. NAIDENKO2 1
Wildlife Conservation Society, Bronx, New York, USA, 2A.N. Severtsov Institute of Ecology and Evolution, Russian Academy of Sciences, Moscow, Russia, 3Pacific Institute of Geography, Far Eastern Branch of the Russian Academy of Sciences, Vladivostok, Russia and 4The Institute of Biology and Soils, Far Eastern Branch of the Russian Academy of Sciences, Vladivostok, Russia
Abstract The rapid explosion of human populations and the associated development of human-dominated landscapes have drastically reduced and fragmented habitat for tigers (Panthera tigris) and leopards (Panthera pardus) across Asia, resulting in multiple small populations. However, Amur tiger (Panthera tigris altaica) habitat in Russia has remained largely interconnected, except for a break between tigers in southwest Primorye and the southern Sikhote-Alin Mountains. This habitat patch in southwest Primorye also retains the last population of Amur leopards (Panthera pardus orientalis). Genetic differentiation of tigers in southwest Primorye and the Sikhote-Alin Mountains along with survey data suggest that habitat fragmentation is limiting movement of tigers and leopards across the Razdolnaya River basin. We looked at historical and recent survey data on tigers and leopards and mapped existing cover types to examine land-use patterns of both large felids and humans in the development strip along the Razdolnaya River. We then used least-cost distance analyses to identify the most effective potential corridor to retain connectivity for large felids between Land of the Leopard National Park and Ussuriskii Zapovednik (Reserve). We identified a single potential corridor that still exists with a total distance of 62.5 km from Land of the Leopard National Park to Ussuriskii Zapovednik, mostly (93%) through forested habitat. We recommend formal recognition of a Razdolnaya ecological corridor and provide specific recommendations for each of 3 proposed management sections. Key words: Amur tiger, ecological corridor, Sikhote-Alin Mountains
INTRODUCTION
Correspondence: Dale G. Miquelle, Wildlife Conservation Society, 2300 Southern Boulevard, Bronx, NY 10460, USA. Email: [email protected]
The rapid explosion of human populations and the associated development of human-dominated landscapes have drastically reduced and fragmented habitat for tigers [Panthera tigris (Linnaeus, 1758)] and leopards (Panthera pardus Linnaeus, 1758) across Asia, resulting
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
389
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
in multiple small populations. Tigers now occupy 7% of their historical range across Asia (Dinerstein et al. 2007) in fragmented small populations. For instance, in the Indian subcontinent, often considered the stronghold of tigers, there are now an estimated 40 isolated landscapes that hold tigers, but the median potential population size of tigers in protected areas of the Indian subcontinent is only 14 (Ranganathan et al. 2008), and genetic differentiation in tiger populations from different sub-regions of the Indian subcontinent already exists (Mondol et al. 2013). While less is known about the impacts of human development on Asian leopards, we do know that some populations, such as that of the Amur leopard (Panthera pardus orientalis Schlegel, 1857) have become small and completely isolated (Miquelle et al. 1999; Hebblewhite et al. 2011). Such small populations are consequently susceptible to a range of threats; for example, the Amur leopard has the lowest genetic diversity of all extant subspecies (Uphyrkina et al. 2002) and is, there-
fore, considered endangered (Nowell & Jackson 1996). In contrast to most of Asia, Amur tiger (P. t. altaica Timminck, 1884) habitat in Russia has remained largely intact, with only an estimated 19% loss in the Sikhote-Alin Mountain ecosystem (Hebblewhite et al. 2014). There is an estimated 155 000 km2 of suitable habitat in Russia that consists of 17 “patches” (Hebblewhite et al. 2014), yet all but 2 are interconnected by continuous forested lands that easily allow movement of tigers and collectively retains a population estimated at 416–492 as of 2005 (Miquelle et al. 2006) that is genetically fully intermixed (Henry et al. 2009; Sorokin et al. 2015). The most significant isolated large tract of suitable habitat in Russia (approximately 3500 km2) exists in southwest Primorye, where a population of tigers and the last population of Amur leopards remain. This habitat patch is linked to the Changbaishan Mountain ecosystem of Jilin and Heilongjiang Provinces of China by
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Figure 1 Map of study area around the Razdolnaya River, showing major land-use categories, the divided highway, railway lines, 49 county boundaries and boundaries of the 2 closest protected areas, Land of the Leopard National Park (left side) and Ussuriskii 50 51 Zapovednik (right side).
390
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
Ecological corridor for Amur tigers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
an extensive forested boundary that allows movement of wildlife. In the Changbaishan Mountains in northeast China there are an estimated 25 000 km2 of potentially suitable tiger habitat (and likely even more suitable leopard habitat) and the opportunity exists for recovery of both species if these lands are properly managed (Hebblewhite et al. 2012). Hence, this habitat patch in southwest Primorye holds critical source populations for both tigers and leopards to recolonize the Changbaishan landscape. From the mouth of the Razdolnaya River to the city of Ussurisk, a development “strip” is growing, including the main highway (which has been upgraded) linking the major cities of the area, the Siberian railway, multiple villages and cites, as well as agricultural lands, multiple secondary roads and small business complexes (Fig. 1). Given that Amur leopards and tigers are known to disperse over hundreds of kilometers (Sludskii & Heptner 1992), it would seem that the distance between suitable forested habitats on each side of the Razdolnaya
River (in places less than 10 km) would be an insignificant barrier to movement. However, evidence already exists in 2 forms that the Razdolnaya development strip acts as a barrier to movement of large carnivores. First, based on microsatellite analyses, Henry et al. (2009) report, and Sorokin et al. (2015) confirm, that clear genetic differentiation exists between the tiger populations in southwest Primorye and Sikhote-Alin. The second piece of evidence is the failure of leopards to disperse across the Razdolnaya River into the Sikhote-Alin Mountains, based on yearly snow track surveys across protected areas and unprotected forest lands on the east side of the river (Miquelle et al. 2010), even though leopards formerly occurred across the southern Sikhote-Alin Mountains (Abramov & Pikunov 1974; Heptner & Sludskii 1992) and their current range extends nearly to the Razdolnaya River (Pikunov et al. 2009; Hebblewhite et al. 2011; Fig. 2). The existence of a subpopulation of Amur tigers that is already genetically distinguishable is significant, as
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 Figure 2 Tiger and leopard tracks reported along routes traveled in winter for surveys conducted from 1997 through 2007, and his- 50 51 toric records of tiger and leopard tracks (not on survey routes) within the study area since 1968.
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
391
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
the landscape appears to retain some permeability. Indeed, both Henry et al. (2009) and Sorokin et al. (this volume) identified some individuals recognizable as descendants from southwest Primorye in Sikhote-Alin Mountains, and vice versa. Hence, an ecological corridor presumably still exists. However, development along this Razdolnaya development strip is progressing. In recognition of this threat, plans for creating a corridor were incorporated into a regional conservation action plan (Darman & Williams 2003) more than a decade ago, and even into regional administration planning maps (http://www.urbanistika.ru/portfolio/zodchestvo-2008/primorskiy/). However, no concrete actions have been taken to initiate creation of such a corridor. With an increasing focus on development in this region, assessment of existing plans, along with delineation and formal creation of a wildlife corridor, is of utmost importance to retain connectivity between these 2 large tiger–leopard conservation landscapes. Various approaches have been used to identify potential ecological corridors (Hilty et al. 2006; Kindlmann & Burel 2008). Least-cost distance analyses that weigh the difficulty of moving through different cover types and then seek the easiest (least costly) route between 2 points have been used to identify links between large carnivore source sites (Wikramanayake et al. 2004; Zimmermann & Breitenmoser 2007; Chetkiewicz & Boyce 2009). More recently, circuit theory analyses, which make use of the intuitive analogy between movement of individuals through a landscape and resistance to movement of electricity, have been applied to map resistance to movement for large carnivores across a landscape (Walpole et al. 2012; Joshi et al. 2013). Least cost distance analyses will often provide a better solution when seeking connectivity between 2 points, while application of circuit theory allows a broader investigation of alternative routes and potential barriers to movement. In this study we seek to document the extent to which the regions on both sides of the development strip have been used by tigers and leopards by using historical records and more recent survey data. In addition, by looking at land cover, ownership and land-use information, we seek to assess “permeability” of the landscape types across the Razdolnaya development strip to identify the best movement corridor between protected areas on each side of the river. Finally, we consider how such a corridor might be protected from further development and what mitigations might be possible to increase the likelihood of movement by large carnivores (and other wildlife) across this region.
392
MATERIALS AND METHODS Study area We centered the study area on the Razdolnaya River, half way along the 55 km between the city of Ussurisk in the north and the mouth of the river in the south (Fig. 1). From that point we delineated a study area of 3025 km2 (55-km diameter) that includes the entire portion of the Razdolnaya River basin, considered a potential barrier to movement of large felids, as well as the outer edges of 2 protected areas (Land of the Leopard National Park to the west, Ussuriskii Zapovednik, or Reserve, to the east) that act as source sites for leopards and tigers in the region (Matyukina et al. 2011) and, therefore, represent appropriate starting and ending locations for least-cost distance analyses. We identified 7 land cover categories, including 28 settlements (villages 5000 residents): Artyom, Volno-Nadeshdenskoye, Razdolnoe and Ussurisk (Fig. 1). There are an abundance of “dacha” lands (cottages situated every 40–100 m interspersed with garden plots) that are mostly occupied only in summer. Wetlands (swamps of sedges and grasses) are extensive, especially towards the mouth of the river (Fig. 1), and agricultural fields (abandoned and still active) are common. Sparse forests (occasional forest patches or individual trees mixed with open meadows) included a mixture of deciduous trees such as Manchurian oak (Quercus mongolica) and Japanese white birch (Betula platyphylla). Riparian deciduous forests include Manchurian alder (Alnus hirsuta), Korean willow (Chosenia arbutifolia) and Maximovich poplar (Populus maximowiczii), and can range in width from only a few meters to several kilometers from the river. Upland secondary forests are a mixture of predominately Manchurian oak and Japanese white birch. Both tigers and leopards prefer a variety of forested habitats comprised of deciduous or mixed deciduous-coniferous species (Miquelle et al. 1999; Hebblewhite et al. 2011, 2014), and, therefore, all forest types (except sparse forests) were combined for analyses. Both species rarely occur in open habitats such as agricultural fields, wetlands and sparse forests (Hebblewhite et al. 2011, 2014). In addition to these land-cover types, the federal highway and the Siberian railway from Vladivostok to Khabarovsk run through the study area, as well as multiple secondary and tertiary roads for access to dacha lands and villages (Fig. 1).
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Ecological corridor for Amur tigers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Forest cover is generally more extensive and closer to the Razdolnaya River Basin on the west side of the river (Fig. 1). The only places where forests extend to the edge of the river basin on the east are at the mouth of the river (north of the Tavrichanka village) and again near Baranovskiy village in the north. Extensive wetlands of the Razdolnaya River basin are up to 8 km wide in the lower reaches of the river, but these wetlands narrow to approximately 5 km at the city of Razdolnoe. Between the villages of Baranovskiy and Partizan the river valley narrows much more dramatically, with a few patches of forest coming to the very edge of the river itself (Fig. 1). On the west side of the Razdolnaya River occasional villages, the primary road to the south and many secondary roads are the major anthropogenic features. Human development is concentrated on the east side of Razdolnaya River (Fig. 1). Dacha lands are extensive on the east side in the south, between the cities of Volno-Nadeshdenskoye and Razdolnoe, and occur in patches again near Partizan (Fig. 1). Razdolnoe city represents a narrow but long (11-km straight line) barrier to tiger movement in the central part of the study area, but further north there are no major developments until the city of Ussurisk at the northern end of the study area (Fig. 1). On the east side of the Razdolnaya River, Ussuriskii Zapovednik includes 408.7 km 2 of high quality tiger habitat where estimates of tiger numbers have ranged from 6 (Kostyria et al. 2010) to 14 individuals (Hernandez et al. 2013). On the west side of Razdolnaya River exists the newly created Land of the Leopard National Park (gazetted in April 2012) at 2799 km2, where estimates of tiger numbers have ranged from 12 (Sugimoto et al. 2012) to 27 (Aramilev & Aramilev 2013), and where most recent camera trap surveys estimated 56 individuals exist in the last remaining population of Amur leopards (Rybin et al., unpubl. data). Human density is 15 individuals/km2 across Primorskii and southern Khabarovskii Krais (Provinces) (the area where tigers occur) but is higher (approximately 22 individuals/km2) in the Razdolnaya study area.
Historical data and recent survey data From the late 1960s to the early 1990s, records of tiger occurrence near Razdolnaya River were recorded by one of the authors (I. G. Nikolaev). Each of these records was reviewed for veracity and authenticity through discussions with the source before inclusion in the database. Although the exact locations of all these
observations could not always be defined, locations on maps represent our “best estimate” based on descriptions by the observer (error likely + 3 km). While these records are no doubt incomplete, they provide some historical background on the presence of tigers along the Razdolnaya River. Over the past 20 years, 2 surveys of tigers have been conducted across the entirety of tiger range in Russia: 1 in the 1995–1996 winter (Matyushkin et al. 1996) and 1 in the 2004–2005 winter (Miquelle et al. 2007) by recording location of tracks in snow along routes placed at regular intervals to maximize the probability of encountering tracks (Hayward et al. 2001; Philips et al. 2005). We also summarize information on location of tiger and leopard tracks from 3 surveys in southwest Primorye during 2000, 2003 and 2007. For all 5 surveys, we report survey routes that were covered within the study area, and the location and numbers of tiger and leopard tracks reported for all these surveys combined. We hypothesized that habitat for tigers and leopards deteriorates closer to the river as the amount and quality of forested habitat declines and anthropogenic disturbances increase, and, therefore, divided survey routes and associated tracks into those close to the river (15 km). We summed routes and associated tracks into sampling units no less than 20 km in length to minimize zero records of tracks, and conducted simple t-tests to assess whether density of tiger and leopard tracks was greater closer to or further from the Razdolnaya River.
GIS coverage Seven land cover types were mapped manually using 4 Landsat 8 satellite images (from 18 September 2013, 4 October 2013, 5 November 2013 and 16 May 2014): wetlands, agricultural lands, small settlements (5000 inhabitants), dacha lands (cottages separated 40–100 m by garden plots), meadows/sparse forests (occasional forest patches or individual trees mixed with open grasslands) and forests (all types). Cells of 500 × 500 m were characterized by these 7 land cover types and the presence of roads (highways, primary and secondary) and railways was plotted.
Identifying potential ecological corridors. Both legislatively and politically, the best chance for re-establishing ecological connectivity between southwest Primorye and the Sikhote-Alin Mountains will be to argue for creation of a single ecological corridor be-
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
393
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
tween 2 legislatively recognized protected areas. Given these non-biological constraints, we chose to use the least cost-distance method, which is considered most appropriate when linking 2 clearly defined source sites. While we recognize that both tigers and leopards move beyond the boundaries of the protected areas, and that other analyses (e.g. applying circuit theory) would inform how animals might move across the entire landscape and what alternative corridors might exist, we focused on identifying the single best corridor that will likely have the greatest conservation impact. We developed an expert-based “friction” model with the help of the CostDistance Tool from the GIS program ArcGis 9.2 to determine the shortest “cost-distance” between cells by applying distance not in geographic units but in cost units. A cost-grid or so called “friction map” delineates impedance which represents the cost involved in moving through any particular cell. Chetkiewicz et al. (2006) recommend using resource selection functions (RSF) to identify areas likely used by animals as an alternative to expert-based friction models. While we used results of an RSF modeling effort (Hebblewhite et al. 2014) to inform the friction grid, the level of resolution used for the RSF (170 km2-sized sampling units) was much coarser than our grid system (500 × 500 m), preventing direct application of the RSF. Instead, we developed values based on expert opinion that were similar to those derived by Li et al. (2010) for a nearby landscape for tigers (Table 1) and, given the similarity of habitat preferences between species, assumed relative values would be similar for both tigers and leopards. Villages and cities were buffered with 2 concentric zones of 500 and 1000 m radius to indicate greater human impact and a greater reluctance of large carnivores to move through such areas. All cells falling in grids that included villages and cities were considered as insurmountable barriers for tigers and leopards and their value was set to 1000 (very high resistance). Values of 400 and 100 were given to cells falling into the 0 to 500 m and 500 to 1000 m distances from villages, respectively, reflecting greater probability of a tiger moving through such cells than through the settlements themselves. Low-use secondary roads were not considered a barrier to movements of tigers, as logging roads are commonly used throughout the Russian Far East as travel corridors by tigers (Matyushkin 1977). Secondary roads with greater traffic volume were given a value of 100, primary roads a value of 300 and the improved highway a value of 600. Based on extensive telemetry data (Goodrich et al. 2010), all forest types were considered to be equally passable for
394
tigers, but previous work (Hebblewhite et al. 2014) demonstrated that more open land covers are less preferred. Therefore, wetlands and meadows/sparse forest cover were give moderately higher friction values (50) while agricultural lands, which are more likely to be open or have lower height vegetative cover, received a greater friction value (200), representing the lesser probability that tigers will pass through them (compared, for instance, to a forest stand). However, probability of tigers using such cover types likely varies seasonally, dependent on the amount and height of vegetative growth, and recent ground fires (which are common in the area). Similarly, dacha lands, and even buffer zones close to settlements, are less often used in winter by humans. Therefore, to better capture the variable permeability of these cover types, we created 2 additional sets of friction values that decreased friction partially (e.g. from 200 to 50 for agricultural fields) or completely (i.e., friction values equivalent to forested cells (Table 1). We also reduced friction values for secondary roads and the railroad (from 100 to 50) to reflect variation in traffic loads. We conducted the same cost-distance analysis with these new values to determine if these changes produced alternative least-cost routes. We used the closest 2 points of the 2 protected areas in the region, Land of the Leopard National Park and Ussuriskii Zapovednik (approximately 50-km apart), as the starting/ending points for this analysis. However, because both these points were in the northern portions of the study area, selection of a least-cost corridor might be biased because the shortest distance between the 2 protected areas would be in the north. Because other starting points might produce different least-cost routes, we also initiated starting points at approximately 5-km intervals north and south of the initial starting points (Fig. 3) as a test of the relative value of the least-cost route selected with initial parameters.
RESULTS Historical data and recent survey data Eleven records of tigers were recorded in the study area in the vicinity of Razdolnaya River between 1968 and 1992, with 2 more occurring more recently (Table 2). These records include 3 reports of female tigers with cubs, all between 1968 and 1969. Based on pad size (Miquelle et al. 2015), it is likely that at least 3 more of these records were of female tigers, suggesting that 6 of 13 records were female tigers. Of the 13 records, 11 were in the 1960s, after which only 2 reports exist
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Ecological corridor for Amur tigers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Table 1 “Friction values” used to define cost for tigers and leopards to move through any given grid within the Razdolnaya River 1 basin, Primorskii Krai, Russia. The 3 sets of friction values attempt to capture the variability of cover type permeability with season 2 and fire events 3 4 Category Variable Friction value 1 Alternative friction values 2 Alternative friction values 3 5 Land cover 6 All forest types 1 1 1 7 8 Wetlands 50 1 1 9 Meadows/sparse forest cover 50 1 1 10 Agricultural fields 200 50 1 11 Settlements 12 1–500 m from settlements 400 400 400 13 14 500–1000 m from settlements 100 100 50 15 Village 1000 1000 1000 16 City 1000 1000 1000 17 Dacha lands 300 100 10 18 Roads 19 20 Forest (logging) roads 1 1 1 21 Secondary roads (gravel) 100 100 50 22 Primary roads 300 300 100 23 Highway 600 600 600 24 Railroad 100 50 50 25 26 27 28 29 in the early 1990s. Presence of cubs in 1968–1969 indieast of Alekseevka village, Fig. 2). The tiger was cap30 cates that this region was still suitable habitat for reprotured and later diagnosed as having canine distemper, 31 duction, but that quality of habitat was somehow comdying some 3 weeks later (Gilbert et al. 2015). Although 32 promised afterwards, with no further records of females tigers with distemper are known to lose their fear of hu33 with cubs, and almost no further records of tigers. mans, there is no evidence that they travel long distanc34 es, suggesting that at least some tigers do occur in close Fifty-five tiger tracks were reported from surveys 35 proximity to the Razdolnaya River. In addition, in 2012, within the study area on the west side of Razdolnaya 36 V. V. Gapanov reported that a male tiger had visited the River (Fig. 2). On the east side of the Razdolnaya River, 37 newly created Center for Rehabilitation and Reintroduc79 tiger tracks were reported during the 2 surveys with38 tion of Tigers and other Endangered Species (just east of in the study area. Tracks become increasingly common 39 the Aleseevka and close to the animal reported in 2010 approaching Ussuriskii Zapovednik and to the north of 40 above), apparently interested in a tigress that was being the Zapovednik (Fig. 2). 41 held there at the time. V. V. Gapanov (pers. comm.) also Although there were notable differences in the den42 indicated that reports of tigers are not uncommon in this sity of tiger tracks close to and far from the Razdolnaya 43 region, but most are poorly documented. These 2 reports River (density 15 km from river = 0.62 + 0.44 tracks/10 45 ly close to the city of Razdolnoe, despite the absence of km), the large confidence intervals resulted in no signif46 regular evidence. icant difference (t = 1.31, df = 52, P = 0.19). 47 A total of 94 leopard tracks were reported with48 Two recent records of tigers outside of formal surveys in this study area on the west side of Razdolnaya Riv49 are significant. On 9 February 2010 a young male tiger er based on 3 surveys conducted from 2000 to 2007. As 50 walked into the “Northern Wonders” hunting base (just with tigers, track densities of leopards were lower clos- 51
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
395
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
er to the river (0.64 + 0.48 tracks/10 km) versus further away (0.78 + 0.39 tracks/10 km), but not significantly so (t = 0.47, df = 39, P = 0.63). Leopard tracks were common in some transects that end in the open meadows and wetlands of the river basin, and leopard tracks were more commonly reported closer to the Razdolnaya River than tiger tracks (Fig. 2). Leopard tracks have not been reported on the east side of the Razdolnaya River for over 40 years, even though the distance from the closest leopard track on the west side was only 9 km from the nearest tiger track on the east side, suggesting that the river basin and associated anthropogenic influences represent a significant barrier to movement of leopards.
fluence permeability, and determined the least costly route between the protected areas on each side of the Razdolnaya River (Fig. 3). The least costly route with the original friction grid (#1 in Table 1) passed through the northern section of the Razdolnaya basin between the villages of Partizan and Baranovskiy (Fig. 3). Changes in the start and end points failed to alter the course of the least-cost pathway (Fig. 3). In this region the river valley narrows considerably, forests come down to the edge of the river on the east side, and there are scattered pockets of scrub forest on the west side that lead to more extensive forests a short distance beyond. The total distance along this route from one protected area to the other is 62.5 km. The vast majority (93%) of this least-cost route is in forest lands, with only 2.1 km in sparse forest/meadows, 1.1 km in agricultural lands and 0.8 km in wetlands. The most costly components of the corridor are close to the river (where forest habitats are rare) and the highway (Fig. 3).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Least cost-distance analysis 16 17 Using the “friction” values from Table 1, we gener18 ated 3 sets of “permeability” grids for the entire study 19 area, assuming changing vegetative conditions will in20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 ≤1.0 37 1.1–50 38 50.1–150 150.1–200 39 200.1–400 40 400.1–550 41 550.1–600 42 600.1–900 43 900.1–950 44 950.1–2100 45 46 47 48 Figure 3 Cost-distance grid based on land-use types (from Fig. 1) and associated “friction” costs (as estimated in Table 1, fric- 49 tion values from 1st column) and the least-cost pathways (with a 500-m buffer) between Land of the Leopard National Park and Us- 50 51 suriskii Zapovednik at six starting points on the east and west sides of the Razdolnaya River.
396
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
Ecological corridor for Amur tigers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Changes in the permeability values (friction costs #2 and #3 in Table 1) of more open habitats changed only a small portion of the corridor, shifting the pathway across the river several kilometers to the south of Baranovskiy village through wetlands, with the majority of the leastcost route remaining the same.
to the Razdolnaya River because few survey routes have been established in the regions closest to the river. Therefore, it is extremely difficult to assess which portions of the forested complexes occasionally (or regularly) contain tigers and leopards, and it is difficult to assess the relative value of different forest patches. Prior surveys demonstrate that leopards are at least occasionally no more than 5 km from the river on the west side, but, nonetheless, have never been reported on the east side. Tiger tracks were also reported within 5 km of the river itself. The cost-distance analysis and development of various friction grids provide an objective means of defining
1 2 3 4 5 6 7 8 DISCUSSION 9 10 Ecological delineation of a corridor 11 Relatively little information exists to clearly de12 fine how often and how close tigers and leopards come 13 14 15 16 17 Table 2 Historic and recent records of tigers within the study area surrounding Razdolnaya River, Primorskii Krai, Russia, in addi18 tion to formal surveys 19 # DD.MM.YYYY Sex-age Location Comments 20 21 1 9.1968 Female with 2 cubs Osinoviyi Creek Cub pad width = 6–7 cm 22 2 9.1968 Visual observation of tiger Barsukovka River 23 3 28.12.1968 Female with 3 cubs Pachekheza, Popovka, and Female was killed; cubs 24 Perevoznaya river basins caught near Tekutevemi. 25 Tracks also seen in 1967 26 4 1968–1969 Female with 2 cubs Popovka River Traveled in direction of 27 Ussuriskii Zapovednik 28 5 Male tiger approximately 5 Pachekheza, Popovka, and Tracks encountered 29 years old Perevoznaya river basins commonly 30 6 1968–1969 Tiger tracks with pad width Kamenushka and Volka Tracks encountered 31 12 cm (male) Rivers commonly 32 7 1968–1969 Tiger tracks with pad width 8 Kamenushka and Volkha Tracks encountered 33 cm (probably female) Rivers commonly, tiger cam from 34 Ussuriskii Reserve 35 8 1968–1969 Tiger tracks with pad width Kamenushka River, Shirokaya Tracks encountered 36 10–12 cm (male) Valley commonly 37 9 1963 Tiger was poached Upper reaches of Batalianza Poacher was fined 38 River 39 40 Depredation on domestic 10 11.1990 Tiger tracks with pad size = Nadezhdenskaya village 9 × 11, total track size = 15 × suburds, 100 m from a store livestock 41 16cm on the banks of Schmidtovka 42 River 43 11 21.11.1992 Fresh tiger track, pad width = “Left bank” of the Razdolnaya Unsuccessful hunting attempt 44 9.5 cm River near the village of on wild boar 45 Razdolnoe 46 12 09.02.2010 Young male tiger walked into Alekseevka, village, Tiger captured and later 47 hunting camp Nadezhdenskii County diagnosed with canine 48 distemper 49 13 winter 2012 Tracks of male tiger Visited Alekseevka Tiger Reported by V. V. Gaponov 50 Rehabilitation Center 51
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
397
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
the most likely potential ecological corridor. Of course, whether tigers and/or leopards have used or might in the future use this corridor is unknown. Nonetheless, the route defined with the primary friction grid likely represents the best option for preserving an ecological linkage between southwest Primorye and the southern Sikhote-Alin Mountains. The slight deviation of this route to the south with friction grids that increased permeability of open habitats (especially wetlands) would likely make crossing the corridor feasible only on a seasonal basis, when extensive growth of sedges and reeds would provide cover. These land-use types are commonly burned in fall and spring, and under such conditions tigers would be less likely to cross barren ground. In winter, after such wetlands are often burned, and when tigers are likely to cross the river on ice, this route would be much less likely used than the more northern route that provides greater cover year round. The area around Razdolnoe city represents another site where the river valley narrows, and good forests exist on both sides of the river, but, unfortunately, the intensively developed city itself, although narrow, is highly unlikely to be crossed by tigers. Similarly, near the mouth of the river, there are reasonable patches of forested habitat on both sides, but to reach the southern Sikhote-Alin Mountains tigers or leopards would have to pass through an extensive complex of dachas and villages (Fig. 1) that lay beyond the forested areas close to the river. Changes in the scale of the friction values for each cover type can also influence outcomes (Rayfield et al. 2010). However, these effects are most prominent with greater fragmentation. In our study area, habitat types are largely linear and parallel with the river itself (forests give way to open habitats close to the river on both sides), making the least-cost decision-making process largely dependent on the width of the more open habitats and absence/presence of human-dominated landscapes (cites, villages and dacha lands) along any given segment. Review of the cost grid (Fig. 3) reinforces the idea that the landscape is actually relatively simple in terms of obstacles to tiger and leopard movement, and even with changing scales the results are likely to be quite similar. Changes in the starting point (Fig. 3) failed to alter the route, providing additional support to the idea that there are few, if any, alternative routes for an ecological corridor. This objective identification of a potential corridor reinforces and is almost identical to the corridor proposed in earlier conservation plans (Darman & Williams 2003). This is the only place between Amur Bay and the
398
city of Ussuriski where the river valley narrows considerably, and hills rise directly from the very narrow river channel. Tracts of forests still come down to the river’s edge in some places, and broken scrub forests exist intermixed with abandoned agricultural lands and meadows. The villages of Baranovskiy and Partizan are very small, and are mostly comprised of dachas. Hence, the site represents the narrowest portion of the Razdolnaya River, and the site with the least human influence. Thus, the least-cost analysis largely reinforces what appears intuitively obvious to people familiar with the landscape.
Management issues surrounding a corridor The absence of any legislature at the federal level in Russia that allows for creation of ecological corridors represents an obstacle to formal recognition of this corridor. However, there is precedent for such legislature in nearby Khabarovskii Krai, which established a suite of ecological corridors (largely for tigers) based on regional legislature. Similarly, rationale for the creation of the Central-Ussuriskii Zakaznik in Primorskii Krai was largely to mirror the efforts of Khabarovskii Krai in securing connectivity along the Strelnikov Ridge, which represents the only possible linkage between Russian tigers and tigers in the Wandashan Mountains of China (Darman & Williams 2003; Miquelle & Zhang 2005). We recommend that management of the proposed Razdolnaya corridor be broken into 3 segments: (i) the relatively narrow section that represents the “core” corridor as it crosses the Razdolnaya River basin and the associated degraded habitat on both sides; (ii) the forest lands on each side beyond the central small corridor that provide linkage to protected areas; and (iii) the federal highway, which represents a significant obstacle to large carnivores (Fig. 4). Management guidelines for each of these 3 zones would be necessary. The core area is comprised of a mixture of private and county-owned lands that could be formally designated as a regional wildlife refuge or ecological corridor to provide some level of protection between the villages of Baranovskiy and Partizan (approximately 3 km in length and 10 km in width, possibly with more land to the western side, where forest loss is greater) (Fig. 4). This core zone would represent a very small plot of land: approximately 30 km2. Mountain lions (Puma concolor Linnaeus, 1771) in the USA have been known to pass through corridors as narrow as 100 m over short distances (
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
doi: 10.1111/1749-4877.12146
ORIGINAL ARTICLE
Identifying ecological corridors for Amur tigers (Panthera tigris altaica) and Amur leopards (Panthera pardus orientalis) Dale G. MIQUELLE,1 Vyachaslav V. ROZHNOV,2 Victor ERMOSHIN,3 Andre A. MURZIN,3 Igor G. NIKOLAEV,4 Jose A. HERNANDEZ-BLANCO2 and Sergie V. NAIDENKO2 1
Wildlife Conservation Society, Bronx, New York, USA, 2A.N. Severtsov Institute of Ecology and Evolution, Russian Academy of Sciences, Moscow, Russia, 3Pacific Institute of Geography, Far Eastern Branch of the Russian Academy of Sciences, Vladivostok, Russia and 4The Institute of Biology and Soils, Far Eastern Branch of the Russian Academy of Sciences, Vladivostok, Russia
Abstract The rapid explosion of human populations and the associated development of human-dominated landscapes have drastically reduced and fragmented habitat for tigers (Panthera tigris) and leopards (Panthera pardus) across Asia, resulting in multiple small populations. However, Amur tiger (Panthera tigris altaica) habitat in Russia has remained largely interconnected, except for a break between tigers in southwest Primorye and the southern Sikhote-Alin Mountains. This habitat patch in southwest Primorye also retains the last population of Amur leopards (Panthera pardus orientalis). Genetic differentiation of tigers in southwest Primorye and the Sikhote-Alin Mountains along with survey data suggest that habitat fragmentation is limiting movement of tigers and leopards across the Razdolnaya River basin. We looked at historical and recent survey data on tigers and leopards and mapped existing cover types to examine land-use patterns of both large felids and humans in the development strip along the Razdolnaya River. We then used least-cost distance analyses to identify the most effective potential corridor to retain connectivity for large felids between Land of the Leopard National Park and Ussuriskii Zapovednik (Reserve). We identified a single potential corridor that still exists with a total distance of 62.5 km from Land of the Leopard National Park to Ussuriskii Zapovednik, mostly (93%) through forested habitat. We recommend formal recognition of a Razdolnaya ecological corridor and provide specific recommendations for each of 3 proposed management sections. Key words: Amur tiger, ecological corridor, Sikhote-Alin Mountains
INTRODUCTION
Correspondence: Dale G. Miquelle, Wildlife Conservation Society, 2300 Southern Boulevard, Bronx, NY 10460, USA. Email: [email protected]
The rapid explosion of human populations and the associated development of human-dominated landscapes have drastically reduced and fragmented habitat for tigers [Panthera tigris (Linnaeus, 1758)] and leopards (Panthera pardus Linnaeus, 1758) across Asia, resulting
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
389
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
in multiple small populations. Tigers now occupy 7% of their historical range across Asia (Dinerstein et al. 2007) in fragmented small populations. For instance, in the Indian subcontinent, often considered the stronghold of tigers, there are now an estimated 40 isolated landscapes that hold tigers, but the median potential population size of tigers in protected areas of the Indian subcontinent is only 14 (Ranganathan et al. 2008), and genetic differentiation in tiger populations from different sub-regions of the Indian subcontinent already exists (Mondol et al. 2013). While less is known about the impacts of human development on Asian leopards, we do know that some populations, such as that of the Amur leopard (Panthera pardus orientalis Schlegel, 1857) have become small and completely isolated (Miquelle et al. 1999; Hebblewhite et al. 2011). Such small populations are consequently susceptible to a range of threats; for example, the Amur leopard has the lowest genetic diversity of all extant subspecies (Uphyrkina et al. 2002) and is, there-
fore, considered endangered (Nowell & Jackson 1996). In contrast to most of Asia, Amur tiger (P. t. altaica Timminck, 1884) habitat in Russia has remained largely intact, with only an estimated 19% loss in the Sikhote-Alin Mountain ecosystem (Hebblewhite et al. 2014). There is an estimated 155 000 km2 of suitable habitat in Russia that consists of 17 “patches” (Hebblewhite et al. 2014), yet all but 2 are interconnected by continuous forested lands that easily allow movement of tigers and collectively retains a population estimated at 416–492 as of 2005 (Miquelle et al. 2006) that is genetically fully intermixed (Henry et al. 2009; Sorokin et al. 2015). The most significant isolated large tract of suitable habitat in Russia (approximately 3500 km2) exists in southwest Primorye, where a population of tigers and the last population of Amur leopards remain. This habitat patch is linked to the Changbaishan Mountain ecosystem of Jilin and Heilongjiang Provinces of China by
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Figure 1 Map of study area around the Razdolnaya River, showing major land-use categories, the divided highway, railway lines, 49 county boundaries and boundaries of the 2 closest protected areas, Land of the Leopard National Park (left side) and Ussuriskii 50 51 Zapovednik (right side).
390
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
Ecological corridor for Amur tigers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
an extensive forested boundary that allows movement of wildlife. In the Changbaishan Mountains in northeast China there are an estimated 25 000 km2 of potentially suitable tiger habitat (and likely even more suitable leopard habitat) and the opportunity exists for recovery of both species if these lands are properly managed (Hebblewhite et al. 2012). Hence, this habitat patch in southwest Primorye holds critical source populations for both tigers and leopards to recolonize the Changbaishan landscape. From the mouth of the Razdolnaya River to the city of Ussurisk, a development “strip” is growing, including the main highway (which has been upgraded) linking the major cities of the area, the Siberian railway, multiple villages and cites, as well as agricultural lands, multiple secondary roads and small business complexes (Fig. 1). Given that Amur leopards and tigers are known to disperse over hundreds of kilometers (Sludskii & Heptner 1992), it would seem that the distance between suitable forested habitats on each side of the Razdolnaya
River (in places less than 10 km) would be an insignificant barrier to movement. However, evidence already exists in 2 forms that the Razdolnaya development strip acts as a barrier to movement of large carnivores. First, based on microsatellite analyses, Henry et al. (2009) report, and Sorokin et al. (2015) confirm, that clear genetic differentiation exists between the tiger populations in southwest Primorye and Sikhote-Alin. The second piece of evidence is the failure of leopards to disperse across the Razdolnaya River into the Sikhote-Alin Mountains, based on yearly snow track surveys across protected areas and unprotected forest lands on the east side of the river (Miquelle et al. 2010), even though leopards formerly occurred across the southern Sikhote-Alin Mountains (Abramov & Pikunov 1974; Heptner & Sludskii 1992) and their current range extends nearly to the Razdolnaya River (Pikunov et al. 2009; Hebblewhite et al. 2011; Fig. 2). The existence of a subpopulation of Amur tigers that is already genetically distinguishable is significant, as
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 Figure 2 Tiger and leopard tracks reported along routes traveled in winter for surveys conducted from 1997 through 2007, and his- 50 51 toric records of tiger and leopard tracks (not on survey routes) within the study area since 1968.
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
391
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
the landscape appears to retain some permeability. Indeed, both Henry et al. (2009) and Sorokin et al. (this volume) identified some individuals recognizable as descendants from southwest Primorye in Sikhote-Alin Mountains, and vice versa. Hence, an ecological corridor presumably still exists. However, development along this Razdolnaya development strip is progressing. In recognition of this threat, plans for creating a corridor were incorporated into a regional conservation action plan (Darman & Williams 2003) more than a decade ago, and even into regional administration planning maps (http://www.urbanistika.ru/portfolio/zodchestvo-2008/primorskiy/). However, no concrete actions have been taken to initiate creation of such a corridor. With an increasing focus on development in this region, assessment of existing plans, along with delineation and formal creation of a wildlife corridor, is of utmost importance to retain connectivity between these 2 large tiger–leopard conservation landscapes. Various approaches have been used to identify potential ecological corridors (Hilty et al. 2006; Kindlmann & Burel 2008). Least-cost distance analyses that weigh the difficulty of moving through different cover types and then seek the easiest (least costly) route between 2 points have been used to identify links between large carnivore source sites (Wikramanayake et al. 2004; Zimmermann & Breitenmoser 2007; Chetkiewicz & Boyce 2009). More recently, circuit theory analyses, which make use of the intuitive analogy between movement of individuals through a landscape and resistance to movement of electricity, have been applied to map resistance to movement for large carnivores across a landscape (Walpole et al. 2012; Joshi et al. 2013). Least cost distance analyses will often provide a better solution when seeking connectivity between 2 points, while application of circuit theory allows a broader investigation of alternative routes and potential barriers to movement. In this study we seek to document the extent to which the regions on both sides of the development strip have been used by tigers and leopards by using historical records and more recent survey data. In addition, by looking at land cover, ownership and land-use information, we seek to assess “permeability” of the landscape types across the Razdolnaya development strip to identify the best movement corridor between protected areas on each side of the river. Finally, we consider how such a corridor might be protected from further development and what mitigations might be possible to increase the likelihood of movement by large carnivores (and other wildlife) across this region.
392
MATERIALS AND METHODS Study area We centered the study area on the Razdolnaya River, half way along the 55 km between the city of Ussurisk in the north and the mouth of the river in the south (Fig. 1). From that point we delineated a study area of 3025 km2 (55-km diameter) that includes the entire portion of the Razdolnaya River basin, considered a potential barrier to movement of large felids, as well as the outer edges of 2 protected areas (Land of the Leopard National Park to the west, Ussuriskii Zapovednik, or Reserve, to the east) that act as source sites for leopards and tigers in the region (Matyukina et al. 2011) and, therefore, represent appropriate starting and ending locations for least-cost distance analyses. We identified 7 land cover categories, including 28 settlements (villages 5000 residents): Artyom, Volno-Nadeshdenskoye, Razdolnoe and Ussurisk (Fig. 1). There are an abundance of “dacha” lands (cottages situated every 40–100 m interspersed with garden plots) that are mostly occupied only in summer. Wetlands (swamps of sedges and grasses) are extensive, especially towards the mouth of the river (Fig. 1), and agricultural fields (abandoned and still active) are common. Sparse forests (occasional forest patches or individual trees mixed with open meadows) included a mixture of deciduous trees such as Manchurian oak (Quercus mongolica) and Japanese white birch (Betula platyphylla). Riparian deciduous forests include Manchurian alder (Alnus hirsuta), Korean willow (Chosenia arbutifolia) and Maximovich poplar (Populus maximowiczii), and can range in width from only a few meters to several kilometers from the river. Upland secondary forests are a mixture of predominately Manchurian oak and Japanese white birch. Both tigers and leopards prefer a variety of forested habitats comprised of deciduous or mixed deciduous-coniferous species (Miquelle et al. 1999; Hebblewhite et al. 2011, 2014), and, therefore, all forest types (except sparse forests) were combined for analyses. Both species rarely occur in open habitats such as agricultural fields, wetlands and sparse forests (Hebblewhite et al. 2011, 2014). In addition to these land-cover types, the federal highway and the Siberian railway from Vladivostok to Khabarovsk run through the study area, as well as multiple secondary and tertiary roads for access to dacha lands and villages (Fig. 1).
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Ecological corridor for Amur tigers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Forest cover is generally more extensive and closer to the Razdolnaya River Basin on the west side of the river (Fig. 1). The only places where forests extend to the edge of the river basin on the east are at the mouth of the river (north of the Tavrichanka village) and again near Baranovskiy village in the north. Extensive wetlands of the Razdolnaya River basin are up to 8 km wide in the lower reaches of the river, but these wetlands narrow to approximately 5 km at the city of Razdolnoe. Between the villages of Baranovskiy and Partizan the river valley narrows much more dramatically, with a few patches of forest coming to the very edge of the river itself (Fig. 1). On the west side of the Razdolnaya River occasional villages, the primary road to the south and many secondary roads are the major anthropogenic features. Human development is concentrated on the east side of Razdolnaya River (Fig. 1). Dacha lands are extensive on the east side in the south, between the cities of Volno-Nadeshdenskoye and Razdolnoe, and occur in patches again near Partizan (Fig. 1). Razdolnoe city represents a narrow but long (11-km straight line) barrier to tiger movement in the central part of the study area, but further north there are no major developments until the city of Ussurisk at the northern end of the study area (Fig. 1). On the east side of the Razdolnaya River, Ussuriskii Zapovednik includes 408.7 km 2 of high quality tiger habitat where estimates of tiger numbers have ranged from 6 (Kostyria et al. 2010) to 14 individuals (Hernandez et al. 2013). On the west side of Razdolnaya River exists the newly created Land of the Leopard National Park (gazetted in April 2012) at 2799 km2, where estimates of tiger numbers have ranged from 12 (Sugimoto et al. 2012) to 27 (Aramilev & Aramilev 2013), and where most recent camera trap surveys estimated 56 individuals exist in the last remaining population of Amur leopards (Rybin et al., unpubl. data). Human density is 15 individuals/km2 across Primorskii and southern Khabarovskii Krais (Provinces) (the area where tigers occur) but is higher (approximately 22 individuals/km2) in the Razdolnaya study area.
Historical data and recent survey data From the late 1960s to the early 1990s, records of tiger occurrence near Razdolnaya River were recorded by one of the authors (I. G. Nikolaev). Each of these records was reviewed for veracity and authenticity through discussions with the source before inclusion in the database. Although the exact locations of all these
observations could not always be defined, locations on maps represent our “best estimate” based on descriptions by the observer (error likely + 3 km). While these records are no doubt incomplete, they provide some historical background on the presence of tigers along the Razdolnaya River. Over the past 20 years, 2 surveys of tigers have been conducted across the entirety of tiger range in Russia: 1 in the 1995–1996 winter (Matyushkin et al. 1996) and 1 in the 2004–2005 winter (Miquelle et al. 2007) by recording location of tracks in snow along routes placed at regular intervals to maximize the probability of encountering tracks (Hayward et al. 2001; Philips et al. 2005). We also summarize information on location of tiger and leopard tracks from 3 surveys in southwest Primorye during 2000, 2003 and 2007. For all 5 surveys, we report survey routes that were covered within the study area, and the location and numbers of tiger and leopard tracks reported for all these surveys combined. We hypothesized that habitat for tigers and leopards deteriorates closer to the river as the amount and quality of forested habitat declines and anthropogenic disturbances increase, and, therefore, divided survey routes and associated tracks into those close to the river (15 km). We summed routes and associated tracks into sampling units no less than 20 km in length to minimize zero records of tracks, and conducted simple t-tests to assess whether density of tiger and leopard tracks was greater closer to or further from the Razdolnaya River.
GIS coverage Seven land cover types were mapped manually using 4 Landsat 8 satellite images (from 18 September 2013, 4 October 2013, 5 November 2013 and 16 May 2014): wetlands, agricultural lands, small settlements (5000 inhabitants), dacha lands (cottages separated 40–100 m by garden plots), meadows/sparse forests (occasional forest patches or individual trees mixed with open grasslands) and forests (all types). Cells of 500 × 500 m were characterized by these 7 land cover types and the presence of roads (highways, primary and secondary) and railways was plotted.
Identifying potential ecological corridors. Both legislatively and politically, the best chance for re-establishing ecological connectivity between southwest Primorye and the Sikhote-Alin Mountains will be to argue for creation of a single ecological corridor be-
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
393
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
tween 2 legislatively recognized protected areas. Given these non-biological constraints, we chose to use the least cost-distance method, which is considered most appropriate when linking 2 clearly defined source sites. While we recognize that both tigers and leopards move beyond the boundaries of the protected areas, and that other analyses (e.g. applying circuit theory) would inform how animals might move across the entire landscape and what alternative corridors might exist, we focused on identifying the single best corridor that will likely have the greatest conservation impact. We developed an expert-based “friction” model with the help of the CostDistance Tool from the GIS program ArcGis 9.2 to determine the shortest “cost-distance” between cells by applying distance not in geographic units but in cost units. A cost-grid or so called “friction map” delineates impedance which represents the cost involved in moving through any particular cell. Chetkiewicz et al. (2006) recommend using resource selection functions (RSF) to identify areas likely used by animals as an alternative to expert-based friction models. While we used results of an RSF modeling effort (Hebblewhite et al. 2014) to inform the friction grid, the level of resolution used for the RSF (170 km2-sized sampling units) was much coarser than our grid system (500 × 500 m), preventing direct application of the RSF. Instead, we developed values based on expert opinion that were similar to those derived by Li et al. (2010) for a nearby landscape for tigers (Table 1) and, given the similarity of habitat preferences between species, assumed relative values would be similar for both tigers and leopards. Villages and cities were buffered with 2 concentric zones of 500 and 1000 m radius to indicate greater human impact and a greater reluctance of large carnivores to move through such areas. All cells falling in grids that included villages and cities were considered as insurmountable barriers for tigers and leopards and their value was set to 1000 (very high resistance). Values of 400 and 100 were given to cells falling into the 0 to 500 m and 500 to 1000 m distances from villages, respectively, reflecting greater probability of a tiger moving through such cells than through the settlements themselves. Low-use secondary roads were not considered a barrier to movements of tigers, as logging roads are commonly used throughout the Russian Far East as travel corridors by tigers (Matyushkin 1977). Secondary roads with greater traffic volume were given a value of 100, primary roads a value of 300 and the improved highway a value of 600. Based on extensive telemetry data (Goodrich et al. 2010), all forest types were considered to be equally passable for
394
tigers, but previous work (Hebblewhite et al. 2014) demonstrated that more open land covers are less preferred. Therefore, wetlands and meadows/sparse forest cover were give moderately higher friction values (50) while agricultural lands, which are more likely to be open or have lower height vegetative cover, received a greater friction value (200), representing the lesser probability that tigers will pass through them (compared, for instance, to a forest stand). However, probability of tigers using such cover types likely varies seasonally, dependent on the amount and height of vegetative growth, and recent ground fires (which are common in the area). Similarly, dacha lands, and even buffer zones close to settlements, are less often used in winter by humans. Therefore, to better capture the variable permeability of these cover types, we created 2 additional sets of friction values that decreased friction partially (e.g. from 200 to 50 for agricultural fields) or completely (i.e., friction values equivalent to forested cells (Table 1). We also reduced friction values for secondary roads and the railroad (from 100 to 50) to reflect variation in traffic loads. We conducted the same cost-distance analysis with these new values to determine if these changes produced alternative least-cost routes. We used the closest 2 points of the 2 protected areas in the region, Land of the Leopard National Park and Ussuriskii Zapovednik (approximately 50-km apart), as the starting/ending points for this analysis. However, because both these points were in the northern portions of the study area, selection of a least-cost corridor might be biased because the shortest distance between the 2 protected areas would be in the north. Because other starting points might produce different least-cost routes, we also initiated starting points at approximately 5-km intervals north and south of the initial starting points (Fig. 3) as a test of the relative value of the least-cost route selected with initial parameters.
RESULTS Historical data and recent survey data Eleven records of tigers were recorded in the study area in the vicinity of Razdolnaya River between 1968 and 1992, with 2 more occurring more recently (Table 2). These records include 3 reports of female tigers with cubs, all between 1968 and 1969. Based on pad size (Miquelle et al. 2015), it is likely that at least 3 more of these records were of female tigers, suggesting that 6 of 13 records were female tigers. Of the 13 records, 11 were in the 1960s, after which only 2 reports exist
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Ecological corridor for Amur tigers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Table 1 “Friction values” used to define cost for tigers and leopards to move through any given grid within the Razdolnaya River 1 basin, Primorskii Krai, Russia. The 3 sets of friction values attempt to capture the variability of cover type permeability with season 2 and fire events 3 4 Category Variable Friction value 1 Alternative friction values 2 Alternative friction values 3 5 Land cover 6 All forest types 1 1 1 7 8 Wetlands 50 1 1 9 Meadows/sparse forest cover 50 1 1 10 Agricultural fields 200 50 1 11 Settlements 12 1–500 m from settlements 400 400 400 13 14 500–1000 m from settlements 100 100 50 15 Village 1000 1000 1000 16 City 1000 1000 1000 17 Dacha lands 300 100 10 18 Roads 19 20 Forest (logging) roads 1 1 1 21 Secondary roads (gravel) 100 100 50 22 Primary roads 300 300 100 23 Highway 600 600 600 24 Railroad 100 50 50 25 26 27 28 29 in the early 1990s. Presence of cubs in 1968–1969 indieast of Alekseevka village, Fig. 2). The tiger was cap30 cates that this region was still suitable habitat for reprotured and later diagnosed as having canine distemper, 31 duction, but that quality of habitat was somehow comdying some 3 weeks later (Gilbert et al. 2015). Although 32 promised afterwards, with no further records of females tigers with distemper are known to lose their fear of hu33 with cubs, and almost no further records of tigers. mans, there is no evidence that they travel long distanc34 es, suggesting that at least some tigers do occur in close Fifty-five tiger tracks were reported from surveys 35 proximity to the Razdolnaya River. In addition, in 2012, within the study area on the west side of Razdolnaya 36 V. V. Gapanov reported that a male tiger had visited the River (Fig. 2). On the east side of the Razdolnaya River, 37 newly created Center for Rehabilitation and Reintroduc79 tiger tracks were reported during the 2 surveys with38 tion of Tigers and other Endangered Species (just east of in the study area. Tracks become increasingly common 39 the Aleseevka and close to the animal reported in 2010 approaching Ussuriskii Zapovednik and to the north of 40 above), apparently interested in a tigress that was being the Zapovednik (Fig. 2). 41 held there at the time. V. V. Gapanov (pers. comm.) also Although there were notable differences in the den42 indicated that reports of tigers are not uncommon in this sity of tiger tracks close to and far from the Razdolnaya 43 region, but most are poorly documented. These 2 reports River (density 15 km from river = 0.62 + 0.44 tracks/10 45 ly close to the city of Razdolnoe, despite the absence of km), the large confidence intervals resulted in no signif46 regular evidence. icant difference (t = 1.31, df = 52, P = 0.19). 47 A total of 94 leopard tracks were reported with48 Two recent records of tigers outside of formal surveys in this study area on the west side of Razdolnaya Riv49 are significant. On 9 February 2010 a young male tiger er based on 3 surveys conducted from 2000 to 2007. As 50 walked into the “Northern Wonders” hunting base (just with tigers, track densities of leopards were lower clos- 51
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
395
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
er to the river (0.64 + 0.48 tracks/10 km) versus further away (0.78 + 0.39 tracks/10 km), but not significantly so (t = 0.47, df = 39, P = 0.63). Leopard tracks were common in some transects that end in the open meadows and wetlands of the river basin, and leopard tracks were more commonly reported closer to the Razdolnaya River than tiger tracks (Fig. 2). Leopard tracks have not been reported on the east side of the Razdolnaya River for over 40 years, even though the distance from the closest leopard track on the west side was only 9 km from the nearest tiger track on the east side, suggesting that the river basin and associated anthropogenic influences represent a significant barrier to movement of leopards.
fluence permeability, and determined the least costly route between the protected areas on each side of the Razdolnaya River (Fig. 3). The least costly route with the original friction grid (#1 in Table 1) passed through the northern section of the Razdolnaya basin between the villages of Partizan and Baranovskiy (Fig. 3). Changes in the start and end points failed to alter the course of the least-cost pathway (Fig. 3). In this region the river valley narrows considerably, forests come down to the edge of the river on the east side, and there are scattered pockets of scrub forest on the west side that lead to more extensive forests a short distance beyond. The total distance along this route from one protected area to the other is 62.5 km. The vast majority (93%) of this least-cost route is in forest lands, with only 2.1 km in sparse forest/meadows, 1.1 km in agricultural lands and 0.8 km in wetlands. The most costly components of the corridor are close to the river (where forest habitats are rare) and the highway (Fig. 3).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Least cost-distance analysis 16 17 Using the “friction” values from Table 1, we gener18 ated 3 sets of “permeability” grids for the entire study 19 area, assuming changing vegetative conditions will in20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 ≤1.0 37 1.1–50 38 50.1–150 150.1–200 39 200.1–400 40 400.1–550 41 550.1–600 42 600.1–900 43 900.1–950 44 950.1–2100 45 46 47 48 Figure 3 Cost-distance grid based on land-use types (from Fig. 1) and associated “friction” costs (as estimated in Table 1, fric- 49 tion values from 1st column) and the least-cost pathways (with a 500-m buffer) between Land of the Leopard National Park and Us- 50 51 suriskii Zapovednik at six starting points on the east and west sides of the Razdolnaya River.
396
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
Ecological corridor for Amur tigers
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
Changes in the permeability values (friction costs #2 and #3 in Table 1) of more open habitats changed only a small portion of the corridor, shifting the pathway across the river several kilometers to the south of Baranovskiy village through wetlands, with the majority of the leastcost route remaining the same.
to the Razdolnaya River because few survey routes have been established in the regions closest to the river. Therefore, it is extremely difficult to assess which portions of the forested complexes occasionally (or regularly) contain tigers and leopards, and it is difficult to assess the relative value of different forest patches. Prior surveys demonstrate that leopards are at least occasionally no more than 5 km from the river on the west side, but, nonetheless, have never been reported on the east side. Tiger tracks were also reported within 5 km of the river itself. The cost-distance analysis and development of various friction grids provide an objective means of defining
1 2 3 4 5 6 7 8 DISCUSSION 9 10 Ecological delineation of a corridor 11 Relatively little information exists to clearly de12 fine how often and how close tigers and leopards come 13 14 15 16 17 Table 2 Historic and recent records of tigers within the study area surrounding Razdolnaya River, Primorskii Krai, Russia, in addi18 tion to formal surveys 19 # DD.MM.YYYY Sex-age Location Comments 20 21 1 9.1968 Female with 2 cubs Osinoviyi Creek Cub pad width = 6–7 cm 22 2 9.1968 Visual observation of tiger Barsukovka River 23 3 28.12.1968 Female with 3 cubs Pachekheza, Popovka, and Female was killed; cubs 24 Perevoznaya river basins caught near Tekutevemi. 25 Tracks also seen in 1967 26 4 1968–1969 Female with 2 cubs Popovka River Traveled in direction of 27 Ussuriskii Zapovednik 28 5 Male tiger approximately 5 Pachekheza, Popovka, and Tracks encountered 29 years old Perevoznaya river basins commonly 30 6 1968–1969 Tiger tracks with pad width Kamenushka and Volka Tracks encountered 31 12 cm (male) Rivers commonly 32 7 1968–1969 Tiger tracks with pad width 8 Kamenushka and Volkha Tracks encountered 33 cm (probably female) Rivers commonly, tiger cam from 34 Ussuriskii Reserve 35 8 1968–1969 Tiger tracks with pad width Kamenushka River, Shirokaya Tracks encountered 36 10–12 cm (male) Valley commonly 37 9 1963 Tiger was poached Upper reaches of Batalianza Poacher was fined 38 River 39 40 Depredation on domestic 10 11.1990 Tiger tracks with pad size = Nadezhdenskaya village 9 × 11, total track size = 15 × suburds, 100 m from a store livestock 41 16cm on the banks of Schmidtovka 42 River 43 11 21.11.1992 Fresh tiger track, pad width = “Left bank” of the Razdolnaya Unsuccessful hunting attempt 44 9.5 cm River near the village of on wild boar 45 Razdolnoe 46 12 09.02.2010 Young male tiger walked into Alekseevka, village, Tiger captured and later 47 hunting camp Nadezhdenskii County diagnosed with canine 48 distemper 49 13 winter 2012 Tracks of male tiger Visited Alekseevka Tiger Reported by V. V. Gaponov 50 Rehabilitation Center 51
© 2015 International Society of Zoological Sciences, Institute of Zoology/ Chinese Academy of Sciences and Wiley Publishing Asia Pty Ltd
397
D. G. Miquelle et al.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
the most likely potential ecological corridor. Of course, whether tigers and/or leopards have used or might in the future use this corridor is unknown. Nonetheless, the route defined with the primary friction grid likely represents the best option for preserving an ecological linkage between southwest Primorye and the southern Sikhote-Alin Mountains. The slight deviation of this route to the south with friction grids that increased permeability of open habitats (especially wetlands) would likely make crossing the corridor feasible only on a seasonal basis, when extensive growth of sedges and reeds would provide cover. These land-use types are commonly burned in fall and spring, and under such conditions tigers would be less likely to cross barren ground. In winter, after such wetlands are often burned, and when tigers are likely to cross the river on ice, this route would be much less likely used than the more northern route that provides greater cover year round. The area around Razdolnoe city represents another site where the river valley narrows, and good forests exist on both sides of the river, but, unfortunately, the intensively developed city itself, although narrow, is highly unlikely to be crossed by tigers. Similarly, near the mouth of the river, there are reasonable patches of forested habitat on both sides, but to reach the southern Sikhote-Alin Mountains tigers or leopards would have to pass through an extensive complex of dachas and villages (Fig. 1) that lay beyond the forested areas close to the river. Changes in the scale of the friction values for each cover type can also influence outcomes (Rayfield et al. 2010). However, these effects are most prominent with greater fragmentation. In our study area, habitat types are largely linear and parallel with the river itself (forests give way to open habitats close to the river on both sides), making the least-cost decision-making process largely dependent on the width of the more open habitats and absence/presence of human-dominated landscapes (cites, villages and dacha lands) along any given segment. Review of the cost grid (Fig. 3) reinforces the idea that the landscape is actually relatively simple in terms of obstacles to tiger and leopard movement, and even with changing scales the results are likely to be quite similar. Changes in the starting point (Fig. 3) failed to alter the route, providing additional support to the idea that there are few, if any, alternative routes for an ecological corridor. This objective identification of a potential corridor reinforces and is almost identical to the corridor proposed in earlier conservation plans (Darman & Williams 2003). This is the only place between Amur Bay and the
398
city of Ussuriski where the river valley narrows considerably, and hills rise directly from the very narrow river channel. Tracts of forests still come down to the river’s edge in some places, and broken scrub forests exist intermixed with abandoned agricultural lands and meadows. The villages of Baranovskiy and Partizan are very small, and are mostly comprised of dachas. Hence, the site represents the narrowest portion of the Razdolnaya River, and the site with the least human influence. Thus, the least-cost analysis largely reinforces what appears intuitively obvious to people familiar with the landscape.
Management issues surrounding a corridor The absence of any legislature at the federal level in Russia that allows for creation of ecological corridors represents an obstacle to formal recognition of this corridor. However, there is precedent for such legislature in nearby Khabarovskii Krai, which established a suite of ecological corridors (largely for tigers) based on regional legislature. Similarly, rationale for the creation of the Central-Ussuriskii Zakaznik in Primorskii Krai was largely to mirror the efforts of Khabarovskii Krai in securing connectivity along the Strelnikov Ridge, which represents the only possible linkage between Russian tigers and tigers in the Wandashan Mountains of China (Darman & Williams 2003; Miquelle & Zhang 2005). We recommend that management of the proposed Razdolnaya corridor be broken into 3 segments: (i) the relatively narrow section that represents the “core” corridor as it crosses the Razdolnaya River basin and the associated degraded habitat on both sides; (ii) the forest lands on each side beyond the central small corridor that provide linkage to protected areas; and (iii) the federal highway, which represents a significant obstacle to large carnivores (Fig. 4). Management guidelines for each of these 3 zones would be necessary. The core area is comprised of a mixture of private and county-owned lands that could be formally designated as a regional wildlife refuge or ecological corridor to provide some level of protection between the villages of Baranovskiy and Partizan (approximately 3 km in length and 10 km in width, possibly with more land to the western side, where forest loss is greater) (Fig. 4). This core zone would represent a very small plot of land: approximately 30 km2. Mountain lions (Puma concolor Linnaeus, 1771) in the USA have been known to pass through corridors as narrow as 100 m over short distances (
