Golestan National Park Zoning Re-Evaluation Based on Higher Species Richness Hotspots

Archive of SID Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) (http://jwb.araku.ac.ir/) DOI: 10.22120/jwb.2019.105061.1059 Research Article

Golestan National Park zoning re-evaluation based on higher species richness hotspots

and Gap Analysis methods. The findings show 1 2 Javad Selyari , Azam Elhami Rad *, that though the park’s conservative zones do 3 4 Morteza Naderi , Kamran Almasieh not adequately overlap with suitable habitats of some species such as gazelle and wolf, 1Department of Environmental Sciences, Faculty of Natural Resources and Environment Islamic fortunately, these zones cover a significant Azad University, Science and Research Branch, proportion of the park’s species richness Tehran, Iran hotspots. Certainly, modification and extension 2*Faculty of Geography and Environmental of the conservative coverage area of zones 1 Sciences, Hakim Sabzevari University, Iran and 2 to the park’s areas of species richness 3 Department of Environmental Sciences, Faculty hotspots containing a richness of 4 to 9 species, of Agriculture and Natural Resources, Arak also modifying and extending the limits of University, Iran these zones in favor of species whose habitat 4Department of Nature Engineering, Agricultural Sciences and Natural Resources University of suitability has lower overlap with zones 1 and 2 Khuzestan, Mollasani, Iran will lead to more efficient conservation in *[email protected] Golestan National Park; therefore, as a result of email: [email protected] revising the zoning of Golestan National Park, Received: 11 March 2019 / Revised: 21 May 2019 / Accepted: 22 this goal can be approached in near future. May 2019 / Published online: 22 May 2019. Ministry of Sciences, Research and Technology, Arak University, Iran. Keywords: Gap Analysis, habitat modeling, Abstract Park designing; Northern Iran forests. In recent decades, the rate of extinction of Introduction species has been increased continually all over Conversion, fragmentation and destruction of the world and lack of effective and efficient habitat threaten the sustainable survival of conservation strategies will lead to extinction wildlife, particularly carnivores, in a way that of a large number of species. Determination of today protection of wildlife populations has habitat suitability is an essential necessity for become necessary and inevitable. Although management and conservation of wildlife. preservation and protection of remaining Suitable habitats play a crucial role in the populations in natural ecosystems is one of the survival and reproduction of species; therefore, most efficient means of protecting endangered paying more attention to these areas in wildlife species (Groves 2003, Rosenzweig 2003), management and conservation will result in However several studies show that current more effective conservation. In this study, the protected areas often fail to reflect the full locations of presence of large carnivores and range of a region’s biodiversity and their herbivores in Golestan National Park have been coverage and inclusion level is lower than the recorded during 2014- 2015 to predict the desired global conservation goals. (Rodrigues presence of the species and determine the et al. 1999, Margules and Pressey 2000). In environmental parameters affecting their fact, in the majority of cases, socio-economic, presence, also to specify species richness aesthetic, or political criteria were given more hotspots in this area. For this purpose, the weight in establishment of protected areas and modeling was performed by using the MaxEnt www.SID.ir

2 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID this has reduced the protective value of these highway), the rise in poaching, intentional and areas in many cases and there is an increasing accidental fires, grazing, and removal of necessity for protection of regions which are shrubs, and opportunistic exploitation of the capable to provide a high level of protection for park (Hasanzadeh Kiabi et al. 1993) and its living organisms in an area. surrounding buffer zones has led to a severe Furthermore, evaluating the efficiency of decline in the populations of the dominant current protected areas is a critical component carnivores and herbivores in the park. of biodiversity conservation. In fact, in the first The zoning of the park was performed in 2002 place, conservation planning and its (Dehdar Dargahi & Makhdum 2002). The performance evaluation will be possible by zoning divided the park into several zones identification of areas containing the highest including a safe zone (zone 1), a protective biodiversity (Maiorano et al. 2006, Brito et al. zone (zone 2), an extensive recreation zone, 2009). However, it is challenging to manage and an intensive recreation zone, with 49.43%, biodiversity when in many cases, the spatial 35.05%, 4.58%, and 0.54% of the area of the distribution of species is hardly known. park allocated to them, respectively. 9.91% of For this reason, species distribution models the park was set aside for restoration, 0.11% serve as essential tools for management and was devoted to special applications, and 0.38% conservation of biodiversity. Through using was established for historical-cultural use. SDMs, field data and species’ presence data, Lastly, 5.69% of GNP’s area was assigned as a quantitative estimations of species’ potential buffer zone, along borders of the safe zone and habitat and comparison of several areas in the protected zone (Dehdar Dargahi & terms of their potential for presence of wildlife Makhdum 2002). Regardless of the criteria that species can be provided. were used in primary zoning of GNP, the On the other hand, having knowledge about criterion of habitat suitability has not been factors which influence wildlife habitat taken into account in selection of several other selection is vital for proper management of protected areas. Certainly GNP has not been an species. Suitable habitat has substantial effect exception in this regard, and accordingly there on survival and reproduction of wildlife and is is an increasing necessity for identification of thus given considerable attention. regions which are capable to provide a high Studying wildlife on large scale is difficult, or level of protection for the most of living in many cases impossible. Therefore, the organisms of an area. models which are able to link the geographical With more than 150 species, small mammals distribution of species and communities to their constitute the majority of mammals in Iran. environment are needed to be used. Habitat However, most studies with a conservation or suitability models can calculate the probability ecological approach have been focused on large of a species occurrence in a particular location mammals (Farhadinia et al. 2010) because that based on environmental conditions. Established large mammals indeed act as umbrella species more than 50 years ago, Golestan National Park in the ecosystem. (GNP) and Biosphere Reserve is the first Similarly, this study aims to investigate the national park in Iran and the Middle East, distribution of large carnivores and herbivores ranking among the top 10 national parks in the in GNP, assess the degree of protection for world in terms of biodiversity (encompassing these species, and perform gap analysis to 1360 species of plants, 71 species of mammals, determine the extent of their suitable habitats 170 species of birds, as well as numerous covered by GNP’s conservation zones. reptilian and amphibian species) and ecosystem diversity. In recent years, problems such as the road passing through the park (the Asian www.SID.ir

3 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID Material and methods Hyrcanian biome in Iran (Varaste Moradi Study area and sampling 2004) and the last safe habitat for various GNP covers 91,895 hectares at the intersection wildlife species (Varasteh Moradi 2005). of three provinces (Golestan, North Khorasan, At first by performing field operation through and Semnan). The park is located at the direct observation method, signs of animals’ easternmost section of the Hyrcanian forests presence (such as track, footprint and run way, (Darvishsefat 2006), 120 km east of Bonjnourd and pellet or dropping for all species, scrape for along the Gorgan-Mashhad highway leopard and brown bear, rubbing for roe deer (Hasanzadeh Kiabi et al. 1993). GNP is located and red deer, wallowing for wild boar, gnawing between 37˚ 16’ to 37˚ 31’ north and 55˚ 43’ to and quills for porcupine) as well as indexes and 66˚ 17’ east (Hasanzadeh Kiabi et al. 1993) GNP experts’ information the highest number (Fig. 1). It has a mountainous landscape with of species presence points during one year (i.e. elevations ranging from 450 to 2411 m above 1613 points) was recorded by using GPS sea level. GNP is the eastern terminus of the technology (Table 1).

Figure 1. Geographic location, elevation and forest cover of GNP (GIS Department, Department of Environment of Golestan Province).

Environmental variables 2012). Each vegetation type was extracted from Type of vegetation cover determines the the land-cover map of the area and the ratio of distribution patterns of animal species (Burton each type was calculated using neighborhood et al. 2012). Vegetation type variables have analysis in ArcGIS. Vegetation types included been used in modeling as indicators of habitat forest, shrub-grassland, forest-shrub, and type, vegetation cover structure, and primary Onobrichys-grass. production (Mueller et al. 2008, Burton et al.

www.SID.ir

4 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID Table 1. Number of presence points recorded for each species in GNP.

Species Threat's category No. of observations Goitered gazelle (Gazella subgutturosa) VU 64 Grey wolf (Canis lupus) LC 87 Wild sheep (Ovis orienalis) VU 154 Roe deer (Capreolus capreolus) LC 81 Red deer (Cervus elaphus) LC 195 Leopard (Panthera pardus) EN 200 Brown bear (Ursus arctos) LC 258 Porcupine (Hystrix indica) LC 237 Bezoar goat (Capra aegagrus) VU 100 Wild boar (Sus scrofa) LC 237

Farmlands are among the alterations made (by of each species and for this reasons these two human) in natural environments which are variables have been prepared. The elevation located with varying distances from the park, map was prepared using a contour map and some even have common borders with GNP. slope was calculated using DEM and Surface Ranger stations can also influence the function. The maps were generated in ArcGIS distribution of species by providing protection 9.3 at 30 m resolution. against poachers. A map which shows the Then variables related to human interferences distance from ranger stations and farmlands and bio characteristics of landscape were was generated using the distance function in prepared in ArcGIS 9.3 as raster layers, for ArcGIS. determination of minimum habitat The landform and topography can influence requirements and modeling of the areas which prey availability, size of home range, and have potential. Overall, 10 information layers availability of movement corridors for were used (table 2)). carnivores. Topography can also affect distance from areas with high human activity and Data analysis and modeling population density as such phenomena are less In this study MaxEnt, a species distribution common on rough terrain (Salvatori and model which utilizes the principle of maximum Linnell 2005, Gavashelishvili and Lukarevskiy entropy, was used to analyze data, generate 2008, Rich et al. 2012). We used a Digital models and prepare habitat suitability maps. Elevation Model (DEM) prepared by USGS to MaxEnt is a presence-only model which provide two variables of elevation and slope. eliminates many of the shortcomings and elevation by affecting temperature, vegetation uncertainties arising from absence points cover, and access of carnivores and herbivores (Phillips et al. 2006). As input, MaxEnt accepts to different areas, and slope by affecting raster layers of species presence data and vegetation cover, soil depth, and access to the environmental variables (all in ASCII format). area, play important role in habitat preferences

Table 2. Input variables for MaxEnt model. Variable Abbreviation Distance from agricultural land agri Elevation Dem Distance from Forest Forest Distance from Ranger Station Gurad_dis Distance from Residential areas Setl_dis Distance from shrub-forest vegetation Type Shrub_forest Distance from shrub-grassland vegetation Type Shrub_grass Distance from Onobrichys-grass vegetation Type Sprs_grass Slope Vrm Distance from Water Water_dis www.SID.ir

5 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID MaxEnt method is based on machine learning discriminate between points where a species is response. It is designed for making predictions present and points where the species is absent using incomplete data. This method estimates (Philips et al. 2006). the distribution with maximum entropy (closest Using the aforementioned methods, 545 and to uniform) from sampling points in 1068 presence points were recorded for comparison with background and considering carnivores and herbivores in GNP, limitations inferred from input data (Grendar respectively. Information layers, including 2001, Phillipes et al. 2004, 2006). Therefore, presence points and environmental variables, the output indicates the goodness of fit between were prepared and entered into MaxEnt. Finally the predictions of the model and presence the habitat suitability layers have been points, compared to a uniform distribution. In this study, the correlation between overlapped in order to determine the species environmental variables and species’ presence richness hotspots in GNP and by comparing points was calculated. This correlation was these hotspots and current zoning of the park compared with the correlation between the conservation gaps can be detected (areas environmental variables and 10,000 random with the highest number of suitability resulting points (background or pseudo-absence points) from the accumulation of overlapping in the study area. In this manner, the relative considered as hotspots) (Rodrigues et al. 2004, entropy between presence points and random Catulloa et al. 2008, Brito et al. 2009, Wilting points was reduced. Finally, a model was et al. 2010, Pous et al. 2011, Kalle et al. 2013, formulated with the best predictive power Almasieh et al. 2016, Naderi et al. 2018). which is capable to specify species’ suitable habitat in its accessible range (Phillips et al. Results 2006, Elith et al. 2011, Libal et al. 2012). A comparison between map of hotspots of The maximum entropy method in MaxEnt by species richness and the current zoning scheme applying the jackknife test is capable to detect exposed the gaps in protection. the most important variables in determining habitat suitability and to evaluate their Overlap between species suitable habitats contribution in prediction of suitable areas. The with zones 1 and 2 in GNP area under the Receiver Operating Table 3 illustrates the degree of overlap Characteristic curve (AUC) was used to between suitable habitat for the studied species evaluate the predictive power of the model. and GNP’s zone 1 (safe zone) and zone 2 AUC is the most prominent threshold-value- (protective zone). The two zones respectively independent metric for evaluating model cover 423.15089 and 282.000299 km2 of the performance (Palialexis et al. 2009). AUC park. equals the probability that the model would

Table 3. Overlap between suitable habitat and conservation zones Species Suitable habitat Suitable habitat Suitable habitat Suitable habitat 2 2 in zone 1 (km ) in zone 2 (km ) in zone 1 (%) in zone 2 (%) Goitered gazelle 1.439 3.858 0.3 1.3 Red deer 109.438 50.3453 25.8 17.8 Roe deer 105.066 58.993 24.8 20.9 Wild sheep 61.068 43.761 14.4 15.5 Bezoar goat 101.345 63.63 23.9 22.5 Wild boar 114.130 70.815 26.9 25.1 Porcupine 107.293 75.601 25.3 26.8 Grey wolf 27.007 39.787 6.3 14.1 Brown bear 134.702 193.268 31.8 68.5 Leopard 119.110 80.393 28.1 28.5

Average 20.8 24.1 www.SID.ir

6 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID According to Table 3, the largest overlap Table 4 presents species richness of large between zones 1 and 2 and suitable habitat is mammals in the protective zone and safe zone predicted for brown bears (Ursus arctos) with (conservative zones) in GNP and Table 5 31.8% and 68.5% of their suitable habitat illustrates area of regions with different levels situated in zones 1 and 2, respectively. Overall, of species richness in conservative and non- a large percentage of conservation zones in the conservative zones in the park. park protect brown bears’ suitable habitat. This can be due to omnivorous diet of this species Hotspots of species richness for large and as a result its distribution in a wide range mammals in GNP area. The smallest overlap was observed for Figure 2 depicts hotspots of species richness in goitered gazelles (Gazella subgutturosa) with a GNP. The map is produced by overlaying mere 0.3% overlap with zone 1 and only 1.3% habitat suitability maps for 10 studied species. overlap with zone 2. This poor overlap may be Figure 3 shows Location of hotspots of species due to the fact that the habitat of this species is richness for large mammals in relation to limited to the marginal regions of the Park. conservation zones in GNP.

Table 4. species richness of large mammals in the protective zone and safe zone in GNP

With a richness of 1 to 3 With a richness of 4 to 6 With a richness of 7 to 9

Area of zone 1 (km2) 263.557 143.441 15.831

Area of zone 2 (km2) 192.915 76.519 12.564

Area of zone 1 (%) 62.2 33.8 3.7

Area of zone 2 (%) 68.4 27.1 4.4

Table 5. Area of regions with different levels of species richness in conservation and non-conservation zones in the park

Regions Surface Matched Matched Matched Matched Matched (km2) with zone 1 with zone with zone with zone with (km2) 2 (km2) 1 (%) 2 (%) other zone (%)

1 to 3 species 573.679 263.31 192.756 45.9 33.6 20.5

4 to 6 species 296.194 143.35 76.418 48.4 25.8 25.8

7 to 9 species 39.434 15.813 12.540 40.1 31.8 28.1

Discussion hotspots of species richness and species’ A range of methods have been proposed for suitable habitats and conservation zones. prioritizing species for conservation, each of Therefore, by selecting focal species, we tried them with a different approach has been to select areas with the highest animal diversity focused on endangered, native, endemic, possible, so that in future revisions of the flagship, key, indicator, umbrella species and zoning of the Park these areas can be added to species with economic, ecological, scientific, or or cut from the current zoning. Table 3 presents cultural significance. the degree of overlap between each species’ In this study, we attempted to revise and suitable habitat and zones 1 and 2 in the park. modify the current zoning in GNP based on Average overlap of suitable habitats with zone maximizing the overlap between coverage of 1 and zone 2 was respectively about 20% and www.SID.ir

7 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID 25% for goitered gazelle (Gazella fall within conservation zones, there are large subgutturosa), roe deer (Capreolus capreolus), patches with high biodiversity that are not Caspian red deer (Cervus elaphus maral), covered with conservative zones. These patches bezoar goat (Capra aegagrus), wild sheep are circled and indicated with arrows in Figure (Ovis orientalis), porcupine (Hystrix indica), 3. In the future, and in the absence of conflicts, wild boar (Sus scrofa), grey wolf (Canis lupus), these patches should be added to current brown bear (Ursus arctos), and leopard conservation zones. (Panthera pardus). The current coverage Pous et al. (2011) used MaxEnt to prioritize demonstrates the inadequacy of the current and assess the efficiency of Morocco’s network zoning and highlights the need to revise current of protected areas to support and conserve conservation zones, particularly for species reptiles and amphibians. After preparing whose suitable habitats have less than 50% species richness maps for reptiles and combined overlap with zones 1 and 2 and amphibians in Morocco and distribution shows previous criteria used for selecting and patterns for each species, they concluded that a bordering of the conservative zones have not significant expansion of protected areas is been sufficient to provide maximum protection needed to meet the global goal of at least 10- for focal species in GNP. These gaps are larger percent representation in order to guarantee the for roe deer (45.7% total overlap), bezoar goat sustainability and survival of each individual (46.4%), grey wolf (20.4%), and wild sheep reptilian and amphibian species. Although it (29.9%). The largest gap is particularly related merely uses mammalian species as pattern, this to G. subgutturosa with only 1.64% of its study indicates as well that a 54-percent gap in suitable habitat included in conservation zones. habitats with an average or higher richness (4-9 Very limited inclusion of the suitable habitat species), and a 28-percent gap in habitats with of G. subgutturosa in zone 1 and zone 2 (with the highest richness (7-9 species) were 0.3% and 1.3% coverage, respectively) could observed which shows the need for inclusion of be due to the fact that limited habitat of this other key and indicator vertebrate species of species is located in the marginal areas of GNP GNP in researches and making new decisions and accordingly for better protection of this based on other species richness areas in order species a revision in zoning, especially in the to have protected areas with the highest species eastern limits of GNP seems necessary. richness and the smallest conservative gap. In Table 4 presents 69% of the area of this regard, species from other taxonomic conservation zones safeguards the suitable classes and their respective richness areas habitat for over 40% of large species (4-9 should also be taken into account. Therefore, species). Meanwhile, only 8.1 percent of considering the aforementioned explanations, conservation zones encompass the suitable the safe zone in GNP, subject to attachment of habitat for 7-9 species (between 70 to 90 some mentioned patches as per Figure 4, covers percent of species). As a result, adjusting the the areas having the highest biodiversity. borders of zones 1 and 2 to include a larger portion of suitable habitats for 7-9 species will Conclusion improve protection efficiency and biodiversity Although the conservation zones of the park conservation. lack sufficient overlap with suitable habitats of Table 5 shows that 74.2% of sites with average some studied species, particularly C. lupus and species richness (4-6 species) and 71.9% of G. subgutturosa, however, zones 1 and 2 cover sites with the highest species richness of large mammals (7-9 species) are covered under the a significant portion of hotspots of species conservation zones in GNP. Shifting the richness in the park. Certainly, adjusting and borders of zones 1 and 2 to include 100% of expanding the protective coverage of zones 1 areas that are suitable habitats for 7-9 species and 2 to include hotspots of species richness would be ideal. (suitable habitats for 4-9 species), as well as Figure 2 depicts hotspots of species richness in extending these zones to species whose GNP. Although significant portions of hotspots suitable habitats have less overlap with www.SID.ir

8 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID

Figure 2. Hotspots of species richness for large mammals.

Figure 3. Location of hotspots of species richness for large mammals in relation to conservation zones in GNP.

zones 1 and 2 will improve protection each zone, may contribute in realization of this efficiency in GNP and enhance representation. goal. It should be noted that as the total area for A revision in zoning, particularly the extent of each zone is proper, a revision in the borders, www.SID.ir

9 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID without expansion or restriction of area, seems under such conditions comprehensive and more appropriate. Through developments in correct decisions regarding the zoning within habitat suitability modeling methods and GNP and selection of new protected areas to considering biodiversity hotspots maps in complement the current network of protected establishment of future protected areas, and areas can be made. Such an approach will pave revision in the borders and zoning of the the way for accomplishing the best possible current protected areas in order to fill level of conservation and eliminating gaps in conservation gaps, one can hope that selection protection. of protected areas will not be merely based on attainment to quantitative results and Acknowledgements fulfillment of 10-percent protection level and The authors would like to express their beside that, achievement to maximum gratitude towards the very well-respected and efficiency in conservation and qualitative principled professor at the Faculty of Natural objectives of protection through using species Resources of University of Tehran, Dr. Kaboli, distribution modeling and biodiversity hotspots and Mr. Mohsen Ahmadi for their will be taken in to account by decision makers unconditional assistance. We would also like in establishment of new protected areas. As to thank GNP’s administrator at the time this other species have different habitat research was carried out, along with park requirements, ideal conditions may only be rangers, administrative personnel, and experts, reached when habitat suitability modeling is in particular Mr. Meysam Madadi, for their performed for other classes of vertebrates, in particular birds and reptiles, in order to help with the field work. supplement the findings of this study. Only

Figure 4. Hotspots of species richness in GNP which are not included in zones 1 and 2.

References the Iranian black bear in Iran. Ursus Almasieh K., Kaboli M., Beie P. 2016. 27(1):18–30. Identifying habitat cores and corridors for Brito J.C., André L., Acosta A.L., Álvares F., www.SID.ir

10 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID Cuzin F. 2009. Biogeography and Blueprint (A Practitioner's Guide To conservation of taxa from remote regions: Planning For Biodiversity), Island Press, An application of ecological-niche based 404 P. models and GIS to North-African Canids. Hasanzade Kiabi B., Zehzad B., Farhang Biological Conservation 142(12):3020– Darreshouri B., Majnunian H., Goshtasb 3029. H. 1993. Golestan National Park. Burton A.C., Sam M.K., Balangtaa C., Department of Environment, 203 P. Brashares J.S. 2012. Hierarchical IUCN, 2012. The IUCN Red List of multispecies modeling of carnivore Threatened Species version (2011)4. responses to hunting, habitat and prey in a (www.iucnredlist.org). West African protected area. PLOS ONE Kalle R., Ramesh Th., Qureshi Q., Sankar K. 7:e38007. 2013. Predicting the distribution pattern of Catulloa G., Masia M., Falcucci A., Maiorano small carnivores in response to L., Rondinini C., Boitani L. 2008. A gap environmental factors in the Western analysis of Southeast Asian mammals Ghats. PLOS ONE, 8(11):e7929. based on habitat suitability models. Libal N.S., Belant J.L., Maraj R., Leopold BIOLOGICAL CONSERVATION B.D., Wang G., Marshall S. 2012. 141:2730–2744. Hierarchical den-site selection of grizzly Darvishsefat A.A. 2006. Atlas of protected bears in southwest Yukon. Ursus 23:226– areas of Iran. Tehran University Press, 175 230. P. Mace G.M., Collar N.J. 2002. Priority-setting Dehdar Dargahi M., Makhdoum M. 2002. in species conservation, in Norris K., Pain Zoning of Golestan National Park. D.J., (Eds.), Conserving bird biodiversity: Environmental Studies, 28(29):71-77. general principles and their application. Elith J., Phillips S., Hastie T., Dudik M., Chee Cambridge: Cambridge University Press Y., Yates C. 2011. A statistical explanation (Conservation Biology Series no.7), pp. of MaxEnt for ecologists. Diversity and 61-73. Distributions, 7:43–57. Maiorano L., Falcucci A., Boitani L. 2006. Farhadinia M.S., Mahdavi A.R., Hosseini- Gap analysis of terrestrial vertebrates in Zavarei F., Baradarani K., Taghdisi M., Italy: priorities for conservation planning Habibi R. 2010. Conservation model for in a human dominated landscape. the Persian leopard: Lessons learned from Biological Conservation 133:455-473. Sarigol National Park, northeastern Iran. Margules C.R., Pressey R.L. 2000. Systematic IUCN Cat News, 53:36-38. conservation planning. NATURE 405:243- Gavashelishvili A. Lukarevskiy V., 2008. 253. Modelling the habitat requirements of Mueller T., Olson K.A., Fuller T.K., Schaller leopard, Panthera pardus in west and G.B., Murray M.G., Leimgruber P. 2008. central Asia. Journal of Applied Ecology In search of forage: predicting dynamic 45(2):579–588. habitats of Mongolian gazelles using Grendar M. 2001. What is the question that satellite-based estimates of vegetation MaxEnt answers? A probabilistic productivity. Journal of Applied Ecology interpretation. AIP Conference 45:649–658. Proceedings, 568(1):83-93 Myers N., Mittermeier R.A., Mittermeier C.G., Groves C.R. 2003. Drafting a Conservation daFonseca G.A.B., Kent J. 2000. www.SID.ir

11 | Journal of Wildlife and Biodiversity 3(4): 1-11 (2019) Archive of SID Biodiversity hotspots for conservation Rodrigues A.S.L., Tratt R., Wheeler B.D., priorities. Nature 403:853–858. Gaston K.J. 1999. The performance of Naderi M., Farashi A., Alipour Erdi M. 2018. existing networks of conservation areas in Persian leopard's (Pantherapardus biodiversity. Proceeding of the Royal saxicolor) unnatural mortality factors Society of London B 266,pp. 1453–1460. analysis in Iran. PLoS ONE Rodrigues A.S.L., Andelman S.J., Bakarr M.I., 13(4):e0195387. Boitani L., Brooks T.M., Cowling R.M., et Palialexis A., Georgakarakos S., Lika K., al. 2004. Effectiveness of the global Valavanis V.D. 2009. Use of GIS, remote protected area network in representing sensing and regression models for the species diversity. Nature 428:640–654. identification and forecast of small pelagic Rosenzweig M.L. 2003. Reconciliation fish distribution. Proceedings of the ecology and the future of species diversity. Second International Conference on Oryx 37(2):194-205. Environmental Management, Engineering, Salvatori V., Linnell J.D.C. 2005. Report on Planning and Economics (CEMEPE 09), the conservation status and threats for wolf June 21–26, Mykonos, Greece. (Canis lupus) in Europe. Council of Europe Phillips S.J., Anderson R.P., Schapire R.E. Report T-PVS 23, 84 P. 2006. Maximum entropy modeling of Varasteh Moradi H. 2004. Population species geographic distributions. dynamism of the Bovidae in Golestan Ecological. Model 190:231-259. National Park. Agricultural Sciences and Phillips S.J., Dudik M., Schapire R.E. 2004. A Natural Resources 11(3):163-172. maximum entropy approach to species Varasteh Moradi H. 2005. Determination of distribution modeling. Proceedings of the sex ratio and age groups in two species 21st International Conference on Machine Cervus elaphus and Capreolus capreolus in Learning, ACM Press. New York, pp. Golestan National Park. Agricultural 655–662. Sciences and Natural Resources 12(4):154- Pous Ph., Beukema W., Weterings M., 161. Dummer I., Geniez Ph. 2011. Area Wilting A., Cord A., Hearn A.J., Hesse D., prioritization and performance evaluation Mohamed M., Traeholdt C., Cheyne S.M., of the conservation area network for the Sunarto S., Jayasilan M.A., Ross J., Moroccan herpetofauna: a preliminary Shapiro A.C., Sebastian A., Dech S., assessment. Biodiversity Conservation Breitenmoser CH., Sanderson J., 20:89–118. Duckworth J.W., Hofer H. 2011. Modeling Rich L.N., Mitchell M.S., Gude J.A., Sime the species distribution of flat-headed cats C.A. 2012. Anthropogenic mortality, (Prionailurus planiceps) an endangered intraspecific competition and prey South-East Asian small feiled. Plos One availability influence territory sizes of 5(3):e9612. wolves in Montana. Mammal 93:722–731.