Scholarly article on topic 'Creating a model of habitat suitability using vegetation and ruggedness for  Ovis ammon  and  Capra sibirica  (Artiodactyla: Bovidae) in Mongolia'

Creating a model of habitat suitability using vegetation and ruggedness for Ovis ammon and Capra sibirica (Artiodactyla: Bovidae) in Mongolia Academic research paper on "History and archaeology"

CC BY-NC-ND
0
0
Share paper
Academic journal
Journal of Asia-Pacific Biodiversity
OECD Field of science
Keywords
{argali / "habitat suitability index model" / ibex / Mongolia}

Abstract of research paper on History and archaeology, author of scientific article — Nanette Bragin, Sukh Amgalanbaatar, Ganchimeg Wingard, Richard P. Reading

Abstract Spatially-explicit wildlife habitat models, such as a Habitat Suitability Index Model (HSIM), are increasingly used to understand optimal environmental conditions needed for a species survival and viability. HSIM compares different environmental variables, including vegetation, slope, and aspect to determine optimal habitat for a species. HSIM can compare a species’ use of resources with availability and determine risks for a species viability. Furthermore, decision makers can use HSIM to make decisions regarding resources needed for species of concern and development of protected areas. We used a geographic information system to create a HSIM for Argali sheep (Ovis ammon) and Siberian ibex (Capra siberica), species of conservation concern in Mongolia. We used vegetation and ruggedness layers and compared argali and ibex use with habitat availability. Argali and ibex presence correlated with three habitat classes: dense rock, low-density shrub, and short grass/forb. We found no significance in correlation for ruggedness.

Academic research paper on topic "Creating a model of habitat suitability using vegetation and ruggedness for Ovis ammon and Capra sibirica (Artiodactyla: Bovidae) in Mongolia"

Accepted Manuscript

Creating a model of habitat suitability using vegetation and ruggedness for Ovis ammon and Capra sibirica, (Artiodactyla: Bovidae) in Mongolia

Nanette Bragin, Sukh Amgalanbaatar, Ganchimeg Wingard, Richard P. Reading

PII: S2287-884X(17)30070-5

DOI: 10.1016/j.japb.2017.06.003

Reference: JAPB 234

To appear in: Journal of Asia-Pacific Biodiversity

Received Date: Revised Date: Accepted Date:

15 January 2017 25 April 2017 10 June 2017

Please cite this article as: Bragin N, Amgalanbaatar S, Wingard G, Reading RP, Creating a model of habitat suitability using vegetation and ruggedness for Ovis ammon and Capra sibirica, (Artiodactyla: Bovidae) in Mongolia, Journal of Asia-Pacific Biodiversity (2017), doi: 10.1016/j.japb.2017.06.003.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Creating a model of habitat suitability using vegetation and rnggedness for Ovis ammon and Capra sibirica, (Artiodactyla: Bovidae) in Mongolia

Nanette Bragina*, Sukh Amgalanbaatarb, Ganchimeg Wingarda, and Richard P. Reading0

a*.Department of Conservation and Research, Denver Zoological Foundation, 2300 Steele Street, Denver, CO 80205. 720-337-1517

bDirector of Ikh Nart Nature Reserve, Mongolia cConservation Biology Consultant, Denver, Colorado

USA E-mail: nbragin@denverzoo.org;

Abstract

Spatially-explicit wildlife habitat models, such as a Habitat Suitability Index Model (HSIM) are increasingly used to understand a species home range, resource use, and optimal environmental conditions needed for survival and viability. An HSIM compares different environmental variables, such as vegetation, slope, and aspect to determine optimal habitat for a species. An HSIM can compare a species' use of resources to what is available and determine risks such as species encroachment from domestic livestock. In addition, decision makers can use the HSIM to inform them of resources needed for species of concern and how to develop protected areas. We used a Geographic Information System (GIS) to create a HSIM for Argali sheep (Ovis ammon) and Siberian ibex (Capra siberica), species of conservation concern in Mongolia, in the Ikh Nart Nature Reserve and surrounding areas. We used vegetation and ruggedness layers and compared argali and ibex use to habitat availability. We found that argali and ibex presence correlated with 3 habitat classes; dense rock, low-density shrub, and short grass/forb. We found no significance in correlation for ruggedness.

Key Words: argali; habitat suitability index model; ibex; Mongolia

Introduction

Spatially-explicit wildlife habitat models are increasingly used in conservation planning. Models can assess habitat preference of targeted animal species, determine protected area status, search for optimal habitat for species re-introduction, identify risks impacting flora and fauna, and provide a framework for decision makers to understand the impact of land use and management decisions (Yamada et al 2003). A Geographic Information System (GIS) can model the spatial and temporal interactions between a species and its environment. Creating a Habitat Suitability Index Model (HSIM) with GIS involves analyzing a species preference in habitat use and creating a numeric scale that indexes most to least preferred resources (Yamada et al 2003). The HSIM can determine appropriate habitat for relocation of or re-introduced species and assess threats to habitat and species survival. It can also help inform decision makers as to the adequacy of a protected area's design for species conservation and, if pertinent, provide guidance for park management and expansion.

We modelled location data for two species of conservation concern, Argali sheep (Ovis ammon) and Siberian ibex (Capra siberica), to help understand habitat correlations (and thus begin to assess habitat selection) in the Ikh Nart Nature Reserve and surrounding area, Mongolia. We used those analyses to develop a Habitat Suitability Index Model (HSIM) for argali and ibex in Ikh Nart that permits identification of preferred habitat.

We used GIS to create a HSIM using vegetation and ruggedness layers. We hypothesized that both argali and ibex would utilize areas of short grass for forage and high ruggedness (or at least near areas of high ruggedness) for escape terrain to avoid predation. However, we also predicted that ibex would select areas with higher levels of ruggedness, as they require escape terrain more than the cursorial argali (Fedosenko and Blank 2005). Our study compares two sympatric species, if and how preferred resources overlap, and if desert species are different from animals in other regions.

The Argali sheep (Ovis ammon), listed as Endangered in Mongolia (Clark et al 2006), inhabits a range that includes much of Mongolia and parts of Central Asia, China and India (Maroney 2005). Ovis ammon, the largest species in the genus, has a stout body, relatively long legs built for running, and massive horns (Fedensko and Blank 2005). Argali prefer topography of foothills, high plateaus, intermountain valleys, gentle

slopes, and rolling steppes in high mountains (Fedensko and Blank 2005; Amgalanbaatar and Reading 2000). In Ikh Nart, our study area, argali are threatened by domestic livestock competition, poaching, and predation (Reading et al 2005).

The Siberian ibex (Capra sibirica) is considered Near-Threatened in Mongolia (Clark et al 2006), yet remains poorly understood due to lack of research. Siberian ibex range through the mountains, canyons, and other rough terrain of central and middle Asia, southern Siberia, and the northwest Himalayas. Fedosenko and Blank (2001) describe the Siberian ibex as the largest and heaviest in the Capra genus, with males weighing up to 130 kg. Agile at climbing sharp rocky slopes and cliffs, they use escape terrain to avoid predators as they cannot run quickly. In Ikh Nart, our study area, ibex are considered threatened by domestic livestock competition, poaching, and predation (Reading et al 2007). Siberian ibex and argali sheep overlap in home ranges in our study area as well as other parts of Mongolia (Reading et al 2006; Fedosenko and Blank 2005).

Methods and materials

Study Area

Established in 1996, the Ikh Nart Nature Reserve (hereafter Ikh Nart) is a protected area, approximately 66,760 ha, in the northwestern Aimag (province) of Dornogobi (N45.723°, E108.645°; Reading et al 2011). Ikh Nart is a transition zone between the steppe and desert-steppe habitats in Mongolia. There is a mix of continental and arid climates with temperatures ranging from -40 °C to 43 °C (Reading et al 2011; Reading et al 2007). The region is dry with low humidity and precipitation, with most rain falling in summer (< 100 mm/year). In the spring winds can reach up to 25 mph (Bragin 2010). There are some permanent cold water springs, while the rest of the reserve contains ephemeral drainages, creek beds, alkaline pools, and ponds (Wingard 2005; Reading et al 2007; Jackson et al 2006). Our study site encompasses 3 areas, the core zone of Ikh Nart, Ikh Nart, and the surrounding area of Ikh Nart based on the extent used in a previous study (Jackson et al 2006) (N45.83°-N45.54°; E108.48°-E108.73°) (Figure 1A).

Methods

We used telemetry data collected for argali (15 males and 24 females) and ibex (10 males and 8 females) from 2003-2008 to calculate home ranges. Animals were fitted with VHF radio collar and tracked using a traditional receiver; a yagi, handheld, two or three-element antenna; and a Global Positioning System (GPS). We only used data from animals with 25 or more telemetry fixes for analyses (Reading et al 2007; Reading et al 2005). We collected a total 3,871 fixes for argali and 1,748 fixes for ibex. We created minimum convex polygons (MCP) and 95% and 50% fixed kernel home ranges (kernel) using ArcGIS 9.3.1 (ArcGIS v. 9.3.1, Environmental Systems Research Institute, Redlands, California, USA) and Hawth's tools (Worton 1989; Beyer 2004; Reading et al 2007). Based on a previous study, the mean home range for argali using MCP = 56.54 ± 3.72 km2; 95% kernel ranges = 75.85 ± 5.32 km2; and 50% kernel ranges = 10.99 ± 1.63 km2. The mean home range for ibex using MCP = 1.34 ± 0.07 km2; 95% kernel ranges = 40.84 ± 2.83 km2; 95% kernel ranges = 15.42 ± 1.08 km2 (Reading et al 2005; Reading el al 2007; Figures 2A & 2B). As the data were clumped, we used least squares cross validation to select a smoothing factor for each kernel polygon. We calculated the area for each polygon using X-tools Pro (X-Pro tools extension for ArcGIS Desktop Copyright C Data East, LLC).

To create the HSIM model, we compared animal resource use to resource availability (Johnson 1980; Boyce et al 2002). We first clipped MCPs and 95% and 50% kernel home range polygons for individual argali and ibex from vegetation and ruggedness layers created earlier (Jackson et al 2006; Bragin et al 2013, Figures 1A & 1B) in ArcMap 10.2 Geographic Information Systems software (Environmental Systems Research Institute, Redlands, CA). The vegetation map produced by Jackson et al (2006) classified Ikh Nart into 7 habitat classes: dense rock, high-density shrub, low-density shrub, semi-shrub, short-grass/forb, tall vegetation, and water (Figure 1A). Water continually showed an insignificant use and so we excluded it from analyses, leaving six vegetation classes. We used slope and aspect to create the ruggedness layer using the following equation (see Bragin et al 2013):

SARI = (STDEV Slope) x Variety of Aspect/ (STDEV Slope) + Variety of Aspect

We divided the ruggedness layer of the study site into 9 classes, ranging from 1 = most rugged to 9 = least rugged (Figure 1B). We combined most rugged categories 1 and 2 as well as least rugged categories 8 and 9 since the tail ends of the rugged categories did not contain a large area.

We calculated percent use for each vegetation and ruggedness class and imported the data into the SYSTAT Software (SYSTAT 11 for windows, SYSTAT Software, 2009). We also calculated the percentage of each vegetation and ruggedness class within the entire study area, within Ikh Nart's boundaries, and within Ikh Nart's Core Area that park officials established to protect ostensibly critical argali and ibex habitat. For each species (argali and ibex), we used mean percent use for all individuals as well as males only and females only. To avoid problems with pseudo-replication, we took the mean of all years for each animal.

We conducted statistical analyses using Pearson's chi-square, Likelihood ratio chi-square, and phi tests to compare argali and ibex use of vegetation and ruggedness classes with the occurrence of each within the core area, reserve, and study site, but since results were very similar, we report only Pearson's chi-square tests here. We insured that all data conformed to the assumptions of each test prior to analyses. We set significance at p < 0.05.

To further refine our model, we used the 95% percent volume kernel for vegetation and ruggedness classes. We found that the 50% volume kernel underestimated home range size whereas the MCP overestimated it. We used the entire study site, rather than only Ikh Nart, to calculate index values, as it encompasses more area to inform conservation decisions (although recently, the Dalanjargalan Soum (county) government established a local protected area contiguous to Ikh Nart that effectively expanded the reserve throughout most of the study area). We divided the percent of each class of vegetation and ruggedness used by argali and ibex by the percent of each class available in the study site (Figures 3A & 3B). To create index values, we divided animal use of each vegetation and ruggedness class by resource availability. We then converted these ratios into whole numbers (1 - 7) as our GIS software will not analyze negative and 0 numbers, with 1 being the least preferred classes and 7 being the most preferred classes (Appendix 1; Figures 3A & 3B). We used these weighted vegetation and ruggedness categories in ArcGIS to create a new layer of ostensible habitat preference (i.e. we only tested occurrence, not actual preference) based on an 80% weight for vegetation and 20% weight for ruggedness (Figures 4A and 4B). We based our weights on the relative strengths of the associations. Since vegetation use differed significantly and ruggedness did not, we gave a stronger weight to vegetation. Although

ruggedness was not significant, it is still important to both species based on past research. We restricted the water class in the weighting since we did not use it in the analysis; giving it a preference value of zero.

For additional verification in ruggedness use, we created 50 m, 100 m, and 200 m buffers around the most used rugged classes (classes 5 and 6) and calculated the areas of use of all rugged classes within those buffers, as we assumed that argali and ibex preferred habitat near escape terrain, but not necessarily in it (i.e. we prioritized the importance of forage in the absence of predation threat, which we assumed was rare). We compared all argali and all ibex to the 3 different buffers. Again, we tested for significance using Pearson chi-square, Likelihood ratio chi-square, and phi tests.

Results

We found that both argali sheep and Siberian ibex occurred in areas with dense rock, low density shrub, and short grass/forb in the study site and Ikh Nart Nature Reserve. They avoided areas with high density shrub, semi-shrub, and tall vegetation (p < 0.05; Figures 3A, 4A and 4B). Ibex occurred in areas with dense rock and short grass forb to a greater extent than argali. Argali occurred in areas with low density shrub more than ibex (Figure 4A). Yet, we found no significant differences between males and females or between argali and ibex in their use of areas with different vegetation classes (p = 1.00, Table 1). We found no significant difference in the proportion of vegetation classes within Ikh Nart's Core Zone and within the home ranges of argali and ibex suggesting this area well matches the needs of these species (p = 0.67 and p = 0.94, respectively; Table 1).

We found no significant use of rugged categories in the core zone, Ikh Nart, or study site for either argali or ibex (p > 0.70, Table 2). Also, we found no significant difference between males and females or species for ruggedness occurrence (p = 1.00, Table 2). Even though not significant, both argali and ibex used areas with a moderate degree of ruggedness (categories 5 and 6), while they avoided areas with the most and least rugged areas (Figures 3B, 4A & 4B). In our further investigation of ruggedness categories through 50, 100, and 200 m buffers, we found no significance (p > 0.20, Table 2) with the exception of one category. Buffering at 50 m showed significant rugged use by ibex, but not argali (p = 0.02 and 0.07, respectively; Table 2).

Discussion

Argali and ibex confirmed our hypothesis that they would occur in areas with short grass/forb and low-density shrubs more than other vegetation types; yet, the lack of correlation with more rugged terrain surprised us. We expected ibex would use high ruggedness areas as a preference, while argali would prefer moderate ruggedness. Overall, our models showed a correlation for more moderate ruggedness than was available to argali and ibex. Perhaps the moderately rugged areas offer a greater access to food sources close to escape terrain. Other researchers found that at least ibex showed preference for more rugged areas. Namgail, et al (2004) found that Tibetan argali, Ovis ammon hodgsoni, selected areas away from cliffs, preferred moderate slopes and avoided steep slopes, but they selected high elevations. Similarly, Singh, et al (2009) found that depending on model, altitude, slope, northness, and Normalized Difference Vegetation Index (NDVI), but not ruggedness, were the most important variables for Tibetan argali. Grignolio, et al (2004) found that for alpine ibex, (Capra ibex ibex), weaning females preferred rugged, rocky areas first, then alpine meadows whereas females without kids preferred Alpine meadows, then rocky areas. Grignolio, et al (2003) reported that male alpine ibex selected alpine meadows, but most frequently used rugged rock and stone ravines, though less than what was available. Although ibex did not select more rugged terrain, they preferentially occurred in areas within 50 m of such terrain, probably because they could access it as escape terrain in the event of predation danger.

Several theories could explain our ruggedness results. First, argali and ibex in our study area may behave differently than these species in other regions of the world. Reading, et al (2005) found that argali in the Gobi steppe do not behave as argali in other regions in showing seasonal movement. Second, our sightings occurred in less rugged areas, as these represent difficult places for telemetry trackers to locate animals. Yet, using 95% kernel home ranges should help avoid some of the potential problems associated with possible source of bias. Third, our location data was biased toward the times of the day (mornings and evening) that argali and ibex likely spend more time grazing, as both species tend to rest in more rugged terrain during the

night and mid-day (G. Wingard, personal communication). Again, using 95% kernel home ranges should at least partially ameliorate this potential problem. We have more recently deployed GPS telemetry collars on both species that should permit us to evaluate the second and third hypotheses in the near future. Fourth, ruggedness, per se might not represent a significant variable in argali and ibex resource selection. Instead, as Singh et al (2009) report, elevation and slope may play a more important role. Finally, dividing the data set by season and exploring preference may produce a clearer picture of Argali and ibex habitat selection in our study area, although Reading et al (2005) found no seasonal movements among argali.

Our results provide insight into resource selection and creates a visual representation for managers of Ikh Nart Nature Reserve and other protected areas with these species. Also, understanding argali and ibex preference for short-grass/forb, low density shrub, and dense rock enables better park management. Park managers can assess resource use by domestic livestock and determine if and to what degree competition exists. As long as managers understand the complexity and accuracy of a model design, they could use our model to inform protected area management, design, expansion, and selection of additional areas suitable for these species.

Conflicts of interest

The authors declare that there is no conflicts of interest.

Acknowledgments

We thank the Denver Zoo for supporting this project and providing equipment, staff, and other resources. Also, we thank the Earthwatch Institute for providing radio and GPS collars.

References

Amgalanbaatar S, Reading RP. 2000. Altay Argali sheep. In Endangered animals, conflicting issues, ed. Richard P. Reading and Brian Miller. Westport: Greenwood Press.

Beyer HL. 2004. Hawth's analysis tools for ArcGIS. http://www.spatialecology.com/htools.

Bragin N. 2010. Creating a Habitat Suitability Index model for Argali sheep (Ovis ammon), and Siberian ibex (Capra sibirica) in Ikh Nartin Reserve, Mongolia. Master Thesis Denver University 1-57.

Bragin N, Singh NJ, Reading RP. 2013. Creating a Ruggedness Layer for Use in Habitat Suitability Modeling for Ikh Nart Nature Reserve, Mongolia. Mong. J. Biol. Sci 11(1-2):19-23.

Boyce MS, Vernier PR, Nielsen SE, et al. 2002. Evaluating resource selection functions. Ecological Modeling 157:281-300.

Clark EL, Munkhbat J, Dulamtseren S, et al (compilers and editors). 2006. Mongolian red list of mammals. Regional red list series Vol. 1. London: Zoological Society of London.

Fedosenko AK, Blank DA. 2001. Capra sibirica. Mammalian Species 675:1-13.

Fedosenko AK, Blank DA. 2005. Ovis ammon. Mammalian Species 773:1-15.

Grignolio S, Parrini F, Bassan B, et al. 2003. Habitat selection in adult males of Alpine ibex, Capra ibex ibex. Folia Zool 52(2):113-120.

Grignolio S, Rossi I, Bassano B, et al. 2004. Seasonal variations of spatial behavior in female Alpine ibex (Capra ibex ibex) in relation to climatic conditions and age. Ethology, Ecology, and Evolution 16:255-264.

Jackson DS, Murdoch J, Mandakh B. 2006. Habitat classification using Landsat 7ETM+ imagery of the Ikh Nart nature reserve and surrounding areas in Dornogobi and Dundgobi Aimags, Mongolia. Mongolian Journal of Biological Sciences 4(2):33-40.

Johnson DH. 1980. The Comparison of Usage and Availability Measurements for Evaluating Resource Preference. Ecology 61(1):65-71.

Maroney RL. 2005. Conservation of Argali, Ovis ammon, in western Mongolia and the Altai-Sayan. Biological Conservation 121:231-141.

Namgail T, Fox JL, Bhatnagar YV. 2004. Habitat segregation between sympatric Tibetan argali Ovis ammon hodgsoni and blue sheep Pseudois nayaur in the Indian Trans-Himalaya. J. Zool., Lond262:57-63

Reading RP, Amgalanbaatar S, Wingard GJ, et al. 2005. Ecology of Argalis in Ikh Nartiin Chuluu, Dornogobi Aymag. Erforsch. biol. Ress. Mongolei (Halle/Saale) (9):77-89.

Reading RP, Bedunah DJ, Amgalanbaatar S. 2006. Conserving biodiversity on Mongolian rangelands: Implications for protected area development and pastoral uses. USDA Forest Service Proceedings RMRS.

Reading RP, Amgalanbaatar S, Kenny D, et al. 2007. Siberian Ibex (Capra [ibex] sibirica) home ranges in Ikh Nart nature reserve, Mongolia: Preliminary findings. Mongolian Journal of Biological Sciences 5(1-2):29-37.

Reading RP, Kenny D, Steinhauer-Burkhart B. 2011. Ikh Nature Reserve Nature Guide No. 4 Mongolia. ECO Nature Edition, Steinhauer-Burkart OHG. pp. 1-64.

Reading RP, Wingard G, Selenge T, et al. 2015. The crucial importance of protected areas to conserving Mongolia's natural heritage. In: Wuerthner G, E Crist, T Butler, editors. Protecting the wild: Parks and wilderness, the foundation for conservation. Washington: Island Press. pp 257-265

Singh NJ, Yoccoz NG, Bhatnagar YV, Fox JL. 2009. Using habitat suitability models to sample rare species in high-altitude ecosystems: a case study with Tibetan argali. Biodivers. Conserv 18(11):2893-2908.

Wingard GJ. 2005. Dietary overlap between Argali sheep and domestic livestock in Mongolia. Master Thesis, University of Montana 1-64.

Worton BJ. 1989. KERNEL METHODS FOR ESTIMATING THE UTILIZATION DISTRIBUTION IN HOME-RANGE STUDIES'. Ecologj 70(1). 1989, pp. 164-168

Yamada K, Elith J, McCarthy M, et al. 2003. Eliciting and integrating expert knowledge for wildlife habitat modeling. Ecological Modelling 165:251-264.

Figure Captions

Figure 1. A, Vegetation classes of northern Ikh Nart Nature Reserve, Ikh Nart's core zone, and the study site. Vegetation classes (dense rock, high density shrub, short grass forb, tall vegetation, water, semi-shrub, and low density shrub) from Jackson et al (2006); B, Ruggedness layer of the northern Ikh Nart Nature Reserve, Mongolia, Ikh Nart's core zone, and the Study Site. Ruggedness from Bragin et al (2010), based on a combination of aspect and slope that created an index from 1 = most rugged to 9 = least rugged.

Figure 2. A, Example of mean home range for argali (Ovis ammon) using minimum convex polygon, 95% kernel ranges, and 50% kernel ranges in square kilometers; B, Example of mean home range for Siberian ibex (Capra sibirica) using minimum convex polygon, 95% kernel ranges, and 50% kernel ranges in square kilometers.

Figure 3. Mean percentage of ibex (Capra sibirica) and argali (Ovis ammon) home ranges falling within different: A, vegetation classes; B, ruggedness categories compared to the percentage of those classes and categories available in and around Ikh Nart Nature Reserve, Mongolia.

Figure 4. A, Habitat suitability index model (HSIM) for argali (Ovis ammon) in and around northern Ikh Nart Nature Reserve, Mongolia; B, Habitat suitability index model (HSIM) for Siberian ibex (Capra sibirica) in and around northern Ikh Nart Nature Reserve, Mongolia. Index values range from 1 to 7, with 1 being less preferred and 7 most preferred. Category 0 represents water that we restricted in our GIS weighting tool.

Table 1. Vegetation preference of Argali sheep (Ovis ammon) and Siberian ibex (Capra sibirica), in Ikh Nart Nature Reserve, Mongolia. Comparisons made using Pearson's X2 test of vegetation within argali and ibex 95% kernel home ranges versus vegetation in the surrounding area (study site), Ikh Nart, and Ikh Nart's core zone; vegetation within male versus female 95% kernel home ranges; and vegetation within all argali versus all ibex 95% kernel home ranges.

Comparison Category a:2 df P

95% kernel vs. study site Argali 24.62 5 <0.001

Ibex 32.47 5 <0.001

95% kernel vs. Ikh Nart Argali 16.24 5 <0.01

Ibex 25.29 5 <0.001

95% kernel vs. core zone Argali 3.20 5 0.67

Ibex 1.27 5 0.94

Male vs. female Argali 0.06 5 1.00

Ibex 0.06 5 1.00

Argali vs. ibex 2.30 5 0.70

Table 2. Ruggedness preference of Argali sheep (Ovis ammon) and Siberian ibex (Capra sibirica), in Ikh Nart Nature Reserve, Mongolia. Comparisons made using Pearson's X test of mean ruggedness 95% kernel home ranges of argali and ibex versus the surrounding area (study site), Ikh Nart, and Ikh Nart's core zone; mean ruggedness within male versus female 95% kernel home ranges; mean ruggedness within all argali versus all ibex 95% kernel home range; and all argali and ibex 95% kernel home ranges versus 0 - 50 m, 0 - 100 , and 0 -200 m buffers of the areas with the highest ruggedness values in the study site.

Comparison Category a:2 df P

95% kernel vs. study site Argali 3.58 6 0.73

Ibex 2.91 6 0.82

95% kernel vs. Ikh Nart Argali 1.85 6 0.93

Ibex 1.30 6 0.97

95% kernel vs. core zone Argali 0.31 6 1.00

Ibex 0.23 6 1.00

Male vs. female Argali 0.02 6 1.00

Ibex 0.01 6 1.00

Argali vs. ibex 0.41 6 1.00

95% kernel vs. 0 - 50 m buffer Argali 11.50 6 0.07

Ibex 15.50 6 0.02

95% kernel vs. 0 - 150 m buffer Argali 4.99 6 0.55

Ibex 7.96 6 0.24

95% kernel vs. 0 - 200 m buffer Argali 2.75 6 0.84

Ibex 4.39 6 0.62

Appendix 1

Vegetation and ruggedness index values used to create a habitat suitability index model for argali (Ovis ammon) and ibex (Capra sibirica) in Ikh Nart Nature Reserve, Mongolia. Assigned values determined by dividing the percent use of each vegetation and ruggedness class used by argali and ibex by the percent of each class available in the study site and converting those values into whole numbers ranging from 1-7, with 1 indicated the least preferred class and 7 indicating the most preferred class. Water = restricted since it was not included in the vegetation analysis.

Assigned value

Variable Category

Argali Ibex

Vegetation Dense rock

7 1 5 2

7 1 6 2

High density shrub Short grass and forb Tall vegetation Water Semi-shrub Low density shrub

Restricted

Restricted

Ruggedness 1

5 2 1 1

5 3 2 1 1

30 25 20 15 10 5 0

7 8 & 9

Ruggedness category

^ 20 15 10 5 0

□ Ibex

□ Argali

□ Study site

Dense Dense Sparse Short rock shrub shrub grass

Semi- Tall veg. shrub

Vegetation class