Changes in ecological conditions may influence intraguild competition: inferring interaction patterns of snow leopard with co-predators

Study area

The study was conducted in the upper catchment of the Bhagirathi River (2,700 to 3,000 m), a tributary of the Ganga also known as the Gangotri landscape (~4,600 km², 2,700 to 5,000 m). It is located in the northwestern part of the state of Uttarakhand in the western Himalayan region of India (Fig. 1). The study area comprises of Trans-Himalayan (Nelang valley) and the greater Himalayan region (Gangotri valley) of Gangotri National Park, and the surrounding greater Himalayan region (Fig. 1). Monitoring data based on weather station situated at 3,700 m showed that temperatures range between −2.3 and 11 °C (2001–2009, Orr et al., 2019). The weather station has recorded a mean annual winter snowfall of ~546 mm (Bhambri et al., 2011). The area between 2,700–3,000 m altitude is represented by temperate vegetation consisting of Oak (Quercus spp), rhododendron, Himalayan cypress (Cupressus torulosa) and deodar (Cedrus deodara). The subalpine zone (3,000–3,600 m) is formed by Betula utilis, Rhododendron sp., Pinus wallichiana, Taxus wallichiana with intermixed broad-leaved trees like Kharsu oak (Quercus semecarpifolia). Alpine area of Greater Himalaya consist of tall, forbes, mixed herbaceous formations, Danthonia grasslands, Kobresia sedge, often mixed with cushioned species and junipers or rhododendron scrubs (Pusalkar & Singh, 2012). Trans-Himalayan valleys (Nelang valley) consist of steppe scrub vegetation dominated by Artemisia spp., Astragalus candolleanus, Caragana sp., Hyssopus officinalis, Oxytropis spp (Pal et al., 2022b). In the Trans-Himalayan region, snow leopard and woolly wolf are the large predators, and potential wild prey include bharal (Pseudois nayaur), Tibetan woolly hare (Lepus oiostolus), and Himalayan marmot (Marmota himalayana) (Pal et al., 2022b). Tibetan argali (Ovis ammon) is rare in the region (Chandola, 2008; Pal et al., 2021a, 2022b). In the greater Himalayan region, snow leopard co-occurs with carnivores species such as Himalayan brown bear (Ursus arctos isabellinus) and common leopard (Panthera pardus) and potential prey species are: bharal, musk deer (Moschus sp.), Himalayan tahr (Hemitragus jemlahicus), goral (Naemorhedus goral), serow (Capricornis thar) and sambar (Rusa unicolor).

There are nine major villages in the study area (Fig. 1), all of which are located below the elevation of 3,000 m. Landscape has one protected area: Gangotri National Park. In summer (June to September) livestock are grazed, except in two valleys within the protected area. Tourism, mountaineering, and pilgrimage attract many visitors between April and November. The Gangotri National Park remains closed for visitors in winter (November to April). The northern boundary of the study area forms the international border with the Tibet region of China. Patrolling camps and small settlements of the Indo-Tibetan Border Police and other security agencies are present. Analysis of seasonal anthropogenic pressure in the UBB showed a low presence of humans and associated activities in winter compared to summer (Pal et al., 2021a). Livestock grazing, developmental and tourism activities, poaching, and advancement of military infrastructure are potential threats to species inside and outside protected areas in the landscape (Chandola, 2008; Bhardwaj, Uniyal & Sanyal, 2010). In addition, studies have observed notable warming trends in the area that may alter resource availability (Alexander et al., 2016), threatening the long-term existence of high altitude species in the region. Glacier retreat has been observed in the Gangotri and Dokriani glaciers (Kumar et al., 2008; Dobhal, Gergan & Thayyen, 2004). The increase in taxa such as Juniperus, Betula, Salix, and Pinus in the high-altitude areas of Gangotri indicates that temperature and precipitation are increasing in this region (Kar et al., 2002).

Field surveys

This study was conducted with the permission and support of the Uttarakhand Forest Department (Letter No. 836/5-6, 702/5-6 and 117/5-6). Camera trapping was carried out using 3 × 3 km grids in different high-altitude valleys of the Upper Bhagirathi basin. Sampling was carried out between October 2015 and March 2019, focusing on two seasons: summer (May to September) and winter (November to March) (Fig. 2). The length of the sessions was chosen to maintain consistency in terms of anthropogenic disturbance and season. Human activities such as grazing, construction, and paramilitary presence increases in summer (May to September). In contrast, livestock grazing is prohibited and the area is covered up in snow from November to March (Pal et al., 2022a, 2022b). Camera traps (C1 Cuddeback, De Pere, WI, USA) were used to assess the seasonal habitat use of species along a 2,500–5,000 m elevation gradient. Camera traps were deployed in locations likely to be used by mammals based on signs such as scats and pugmark/pellets, at a height of approximately 30–45 cm above the ground. Due to logistical constraints, each camera site could monitored only once per season. Data for summer 2017 (66 sites) and winter 2017–2018 (62 sites) for subalpine and alpine habitats (2,700–3,500 m) were used to analyze space use overlap between snow leopard and common leopard. Data for summer 2018 (47 sites) and winter 2018–2019 (43 sites) for the Trans-Himalayan region (3,200–5,000 m) were used to analyze space use overlap between snow leopard and woolly wolf.

Data analysis

Spatial overlap

In the Trans-Himalayan region, the snow leopard was captured on 84 occasions and the woolly wolf was captured on 22 occasions in summer. Models with habitat variables (slope, ruggedness, elevation, vegetation index) did not converge; therefore, only the interaction model was used to understand the relationship. The Species Interaction Factor (SIF) for woolly wolf and snow leopard was 0.77 (HDI: 0.2 to 1.3) (Fig. 3A). Since this is not statistically different from one, the model suggests that both species occur independently. The occupancy of the snow leopard changed only slightly in the presence of the woolly wolf; occupancy ( $\psi$) changed from 0.52 ± 0.08 to 0.37 ± 0.1 (Fig. 3A). During the winter, the snow leopard was captured 57 times and the woolly wolf 48 times. The SIF for the woolly wolf and snow leopard was 0.64 (HDI: 0.19–1.09), again not statistically different from one (Fig. 3B). The model suggests that species occur independently and do not impact the distribution of each other. The presence of snow leopard’s occupancy ( $\psi$) changed slightly from 0.51 ± 0.09 to 0.21 ± 0.12 in the presence of woolly wolf (Fig. 3B).

In subalpine and alpine regions, the snow leopard was captured on 29 occasions, and common leopard was captured on 33 occasions in summer. In winter, the snow leopard was captured on 57 occasions, and the common leopard was captured on 41 occasions. Models with habitat variable failed to converge, therefore only the interaction model was used to understand the relationship. During summer, SIF for common leopard and snow leopard was 0.78 (HDI: 0.03 to 1.65) (Fig. 4A). The interaction factor was not statistically different from one, suggesting that both species occur independently. In the presence of common leopard, the occupancy ( $\psi$) of snow leopard showed minor change, from 0.22 ± 0.06 to 0.15 ± 0.1. In contrast to summer, SIF for common leopard and snow leopard (0.3; HDI: 0.02–0.75) was found significant in winter (Fig. 4B). The model suggests that individuals of these species occurred together in this area less than expected. In the presence of common leopard, snow leopard occupancy ( $\psi$) changed from 0.47 ± 0.08 to 0.13 ± 0.08 in winters (Fig. 4B).

Diet

For wolves, the undigested matter was present in the following decreasing order: hair (100%, present in all scats), bone (81.25%), hoof (18.75%), claws (3.12%), feathers (9.75%), remains of grass (6.25%) and eggshells (6.25%). The mean percentage frequency of occurrence and relative biomass was highest for bharal, followed by livestock, marmot, birds, and small prey (Table 1).

A total of 54 scats were analyzed for snow leopard collected from greater Himalayan (39) and Trans-Himalayan regions (15). Only 11.11% of scats had more than one prey species. The undigested matter was present in the following decreasing order: hair (100%, present in all scats), bone (59.2%), remains of grass (37.03%), hoof (11.11%), claws (7.4%), feathers (1.8%) and teeth (1.85%). In the greater Himalayan habitat, the frequency of occurrence of bharal and livestock was similar, followed by small rodents, musk deer, Himalayan tahr, Himalayan langur, and birds (Table 2). In Trans-Himalaya, snow leopard scat analysis indicated a much narrower range of prey: bharal followed by livestock, marmots, and small prey (rodents and birds) (Table 1).

In common leopard, only two scats out of 24 (4.8%) had more than one prey species. The undigested matter was present in the following decreasing order: hair (100%, present in all scats), bone (75%), remains of grass (20.8), hoof (4.16%), claws (8.33%), and teeth (4.16%). Rodents were the most frequently found prey item, followed by livestock (Table 2). Wild ungulates in the landscape were also found in leopard’s diet (Table 2). The mean relative biomass was highest for livestock (42.17%), followed by sambar and rodents. Wild ungulates contribute 44.90% of biomass to the leopard’s diet.

Pianka Index for dietary niche overlap between snow leopard and woolly wolf in the Trans-Himalayan valley was 0.8, and 0.75 for the snow leopard and common leopard in the greater Himalayan region.

Activity pattern

In summer, snow leopard showed bimodal activity patterns in both the greater Himalayan (N = 62) and Trans-Himalayan region (N = 104) (Fig. 5). In winters, a similar bimodal activity pattern is observed from greater (N = 466) and Trans-Himalayan region (N = 258) (Fig. 5). In contrast, the common leopard (N = 59) did not show a clear activity pattern (cathemeral activity pattern) in summer (Fig. 5). In winter (N = 206), the common leopard showed a bimodal activity pattern with high activity during dawn and dusk (Fig. 5). The woolly wolves were active throughout the day but showed peak activity during early morning and mid-night in summers (N = 48) (Fig. 5). In winters (N = 391), wolves showed a cathemeral activity pattern (Fig. 5). The activity overlap between snow leopard and common leopard showed 74% overlap in summer and 83% in winters (Fig. 5). In the case of wolf and snow leopard, the temporal overlap was 74% in summer and 80% in winter (Fig. 5).

Snow leopard and woolly wolf interactions

Our hypothesis on the negative effect of wolves on snow leopard space use was not supported on spatial scale. Snow leopard and wolf were found to co-occur independently. Species with similar ecological needs can coexist using different habitat features (Gause, 1934; MacArthur, 1958). We could not directly account for the effect of topographic features on co-occurrence patterns, and used the habitat use information (Pal et al., 2022a, 2022b) to understand habitat preferences and plausible mechanisms of co-occurrence. An assessment of habitat use over 4 years in the Upper Bhagirathi basin have shown that snow leopard and woolly wolves differ in their preference for distinct topographic features. Snow leopard prefer steep terrain (Pal et al., 2022a), a pattern well documented in other studies (Jackson & Ahlborn, 1989; Chundawat, 1990; Jackson, 1996). Rugged areas such as cliffs, broken terrain, and overhanging rock faces are considered important for marking, resting, hunting, and escape shelter for the snow leopard, and the cursorial hunting mode of wolves is best suited to less rugged terrain (Chetri, Odden & Wegge, 2017; Pal et al., 2022b; Kachel, Karimov & Wirsing, 2022). Both the predators may use different habitats to maximise the efficacy of their contrasting hunting modes and access to prey, or snow leopard may use marginal habitats to avoid interference interactions with wolves (Lovari et al., 2013a; Kachel, Karimov & Wirsing, 2022). The preliminary analysis of diet overlap indicated a high overlap between woolly wolves and snow leopard. Such overlap is likely due to the absence of diverse prey species in the region. Both snow leopard and woolly wolves in the Trans-Himalayan region are dependent on bharal, the only wild ungulate found in the area. Even though livestock is available for four months (June to September), it significantly contributes to the diets of wolves and snow leopard.

Temporal partitioning is common in carnivore species pairs with distinct dominance hierarchies (Palomares & Caro, 1999). However, such patterns were lacking in the case of wolves and snow leopard. During the summer, both species showed similar crepuscular activity patterns, and both demonstrated minimal activity during peak human activity hours (see Pal et al., 2021a; Pal et al., 2022b). Snow leopards had a similar crepuscular activity pattern across the habitat types in both seasons. Radio collar investigations have also revealed crepuscular activity in snow leopards (Jackson & Ahlborn, 1989; Johansson, 2017). On the other hand, the wolf does not appear to have a distinct bimodal activity pattern in winter. Human activity is relatively low in Nelang during the winter, which may have encouraged wolves to shift their activity to the daytime (Pal et al., 2022b). Conversely, snow leopard continued to follow the crepuscular activity pattern and showed lesser overlap with wolves in winter.

Snow leopard and common leopard interactions

Contrary to our expectation, we found snow leopard and common leopard temporarily and spatially separated during the summer season. However, in the winter, the common leopard negatively influenced the space use of the snow leopard. Common leopards in the Upper Bhagirathi basin were photo captured up to 3,500 m (subalpine habitat). Snow leopards were photographed mainly in alpine habitats, but also in subalpine and temperate habitats of the basin (2,700–3,500 m). In the summer, snow leopards may move to higher elevations with less snow and occur independently of common leopard presence. During winter, seasonal movement is widely documented in prey species of snow leopard (Schaller, 1998; Anwar & Minhas, 2008; Dendup & Lham, 2018; Pal et al., 2021a, 2021b) which are known to shift to lower elevations due to excessive snowfall. Such behavior is also expected in snow leopards (following prey species) in the winter season, where chances of encounters and competition for space increase with common leopard. In the greater Himalayan region, with a diverse range of species available, snow leopards and common leopard feed on various wild ungulates. Bharal and musk deer are the wild prey found common in the diet of both the felids, but the common leopard’s diet was more diverse than snow leopard. High dietary overlap between snow leopard and common leopard is primarily due to the high proportion of livestock in the diet of both species. In the absence of livestock during winter, competition among predators for prey is likely to increase.

In the greater Himalaya, snow leopard showed the crepuscular activity pattern (similar to the Trans-Himalayan habitat) in both seasons. It is known that under certain circumstances carnivores adjust their activities to the availability of prey (Jenny & Zuberbuhler, 2005; Salvatori et al., 2021). The common leopard showed a similar pattern in summer, showing diurnal activity which corresponds closely to livestock grazing (see Pal et al., 2021a). In winter, the common leopard showed a bimodal activity pattern similar to that of the snow leopard, with a high overlap in activity patterns.

Drivers of intraguild interactions

According to the competitive exclusion principle (Gause, 1934), sympatric predators should partition dietary, spatial, and/or temporal niche dimensions to achieve some level of resource independence that allows stable coexistence (Chesson, 2000). Similar dietary needs and activity patterns between the wolf and snow leopard expect considerable competition between these species. However, habitat heterogeneity and differences in habitat use are likely to limit the competition between snow leopard and wolves, facilitating their co-occurrence in the region. This is consistent with the results of Salvatori et al. (2021) from Mongolia, which found a similar co-occurrence pattern of snow leopard and wolf. A recent study from the Pamir mountains in Tajikistan, conducted at finer scale, found overlapping temporal and dietary niches, but a differential use of habitat type between wolves and snow leopard (Kachel, Karimov & Wirsing, 2022). In the case of snow leopard and common leopard, the negative spatial interaction was limited to scarce resource season, when livestock is absent, and space available is limited due to heavy snow. A study on the interaction between two leopard species in the Sagarmatha National Park showed a high overlap in prey categories and sizes but habitat separation between these two felids (Lovari, Ventimiglia & Minder, 2013b). We believe that, in our case, limited prey resources (lack of livestock) and similar activity patterns in winters may result in intense competition, causing these species to avoid each other on a spatial scale. Melting snow in the higher elevations in summers opens up rugged and alpine habitats and ridgelines for snow leopards and other prey species. Heavy snow in higher elevation locations or the absence of anthropogenic influences in winter cause prey species to shift/expand to lower elevation places. Such a shift in prey species is likely to promote snow leopard utilization of low elevation habitat, increasing spatial overlap between the two leopard species.

Body size appears to be the most crucial determinant of aggressive interaction in carnivores (Donadio & Buskirk, 2006). In the case of large and small differences, intense competition is less likely to occur, whereas, at intermediate differences, intraguild killings are more common (Donadio & Buskirk, 2006). The degree of competition is also influenced by taxonomic relatedness. Species within a family tend to have high similarities in life-history traits than species from other families (Gittleman, 1985; Van Valkenburgh, 1991). As a result, competition is expected to be more lethal within than among families (Donadio & Buskirk, 2006). Furthermore, in seasonal environments, the majority of agonistic encounters between carnivore species occur when food is scarce (Palomares & Caro, 1999). We aimed to understand the interactions of large predators under varying conditions by examining seasonal habitat conditions and prey availability. Wolf and snow leopard, carnivores belonging to different families, appear to co-occur independently in summer and winter habitat conditions. Even though they both rely on the same prey species in the area, the heterogeneous habitat conditions and preference for different terrain types promote their co-existence. On the contrary, snow and common leopard, carnivores from the same family, showed contrasting patterns in distinct ecological settings. Resource-rich seasons allow for minimal competition in terms of prey and accessible area; however, when resources are scarce, both species appear to compete for space. In the latter circumstances, a species with dominance and a more adaptive, broader dietary and spatial niche, such as the common leopard in our case, is likely to predominate.

It is likely that we may not have detected all possible finer-scale intra-guild carnivore adaptations. Research with more scat samples can help understand the full spectrum of diet, seasonality in prey dependence, and competition for prey. The effect of various topographic features, such as topographic complexity, elevation, and prey availability, could be conclusively quantified with a larger dataset. Further research is needed to understand the other factors that can influence intraguild relations, such as the interactive effect of density thresholds, anthropogenic settings, and the availability and spatial distribution of feeding resources (Palomares et al., 1996; Karanth et al., 2017; Sivy et al., 2017). Investigating the interactions of snow leopards with co-predators in various abiotic and biotic settings can help to provide such understanding.