Introduction
Large carnivores such as the tiger Panthera tigris are apex predators and integral to maintaining ecological balance by controlling prey populations, suppressing mesopredators and providing protection as an umbrella species in diverse habitats (Seidensticker & McDougal, Reference Seidensticker and McDougal1993; Sunquist et al., Reference Sunquist1999). However, tiger populations and those of their prey have declined dramatically across the species’ 13 range countries as a result of habitat loss, human–wildlife conflict, poaching and urbanization (Schaller, Reference Schaller1967; Karanth & Sunquist, Reference Karanth and Sunquist1995). Loss of prey is a major factor directly affecting the viability of tiger populations by lowering carrying capacity, adult female density and cub survivorship (Karanth & Stith, Reference Karanth, Stith, Seidensticker, Christie and Jackson1999; Carbon & Gittleman, Reference Carbone and Gittleman2002; Karanth et al., Reference Karanth, Nichols, Kumar, Link and Hines2004; Jhala et al., Reference Jhala, Gopal, Mathur, Ghosh, Negi and Narain2021). India, one of the tiger range countries, harbours 75% of the global tiger population, primarily in 58 tiger reserves (Qureshi et al., Reference Qureshi, Jhala, Yadav and Mallick2023). Studies on tiger diet across India have shown that medium-sized to large ungulates (> 5 kg) are their preferred prey (Karanth & Sunquist, Reference Karanth and Sunquist1995; Ramesh et al., Reference Ramesh, Snehalatha, Sankar and Qureshi2009; Sankar et al., Reference Sankar, Qureshi, Nigam, Malik, Sinha and Mehrotra2010), including in particular the chital Axis axis, chousingha Tetracerus quadricornis, nilgai Boselaphus tragocamelus, sambar Rusa unicolor and wild pig Sus scrofa. Any reduction in the abundance of these prey species in a tiger reserve will also reduce the carrying capacity of the reserve for a breeding tiger population (Karanth et al., Reference Karanth, Nichols, Kumar, Link and Hines2004).
Long-term population monitoring of prey is vital for assessing predator–prey dynamics and their impact on ecosystems, and for implementing adaptive management (Serrouya et al., Reference Serrouya, McLellan, Boutin, Seip and Nielsen2011; Banerjee et al., Reference Banerjee, Jhala, Chauhan and Dave2013; Hou et al., Reference Hou, Reyna-Hurtado, Omeja, Tumwesigye, Sarkar and Gogarten2021; Lindenmayer et al., Reference Lindenmayer, Lavery and Scheele2022). In the Russian Far East, multi-decadal studies of prey dynamics have revealed that the viability of tiger populations depends not only on the abundance of a single prey species but also on the diversity and stability of ungulate communities (Miquelle et al., Reference Miquelle, Goodrich, Smirnov, Stephens, Zaumyslova and Chapron2010); synchronous declines in prey directly threatened tiger persistence. By quantifying these relationships using long-term data, managers adopted a holistic approach, prioritizing habitat connectivity and anti-poaching measures for both tigers and their prey base, leading to a recovery of ungulate populations and, consequently, tiger numbers (Miquelle et al., Reference Miquelle, Smirnov, Zaumyslova, Soutyrina and Johnson2015). However, few tiger reserves in India use prey numbers to assess the number of tigers that a reserve can potentially support (Dahal et al., Reference Dahal, Amin, Lamichhane, Giri, Acharya and Acharya2023).
The central Indian landscape, spanning the semi-arid zone of Rajasthan state, the central Indian plateau (Maharashtra, Madhya Pradesh, Chhattisgarh and Jharkhand states) and parts of Odhisha and Telangana states, is a global priority landscape for tiger conservation, with > 30% of India’s tiger population (Dutta et al., Reference Dutta, Sharma and Defries2018; Qureshi et al., Reference Qureshi, Jhala, Yadav and Mallick2023). Kawal Tiger Reserve, in northern Telangana, has significant potential to support tiger conservation. Its location offers ecological connectivity to key protected areas such as the Tadoba–Andhari Tiger Reserve in Maharashtra (Yadav et al., Reference Yadav, Tiwari, Mallick, Garawad, Talukdar and Sultan2023). However, apart from occasional transient individuals, Kawal does not have a resident population of tigers despite being one of the larger tiger reserves in India (Qureshi et al., Reference Qureshi, Jhala, Yadav and Mallick2023). The landscape in Kawal is interspersed with villages and agricultural fields and faces intense anthropogenic pressures such as livestock grazing, poaching and habitat fragmentation, which collectively degrade habitat quality and present major challenges to tiger recovery (Jhala et al., Reference Jhala, Mungi, Gopal and Qureshi2025).
In this study we aim to set a realistic target for tiger recovery in Kawal by using prey availability as a key indicator. We estimate the density of principal tiger prey species in the core of the Reserve and assess any trend across seven surveys during 2010–2022, and estimate the carrying capacity of the Reserve for tigers. We also recommend management interventions to enhance Kawal’s potential as a sustainable tiger habitat.
Study area
Kawal lies along the northern banks of Godavari River in Telangana State, India (Fig. 1). The reserve comprises a core area of 893 km2 surrounded by a buffer zone of 1,123 km2. The present core area was previously a wildlife sanctuary and was upgraded to a tiger reserve in 2012, enhancing its protection and infrastructure. The terrain features undulating hills, valleys, plateaus and plains, at altitudes of 152–610 m. The Climate is tropical, with temperature ranging from 5 °C in winter to 43 °C in summer, and a mean annual rainfall of c. 1,040 mm (Siddiqui, Reference Siddiqui2010). The vegetation is classified as tropical teak-bearing dry deciduous forests (Champion & Seth, Reference Champion and Seth1968) dominated by Tectona grandis. Other dominant tree species include Anogeissus latifola, Boswellia serrata, Lannea coromandalica, Madhuca indica, Soymmida febrifuga and Terminalia tomentosa.
The core of Kawal Tiger Reserve, Telangana, India, showing the locations of the line transects and major roads.
The principal ungulate tiger prey species in Kawal are chital, sambar, nilgai, wild pig and chousingha. Other herbivores include the blackbuck Antilope cervicapra, gaur Bos gaurus, chinkara Gazella benettii, and primates include the Hanuman langur Semnopithecus entellus and rhesus macaque Macaca mulatta. Large carnivores in the reserve include the dhole Cuon alpinus, leopard Panthera pardus, sloth bear Melursus ursinus and transient tigers. Mesocarnivores and small carnivores include the golden jackal Canis aureus, honey badger Mellivora capensis, jungle cat Felis chaus and rusty-spotted cat Prionailurus rubiginosus. The tribes living in and around Kawal (Gond, Lambadi, Kollam and Nayakpod) are dependent on agriculture and forest produce.
Methods
Field study design and data collection
We estimated prey density on seven occasions during 2010–2022. We surveyed 28 line transects, each 4 km long, using square sampler design, with sampling locations positioned 4 km apart and transect starting points determined randomly (Buckland, Reference Buckland2001). Surveys were carried out during the summer (March–May) by the same group of observers. Systematic random sampling design was used, with square samplers chosen for their logistical advantages (Karanth et al., Reference Karanth, Nichols, Kumar, Link and Hines2004, Strindberg et al., Reference Strindberg, Buckland and Thomas2004). Two trained observers surveyed the transects during peak animal activity times (6.00–8.30 and 16.00–18.30). Upon detecting prey species, they recorded group size, and the location, perpendicular distance and angle to the centre of the group, using a GPS, range finder and compass, respectively (Karanth et al., Reference Karanth, Thomas and Kumar2002). We surveyed each transect 6–8 times over 1–2 months on each survey occasion.
Data analysis
Herbivore density
Given that tiger prey species mostly occur in groups, the cluster constituted the unit of detection. Individual density was estimated from group density using:
$$\widehat D = C\widehat E\left( s \right)/\widehat p2wL,$$
where
$\widehat D$
is the density of animals, C is the number of clusters detected,
$\widehat E$
(s) is the estimated group size,
$\widehat p$
is the probability of detection, w is the effective half width of the transect, and L is the length of the transect.
For each survey occasion, we followed the analytical framework of Buckland (Reference Buckland2001) and Thomas et al. (Reference Thomas, Buckland, Rexstad, Laake, Strindberg and Hedley2010). Exploratory analysis involved interpreting histograms and initial model fitting to identify issues such as heaping or outliers, using Distance 7.0 (Thomas et al., Reference Thomas, Buckland, Rexstad, Laake, Strindberg and Hedley2010; Supplementary Figs 1–5). Initial analysis involved fitting detection function models (uniform, half-normal, hazard-rate, negative exponential) with varying group intervals and shape criteria (cosine or simple polynomial). Model selection was based on Akaike’s information criterion (AIC) and the Kolmogorov–Smirnov goodness-of-fit test. Heaping issues were mitigated with appropriate grouping, and right truncation was used to eliminate outliers, to reduce the variance.
We used AIC and goodness-of-fit values to select the best model for each species in each survey (Buckland et al., Reference Buckland, Anderson, Burnham, Laake, Borchers and Thomas2004). Using the best-fit model, we estimated species-specific density, group density, mean cluster size, effective strip width and detection probability for each occasion (Awasthi et al., Reference Awasthi, Kumar, Qureshi, Pradhan, Chauhan and Jhala2016). We grouped species-specific herbivore densities across years. In our seven surveys, we gathered an average of ≥ 30 observations per species per year, except for sambar and chousingha, for which detections were < 30 in all years, which is less than the minimum for reliable Distance analysis (Buckland, Reference Buckland2001). Blackbuck, gaur and chinkara were excluded from this analysis as there were < 10 observations of these species. To examine annual changes and trends in ungulate densities, we examined overlap of 95% CIs (Payton et al., Reference Payton, Greenstone and Schenker2003; Cumming & Finch, Reference Cumming and Finch2005) and also used a permutation ANOVA (9,999 samples), which is recommended for analysing wildlife population trends because of its robustness to non-normal data and in the case of small sample sizes (Anderson, Reference Anderson2014; Fieberg et al., Reference Fieberg, Vitense and Johnson2020).
Carrying capacity for tigers
We estimated the carrying capacity for tigers in Kawal’s core area at the time of each of the seven surveys. Firstly, we estimated carrying capacity using the simple two-parameter, one-variable (prey abundance) model described by Karanth et al. (Reference Karanth, Nichols, Kumar, Link and Hines2004):
$${T_K} = A*{P_n}^b,$$
where T K is the number of tigers/unit area, P n is the prey abundance in the same area, A is the proportion of prey population killed each year by tigers, and the exponent b (≤ 1.0), which facilitates a non-linear relationship between prey availability and tiger numbers, for which we used the estimate of 0.514 for several tiger reserves in India (Karanth et al., Reference Karanth, Nichols, Kumar, Link and Hines2004). After estimating the density of tigers per 100 km2, we multiplied it by 8.93 to estimate the potential number of tigers that Kawal’s core area could support on each survey occasion.
Secondly, we used the model of Miquelle et al. (Reference Miquelle, Smirnov, Merrill, Myslenkov, Quigley, Hornocker, Seidensticker, Christie and Jackson1999), which derives the carrying capacity of tigers from available prey biomass using a simple linear equation:
$${T_M} = \left( {{P_b} + 256.3} \right)/456.5,$$
where T M is tiger density (individuals/100 km2), and P b is cumulative prey biomass (kg/km2). Following Hayward et al. (Reference Hayward, Jędrzejewski and Jedrzejewska2012), three-quarters of the adult female body mass was assigned to each prey species to approximate the realized size of kills, accounting for the predation of juveniles and subadults. These standardized prey masses were used to examine size-related patterns in prey use (Hayward et al., Reference Hayward, Jędrzejewski and Jedrzejewska2012; Supplementary Tables 1–14).
The abundance-based approach correlates prey density with tiger numbers through predation rates and scaling factors, which is particularly effective for ecosystems with uniform prey sizes. The biomass-based method evaluates nutritional contributions by accounting for prey body mass differences, which is important in landscapes with both small and large ungulates. This two-model framework addresses the inherent limitations of each approach used in isolation. Abundance models overlook the metabolic demands of prey, and biomass models do not account for hunting success rates. Integrating these methods and cross-validating the results against established tiger–habitat relationships enhances the ecological reliability of the predictions. This robust output is designed to inform adaptive management practices (Miquelle et al., Reference Miquelle, Goodrich, Smirnov, Stephens, Zaumyslova and Chapron2010).
Results
Herbivore density
The chital population in Kawal, one of the major prey species in the landscape, grew from 1.94 individuals/km2 in 2010 to 6.08 individuals/km2 in 2022. The density of wild pigs fluctuated, ranging from 4.99 individuals/km2 in 2020 to 22.82 individuals/km2 in 2013 (Fig. 2). Nilgai, sambar and chousingha populations remained stable throughout the study period. Annual estimates that have overlapping confidence intervals indicate no significant change in density, and this was the case in most years for each species; for example, chital densities in 2010 and 2013 overlap, suggesting no clear change in this period. However, the 2022 estimate for the chital did not overlap with the estimate for 2010, suggesting a real increase (Schenker & Gentleman, Reference Schenker and Gentleman2001), and the lack of overlap of the peak estimate for wild pig 2013 and its lowest estimate, in 2016, suggests a significant decline over this period (Buckland et al., Reference Buckland, Rexstad, Marques and Oedekoven2015).
Estimated densities (± 95% CI) of the five principal large herbivore species, all of which are potential tiger Panthera tigris prey, in Kawal Tiger Reserve, Telangana (Fig. 1), in seven surveys from 2010 to 2022. Only the trends for chital and wild pig are significant (Table 1). Note the differences in the y-axis scales.
Species-level analyses identified significant annual trends for chital (P = 0.034) and wild pig (P = 0.012), consistent with the non-overlapping confidence intervals. Kendall’s W (> 0.85 for both species) confirmed strong temporal consistency in these trends, while accounting for repeated measures (Table 1).
Results of permutation ANOVA (9,999 resamples) and Kendall’s W tests to assess trends in mean ungulate densities ± SE (individuals/km2) from 2010 to 2022, with per cent change and trend.
Estimated prey biomass varied markedly across years, with wild pig and nilgai consistently contributing the greatest biomass, sambar and chital having intermediate and fluctuating biomass, and the biomass of chousingha remaining low throughout the study period (Table 2).
Estimated biomass density (kg/km2, with 95% CI) of the five principal tiger prey in Kawal Tiger Reserve from seven surveys during 2010–2022. To account for the proportion of subadults in the prey population, we corrected the female body weight by a factor of 0.75, following Hayward et al. (Reference Hayward, Jędrzejewski and Jedrzejewska2012), and multiplied it by the respective year’s estimated density of the species (Fig. 2).
Carrying capacity for tigers
The prey abundance-based model predicted a tiger density ranging from 2.67 (95% CI 1.44–5.00) individuals/100 km2 in 2010 to 4.62 (95% CI 2.42–8.84) in 2022. The corresponding estimated tiger carrying capacity ranged from 23.85 (95% CI 12.88–44.60) in 2010 to 41.29 (95% CI 21.63–78.87) in 2022, a 73% increase over this period.
The prey biomass-based model predicted a tiger density ranging from 2.64 (95% CI 1.70–4.40) in 2010 to 3.99 (95% CI 2.40–7.01) in 2022. The corresponding estimated tiger carrying capacity ranged from 23.55 (95% CI 15.19–39.32) in 2010 to 35.66 (95% CI 21.40–62.53) in 2022, a 51% increase (Fig. 3).
The estimated carrying capacity (with 95% CIs) for tigers in Kawal Tiger Reserve, based on prey data from seven surveys during 2010–2022 (Fig. 2). Two models were used to estimate carrying capacity: (a) prey abundance (individuals/km2) and (b) prey biomass (kg/km2; see text for details).
Discussion
Herbivore density
The long-term distance sampling estimates of five wild ungulate species in Kawal across seven survey occasions revealed varying density trends (Fig. 2). The wide annual variations in CIs can be caused by habitat heterogeneity and small sample sizes (≤ 40), which could be addressed with hierarchical modelling and improved sample coverage (Thomas et al., Reference Thomas, Buckland, Rexstad, Laake, Strindberg and Hedley2010).
The implementation of grassland management policies, increased protection and water source development under the Tiger Conservation Plan for Kawal Tiger Reserve could have enhanced habitat quality, likely contributing to the increase in the chital population. However, excessive habitat manipulation, such as increased artificial water sources, can have unintended consequences and negatively affect other species such as chousingha (Owen-Smith, Reference Owen-Smith1996; Sutherland et al., Reference Sutherland, Ndlovu and Pérez-Rodríguez2018). The population of the sambar is lower than in the nearby Tadoba–Andhari Tiger Reserve (density 3.20, 95% CI 1.78–5.75 individuals/km2; Habib et al., Reference Habib, Nigam, Banerjee, Ramgaokar, Annabathula and Jayramegowda2023). The static densities of the nilgai and chousingha during 2010–2022 indicate that management practices in Kawal have not been conducive to an increase in their populations. We recommend that management should take a holistic approach to support the growth of all the Reserve’s antelope populations.
Estimates of the annual density of wild ungulates in Kawal have wide confidence intervals, particularly for wild pig, and therefore need to be interpretated cautiously (Cumming, Reference Cumming2014). Nevertheless, temporal fluctuations around mean density estimates may reflect underlying natural variation or indicate anthropogenic influences. When interpreted carefully, such patterns can provide useful insights into potential pressures and can be of value for designing conservation interventions to strengthen ungulate populations.
Carrying capacity for tigers
Estimated prey densities for 2022 suggest that the core area of Kawal Tiger Reserve could support approximately 35–41 tigers. Although this potential carrying capacity is lower than reported from other reserves in the central Indian landscape (Bandhavgarh, Madhya Pradesh: 132 ± SE 1.07; Kanha, Madhya Pradesh: 105 ± SE 0.49; Tadoba–Andhari Tiger Reserve, Maharashtra: 97 ± SE 0.22; Qureshi et al., Reference Qureshi, Jhala, Yadav and Mallick2023), there is sufficient prey to support the establishment and growth of a recovering tiger population. The high chital density resulted in estimation of a higher tiger carrying capacity by the prey abundance-based model than by the prey biomass-based model, where the greatest contribution was from the larger-bodied sambar and nilgai. Because large-bodied prey provides greater energetic returns for tigers, limited growth in large-bodied species may slow early population recovery despite the presence of multiple ungulate species. This highlights the importance of conserving and enhancing populations of large-bodied ungulates within the Reserve, such as gaur and sambar. Although the two models produced similar estimates of carrying capacity (35–41 tigers), year-to-year variation in ungulate abundance, especially fluctuations in wild pig, affected model outputs, indicating sensitivity to prey species composition rather than substantial differences between model predictions.
Despite being a global priority tiger landscape, the central Indian landscape is highly fragmented, with dense human settlements, a road and rail network with a high volume of traffic, and mines and urban expansion (Dutta et al., Reference Dutta, Sharma and Defries2018; Thatte et al., Reference Thatte, Joshi, Vaidyanathan, Landguth and Ramakrishnan2018). This fragmentation affects tiger movement between reserves and has degraded habitat quality outside the reserves, creating dispersal bottlenecks (Joshi et al., Reference Joshi, Vaidyanathan, Mondol, Edgaonkar and Ramakrishnan2013). The lack of tigers in Kawal could be a result of weak connectivity with nearby source populations. In a separate tiger monitoring programme (authors, unpubl. data), we collected tiger presence data from 2019 to 2025 using the Spatial Monitoring and Reporting Tool software (Cronin et al., Reference Cronin, Dancer, Long, Lynam, Muntifering and Palmer2021). Preliminary analysis suggests tiger movement from Maharashtra to Kawal in the north-east (Tadoba–Andhari Tiger Reserve to Kawal through Kagaznagar–Mancherial Forest Divisions), and in the north (Tadoba–Andhari Tiger Reserve to Kawal through Asifabad Forest Division) and north-west (Tipeshwar to Kawal through Adilabad–Ichoda-Nirmal Forest Divisions). The data indicate frequent tiger movement from Tadoba–Andhari Tiger Reserve to Kagaznagar (Fig. 4), which then connects to Indravati Tiger Reserve in Chhattisgarh (Yadav et al., Reference Yadav, Tiwari, Mallick, Garawad, Talukdar and Sultan2023). The dispersal path to Kawal, however, is potentially hindered by roads (e.g. NH 44), rail (e.g. Nagpur–Hyderabad railway network), dense human habitations, an irrigation project and open-cast mines in KB Asifabad and Mancherial Districts. Furthermore, human–tiger conflict is high in this region, leading to both human and tiger mortality, and negative attitudes towards tiger conservation (Dhanwatey et al., Reference Dhanwatey, Crawford, Abade, Dhanwatey, Nielsen and Sillero-Zubiri2013). Another factor potentially hindering tiger recovery in Kawal is that only two transient females (of 15 tigers) have been recorded entering the core area since 2015, from Maharashtra. However, a breeding tiger population, including three females with cubs and 3–5 additional females, has been documented in adjacent corridors (Siddiqui & Vasudev, Reference Siddiqui and Vasudev2017).
Potential corridors for tiger movement connecting the core of Kawal Tiger Reserve to adjacent source populations, with the locations of direct (camera-trap images, direct sightings, cattle kill) and indirect (pugmarks, scats) tiger signs recorded by field researchers, forest guards and biologists during 2019–2024.
Conclusion
Long-term prey monitoring is vital, serving as a key ecological indicator for conservation planning and for setting realistic recovery goals for the tiger. However, our findings show that tiger recovery in Kawal is not principally contingent on prey recovery. Rather, it depends on strengthening corridor connectivity and mitigating human–wildlife conflicts (Naves et al., Reference Naves, Wiegand, Revilla and Delibes2003; Jhala et al., Reference Jhala, Gopal, Mathur, Ghosh, Negi and Narain2021). The corridors connecting Kawal with nearby source populations need to be managed holistically, involving local communities and local governing bodies such as Gram-Sabha (village governing bodies). The c. 30 villages within Kawal’s core area rely on forest resources, and this could impact wildlife (Yadav et al., Reference Yadav, Tiwari, Mallick, Garawad, Talukdar and Sultan2023). Integrating community needs through eco-development, revenue sharing, alternative income sources and timely compensation for wildlife damages is crucial for conservation success. The highways and rail tracks that pass through corridors (especially those through Kagaznagar Forest Division) require overpasses or underpasses to facilitate tiger movement (Habib and Saxena, Reference Habib and Saxena2024).
Livestock grazing, which can lead to disease transmission in shared resources such as water bodies and grasslands (Caron et al., Reference Caron, Miguel, Gomo, Makaya, Pfukenyi and Foggin2013), remains an issue in Kawal. Free-ranging dogs, coupled with their use by local communities for hunting, alongside snaring, pose a significant threat to wildlife and can severely undermine species recovery efforts (Vanak & Gompper, Reference Vanak and Gompper2009; Doherty et al., Reference Doherty, Dickman, Nimmo and Ritchie2015; Groenenberg et al., Reference Groenenberg, Crouthers, Yoganand, Banet-Eugene, Bun and Muth2023). Designation of grazing land and strict monitoring of hunting could reduce these threats. To attain a target of > 30 tigers in Kawal (c. 20 breeding females; Bisht et al., Reference Bisht, Banerjee, Qureshi and Jhala2019) may require assisted dispersal of tigers, particularly females (Ash et al., Reference Ash, Cushman, Kaszta, Landguth, Redford and Macdonald2023), under the National Tiger Conservation Authority’s guidance.
Author contributions
Study design, concept, fieldwork: IS, VA; data analysis: IS; writing: NB, IS; scientific input, technical support: KB, JLK.
Acknowledgements
We thank K. Ullas Karanth, N. Samba Kumar and Devcharan Jathanna for providing technical input and assisting in study design and data collection during the initial years of our study; the National Centre for Biological Sciences, Bengaluru, India, for supporting us in the first year; the Telangana State Forest Department for granting permission to conduct our research; our field staff, including K. Shankar, supported by Wildlife Conservation Society–India staff G. Harshvardhan, M. Vasmshi, Chenduraju, Preethi Sridharan and Jakka Amarnath, and volunteers from Hyderabad Tiger Conservation Society; and the Wildlife Conservation Society, Gland Fosun Foundation, and the U.S. Fish & Wildlife Service for their financial support.
Competing interests
None.
Ethical standards
This study followed the Oryx guidelines on ethical standards and research standards in India, including appropriate field protocols and research permissions. Permissions were obtained from the Telangana State Forest Department to conduct field surveys in Kawal. Data collection methods were non-invasive and did not involve directly handling animals. Observers maintained a safe distance from animals to prevent stress or behavioral changes. All field staff were trained to handle wildlife encounters responsibly and to avoid disrupting natural processes during data collection.
Supplementary material
The supplementary material for this article is available at doi.org/10.1017/S003060532510224X
Open access
