Abstract
Introduction:
Understanding how apex predators allocate space-use intensity within their home ranges is essential for linking individual-level movement decisions to broader ecological and conservation management. This study presents the first investigation of recursive movement ecology in tiger from the Indian subcontinent, examining the environmental determinants of site revisitation frequency and patch residence time in the tropical dry deciduous forests of Panna Tiger Reserve, central India.
Methods:
Using continuous GPS telemetry data of 12 resident tigers across three seasons (summer, monsoon, winter), recursive-use patches were quantified at a 250-m analytical neighbourhood, and revisitation rate and residence time were modelled separately using mixed-effects fine-scale habitat-use models.
Results:
Revisited sites were non-randomly distributed with respect to all measured covariates, although effect sizes differed: tree canopy height produced the strongest univariate separation between revisited and available sites, whereas distance to villages and water showed intermediate separation, while enhanced vegetation index and topographic position showed weaker separation. Females had consistently higher descriptive revisitation and residence-time values than males, and both metrics were highest in summer and lowest during monsoon. The mixed-effects models revealed persistent sex- and season-specific non-linearity. Topographic position showed the most consistent negative linear association around average conditions, but the complete response curves demonstrated cohort-specific curvature. Associations with water, vegetation greenness, canopy structure, and village distance differed between revisitation and residence time, showing that frequently returned-to sites were not necessarily the places where tigers remained longest. Fixed environmental effects explained a modest proportion of variation relative to models including tiger identity, emphasizing substantial individual heterogeneity.
Discussion:
These results identify recursive space use as a distinct dimension of tiger habitat use and show that fine-scale site fidelity in dry deciduous forests reflects context-dependent associations with structural vegetation, hydrology, topography, and human-use interfaces rather than a single habitat optimum.
Introduction
Animal movement data offer a robust framework for comprehending not only the spatial locations occupied by individuals within a landscape, but also the trajectories they follow, the behavioral states they exhibit, and the environmental factors influencing their habitat use and movement decisions (; ). Within their home ranges, animals rarely utilize space uniformly; rather, they exhibit recurrent visitation to specific locations, characterized by structured and predictable patterns. This generates recursive space-use patterns, indicative of long-term landscape familiarity (, ; ). Such repeated utilization of familiar areas confers several advantages, including reduced uncertainty in novel environments, mitigated predation and competitive pressures, and avoidance of energetically costly terrains. Consequently, it optimizes the trade-off between energy expenditure and resource acquisition (; ).
Revisited sites are often of considerable ecological significance, as they represent locations consistently providing essential resources, such as productive foraging sites, water sources, dens, nests, resting areas, and other functionally critical habitat features (). The dynamics of site revisitation are closely associated with the tenets of Optimal Foraging Theory, which posits that animals optimize their time and movement allocation to maximize net energetic gain while minimizing the costs of resource search and exploitation (). Within this theoretical construct, the repeated utilization of specific locations is interpreted as an adaptive response to spatial heterogeneity in resource availability. Similarly, the quantification of recursive movement patterns facilitates the identification of these ecologically relevant sites and enables inferences regarding the underlying behavioral and environmental processes driving their repeated utilization, encompassing responses to seasonal resource fluctuations (). However, high recursive space use does not necessarily indicate resource-rich habitat alone; it may also identify familiar locations retained through spatial memory (), recurrent territorial or scent-marking nodes (), patches with predictable prey encounters (), or movement bottlenecks created by landscape resistance (). Alternatively, empirical investigations have shown that recurrent high-use areas and extended residence times result from the synergistic interplay of habitat characteristics, interspecific interactions, and temporal activity patterns (). Spatial variations in revisitation frequency and visit duration are frequently correlated with localized habitat conditions and competitor presence, suggesting that fine-scale niche partitioning can manifest through the repeated, non-random reuse of particular locations. Furthermore, visitation frequency serves as a recognized proxy for site importance within an individual’s home range, as resource-rich and energetically favorable patches generally exhibit more frequent and sustained use compared to less suitable areas (; ).
Residence time represents another valuable metric, quantifying the temporal duration an individual allocates to a specific patch during each visit. This metric provides an estimation of patch profitability and exploitation intensity, having long been recognized as a foundational component of foraging theory (; ). According to patch-use models, animals are expected to prolong their tenure in locations where the rate of resource acquisition exceeds the anticipated gain from relocating. Recent developments in movement ecology have extended this perspective to encompass broader spatio-temporal scales, emphasizing that animals frequently track resources exhibiting temporal fluctuations but predictable recurrence at specific, enduring sites (). As a result, a combined analysis of revisitation frequency and residence time offers a robust framework for identifying locations that are recurrently utilized due to their consistent provision of ecological advantages across both short and long temporal scales.
The return of animals to regenerating or seasonally renewed resource patches may be mediated by spatial memory, which facilitates the retention of information regarding the location and quality of previously advantageous areas (). Alternatively, recurrent site utilization could stem from the active selection of habitats, guided by environmental cues that reliably signal resource availability or favorable conditions (). Memory-dependent movement and habitat selection are increasingly acknowledged as synergistic mechanisms influencing recursive space utilization, particularly relevant for species thriving in dynamic and heterogeneous environments. Nevertheless, there remains a notable scarcity of information concerning the precise manner in which specific environmental attributes mediate both the frequency of site revisitation and the duration of animal occupancy within those sites. This knowledge gap is especially relevant in the context of large carnivores, whose movement decisions are influenced not solely by prey distribution but also by interspecific competition, anthropogenic disturbance, and energetic constraints imposed by terrain. Tiger (Panthera tigris) constitutes the dominant apex predator within the dry deciduous forests of central India, an area where its spatial ecology has been subjected to extensive investigation. Prior research has delineated broad-scale patterns of habitat utilization, indicating a preferential selection of forested habitats by tigers and a general avoidance of agricultural landscapes and human settlements (, ; ; ; ). This habitat selection is further modulated by factors such as prey abundance, water accessibility, topographical complexity, and the spatial distribution of competing carnivores, collectively leading to highly structured patterns of space utilization within their respective home ranges.
Over the past decade, revisitation ecology has emerged as an established analytical framework for understanding the spatial organization of animal behavior across diverse taxa, including large herbivores (Seidel et al., 2019; ) and carnivores (; ). Nevertheless, a dedicated investigation of recursive space use and revisitation behavior in tigers remains scarce, and has not previously been applied for tigers in the Indian subcontinent. Traditional occurrence or location-based analyses generally give equal analytical weight to presence records and therefore do not distinguish sites that are returned to repeatedly from sites occupied only briefly (). This methodological approach consequently overlooks the ecological significance of sites characterized by repeated visitation or prolonged occupancy. In contrast, revisitation metrics offer a method for quantifying site importance by explicitly integrating both visit frequency and duration, thereby establishing a direct link between movement decisions and underlying resource dynamics. To address this recognized knowledge gap, the present study investigated the environmental determinants of revisited sites within tiger home ranges, addressing three principal questions: (1) whether revisited locations exhibit non-random distribution patterns in relation to environmental conditions; (2) whether revisitation rate and residence time responded differently to the same covariates; and (3) how these relationships changed between sexes and seasons. We predicted that recurrent-use sites would be non-randomly associated with structural vegetation, hydrological, topographic, and anthropogenic gradients; that revisitation and residence time would not show identical responses because they represent different behavioral dimensions; and that seasonal changes in water and vegetation would alter the strength or shape of these associations. Lastly, we expected the underlying environmental characteristics of intensely used sites to exhibit relative consistency, thereby reflecting the persistent ecological attributes that confer disproportionate importance to these locations within the tiger’s home range.
Materials and methods
Study area
Panna Tiger Reserve (PTR) is situated within the Vindhyan landscape of central India, encompassing a core area of 542 km² and a buffer area of 1,032 km². Additionally, the reserve incorporates the contiguous Gangau Wildlife Sanctuary (33.47 km²) and spans the Panna, Chhatarpur, and Damoh districts of Madhya Pradesh (Figure 1). Elevations within PTR range from 164 to 555 meters above sea level (MASL), characterized by a rugged topography comprising plateaus, deep gorges, rocky escarpments, and caves. The lowest elevations are found within the Ken River basin, which constitutes the sole perennial water source within the reserve. This river traverses approximately 55 km through the protected area, flowing in a south-north direction (). The vegetation in PTR is classified as tropical dry deciduous forest, broadly corresponding to Group 5: Tropical Dry Deciduous Forests as per . This vegetation type is predominantly composed of tropical dry deciduous forests, with dominant species including Teak (Tectona grandis), Tendu (Diospyros melanoxylon), and Kardhai (Anogeissus pendula). These formations are interspersed with mixed forests, localized bamboo patches, grasslands, and riparian vegetation along the Ken River. The region experiences a subtropical dry climate characterized by three distinct seasons. The hot and dry summer (March-June) exhibits mean maximum temperatures up to 45 °C and relative humidity between 15-20%. This is followed by a relatively humid monsoon season (July-October), which receives an average annual precipitation of approximately 1,100 mm. The cool, dry winter (November-February) is marked by mean minimum temperatures that can decline to 5 °C, with average relative humidity around 15%. Within this ecosystem, tiger operates as the apex predator, while the Leopard (Panthera pardus) is present as a sympatric co-predator. The carnivore guild additionally comprises several meso-carnivores and omnivores, including the Golden Jackal (Canis aureus), Striped Hyena (Hyaena hyaena), Dhole (Cuon alpinus), Indian Wolf (Canis lupus), Indian Fox (Vulpes bengalensis), and the Sloth Bear (Melursus ursinus). The primary ungulate prey base consists of Chital (Axis axis), Sambar (Rusa unicolor), Nilgai (Boselaphus tragocamelus), Chinkara (Gazella bennettii), Four-horned Antelope (Tetracerus quadricornis), and Wild Boar (Sus scrofa). The buffer zone of PTR presents a heterogeneous mosaic landscape, characterized by forest patches interspersed with agricultural fields and human settlements. Within the buffer boundary of the reserve, a total of 63 villages is situated, collectively supporting an estimated human population of approximately 30,000 individuals.
Figure 1
Data collections
Radio telemetry
The telemetry data used in this study were collected as part of the long-term “Tiger Reintroduction and Recovery Programme” in PTR from 2009-2020. As part of this program, 25 tigers (six male, 19 female) were collared under permit by the Madhya Pradesh Forest Department (Permit No. Exp./2009/1205, dated August 31, 2009), under Section 11(1A) of the Indian Wildlife (Protection) Act, 1972. Each animal was chemically immobilized by a skilled wildlife veterinarian and fitted with a radio collar using the Hellabrunn mixture (125 mg xylazine and 100 mg ketamine/ml; ), projected with the Tele-Inject (Model 4V.31). Individuals were fitted with GPS and VHF collars manufactured by Telonics® Inc, African Wildlife Tracking® Inc and Vetronic Aerospace®. GPS collars were programmed to acquire locations at hourly intervals, whereas VHF-collared animals were monitored intensively by trained field staff using vehicle-based radio tracking. Of the 25 tigers collared during the long-term monitoring program, 12 resident individuals (three males and nine females, including founder individuals; ) met the predefined inclusion criteria for temporal continuity and seasonal coverage. These individuals contributed repeated seasonal datasets comprising 18 male*summer, 15 male*monsoon, 17 male*winter, 41 female*summer, 37 female*monsoon, and 33 female*winter individual–season records (Supplementary Table S1). Data were categorized into three seasons based on habitat and available resources, and data were only considered if at least 60 consecutive days of contiguous tracking occurred during each season.
Analytical methods
Missing locations
While the GPS collars were programmed to obtain location fixes at hourly intervals, missing location fixes are an intrinsic characteristic of wildlife telemetry data as GPS signals are interrupted by canopy cover, topographical features, temporary satellite failure or equipment malfunction (). As movement data typically needs to be regularly sampled for most modern analytical procedures (), we used the ctmm package in R (Fleming and Calabrese, 2023; ) for estimation of locations based on the models of movement with estimated states. Individual-specific continuous-time movement models were fitted so that predicted locations were conditioned on each animal’s estimated movement process and temporal autocorrelation structure. This approach preserves individual variation in movement behavior and avoids the assumptions of deterministic straight-line interpolation.
Predictor variables
To explore the factors influencing recursive site use, we developed spatial variables representing resource presence and anthropogenic influence. We digitized major water resources in PTR (perennial and seasonal streams and rivers, lakes, artificial and natural waterholes) for each season separately based on availability of water, creating Euclidean distance raster surfaces for accessibility to a critical resource in the dry deciduous ecosystem. Seasonal vegetation productivity measures were developed using Enhanced Vegetation Index (EVI; derived from MODIS/061/MOD13Q1) calculated via Google Earth Engine for the year 2016. EVI is a well-used measure of photosynthetic activity that often correlates with prey availability and foraging success in a region (), however in tropical dry deciduous forests such as PTR, vegetation phenology undergoes marked seasonal change due to a strong dry season with associated canopy drop. To characterize vertical forest structure, we included Tree Canopy Height (TCH), derived from the Global Forest Canopy Height dataset (). TCH was treated as a structural habitat descriptor rather than a direct measure of habitat quality, as taller-canopy patches may also reflect differences in canopy continuity, microclimatic buffering, vegetation structure, moisture conditions, and disturbance history. Terrain variability was quantified using the Topographic Position Index (TPI) developed from the SRTM Digital Elevation Model (USGS/SRTMGL1_003) via Google Earth Engine which helps delineate slopes and areas of differing elevation. Anthropogenic influence was estimated as distance from village settlements and agriculture patches (digitized from a land-use and settlements layer) using a Euclidean distance raster. These measures have proven to effectively represent anthropogenic gradients in terms of their impact on the use of space by large carnivores. All environmental raster were projected to a common coordinate reference system. For each used and available location, covariate values were summarized as the mean within a 250-m-radius circular neighborhood, matching the spatial radius used to define revisitation sites. Habitat associations were therefore evaluated at the scale of the revisitation neighborhood rather than at the resolution of individual raster cells. The extracted covariates were subsequently z-standardized before model fitting. To evaluate potential multicollinearity among predictors, we calculated Pearson’s pairwise correlation coefficients using the corrplot package in R (; ), and excluded combinations exhibiting strong correlation (|r| ≥ 0.70) following the recommendations of ; Supplementary Figures S1–3). All spatial layers were prepared in R using terra package ().
Revisitation analyses
The quantification of site revisits requires a defined spatial scale on which sites are established. Recursion methods define sites as points from which a circle of specified radius is drawn, and the algorithm records revisits as points inside a circle after leaving. The choice of circle radius affects the recorded revisitation values: a small circle will under-estimate revisitation rates by only considering limited spatial locations, while a large circle will lose spatial information by flattening unique landscapes (). For this study, we tested multiple radius options (100, 200, 250, 300, 400, and 500 m) and chose the radius that yielded the highest variance in revisit frequency, which we found to be 250 m. We used this radius in all analyses using the recurse package in R (; ; Supplementary Figure S4). This radius is interpreted as an analytical scale for identifying localized recursive-use patches, rather than as a direct representation of a specific tiger behavior such as hunting, resting, or territorial marking. Further, we used an inter-visit time of 12 h to limit recordings to biologically relevant revisits, meaning only times when an individual stayed outside of a circle for more than 12 hours before returning to it counted as a revisit (; ). Analyses were conducted separately for each sex and each season to detect potential variations.
Since there was variance in monitoring lengths between season-specific datasets (between 3-4 months), revisitation frequencies were normalized to the number of tracked days for each individual/season to account for differing lengths, generating a per-day revisit rate for each site. As there was a circle drawn for each location fix, a very large proportion of sites had very low revisitation frequency and a few were highly visited by individuals, leading to a highly right-skewed dataset. For subsequent analyses, we identified the best error distribution for fitting data using fitdistrplus package (; ) and determined that in most cases, the normal, log-normal, Weibull, and gamma distributions were reasonable, but the log-normal distribution often best fits data, therefore we used this and log-transformed revisitation rates for modeling (Supplementary Table S2; Supplementary Figures S5, S6).
Environmental effects on site selection (randomness)
To assess whether revisitation to a given tiger habitat occurred non-randomly with respect to landscape characteristics, a two-stage statistical analysis was applied. We defined spatial extents for available landscape by the 95% AKDE home range contour (generated to accommodate for directional autocorrelation and to avoid underestimation of space use as occurs with traditional kernel methods ) for each individual tiger. We generated a random 1:10 (used: available) distribution of background points within each respective 95% AKDE polygon of each tiger to maintain a robust statistical comparison. We extracted environmental data around both used (assigned 1) and available (assigned 0) coordinate data points at a radius of 100 m around each data point. We used a 2-sample Kolmogorov-Smirnov test (KS test; ) to evaluate the null hypothesis, the KS test is a non-parametric statistic which quantifies the probability that the two samples were drawn from the same distribution and assesses the distance between cumulative probability functions. Significant p-values (p<0.05) would cause us to reject the null hypothesis; the D-statistic from the KS test was then used as a measure of effect size, to rank how the different environmental covariates influence the revisitation frequency at a habitat. All spatial and statistical analysis was carried out using R ().
Fine-scale site use modeling
High revisitation frequency is widely interpreted as an indicator of site fidelity and repeated use of ecologically important locations (; ; ). To examine fine-scale variation in repeated site use within individual home ranges, we modelled revisitation intensity as a continuous response rather than classifying locations simply as used or available. This approach allowed highly revisited sites to be distinguished from locations receiving only occasional use. The log-transformed revisitation rate was fitted as the response variable using Generalized Linear Mixed-Effects Models (GLMMs) in the lme4 package in R (), with environmental covariates included as fixed effects and individual tiger identity specified as a random intercept to account for repeated observations. Residence time was analyzed separately as a complementary response representing the duration of occupancy at revisited sites.
Results
Revisitation analyses
The recursive analysis of 12 residential tigers revealed distinct seasonal and sex-specific patterns in site revisitation and patch residence time in PTR. Across all seasonal strata, female tigers exhibited consistently higher revisitation rates and prolonged patch residence times compared to males (Table 1). Revisitation frequency was highest during the summer for both sexes (♀:11.14; ♂:5.70) and reached a seasonal low during the monsoon (♀: 8.88; ♂: 3.68), while moderate revisitation during winter (♀:9.43; ♂:5.3). A similar pattern was observed in patch residence dynamics, females maintained greater site occupancy, highest during the summer (♀: 14.36 h), whereas males exhibited reduced residence durations, with summer values (♂: 9.80 h) remaining significantly lower than their female counterparts. Females spend more times (♀ winter: 13.78 h; ♀ monsoon: 13.44 h), compared to males(♂ winter: 7.45 h; ♂ monsoon: 8.31 h; Supplementary Table S3; Supplementary Figures S7, S8).
Table 1
| Cohort | Behavioral metric | Mean value | Standard error | Range |
|---|---|---|---|---|
| Male: Summer | Patch revisitation rate (counts) | 5.7 | 0.008 | 1 - 28 |
| Patch residence time (hrs) | 9.80 | 0.025 | 0.002 - 175.26 | |
| Female: Summer | Patch revisitation rate (counts) | 11.14 | 0.007 | 1 - 53 |
| Patch residence time (hrs) | 14.36 | 0.02 | 0.0001 - 631.44 | |
| Male: Monsoon | Patch revisitation rate (counts) | 3.68 | 0.007 | 1 - 21 |
| Patch residence time (hrs) | 8.31 | 0.03 | 0.00007 - 135.47 | |
| Female: Monsoon | Patch revisitation rate (counts) | 8.88 | 0.009 | 1 - 49 |
| Patch residence time (hrs) | 13.44 | 0.035 | 0.00004 - 654.91 | |
| Male: Winter | Patch revisitation rate (counts) | 5.3 | 0.01 | 1 - 34 |
| Patch residence time (hrs) | 7.45 | 0.02 | 0.0001 - 106.98 | |
| Female: Winter | Patch revisitation rate (counts) | 9.43 | 0.008 | 1 - 47 |
| Patch residence time (hrs) | 13.78 | 0.027 | 0.00001 - 295.59 |
Summary of revisitation frequency and residence time for male and female tigers across seasonal cohorts in PTR.
Environmental effects on site selection (randomness)
To evaluate the hypothesis that these concentrated usage patches occur non-randomly, the Kolmogorov-Smirnov (KS) tests results yielded significant differences across all tested variables (p < 0.001), rejecting the null hypothesis of random selection (Table 2). However, the magnitude of separation varied considerably among variables. TCH showed the largest and most consistent D-statistics across all sex-season cohorts (D = 0.370-0.453), indicating the strongest univariate distinction between revisited and available sites. Distance to villages and water showed intermediate separation, generally stronger in males, whereas EVI and TPI had comparatively small D-values despite statistical significance. These patterns suggest that canopy structure, followed by anthropogenic and water-related gradients, contributed most strongly to the observed differences in revisited-site characteristics.
Table 2
| Season | Gender | Predicted variables D statistics (p values) | ||||
|---|---|---|---|---|---|---|
| Dist. to water | Dist. to village | EVI | TCH | TPI | ||
| Summer | Male | 0.189 (<0.001) | 0.266 (<0.001) | 0.037 (<0.001) | 0.391 (<0.001) | 0.070 (<0.001) |
| Female | 0.085 (<0.001) | 0.135 (<0.001) | 0.030 (<0.001) | 0.370 (<0.001) | 0.085 (<0.001) | |
| Monsoon | Male | 0.244 (<0.001) | 0.315 (<0.001) | 0.034 (<0.001) | 0.430 (<0.001) | 0.047 (<0.001) |
| Female | 0.177 (<0.001) | 0.142 (<0.001) | 0.032 (<0.001) | 0.375 (<0.001) | 0.029 (<0.001) | |
| Winter | Male | 0.198 (<0.001) | 0.263 (<0.001) | 0.042 (<0.001) | 0.453 (<0.001) | 0.049 (<0.001) |
| Female | 0.177 (<0.001) | 0.139 (<0.001) | 0.035 (<0.001) | 0.381 (<0.001) | 0.056 (<0.001) | |
Results of Kolmogorov-Smirnov (KS) tests comparing environmental covariate distributions between revisited and available locations.
Fine-scale site use modeling
Fine-scale site use varied markedly between revisitation and residence time and was strongly contingent on sex and season (Figure 2; Supplementary Tables S4, S5). The most consistent pattern was associated with TPI: the linear TPI coefficient was negative for every sex*season cohort and for both response variables, indicating that, around average environmental conditions, repeated use and prolonged occupancy were more strongly associated with relatively lower topographic positions than with elevated terrain. Associations with water were more context dependent. Revisitation generally declined with increasing distance from water in females during summer, both sexes during monsoon and winter, whereas males in summer showed a weak positive linear association. Most water models also contained appreciable quadratic components, demonstrating that the relationship changed across the water-distance gradient rather than remaining uniformly linear. Canopy structure showed a clear seasonal shift in revisitation: TCH was positively associated with revisitation during summer and monsoon but became negative in winter for both sexes. Residence time was generally positively associated with TCH, although the linear effect was not clearly supported for males in winter, and the direction of the quadratic term varied among cohorts. Village-distance responses also differed between the two behavioral metrics. Revisitation increased with distance from villages in both summer cohorts, monsoon females, and winter males, but decreased in monsoon males and winter females. In contrast, residence time increased with distance from villages during summer but declined during monsoon and winter, with the strongest association observed for winter females (β=-0.122), indicating longer residence closer to villages within the sampled gradient. EVI was positively associated with revisitation in most cohorts, except females in monsoon, while residence time responses were more variable; the strongest positive association occurred in winter females (β=0.106).
Figure 2
Quadratic terms were supported for most predictor by cohort combinations, confirming that many environmental relationships changed in strength or direction across the observed gradients (Supplementary Tables S6, S7; Supplementary Figures S9–14). The accompanying response curves were therefore used to interpret the fitted models; mathematical vertices were treated as descriptive turning points and not as ecological thresholds or optima, particularly when they occurred near the edge of the data range. Model diagnostics showed no evidence of unusual dispersion across cohorts (dispersion ratios 0.999 to 1.000, all p≥0.86; Supplementary Tables S4, S5; Supplementary Figures S15–17). Additionally, marginal R2 values remained modest for both revisitation (0.009 to 0.041) and residence time (0.006 to 0.039), whereas conditional R2 values were higher (0.125 to 0.204 and 0.055 to 0.136, respectively), indicating that individual-level heterogeneity accounted for considerably more variation than the measured environmental covariates alone.
Discussion
The central finding of this study is not that a single habitat attribute determines recursive space use, but that return frequency and time spent express different ways in which tigers repeatedly use places within their home ranges. Revisitation identifies locations that enter an animal’s movement sequence repeatedly, whereas residence time identifies locations at which movement is locally concentrated for longer periods (; ). The two metrics therefore need not identify the same sites or respond in the same direction to an environmental gradient. This distinction was clear in the cohort-specific curves: several covariates were associated with frequent returns but not with long stays, or with long stays only during particular seasons. Such divergence is consistent with recurrent-use sites performing multiple functions, including route junctions, territorial or scent-marking locations, access points to resources, resting areas, or patches used intensively on fewer occasions (; ). Our data do not distinguish these behaviors directly, but modeling both responses separately avoids assigning a single ecological meaning to all high-use locations. Comparable separation between visit frequency and duration has been reported for reintroduced African lions, where fine-scale use revealed features of home-range organization (). Similarly, jaguars differed in how often they revisited landscape edges and how long they remained in forests, agricultural areas and near roads (). Recursive movement therefore offers more than a map of concentrated use; it reveals how different places are incorporated into the behavioral organization of a home range.
Revisitation patterns and environmental non-randomness
Females consistently showed higher mean revisitation rates and longer residence times than males in all three seasons, while both sexes reached their highest values during summer and their lowest revisitation rates during the monsoon. The consistency of this pattern suggests that females concentrated their movements within a more repeatedly used network of sites. Such a pattern agrees with the socio-spatial organization of tiger populations. Female home ranges are generally arranged around dependable access to prey, secure resting areas and the energetic demands of reproduction, whereas males occupy larger territories that commonly overlap the ranges of several females and require more extensive territorial movements (; ; ; ). Females may therefore return more often to a comparatively stable set of familiar patches, while males distribute their movements across a broader system of routes, boundaries and female territories. The present result is best understood as a population-level expression of these contrasting spatial strategies rather than as evidence for one specific behavior. Reproductive status, cub presence, kill sites and scent-marking were not measured, but the repeated appearance of the sex difference across seasons gives the pattern clear ecological relevance. Similar sex-specific adjustments occur in other large felids: female lions in a Tanzanian pastoral landscape showed more consistent avoidance of people and stronger reliance on vegetation cover near settlements than males, illustrating how sex can modify the way environmental risk is negotiated within a shared landscape ().
The seasonal pattern suggests that the concentration of recursive use was reorganized as the dry-deciduous landscape changed through the year. Summer in PTR brings extreme temperatures, widespread leaf fall and progressive contraction of surface water. These changes are likely to sharpen contrasts among habitat patches, so that a limited number of shaded, water-linked or structurally secure locations become disproportionately important. Seasonal changes in prey distribution and human activity around remaining water sources may further alter which sites are repeatedly checked and where tigers remain for longer periods. In the monsoon, water and green vegetation become more widely distributed, potentially reducing the need to return to a small number of spatially restricted patches. At the same time, dense regrowth, flooded drainage lines and redistribution of prey may redirect movement through a different set of routes (; ; ). The decline in monsoon revisitation therefore need not imply reduced habitat value. It is more consistent with a diffusion of use across a temporarily more homogeneous resource landscape, whereas summer appears to concentrate recursive movement around fewer locations. This seasonal reorganization parallels findings from African savannas, where lion use of cover, water and human-used areas changes with season and local risk, and from European brown bears, whose movement and space use vary with resource phenology, human pressure and reproductive class rather than following a single annual habitat rule (; ).
The KS comparisons provide a complementary view of this spatial organization. Revisited sites were environmentally non-random, but the magnitude of separation from available locations differed markedly among covariates. TCH produced the largest D-statistics in every sex*season cohort, distances to villages and water showed intermediate separation, and EVI and TPI showed comparatively modest univariate differences. Given the large number of telemetry-derived locations, the D-statistic carries greater ecological meaning than the uniformly small p-values because it describes the magnitude of separation between the used and available environmental distributions. The result indicates that recursive-use sites were not simply a representative sample of each home range; they occupied a recognizably different part of environmental space.
The KS comparisons and mixed-effects models nevertheless describe different stages of habitat organization. The KS analysis asks whether the environmental settings of revisited locations differ from those available within individual home ranges. The continuous models then ask why some recursive-use locations are revisited more frequently or occupied for longer than others after accounting for tiger identity and the simultaneous effects of the measured covariates. TCH can therefore be the strongest univariate discriminator of where recursive-use sites occur without showing the same directional association with how intensively each site is used. Rather than being inconsistent, these results suggest a hierarchical process. Forest structure helps establish the environmental template within which recursive-use patches occur, while water, productivity, topography, anthropogenic proximity, season and individual movement history influence the frequency and duration with which particular patches are used.
TCH should consequently be interpreted as a structural signature of recursive habitat rather than as an isolated ecological mechanism. Taller-canopy stands may differ from shorter vegetation in canopy continuity, stand maturity, moisture retention, understory development, disturbance history and near-ground microclimate. These associated properties can jointly influence concealment, movement conditions, prey accessibility and the suitability of resting locations. Studies in Sumatra found that tigers spent time in areas with high vegetation cover and less human disturbance, while earlier study in PTR identified dense-canopy, riverine forest, bamboo-associated and mixed-forest habitats as important components of fine-scale tiger selection (; ). Similar context dependence is evident elsewhere. Lions in the Ngorongoro Conservation Area relied more strongly on vegetation cover when they were close to people, whereas jaguars in South America spent longer in large natural forest patches and moved more slowly through drainage systems (; ). Across Europe, brown-bear space use is also reduced where human footprint and forest disturbance are high, demonstrating that vegetation structure and disturbance often operate together rather than as independent gradients (). These comparisons support the interpretation that canopy height in PTR is a major feature of structural conditions repeatedly incorporated into tiger movement. The key result is not that one canopy height constitutes an optimum, but that recurrent-use sites differed consistently from availability along the vertical-structure gradient.
Fine-scale site use modeling
The mixed-effects response curves add a more behaviorally specific layer to this broad environmental separation. Many relationships were curved but relatively broad, with gradual changes in slope rather than narrow peaks. A significant quadratic coefficient should therefore be read as evidence that the effect of a covariate changes across its observed range, not automatically as proof of a sharply defined ecological optimum. This interpretation is particularly important where the mathematical vertex lies near the edge of the sampled gradient or where uncertainty increases toward the extremes. The response curves are most informative as descriptions of how predicted recursive use changes within the environmental conditions actually represented by the 250-m neighborhoods.
Topography and water together formed the most coherent spatial template, although neither produced a universal monotonic response. The negative linear TPI coefficient in every sex*season cohort indicates a common tendency, around average environmental conditions, for revisitation and residence time to decline as locations shifted toward relatively higher topographic positions. In PTR, lower positions often correspond to drainage lines, valley bottoms, riparian vegetation, moisture-retaining areas and relatively gentle movement routes. These features may combine easier travel with vegetation cover and predictable movement opportunities for prey. TPI therefore appears to capture a landscape complex rather than elevation alone. The complete curves, however, included U-shaped responses and peaks at low or intermediate TPI values, indicating that topographic effects changed toward the ends of the sampled gradient. The result is best expressed as a broad affinity for lower landscape positions, modified by seasonal and the behavioral metric being considered. Comparison with European lynx in the Slovenian Dinaric Mountains and the Bohemian Forest selected rocky outcrops, steep slopes and rugged terrain, especially for daytime resting, while using linear infrastructure differently during active night-time movement (). It illustrates that terrain features acquire value through the opportunities they provide to a particular predator in a specific landscape. Rugged relief can offer lynx secure resting and stalking locations, whereas lower drainage-associated terrain in PTR may provide tigers with connected cover, accessible movement routes and seasonally persistent moisture. The contrast reinforces an important interpretation of the present analysis: topographic variables do not carry a fixed ecological meaning across carnivores. Their effect emerges from the interaction between morphology, hunting mode, behavioral state and landscape structure.
Distance to water was still more contingent. The curves do not support a simple seasonal reversal in which tigers universally avoided water in summer and selected it in winter. Summer responses for both sexes were generally U-shaped, with lower predicted use at intermediate distances and greater use toward one or both ends of the sampled gradient. Winter female revisitation declined strongly with increasing distance to water, and male residence time showed a similar decline, while other responses remained curved. Monsoon patterns again differed between revisitation and residence time. These shapes suggest that water-associated habitat is spatially heterogeneous. A water source may provide drinking access, riparian cover and prey encounters, but it may also attract people, livestock, vehicles (tourism) or competing carnivores. Locations farther away can still be functionally connected to water through shaded resting cover, drainage networks or habitual routes between water and other resources. Studies of African lions demonstrate this complexity, in semi-arid Hwange landscape, waterholes acted as prominent nodes within lion movement networks, and lions concentrated search activity near them because water aggregated prey across the landscape (). Yet the value of an individual waterhole depended not only on prey abundance, but also on surrounding vegetation, catchability, territorial behavior and disturbance (; ). In Latin America, jaguars frequently revisited drainage edges and moved more slowly within natural drainage systems, whereas roads facilitated faster movement and shorter use (). The PTR results extend this broader pattern by showing that hydrological features can organize recursive movement without producing a single directional response. Water distance influenced whether sites were repeatedly incorporated into tiger movements and whether they supported longer stays, but its role changed with sex and season.
Canopy structure showed a similarly revealing distinction between the environmental occurrence of revisited sites and the intensity of their use. Revisitation generally increased with canopy height during summer and monsoon, while winter relationships became weaker, negative or distinctly curved. Residence time increased with canopy height in male and female summer cohorts, whereas monsoon and winter curves often peaked at intermediate values or changed direction across the observed range. Tall or continuous-canopy neighborhoods may be particularly valuable during summer because they retain comparatively buffered microclimates and structural cover when deciduous leaf loss exposes much of the surrounding landscape (). During the monsoon, the wider recovery of vegetation may redistribute cover beyond permanently taller stands, while winter conditions may alter the balance among concealment, solar exposure, prey movement and ease of travel. The absence of a common threshold is itself informative. It indicates that canopy height is not used as a fixed habitat rule; rather, its value is re-weighted as environmental conditions change. This mirrors patterns in African lions, where vegetation cover was the strongest habitat predictor but its effect depended on season, time of day and distance from people (). It also resembles European lynx, for which the selection of rocky and rugged microhabitats changed with their availability and differed between resting and active periods (). Structural habitat therefore functions through context. In PTR, canopy height appears to define where many recursive-use patches occur, while season and behavioral purpose determine how intensely an individual patch is used.
EVI described a different ecological dimension and should not be treated as interchangeable with canopy height. Whereas canopy height represents comparatively stable vertical structure, EVI reflects photosynthetic activity and green vegetation condition at the time represented by the satellite imagery (). Revisitation increased with EVI around average conditions in most cohorts, but the complete curves included plateaus and broad peaks. Residence-time responses were even more variable. Male summer revisitation was lowest at intermediate EVI, while residence time reached a peak at a higher value; female monsoon revisitation peaked at intermediate greenness, whereas residence time declined across the same gradient. Winter patterns also differed between the sexes. This dissociating is ecologically credible in a dry forest, where greenness can simultaneously increase forage for ungulates, improve concealment and create dense vegetation that alters visibility or movement (). Seasonal herbivore distributions in central Indian forests are dynamic, and tiger abundance and space use are closely tied to prey at broader scales (; ). Yet a green patch need not contain accessible prey, and an area of high prey abundance need not favor prolonged residence if dense vegetation restricts movement or visibility. The EVI relationships therefore indicate that recursive tiger use tracked seasonal variation in vegetation productivity, but did so differently for recurrent passage and extended occupancy. This distinction strengthens the ecological result: vegetation productivity was not simply selected or avoided; it altered the functional role of sites within the movement network.
Distance to villages provided the clearest example of why human influence is better understood as an interface than as a linear gradient of attraction or avoidance. During summer, revisitation and residence time for both sexes showed broad maxima at intermediate distances from villages. Use was reduced immediately beside settlements, but it did not increase continuously toward the most remote parts of the landscape. During monsoon and winter, residence time generally declined with village distance, while revisitation retained sex-specific and curved responses. These patterns suggest that the village-distance gradient also represents variation in forest-agriculture edges, local water sources, roads, livestock routes, prey movement and the timing of human activity. Longer residence toward the settlement-proximate end of the sampled gradient should therefore not be interpreted as attraction to people. It is more likely to reflect selective use of particular habitat elements within a heterogeneous human-forest boundary. Comparable behavioral accommodation is widespread among large carnivores. Tigers in Nepal used human-dominated areas by separating their activity from people in time, rather than excluding those areas entirely (). Leopards in western India persisted within densely settled agricultural landscapes by using cover, nocturnal activity and fine-scale avoidance of direct encounters (). In Tanzania, lions altered their reliance on vegetation cover according to settlement proximity and time of day, with substantial variation among individuals (; ). Jaguars repeatedly used natural and agricultural edges but reduced revisitation within extensive agriculture and moved rapidly near roads (). European lynx similarly avoided human-used paths and brown bears across Europe compressed their space use where human footprint and forest disturbance were greater (; ; ). The PTR results fit this broader picture: large carnivores do not respond to human landscapes through simple distance-based exclusion. They adjust when, where and for how long particular parts of the interface are used.
The contrast between marginal and conditional R2 values adds an important biological dimension to these environmental patterns. Fixed covariates captured shared tendencies across the population, but model fit increased substantially when tiger identity was included. Under the Nakagawa framework, this difference indicates persistent among-individual heterogeneity in recursive space use ().
This result is consistent with comparative evidence. Lions living alongside pastoralists varied widely in their tolerance of people even after common environmental effects were considered, while Eurasian lynx showed substantial individual variation in their selection of rocky terrain and infrastructure (; ). Continental analyses of brown bears and Eurasian lynx have likewise found that habitat relationships vary among regions and individuals, making locally conditioned responses more realistic than a single range-wide selection function (; ). In PTR, individual identity filtered the common environmental gradients through each tiger’s territory and experience. The biological picture is therefore one of shared tendencies embedded within individual movement traditions, rather than a universal habitat prescription.
Overall, the fine-scale models identify a structured but flexible system of recursive space use. Revisitation and residence time remained complementary because recurrent passage and prolonged occupancy expressed different ways of using the same landscape. TCH provided the clearest environmental signature separating revisited from available sites, but its relationship with use intensity changed across seasons and cohorts. Water, vegetation productivity, topographic position and village proximity repeatedly influenced recursive use, yet they did so through curved, sex-specific and seasonally re-weighted relationships rather than fixed thresholds. Individual identity accounted for a substantial share of the remaining variation, showing that common environmental effects were interpreted through the spatial history and local circumstances of each tiger. The contribution of this study is therefore not the definition of a single optimal tiger habitat at 250 m. It is the demonstration that familiar places form a differentiated network within the home range: some are returned to repeatedly, some support prolonged occupancy, and their importance changes as the dry-deciduous landscape moves from seasonal scarcity to seasonal abundance.
Caveats and limitations
Several limitations define the scope of these inferences. First, the biologically independent sample comprised 12 resident tigers, including only three males. Intensive tracking and repeated seasonal records improve estimation of within-individual and seasonal patterns, but do not replace replication among animals; sex specific interpretations, particularly for males, should therefore remain cautious. Second, the 250-m radius was selected as an operational scale that maximized variation in revisitation among candidate radii. It represents the spatial support of the analysis, not the known size of a resting, hunting, kill, or marking site, and results may change at other radii. Third, environmental layers were summarized within the same 250-m neighborhood, but their native resolutions, temporal frequencies, and ecological meanings differed. Tree canopy height is a structural proxy, EVI represents greenness, distance to village is an indirect disturbance proxy. Most importantly, direct measures of wild and domestic prey abundance, prey movement, kills, reproductive state, human and vehicle activity, fire, roads, and tourism were unavailable. These omissions prevent causal attribution of the observed associations. Finally, quadratic vertices were estimated from fitted curves and should not be interpreted as transferable thresholds, especially near sparse portions of the covariate range. Validation across additional individuals, dry-forest reserves, spatial radii, and independently measured behavioral states will be required before generalizing the response shapes beyond PTR.
Conservation implications
Recursive-use maps can add a useful management layer to conventional occurrence and home-range products as it distinguishes places repeatedly entered from places occupied for long periods. Management should not, however, convert the present curves into universal prescriptions for a particular canopy height, water distance, or village buffer. A stronger application is to identify a portfolio of recurrent-use neighborhoods across seasons and then determine why they are important through field validation. Sites with both high revisitation and long residence may warrant particular attention, but frequently revisited travel nodes, drainage crossings, or water-access routes can also be functionally important even when residence is short. Maintaining structural heterogeneity in dry deciduous forest, protecting drainage and riparian networks, limiting intense or predictable disturbance around recurrent-use sites, and retaining permeable forest cover across village interfaces are consistent with the observed associations. Priority sites should be checked against prey surveys, kill and resting locations, fire and tourism records, livestock use, and human-activity schedules before intervention. This evidence-led approach preserves the principal conservation value of recursive analysis: it identifies where repeated use is concentrated and directs management to the ecological processes that require confirmation.
Conclusion
This study provides the first explicit assessment of recursive space use of tigers in the Indian subcontinent and demonstrates that locations within home ranges are not used uniformly. Females had higher descriptive revisitation and residence-time values than males, recursive use was most concentrated in summer, and revisited sites were environmentally non-random. Yet the strongest conclusion lies in the divergence between metrics and among cohorts: canopy structure, vegetation greenness, water distance, topographic position, and village distance were associated with repeated return and prolonged occupancy in different, often non-linear ways. TCH was the strongest univariate discriminator of revisited versus available sites, but it was not a universal directional predictor of use intensity, and the measured environmental effects were outweighed by substantial individual heterogeneity. Recursive movement analysis therefore complements occurrence-based habitat studies by revealing how often sites are re-entered and how long they are occupied. Within the limits of a 12-individual study in one dry deciduous reserve, the results support management that protects a seasonally diverse network of recurrent-use sites and validates their behavioral functions rather than assuming a single optimal habitat configuration. Collectively, the findings establish recursive movement ecology as a powerful approach for identifying fine-scale behavioral refugia and ecologically crucial habitats critical for tiger conservation and landscape-level management.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
All 25 tigers (6 male, 19 female) were collared under permit by the Madhya Pradesh Forest Department (Permit No. Exp./2009/1205, dated August 31, 2009), under Section 11(1A) of the Indian Wildlife (Protection) Act, 1972. The study was conducted in accordance with the local legislation and institutional requirements.
Author contributions
SD: Software, Writing – original draft, Writing – review & editing, Methodology, Formal analysis, Validation, Data curation, Visualization, Conceptualization. RK: Writing – review & editing, Resources, Funding acquisition, Investigation, Supervision, Conceptualization, Project administration.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This research was funded by the Madhya Pradesh Forest Department (MPFD) and National Tiger Conservation Authority (NTCA) No. WII/KR/PROJECT/PTRP/2013-14/012 (B), National Water Development Agency (NWDA) No. WII/KR/PROJECT/PLMP/2017-18/F (1).
Acknowledgments
We would like to express our deep gratitude to the Wildlife Institute of India, especially Director, Dean and Research Coordinator of the Wildlife Institute of India for the opportunity to undertake the study. We are thankful to the Field Directors of PTR (Mr Lalan Kumar Choudhury, Mr R.S. Murthy, Mr Alok Kumar, Mr Vivek Jain, Mr Ravi Kant Mishra, Mr K.S. Bhadoria, and Mr Uttam Kumar Sharma) and the veterinary officer Dr Sanjeev Kumar Gupta for their continuous support and encouragement. We thank Ms. Suzanne Jacob and Mr. Hritik Raj for their assistance during data analysis. We would also like to extend our thanks to Mr. Dnyanesh S. Rathod, Ms. Kamana Pokhariya and our field staff, Mr. Darshan Singh, Ms. Anjana Raikwar, Mr. Pappu Gadariya, and Mr. Shiv Pratap Barman, for their assistance in conducting the extensive fieldwork.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fevo.2026.1929428/full#supplementary-material
References
1
AlegreV. B.KandaC. Z.de Faria OshimaJ. E.NiebuhrB. B.MoratoR. G.ThompsonJ. J.et al. (2024). Jaguar at the edge: movement patterns in human-altered landscapes. Perspect. Ecol. Conserv.22, 358–366. doi: 10.1016/j.pecon.2024.10.002
2
AllenM. L.GuntherM. S.WilmersC. C. (2017). The scent of your enemy is my friend? The acquisition of large carnivore scent by a smaller carnivore. J. Ethol.35, 13–19. doi: 10.1007/s10164-016-0492-6
3
AthreyaV.OddenM.LinnellJ. D.KrishnaswamyJ.KaranthU. (2013). Big cats in our backyards: persistence of large carnivores in a human dominated landscape in India. PloS One8, e57872. doi: 10.1371/journal.pone.0057872
4
BarkerK. J.ColeE.CourtemanchA.DeweyS.GustineD.MillsK.et al. (2023). Large carnivores avoid humans while prioritizing prey acquisition in anthropogenic areas. J. Anim. Ecol.92, 889–900. doi: 10.1111/1365-2656.13900
5
BatesD.MächlerM.BolkerB.WalkerS. (2015). Fitting linear mixed-effects models using lme4. J. Stat. Software67, 1–28. doi: 10.18637/jss.v067.i01
6
BenhamouS.Riotte-LambertL. (2012). Beyond the utilization distribution: identifying home range areas that are intensively exploited or repeatedly visited. Ecol. Modell.227, 112–116. doi: 10.1016/j.ecolmodel.2011.12.015
7
Berger-TalO.Bar-DavidS. (2015). Recursive movement patterns: review and synthesis across species. Ecosphere6, 1–12. doi: 10.1890/es15-00106.1
8
BertiE.RosenbaumB.VollrathF. (2025). Energy landscapes direct the movement preferences of elephants. J. Anim. Ecol.94, 908–918. doi: 10.22541/au.168373276.62196439/v1
9
BoyerD.WalshP. D. (2010). Modelling the mobility of living organisms in heterogeneous landscapes: does memory improve foraging success? Philos. Trans. R. Soc. A. Mathematical Phys. Eng. Sci.368, 5645–5659. doi: 10.1098/rsta.2010.0275
10
BracisC. (2024). recurse: Computes revisitation metrics for trajectory data (R package version 1.4.0), CRAN. doi: 10.32614/CRAN.package.recurse
11
BracisC.BildsteinK. L.MuellerT. (2018). Revisitation analysis uncovers spatio‐temporal patterns in animal movement data. Ecography41, 1801–1811. doi: 10.1111/ecog.03618
12
CarterN. H.ShresthaB. K.KarkiJ. B.PradhanN. M. B.LiuJ. (2012). Coexistence between wildlife and humans at fine spatial scales. Proc. Natl. Acad. Sci.109, 15360–15365. doi: 10.1073/pnas.1210490109
13
ChampionH. G.SethS. K. (1968). A Revised Survey of the Forest Types of India ( Manager of publications).
14
CharnovE. L. (1976). Optimal foraging, the marginal value theorem. Theor. Popul Biol.9, 129–136. doi: 10.1016/0040-5809(76)90040-x
15
ČončŠ.OliveiraT.BelottiE.BufkaL.ČerneR.HeurichM.et al. (2024). Revealing functional responses in habitat selection of rocky features and rugged terrain by Eurasian lynx (Lynx lynx) using LiDAR data. Landscape Ecol.39, 121. doi: 10.1007/s10980-024-01923-y
16
DaviesA. B.TamblingC. J.KerleyG. I.AsnerG. P. (2016). Effects of vegetation structure on the location of lion kill sites in African thicket. PloS One11, e0149098. doi: 10.1371/journal.pone.0149098
17
DavisK. T.DobrowskiS. Z.HoldenZ. A.HigueraP. E.AbatzoglouJ. T. (2019). Microclimatic buffering in forests of the future: the role of local water balance. Ecography42, 1–11. doi: 10.1111/ecog.03836
18
DejeanteR.LoveridgeA. J.MacdonaldD. W.MadhlamotoD.Chamaillé-JammesS.ValeixM. (2026). Are all waterholes equal from a lion’s view? Exploring the role of prey abundance and catchability in waterhole visitation patterns in a savannah ecosystem. Peer Community J.6, e32. doi: 10.24072/pcjournal.705
19
De KnegtH. J.PretoriusY.van LangeveldeF.de BoerW. F.GortG.SkidmoreA. K.et al. (2010). Visited sites revisited-site fidelity in African elephants. Beyond here now: Herbivore Ecol. A. spatialtemporal context. Wageningen University Wageningen91 (8), 35–45. doi: 10.1017/9781316534380
20
Delignette-MullerM. L.DutangC. (2015). fitdistrplus: An R package for fitting distributions. J. Stat. Software64, 1–34. doi: 10.18637/jss.v064.i04
21
DinhE. T.CauvinA.OrangeJ. P.ShumanR. M.WiselyS. M.BlackburnJ. K. (2020). Living la Vida T-LoCoH: Site fidelity of Florida ranched and wild white-tailed deer (Odocoileus virginianus) during the epizootic hemorrhagic disease virus (EHDV) transmission period. Mov. Ecol.8, 14. doi: 10.1186/s40462-020-00200-2
22
DormannC. F.ElithJ.BacherS.BuchmannC.CarlG.CarréG.et al. (2013). Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography36, 27–46. doi: 10.1111/j.1600-0587.2012.07348.x
23
DuttaS.KrishnamurthyR. (2024a). Multiphasic movement and step-selection patterns of dispersed tigers in the central Indian landscape. PloS One19, e0309517. doi: 10.1371/journal.pone.0309517
24
DuttaS.KrishnamurthyR. (2024b). Successful conservation translocation: Population dynamics of tiger recovery in Panna Tiger Reserve, Central India. Ecol. Solutions Evidence5, e12337. doi: 10.1002/2688-8319.12337
25
DuttaS.KrishnamurthyR. (2025). Tigers in their new territory: intraspecific interactions among the reintroduced tiger population in Panna Tiger Reserve, Central India. Sci. Rep.15, 1488. doi: 10.1038/s41598-024-83278-7
26
DuttaS.MaheswaranG.KrishnamurthyR. (2026). Interspecific interaction among mammals in Panna Tiger Reserve, central India: Prey-predator dynamics. Global Ecol. Conserv.67, e04145. doi: 10.1016/j.gecco.2026.e04145
27
ErwinR. M. (1989). Predator-prey interactions, reosurce depression and patch revisitation. Behav. Processes18, 1–16. doi: 10.1016/s0376-6357(89)80002-6
28
FaganW. F.LewisM. A.Auger‐MéthéM.AvgarT.BenhamouS.BreedG.et al. (2013). Spatial memory and animal movement. Ecol. Lett.16, 1316–1329. doi: 10.1111/ele.12165
29
FlemingC. H.FaganW. F.MuellerT.OlsonK. A.LeimgruberP.CalabreseJ. M. (2015). Rigorous home range estimation with movement data: a new autocorrelated kernel density estimator. Ecology96, 1182–1188. doi: 10.1890/14-2010.1
30
FlemingC. H.CalabreseJ. M. (2023). ctmm: Continuous-time movement modeling (R package version 1.2.0). doi: 10.32614/CRAN.package.ctmm
31
FrairJ. L.NielsenS. E.MerrillE. H.LeleS. R.BoyceM. S.MunroR. H.et al. (2004). Removing GPS collar bias in habitat selection studies. J. Appl. Ecol.41, 201–212. doi: 10.1111/j.0021-8901.2004.00902.x
32
FullerA.HetemR. S.MaloneyS. K.MitchellD. (2014). Adaptation to heat and water shortage in large, arid-zone mammals. Physiology29, 159–167. doi: 10.1152/physiol.00049.2013
33
GoodrichJ. M.MiquelleD. G.SmirnovE. N.KerleyL. L.QuigleyH. B.HornockerM. G. (2010). Spatial structure of Amur (Siberian) tigers (Panthera tigris altaica) on Sikhote-Alin Biosphere Zapovednik, Russia. J. Mammal.91, 737–748. doi: 10.1644/09-mamm-a-293.1
34
GuptaD.KrishnamurthyR. (2023). Spatial density patterns of herbivore response to seasonal dynamics in the tropical deciduous forest of central India. Biotropica55, 430–443. doi: 10.1111/btp.13197
35
GurarieE.BracisC.DelgadoM.MeckleyT. D.KojolaI.WagnerC. M. (2016). What is the animal doing? Tools for exploring behavioural structure in animal movements. J. Anim. Ecol.85, 69–84. doi: 10.1111/1365-2656.12379
36
HafnerS.WiesnerH.von HegelG.HalmS.ErhardtW. (1989). The antagonism of ketamine/xylazine anesthesia (" Hellabrunn mixture") in wild zoo ruminants. Berliner und Munchener Tierarztliche Wochenschrift102, 181–184. doi: 10.5282/ubm/epub.8258
37
HertelA. G.ParresA.FrankS. C.RenaudJ.SelvaN.ZedrosserA.et al. (2025). Human footprint and forest disturbance reduce space use of brown bears (Ursus arctos) across Europe. Global Change Biol.31, e70011. doi: 10.1111/gcb.70011
38
HijmansR.BrownA.BarbosaM. (2026). terra: Spatial Data Analysis (R package version 1.9-40), CRAN. doi: 10.32614/CRAN.package.terra
39
JacobS. A.De BoerW. F.De KnegtH. J.EngelenC. C.HenleyM. (2025). Environmental determinants of foraging site revisitation by African elephants (Loxodonta africana). Ecol. Evol.15, e71506. doi: 10.1002/ece3.71506
40
JanssonI.ParsonsA. W.SinghN. J.FaustL.KissuiB. M.MjingoE. E.et al. (2024). Coexistence from a lion’s perspective: Movements and habitat selection by African lions (Panthera leo) across a multi-use landscape. PloS One19, e0311178. doi: 10.1371/journal.pone.0311178
41
JhalaY.GopalR.MathurV.GhoshP.NegiH. S.NarainS.et al. (2021). Recovery of tigers in India: Critical introspection and potential lessons. People Nat.3, 281–293. doi: 10.1002/pan3.10177
42
KapotaD.DolevA.SaltzD. (2017). Inferring detailed space use from movement paths: A unifying, residence time‐based framework. Ecol. Evol.7, 8507–8514. doi: 10.1002/ece3.3321
43
KaranthK. U.KumarN. S.KaranthK. K. (2020). Tigers against the odds: Applying macro-ecology to species recovery in India. Biol. Conserv.252, 108846. doi: 10.1016/j.biocon.2020.108846
44
KaranthK. U.NicholsJ. D.KumarN. S.LinkW. A.HinesJ. E. (2004). Tigers and their prey: predicting carnivore densities from prey abundance. Proc. Natl. Acad. Sci.101, 4854–4858. doi: 10.1073/pnas.0306210101
45
KaranthK. U.SrivathsaA.VasudevD.PuriM.ParameshwaranR.KumarN. (2017). Spatio-temporal interactions facilitate large carnivore sympatry across a resource gradient. Proc. R. Soc. B. Biol. Sci.284, 20161860. doi: 10.1098/rspb.2016.1860
46
KaysR.CrofootM. C.JetzW.WikelskiM. (2015). Terrestrial animal tracking as an eye on life and planet. Science348, aaa2478. doi: 10.1126/science.aaa2478
47
KlappsteinN. J.PottsJ. R.MichelotT.BörgerL.PilfoldN. W.LewisM. A.et al. (2022). Energy-based step selection analysis. J. Anim. Ecol.91, 946–957. doi: 10.1111/1365-2656.13687
48
LoveridgeA. J.ValeixM.ElliotN. B.MacdonaldD. W. (2017). The landscape of anthropogenic mortality: how African lions respond to spatial variation in risk. J. Appl. Ecol.54, 815–825. doi: 10.1111/1365-2664.12794
49
LyonsA. J.TurnerW. C.GetzW. M. (2013). Home range plus: a space-time characterization of movement over real landscapes. Mov. Ecol.1, 2. doi: 10.1186/2051-3933-1-2
50
MasseyF. J.Jr. (1951). The Kolmogorov-Smirnov test for goodness of fit. J. Am. Stat. Assoc.46, 68–78. doi: 10.1080/01621459.1951.10500769
51
McRaeB. H.DicksonB. G.KeittT. H.ShahV. B. (2008). Using circuit theory to model connectivity in ecology, evolution, and conservation. Ecology89, 2712–2724. doi: 10.1890/07-1861.1
52
NaglD.BreitenmoserU.HackländerK.RyserA.ZimmermannF.SignerS.et al. (2022). Long‐term changes in habitat selection and prey spectrum in a reintroduced Eurasian lynx (Lynx lynx) population in Switzerland. Ecol. Evol.12, e8614. doi: 10.1002/ece3.8614
53
NakagawaS.NobleD. W.SeniorA. M.LagiszM. (2017). Meta-evaluation of meta-analysis: ten appraisal questions for biologists. BMC Biol.15, 18. doi: 10.1186/s12915-017-0357-7
54
NonacsP. (2001). State dependent behavior and the marginal value theorem. Behav. Ecol.12, 71–83. doi: 10.1093/oxfordjournals.beheco.a000381
55
OeserJ.HeurichM.Kramer‐SChadtS.MattissonJ.KrofelM.Krojerová‐ProkešováJ.et al. (2023). Integrating animal tracking datasets at a continental scale for mapping Eurasian lynx habitat. Divers. Distrib.29, 1546–1560. doi: 10.1111/ddi.13784
56
Owen-SmithN.FryxellJ. M.MerrillE. H. (2010). Foraging theory upscaled: the behavioural ecology of herbivore movement. Philos. Trans. R. Soc. B. Biol. Sci.365, 2267–2278. doi: 10.1098/rstb.2010.0095
57
PotapovP.LiX.Hernandez-SernaA.TyukavinaA.HansenM. C.KommareddyA.et al. (2021). Mapping global forest canopy height through integration of GEDI and Landsat data. Remote Sens. Environ.253, 112165. doi: 10.1016/j.rse.2020.112165
58
Puehringer-SturmayrV.KrejciJ.SchusterR.KleindorferS.KotrschalK.FrigerioD.et al. (2023). Space use and site fidelity in the endangered Northern Bald Ibis Geronticus eremita: Effects of age, season, and sex. Bird Conserv. Int.33, e10. doi: 10.1017/s0959270922000053
59
R Core Team (2025). R: A language and environment for statistical computing (Vienna, Austria: R Foundation for Statistical Computing). doi: 10.32614/R.manuals
60
SarkarM. S.KrishnamurthyR.JohnsonJ. A.SenS.SahaG. K. (2017). Assessment of fine-scale resource selection and spatially explicit habitat suitability modelling for a re-introduced tiger (Panthera tigris) population in central India. PeerJ5, e3920. doi: 10.7717/peerj.3920
61
SeidelD. P.LinklaterW. L.KilianW.du PreezP.GetzW. M. (2019). Mesoscale movement and recursion behaviors of Namibian black rhinos. Movement Ecology7 (1), 34. doi: 10.1186/s40462-019-0176-2
62
SimcharoenA. C. H. A. R. A.SaviniT.GaleG. A.SimcharoenS.DuangchantrasiriS.PakpienS.et al. (2014). Female tiger Panthera tigris home range size and prey abundance: important metrics for management. Oryx48, 370–377. doi: 10.1017/S0030605312001408
63
SimsD. A.RahmanA. F.CordovaV. D.El‐MasriB. Z.BaldocchiD. D.FlanaganL. B.et al. (2006). On the use of MODIS EVI to assess gross primary productivity of North American ecosystems. J. Geophysical Research: Biogeosciences111, G04015. doi: 10.1029/2006jg000162
64
SmithJ. L. D. (1993). The role of dispersal in structuring the Chitwan tiger population. Behaviour124, 165–195. doi: 10.1163/156853993x00560
65
SunartoS.KellyM. J.ParakkasiK.KlenzendorfS.SeptayudaE.KurniawanH. (2012). Tigers need cover: multi-scale occupancy study of the big cat in Sumatran forest and plantation landscapes. PloS One7, e30859. doi: 10.1371/journal.pone.0030859
66
SunquistM. E. (1981). The social organization of tigers (Panthera tigris) in Royal Chitawan National park, Nepal. Smithsonian Contribution to Zoology; no 336. doi: 10.5479/si.00810282.336
67
SunquistM.SunquistF. (2002). Wild Cats of the World ( University of Chicago Press).
68
ValeixM.LoveridgeA. J.DavidsonZ.MadzikandaH.FritzH.MacdonaldD. W. (2010). How key habitat features influence large terrestrial carnivore movements: waterholes and African lions in a semi-arid savanna of north-western Zimbabwe. Landscape Ecol.25, 337–351. doi: 10.1007/s10980-009-9425-x
69
Van MoorterB.BunnefeldN.PanzacchiM.RolandsenC. M.SolbergE. J.SætherB. E. (2013). Understanding scales of movement: animals ride waves and ripples of environmental change. J. Anim. Ecol.82, 770–780. doi: 10.1111/1365-2656.12045
70
Van MoorterB.RolandsenC. M.BasilleM.GaillardJ. M. (2016). Movement is the glue connecting home ranges and habitat selection. J. Anim. Ecol.85, 21–31. doi: 10.1111/1365-2656.12394
71
Van MoorterB.VisscherD.BenhamouS.BörgerL.BoyceM. S.GaillardJ. M. (2009). Memory keeps you at home: a mechanistic model for home range emergence. Oikos118, 641–652. doi: 10.1111/j.1600-0706.2008.17003.x
72
WallisDeVriesM. F.LacaE. A.DemmentM. W. (1999). The importance of scale of patchiness for selectivity in grazing herbivores. Oecologia121, 355–363. doi: 10.1007/s004420050939
73
WarrierR.NoonB. R.BaileyL. (2020). Agricultural lands offer seasonal habitats to tigers in a human‐dominated and fragmented landscape in India. Ecosphere11, e03080. doi: 10.1002/ecs2.3080
74
WeiT.SimkoV. (2017). R package ‘corrplot’: Visualization of a correlation matrix (Version 0.84). CRAN.
75
YiuS. W.ParriniF.KarczmarskiL.KeithM. (2017). Home range establishment and utilization by reintroduced lions (Panthera leo) in a small South African wildlife reserve. Integr. Zool.12, 318–332. doi: 10.1111/1749-4877.12243
76
YoungmannJ. L.HintonJ. W.BaknerN. W.ChamberlainM. J.D'AngeloG. J. (2022). Recursive use of home ranges and seasonal shifts in foraging behavior by a generalist carnivore. Ecol. Evol.12, e9540. doi: 10.1002/ece3.9540
Summary
Keywords
behavioral plasticity, optimal foraging, recursive space use, resource selection, site fidelity
Citation
Dutta S and Krishnamurthy R (2026) Revisitation dynamics and fine-scale habitat use of tigers in a tropical dry deciduous forest of central India. Front. Ecol. Evol. 14:1929428. doi: 10.3389/fevo.2026.1929428
Received
06 July 2026
Revised
02 August 2026
Accepted
11 August 2026
Published
08 September 2026
Volume
14 - 2026
Edited by
Sergei Petrovskii, University of Leicester, United Kingdom
Reviewed by
Qi Liu, Central South University Forestry and Technology, China
Paras Singh, Biodiversity Conservation Society Nepal, Nepal
Updates
Copyright
© 2026 Dutta and Krishnamurthy.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Ramesh Krishnamurthy, ramesh@wii.gov.in
†ORCID: Supratim Dutta, orcid.org/0009-0007-1009-7845; Ramesh Krishnamurthy, orcid.org/0000-0001-8797-8615
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.