Assessment of human-wildlife conflict through GIS And RS in India

Introduction

India supports large populations of elephants, tigers and leopards while also having a very high human population. The close sharing of landscapes by people and wildlife has made humanwildlife conflict (HWC) an important concern for wildlife conservation, agriculture and the safety of rural communities. HWC in India is generally divided into three categories crop raiding by herbivores, livestock depredation by carnivores and human injury or death from direct encounters (Karanth, et al., 2012) [14] . Crop raiding, mainly by elephants and wild boar, is the most widespread form of conflict and affects farmers across almost every state with forest cover nearby. Particularly elephants can destroy an entire season crop in a single night and damage tends to be worst in villages close to forest edges (Naha, et al., 2020) [18] . In Central India, some of this damage goes unrecorded because farmers do not always file for compensation (Kshettry, et al., 2021) [15] . Livestock depredation is mostly caused by leopards and tigers preying on cattle, goats and other domestic animals grazing near forest boundaries. This form of conflict is closely tied to how livestock is managed weak enclosures and free grazing near forest edges increase the risk significantly (Donikar et.al.,2011) [8] . Leopard presence alone does not always predict how severe depredation will be since some high-density leopard landscapes report comparatively low livestock losses (Athreya et al., 2020) [2] . Human injury and death are the least common but most serious form of conflict, usually resulting from sudden encounters with elephants, tigers or leopards near forest-edge areas, water sources or during resource collection activities such as firewood gathering or non-timber forest product (Dhanwatey et al.,2013) [7] . Despite being less frequent than crop, livestock losses and human casualties account for the largest overall economic and social cost of HWC in India (Gulati et al., 2021) [10] . For a long time, HWC in India was studied mainly through household surveys, incident counts and compensation records (Karanth et al., 2012) [14] . These methods are useful for understanding local experiences but cannot easily predict where conflict is likely to occur next. Over the past two decades Geographic Information Systems (GIS) and remote sensing (RS) have made this possible, using tools such as kernel density estimation, hotspot analysis and MaxEnt modelling combined with land-use/land-cover (LULC) data. These methods have been applied successfully in places like Junnar (Habib et al., 2020) [11] , Keonjhar (Tripathy et al., 2021) [26] , Jharkhand (Roy et al., 2025) [25] , Assam (Athira et al., 2026) [1] and the Southern Western Ghats (Pavithra et al., 2025) [20] . This progress, however is uneven. Vidarbha, a region with welldocumented conflict around Tadoba-Andhari Tiger Reserve (Bayani et al., 2016; Dhanwatey et al., 2013) [4,7] and Pauni Wildlife Sanctuary (Jha, Jha, & Dahake, 2022) [13] , has only one GIS based study, conducted in Chandrapur (Habib et al., 2022) [12] . Akola district, part of this same region, has no GIS based study for any of the three conflict categories, though descriptive farmer perception evidence is now available for crop raiding specifically (Khekade et al., 2025) [16] .

Objective

To review GIS and remote sensing-based mapping of humanwildlife conflict across India, Maharashtra, Vidarbha and Akola.

Methodology

We collected English-language scholarly publications (journal articles, institutional reports and edited-volume book chapters) that referred to geospatial (GIS/remote sensing) approaches to mapping human-wildlife conflict (HWC) in India, as well as descriptive studies documenting crop raiding, livestock depredation and human injury/death in the absence of spatial modelling. We included articles focused primarily on GIS/RSbased hotspot mapping and those that documented conflict through non-spatial means such as household surveys or compensation-claim analysis, so that the two approaches could be compared directly. We used the search engines Google Scholar, Web of Science and institutional repositories including the Wildlife Institute of India (WII) and the Indian Institute of Forest Management (IIFM). We used the following keywords singly and in combination: human-wildlife conflict, human-elephant conflict, humanleopard conflict, crop raiding, livestock depredation, human injury wildlife, GIS, geospatial, remote sensing, hotspot mapping, kernel density estimation, MaxEnt and land-use/landcover. We added geographic identifiers to these search terms, including India, Maharashtra, Vidarbha and Akola, as well as named landscapes appearing in the literature, such as Junnar, Chandrapur, Keonjhar, Jharkhand, Assam and the Western Ghats. We examined and included where appropriate related articles identified through citation linking on Google Scholar. We also searched for relevant literature cited within the articles returned by our initial searches (backward/snowball citation sampling). No lower date restriction was applied included studies span 2004 to 2026. Publications were excluded if they addressed HWC outside India, focused solely on non-target taxa or were opinion pieces or commentaries without original data. For each publication, we documented author names, year of publication, conflict category addressed (crop raiding, livestock depredation and human injury/death) and spatial scale addressed (India, Maharashtra, Vidarbha and Akola). We classified each publication methodology as either GIS/spatially explicit employing kernel density estimation, Optimized or Emerging Hotspot Analysis, MaxEnt species distribution modelling, or LULC-based covariate regression, typically implemented in ArcGIS or an equivalent platform or descriptive/non-spatial, relying on household surveys, damage assessments, incidentcount reporting without formal spatial modelling. We further recorded for each publication the specific spatial technique used (where applicable) the landscape or protected area studied the species implicated in conflict and the time period covered by the data. Locations of GIS-based studies were noted by district and state to allow comparison across the four nested spatial scales examined in this review.

Results and Discussion

Overview of geospatial coverage by scale Across the three conflict categories a consistent scale-dependent pattern emerges. GIS/RS-based hotspot mapping is well established at the national scale, present but narrower at the state scale, limited to a single district at the regional (Vidarbha) scale and now partially represented at the district (Akola) scale. This pattern held across all three conflict categories, suggesting it reflects a structural gap in research effort rather than a feature specific to any one conflict type.

Crop Raiding

Table 1 presents that the crop raiding id best documented at the national level. In India, the species causing crop damage is the Asian elephant studied and used MaxEnt modelling by Naha et al. (2019, 2020) [18,19] , Rani et al. (2024) [24] , while Pavithra et al. (2025) [20] used kernel density estimation combined with kriging and these studies used satellite based data such as Landsat and Sentinel land-use/land-cover change data, along with tree loss data, elevation data and road network data as inputs, the key findings were that villages within 1.5 km of the forest edge are affected the most and that mapped a total of 34,596 conflict incidents across the Southern Western Ghats, with these studies covering large landscapes including North Bengal, the Rajaji Corbett landscape in Uttarakhand and the Southern Western Ghats Pavithra et al. (2025) [20] . In Maharashtra, Pawar & Mule (2025) [21] analyzed 19,050 humanwildlife incidents across 2,659 villages in the Marathwada region, using negative binomial regression modelling with NDVI, distance to waterbody and elevation as spatial covariates to map conflict risk hotspots and pattern of compensation access, covering an area of 64,590 km 2 of Maharashtra identified in this review. In Vidarbha, the species causing crop damage are wild boar, sambar, chital, and nilgai, studied by Bayani et al. (2016) [4] , who has not use any GIS or remote sensing method and instead used distance-decay analysis, measuring how crop damage changed based on distance from the forest boundary, using field damage surveys collected along the Tadoba-Andhari Tiger Reserve boundary as their only data and finding that more than half of all crop damage occurred near that reserve boundary, with their study covering only the area immediately around the Tadoba-Andhari Tiger Reserve. Akola Khekade et al. (2025) [16] surveyed 120 farmers across four villages in patur tahsil using structured interview findings that wild boar (100%) and blue bull (94.17%) were the dominant species causing damage while 93.06% observed yield reduction, with the study covering villages in Akola district [16] Bayani et al., 2016 [4] Pawar & Mule 2025 [21] Naha et al.,2019 [19] Spatial /mapping method used Khekade et al., 2025 [16] Bayani et al., 2016 [4] Pawar & Mule 2025 [21] Naha et al., Rani et al., 2024 [24] & Pavithra et al.,2025 [20] Data inputs used Khekade et al., 2025 [16] Bayani et al., 2016 [4] Pawar & Mule 2025 [21] Naha et al., 2019& 2020 [18,19] ; Pavithra et al.,2025 [20] Key spatial findings Khekade et al., 2025 [16] Bayani et al., 2016 [4] Pawar & Mule 2025 [21] Naha et al., 2019 & 2020 [18,19] Pavithra et al.,2025 [20] Scale/extend of study Khekade et al., 2025 [16] Bayani et al., 2016 [4] Pawar & Mule 2025 [21] Naha et al.,2020 [18] ; Rani et al., 2024 [24] ; Pavithra et al.,2025 [20] Livestock Depredation Table 2 shows that, at the national level large carnivores and big herbivores are studied together for livestock killing. In India, the species involved are large carnivores and megaherbivores studied broadly by Ramesh et al. (2020 & 2022) [22,23] , who used spatial predation-risk modelling with landscape-level ecological covariates across the Eastern and Western Ghats, finding mapped predation-risk patterns and the drivers of humanmegaherbivore interaction, with their study covering the Eastern and Western Ghats across multiple states. In Maharashtra, the species involved is the leopard, studied by Habib et al. (2020) [11] using Optimized and Emerging Hotspot Analysis in ArcGIS, building on the earliest GIS-based leopard-conflict study in India done by Athreya et al. (2004) [3] , using 20 years of depredation records from the Junnar Forest Division as their data input and finding that a high leopard density of 10 per 100 square kilometers still resulted in a low livestock loss of 0.45 per square kilometer per year with zero human deaths at one site, as shown by Athreya et al. (2020) [2] , while Donikar et al. (2011) [8] found that poor livestock enclosure quality, not leopard density was the main driver of depredation in Ratnagiri district, with these studies covering the Junnar Forest Division and Ratnagiri district. Additionally, Krithivasan et al. (2009) [17] has documented wolf-livestock conflict in Ahmednagar district using GIS based mapping of attack locations combined with wolf density estimation, introducing wolf as third Maharashtra species alongside leopard and also contribute carnivores side conflict patterns from their Marathwada wide analysis, covering the junnar forest division, Ratnagiri Ahmednagar district and the Marathwada region. In Vidarbha, the species involved are leopard and tiger studied by Habib et al. (2022) [12] using the same ArcGIS hotspot analysis method, with depredation incident records from the Chandrapur Forest Division as their data input, finding clear depredation hotspots identified across the Chandrapur landscape, with their study covering Chandrapur district. In Akola, no study on livestock depredation was found at all so there is no species, method, data input, finding or study area available for this district. [12] Human Injury Tables 3 shows that human injury is the most spatially advanced category at the national scale, In India, the species involved is the Asian elephant, studied by Tripathy et al. (2021) [26] , Roy et al. (2025) [25] , and Athira et al. (2026) [1] ; Tripathy et al. (2021) [26] used kernel density estimation with temporal-phase mapping in Keonjhar, using fatality records across three time periods as their data input and found 345 humans deaths between 2001 and 2018, with the at risk population rising from 14,724 to 65,444 people over that time, while Roy et al. (2025) [25] and Athira et al. (2026) [1] used Landsat land-use/land-cover data combined with kernel density estimation in Jharkhand and Assam, finding 1,340 deaths among 1,740 incidents in Jharkhand and 1,468 deaths among 1,806 incidents in Assam and Gulati et al. (2021) [10] additionally found that human casualties are the single biggest cost of human-wildlife conflict across India overall, with these studies covering Keonjhar Forest Division, Jharkhand state and Assam state. In Maharashtra, the species involved is the leopard, studied by Habib et al. (2020) [11] using Emerging Hotspot Analysis in ArcGIS with 20 years of attack records from the Junnar Forest Division, finding clear human-attack hotspots identified at Junnar and by Deshmukh and Jaybhaye (2024) [6] , who conducted a household survey without GIS methods in Kothale village, documenting village-level conflict patterns there, with these studies covering the Junnar Forest Division and Kothale village. In Vidarbha, the species involved are tiger and leopard, studied by Dhanwatey et al. (2013) [7] , who did not use GIS and instead used case records from around the Tadoba-Andhari Tiger Reserve periphery, finding that attacks cluster near forest-edge resource-collection activities such as firewood gathering and by Habib et al. (2022) [12] , who used ArcGIS mapping with incident records from Chandrapur, finding attack hotspots across the Chandrapur landscape, with these studies covering the Tadoba-Andhari periphery and Chandrapur district. In Akola, no study on human injury or death was found at all, so there is no species, method, data input, finding, or study area available for this district.

Discussion

The consistent presence of Habib et al. (2020, 2022) [11,12] across both the livestock and human-injury tables indicates that GIS adoption in Maharashtra and Vidarbha has been driven largely by a single research program (Wildlife Institute of India, Junnar and Chandrapur), rather than broad-based regional uptake. While crop raiding has now an additional, independent entry point through Pawar & Mule (2025) [21] in Marathwada region showing that GIS based crop raiding research is beginning beyond the WII program specifically. This pattern also suggests a direct, low-cost extension pathway since the Chandrapur methodology (Habib et al., 2022) [12] was itself adapted from the earlier Junnar framework (Habib et al., 2020) [11] and since Pawar & Mule (2025) [21] have now demonstrated a working regression-based crop raiding model in Marathwada, either approach could reasonably be extended to Akola using Khekade et al. (2025) [16] & Bayani et al. (2016) [4] . Beyond extending existing spatial methods, recent conservation scholarship argues that human-wildlife conflict should also be addressed through adaptive, governance-level frameworks rather than site-specific hotspot mapping alone (Batar, 2026) [5] , suggesting that any future Akola-focused model should be paired with broader institutional and community-level mitigation planning not spatial modelling alone.

A limitation of this comparison is that the absence of a Maharashtra scale crop-raiding study and the absence of any Akola-scale study across all three categories, reflects only what was retrievable through this review's search strategy and does not rule out the existence of unpublished Forest Department records or internal assessments not indexed in the databases searched.

Research gap

Lack of study on GIS or remote sensing has been done for Akola district for livestock depredation or human injury/death this remains the main gap this review finds, though descriptive, farmer-perception data is now available for crop raiding (Khedkar et al., 2025). Crop raiding has recently been mapped using GIS in Maharashtra at the Marathwada scale (Pawar & Mule, 2025) [21] ,

References

1. Donikar et al., 2011 [8] ; Krithivasan et al., 2009 [17] Ramesh et al., 2020 & 2022 [22, 23] Spatial /mapping method used - Habib et al., 2022 [12] Habib et al., 2020 [11] ;
2. Krithivasan et al., 2009 [17] Athreya et al., 2004 [3] Ramesh et al., 2020 & 2022 [22, 23] Data inputs used - Habib et al., 2022 [12]
3. Krithivasan et al., 2009 [17] Ramesh et al., 2020 & 2022 [22, 23] Key spatial findings - Habib et al., 2022 [12] Athreya et al., 2020 [2] Donikar et al., 2011 [8] ;
4. Scale/extend of study - Habib et al., 2022 [12]
5. Athira NG, Pandey RK, Roy K, Dutta A, Mittal D, Nigam P, et al. Landscape determinants of human-elephant conflict in Assam, India: insights from two decades of spatial analysis. PeerJ. 2026;14:e21082.
6. Athreya V, Isvaran K, Odden M, Linnell JD, Kshettry A, Krishnaswamy J, et al. The impact of leopards (Panthera pardus) on livestock losses and human injuries in a humanuse landscape in Maharashtra, India. PeerJ. 2020;8:e8405.
7. Athreya VR, Thakur SS, Chaudhuri SUJOY, Belsare AV. A study of the man-leopard conflict. Junnar Forest Division: Junnar, India; 2004.
8. Bayani A, Tiwade D, Dongre A, Dongre AP, Phatak R, Watve M. Assessment of crop damage by protected wild mammalian herbivores on the western boundary of TadobaAndhari Tiger Reserve (TATR), Central India. PLoS One. 2016;11(4):e0153854.
9. Batar AK. A transformative conservation framework for advancing human-wildlife dynamics in India. Discover Conserv. 2026;3(1):5.
10. Deshmukh SN, Jaybhaye RG. Assessment of humanwildlife conflict: a case study from Kothale Village in the Kalsubai-Harishchandragad Wildlife Sanctuary (KHWS), Maharashtra, India. Asian J Environ Ecol. 2024;23(12):187198.
11. Dhanwatey HS, Crawford JC, Abade LA, Dhanwatey PH, Nielsen CK, Sillero-Zubiri C. Large carnivore attacks on humans in central India: a case study from the TadobaAndhari Tiger Reserve. Oryx. 2013;47(2):221-227.
12. Donikar RP, Patil VK, Narkhede SS, Rane AD, Mokat DN, Bhave SG. Circumstantial and response attitudes of people affected with livestock depredation by leopards Panthera pardus Linnaeus in Ratnagiri District, Maharashtra, India. J Bombay Nat Hist Soc. 2011;108(1):18.
13. Gureja N, Kumar A, Saigal S. Human-wildlife conflict in India. New Delhi: National Biodiversity Strategy and Action Plan (NBSAP); 2019.
14. Gulati S, Karanth KK, Le NA, Noack F. Human casualties are the dominant cost of human-wildlife conflict in India. Proc Natl Acad Sci U S A. 2021;118(8):e1921338118.
15. Habib B, Khandekar V, Nigam P, Mondol S, Jayaramegowda R, Ghanekar R, et al. Identification of human-leopard hotspot: prioritizing the mitigation measures in Junnar Forest Division, Pune, Maharashtra. Wildlife Institute of India; 2020.
16. Habib B, Nigam P, Praveen NR, Ravindran A. Patterns of human-wildlife conflict in Chandrapur, Maharashtra, India. Wildlife Institute of India; 2022.
17. Jha AK, Jha S, Dahake A. Assessing of conflict and mitigation measures between wildlife conservation and agro-pastoral practices in Pauni Wildlife Sanctuary in Bhandara District of Maharashtra. In: Dubey YK, editor. Managing human-wildlife conflict: an approach to mitigation and coexistence. Bhopal: Indian Institute of Forest Management; 2022. p. 174-183.
18. Karanth KK, Gopalaswamy AM, DeFries R, Ballal N. Assessing patterns of human-wildlife conflicts and compensation around a central Indian protected area. PLoS One. 2012;7(12):e50433.
19. Kshettry A, Bhave N, Das P, Athreya V. Mahakal blessed my crop: community dynamics and religious beliefs influence efficacy of a wildlife compensation program. Front Conserv Sci. 2021;2:727696.
20. Khekade SS, Deshmukh HK, Harne SS, Katole RT, Walke RD, Palaspagar PR, et al. Perception of farmers towards crop raiding by wild animals in Patur Tahsil of Akola District, Maharashtra. 2025.
21. Krithivasan R, Athreya V, Odden M. Human-wolf conflict in human dominated landscapes of Ahmednagar District, Maharashtra. Rufford Small Grants Foundation for Nature Conservation; 2009. p. 1.
22. Naha D, Dash SK, Chettri A, Roy A, Sathyakumar S. Elephants in the neighborhood: patterns of crop-raiding by Asian elephants within a fragmented landscape of Eastern India. PeerJ. 2020;8:e9399.
23. Naha D, Sathyakumar S, Dash S, Chettri A, Rawat GS. Assessment and prediction of spatial patterns of humanelephant conflicts in changing land cover scenarios of a human-dominated landscape in North Bengal. PLoS One. 2019;14(2):e0210580.
24. Pavithra GM, Venkataramana GV, Pasha SV, Reddy CS, Swarada B, Dadhwal VK. Geospatial identification of human-wildlife conflict hotspots in the Southern Western Ghats. J Indian Soc Remote Sens. 2025;53(12):4065-4080.
25. Pawar P, Mule M. Farmer-wildlife coexistence: decoding the pattern of human-wildlife conflict and compensation in the human-dominated landscape of Marathwada, western India. Wildl Res. 2025;52(9):WR24080.
26. Ramesh T, Kalle R, Milda D, Gayathri V, Thanikodi M, Ashish K, et al. Patterns of livestock predation risk by large carnivores in India's Eastern and Western Ghats. Glob Ecol Conserv. 2020;24:e01366.
27. Ramesh T, Milda D, Kalle R, Gayathri V, Thanikodi M, Ashish K, et al. Drivers of human-megaherbivore interactions in the Eastern and Western Ghats of southern India. J Environ Manage. 2022;316:115315.
28. Rani M, Panda D, Allen ML, Pandey P, Singh R, Singh SK. Assessment and prediction of human-elephant conflict hotspots in the human-dominated area of Rajaji-Corbett landscape, Uttarakhand, India. J Nat Conserv. 2024;79:126601.
29. Roy K, Pandey RK, Ganesan AN, Dutta A, Mittal D, Nigam P, et al. Two decades of human-elephant conflict in Jharkhand: spatial and ecological drivers of human fatalities. Ecol Evol. 2025;15(12):e72679.
30. Tripathy BR, Liu X, Songer M, Kumar L, Kaliraj S, Chatterjee ND, et al. Descriptive spatial analysis of humanelephant conflict (HEC) distribution and mapping HEC hotspots in Keonjhar forest division, India. Front Ecol Evol. 2021;9:640624.