Introduction
The balancing act of protecting the livelihood of individuals and communities while conserving wildlife species and habitat is a challenge I’ve always found fascinating. Human-wildlife conflict develops between many different species and stakeholders worldwide but seem to be particularly contentious between livestock producers and carnivore populations that prey on their stock. Carnivore-livestock conflict is a multifaceted issue that incorporates not just identifying or predicting areas of conflict, but also carnivore research techniques, landscape ecology, habitat evaluation and mapping, conservation practices and policy, developing mitigation strategies, and evaluating those strategies in specific situations. This literature review addresses how various GIS techniques and applications, spatial data, and analyses are used in studying and mitigating such conflict worldwide.
Annotated Bibliography
Akyazi, I., Y. Z. Ograk, E. Eraslan, M. Arslan, and E. Matur. 2018. Livestock guarding behaviour of Kangal dogs in their native habitat. Applied Animal Behaviour Science 201:61–66.
Livestock guarding dogs (LGD) play a significant role in non-lethal carnivore-livestock conflict mitigation. Specific traits in guard dog breeds make them ideal for this job, including attentiveness, trustworthiness, and protectiveness. Akyazi et al. studied the behavior of Kangals (a well-known LGD breed) in their native home of Sivas, Turkey using GPS data of dog, shepherd, and herd movements. The shepherd, dogs, and one sheep in the herd were fitted with GPS receivers to record their movements continuously, and the regular routine of grazing at night and coming back early morning was maintained. The mean speeds of the dogs, sheep, and shepherd were calculated, and events of moderate and high speeds by the dogs were classified as “running” or “sprinting”. The movements of the dogs and shepherd, their distances from each other, and their proximity ratio to the sheep were determined. The proximity ratio of dogs to sheep showed the dogs stayed closer to the shepherd rather than the herd, suggesting Kangals may establish closer relationships to the shepherd than the sheep. This study was a novel approach to studying LGD behavior, and while limitations of the study were acknowledged, the development of a spatial analysis metric for measuring effectiveness of LGDs is an important component of identifying best conflict mitigation practices.
Behdarvand, N., M. Kaboli, M. Ahmadi, E. Nourani, A. Salman Mahini, and M. Asadi Aghbolaghi. 2014. Spatial risk model and mitigation implications for wolf–human conflict in a highly modified agroecosystem in western Iran. Biological Conservation 177:156–164.
The Hamedan province in western Iran has seen increased wolf attacks on humans and livestock in the past several years which were classified as predatory. Successful wolf-human mitigation centers on a low level of conflict, and that these conflicts appear in a non-random pattern makes predicting areas of high risk a very significant tool in conflict mitigation. The authors of this study developed risk maps from data on livestock and human wolf attacks from 2001 to 2010 using Maximum Entropy Analysis (Maxent software). This method is considered one of the most effective presence-only models, estimating species distribution based on the probability of closest to uniform distribution and a specified set of environmental factors. When compared with data on wolf attacks in 2011 and 2012, the Maxent models indicated high levels of predictive power for both types of attacks (humans slightly better than livestock). Based on correlation results between the environmental factors and attacks, the models suggest the surge was caused by human alteration of natural habitat; these areas are aggregations of suitable resources that attract carnivores despite their traditional aversion to populated areas. Local governments can use the risk maps and knowledge of environmental variables to target their education efforts for husbandry and safety, and the spatial modeling methodology can be repeated in other countries who similarly lack robust biological data.
van Bommel, L., and C. N. Johnson. 2014. Where Do Livestock Guardian Dogs Go? Movement Patterns of Free-Ranging Maremma Sheepdogs. PLoS ONE 9:e111444.
Livestock guardian dogs (LGD) act as protectors and deterrents for predators and can be trained to stay within a fenced area or allowed to roam freely; this second strategy comes with more unknowns, such as the movement of the LGD away from or around the herd and the potential for territorial behaviors. This study tracked movements of free-ranging Maremma sheepdogs on three properties in Australia. The data was collected during fall and winter months using GPS collars on the dogs at all three properties, and a representative group of livestock at two properties. Spatial analysis was performed using fixed kernel density distributions and isopleths based on 10% density increases, calculated tortuosity (ratio of the actual path taken to the straight distance between two points), and minimum convex polygons of the area used each day. Results showed the LGDs spent most of their time with the herd and occupied ranges that extended up to 2 km beyond livestock paddocks. They moved in a sequential fashion that mirrored the herd, and the partial overlap with the other LGDs allowed the group to occupy a larger area than each individual could. When the dogs did move away from the livestock it was frequently at night and in direct and fast-paced movements, speculated to be chasing off predators or related to maintaining larger territories. Free-ranging LGDs can be a useful and effective management tool given a property size large enough to encompass the dogs’ range.
Broekhuis, F., S. A. Cushman, and N. B. Elliot. 2017. Identification of human-carnivore conflict hotspots to prioritize mitigation efforts. Ecology and Evolution 7:10630–10639.
This study identified hotspots of livestock loss by carnivores within bomas (livestock enclosures) in Maasailand, East Africa. As their home ranges shrink large carnivores are increasingly coming in contact with humans, and retaliatory killing of large carnivore species in Eastern Africa has been driving multiple species declines. Broekhuis et al. conducted interviews with community members in and around the Maasai Mara ecosystem in Kenya to describe the extent of the conflict and map hotspots based on the likelihood of losing livestock within bomas. They also looked at livestock loss in relation to livestock husbandry and used those results in conjunction with environmental factors to identify high risk areas. In the short-term overall livestock predation was higher than annual global figures at 3.4% of cattle and 3.6% of small stock during a three-month period. The results suggest that extensive closed vegetation landscape and proximity to protected areas influences carnivore predation, so these factors were used to map areas of higher and lower predation risk. Boma weakness was also significant for increased predation within the boma, and the authors recommend improvement to weakly constructed bomas be a priority over other deterrents. Using interviews results as attribute data adds detail and context to spatial data associated with livestock-carnivore conflict studies.
Davie, H. S., J. D. Murdoch, A. Lhagvasuren, and R. P. Reading. 2014. Measuring and mapping the influence of landscape factors on livestock predation by wolves in Mongolia. Journal of Arid Environments 103:85–91.
Landscape factors are often combined with carnivore density, prey biomass, or other variables to produce a detailed picture of predation risk, but when data on other variables is limited landscape factors can be used individually. In Mongolia the scarcity of grey wolf distribution data, the narrowness of wild prey options, and the migratory nature of the pastoral herdsmen make landscape variables the best place to start building predation risk models. Davie et al. used interview data of GPS marked predation sites input in ArcGIS to identify key landscape variables. The model was built using a partitioned Mahalanobis D2 (k) analysis, a simple but accurate technique for modeling presence only data that partitions the calculated distances between characteristics of known and unknown sites and selects those variables that do not vary across known locations. Livestock predation was most likely to occur where wolves were less likely to be detected, with risk increasing further away from ger camps (pastoral nomadic dwellings) and in areas of tall vegetation and shrubland and decreasing in open plains. Wolf distribution data or livestock husbandry practices should be added to improve the model, but this study demonstrates the range of spatial data that can be used to create predation risk models and the importance of tailoring spatial analysis to the available data in the context of unique ecosystems.
García-Rangel, S., and N. Pettorelli. 2013. Thinking spatially: The importance of geospatial techniques for carnivore conservation. Ecological Informatics 14:84–89.
The Carnivora taxa’s large home range sizes, wide global distribution, and charismatic nature draw conservation support for not just single species efforts but for entire ecosystems struggling with decline to which they belong (“keystone species”). They act as indicators of ecosystem decline and have direct and indirect effects on landscape and trophic ecology. GI systems and remote sensing create opportunities for carnivore studies traditionally limited by the taxa’s elusive nature, low population density, and large ranges. In ecology, remote sensing is used primarily for integrated ecosystem measurements, land cover classification, and detecting system changes. Ecologists use several GIS databases including biome and taxa distribution, land cover data, human impact, and climactic conditions. Specific habitat features and primary productivity can be compiled to create spatial distributions of suitable carnivore habitat, while examining spatial patterns of poaching and landcover change indicate possible areas of decline. Presence-absence models are a significant piece of geospatial carnivore studies, where GPS collared individuals provide data on species-environment relationships in more detail than is possible from presence only models (carnivore sightings). The standard financial and technological constraints associated with GIS and remote sensing apply to these studies, as well as the risk of digitizing error and the problem of combining spatial autocorrelation principles with independently characterized statistics. Future research directions for geospatial applications in carnivore conservation include multi-scale habitat selection and sustainability, incorporating human-modified landscapes into fragmentation and connectivity assessments, and landscape genetic studies.
Khorozyan, I., A. Ghoddousi, M. Soofi, and M. Waltert. 2015. Big cats kill more livestock when wild prey reaches a minimum threshold. Biological Conservation 192:268–275.
This study looked at livestock predation as a function of wild prey biomass density in big cat species around the world. Cases on lions, leopards, tigers, cheetahs, snow leopards, and pumas were compiled from peer reviewed journal articles that included information on livestock predations, prey biomass, cat density, and prey density. Linear regression was run against a model for the dependence of sheep and goat predation on prey biomass and a model for the dependence of cattle predation on prey biomass. Threshold values of prey biomass were estimated from the models by setting a 50:50 chance outcome equating to predation or no predation. The cases were used to test the model and thresholds, then mapped in QGIS and ranked from low to high predation risk. The probability of cattle predation by big cat species increases significantly when the prey biomass hits below the threshold of 812.41 kg/km2 and for sheep and goats when it hits below 544.57 kg/km2. The prey biomass thresholds can be used to predict livestock predation by big cats and act as part of an early warning system for preventative livestock-felid conflict management. This model can be applied to any big cat species range across the world, using the specific biomass data of a local area as the input.
Kuiper, T. R., A. J. Loveridge, D. M. Parker, P. J. Johnson, J. E. Hunt, B. Stapelkamp, L. Sibanda, and D. W. Macdonald. 2015. Seasonal herding practices influence predation on domestic stock by African lions along a protected area boundary. Biological Conservation 191:546–554.
Areas of livestock grazing that border protected areas (PAs) are particularly susceptible to livestock predation, especially by large carnivores like lions. Kuiper et al. added a temporal aspect to their livestock-carnivore spatial analysis study by looking at cattle predation by lions in relation to seasonal herding patterns in Zimbabwe. The movements of cattle were recorded using GPS data collars and depredation surveys were conducted monthly. The distance of cattle to the boundary of a PA and from the home enclosure were calculated in ArcMap, and the monthly differences were input into mixed effects analysis models. The cattle are herded further from home and into more woody areas during the wet growing season to keep them from grazing on the new crops, and predation rates by lions are subsequently higher. Once the crops are harvested the cattle are kept closer to graze in the fallow fields. The combined distance from the PA boundaries and the nearby presence of people results in a lower rate of predation by lions. Predicting livestock-carnivore conflict seasonally gives another dimension to mitigation plans, which can be tailored to different seasons. In this case, plans can focus on livestock protection during the wet season by incorporating human and LGD presence and avoiding woody areas.
Laporte, I., T. B. Muhly, J. A. Pitt, M. Alexander, and M. Musiani. 2010. Effects of Wolves on Elk and Cattle Behaviors: Implications for Livestock Production and Wolf Conservation. PLoS ONE 5:e11954.
This study examined the effect of wolf presence on cattle and elk populations to compare anti-predator responses between the two species. Domesticated cattle may have lost anti-predator behaviors, but if they do exhibit them it could result in a loss of fitness. By fitting wolves, cattle, and elk located in southwest Alberta, Canada with GPS and satellite radio collars, Lapotre et al. identified several variables characterizing ani-predator responses in both cattle and elk including speed and path directness, distance to forested areas, slope, and terrain ruggedness. Cattle responses also included time spent with heads up and distance to others in the herd. GIS queries defined wolf presence as when their telemetry location was in a cattle pasture, within a 1.5 km buffer of a cattle location, or within a home range of an elk location. The variables were analyzed in relation to wolf presence events using statistical and spatial analysis. Both species modify their behavior in response to predator detection; elk utilize terrain and changes in direction, whereas cattle showed a variety of responses that may indicate inexperience with predators or less developed anti-predator instincts. Looking at the indirect spatial effects of carnivores on livestock is an important component of developing a better understanding of livestock-carnivore conflict.
McInturff, A., J. R. B. Miller, K. M. Gaynor, and J. S. Brashares. 2020. Patterns of coyote predation on sheep in California: A socio-ecological approach to mapping risk of livestock-predator conflict. Conservation Science and Practice:e175.
Livestock-predator conflict combines socio-ecological aspects with ecological processes but is predominantly studied using predation risk modeling. This technique is easy to interpret but doesn’t address risk perception, or the beliefs held by producers about levels of predation risk in the production landscape. This article suggest risk perception needs to be combined with predation risk modeling for a more effective mitigation strategy. The authors present a case study of sheep predation by coyotes in California, where maps of predation risk modeling, producer risk perceptions, and perceived environmental factors were developed. Areas in all maps were classified as low, medium, or high risk, then were subtracted from each other using the ArcGIS raster calculator. There were big differences between the predation risk model and risk predation map (only 8.7% agreement), and more similarities between risk perception and producers’ linking environmental factors with predation risk (62.0% agreement). These results demonstrate how socio-ecological aspects can be classified and studied using spatial data and analysis. Applying predation risk models and risk perception in tandem is more beneficial for management strategies than relying on just one. The producer’s perceived risk of an area informs their husbandry decisions and effects where predation can occur and conclude perceived risk models are beneficial at larger scales while predation risk models offer fine scale (within pastures) management practices.
Miller, J. R. B. 2015. Mapping attack hotspots to mitigate human–carnivore conflict: approaches and applications of spatial predation risk modeling. Biodiversity and Conservation 24:2887–2911.
Predation risk modeling is a key spatial statistical tool in identifying conflict hotspots between carnivores and livestock. Initially designed to look at ecological interactions between predators and their wild prey, predation risk modeling for conflict hotspots focuses on the relationship between carnivore hunting strategies and livestock avoidance/response strategies. Miller’s review of the literature, encompassing 18 studies spanning 6 continents and 10 years, critiques approaches to building predation risk models, what landscape attributes were associated with hotspots, and what (if any) impact the results had on management strategies. Models were based on correlation modeling, spatial association, and spatial interpolation. Miller identified 3 levels of management that the predation risk models were shared at: livestock owners and households effected by livestock predation, government agencies and conservation groups, and one study had a successful impact on the policy-making level. Based on the results of the literature review, Miller created a 6 step process for integrating predation risk modeling into conservation efforts and identified key areas for improvement. Models need to be updated as new behavioral data comes in, models should be validated with independent data, new technology needs to be utilized to share model results which needs to be readily available for policymakers, and models should be utilized in areas of high livestock depredation across the world.
Moa, P. F., I. Herfindal, J. D. C. Linnell, K. Overskaug, T. Kvam, and R. Andersen. 2006. Does the spatiotemporal distribution of livestock influence forage patch selection in Eurasian lynx Lynx lynx? Wildlife Biology 12:63–70.
Moa et al. analyzed the overlap between distributions of lynx, roe deer, semi-domesticated reindeer, and sheep to determine the effect of livestock on lynx forage patch selection in Norway. Data from 10 radio collared lynx was inputted into a GIS and 100% minimum convex polygon (MCP) home ranges and seasonal variability in space use were calculated for each lynx. Ungulate distribution was determined from previously collected data and provided by sheep farmers and reindeer herders. Distributions of roe deer, reindeer, and sheep were mapped for winter, summer, and both winter and summer combined. Logistic regression models were run on the distribution data and results showed that contrary to expectations lynx did not show any preference toward sheep patches but did show clear preference toward roe deer patches for both summer and winter seasons. Predation on livestock is more about chance encounters rather than the lynx actively seeking out livestock. This study has implications for conflict mitigation in Norway; if potential conflict is based more on chance encounters, non-lethal deterrents should help keep depredation down (carnivores that actively seek out livestock aren’t always phased by deterrents). Describing the behavior of different carnivore populations using spatial analysis is a valuable component of mitigation planning.
Morehouse, A. T., J. Tigner, and M. S. Boyce. 2018. Coexistence with Large Carnivores Supported by a Predator-Compensation Program. Environmental Management 61:719–731.
Compensation programs for livestock depredation are a popular indirect mitigation strategy for carnivore-livestock conflict, providing monetary compensation for livestock producers who have lost stock to carnivores. The compensation program for the province of Alberta, Canada is more than 50% funded by levies from hunting and fishing licenses; there has been recent push-back from sportsmen due to increased compensation program costs. This study compared spatial patterns of livestock depredation and compensation to hunting effort patterns to evaluate the finances associated with the compensation program. Compensation records from the Alberta Conservation Association were assigned postal addresses and were input into ArcMap along with hunter effort data. The highest levels of depredation events were located on private land that bordered public lands, which was also where the greatest hunter effort was distributed. Private lands appear to support habitat for wild ungulates as well as carnivores, and compensation programs that draw revenue from hunting and fishing indirectly boost the maintenance of wildlife populations on private land. Conflict mitigation strategies have little weight unless they are evaluated for their effectiveness in different situations, and using spatial data is an important evaluation technique.
Mori, E., S. Bagnato, P. Serroni, A. Sangiuliano, F. Rotondaro, V. Marchianò, V. Cascini, L. Poerio, and F. Ferretti. 2020. Spatiotemporal mechanisms of coexistence in an European mammal community in a protected area of southern Italy. Journal of Zoology 310:232–245.
This article explored patterns of coexistence between a large carnivore species (grey wolf), its wild prey species, livestock, mesocarnivores, and humans by looking at spatial and temporal overlap. Pollino National Park in southern Italy is home to grey wolves and three wild ungulate species and has a large amount of cattle left unattended to graze in the park’s interior. Data was collected over 5 years using camera traps placed at 42 stations in the study site, the locations of which were fixed georeferenced locations randomly selected beforehand in Q-GIS. Analysis for spatiotemporal patterns was run using the R package Overlap. Temporal overlap was high between wolves, wild prey species, and mesocarnivores and lower for humans and livestock. Spatial overlap was moderate between wolves and livestock and low between wolves and wild prey species. Wolves appear to maximize encounters with their top prey species by shifting temporal behaviors and prey species employ spatial avoidance strategies reducing spatial overlap. The moderate spatial and low temporal overlap of wolves and cattle suggest that although cattle are the second top component in this population’s diet its likely an opportunistic food source and may be related to scavenging opportunities. Combining data collection and management in a GIS with detailed statistical analysis is a powerful technique for carnivore-livestock conflict management research.
O′Neil, S. T., K. C. Rahn, and J. K. Bump. 2014. Habitat Capacity for Cougar Recolonization in the Upper Great Lakes Region. PLOS ONE 9:e112565.
Natural and artificial (translocated) carnivore recolonizations have been occurring more frequently as conservation efforts stabilize carnivore populations. O’Neil et al. used GIS to create a cougar habitat model for the Upper Great Lakes Region in anticipation of cougar recolonization in the Midwest. An expert-assisted spatial habitat model was applied to the study area and was validated using confirmed cougar sighting data in both states. The model was built by using weighted parameters deemed essential for cougar habitat by expert-based rankings. Data for these parameters was acquired from the US Census Bureau, National Elevation Dataset, and the Michigan and Wisconsin land cover and water recourse datasets. After establishing the northern regions of both states as areas of suitable habitat, O’Neil et al. added in a potential prey biomass model and determined more than 500 cougars could inhabit the area overall. They caution the analysis is exploratory because the limited cougar data was likely from transient individuals and the nature of incidental sightings varies geographically. In the context of livestock-carnivore conflict, studies like this provide information on GIS techniques for selecting low conflict areas that can be recolonized naturally or through translocation of problem subpopulations or individuals.
Reyna‐Sáenz, F., M. M. Zarco‐González, O. Monroy‐Vilchis, and X. Antonio‐Némiga. 2020. Regionalization of environmental and anthropic variables associated to livestock predation by large carnivores in Mexico. Animal Conservation 23:192–202.
While studies of livestock-carnivore conflict over broad areas are helpful at the national or international level, factors influencing livestock predation usually fluctuate at the regional or local scale, making generalized solutions less effective. This study identified environmental, anthropic, and livestock management variables influencing livestock predation by jaguars, pumas, and brown bears at a regional level within Mexico. Data was collected primarily from institutional and non-institutional documentation. A buffer the size of the species home range was drawn around each predation record to account for influences from the surrounding area, and variables were assessed within each buffer zone. Factorial analysis was used to determine which variable had the greatest contribution for each record, and clusters were developed, mapped, and uniquely described based on the variables with the highest scores. Four livestock management variables were most significant over all three species, with various environmental and anthropic variables contributing. Management strategies can be prioritized based on the influential variables for each cluster; for example, cattle confinement was proposed for areas with high percentages of extensive grazing. Running regionalized spatial analysis on livestock predation data identifies conflict hotspots and prioritizes mitigation strategies based on local and regional influences.
Santos, M. J., L. M. Rosalino, M. Santos-Reis, and S. L. Ustin. 2016. Testing remotely-sensed predictors of meso-carnivore habitat use in Mediterranean ecosystems. Landscape Ecology 31:1763–1780.
Traditional methods of classifying ecosystem type, structure, and function are useful inputs for usage models but can be time consuming and incomplete at different scales. Remote sensing opens the door for more efficient and comprehensive datasets from which to classify these inputs. Santos et al. assessed if combining ecosystem type, function, and structure data from remote sensing produced better usage results for mesocarnivores than ecosystem type alone. Radio tracking data was collected from genets, European badgers, and stone martins in southern Portugal and compared with aerial photography and Landsat data of vegetation type, productivity, structure and drought stress effects. They used univariate generalized linear mixed models (GLMM) and Principal Component Analysis (PCA) to select the metrics that best explained the individuals’ locations and selected the model with the top performance. The best model for badgers and stone martins was a combination of the three ecosystem variables as predicted, but the best for genets was based on ecosystem type alone. This result could be explained by genets being habitat generalists. While combining ecosystem type, function, and structure data from remote sensing is more informative for predicting usage, it may be limited to finer spatial scales. Access to Landsat and other pre-exiting remote sensing databases provides mitigation conflict researchers and policy makers the ability to make better informed decisions.
Weise, F. J., J. R. Lemeris, S. J. Munro, A. Bowden, C. Venter, M. van Vuuren, and R. J. van Vuuren. 2015. Cheetahs (Acinonyx jubatus) running the gauntlet: an evaluation of translocations into free-range environments in Namibia. PeerJ 3:e1346.
Translocating identified conflict individuals is a common practice for mitigating carnivore conflict, but like all mitigation strategies has varying levels of success and needs situational evaluation. This study evaluated the outcome of cheetah translocations in Namibia by releasing trapped cheetahs with GPS and VHF radio collars and developing indicators of translocation success. Exploratory movements of released cheetahs were calculated using Minimum Convex Polygon (MCP) values, and site fidelity was measured by calculating percentage overlap of the daily location of the cheetah with the release reserve/protected area. Survivorship estimates and prey diet were also determined. Success of the translocation was strongly determined by the ability of an individual to survive the first year; survivorship was lowest in the first three months and overall site fidelity was low. The authors used these results and the geoprocessing tool CaTSuiT in ArcGIS to model area suitability for receiving translocated cheetahs across conservation areas. There were very few sites able to accommodate the vast exploratory distances observed in translocated cheetahs without them moving onto private land. In this situation translocations should be used as a last resort and conservation efforts should focus on mitigation on site and increasing tolerance for cheetahs. Evaluating translocation as a mitigation strategy is done most effectively with both spatial data and analysis in a GIS.
Whittington-Jones, B. M., D. M. Parker, R. T. F. Bernard, and H. T. Davies-Mostert. 2011. Habitat Selection by Transient African Wild Dogs (Lycaon pictus) in Northern KwaZulu-Natal, South Africa: Implications for Range Expansion. African Journal of Wildlife Research 44:135–147.
African wild dogs are an IUCN listed endangered species whose populations are threatened by habitat fragmentation, persecution in response to perceived or real threats, and human resource expansion. This study looks at modeling suitable habitat for African wild dogs to determine best locations for range expansion and connectivity, a key factor in conservation of larger carnivores who live in pack structures with transient individuals moving between them. Radio collar data of individuals from 9 packs in 3 protected areas in the KwaZulu-Natal province was collected and combined with a smaller number of reported sightings to develop Maximum Entropy distribution models. Environmental variable constraints for the Maxent model were extracted from raster and vector datasets in ArcMap. Transient individuals and pack structures showed preference for low-lying elevation such as shallow valleys and land cover including woodland habitat while avoiding areas of high human activity. Overlaying the results with a map of protected areas showed a good amount of suitable habitat for range expansion falls in protected parks and game reserves. This article demonstrates how GIS can be used to inform conservation plans that also act as preventative mitigation strategies for a species that could be under significant threat from future carnivore-livestock conflict.
Winterbach, H. E. K., C. W. Winterbach, and M. J. Somers. 2014. Landscape Suitability in Botswana for the Conservation of Its Six Large African Carnivores. PLoS ONE 9:e100202.
The purpose of this study was to develop national scale maps of habitat suitability for Botswana’s large carnivore guild (lion, leopard, spotted hyena, brown hyena, cheetah, and African wild dog) based on wild ungulate biomass overlaid on a base of land classifications, and use wild prey biomass as a percentage of the total biomass to evaluate areas of high risk in agricultural land. Biomass for large and small prey and livestock was calculated from areal grid surveys conducted by Botswana Department of Wildlife and National Parks. Large prey biomass was primarily located in the conservation zone land classifications and was lower in agricultural zones, whereas small prey biomass was higher in agricultural zones. Carnivores that hunt large prey also hunt large livestock and are frequently killed in retaliation. They are less likely to range outside of conservation zones, so their management should focus on boundaries between conservation and agricultural zones and population connectivity. Carnivores preferring smaller prey and livestock are more likely to persist in agricultural zones and need mitigation strategies focused on conflict hotspots and connectivity. This study demonstrates the development of national conservation strategies using spatial data display and analysis.
Supplementary Literature
The following sources were compiled and reviewed in the initial literature search but excluded from the bibliography due to age of research, repeated techniques already described in the articles above, or subject material that had only minimal connections to carnivore-livestock conflict or spatial/GIS applications. They can be used as additional reference material, particularly the studies that repeat the same analysis techniques on different systems.
Angelieri, C. C. S., C. Adams-Hosking, K. M. P. M. de B. Ferraz, M. P. de Souza, and C. A. McAlpine. 2016. Using Species Distribution Models to Predict Potential Landscape Restoration Effects on Puma Conservation. PLOS ONE 11:e0145232.
Kissling, W. D., N. Fernández, and J. M. Paruelo. 2009. Spatial risk assessment of livestock exposure to pumas in Patagonia, Argentina. Ecography 32:807–817.
McNeill, A., L. Leung, M. Goullet, M. Gentle, and B. Allen. 2016. Dingoes at the Doorstep: Home Range Sizes and Activity Patterns of Dingoes and Other Wild Dogs around Urban Areas of North-Eastern Australia. Animals 6:48.
Naha, D., S. Sathyakumar, and G. S. Rawat. 2018. Understanding drivers of human-leopard conflicts in the Indian Himalayan region: Spatio-temporal patterns of conflicts and perception of local communities towards conserving large carnivores. PLOS ONE 13:e0204528.
Nelson, A. A., M. J. Kauffman, A. D. Middleton, M. D. Jimenez, D. E. McWhirter, and K. Gerow. 2016. Native prey distribution and migration mediates wolf (Canis lupus) predation on domestic livestock in the Greater Yellowstone Ecosystem. Canadian Journal of Zoology.
Rostro-García, S., L. Tharchen, L. Abade, C. Astaras, S. A. Cushman, and D. W. Macdonald. 2016. Scale dependence of felid predation risk: identifying predictors of livestock kills by tiger and leopard in Bhutan. Landscape Ecology 31:1277–1298.
Ruda, A., J. Kolejka, and T. Silwal. 2018. GIS-Assisted Prediction and Risk Zonation of Wildlife Attacks in the Chitwan National Park in Nepal. ISPRS International Journal of Geo-Information 7:369.
Smith, J. B., C. K. Nielsen, and E. C. Hellgren. 2016. Suitable habitat for recolonizing large carnivores in the midwestern USA. Oryx 50:555–564.
Soh, Y. H., L. R. Carrasco, D. G. Miquelle, J. Jiang, J. Yang, E. J. Stokes, J. Tang, A. Kang, P. Liu, and M. Rao. 2014. Spatial correlates of livestock depredation by Amur tigers in Hunchun, China: Relevance of prey density and implications for protected area management. Biological Conservation 169:117–127.
Contact Information
This literature review was compiled for GIScience I: Introduction to Geographic Information Science (GEOG 560) for fall term 2020 at Oregon State University. For questions and comments please email Bridget Regan
