Studying Wildlife Activity to Mitigate Conflict with Agriculture

Mitigation of wildlife-agriculture conflict by studying patterns of animal damage to crops

August 2, 2020


Background. Arable land covers about 10% of the world's land area (World Bank), and nearly 20% of the land in Israel (Central Bureau of Statistics). This significant portion of terrestrial landscape is shared with wildlife, inherently involving conflicts between animal conservation and agriculture (Lemly et al., 2000). Mammals and birds may damage crops directly by eating and trampling (e.g.,wild boars), or indirectly by damaging infrastructure, mainly irrigation systems. Farmers dealing with animal damage, find themselves investing resources in preventive actions (e.g., fencing), fixing damage, or hurting wildlife. Recently in India, for instance, the killing of an elephant with explosive-stuffed fruit, raised awareness to this wildlife-agriculture conflict. Nevertheless, studies focusing on understanding the agro-ecological system in order to mitigate this conflict are still rare, and even the scales of financial costs scarcely everestimated. Locally, wild boars and jackals are among the most notorious mammals known for causing agricultural damages. Malkinson et al. (2006) estimated that wild boars alone cause an annual damage at the range of 2M ILS per year, in the Golan Heights region alone. Consequently, it is likely that at the national scale, wildlife damage to Israeli agriculture is at the scale of dozens of million ILS per year.

With such a considerable impact, it is urgent to improve our knowledge of wildlife behavior, and consequently, optimize solutions to mitigate animal-agriculture conflict. The generality of the problem and expenses involved highlight the potential benefits that may arise from even mild local improvements. Tracking wild animal behavior and movement is developing fast over the last decades, facilitated not only by the growing recognition of importance but also by technological advances and miniaturizations, such as GPS light-weight tags (Kays et al., 2015, Wilmers et al., 2015), and the use of motion sensitive trail cameras (Cutler et al., 1999). Here, we propose to use the latter for investigating wildlife-agriculture conflict and quantifying patterns of animal damage to crops. Specifically, we ask (1) which animals are active at a focal agricultural field? (2) At what time of the day are they active and hence more likely to damage crops? (3) What type of damage these animals cause? We predicted that nocturnal mammals damaging the irrigation system cause the major damage, and that agricultural field damage increase in correlation with scarcity of available drinking water for wildlife, or with increased summer temperatures.


Figure 1..Figure 1. An aerial image of the research site. The orchard size is 2600 m2 (0.26 ha) with 113 trees in 9 rows planted on an east-west alignment. Red angles mark cameras fixed to a tree and watching northward. Video cameras are marked yellow, and facing to the nearest water sprinkler, The different areas of camera placements are marked with a semi-transparent color rectangles – North, East, Inner cameras, South trail + West trail. Two black arrows at the bottom side of the map are pointing the location of additional cameras: a camera on the South trail, and a camera at the periphery, 500m apart.


To test our predictions, we placed 18 trail cameras (Dark OPS HD Pro X, Browning Trail Cameras, Morgan, UT) in a citrus orchard (Ora, a mandarin cultivar) located near kibbutz Haogen (32°22'N 34°55'E), Israel (Fig. 1). Within such a spatially confined array, observations are signifying animal activity rather than occupancy or abundance. Thirteen cameras were programmed to take still images at five separated sub-sections of our site: the northern part of the orchard, eastern side, south and west bordering trails, within the mandarin orchard, and in the periphery, 500m apart. Five additional cameras were set to capture short videos for identifying animal behavior at the vicinity of irrigation sprinklers. Some of the cameras were relocated while others were added to fill in gaps in the recordings. We deployed the cameras for two months (March-May 2020) and surveyed them every 10 days. For data analysis, we considered a single observation as one animal at a time per camera. For each observation, we manually recorded time, species, number of individuals, temperature, the direction the animal seems to be heading to, and any other notable behavior.


A  B 
 C D 
 E  F


Figure 2.Animal activity throughout the entire time of research at the regional scale – all 18 trail cameras. A) Total activity of birds percentage of total observations are given for the most active seven species of birds; B) Regional activity of mammals  – percentage of total observations are given for the most active four species of mammals; C) Golden jackals (Canis aureus) were the most active animal with 42% of total observations; D) A cape hare (Lepus capensis) in a daily activity near camera #3; E) Indian crested porcupine (Hystrix indica) passing by an irrigation hose; a wild boar (Sus scrofa) female walking in the orchard. The male's eye is shining at the back; G) Daily activity of Mammals is typically nocturnal, vs. birds whichtypically exhibit diurnal activity. To normalize data to the sampling effort, we divided the total number of observations per camera, by the number of days the camera was located and operated



Altogether, we had 855 active camera days (including 143 days of video recordings), with 45±15.7 days per camera resulting in an average of 5.7±3.8 observations per camera per day.Our dataset contained 5230 observations as images or videos, with 91.5±14.4% of the images detecting no wildlife, likely due to humanactivity or random vegetation movements which activated the motion sensor. We recorded a total of 11 mammal species, 24 bird species, and 2 species of reptiles. Despite the higher diversity of birds,a majority of 68.9% of thedetected observationswere of mammals (Figure 2). Unsurprisingly, birds dominate day hours, and mammals are more active at nighttime (Fig. 2G). A similar analysis is focused at the type of activity of mammals, as they were the most dominant and relevant taxon (Fig. 4). 484 short videos documented behaviors of passing by, standing, and foraging - none recorded any damage-related behaviors (e.g., biting the irrigation system).


Figure 3. Atypical activity time pattern of mammals captured by trail cameras, demonstrating their strong nocturnal preference. We excluded from this figure two mammalian species with less than 10 observations in total – the European badger (Meles meles) and the Egyptian fruit bat (Rousettus aegyptiacus). The Mongoose (Herpestes ichneumon) was exceptional by being the only diurnal mammal.


Discussion and conclusion.

Remote photography is a relatively accessible and powerful tool to gain basic understanding of animal activity types and patterns in time and space in an agriculture ecosystem. Our results from two months of camera tracking agree with normal expectations for wildlife composition and activity in the region. Mammals were mainly nocturnal, with jackals as the most frequently observed species, followed by Cape hare. Birds were largely diurnal with higher species richness dominated by synanthropic species (wild species living closely to humans). These included hooded crow (Corvus cornix), spur-winged lapwing (Vanellus spinosus), and blackbird (Turdus merula) being the most commonly observed species. To conclude, at this time of the year, we did not record any noticeable agricultural damage caused by wildlife. Incidentally, we have recorded an interesting behavior of jackals carrying fish in their mouth (from nearby fishponds), which may point at direct damage to aquaculture, but in a different context. However, our data cannot determine whether the jackals hunt on fresh fish, or scavenge on carcasses, providing a sanitation service to the ecosystem.  In addition, we have recorded birds foraging on invertebrates –which indicates their potential service ofbiological pestcontrol. These findings suggest that in order to fully understand the feedback between wildlife and agriculture one should consider both conflicts and potential benefits simultaneously, ideally while quantifying financial impact. We hope to expand our work and include direct animal tracking for improving the estimation and generality of agricultural damage and contributions.


Cutler, T. L., & Swann, D. E. (1999). Using remote photography in wildlife ecology: a review. Wildlife Society Bulletin, 571-581.‏
Kays, R., Crofoot, M. C., Jetz, W., & Wikelski, M. (2015). Terrestrial animal tracking as an eye on life and planet. Science, 348(6240), aaa2478-1-aaa2478-9.‏

Lemly, A. D., Kingsford, R. T., & Thompson, J. R. (2000). Irrigated agriculture and wildlife conservation: conflict on a global scale. Environmental Management. 25 (5): 485-512.‏

Malkinson, D., Saltz D., and Kaplan D. (2006), Monitoring and modeling wild boar population (Sus scrofa) for mitigation of damage in crops at the Southern Golan heights. Summarizing report for the Ministry of Science (Hebrew) 

Wilmers, C. C., Nickel, B., Bryce, C. M., Smith, J. A., Wheat, R. E., &Yovovich, V. (2015). The golden age of bio‐logging: how animal‐borne sensors are advancing the frontiers of ecology. Ecology, 96(7), 1741-1753.‏

* Department of Natural Resources, Institute of Plant Sciences, Agricultural Research Organization, RishonLeZion, Israel
**. Ecolo-GIS, POB 133, Berotaim 4285000, Israel;
**. School of Zoology, Faculty of Life Sciences, Tel Aviv University, Tel Aviv, Israel.  

  PDF file of the article

Related Articles