Pattern of crop raiding by wild large mammals and the resultant impacts vary with distances from forests in Southwest Ethiopia

Abstract Crop raiding is a major form of human‐wildlife interaction mainly in the ecotone areas of human‐modified natural landscapes. The aim of this study was to examine the spatial pattern of crop raiding and the resultant impacts on how farmers perceive forests at different distances from Yayu Coffee Forest Biosphere Reserve which is located in southwest Ethiopia. For this, thirty transects (each 1 km long) were laid out at 200 m interval parallel to forest edges: ten transects close to forest (<0.5 km), ten at intermediate (0.5–1 km), and ten transects were taken far from forest (>1 km). Along each transect, 2–6 households were randomly selected and interviewed using semistructured questionnaire. The perception of the respondents on forests at different distances from forest edges was analyzed using Pearson's Chi‐square test. The variation in the amount of damage among these three locations was tested using one‐way ANOVA. Four wild large mammals including olive baboon, vervet monkey, bush pigs, and crested porcupine were identified as top crop raiders in the area. The frequencies of occurrence of crop raiders decreased with increasing distance from forest edges. Similarly, the amount of damage in maize fields was higher close to forests when compared with that of either at intermediate or far from forest edges (p < .001). Eighty‐one percent of the households living close to the forests perceive that forest is a threat to their survival. Overall, our results imply that strategies need to be sought in order to minimize the socio‐ecological impacts of crop raiders mainly in locations close to forest edges.

and blue monkeys were reported as a worst crop raiders in both field crop and homegardens (Lemessa et al., 2013;Ango et al., 2014).
However, the pattern of the frequency of occurrence or raiding and the corresponding crop damage may vary with distance gradient from natural habitats into human-modified landscapes (Lemessa et al., 2013;Madden, 2008). Moreover, the intensity of crop raiding depends on type of crop raider species, crop species grown, and season (Mwamidi et al., 2012). The study was executed by taking thirty transects (each 1 km long) which were laid out with 200 m interval from forest edge toward agricultural landscape. Moreover, these transects were categorized into three locations: ten transects close to forest (<0.5 km), ten intermediate (0.5-1 km), and ten transects far from forest (>1 km). Accordingly, from each transect, 2-6 households (i.e., 48 close to forest, 38 intermediate, and 38 far from forest households) and in total 124 households were randomly selected to undertake a household questionnaire using semistructured questionnaire. Altogether, the results showed that the pattern of crop raiding, the extent crop damage and how farmers perceive forests vary with distances from the forest edge surrounding Yayu Coffee Forest Biosphere reserves.

| Study area
The Yayu Coffee Forest Biosphere Reserve is located in Illubabor Zone of Oromia National Regional State, southwest Ethiopia between the geographical coordinates of 8°10ˈ0″N-8°20ˈ0″ N latitudes and 35°40ˈ0″ E-36°0ˈ0″ E longitudes (Figure 1). The elevational range of the study area is between 1,140 and 2562 m a.s.l. (Gole et al., 2008). Yayu Coffee Forest belongs to the moist evergreen afromontane vegetation type F I G U R E 1 The map of the study area in relation to the map of Ethiopia produced from satellite image in Google Earth (Image©2020 CNES/Airbus), white dots are the location of the households used for interview, dark areas are forests, and white gray fields are agricultural areas in the landscape (Friis et al., 2010) and was designated as biosphere reserve in 2010 by UNESCO (Gole et al., 2009). The total area of the Yayu Coffee Forest biosphere reserve is 16.7 Km 2 (Fekensa et al., 2016). The major dominant woody species of the area include Albizia spp., Cordia Africana, Pouteria adolfi-friedericii, Ficus spp., and Polyscias fulva (Senbeta, 2006) and the common wild large mammals found in the study area are lions, buffalos, bush pigs, warthog, forest giant hog, porcupines, olive baboons, and vervet monkeys (Lemessa et al., 2013;Quirin, 2005).
The area gets the mean annual precipitation of 2,100 mm that ranges between 1,400 and 3,000 mm with a unimodal pattern (Friis et al., 2010; The area has hot and humid climate and the mean annual temperature of 20°C where the minimum and maximum ranges between 12°C and 29°C Gole et al., 2009).
The local people largely depend on agriculture for their livelihoods and a mixed farming system is exercised in the area where production comprises coffee and cereal crops, animal husbandry, beekeeping, and spices. The major cereal crops include maize, sorghum, teff, wheat, burley, and millet, while coffee is the major cash crop accounting over 60% of the households' annual income (Gole, 2003;Fekensa et al., 2016).

| Data collection
The first 1 km long transect was laid out along the edges of Yayu Coffee Forest. In the next, 29 transects (each 1 km long) were arranged with 200 m interval in parallel to the first transect and to each other in the agricultural landscape. These transects were categorized into three locations: ten transects close to forest (<0.5 km), ten intermediate (0.5-1 km), and ten transects far from forest (>1 km) in similar method used by Lemessa et al. (2013). By walking along each transect in zigzag way turning right and left, the households were randomly selected for questionnaire survey. Accordingly, from each three locations from forest edges (i.e., 48 households close to, 38 intermediate and 38 far away from forest households) or 2-6 households from each thirty transects and in total 124 households were interviewed using semistructured questionnaire. This household level interview was executed during the crop growing and maturity seasons from 1 February 2019 to 30 July 2019.
The major data collected were the type of crop raiders and the frequency of occurrences per year, crop species liable to be raided, extent of crop damage and how the households perceive the about the Yayu Coffee Forest. The scientific names of the large mammals were identified using field guide of large mammals of Africa after their local or common names in Afaan Oromo were collected during the questioner survey (Stuart, 2006). Moreover, after the maize crop was identified as the most liable to crop raiding, one maize field of each of 22, 21, and 21 households used for questionnaire survey was, respectively, selected from the three locations close to forest, intermediate, and far away from forest edge at random to estimate the extent of crop raiding. Correspondingly, four permanent plots (size: 4 × 4 m each) were randomly established in each of the 64 maize sown fields of the households (0.5 ha each on average) in March 2019 and at the maturity stage the number of damaged stems due to crop raiding were recorded in July, 2019. Moreover, the amount of annual yield loss estimated by respondents was compared with the yield loss we computed based on the secondary data of the average annual yield of maize from Yayu District Agricultural Development Office of the study area (YDADO, 2019). Here, the percent of annual yield loss due to crop raiding was calculated in percent from the difference between the amount of annual maize yield from the field survey (actual yield of maize) and average annual yield of maize obtained from YDADO.

| Data analyses
The perception of the households on forests in relation to the impact of crop raiding at the three locations from forest edges (i.e., close to forest, intermediate, and far away from forest edges) was tested using Pearson's Chi-square test. The number of maize stems attacked by crop raiders was used as a proxy indicator for the amount of damage by crop raiders. The number damaged maize stems were square-root transformed to meet the assumption of homoscedasticity of the variance. To test for the variation in the amount of damage among the three locations (close to, intermediate, and far from forest edges) one-way ANOVA was run and after the significant variation was found among the three locations, a multiple comparison of the means was made using HSD Tukey test with bonferroni p-value adjustment. All data were analyzed using R-statistical program version 4.0.2 (R Core Team, 2020).

| Crop raiders and crops most raided in relation to distances from forest edges
Four wild large mammal pests, such as olive baboon (Papio anubis), vervet monkey (Chlorocebus aethiops), bush pig (Potamochoerus larvatus), and crested porcupine (Hystrix cristata) were identified by all of the respondents as the worst crop raiders in the area. According to the respondents, the crops that are most raided by these and other wild large animals are maize, sorghum, potato, mango, avocado, and teff. In this connection, the respondents close to forests (67%), at intermediate distance (34%), and the respondents at far away from forests (42%) mentioned that from among the crops identified above, maize is reported as the most liable crop to be raided by wild large mammals in the area (Table 1).

| The pattern of crop raiding with distances to forest edges
The frequency of occurrence of crop raiders in agricultural field varies spatially and decreases with increasing distance from the forest edge (Figure 2). The severity of crop raiding is dependent on the different locations from the forest edges (χ 2 = 8.04, df = 2, p = .018). The households close to the forests face severe crop raiding impact than those living far away from Yayu Coffee Biosphere Reserve. Here, the respondents living close to forests (89.6%) and at intermediate distances from forests (28.9%) estimated that the annual maize yield loss due to crop raiding ranges between 20% and 40%, while 100% of the respondents dwelling at far away from forests indicated that the annual maize is <5% (Table 2) The perception of respondents on forests vary among the different locations from forest edges that respondents living close to forests (81%), at intermediate distance from forests (65.8%), and respondents situated at far away from forests (52.6%) of, explained that forest is home for wild large mammals and hence is a threat to our survival (Table 3). However, how the respondents perceive the trend of crop damage by wild large mammals vary among respondents in that 91.1% (113) of the respondents stated that the extent of damage is increasing while 8.9% (11) of the respondents assume that it is not possible to predict.

| D ISCUSS I ON
Understanding how crop raiding varies across spatial scales from forest edges is critically essential in subsistence farming system to frame harmonized development-conservation schemes. We show  (Ango et al., 2017;Geleta et al., 2019;Lemessa et al., 2013;Leta et al., 2016). However, there is a variation in the magnitude of distances that crop raiders move out from their natural habitats into agricultural fields including in homegardens for raiding different kinds of crops. Our study indicated that not only the frequency of occurrence and extent of crop damage close to the forest is higher for diurnal crop raiders such as olive baboons, and   (Table 2). On the hand, 100% of the respondents at far away from forests asserted that the yield loss due to crop raiding is <5%/ of the total expected annual yield per ha close to forests and 80 kg (11.5%) at away from forests in southwest Ethiopia. Moreover, the present estimation is also close to what was reported from Uganda that the seasonal amount of yield loss by wildlife reaches up to 50% (Wallace & Hill, 2007). gest that strategies need to be sought in order to minimize the socio-ecological impacts of crop raiders mainly in locations close to forest edges.

ACK N OWLED G M ENTS
We would like to thank College of Agriculture and Veterinary Medicine, Jimma University (JUCAVM) for financial and logistics support. We are also grateful to our respondents for they are willing for sharing us their experiences. Finally, our gratitude goes to the field assists who wholeheartedly participated in data collection during the study period.

CO N FLI C T O F I NTE R E S T
None declared.

E TH I C A L A PPROVA L
This study is based on the questionnaire survey and observation or assessment of the crops damaged by large mammal pests and did not involve human subjects, experimentation with animals or collection of specimens. We all authors abide by the Ethical standards stated in the Ecology and Evolution journal Guidelines for Authors.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data used in the manuscript were deposited in Dryad repository