Tezuka Shiori, Tanaka Mii, Naganuma Tomoko, Tochigi Kahoko, Inagaki Akino, Myojo Hiroaki, Yamazaki Koji, Allen Maximilian L, Koike Shinsuke
Faculty of Agriculture, Tokyo University of Agriculture and Technology, Fuchu, Tokyo 183-8509, Japan.
Institute of Global Innovation Research, Tokyo University of Agriculture and Technology, Fuchu, Tokyo, 183-8509, Japan.
J Mammal. 2022 Nov 25;104(1):184-193. doi: 10.1093/jmammal/gyac101. eCollection 2023 Feb.
In recent years, animal-borne video cameras have been used to identify the food habits of many species. However, the usefulness and difficulties of identifying food habits from animal-borne video cameras have not been sufficiently discussed in terrestrial mammals, especially large omnivores. The aim of this study is to compare the video analysis of foraging behavior by Asian black bears () acquired by camera collars with estimates from fecal analysis. We attached GPS collars equipped with video cameras to four adult Asian black bears in the Okutama mountains in central Japan from May to July 2018 and analyzed video clips for foraging behavior. Simultaneously, we collected bear feces in the same area to determine food habits. We found that using video analyses was advantageous to recognize foods, such as leaves or mammals, that were physically crushed or destroyed while bears chewed and digested foods, which are difficult to identify to species using fecal analyses. On the other hand, we found that camera collars are less likely to record food items that are infrequently or quickly ingested. Additionally, food items with a low frequency of occurrence and short foraging time per feeding were less likely to be detected when we increased the time between recorded clips. As one of the first applications of the video analysis method for bears, our study shows that video analysis can be an important method for revealing individual differences in diet. Although video analysis may have limitations for understanding the general foraging behavior of Asian black bears at the present stage, the accuracy of food habit data from camera collars can be improved by using it in combination with established techniques such as microscale behavior analyses.
近年来，动物携带式摄像机已被用于识别许多物种的食性。然而，在陆生哺乳动物，尤其是大型杂食动物中，利用动物携带式摄像机识别食性的实用性和困难尚未得到充分讨论。本研究的目的是比较通过相机项圈获取的亚洲黑熊觅食行为的视频分析结果与粪便分析的估计结果。2018年5月至7月，我们在日本中部奥多摩山区给四只成年亚洲黑熊佩戴了配备摄像机的GPS项圈，并分析了觅食行为的视频片段。同时，我们在同一区域收集熊粪便以确定食性。我们发现，使用视频分析有助于识别熊在咀嚼和消化食物时被物理碾碎或破坏的食物，如树叶或哺乳动物，而这些食物很难通过粪便分析确定其种类。另一方面，我们发现相机项圈不太可能记录很少或很快被摄入的食物。此外，当我们增加录制片段之间的时间间隔时，出现频率低且每次进食觅食时间短的食物被检测到的可能性较小。作为视频分析方法在熊类中的首批应用之一，我们的研究表明，视频分析可以成为揭示个体饮食差异的重要方法。尽管目前视频分析在理解亚洲黑熊的一般觅食行为方面可能存在局限性，但通过将其与微观行为分析等既定技术结合使用，可以提高相机项圈食性数据的准确性。
