Agroecosystems are experiencing a biodiversity crisis. Biodiversity monitoring is needed to inform conservation, but existing monitoring schemes lack standardisation and are biased towards birds, insects and plants. Automated monitoring techniques offer a promising solution, but while passive acoustic monitoring and remote sensing are increasingly used, the potential of camera traps (CTs) in farmland remains underexplored. We reviewed CT publications from the last 30 years and found only 59 articles that sampled farmland habitats in Europe. The main research topics addressed management or (avian) conservation issues, such as monitoring wildlife‐livestock interactions, nest predation, and the use of feeders and water troughs. Fewer studies employed landscape‐wide approaches to investigate species' habitat use or activity patterns over large agricultural areas. We discuss existing barriers to a more widespread use of CTs in farmland and suggest strategies to overcome them: boxed CTs tailored for small mammals, reptiles and amphibians, perch‐mounted CTs for raptor monitoring and time‐lapse imagery can help in overcoming the technical challenges of monitoring (small) elusive species in open habitats where misfires and missed detections are more frequent. Such approaches would also expand the taxonomic coverage of farmland monitoring schemes towards under‐surveyed species and species groups. Moreover, the engagement of farmers in CT‐based biodiversity monitoring programmes and advances in computer vision for image classification provide opportunities for low‐cost, broad‐scale and automated monitoring schemes. Research priorities that could be tackled through such CT applications include basic science topics such as unravelling animal space use in agricultural landscapes, and how this is influenced by varying agricultural practices. Management‐related research priorities relate to crop damage and livestock predation by wildlife, disease transmission between wildlife and livestock, effects of agrochemicals on wildlife, and the monitoring and assessment of conservation measures. Altogether, CTs hold great, yet unexplored, potential to advance agroecological research.

中文翻译：

---

相机陷阱在农田生物多样性监测中的未开发潜力：当前实践和突出的农业生态问题


农业生态系统正在经历生物多样性危机。需要生物多样性监测来为保护提供信息，但现有的监测计划缺乏标准化，并且偏向于鸟类、昆虫和植物。自动监测技术提供了一个很有前途的解决方案，但尽管被动声学监测和遥感的使用越来越多，但农田中相机陷阱 （CT） 的潜力仍未得到充分开发。我们回顾了过去 30 年的 CT 出版物，发现只有 59 篇文章对欧洲的农田栖息地进行了采样。主要研究主题涉及管理或（鸟类）保护问题，例如监测野生动物与牲畜的互动、巢穴捕食以及喂食器和水槽的使用。较少的研究采用景观范围的方法来调查物种在大片农业区域的栖息地利用或活动模式。我们讨论了在农田更广泛使用 CT 的现有障碍，并提出了克服这些障碍的策略：为小型哺乳动物、爬行动物和两栖动物量身定制的盒装 CT、用于猛禽监测的栖息式 CT 和延时图像可以帮助克服在开放栖息地监测（小型）难以捉摸的物种的技术挑战，因为在这些栖息地中，失火和漏检更频繁。这种方法还将扩大农田监测计划的分类覆盖范围，以涵盖调查不足的物种和物种群。此外，农民参与基于 CT 的生物多样性监测计划以及用于图像分类的计算机视觉的进步为低成本、大规模和自动化的监测计划提供了机会。 可以通过此类 CT 应用解决的研究重点包括基础科学主题，例如揭示农业景观中的动物空间利用，以及不同的农业实践如何影响这一点。与管理相关的研究重点涉及野生动物对农作物的破坏和牲畜捕食、野生动物和牲畜之间的疾病传播、农用化学品对野生动物的影响以及保护措施的监测和评估。总而言之，CT 在推进农业生态学研究方面具有巨大但尚未开发的潜力。