Wildlife detection dogs effectively survey a terrestrial amphibian, but differ among individuals, weather and habitat

The endangered natterjack toad (Epidalea calamita) faces ongoing habitat loss, requiring effective conservation measures. Field surveys must therefore address their highly variable detectability in terrestrial habitats. We investigated the factors influencing capture success comparing two methods: artificial cover boards (ACB) and wildlife detection dogs (WDD). We first analysed environmental and training factors influencing toad detectability for four detection dogs with varying experience using a binomial generalized linear model (GLM). Then, we compared capture success of ACBs and WDD transects (deploying two dogs) considering weather and habitat type using negative binomial GLMs. Across 200 tests, each detection dog displayed a unique learning curve, with performance influenced mainly by dog behaviour, wind, test blindness and insect presence. Detection rates across dogs levelled off at 87.2 % (75.7 % – 94.8 %). Capture success of both methods under real-deployment conditions was separately analysed for juvenile and (sub-)adult toads by calculating a success rate. For juvenile toads, capture success was primarily influenced by capture method, habitat type and study year, while adult toads capture success mainly depended on habitat type and precipitation. Overall, WDDs (mean rate juvenile toads = 4.15, mean rate adult toads = 5.61) showed higher success rates than ACBs (mean rate juvenile toads = 0.51, mean rate adult toads = 4.73), particularly in dense and versatile habitats. This study provides practical guidance for the selection and implementation of survey methods for E. calamita, identifying critical factors to consider when designing a study to maximise survey success. Dog handlers should be able to adapt search and training strategies to their dogs’ individual pace and limitations and keep environmental influences in mind. Both ACBs and WDDs are suitable methods for detecting E. calamita in terrestrial habitats, however, detection dogs are particularly useful in densely vegetated areas, for finding juveniles and for covering larger regions more efficiently. Success rates for adult toads increase in favourable habitat and climatic conditions, such as warmer temperatures and sufficient moisture. Our findings provide a framework that may be extended to the monitoring and conservation of other amphibian species.

濒危黄条背蟾蜍（Epidalea calamita）正面临持续的栖息地丧失，亟需有效的保护措施。因此，野外调查需应对该物种在陆地生境中检测率高度可变的问题。本研究对比人工遮盖板（artificial cover boards, ACB）与野生动物检测犬（wildlife detection dogs, WDD）两种方法，探究影响其捕获成功率的关键因素。研究首先采用二项广义线性模型（binomial generalized linear model, GLM），针对4只经验水平各异的检测犬，分析影响蟾蜍可检测性的环境与训练因素；随后结合天气与生境类型，采用负二项广义线性模型，对比了ACB与WDD调查样线（使用2只检测犬）的捕获成功率。在总计200次测试中，每只检测犬均展现出独特的学习曲线，其表现主要受犬只行为、风速、测试盲态以及昆虫出没情况的影响；各检测犬的检测率最终趋于稳定，达87.2%（75.7%~94.8%）。我们分别针对幼体与（亚）成体蟾蜍，计算了两种方法在实际部署场景下的捕获成功率。结果显示，幼体蟾蜍的捕获成功率主要受捕获方法、生境类型与研究年份影响，而成体蟾蜍的捕获成功率则主要取决于生境类型与降水量。整体而言，WDD的平均成功率（幼体蟾蜍：4.15，成体蟾蜍：5.61）高于ACB（幼体蟾蜍：0.51，成体蟾蜍：4.73），在植被茂密且生境多样的区域尤为显著。本研究为黄条背蟾蜍的调查方法选择与实施提供了实用指导，明确了设计最大化调查成效的研究方案时需考量的关键因素。检测犬训导员应根据犬只的个体节奏与局限性调整搜索与训练策略，同时兼顾环境因素的影响。ACB与WDD均适用于陆地生境中的黄条背蟾蜍检测，但检测犬在植被茂密区域、搜寻幼体以及更高效覆盖大范围区域时优势尤为突出。成体蟾蜍的检测成功率在适宜的生境与气候条件下会提升，例如较高的环境温度与充足的湿度。本研究结果可为其他两栖动物的监测与保护提供可推广的研究框架。