Accounting for Imperfect Detection Is Critical for Inferring Marine Turtle Nesting Population Trends

Assessments of population trends based on time-series counts of individuals are complicated by imperfect detection, which can lead to serious misinterpretations of data. Population trends of threatened marine turtles worldwide are usually based on counts of nests or nesting females. We analyze 39 years of nest-count, female-count, and capture-mark-recapture (CMR) data for nesting loggerhead turtles (Caretta caretta) on Wassaw Island, Georgia, USA. Annual counts of nests and females, not corrected for imperfect detection, yield significant, positive trends in abundance. However, multistate open robust design modeling of CMR data that accounts for changes in imperfect detection reveals that the annual abundance of nesting females has remained essentially constant over the 39-year period. The dichotomy could result from improvements in surveys or increased within-season nest-site fidelity in females, either of which would increase detection probability. For the first time in a marine turtle population, we compare results of population trend analyses that do and do not account for imperfect detection and demonstrate the potential for erroneous conclusions. Past assessments of marine turtle population trends based exclusively on count data should be interpreted with caution and re-evaluated when possible. These concerns apply equally to population assessments of all species with imperfect detection.

基于个体数量时序统计的种群趋势评估，常因检测不完善问题变得复杂，甚至可能引发数据的严重误读。全球受胁海洋海龟的种群趋势评估，通常依托巢穴或筑巢雌性个体的数量统计开展。本研究针对美国佐治亚州瓦萨夫岛（Wassaw Island）的筑巢蠵龟（Caretta caretta），分析了其长达39年的巢穴计数、雌性个体计数以及标记重捕（Capture-Mark-Recapture, CMR）数据。未对检测不完善情况进行校正的年度巢穴与雌性个体计数结果，显示种群丰度呈现显著的正向增长趋势。然而，通过对CMR数据开展考虑检测不完善动态变化的多状态开放稳健设计建模分析后发现，在这39年间，筑巢雌性个体的年度种群丰度基本保持稳定。这种结果分歧可能源于调查方法的改进，或是雌性个体在繁殖季内筑巢位点保真度的提升——二者均可提高检测概率。本研究首次在海洋海龟种群中，对比了考虑与未考虑检测不完善情况的种群趋势分析结果，并证实了忽略检测不完善可能得出错误结论的风险。以往仅基于计数数据开展的海洋海龟种群趋势评估，应当谨慎解读，并在可行条件下重新评估。这一警示同样适用于所有存在检测不完善问题的物种的种群评估工作。