Department of Mathematics and Statistics, University of Otago, P.O. Box 56, Dunedin, 9054, New Zealand.
Ecology. 2019 Feb;100(2):e02580. doi: 10.1002/ecy.2580.
Spatially explicit capture-recapture (SECR) models have emerged as one solution to the problem of estimating the population density of mobile and cryptic animals. Spatial models embody assumptions regarding the spatial distribution of individuals and the spatial detection process. The detection process is modeled in SECR as a radial decline in detection probability with distance from the activity center of each individual. This would seem to require that home ranges are circular. The robustness of SECR when home ranges are not circular has been the subject of conflicting statements. Ivan et al. previously compared the SECR density estimator to a telemetry-scaled non-spatial estimator. I suggest that the apparent non-robustness of SECR in their study was a simulation artefact. New simulations of elliptical home ranges establish that the SECR density estimator is largely robust to non-circularity when detectors are spread in two dimensions, but may be very biased if the detector array is linear and home ranges align with the array. Transformation to isotropy reduces bias from designs of intermediate dimension, such as hollow square arrays. Possible alignment of home ranges should be considered when designing detector arrays.
空间显式捕获-再捕获（SECR）模型已成为估计移动和隐匿动物种群密度的一种解决方案。空间模型体现了个体空间分布和空间探测过程的假设。在 SECR 中，探测过程建模为从每个个体的活动中心到距离的探测概率呈径向下降。这似乎需要家域是圆形的。当家域不是圆形时，SECR 的稳健性一直是相互矛盾的说法的主题。Ivan 等人。先前将 SECR 密度估计值与遥测比例的非空间估计值进行了比较。我认为，在他们的研究中，SECR 的明显不稳健性是模拟假象。对椭圆形家域的新模拟表明，当探测器在二维空间中分布时，SECR 密度估计值在很大程度上不受非圆形性的影响，但如果探测器阵列是线性的并且家域与阵列对齐，则可能会产生很大的偏差。各向同性的转换可以减少中等维度设计的偏差，例如空心正方形阵列。在设计探测器阵列时，应考虑家域的可能对齐。
