Weldy Matthew J, Wilson Todd M, Lesmeister Damon B, Epps Clinton W
Department of Fisheries and Wildlife, Oregon State University, Corvallis, OR, United States of America.
Pacific Northwest Research Station, U.S.D.A. Forest Service, Corvallis, OR, United States of America.
PeerJ. 2019 Oct 3;7:e7783. doi: 10.7717/peerj.7783. eCollection 2019.
Live trapping is a common tool used to assess demography of small mammals. However, live-trapping is often expensive and stressful to captured individuals. Thus, assessing the relative tradeoffs among study goals, project expenses, and animal well-being is necessary. Here, we evaluated how apparent bias and precision of estimates for apparent annual survival, abundance, capture probability, and recapture probability of Humboldt's flying squirrels () varied with the number of secondary trapping occasions. We used data from forested sites trapped on 12 consecutive occasions annually in the HJ Andrews Experimental Forest (9 sites, 6 years) and the Siuslaw National Forest (seven sites, three years) in Oregon. We used Huggins robust design models to estimate parameters of interest for the first 4, 8, and 12 trapping occasions. We also estimated the effect of attaching Tomahawk traps to tree boles on site- and year-specific flying squirrel capture frequencies. Our estimates with 12 occasions were similar to those from previous studies. Abundances and capture probabilities were variable among years on both sites; however, variation was much lower on the Siuslaw sites. Reducing the length of primary trapping occasions from 12 to 8 nights had very little impact on parameter estimates, but further reducing the length of primary trapping occasions to four nights caused substantial apparent bias in parameter estimates and decreased precision. We found that attaching Tomahawk traps to tree boles increased the site- and year-specific capture frequency of flying squirrels. Our results suggest that live-trapping studies targeting Humboldt's flying squirrels in the Pacific Northwest of the United States could reduce per-site costs and stress to captured individuals without biasing estimates by reducing the length of primary trapping occasions to 8 nights. We encourage similar analyses for other commonly-trapped species in these and other ecosystems.
活捉是评估小型哺乳动物种群统计学特征的常用手段。然而,活捉往往成本高昂,且会给被捕获个体带来压力。因此,有必要评估研究目标、项目成本和动物福祉之间的相对权衡。在此,我们评估了洪堡飞鼠()的表观年生存率、种群数量、捕获概率和重捕概率估计值的表观偏差和精度如何随二次诱捕次数的变化而变化。我们使用了俄勒冈州HJ安德鲁斯实验森林(9个地点,6年)和锡厄斯劳国家森林(7个地点,3年)每年连续12次诱捕的森林地点的数据。我们使用哈金斯稳健设计模型来估计前4次、8次和12次诱捕时感兴趣的参数。我们还估计了将战斧式陷阱附着在树干上对特定地点和年份的飞鼠捕获频率的影响。我们12次诱捕的估计结果与之前的研究相似。两个地点的种群数量和捕获概率在不同年份都有所变化;然而,锡厄斯劳地点的变化要小得多。将初次诱捕时间从12晚减少到8晚对参数估计的影响很小,但进一步将初次诱捕时间减少到4晚则会导致参数估计出现明显偏差并降低精度。我们发现,将战斧式陷阱附着在树干上会增加特定地点和年份的飞鼠捕获频率。我们的结果表明,针对美国太平洋西北部洪堡飞鼠的活捉研究可以通过将初次诱捕时间减少到8晚来降低每个地点的成本和对被捕获个体的压力,同时不会使估计产生偏差。我们鼓励对这些及其他生态系统中其他常见捕获物种进行类似分析。