作者: Aimee P. McIntyre , Jay E. Jones , Eric M. Lund , Frithiof T. Waterstrat , Jack N. Giovanini
DOI: 10.1016/J.FORECO.2012.08.039
关键词:
摘要: Abstract Conservation and management programs use populations of sentinel taxa, such as stream-associated amphibians, indicator species due to their perceived sensitivity environmental change. Estimating population size with traditional tools mark–recapture estimators may be impractical for forest-dwelling many which are cryptic have low detection probabilities. In addition, sampling techniques can alter habitat conditions, particularly repeat sampling, compromise inferences about impacts. We used simulated empirical data N-mixture models estimate probabilities abundances two amphibian genera, giant Dicamptodon torrent Rhyacotriton salamanders, sampled a less-invasive approach than other methods. surveyed forested headwater streams located in western Washington, USA, salamanders 7 July–27 August 2008. assessed model changes animal abundance (5 15) probability (0.05–0.5), well study design alternatives including number sample plots (50–150) visits (2–4). also evaluated the effects stream temperature order on using collected from USA. Precision estimates improved occasions increased. Variability estimated sizes decreased higher probability, although had little effect precision estimates. Detection ranged 0.07 0.65 0.06–0.67 salamanders. Giant salamander was positively associated regardless order, second- third-order first-order streams. varied temperature, interaction those covariates, increasing but showing flat or decreasing trend models, paired careful consideration alternatives, produce robust obviate reliance indices relative rare sensitive taxa. Use less invasive technique amphibians resulted sufficient captures suitable performance. Our simulation results inform direct efficient allocation field effort.