Alien fish eradication from high mountain lakes by multiple removal methods: Estimating residual abundance and eradication probability in open populations

Invasive alien species are among the most important threats to biodiversity. Plans for their eradication have been implemented worldwide but estimating residual population size and eradication probability to assess removal success is complicated by the imperfect detection of residual individuals. Most methods to assess residual abundance and eradication probability rely on the often unrealistic assumption that a population is closed to mortality and recruitment processes during the implementation of removal actions. We extended existing removal models and developed a novel analytical approach to estimate residual population size and derive eradication probability in open populations while accounting for multiple removal methods. We apply this approach to 20 eradication projects in Europe and the United States that used mechanical methods to return high mountain lakes to their original fishless condition. The new removal model incorporates (a) a mechanistic description of the ecological process underlying survival and recruitment probabilities during the eradication period and (b) the use of multiple, concomitant removal methods (i.e. electrofishing, gillnetting, fyke‐netting) at multiple sites simultaneously. We used a subset of ‘control’ lakes where eradication success was confirmed by more than 5 years of post‐removal surveys to validate the model. For these lakes, eradication success, evaluated by whether the 95% Bayesian credible interval for estimated residual population size encompassed values of