Search

We consider optimal conservation strategies for an endangered population. We assume that juvenile survival is affected by unpredictable environmental fluctuation and can be improved by costly conservation effort. The initial population size is not accurately known at the time that the conservation effort level is chosen, but the uncertainty of its estimate can be reduced by a costly monitoring effort. In a previous paper, we analysed the optimal management strategy that minimizes a weighted sum of extinction probability and economic costs when only a single year is considered. Here we examine the case in which the conservation period lasts for several years by dynamic programming with incompletely observed process states. We study the optimal levels of the conservation and the monitoring efforts, and their dependence on the length of the conservation period and other parameters. The main conclusions are: (1) The optimal conservation effort in the first year depends on the accuracy of the information on the population size in the first year, but is almost independent of the accuracy of the information in later years. (2) When the risk of population extinction is small, the optimal conservation effort increases with the uncertainty of the population size. In contrast when the population is endangered, the optimal conservation effort decreases with the uncertainty of the population size. (3) The optimal conservation and monitoring efforts both increase with the length of the conservation period, provided that the population is relatively safe. However, if the population is endangered, both types of effort become smaller when the conservation period increases. [Copyright &y& Elsevier]

Ecological niche models (ENMs) are widely used in spatial prioritization for biodiversity conservation (e.g. selecting conservation areas). However, it is unclear whether ENMs are always beneficial for such purposes. We quantified the benefit of using ENMs in conservation prioritization, comparing the numbers of species covered by conservation areas selected on the basis of probabilities estimated by ENMs (ENM approach) and those selected on the basis of raw observation data (raw-data approach), while controlling survey range, survey bias, and target size of conservation area. We evaluated three ENM algorithms (GLM, GAM, and random forests). We used virtual community data generated by simulation for the evaluation. ENM approach was effective when survey bias is strong, survey range is narrow, and target size of conservation area is moderate. The percentage of cases in which the ENM approach outperformed the raw-data approach ranged from 0.0 to 33% (GLM), 31% (GAM), and 75% (random forests) depending on conditions. The number of rare species (< 20 presence records) included in the conservation area based on the ENM approach was less than, or the same as, that of the raw-data approach. The unexpectedly limited cases in which the ENM approach was effective in the present research may depend on the conservation target we used (to cover as many species as possible in conservation area). Our results highlight urgent need for evaluating ENM's effectiveness under other conservation targets for wise use of ENM in conservation prioritization. [ABSTRACT FROM AUTHOR]