Using species distribution and metapopulation models to prioritize conservation actions for threatened species

全球保护工作远远不足以阻止当前的灭绝危机。保护机构必须对资源进行优先排序，通常是根据一个物种在没有干预的情况下灭绝的估计概率来分配等级或威胁类别。这种估计取决于对威胁和栖息地可用性的评估。然而，新的研究表明，如果在单一的空间或时间尺度上进行威胁评估，可能会有偏差，而且以前假设的来源栖息地实际上可能是汇。此外，灭绝概率可能被低估，特别是对于孤立种群和下降缓慢的物种。因此，许多物种目前可能被忽视或管理不善。在空间和时间的多个尺度上预测物种分布的模型，以及在数学上描述相互联系的种群的元种群模型，有可能极大地改善威胁和栖息地评估，以及灭绝概率的估计。在受威胁物种群体中使用这种模型还可以确定最紧急的保护行动，以帮助防止灭绝。然而，构建这种模型所需的监测数据的成本必须与根据现有知识采取行动的成本相平衡。

Global conservation efforts fall far short of what is required to halt the current extinction crisis. Conservation agencies must prioritize resources, usually by assigning ranks or threat categories based on the estimated probability that a species will go extinct without intervention. Such estimates depend on assessments of both threats and habitat availability. However, new research has shown that threat assessments can be biased if conducted at a single spatial or temporal scale, and that previously-assumed source habitats may actually be sinks. In addition, extinction probability may be underestimated, especially for species in isolated populations and whose decline is slow. Thus, many species may be currently neglected or mismanaged. Models predicting species distributions at multiple scales in space and time, and metapopulation models that mathematically depict interconnected populations, have the potential to greatly improve threat and habitat assessments, and estimates of extinction probability. Using such models across groups of threatened species may also allow the most urgent conservation actions to be determined, to help prevent extinctions. However, the costs of the monitoring data necessary for constructing such models must be balanced against the costs of acting on current knowledge. My proposed research will quantitatively determine the utility of multiple-scale, dynamic species distribution and metapopulation models for refining extinction risk estimates, and guiding conservation. My students and I will use uniquely detailed datasets on populations and proposed conservation actions from New Zealand and New South Wales, as well as smaller-scale datasets we collect from biodiversity hotspots in Canada, to: 1) Develop models to determine whether current extinction risk estimates are biased for certain groups of species, and whether accounting for this bias alters the threat ranking among species; 2) Determine whether using our models to prioritize actions rather than species can lead to more biodiversity conserved, and allow actions such as habitat protection and restoration to be staggered through time; and 3) Determine when monitoring data needed to derive such models are worth gathering, and which groups of species should be monitored while others are acted upon. This research will help to resolve persistent debates about extinction probabilities for threatened species, and will test the utility of an action-based prioritization approach to conservation, which is a fundamental departure from the species-based ranking approach typically used in threatened species conservation programs. It will also address the crucial question facing conservation agencies as to when to monitor species and when to act on existing information, and improve the biological basis for conservation decision-making, potentially leading to more species saved from extinction.