利用网络度量指导入侵物种的大规模管理
Guiding large-scale management of invasive species using network metrics
摘要:
复杂的社会环境相互依存关系驱动生物入侵,造成大规模空间破坏。对于广泛的入侵,基于优化方法的管理活动的目标可能会失败,由于计算或数据的限制。在这里,我们评估了一种替代方法,包括复杂性,通过将入侵表示为网络,并使用网络结构通知管理位置。我们比较了最佳与网络引导的入侵物种管理在一个小规模的,考虑选址的船只净化站,目标是160万船民运动在美国明尼苏达州的9,182个湖泊。研究58个县的性能,我们发现,当入侵状态和船民运动的全部信息是已知的,性能最好的网络引导指标实现了中位数和下四分位数的性能100%的最佳。我们还发现,使用不同的网络指标或较少的信息(中位数高于80%,下四分位数高于大多数指标的最佳值的60%),性能仍然相对较高,但变化更大,特别是在下四分位数。此外,在不同湖泊数量的县,性能一般是稳定的,这表明大规模入侵管理的可行性。
英文摘要:
Complex socio-environmental interdependencies drive biological invasions, causing damages across large spatial scales. For widespread invasions, targeting of management activities based on optimization approaches may fail due to computational or data constraints. Here we evaluate an alternative approach that embraces complexity by representing the invasion as a network and using network structure to inform management locations. We compare optimal versus network-guided invasive species management at a landscape-scale, considering siting of boat decontamination stations targeting 1.6 million boater movements among 9,182 lakes in Minnesota, USA. Studying performance for 58 counties, we find that when full information is known on invasion status and boater movements, the best-performing network-guided metric achieves a median and lower quartile performance of 100% of optimal. We also find that performance remains relatively high using different network metrics or with less information (median above 80% and lower quartile above 60% of optimal for most metrics), but is more variable, particularly at the lower quartile. Additionally, performance is generally stable across counties with varying lake counts, suggesting viability for large-scale invasion management.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。