Scaling up keystone effects from simple to complex ecological networks
Scaling up keystone effects from simple to complex ecological networks
复制标题
DOI：
10.1111/j.1461-0248.2005.00838.x
复制
发表时间：
2005-12-01
期刊：
影响因子：
8.8
通讯作者：
Martinez, ND
中科院分区：
文献类型：
--
作者：
Brose, U;Berlow, EL;Martinez, ND
预测物种损失的后果需要将我们对少数相互作用物种的简单动态系统的传统理解扩展到自然生态系统中发现的更复杂的生态网络。尤其重要的是扩大我们对“关键”物种的丧失如何以及在什么条件下导致许多其他物种大量减少的有限理解。在这里，我们将探讨这些关键效应如何在从简单到更复杂的系统逐步扩大的模拟中变化。对4到7个相互作用物种的简单模拟表明，距离最多4个环节的物种可以强烈地改变基石效应，并使基石损失的后果在更现实的复杂群落中潜在地不确定。我们没有发现不确定性，而是发现多达32个物种的更复杂的网络通常会缓冲远距离影响，这样，在keystone子系统的两个环节中，通过令人惊讶的局部“自上而下”、“自下而上”和“水平”约束，可以很好地预测keystone效应的变化。研究结果表明：(1)keystone对竞争优势的强烈抑制对下级竞争者的影响较弱；(2)目标物种的群落环境决定了是否实现强keystone效应；(3)复杂群落的简单、可测量和局部属性可以解释许多经验观察到的楔石效应变化；(4)增加网络复杂性本身并不会使强keystone效应的预测变得更加复杂。
Predicting the consequences of species loss requires extending our traditional understanding of simpler dynamic systems of few interacting species to the more complex ecological networks found in natural ecosystems. Especially important is the scaling up of our limited understanding of how and under what conditions loss of 'keystone' species causes large declines of many other species. Here we explore how these keystone effects vary among simulations progressively scaled up from simple to more complex systems. Simpler simulations of four to seven interacting species suggest that species up to four links away can strongly alter keystone effects and make the consequences of keystone loss potentially indeterminate in more realistically complex communities. Instead of indeterminacy, we find that more complex networks of up to 32 species generally buffer distant influences such that variation in keystone effects is well predicted by surprisingly local 'top-down', 'bottom-up', and 'horizontal' constraints acting within two links of the keystone subsystem. These results demonstrate that: (1) strong suppression of the competitive dominant by the keystone may only weakly affect subordinate competitors; (2) the community context of the target species determines whether strong keystone effects are realized; (3) simple, measurable, and local attributes of complex communities may explain much of the empirically observed variation in keystone effects; and (4) increasing network complexity per se does not inherently make the prediction of strong keystone effects more complicated.