Conservation of species interaction networks

Recent work has shown that antagonist (e.g. predator-prey food web) and mutualist (e.g. pollinator-plant) network structure can be altered by global environmental change drivers, and that these alterations may have important ecosystem-level consequences. This has prompted calls for the conservation of network structure, but precisely which attributes of webs should be conserved remains unclear. Further, the extent to which network metrics characterise the spatiotemporally-variable dynamic structure of interacting communities is unknown. Here, we summarise the attributes of web structure that are predicted to confer stability or increased function to a system, as these may be of greatest interest to conservation biologists. However, empirical evaluation of these effects is lacking in most cases, and we discuss whether stability is even desirable in all contexts. The incorporation of web attributes into conservation monitoring requires that changes in these attributes can be recorded (sampled) with relative ease. We contrast the sensitivity of metrics to sampling effort, and highlight those (such as nestedness and connectance) that could easily be incorporated into conservation monitoring. Despite our growing understanding of the characteristics of food webs that confer stability and function, numerous practical challenges need to be overcome before the goal of conserving species interaction networks can be achieved. (C) 2009 Elsevier Ltd. All rights reserved.