Ecological modules and roles of species in heathland plant–insect flower visitor networks
Article first published online: 4 NOV 2008
© 2008 The Authors. Journal compilation © 2008 British Ecological Society
Journal of Animal Ecology
Volume 78, Issue 2, pages 346–353, March 2009
How to Cite
Dupont, Y. L. and Olesen, J. M. (2009), Ecological modules and roles of species in heathland plant–insect flower visitor networks. Journal of Animal Ecology, 78: 346–353. doi: 10.1111/j.1365-2656.2008.01501.x
- Issue published online: 9 FEB 2009
- Article first published online: 4 NOV 2008
- Received 11 July 2008; accepted 9 October 2008; Handling Associate Editor: Guy Woodward
- coevolutionary units;
- community structure;
- mutualistic webs;
- network topology
- 1Co-existing plants and flower-visiting animals often form complex interaction networks. A long-standing question in ecology and evolutionary biology is how to detect nonrandom subsets (compartments, blocks, modules) of strongly interacting species within such networks. Here we use a network analytical approach to (i) detect modularity in pollination networks, (ii) investigate species composition of modules, and (iii) assess the stability of modules across sites.
- 2Interactions between entomophilous plants and their flower-visitors were recorded throughout the flowering season at three heathland sites in Denmark, separated by ≥ 10 km. Among sites, plant communities were similar, but composition of flower-visiting insect faunas differed. Visitation frequencies of visitor species were recorded as a measure of insect abundance.
- 3Qualitative (presence–absence) interaction networks were tested for modularity. Modules were identified, and species classified into topological roles (peripherals, connectors, or hubs) using ‘functional cartography by simulated annealing’, a method recently developed by Guimerà & Amaral (2005a).
- 4All networks were significantly modular. Each module consisted of 1–6 plant species and 18–54 insect species. Interactions aggregated around one or two hub plant species, which were largely identical at the three study sites.
- 5Insect species were categorized in taxonomic groups, mostly at the level of orders. When weighted by visitation frequency, each module was dominated by one or few insect groups. This pattern was consistent across sites.
- 6Our study adds support to the conclusion that certain plant species and flower-visitor groups are nonrandomly and repeatedly associated. Within a network, these strongly interacting subgroups of species may exert reciprocal selection pressures on each other. Thus, modules may be candidates for the long-sought key units of co-evolution.