Detecting phylogenetic signal in mutualistic interaction networks using a Markov process model. Issue 10 (17th April 2014)

• Record Type: Journal Article
• Authors: Minoarivelo, H. O.; Hui, C.; Terblanche, J. S.; Kosakovsky Pond, S. L.; Scheffler, K.
• Is Part Of: Oikos. Volume 123:Issue 10(2014)
• Page Start: 1250
• Page End: 1260
• Publication Date: 2014-04-17
• Subjects: Ecology -- Periodicals; 570
• DOI: 10.1111/oik.00857
• Languages: English
• ISSNs: 0030-1299

Abstract

<abstract abstract-type="main" xml:lang="en"> <title> <x xml:space="preserve">Abstract</x> </title> <p>Ecological interaction networks, such as those describing the mutualistic interactions between plants and their pollinators or between plants and their frugivores, exhibit non‐random structural properties that cannot be explained by simple models of network formation. One factor affecting the formation and eventual structure of such a network is its evolutionary history. We argue that this, in many cases, is closely linked to the evolutionary histories of the species involved in the interactions. Indeed, empirical studies of interaction networks along with the phylogenies of the interacting species have demonstrated significant associations between phylogeny and network structure. To date, however, no generative model explaining the way in which the evolution of individual species affects the evolution of interaction networks has been proposed. We present a model describing the evolution of pairwise interactions as a branching Markov process, drawing on phylogenetic models of molecular evolution. Using knowledge of the phylogenies of the interacting species, our model yielded a significantly better fit to 21% of a set of plant–pollinator and plant–frugivore mutualistic networks. This highlights the importance, in a substantial minority of cases, of inheritance of interaction patterns without excluding the potential role of ecological novelties in forming the current network architecture. We suggest that our model can be used as a null model for controlling evolutionary signals when evaluating the role of other factors in shaping the emergence of ecological networks.</p> </abstract>