The nested assembly of plant-animal mutualistic networks

Abstract

Most studies of plant-animal mutualisms involve a small number of species. There is almost no information on the structural organization of species-rich mutualistic networks despite its potential importance for the maintenance of diversity. Here we analyze 52 mutualistic networks and show that they are highly nested; that is, the more specialist species interact only with proper subsets of those species interacting with the more generalists. This assembly pattern generates highly asymmetrical interactions and organizes the community cohesively around a central core of interactions. Thus, mutualistic networks are neither randomly assembled nor organized in compartments arising from tight, parallel specialization. Furthermore, nestedness increases with the complexity (number of interactions) of the network: for a given number of species, communities with more interactions are significantly more nested. Our results indicate a nonrandom pattern of community organization that may be relevant for our understanding of the organization and persistence of biodiversity.

Fig. 1.

Plant–animal mutualistic interaction matrices. Numbers label plant and animal species, which are ranked in decreasing number of interactions per species. A filled square indicates an observed interaction between plant i and animal j. a–c correspond to perfectly nested, random, and real mutualistic matrices [plant–pollinator network of Zackenberg (J.M.O. and H. Elberling, unpublished work)], respectively. Values of nestedness are N = 1(a), N = 0.55 (b), and N = 0.742 (P < 0.01) (c). The box outlined in a represents the central core of the network, and the line in c represents the isocline of perfect nestedness. On a perfectly nested scenario, all interactions would lie before the isocline (on the left side).

Fig. 2.

Nestedness values for seed dispersal (SD, circles), pollination (P, squares), and food webs (FW, diamonds). (a) Mean and SE of nestedness for the three types of networks. Seed-dispersal and pollination matrices have similar nestedness, significantly higher than consumer–resource webs. (b) Nestedness vs. species richness for all data sets. Each point corresponds to a specific community and is solid if the level of nestedness is significant at the P < 0.05 level and empty otherwise. The arrow indicates the plant–pollinator network shown in Fig. 1c.

Fig. 3.

Number of interactions (L) vs. number of species (S) for the mutualistic networks (pollination and seed dispersal). The continuous line is the best fitto data. The broken line represents the x = y axis. As noted, L increases slightly faster than S (slope = 1.139). All communities can be classified in two groups: networks with fewer interactions than expected (negative residuals) and networks with more interactions than expected (positive residuals). (Inset) The average and SE of relative nestedness (N*) for the communities with positive and negative residuals. Networks with positive residuals, that is, with more interactions than expected for a specific number of species, are significantly more nested than networks with fewer interactions than expected.