How ecological network structures are influenced by species coexistence, community stability and perturbations is a topic of debate. It seems that one overlooked correlate of nested structures is species abundances. See Letter p.449
Understanding the mechanisms that shape biodiversity is one of the main goals of ecology. Network approaches, which integrate species and the interactions among them into a single framework, have proved enlightening, revealing distinct 'architectural' patterns that are strongly associated with particular ecological interactions. For mutualistic networks — those in which the interactions benefit both partners, such as between a plant and its pollinator, or a fish and a cleaner fish — the pervasive pattern seems to be a nested one, whereby specialist species (which have few partners) interact with a subset of the many partners of more generalist species. The origin and implications of nestedness remain strongly debated. On page 449 of this issue, Suweis et al.1 bring an innovative and intriguing contribution to this topic by demonstrating strong relationships among species abundances, nested architecture and community stability.
Nestedness is a pattern characterized by several features (Fig. 1), including a skewed distribution of the number of interacting partners per species, with many specialist species and few extremely generalist species. Nestedness also implies asymmetric specialization, such that specialist species tend to interact with generalist ones. Finally, the generalist species in the nested network form a single, highly connected core, making the networks very cohesive.
Three main hypotheses have been proposed to explain the biology behind this seemingly highly organized structure. One is that nestedness is 'neutral', meaning that all interactions between individuals are equally likely. Species abundances in many communities are well described by a log-normal distribution, with many rare species and a few common ones. Under this hypothesis, differences in species abundance result in differences in interactions at the species level: abundant species are expected to interact more frequently and with more species than rare species, and rare species tend to interact with abundant species rather than with other rare species. However, the empirical correlation between species abundances and species generalism is not easy to interpret2. Do species become generalists because they are more abundant, or are they more abundant because they are generalists and therefore can access more resources?
The second hypothesis suggests that nestedness affects ecological dynamics, particularly species coexistence and community stability. A simple argument supporting this hypothesis is that it is much safer for specialist species to interact with generalist species than with other specialists, because generalist species are expected to have less-fluctuating population dynamics and so to be more reliable partners. Such constraints on community persistence or stability could therefore be a driving force shaping interaction networks. However, no consensus on this topic has been reached among several investigations3,4,5,6 in recent years of the links between network nestedness and community dynamics in mutualistic species.
According to the third hypothesis, nested architecture may be shaped by the (co-)evolutionary dynamics of species interacting within a community. There are many examples of interspecies interactions affecting the fitness of individuals, and of the evolution of species traits controlling the identity of potential interaction partners. Closely related species in mutualistic interaction networks tend to have similar interacting partners, which emphasizes the idea that evolutionary history has an impact on the structure of mutualistic networks7. But, so far, no precise evolutionary process has been directly related to a nested structure.
Suweis et al. have drawn these three hypotheses together by demonstrating a two-step relationship between species abundances in a community and the nestedness of the interaction network that depicts that community. Using analytical and simulation approaches, the authors first show that, under stationary conditions that have a constant number and strength of mutualistic interactions, 'interaction swaps' (an exchange of interactions between two species couples) that lead to an increase in the abundance of the species also increase the total abundance of the community. Second, the researchers demonstrate that total community abundance is positively related to the nestedness of the network. This connection opens up fascinating perspectives.
To demonstrate the implications of their findings, the authors show that, under the condition that exchanges result in increased species abundance, iterative swapping ultimately converts random networks, with randomly distributed interactions among species, into nested networks. The interpretation of this is that any process that maximizes species abundance through changes in interspecies interactions will lead to a nested network. The question thus becomes, what biological process could select for higher population size? Selection at the population level involves group-selection processes such as hard selection8,9. More work is needed to unravel the microevolutionary processes that affect network architecture, but this line of research seems promising.
Suweis and colleagues further demonstrate that the population size of the rarest species in the community is positively related to community resilience — the speed at which community dynamics return to equilibrium after a small perturbation. These results fuel the current debate about the relationship between network architecture and community stability3,4,5,6,10,11 by introducing the distribution of species abundance as a key element. Again, however, the processes through which the abundance of the rarest species relates to community resilience remain to be identified. They may involve the rarest species directly, or may emerge from other mechanisms affecting both the rarest species and community resilience.
Last but not least, the relationship found by Suweis et al. between network nestedness and total community abundance goes both ways. Abundance is correlated with biomass, which is one of the main variables used in studies of biodiversity and ecosystem functioning, so the two-way relationship provides a bridge between the authors' results and the rich literature on these topics. We already know that the structure of food webs, for example, can affect the relationship between biodiversity and ecosystem function12. But little is known about the impact of mutualistic networks on the functioning of ecological communities. Like all exciting pieces of research, Suweis and colleagues' work raises more questions than it answers.
Suweis, S., Simini, F., Banavar, J. R. & Maritan, A. Nature 500, 449–452 (2013).
Santamaría, L. & Rodríguez-Gironés, M. A. PLoS Biol. 5, e31 (2007).
Bastolla, U. et al. Nature 458, 1018–1020 (2009).
James, A., Pitchford, J. W. & Plank, M. J. Nature 487, 227–230 (2012).
Thébault, E. & Fontaine, C. Science 329, 853–856 (2010).
Allesina, S. & Tang, S. Nature 483, 205–208 (2012).
Rezende, E. L., Lavabre, J. E., Guimarães, P. R., Jordano, P. & Bascompte, J. Nature 448, 925–928 (2007).
Saccheri, I. & Hanski, I. Trends Ecol. Evol. 21, 341–347 (2006).
Goodnight, C. J. Phil. Trans. R. Soc. B 366, 1401–1409 (2011).
Saavedra, S. & Stouffer, D. B. Nature 500, E1–E2 (2013).
James, A., Pitchford, J. W. & Plank, M. J. Nature 500, E2–E3 (2013).
Thébault, E. & Loreau, M. Proc. Natl Acad. Sci. USA 100, 14949–14954 (2003).
About this article
Perspectives in Plant Ecology, Evolution and Systematics (2017)
Ecology Letters (2016)
Ecology and Evolution (2015)
Nature Communications (2015)
Bárcena Volcano, 1952: a 60-year report on the repopulation of San Benedicto Island, Mexico, with a review of the ecological impacts of disastrous events
Pacific Conservation Biology (2015)