Keeping the community in ecology
The concept of the community is a beautiful thing. However, there has been recent and historical debate surrounding its use as a concept in ecology. The debate is legitimate and useful. Effective dialogue about terms and concepts promotes reassessment of previous work in the light of new frameworks or criteria that emerge as we consider the various methods available for studying interacting groups of individuals from different species. At one end of the scale continuum, we can focus our efforts only on regional processes (Ricklefs 2008), and at the other end of the continuum, we can focus on attributes of individual species such as traits. Nonetheless, when scaling up or down from our preferred scale of study, we often invoke filters or sets of processes that describe the forces which regulate the identities of species that occur together in a given place (Brooker et al. 2009). In this issues's Editor's Choice paper "Ecology's cruel dilemma, phylogenetic trait evolution and the assembly of Serengeti plant communities" authors T. Michael Anderson, Joey Shaw and Han Olff provide a perfect example of reconciliation between these different approaches, and elegantly illustrate that small-scale analyses can inform understanding of larger patterns.
Phylogenetics and gradients
Current phylogenetic studies often have the capacity to elucidate not only evolutionary processes that shape sets of traits but to also reveal interactions that are important in shaping communities. Phylogenetic patterns alone, however, can be misleading depending upon whether trait evolution has been convergent, random or divergent (Cavender-Bares et al. 2004; Webb et al. 2002). In a recent review, the implications and limitations of phylogenetic tools as they relate to communities are described, and the importance of filtering or clustering in response to local conditions is clearly illustrated (Vamosi et al. 2009). Anderson et al. (2011) take this to the next level by identifying a large-scale environmental gradient in Serengeti grasslands and applying phylogenetic analyses to the specific leaf area (SLA) and maximum height for 52 grass species. A large number of sites were sampled (133), at a local scale (1-m2 plots), with the 52 focal species comprising a significant proportion of the community (> 86%). The two traits examined also allowed inference of life-history implications and allocation strategies associated with a gradient analysis. A set of a priori predictive models were developed for this data set and presented visually to illustrate how trait distances and mean phylogenetic distance (MPD) interact in a multivariate context including both direct and indirect effects to detect patterns on gradients and contrast various drivers of community assembly. Trait evolution was carefully assessed, the countervailing effects of phylogenetic patterns on community assembly were examined using structural equation modeling of respective factors, and mechanistic and visual insights into the complex determinants of community composition were presented.
Univariate analyses of environmental gradients (rainfall, proportion of rain falling in the dry season, elevation) generally failed to describe MPD. Selection of only non-random MPD plots did, however, conform, to a limited extent, to dominant assembly theory in that plots with relatively more under-dispersed MPDs (i.e. trait distances between species were low) were associated with more stressful conditions, whilst over-dispersed plots were associated with moist, less-seasonal, lower elevation sites . Hence, more similar species were found at higher elevation sites or lower rainfall due to filtering by the environment whilst filtering by biotic interactions—competition—at the other end of the gradient generates a community composed of a set of less similar species. This is a fascinating finding as it uses an environmental gradient to infer interactions parallel to studies that directly measure plant–plant interactions and responses to stress. Structural equation modelling (SEM) was also used to assess which set of potential interactions, combined, best fit the a priori models proposed. The variation in MPD was best modeled by trait distances between species within plots. This means that neither the large-scale environmental gradient (i.e. regional-scale processes) nor species richness at the plot level (i.e. local-scale processes) directly predict MPD but that the traits of the species in each context and assemblage of species filter the species via the environment to determine the assembly. This demonstrates that very small-scale interactions can drive community and regional patterns of species in space or time. Admittedly, neither the story nor the data were so clear in this study to warrant a conclusion so definitive, but the merits of combining the gradients we use in direct-interaction plant ecology studies with phylogenetics are evident. A broader implication for community ecology is that natural gradients are complex and that sets of various factors both biotic and abiotic are always interacting to shape communities. This does not mean that studies of communities are futile but that novel tools and appropriate application of analyses such as SEMs will provide insights into the filters that are functioning at multiple scales.
Reconciliation and novelty of this paper
There are several interesting implications of this paper both for community ecology and for publishing in general. Multiple-scale studies, use of SEMS, application of phylogenetics to gradient analyses, and embracing complexity are all very productive approaches for achieving synthesis and identifying future directions for research in plant ecology. Reconciliation of the influences of forces acting at different scales is a possibility in plant ecology with appropriate application of these tools to examine important hypotheses. I am very pleased that the top plant ecology journal not only published this paper but also selected it as one the best in this issue. Many of the statistics are not significant, counteracting and complex forces are described, at many points in the ‘story’ or narrative the reader is not provided with a simplified or ‘punchy’ statement, and more importantly, the results are not over-interpreted by the authors. Hence, this paper is an excellent example of avoidance of the ‘file-drawer’ problem (Csada et al. 1996) and over-interpretation (Lortie & Dyer 1999). This paper should serve as an inspiration to all of us not just to publish our clean, significant work that supports dominant theory but also to publish our more challenging studies.
Associate Editor, Journal of Ecology
- Anderson, T.M, Shaw, J., & Olff, H. (2011) Ecology’s cruel dilemma, phylogenetic trait evolution and the assembly of Serengeti plant communities. Journal of Ecology, 99, 797-806.
- Brooker, R.W., Callaway, R.M., Cavieres, L.A., Kikvidze, Z., Lortie, C.J., Michalet, R., Pugnaire, F., Valiente-Banuet, A., & Whitham, T.G. (2009) Don’t diss integration: A comment on Ricklefs’s disintegrating communities. The American Naturalist, 174, 919-927.
- Cavender-Bares, J., Ackerly, D.D., Baum, D.A., & Bazzaa, F.A. (2004) Phylogenetic overdispersion in Floridian oak communities. The American Naturalist, 163, 823-843.
- Csada, R.D., James, P.C., & Espie, R.H.M. (1996) The "file drawer problem" of non-significant results" does it apply to biological research? Oikos, 76, 591-593.
- Lortie, C.J. & Dyer, A.R. (1999) Over-interpretation: avoiding the stigma of non-significant results. Oikos, 87, 183-184.
- Ricklefs, R.E. (2008) Disintegration of the ecological community. The American Naturalist, 172, 741-750.
- Vamosi, S.M., Heard, S.B., Vamosi, J.C., & Webb, O.C. (2009) Emerging patterns in the comparative analysis of phylogenetic community structure. Molecular Ecology, 2009, 572-592.
- Webb, C.O., Ackerly, D.D., McPeek, M.A., & Donoghue, M.J. (2002) Phylogenies and community ecology. Annual Review of Ecology, Evolution, and Systematics, 33, 475-505.
Log in for Access
If you are a BES member click here for instructions on how to access all BES journal content
to BES for access) (Note: you will be redirected