Selection at the Level of the Community: The Importance of Spatial Structure

Johnson, Craig R., and Maarten C. Boerlijst. “Selection at the Level of the Community: The Importance of Spatial Structure.” Trends in Ecology & Evolution 17, no. 2 (February 1, 2002): 83–90. doi:10.1016/S0169-5347(01)02385-0.

Johnson and Boerlijst review the role of spatial structure in community selection, highlighting several important scenarios in which spatial structure arises and areas where the field is lacking. The particular type of spatial structure they are interested in is spatial self-structuring, in which nonrandom spatial patterns at scales larger than that which individuals are interacting emerge solely because of local interactions and dispersal, and are not explicitly incorporated in a model. They term this “order for free”, and argue it is one mechanism by which community selection controls selection among individuals from overriding the “integrity”, or persistence of a property or spatial pattern, of the system.

What is especially interesting about these spatial patterns is their ability to create a “trade-off” within the overall community that may be detrimental at the individual level. This can happen when the selection of a trait at the individual level reaches a threshold at which it creates instability within the overall landscape, which causes populations with this trait to go extinct, thereby selecting for traits below the threshold. A particularly well-studied example of this trade-off is virulence in parasites, which is selected for intermediate levels. Further, they note that vector-borne pathogens tend to be more virulent than non-vector borne pathogens, suggesting that this is perhaps because the selection at the community level that limits virulence is not as strong, but not explaining the exact mechanism.

With regards to experimental evidence of this spatial patterning, they mention the difficulty in identifying community vs. individual level selection in experiments, as most experiments to dates demonstrate a phenomenon, not a mechanism. To conclude that community selection is occurring, the investigator must show opposing directions of selections at the individual and community level, something that has only been shown in extremely simplified systems such as serial passage experiments of pathogens.

This paper illustrates a multi-scale model because these phenomenon occur at a community scale, but are a direct result of only processes at the local scale. Additionally, it is possible for the community level selection to feed back to the local scale. Perhaps the most interesting aspect of the paper is the existence of these emergent spatial structures. Often in theoretical modeling, the results can be somewhat intuitive given a thorough understanding of the system, but this reinforces the idea the combination of local processes, the results of which are easily intuited, can lead to unexpected results at a larger scale.