Document: JOH-3-92-19

Modeling commnunity assembly: A synthesis of approaches.

SABUCO, J.* ^1,2

Illinois State University, Normal, IL USA ¹
ELdertree Stoneoak & Brookings, Inc., Flossmoor Illinois, 60422, USA ²

Abstract:
Ecologists have long debated whether stochastic or biological forces drive community assembly. There are at least 13 different models that explain the assembly process and the resulting species abundance curves. I have made two observations: 1) All samples of communities that I have tested have asymptotic decay rates when plotted against the number of species in a community, and the range of variation has a distinct pattern; 2) 36 to 78% of all samples in a variety of community types have either a Negative Exponential (NE) or a Modified Zipf-Mandelbrot (MZM) distribution. I built a model to attempt to resolve these observations and to incorporate the stochastic and biological forces that have been described and supported to date. This model, now incorporated into a computer simulator called Eco-Assembler, includes a complex set of algorithms (using fuzzy logic and multiple layers of interaction) that incorporate community resources, habitat space, competition, dispersal ability, facilitation, inhibition, matrix dispersal rate, reproduction, disturbance, sensitivity to disturbance, mortality, and degradation of the matrix. I used actual species abundances from communities I had sampled to initiate the model, and then compared the model results to actual samples from those same communities. I found that the best fit of the abundance distributions agreed in type frequency (average r = 0.84) and consistency of residual values (average r = 0.81). In addition, the communities demonstrated the asymptotic decay curves. I then sought to determine whether the model could be simplified and still obtain the same results. I hypothesized that the interaction of the high number of complex processes in the model would reduce to simple overall processes. I was able to simplify the model and show not only the asymptotic decay curves and the prevalence of the NE and MZM distributions, but an interesting alternation from sample to sample between these two distributions as well. It appears that the complexity of community assembly can be reduced to a few simple rules governed by simple additive and subtractive neighborhood interactions based on stochastic juxtaposition of species during the invasion process, with community resources and differentiation of habitat space as limiting factors.

Keywords: community assembly, ecological modeling, species abundance distribution

This abstract is being presented at: 10:15 AM in session:
Oral Session #56: Metapopulation Analysis.