Most studies on ecological succession, and many hypotheses for mechanisms of succession, focus on terrestrial plant communities.  Many of these hypotheses could also apply to animal communities, but have not been well-tested.  We used aquatic macroinvertebrate container communities (primarily larval dipterans) as a model system for investigating succession because of their small size, simple community structure, and potentially rapid turnover of species.   To determine whether succession occurs in container systems, we arranged 128 identical water-filled containers in groups of 4 in a forested area.  One-fourth of these containers were destructively sampled each week, and the abundances of 6 dipteran taxa were counted for each container.   We also tested whether Egler’s hypothesis of initial floristic composition – or for our study, initial faunistic composition (IFC) – could explain changes in community composition.  After one week for initial oviposition, we covered half of our containers and then compared community composition of covered and uncovered containers for the remainder of the study.  We used MANOVA and profile analysis to compare relative abundances of taxa across treatments (open vs. closed) and week, and all relevant interactions.  

Results/Conclusions:

Relative abundances of taxa were significantly affected by week (p<0.0001) and by week*treatment interaction (p=0.0287), indicating that patterns of succession differed for covered and uncovered treatments.  Diversity was lower in closed containers than in open containers (p=0.0108), with 2 taxa present in only open containers during weeks 3 and 4.   Thus, the data do not support the IFC hypothesis.  To determine whether changing abiotic conditions may affect community composition, we also tested whether temperature, volume, and dissolved sodium concentrations were related to either diversity or to relative abundances of any taxa.  None of these covariates proved significant.  We describe plans for additional experiments on succession in container systems that will test other mechanisms of succession, such as Tilman’s resource-ratio hypothesis or the facilitation/tolerance/inhibition hypotheses proposed by Connell and Slatyer.