Water-filled containers are frequently colonized by aquatic insect larvae.  The relative abundances of species in containers also change over time, with a predictable pattern of species turnover. These changes in larval species abundances could be products of ecological succession, or could be the result of seasonal changes in populations of ovipositing adults.  To test the hypothesis that container age, independent of seasonal changes in adult populations, contributes to community change in containers, we established two groups of water and detritus-filled 1-gallon buckets in an oak hickory forest.  One group was established in late May and subsampled weekly for larvae for 10 weeks.  The second group was established in early July and subsampled weekly for 3 weeks. We used principal components analysis (PCA) to describe community composition of each set of containers in 5 PCs.  We then analyzed each PC using repeated-measures MANCOVA to determine whether community composition changed over time (succession was occurring) in both groups of containers.   We also ran separate MANCOVAS to determine whether (A) succession patterns differed between container groups of the same chronological age at different season (May vs. July), (C) succession patterns differed between container groups of different chronological ages (1-3 weeks vs 7-9 weeks of age) during the same season (July). 

Results/Conclusions

Containers of the same chronological age during different season showed significant change in community composition over time (PC2 p=0.0632, PC3 p=0.0190), which supports that succession was occurring in these containers; however, the succession patterns of these two container groups were not significantly different by season (p>0.05 for all 5 PCs).  Containers of different chronological ages within the same season also showed significant succession patterns (PC1 p=0.0018, PC5 p= 0.0673), and the succession trajectories also differed between these groups by chronological age (PC1 p=0.0006, PC5 p=0.0548).  This means that the containers established in July did not acquire the insect communities already present in the May containers, but instead acquired communities similar to those of the May group when it was first established.   Our data support the hypothesis that community change observed in newly-established containers results at least in part from succession (i.e., age), and independent of seasonal differences in adult populations.