Dispersal allows species to immigrate and emigrate to and from habitat patches and is an important factor in determining community structure. The influence of species dispersal in a metacommunity is poorly understood, particularly its effect at the local community level. We aimed to address this deficiency by evaluating the potential influence of dispersal on local community structure in a rock pool metacommunity. Short term dispersal was quantified over an 11 day period by intercepting propagules dispersing via overflowing pool water and via wind. The composition of dispersing organisms was compared to natural local communities in the rock pools surveyed annually on 11 occasions. On average, the composition and abundance of dispersing species was approximately 54.1 ± 9.3% (mean ± SD) similar to the established pool community. Some species were more abundant among dispersers than they were in the pool community. This may be attributed to several factors including a variation in tolerance to environmental conditions, dispersal capacity, and local scale species interactions (predation and competition). In general, we found considerable similarity between short term dispersal and long term local community structure across a metacommunity. Differences in abundance patterns between the resident rock pool community and the dispersal assemblage emphasize that dispersal, a regional process, must interact with local factors in structuring communities.
Authors:T. W. Therriault, T. W. Therriault, J. Kolasa and J. Kolasa
Habitat variability is largely an external mechanism influencing community variability by affecting abundances and precipitating other community changes but the nature of this influence is poorly understood. The absence of systematic quantitative studies appears to be a major reason for this deficiency. To address the problem, we have evaluated community and population variability in invertebrate communities collected from 49 coastal Jamaican rock pools with contrasting levels of habitat variability. We calculated a multivariate index of habitat variability based on temporal changes in physicochemical variables. Variability in diversity indices (Simpson.s and Shannon-Wiener), evenness (2 measures), and species richness represented community variability while species rank correlations and community constancy represented changes in community structure. Additionally, we analyzed the impact of three habitat generalists (harpacticoid copepod (Nitocra spinipes Boeck), cyclopoid copepod (Orthocyclops modestus Herrick), and the ostracod (Potamocypris sp.)) on overall community variability. As habitat variability increased, both community and population variability increased. Community structure (ranked abundances) was more variable in variable habitats compared to non-variable habitats but communities in these variable habitats retained greater constancy of composition suggesting that highly variable habitats are dominated by a few species with good dispersal abilities. Rare species may come and go, but the dominant species persist in these habitats. Habitat generalists influenced temporal community variability differently, especially evenness (based on the Shannon-Wiener index). Positive relationships were found between the variability in evenness and population variability of the ostracod and cyclopoid copepod. A negative relationship was found between the variability in evenness and the variability of harpacticoid copepods. Our study suggests that individual communities or assemblages respond independently and asynchronously to environmental factors, a view originally proposed by Gleason (1917).We conclude that the form of community structure in variable habitats remains constant. The species composition and relative abundances can change over time but the relative abundance of the dominant species stays high and the remaining species, regardless of their numbers,make relatively small contributions to the overall community variability pattern.
The Energy-Limitation Hypothesis (ELH) predicts that species richness is an increasing function of abundance. In contrast, the Niche-Limitation Hypothesis (NLH) predicts that high abundances become easier to attain as species richness increases. We tested the NLH and ELH using aquatic invertebrate communities of tropical rock pools. These rock pools span a gradient from very low density pools (N <200 individuals/liter) to very high density pools (N ≯ 5001). Only in low density rock pools were species richness and abundance positively related. In intermediate abundance rock pools no relationship between species richness and abundance was observed, and in high abundance rock pools the relationship between species richness and abundance was strongly negative. These patterns are inconsistent with the predictions of the ELH, but not with predictions of the modified NLH which adds carrying capacity to the model.
The contribution of local (e.g., competition) and regional (e.g., dispersal) processes in the structure of communities remains an unresolved issue. In general, a tendency to assume local processes to be deterministic and regional to be stochastic dominates, although it is challenged. Fortunately, it can be cast as a testable proposition: if correct, the degree of determinism in the final community structure might indicate which process is more prominent in the control of community structure. However, recent findings have also suggested that stochastic patterns can arise from local processes and that dispersal can homogenize communities, which would make them appear deterministic irrespective of the mechanism involved. To evaluate these competing expectations we conducted an experiment where the initial communities had the same composition and species abundances. We hypothesized that if local processes dominate, then arrays of communities will show divergence of community structures whether connected by dispersal or not (i.e., being fully isolated). Alternatively, if regional processes dominate, the dispersal connected communities should converge while isolated ones should not. We found, however, that both groups of experimental communities showed similar patterns of change - a decline in similarity and a tendency to diverge. This suggests that biological interactions, demographic stochasticity, or both, exert noticeable control over community structure such that they reduce similarity among replicate communities and diversify their final states. We speculate that these mechanisms enhance potential for species additions, particularly in conjunction with factors such as dispersal and the size of the regional species pool.
Hubbell (2001) proposes that random demographic processes (i.e., neutral dynamics) can explain observed levels of variation in the richness and abundance of species within and among communities. Hubbell's neutral models have drawn attention because they reproduce several characteristic features of natural communities. But neutral models are criticized for ignoring nonrandom processes known to cause species densities to fluctuate. We parameterized neutral models using the population counts of 64 species of aquatic invertebrates collected from 49 discrete rock pools over a 13 year period.We used Hubbell's numerical modeling approach to evaluate the effect of natural population fluctuations on the parameter settings. We also analyzed the effect of observed variation on the species proportional abundance predicted by neutral models. We find that observed levels of variation in abundance are much higher than predicted by neutral models, forcing estimates of themigration probability, m, and fundamental biodiversity parameter, ?, to fluctuateover time. Much of the observed variation is mediated by predator-prey interactions. Low predator densities are associated with fewer species and less even relative abundances of species, resulting in lower estimates of m and ? comparedto periods of high predator densities. Our results show that by assuming an identical survival probability for all species, neutral models misrepresent substantial aspects of community dynamics.