The objective of this research was to improve our understanding of how changes in the environment affect ecological processes. Change detection is often confounded by the large variation found in ecological data due to the difficulty of finding replicates in nature. Intertidal communities were chosen for studies of biophysical interactions because the physical gradients are very strong, thus creating complex systems within spatial scales that are easily sampled. The selection of replicate beach habitats was the first step in designing a sampling protocol for comparative analyses of nearshore community structure. A high resolution shoreline partitioning model was developed to quantify the physical attributes of homogeneous shoreline segments and to statistically cluster replicate segments. This model was applied at 3 locations in Washington State. A portion of the south shore of San Juan Island was partitioned and the physical attributes quantified. Three groups of rocky segments differing only in slope angle were selected for biological sampling. The objective was to test the fidelity of macroalgal and invertebrate populations to replicate bedrock shore segments. The results showed that community structure and population abundances were more similar within groups of replicate segments (similar slopes) than among groups (different slopes). In South Puget Sound, community structure was compared to test for a deterministic organization of communities among replicate soft sediment beaches in an estuary. The results showed that replicate beach segments support similar communities, that communities become less similar as the distance between replicates increases, and that replicates within or among nearshore cells with similar temperature and salinity support communities that are more similar than replicates among cells with different water properties regardless of distance. On the outer Olympic coast, community comparisons were made among 9 sand beaches over a shoreline distance of 250 km. The results show that these communities are similar within segments and within nearshore cells, but because of population abundance fluctuations, the communities were different among cells and among years. This study shows that processes determining patterns in nearshore habitats can be quantified, which is a significant contribution to studies of habitat distribution and the siting of marine preserves. / Graduation date: 2000

As ecologists are being called upon to predict the consequences of human perturbations to natural communities, an important goal is to understand what factors drive variability or consistency in nature. In the rocky intertidal of San Juan Island, Washington, a comparative experimental approach was used to investigate spatial and temporal variation in community organization. The effect of predation on B. glandula varied dramatically over small spatial scales between microhabitats but was remarkably consistent over time withing a given microhabitat. The effects of predation on S. cariosus varied over time within the same microhabitat. By repeating previous landmark experiments at our study site, and replicating these experiments across microhabitats, the domain of applicability of previous experimental information was greatly expanded. In an early successional assemblage on the Oregon coast, I tested the hypothesis that, when the direct effect of one species on another increases in relative strength, its total effect (direct + indirect) is less variable or conditional than if the link between those species is weak. The effect of strong predation by whelks was less sensitive to the presence of additional species and more consistently dampened natural variation between experimental starting dates and between individual replicates within a given experiment. In contrast, the outcome of weak predation was more spatially and temporally variable in sign, and whether it magnified or dampened differences between individual replicates varied between experiments. Consequently the mean total effect of weak predation generally did not differ significantly from zero. However, in some cases, the range of variation (both within and between experiments) in the effect of weak predation exceeded the magnitude of the strongest total effect observed. Longer term results of the experiments on the Oregon coast examined the role of historic factors in influencing the degree to which successional paths are canalized and repeatable or contingent and variable. Succession in mid-intertidal patches in the mussel bed displayed complex patterns of historic effects that varied between species and between different stages of succession. Despite its potential complexity, this system exhibited some consistent and repeatable patterns of succession. Some important canalizing, or noise-dampening forces in this system included: 1) physiological and life-history constraints, 2) compensatory responses of functionally redundant species, and 3) strong interactions between species. / Graduation date: 1996