Fishing, other human activities, and natural perturbations can alter the species composition and size structure of fish communities in coastal ecosystems. Normalized biomass size spectra (NBSS) and other metrics based on size and abundance of fish communities are sensitive to effects of fishing and have been proposed as useful tools for ecosystem-based management. However, these approaches based on size and abundance are unevaluated at temporal and spatial scales relevant for management within estuaries. Because individual species have important ecological and economic value, tracking temporal and spatial changes in the species composition of the fish communities using multivariate analyses, such as principal component analysis (PCA), can facilitate interpretation of patterns observed in the NBSS. A goal of my dissertation was to determine if indicators suitable for ecosystem-based management can be derived from NBSS parameters and other metrics based on size and abundance for estuarine fish and plankton communities at relatively small temporal and spatial scales. Additionally, I sought to elucidate effects of temporal and spatial variability in species composition on community size structure of estuarine fish communities by combining multivariate and NBSS analyses. Analyzing data from multiple fisheries-independent surveys and water quality monitoring programs, the objectives of my dissertation were 1) to describe and quantify the size distribution and community composition of fish and plankton in Chesapeake Bay at temporal scales ranging from months to over a decade and at spatial scales ranging from 18 km to 100 km, 2) to evaluate long-term trends in abundance, size distribution, and species composition of fish communities in Chesapeake Bay and Pamlico Sound, and 3) to analyze environmental variables and their effects on community structure and size distribution of biological communities in the Chesapeake and Pamlico Sound estuaries. Results supported the conclusion that NBSS combined with traditional community analyses permits detection of changes in ecosystem status, facilitates identification the species associated with the observed variability, and provides a framework to establish management reference points.