Habitat area more consistently affects seagrass faunal communities than fragmentation per se

Abstract

Seminal ecological theories, island biogeography and the single large or several small (SLOSS) reserve debate, examine whether large contiguous habitats conserve biodiversity better than multiple smaller patches. Today, delineating the ecological effects of habitat area versus configuration in a fragmentation context remains difficult, and often confounds efforts to understand proximate and ultimate drivers of community change in response to habitat alteration. We examined how the major components of fragmentation, habitat division versus area loss, independently influence faunal communities using landscapes constructed from artificial seagrass at scales relevant for juvenile estuarine nekton. We deployed 25 unique, 234-m² landscapes designed along orthogonal axes: habitat percent cover (i.e., area) and fragmentation per se (i.e., patchiness) to examine their effects on faunal density, community composition, and probability of bait-assay consumption. Faunal sampling occurred in both artificial seagrass and interspaced sandflat matrix. We also examined whether larval-settler density drove faunal density patterns across landscapes. Further, we assessed the relative importance of landscape-scale parameters versus fine-scale complexity–canopy height and epiphyte biomass–in determining faunal densities. We most consistently observed increasing epibenthic fish and macroinvertebrate density with increasing seagrass percent cover. Fragmentation per se only negatively affected epibenthic faunal density within the matrix at low seagrass coverage. Bait consumption increased with seagrass cover, suggesting larger habitats are relative foraging hotspots. Alternatively, benthopelagic fish density was unaffected by habitat parameters, reflecting lower seagrass reliance, or increased matrix tolerance. Community compositions did not vary across landscapes, suggesting that abundant species used landscapes indiscriminately. Finally, the relative importance of habitat parameters shifted across faunal guilds and life stages. Landscape percent cover most affected epibenthic faunal density, but not benthopelagic fish density, and neither pattern was related to settler density. Further, only fine-scale complexity influenced settler densities. Collectively, our results indicate habitat area is a primary, positive driver of faunal densities and generalist consumption, and therefore should be prioritized in seagrass conservation. However, sampling across spatial scales and habitat types revealed nuances in habitat use patterns among faunal guilds and life stages that were not solely area-dependent, illustrating that a variety of landscape configurations support essential nursery functions.