Spatial, temporal and interactive variability of infauna in Swedish coastal sediments

The relative sizes and importance of spatial, temporal and interactive variability of soft-sediment assemblages in Swedish coastal waters were estimated using a hierarchical sampling programme. Spatial scales ranged between 100-1 and 103 m and samples were collected at three times within 1 year. Variability in diversity of assemblages, as well as abundances of the 15 most common taxa, were estimated and tested using ANOVA. Nonparametric multivariate analyses of variance (NP-MANOVA) were used to test and estimate the variability of whole assemblages. Univariate analyses revealed that all variables showed significant spatial and temporal patterns but that these differed among individual taxa. A general finding was that there was much interactive variability between times of sampling and the two spatial scales tested. This means that abundance and diversity of infauna changed differently among bays located a few kilometers apart and among locations within bays located hundreds of meters apart. Thus, spatial patterns of assemblages changed among times of sampling. Conclusions were similar for univariate as well as multivariate analyses, although multivariate patterns appeared slightly more persistent. Furthermore, estimation of variance components showed a substantial variability among cores within locations. The results provide the first coherent analysis of interactive scale-dependent variability in these geographic areas and one of few in soft sediments. Results are discussed in relation to previously observed patterns of spatio-temporal variability in soft sediments and in a context of potential processes responsible for creating patterns at particular spatial scales. Results were also used for cost-benefit analyses in order to compare precision of optimal solutions to what was observed and to optimise future sampling of these assemblages.