POPULATION-DYNAMICS OF THE MUSSEL MYTILUS-EDULIS ALONG ENVIRONMENTAL GRADIENTS - SPATIAL VARIATIONS IN DENSITY-DEPENDENT MORTALITIES

1. From September 1976 to September 1983, we examined how studies of the 12 individual mussel beds in the Exe estuary could enhance our understanding of the dynamics of the whole intertidal mussel Mytilus edulis L. population. The 12 beds varied considerably in their distance up-river from the estuary mouth, muddiness of their substrata and level on the shore, the three main features defining the environmental gradient for mussels in the Exe. 2. Since adult (>1-year old) densities over the whole estuary were already known to be strongly regulated during the first year, we focused on spat (0-year) mussels. Mortality and recruitment were measured as k-values over three first year stages; k3 was a notional mortality and measured the inability of the beds to support as many spat as they produced, without any implications regarding the provenance of the larvae that did settle on the beds in March, the main settlement period; k4 was the first summer mortality, or recruitment, of larvae from March to September; k5 was the overwinter mortality by the following March when mussels became adult. 3. Mortality during recruitment, either in the water column or on the bed at or following settlement, was the key factor on all beds, though its timing varied. On eight beds, the difference between egg production and spat settlement in spring (k3) was the key factor whereas, on four, it was net recruitment during the first summer (k4). There was some evidence of synchrony between beds in annual variations in the three k-values. 4. Density dependence was detected directly in all three stages on only a few beds because of a narrow range of initial densities combined with very variable k-values and a small sample of years. Combining data from all beds suggested density dependence was widespread, especially in stages k4 and k5. The best model ascribed a strong, common, density-dependent slope to all beds in each stage, but with each bed having different intercepts. 5. The intercepts on the x-axes measured the critical density above which there was density-dependent mortality and below which there was density-dependent recruitment. Critical densities were shown by regression analyses to be correlated consistently with adult density, with additional negative associations with substratum muddiness in k3 and k4, and distance up-river in k3. Higher densities of young mussels were therefore recruited (k3 and k4) or survived (k5) on beds where adult densities were already high. 6. These density-dependent relationships resulted in a temporal and spatial compensation in the pattern of settlement and mortality between beds. In particular, k4 acted as a spatial compensatory process between the downstream beds, that suffered losses, and the upstream beds that had continuing settlement throughout most summers; thereby reducing variations in spat densities between beds within years. Also, density-dependent losses or recruitment in k4 reduced the variation in spat recruitment over the whole estuary between years and, therefore, acted as a compensatory process over time. 7. The residual, density-independent variation in k-values around the density-dependent relationships co-varied between years over 10 of the 12 beds. Regression analysis suggested that these fluctuations in k3-k5 were related to annual variations in the weather to which the spat were exposed at low water.