Progressive restoration of a shallow lake: A 12-year experiment in isolation, sediment removal and biomanipulation

1. Cockshoot Broad (mean depth 1 m, area of open water 3.3 ha) was isolated from the sewage-effluent contaminated River Bure in 1982. About 70 cm of sediment were pumped out at the same time and settled in a remote lagoon. The total phosphorus and phytoplankton chlorophyll a concentrations rapidly fell and large numbers of large-bodied Daphnia were present. The zooplanktivorous fish stock was probably then very low. Submerged macrophytes colonized one area (the Dyke) heavily but more slowly over the main basin. 2. These conditions lasted until 1985 by which time the fish stocks had increased, the Daphnia declined, and plants disappeared from the main basin. Plant communities remained stable in the Dyke. Chlorophyll a increased in the main basin and remained at a high value until 1988. 3. In early 1989, virtually all of the fish community was manually removed. Daphnia numbers recovered, chlorophyll a concentrations fell and plants began to recolonize the main basin. Recolonizing fish were removed in successive winters and chlorophyll has remained low and Daphnia high except in 1992. In that year, an estuarine mysid, Neomysis integer, colonized the Broad when the dams separating it from the river were overtopped. This was a consequence of low fluvial flows in a drought period and was apparently only a temporary set-back. There was also a loss of macrophytes from the dyke in 1991/92. This is thought to have resulted from bird damage to the overwintering inocula in a cold spring. 4. In 'low Daphnia-high fish' years, there was a significant positive correlation between total P and chlorophyll, but not in 'high Daphnia-low fish' years. Overall, however, there was no correlation between chlorophyll and nutrients but a strong inverse correlation with Daphnia numbers. The nutrient conditions supporting phytoplankton and macrophyte dominance overlapped and the two conditions are seen as alternative stable states. The results are discussed in the context of a general hypothesis of change with eutrophication in shallow lakes.