The role of phosphorus release by roach [Rutilus rutilus (L.)] in the water quality changes of a biomanipulated lake

SUMMARY 1. Following fish removal, the water quality in biomanipulated lakes often improves concomitant with decreased phosphorus (P) levels. Because the decrease in P concentrations derives most probably either directly or indirectly from fish, which are the main target of biomanipulation, this study examined the P release of 0+, 1+ and 2+ roach Rutilus rutilus (L.)] and changes in the P release during summer in a shallow eutrophic lake in Finland. 2. The P release was separated into P derived from benthic and littoral food items and into recycled P derived from feeding on zooplankton, to estimate the contribution of net P additions to the water column by the fish to the increase in P concentrations of the lake water (75–110 mg P m^?3) in summer 1991–96. 3. Individual P release of roach by both egestion and excretion was estimated with a bioenergetics model. The size of the roach population was estimated with a depletion method and the proportions of different age groups from catch samples, using a programme separating mixtures of normal distributions. The sensitivity of the release estimates to variation in the growth data was estimated with the jackknife technique. 4. The biomass‐specific P release by 0+ roach (0.36–0.54 mg P g^?1 day^?1) was higher than that by older roach (0.07–0.16 mg P g^?1 day^?1) throughout the summer. The P release by the whole roach population deriving from benthic and littoral food items (0.7–2.7 mg m^?3 during July to August, representing a net addition to the water column) was 5–19 times lower in 1991–96 than the recycled P release deriving from zooplankton (8.9–25.7 mg m^?3), and too low to explain the increase in the P concentration of the lake water during the summer. Because the biomass‐specific P release and roach diet composition vary with fish age, it is important to consider the age structure of fish populations to obtain correct estimates of P release and net additions to the water column. 5. The removal of roach by fishing diminished the roach stock greatly, but the fish‐mediated P release to the water column changed little. This effect was because of the high compensation capacity of the roach population, leading to high recruitment of young fish with higher biomass‐specific P release rates. 6. External loading is very low during summer months and therefore it cannot explain the increase in the P concentration of water during that time. Internal loading from the sediment might be as high as 10.2 mg P m^?2 day^?1, i.e. 50 times higher than the maximum net P addition by the total roach population.