Search

The importance of trophic interactions for determining the distributions, abundances, and taxonomic compositions of organisms in ecosystems has long been studied and debated. Here we test the effect of a trophic cascade on diatom (Asterionella formosa) blooms in subtropical, eutrophic Lake Taihu, China. A long-term data series (2005–2015) on planktivorous fish, zooplankton and diatoms has been analysed. Structural equation modelling is used to test our hypotheses about the influences of top-down and bottom-up forces on A. formosa. Since 2009, a spring bloom of A. formosa has occurred in the lake, coinciding with a reduction of Daphnia galeata biomass and of total cladoceran biomass following a marked increase in the stock of planktivorous fish (bighead carp and silver carp). Light, phosphorus (total and soluble reactive phosphate) and silica did not act as limiting factors for the growth of A. formosa. Structural equation modelling analysis showed that top-down effects of cladocerans on A. formosa biomass were more important than bottom-up effects (wind speed and soluble reactive phosphate). Although A. formosa was negatively correlated with total nitrogen, total with other inorganic nitrogen (NO 3 and NH 4) did not restrict the proliferation of diatom blooms after 2007. These results suggested that the substantial reduction of the Daphnia population caused a diatom bloom through a trophic cascade by planktivorous fish. Our study provides new insight into the effects of trophic interactions on diatom bloom formation in natural freshwater ecosystems.

Filter-feeding silver carp (Hypophthalmichthys molitrix) has been extensively used to control algae blooms in Chinese lakes. However, the benefit of stocking silver carp to restore eutrophic lakes remains controversial and the reasons for the different responses observed are unclear.In this study, we conducted an outdoor experiment to assess the impact of the biomass level of four silver carp (c. 0, 25, 50, and 100 g/m(3)) on phytoplankton through direct grazing and fish-mediated nutrient recycling, respectively. We used a nested design with nutrient-permeable chambers with or without fish placed inside ponds. In this way, the phytoplankton incubated outside the chambers was affected by both consumer grazing and nutrient recycling, whereas the phytoplankton inside the fish-free chambers was only affected by consumer-mediated nutrient recycling.Outside the chambers, the concentrations of total nitrogen, dissolved total nitrogen, and total phosphorus did not differ significantly among the treatments, whereas the dissolved total phosphorus level was significantly higher in the fishless treatment. The concentration of chlorophyll a (Chl a), phytoplankton biomass and the ratio of nanophytoplankton Chl a to total phytoplankton Chl a in the fish-present treatments were significantly higher than in the fishless treatment and the phytoplankton was dominated by small-sized species. While the zooplankton biomass and zooplankton:phytoplankton biomass ratio were significantly lower in the fish-present treatments, indicating a strong fish predation on zooplankton. As for the water inside the chambers, there was no significant difference among the four treatments for any of the measured variables except total phosphorus.Our results suggest that silver carp indirectly stimulated phytoplankton growth by reducing the zooplankton grazing rather than through nutrient recycling, and that fish grazing on phytoplankton could not compensate for the loss of grazing by zooplankton, leading to higher phytoplankton biomass when silver carp was present. The benefit of using filter-feeding fish as a biomanipulation tool to create clear water is therefore questionable unless control of nanophytoplankton can be achieved.