Changes in microcystin production by Microcystis aeruginosa exposed to phytoplanktivorous and omnivorous fish

With direct exposure to phytoplanktivorous fish (Hypophthalmichthys molitrix), increased mass-specific microcystin production occurred in three monoclonal Microcystis aeruginosa strains (NIES 44, 88 and 99). Total mass-specific microcystin content of NIES 44 exposed to H. molitrix was over 50 times higher than controls (a mean value of 16.2 mug g(-1)-dry cell in controls versus 878.6 mug g(-1)-dry cell in treatments). Up to nine times higher microcystin levels were detected in NIES 88 exposed to H. molitrix compared to controls (a mean value of 553 in controls versus 5145 mug g(-1)-dry cell in treatments). The microcystin levels of all strains were significantly different between controls and H. molitrix treatments (P < 0.01 for NIES 44 and 88; P < 0.05 for NIES 99). The microcystin response to the omnivorous Carassius gibelio langsdorfi was weaker than that of H. molitrix, though the levels in all strains exposed to the fish were higher than in controls and a significant difference in microcystin production between controls and omnivorous fish treatments occurred for NIES 44 (a mean value of 6.9 in controls versus 41.5 mug g(-1)-dry cell in treatments; P < 0.01) and NIES 88 (a mean value of 359.8 versus 480.4 mu g g(-1)-dry cell; P < 0.05). Microcystis cells were observed in the both fish faeces and gut contents, and microcystin was also detected in the body tissues (from 0.6 to 2.5 mug g(-1)-dry weight) and faeces of both fish species on the final day of experiment, although 98% of fish in three strains of Microcystis cultures had lost weight (mean +/- S.E. fish growth rate with M. aeruginosa; -0.90 +/- 0.06% per day, n = 96). This study showed that several M. aeruginosa strains increased toxin production when exposed to fish, especially phytoplanktivorous species, even though fish appeared not to feed vigorously on toxic Microcystis, and supports the hypothesis that this response is a fish-induced defence mediated by physical contact associated with feeding or by chemical cues (e.g. kairomones). (C) 2004 Elsevier B.V. All rights reserved.