A direct test of cyanobacterial chemical defense: Variable effects of microcystin-treated food on two Daphnia pulicaria clones

To determine the direct effects of microcystin on the fitness of herbivorous zooplankton, we experimentally added microcystin-LR to freeze-dried cells of palatable Chlorella and fed these compound-treated cells to two clones of Daphnia pulicaria that had shown differing responses to a diet containing a strain of Microcystis aeruginosa that produces microcystin. The Daphnia that performed better on a diet containing live Microcystis showed reduced population growth when exposed to microcystin-LR-treated Chlorella, whereas the Daphnia that performed poorly on the diet containing live Microcystis was not affected by the experimental diet containing microcystin-LR. This is the first study to unambiguously show that some Daphnia strains are and some strains are not harmed by the consumption of microcystin-LR. These surprising results were not generated by interference from lipophilic secondary metabolites in Microcystis. When the crude lipophilic extract of Microcystis was added to dried Chlorella cells, it enhanced the fitness of both Daphnia clones. We hypothesize that the Daphnia clone more tolerant to live cells may upregulate resistance when cued by the presence of the live Microcystis cells, but not by microcystin-LR alone. Alternatively, the Daphnia clone that grew well on a diet containing live Microcystis may sequester compounds from Microcystis that defend the cyanobacteria from autotoxicity; these compounds would have been unavailable to Daphnia consuming freeze-dried Chlorella treated with microcystin-LR alone. Thus, microcystin-LR can suppress Daphnia fitness when consumed; however, the effects of microcystin vary across clones of herbivorous zooplankton and the consequences of this variance should not be overlooked when considering zooplankton-cyanobacteria interactions. Blooms of cyanobacteria threaten aquatic communities and global water supplies (Paerl 1988) because they produce secondary metabolites that can harm or kill fishes, livestock, and humans (Carmichael 1992). Some of the more commonly studied cyanotoxins include cyclic peptides that target mammalian livers (e.g., microcystin and cylindrospermopsin), nerve synapses (e.g., anatoxin-a), and gastrointestinal tracts (e.g., lyngbyatoxin-a) (Carmichael 1992; Chorus and Bartram 1999; Zurawell et al. 2005). Despite a large correlative literature suggesting that cyanotoxins may deter feeding by zooplankton (Lampert 1987; Watanabe et al. 1996; Zurawell et al. 2005), the effect of cyanobacterial secondary metabolites on ecologically relevant organisms, such as herbivorous zooplankton, is equivocal (Wilson et al. 2006a). There are no direct empirical tests of how cyanobacterial secondary metabolites affect grazers consuming these compounds. Cyanobacteria have been proposed to lower zooplankton grazing rates in three basic ways: (1) by occurring as large colonial and filamentous morphologies that clog filtering appendages or are otherwise inedible, (2) by being nutritionally deficient, and (3) by producing toxic secondary metabolites (Porter and Orcutt 1980; Lampert 1987). There are direct tests of the effects of morphology (FerrãoFilho and Azevedo 2003) and nutrition (DeMott and Müller-Navarra 1997; von Elert and Wolffrom 2001) on zooplankton feeding, but no study has unambiguously assessed how consuming cyanobacterial secondary metabolites affects zooplankton fitness. Past experiments have compared zooplankton performance when (1) immersed in media containing dissolved cyanotoxins (Reinikainen et al. 2001) or cyanobacterial extracts (Wheeler et al. 1942), (2) fed diets containing cyanobacteria versus being starved (Arnold 1971; reviewed by Wilson et al. 2006a), (3) fed diets containing cyanobacteria versus foods supporting better growth (Arnold 1971; reviewed by Wilson et al. 2006a), (4) fed diets of two conspecific cyanobacteria that are or are not toxigenic to mammals (Smith and Gilbert 1995), and (5) fed diets consisting of a wild-type cyanobacterium containing microcystins or its mutant that lacks the ability to produce microcystins but may produce other toxic, nonmicrocystin oligopeptides (Rohrlack et al. 2005). Results from these studies show that cyanobacteria are poor food for grazers, but do not unambiguously demonstrate the effects of consuming cyanobacterial secondary metabolites (see review by Wilson et al. 2006a), because cyanotoxins are either present in unrealistic forms (dissolved in water rather than bound in cells) or are potentially confounded by other factors (i.e., cyanobacterial species or strains with and without a toxic secondary 1 Present address: Cooperative Institute for Limnology and Ecosystems Research, University of Michigan, School of Natural Resources and Environment, NOAA/GLERL, 2205 Commonwealth Blvd., Ann Arbor, Michigan 48105-2945. 2 Corresponding author ([email protected]).