INDUCTION OF SEAWEED CHEMICAL DEFENSES BY AMPHIPOD GRAZING

Grazing by the generalist amphipod Ampithoe longimana induced increase d concentrations of defensive secondary metabolites in the brown alga Dictyota menstrualis and made the seaweed less susceptible to further attack by the amphipod. Although A. longimana preferentially consumes D. menstrualis, its feeding rates can be reduced significantly by high concentrations of diterpenoid dictyols produced by the alga. In 1991, D. menstrualis from sites with high numbers of A. longimana had highe r levels of grazing scars, higher concentrations of dictyols, and were less palatable to A. longimana than plants from sites with few amphip ods. Among-site differences in palatability to amphipods did not corre late with plant differences in protein, nitrogen, or carbon content. W ithin a site, plants that had apparent amphipod grazing scars were sig nificantly less palatable to A. longimana than neighboring undamaged p lants. Controlled field experiments manipulating A. longimana densitie s supported the hypothesis that feeding by this amphipod induced eleva ted chemical defenses in the alga. Compared to undamaged control plant s, amphipod-damaged plants had 19-34% more of three diterpenoid second ary metabolites and were 50% less palatable to amphipods. Soluble prot ein and thallus toughness were unaffected by amphipod grazing and thus could not have caused the differences in palatability. High-pressure liquid chromatography evaluation of adventitious branches growing from blade margins at sites of amphipod grazing scars showed that these br anches had significantly elevated levels of two diterpenoids relative to normal blade apices or middles. Thus, the amphipod-induced resistan ce to further attack occurs through an increase in chemical defenses, and these defenses are, to some extent, localized within the plant tha llus. Among-site differences in amphipod densities, grazing scars, sea weed defensive chemistry, and plant palatability that we documented in 1991 varied considerably during 1992 and 1993, suggesting that these interrelationships may be complex. In 1992, A. longimana densities did not differ between sites, and there were no between-site differences in palatability or concentrations of deterrent secondary metabolites. In 1993, however, A. longimana densities did differ between sites, but between-site differences were less dramatic than in 1991. Some second ary metabolites were slightly, but significantly, increased at the sit e with higher densities of A. longimana, but this had no effect on A. longimana feeding. It has been long recognized that marine herbivores are active participants in seaweed-herbivore interactions and can grea tly influence the structure of benthic algal communities. Our findings suggest that seaweeds are not passive participants in these interacti ons, but can actively alter their susceptibility to herbivores in ecol ogical time. Induced responses to herbivory help explain both spatial (i.e., within-thallus, within-site, and among-site) and temporal varia tion in the chemical defenses of D. menstrualis.