Limnol. Oceanogr., 44(8), 1999, 2005–2011

To evaluate the use of pigments as tracers for determining copepod grazing rates and selectivity, we examined the stability of several biomarker pigments during copepod feeding incubations. During these incubations, we measured changes in phytoplankton-derived chlorophylls and carotenoids in the particulate and dissolved pools. Budgets were calculated to determine changes in pigment concentrations in the food, copepod fecal pellets, copepod guts, and dissolved/colloidal fraction. For each of six algal diets, adult female Acartia tonsa copepods were fed limiting and saturating food concentrations, approximately 100 and 500 mg C L21, respectively. Thalassiosira weissflogii, Rhodomonas lens, Chroomonas salina, Dunaliella tertiolecta, and two Tetraselmis strains were used in the feeding experiments to investigate the fate of fucoxanthin, alloxanthin, lutein, chlorophyll (Chl) b, and Chl a. In all experiments using saturating food concentrations, the dissolved/colloidal pool contained no more than 3% (usually less than 1%) of any pigment, whereas in experiments using limiting food conditions, pigments were undetectable in the dissolved/colloidal pool. Pheopigments were present in fecal pellets and copepod guts in most of the experiments. Chl a destruction (conversion to colorless products) was variable among the different experiments, depending on algal species and food concentration and, in most cases, Chl a was destroyed to a greater extent than the carotenoids. In all cases, pigment destruction was higher when copepods were fed limiting rather than saturating food concentrations. These data attribute the variability in pigment destruction to algal species and concentration, and suggest caution when pigments are used as tracers of herbivory. In such studies, assumptions about conservative behavior, even for the carotenoids, would need to be verified for each set of experimental conditions and grazers. Phytoplankton pigments are commonly used as taxonomic markers in aquatic studies (e.g., Carpenter et al. 1986; Levinton and McCartney 1991; Barlow et al. 1995; Bianchi et al. 1996). Chlorophyll (Chl) a, found in all phytoplankton, is used as an index of total biomass, while other pigments (e.g., Chl b and certain carotenoids) can be used as markers for phytoplankton community composition because some pigments are taxon specific (see reviews by Goodwin 1971; Jeffrey 1980; Millie et al. 1993; Jeffrey et al. 1997). Analysis of pigments in the guts of grazers has become a popular tool for measuring the feeding activity of planktonic herbivores (Mackas and Bohrer 1976; Baars and Oosterhuis 1984; Dagg and Walser 1987; Kleppel et al. 1988; Dam and Peterson 1991; Quiblier et al. 1994, 1996a,b). Although most such studies have used only Chl a to estimate grazing rates, both chlorophylls and carotenoids can be measured using highperformance liquid chromatography (HPLC). This information makes it possible to evaluate selectivity of grazers for different phytoplankton. However, in order to use these pigments as quantitative tracers for zooplankton grazing, the extent and variability of pigment destruction during feeding and digestion must be quantified. Throughout the present text, the term degradation refers to the breakdown of pigments into detectable derivatives (e.g., pheopigments), whereas the term destruction refers to the conversion of pigment into colorless forms, undetectable by light absorption. It is well established that Chl a degradation to pheopigments occurs during grazing, although there is no consensus as to the extent to which destruction into colorless products occurs; the range of values for conversion by copepods of Chl a and pheopigments into colorless products has been reported from 0 to 100% (see review by Dam and Peterson 1988). Most studies have found chlorophyll breakdown products (i.e., pheopigments) in copepod guts and fecal pellets (Downs 1989; Head and Harris 1992, 1994); however, not all of the chlorophyll is always recovered as pheopigments. Downs (1989) found that the copepod Calanus pacificus converted on average 63% of ingested Chl a to pheophorbide and pheophytin, while the rest was destroyed. Although various estimates of chlorophyll conversions have been reported, an average of ;30% pigment loss (conversion to undetectable, colorless products) is assumed in many gut fluorescence studies (e.g., Dam and Peterson 1988). The inconsistent results among studies of Chl a loss during grazing may be the result of the physiology and trophic history of the animals (Mayzaud and Razouls 1992; Tirelli and Mayzaud 1998). Penry and Frost (1991) showed that copepods that had been acclimated for long periods of time (several days) to high algal food concentrations destroyed more pigment than those acclimated to low food. Gut residence time that is negatively correlated with food concentration (Dagg and Walser 1987) has also been hypothesized to affect the extent of pigment alteration (Penry and Frost 1991), presumably because longer gut residence time allows for more breakdown, although elsewhere it has been suggested that destruction may actually occur prior to entry of food into the guts (Head and Harris 1996). Some studies (e.g., Penry and Frost 1991; Head and Harris 1996) have demonstrated that the extent of Chl a destruction varies with the ingestion rates of the copepods. Therefore, animal physiology, trophic history, and potentially other variables (e.g., phytoplankton physiology, species, light environment, etc.) may all play a role in the extent of pigment destruction that occurs during grazing. With the availability of HPLC methods, pigments other than Chl a, particularly carotenoids, have been used to determine which phytoplankton groups are being grazed. However, little is known about carotenoid destruction during copepod feeding. If carotenoids are destroyed during feeding, then grazing rates calculated from gut contents will be underestimated (given the conventional assumption of no ca-