Over the course of a bloom, the quantity and quality of algae available for copepods to eat may vary dramahcally The effects of varying phytoplankton concentration and quality on fecal pellet produchon rates and fecal pellet charactenstics of the calanoid copepod Acartia tonsa were examined in the laboratory Bloom conditions were siinulated by feeding copepods cells of the diatom Thalassioslra weissflogii grown in batch cultures Experiments were run when algae were rapidly growing (early bloom) and when algae were in stationary growth phase (late bloom) at concentrations ranging from 0 to 3200 cells ml-' Daily pellet production rates and volume, density, and carbon and nitrogen content of pellets were measured Results show that for a given algal growth stage pellet production rate and pellet volume increase linearly w~t h cell concentration up to a crit~cal point beyond whlch no further increases in these parameters occur In general, (1) maximum pellet volumes were higher when copepods ate late-bloom cells, (2) pellet densities were higher when copepods ate early-bloom cells and (3) volume-specific pellet carbon was higher when copepods ate early-bloom cells Sinking rates of pellets were estlrnated from an empincal model based on pellet volume and density Sinking rates of pellets were lowest at low concentralons (200 cells ml-l) concentrahons of early-bloom cells Sinking rates were highest at high concentrations (2000 cells ml-l) of early-bloom cells S i n l n g rates of fecal pellets were relatively high regardless of food concentration when copepods ate late-bloom cells Daily volume-speclfic fecal matter produced per copepod was also calculated Maxlmum values of total fecal matter index were h~g h e r when copepods ate early-bloom cells Our results indicate that copepod fecal pellets change in a predictable manner over the course of a bloom We conclude that copepod fecal pellets may play different roles in recychng and vertically transporting carbon over the course of a phytoplankton bloom KEY WORDS: Fecal pellets. Phytoplankton blooms. Carbon flux. Food quality. Zooplankton

We estimated in situ ingestion rates of females of the calanoid copepod Temora longicornis in Long Island Sound, New York, USA, at roughly weekly intervals, between February and August from 1985 to 1987 using the gut fluorescence technique. Variability in gut pigment content (GPC) and ~ngestion rates (IR) was examined in relation to changes in the concentration of total chlorophyll and of chlorophyll in the > 20 vm size fractions, and changes in body size. Correlations of GPC and IR with chlorophyll concentration were highest for the > 20 Km size fraction and lowest for total chlorophyll. Saturation of GPC and IR was apparent only when phytoplankton > 10 pm or > 20 pm was considered as available food. These results lend support to the notion that the degree to which herbivorous coastal copepods are food-limited may be strongly dependent on the structure of the phytoplankton assemblage Prosome length of female T long~corn~s decreased by almost a factor of 2, and dry weight by a factor of 4, from February to August Correlations between weight-specific GPC, welght-specific IR and chlorophyll concentration in all size fractions were poorer than in the cases of GPC and IR on a per mdlvidual basis However, a multivariate regression analysis indicated that most of the vanance in GPC and IR was explained by the different size fractions of chlorophyll, not body weight. Therefore, the use of weight-specific indices may not always be justified even when variation in body size is pronounced. We suggest that multivanate analysis or other statistical techniques b e employed, instead of weightspecific indices, when body size may be a confounding vanable.

The instantaneous amount of food in an animal's gut, or gut content (G), is a function of its ingestion rate (I) minus its gut evacuation rate (E). We present simple models describing the short-term (minutes to hours) dynamics of G, incorporating I and E simultaneously. The models apply to situations in which animals start feeding after a period of starvation. Two possibilities are considered: (1) Both I and K (the coefficient of gut evacuation) remain constant during the feeding period. (2) K remains constant, but I decreases, after satiation, from the initial value (I,) to an equilibrium value (I,). Results from the models show that although gut content is usually asymptotic with time, the shapes of the curves vary considerably depending on the relationship between I, and Kin the first instance, and I,, I, and K in the second instance. The relationship between the time of first defecation (tl) and the gut passage time (K-') is critical in these models. Results from the models are compared to data from published experiments in which gut content of copepods was monitored at intervals of minutes to hours. We conclude that short-term variability in feeding behavior may not always be inferred solely from changes in gut content with time.