Poststarvation feeding and swimming activity in Calanus pacijkus and Metridia pacijica

Gut fluorescence analysis of Calanus paczjkus and Metridia pacijica is used to describe variation in food intake rate at time scales of one to several hours. The animals are starved before each experiment, then exposed continuously to phytoplankton cultures. Variability of gut fullness is large (order IO-fold) between individual copepods captured at the same time and is about proportional to the within-time-period average. The gut fullness curve shows a strong initial peak l2 h after food is supplied and drops to about 20% of the peak level after 7-9 h of exposure to high food levels. For Calanus gut fullness and swimming activity at the time of capture are strongly correlated. Active animals have on average 2-5 times as much pigment in their guts as inactive animals captured at the same time. However, inactive animals resume swimming and feeding activity in l-3 h. The changes in gut fullness appear to involve episodic on-off switching of feeding activity as well as regulation of maximum instantaneous feeding rate and of either or both satiation and hunger thresholds. The feeding rate of herbivorous copepods is not uniform with time. Twenty-four-hour periodicities have been demonstrated for many species (e.g. Gauld 1953; Petipa 1958; Mackas and Bohrer 1976; Dagg and Grill 1980). At much shorter time scales (order 0. l-l 0 s), microcinematographic studies (Rosenberg 1980; Cowles and Strickler 1983) have shown that animals alternate brief bouts of slow swimming/feeding with “breaks” during which swimming and feeding activities cease. There is also substantial evidence that planktonic animals may reduce or interrupt their feeding activity over intermediate time scales (up to a few hours) in response to previous feeding success. Populations of starved animals show much higher average ingestion rates during the first few hours after exposure to food (Mullin 1963; McAllister 1970; Runge 1980). Their food intake is initially 1.5-3 x that of animals preconditioned to the experimental level of food availability, but after several hours approaches the feeding rate of preconditioned animals (Runge 1980). A different line of evidence is provided by field studies (Pearre 1973; Mackas and Bohrer 1976; Boyd et al. 1980; Simard et al. 1985) in which animals with full guts were found at depths well below their probable food source. The interpretation of these results was that satiated animals stop feeding and modify their swimming and migratory activity. Laboratory studies (T. Cowles pers. comm.; E. Anderson pers. comm.) also suggest that swimming activity is linked to food availability. A poststarvation feeding peak and subsequent decline to an equilibrium feeding rate could occur by either or a combination of two mechanisms (Fig. 1). Each individual animal could be feeding continuously but gradually reducing its rate of food intake (Fig. la, b), or the animals could be switching between feeding and nonfeeding modes of behavior (Fig. lc, d), with the variation in average feeding rate determined by the allocation of time between the two modes. If the feeding mode was cued by hunger and the nonfeeding mode by some threshold level of satiation (possibly variable between individuals) then a population of starved animals would tend to have an initially synchronous period of active feeding by all of its members, followed by a gradual transition to feeding episodes randomly phased between individuals. To account for the observed duration of the poststarvation feeding peak and to allow significant rearrangement of the vertical distribution of full animals, we need a time scale for any alternation between feeding and nonfeeding modes of order l-2 h or longer. Because the previous laboratory studies of the poststarvation feeding response measured the average response of many animals per experimental treatment (typically lo15) and integrated the food intake over a relatively