Pepin: The encounter rate of Mallotus villosus with fish eggs 



205 



2000; Koster et al., 2001; Garrison et aL, 2002). Indi- 

 vidual-based models have often served as a tool to infer 

 the potential effect of predators on early life stages 

 (Cowan et al., 1999; Paradis et al., 1999; Werner et al., 

 2001) where the likelihood of a prey being eaten by a 

 predator is the product of the probabilities of encounter, 

 attack, and capture. Although the latter two components 

 of the predation process have been estimated in the 

 laboratory for a number of species (see Paradis et al., 

 1996, for a review), encounter rates are often modeled 

 as the product of predator's swimming speed and reac- 

 tive distance with limited substantive data to support 

 the overall estimate. Furthermore, encounter rates are 

 assumed to occur randomly and thus follow a Poisson 

 distribution; yet there have been few attempts to vali- 

 date this assumption (but see Huse and Toresen, 2000). 

 Given that encounter rates represent a key parameter 

 required to estimate the potential impact of predators 

 on early life stages, it is surprising that knowledge of 

 this element in the field is so limited. 



The goal of the present study was to estimate effective 

 encounter rates of adult capelin with fish eggs and larvae 

 in coastal Newfoundland. The approach was to obtain 

 a high number of specimens from a series of intensive 

 diurnal collections to provide an accurate estimate of 

 the probability distribution of prey numbers from which 

 encounter rates could be derived. The primary focus of 

 this analysis was on the presence and frequency of occur- 

 rence of fish eggs as prey for capelin, because they per- 

 sist longer in stomachs than do fish larvae (Hunter and 

 Kimbrell, 1980; Folkvord, 1993). I used estimates of prey 

 availability from plankton samples combined with the 

 frequency of occurrence of eggs in stomachs, in relation 

 to the size of capelin and time of day, to estimate effec- 

 tive encounter rates (volume searched) by capelin. Fish 

 eggs are well suited to this objective: they are completely 

 passive and thus the probabilities of their avoidance of 

 both plankton nets and predators can be considered as 

 nil. It is thus possible to measure accurately their abun- 

 dance in the field, given appropriate choice of mesh size, 

 and their numbers in predator stomachs can provide a 

 measure of the effective volume searched, which is the 

 product of the probabilities of encounter and attack. 



Materials and methods 



The study was conducted in Trinity Bay, Newfound- 

 land, Canada, between 17 and 31 July 2000, as part 

 of a larger program dealing with the dynamics of ich- 

 thyoplankton (Baumann et al., 2003). Five sites were 

 selected at which to conduct diurnal sampling of capelin 

 based on the observation of high juvenile and adult 

 capelin densities from hydroacoustic returns. Sampling 

 was conducted within a 3-km radius of each site and at 

 4-hour intervals over a single twenty-four hour interval. 

 Each site was treated as an independent observation of 

 the diurnal feeding patterns of capelin. As a result of 

 poor catches, sampling was stopped after two sampling 

 periods at the two sites located at the head of the bay 

 (Table 1). 



Capelin were sampled using an international young 

 gadoids pelagic trawl (lYGPT) mid-water-trawl towed 

 in a single oblique haul from the surface to a 100-m 

 depth. Each tow lasted approximately 30 minutes. Up 

 to 50 juvenile and adult capelin were taken at random 

 from the catch. When capelin were highly numerous, 

 sample size was increased to 100. For each specimen, 

 total length (TL) was measured to the nearest mm, 

 the entire intestinal tract was removed by cutting the 

 oesophagus as close to the mouth as possible, and the 

 tract was then preserved in individual numbered vials 

 containing 5% formaldehyde. A total of 1052 specimens 

 were obtained from these collections. After each trawl 

 haul, a plankton sample was taken by using a 60-cm 

 bongo net (fitted with 333-im mesh net and flowmeters) 

 towed obliquely from the surface to 100 m along the 

 same path as the mid-water-trawl, and the sample was 

 preserved in 2% formaldehyde solution. 



Stomachs were processed by slitting them open and 

 scraping out the contents onto preweighed aluminium 

 weight boat. The contents were blotted dry to remove 

 all excess liquid and weighed to the nearest 0.01 mg 

 by using a Cahn microbalance. Stomach contents from 

 each capelin specimen were placed into individual petri 

 dishes and examined for the presence, number, and 

 taxonomic identification of pelagic fish eggs only. An egg 

 was counted only when the specimen was complete and 



