We investigated factors affecting sex change in the Pandalid shrimp Pandalus borealis on the Scotian Shelf. Transition from male to female occurred at different sizes and ages, and could not be related to a minimum size or age. Our data did not show a positive relationship between abundance of older females and shrimp size at sex transition, nor a negative relationship between male:female sex ratio and shrimp size at transition, that would be expected if the population was compensating for decreases in reproductive capacity as predicted by sex allocation theory. Size at transition was inversely related to female density, which was attributable to density dependent growth affecting all stages. Density dependant growth also appeared to explain previously reported results which had been used to support sex allocation theory. Density was the most important factor determining individual growth at high densities while at low densities other factors, including temperature, were also an important determinant of growth. We conclude that a decrease in the size at transition can be indicative of a healthy, as well as a declining stock, and must be fully understood before it can be used as an indicator of stock status. Possible alternative mechanisms for regulating sex change in Pandalids are discussed.

Fisheries and OCCLB~S Canada, Resource Bronch, P . 0 . Box 550, Nakifkx, M.S. BdJ 2S7 JAMIESON, G . S . , W. STONE, AND M. E~E R . 1982. Predation of sea scallops (Fkacopecresa magelkanicus) by lobsters (Nontarusumeric~~nus) and rock crabs (Cancer irror~fu.~) in underwater cage enclosures. Can. S. Fish. Aquat. Sci. 39: 499-585.Predation studies in underwater cage enclosures on natural subs~dtes were conducted in 1979 and 1980 to assess the propensity of adult American lobsters, Homai-us arneric*arae[s, and rock crabs, Cancer irroratus, to prey upon sea scallops, P/acopecten rnagebbanictss. In % 979. the initial prey by both predator species was large, embedded horse mussels (Mohdic~lus modiobus). H n the enclosure containing lobsters, this was followed by lobster cannibalism and, finally, when lobster density was reduced, predation on Barge sea scallops (88-1 10 mrn shell height). No cannibalism or predation on large scallops occurred in the rock crab enclosure.In 19869, small scallops (40-55 rnm shell height) were presented to lobsters and rock crabs; predation occurred regardless of predator density. Large sea scallops and unembedded horse mussels were consumed when crustacean densities approached natural levels. These studies demonstrate that small sea scallops and embedded horse mussels can be readily consumed by lobsters and roc86 crabs under field conditions, but that natural predation of large scallops by either rock crabs or lobsters is of lesser magnitude.