Large numbers of invertebrates were collected from the lower layer of seasonal ice in the nearshore Beaufort Sea, off the northern coast of Alaska in spring 1980. Over 90 % of the individuals found belonged to meiofaunal taxa. Although density within the ice was low compared to that typically found in sediments, it was comparable to the highest previous estimate from sea ice. Densities were lowest in samples taken in April and reached a maximum of around 50 000 individuals m-2 in June. Nematodes were numerically dominant, but copepods and turbellanans were also abundant. Population structures of the 2 most numerous copepod species were examined and found to differ greatly.Cyclopina gracilis appeared to reproduce continuously during the study; both adults and young copepodites were always present. The population of Harpacticus sp. initially consisted only of juveniles; the growth of a single cohort was observed. Harpacticus sp. mated in the ice, but gravid females did not appear during the sampling period. Although the ephemeral seasonal sea ice habitat would seem to be very different from typical habitats of benthic copepods, these 2 species did not appear to possess any special strategies for living there.

Dispersal of meiobenthic copepods was examined at 2 subtidal sites subject to moderate flow conditions, and in a laboratory experiment under conditions of no-flow. Small-scale spatial heterogeneity In settlement into containers filled with azoic sediment was found for several species in 3 field experiments. The species composition in the colonization containers was different from that observed in the surrounding sediments. Juvenile copepodites of several species were disproportionally abundant in the containers compared to the sediment, suggesting that juveniles are more hkely to disperse than adults. An additional field experiment was conducted using treatments varying in food level. Several species positively responded to microbially-enriched containers, probably by a decreased emigration rate compared to that in less enriched treatments. Results of a laboratory study were similar to that in the field food-level experiment. A model is presented that depicts the roles of active and passive aspects in dispersal at these study sites.