Trematode infection was studied in sympatric populations of the periwinkles Ljttorina saxatilis and L. obtusata in 2 regions of Kandalaksha Bay of the White Sea to assess host-parasite interactions at the population level. Twenty-seven spatially separated populations were each surveyed in 1984-1994; 2 heavily infected populations were investigated annually over a 16 yr period. Ten trematode species were found in the periwinkle populations. The closest association in spatial distribution and temporal dynamics was observed between 3 ecologically and morphologically similar trematodes of the 'pygmaeus' group: Microphalluspjriformes, M. pygmaeus and M. pseudopygmaeus. For these 3 species, the prevalences were closely associated in the 2 host species when spatially separated sites from the 2 studied regions were considered, while in the 2 populations studied over the 16 yr period, a correlation was only observed between the infection levels of L. saxatilis and L. obtusata by either M. piriformes and immature microphallids. Likewise, within each host species, significant correlations were revealed between the prevalence of the different microphallids of the 'pygmaeus' groups. However, they were fewer and weaker when the long-term dynamics of infection in the 2 heavily infected populations were considered. Most other trematodes did not show significant association in prevalence either within or between the 2 host species on spatial and temporal scales.

Five populations of the rough winkle Littorina saxatilis (Olivi) with different levels of trematode infection were investigated in the White Sea, NW Russia: 2 control populations N1 and N2 (ca 5'K) infection), 1 moderately infected population M (ca 10 to 15% infection) and 2 heavily infected populations H1 and H2 (50 to 80% infection). Trematode prevalence was the lowest in snails collected from macrophytes, intermediate on gravel and highest on stone surfaces at all collection sites. The distribution of infection in different age groups of snails was random in the control populations, but prevalence increased with age in moderately and heavily infected populations. Demographic structure was variable. The following parameters were typical for all populations regardless of infection level: (1) total sex ratios were close to 1:l or biased slightly towards females; (2) the relative abundance of males decreased with the age of snails; (3) no sex-specific ditterences in infection prevalence were found in any population, and (4) females became mature at 3 to 4 yr However, the following demographic parameters showed significant differences that were correlated with trematode ~nfection: (1) fewer juveniles were found in populations H1 and H2, probably a result of reduced recruitment due to severe infection; and (2) fecundity of uninfected females increased with the increase in trematode infection at the collection sites and was the lowest in populations N1 and N2, intermediate in population M and the highest in population HI. These results suggest an adaptation that ensures periwinkle survival under conditions of heavy infection pressure.