Phenetic Variation and Habitat Differences in Gasterosteus Populations of the Queen Charlotte Islands

Abstract

Moodie, G. E. E., and T. E. Reimchen (Department of Biology, University of Winnipeg, Winnipeg, Manitoba, R3B 2E9, Canada) 1976. Phenetic variation and habitat differences in Gasterosteus populations of the Queen Charlotte Islands. Syst. Zool. 25:49-61.-Lateral plate polymorphism and asymmetry, gill raker number and asymmetry, body size and certain other traits were compared in 25 Gasterosteus aculeatus populations. Fluctuating asymmetry and mean number of lateral plates as well as body size and proportions were correlated with the presence of predatory fish. In the absence of other fish species, gill raker number was correlated with lake area. Fluctuating asymmetry of the lateral plates and gill rakers did not appear to decrease with age; in three populations the converse occurred. Directional asymmetry was present in six populations. Fluctuating asymmetry in these populations does not seem to conform well to current concepts concerning developmental stability and asymmetry. [Gasterosteus; phenetics; habitat variation; predation; asymmetry.] Several recent studies have documented the great variability of populations of the threespine stickleback (Gasterosteus aculeatus L.) and the adaptive value of certain morphological traits (Semler, 1971; Moodie, 1972a, b; Hagen and Gilbertson, 1972). Much, however, remains to be learned about the adaptive value of such things as lateral plate polymorphism (low, partial or complete), plate numbers and symmetry, body size and proportions. In addition, little is known of how the phenotype is influenced by selective agents other than predation and salinity. In the Queen Charlotte Islands there are numerous fresh water populations of G. aculeatus. Although they are within limited distance of each other, these populations show a range of variation (Fig. 1) practically matching the limits known for this widely distributed species. On these islands, populations may be isolated in small closed lakes or be in large lakes connected to the nearby sea. Some lakes contain only G.aculeatus: others contain G. aculeatus as well as up to four potential (fish) predators and competitors. The purpose of this study was to determine whether or not phenetic variation was correlated with competition, isolation and habitat or population size as well as predation. Most of these factors had to be measured indirectly. Most samples came from Graham Island (530 49' N, 1320 02' W). Two collections (Awun and Ain Lakes) were obtained in 1929; the others were collected by ourselves between 1966 and 1972. Most of the sites are officially named. To others we applied names which are enclosed in quotation marks when they first appear below. Sticklebacks were usually captured with seines, but a few lakes could be sampled only by using unbaited minnow traps. Difficult conditions in Yakoun and Clearwater Lakes forced us to utilize fish succumbing to post-reproductive mortality. (It is possible such samples do not represent the portion of the population one would obtain by seining. However, in Drizzle Lake we compared a sample taken by seining with one composed of victims of post-reproductive mortality. There were no significant differences in the means or variances of the plate and raker counts; sample sizes were 37 and 308.) Gillnets were set wherever possible to help determine the presence or absence of larger species. Many of the lakes can be visited only by aircraft or on foot through dense vegetation. This created two problems. Firstly, our