Most species with lengthy migrations display some degree of natal homing; some (e.g. migratory birds and anadromous salmonids) show spectacular feats of homing. However, studies of the sea lamprey (Petromyzon marinus) indicate that this anadromous species locates spawning habitat based on pheromonal cues from larvae rather than through philopatry. Previous genetic studies in the anadromous Pacific lamprey (Entosphenus tridentatus) have both supported and rejected the hypothesis of natal homing. To resolve this, we used nine microsatellite loci to examine the population structure in 965 Pacific lamprey from 20 locations from central British Columbia to southern California and supplemented this analysis with mitochondrial DNA restriction fragment length polymorphism analysis on a subset of 530 lamprey. Microsatellite analysis revealed (i) relatively low but often statistically significant genetic differentiation among locations (97% pairwise F(ST) values were <0.04 but 73.7% were significant); and (ii) weak but significant isolation by distance (r(2) = 0.0565, P = 0.0450) but no geographic clustering of samples. The few moderate F(ST) values involved comparisons with sites that were geographically distant or far upstream. The mtDNA analysis--although providing less resolution among sites (only 4.7%F(ST) values were significant)--was broadly consistent with the microsatellite results: (i) the southernmost site and some sites tributary to the Salish Sea were genetically distinct; and (ii) southern sites showed higher haplotype and private haplotype richness. These results are inconsistent with philopatry, suggesting that anadromous lampreys are unusual among species with long migrations, but suggest that limited dispersal at sea precludes panmixia in this species.

Phylogenetic structure of four Lampetra species from the Pacific drainage of North America (western brook lamprey Lampetra richardsoni, Pacific brook lamprey Lampetra pacifica, river lamprey Lampetra ayresii and Kern brook lamprey Lampetra hubbsi) and unidentified Lampetra specimens (referred to as Lampetra sp.) from 36 locations was estimated using the mitochondrial cytochrome b gene. Maximum parsimony and Bayesian inferences did not correspond with any taxonomic scheme proposed to date. Rather, although L. richardsoni (from Alaska to California) and L. ayresii (from British Columbia to California) together constituted a well-supported clade distinct from several genetically divergent Lampetra populations in Oregon and California, these two species were not reciprocally monophyletic. The genetically divergent populations included L. pacifica (from the Columbia River basin) and L. hubbsi (from the Kern River basin) and four Lampetra sp. populations in Oregon (Siuslaw River and Fourmile Creek) and California (Kelsey and Mark West Creeks). These four Lampetra sp. populations showed genetic divergence between 2.3 and 5.7% from any known species (and up to 8.0% from each other), and may represent morphologically cryptic and thus previously undescribed species. A fifth population (from Paynes Creek, California) may represent a range extension of L. hubbsi into the Upper Sacramento River.

Mitochondrial DNA variation among 1246 individuals of Pacific lamprey (Entosphenus tridentatus) from 81 populations spanning 2600 km from the Skeena River, British Columbia, to the Ventura River, California, was surveyed using five restriction enzymes. A total of 29 composite haplotypes was detected in two gene fragments (ND2 and ND5). The three most common haplotypes, occurring in 91% of all samples, were present at similar frequencies in all regions. Samples were divided into six biogeographic regions based on sample distribution and geographical landmarks to assess geographic genetic structure. Analysis of molecular variance indicated that 99% of the genetic variation was explained by variability within drainages. The lack of geographical population structure is likely related to a life-history pattern that includes a prolonged larval freshwater stage, migration to oceanic feeding and return to fresh water to spawn. The lack of strong natal homing apparently promotes gene flow among drainages and regions.