Lowe et a I.: Geographic variation in genetic and growth patterns of Pleurogrammus monopterygius 
51 1 
survey of Atka mackerel intended to help understand 
the level and pattern of reproductive isolation among 
areas thought to be reflected in the areal growth dif- 
ferences. Samples of Atka mackerel were collected dur- 
ing the spawning season when groups of this species 
are locally aggregated, presumably providing the best 
separation of potentially reproductively isolated groups. 
Our results show that Atka mackerel have above- 
average levels of genetic variation for a marine fish. 
The average heterozygosity per locus for Atka mack- 
erel is 0. 137, well above the average reported by Ward 
et al. (1994) for 57 species of marine fish (0.064 ± 
0.004). Assuming that the majority of the allozyme 
variation is selectively neutral (Kimura, 1968), we 
believe that large levels of genetic variation are most 
parsimoniously explained by large, historically stable 
populations. An alternative explanation is that large 
levels of variability reflect a response to inhabiting a 
heterogeneous habitat (Avise, 1994). Nonetheless, 30 
polymorphic loci provide substantial statistical power 
to assess the level of between-sample differentiation. 
Between-sample variation was extremely low 
among the four samples of Atka mackerel. In a study 
where electrophoretic data for marine, freshwater, 
and anadromous fishes were compared. Ward et al. 
(1994) calculated an average G gT ( roughly equiva- 
lent to F gT ) of 0.062 for 57 species of marine fish (com- 
pared with 0.22 for freshwater species of fish). The 
F gT for Atka mackerel (0.004) is far smaller than this 
average, providing evidence that a large amount of 
gene flow is occurring throughout the range repre- 
sented by these samples. Very low genetic distance 
values between samples were observed; the largest 
distance value between two samples was 0.00017, 
indicating little or no stock differentiation. Further- 
more, the nonsignificant Hardy- Weinberg test of the 
four samples pooled together did not indicate any 
between-sample heterogeneity. In light of these two 
genetic results, we can not reject the null hypothesis 
that our samples came from a single genetically ho- 
mogenous population of Atka mackerel. There was 
also no apparent gradual differentiation throughout 
the Aleutian Archipelago corresponding to the clinal 
geographic variation seen in the growth data. Stock 
delineations based on the age-length and weight- 
length relationships were not supported by the 
allozyme data. 
Concordance of 
electrophoretic and growth data sets 
A lack of congruence between genetic and life his- 
tory characteristics was also found for Pacific ocean 
perch by Seeb and Gunderson (1988). They analyzed 
data from the Washington coast to the Bering Sea. 
Stock delineations based on age structure, age-length 
relationships, and ages at maturity were not sup- 
ported by the allozyme data. Although they did not 
find clear genetic stock differentiation, they did find 
a cline of gene frequencies within the Gulf of Alaska 
and significant allele frequency differences between 
the extremes of the geographic range for some loci. 
There are at least three possible explanations for 
the lack of genetic stock differentiation for Atka 
mackerel in the Aleutian Islands. First, the genetic 
technique surveyed invariant gene loci when in fact 
genetic differences may exist among the sampled 
stocks. As in any genetic study, the absence of ge- 
netic differentiation does not preclude the possibil- 
ity that true genetic differences exist, and other ge- 
netic techniques may be employed to further exam- 
ine this possibility (Avise, 1994). Second, the species 
could perform a major spawning migration encom- 
passing the entire population, with mixing and 
spawning in one area. Third, separate spawning loca- 
tions are used but gene flow occurs among locations 
through the dispersal of pelagic larvae and juveniles. 
We consider the third explanation most likely given 
the growth differences seen in adult Atka mackerel. 
The Aleutian Archipelago encompasses several 
wide and deep straits that could form barriers to sig- 
nificant movement across the island chain for adult 
demersal fish residing over the continental shelf. The 
pelagic behavior of larval and juvenile fish and their 
presumed distribution in the open ocean habitat, 
however, would make them susceptible to wide dis- 
persal by currents, thus accounting for the genetic 
homogeneity among samples. A schematic diagram 
of the basic surface currents in the north Pacific is 
shown in Figure 6 (McAlister and Favorite, 1977). 
The Alaska stream flows westward throughout the 
Aleutian Archipelago, with significant northward 
flow to the Bering Sea through Amukta, Amchitka, 
and Buldir passes. The North Aleutian current flows 
eastward to Umnak Island providing a thorough 
mixing mechanism across the Aleutian Island chain. 
The currents could thus provide the mechanism for 
sufficient gene flow through the Aleutian chain to 
actually prevent genetic differentiation. It is pre- 
sumed that adults do not undertake large-scale mi- 
grations and assume a fairly localized existence, 
which might thus account for the differences in 
length, age, and weight comparisons of the adults. 
Observed spawning areas in the Aleutian, Shumagin, 
and Commander Islands, which have been histori- 
cally referenced in the literature (Turner, 1886; 
Rutenberg, 1962), are thus presumed to attract 
nearby resident schools that migrate inshore. 
Although there are no major currents flowing east- 
ward from the Aleutian Islands to the Gulf of Alaska, 
