THREESPINE STICKLEBACK OF KODIAK ISLAND 



541 



When the female is approaching the time of 

 spawning, the belly becomes distended visibly. 

 In the early part of the year, before the breeding 

 characters are evident, the two sexes look much 

 alike. 



Morphological Variation 



Throughout its geographical range, the species 

 Gasterosteus aculeatus is subject to a large 

 amount of morphological variation. There are 

 minor differences in color and color pattern and 

 in body shape. More pronounced variations oc- 

 cur in numbers of vertebrae, fin rays, and lateral 

 bony plates. 



Within the species, the number of lateral plates 

 varies from none to about 35. Most of the long 

 list of specific, subspecific, and racial names 

 which have been used have to do with the degree 

 of armature. 



It has long been recognized that populations 

 which are year-round inhabitants of fresh waters 

 have few plates, while those that live in salt 

 water, and enter fresh water only for spawning, 

 are many-plated. Bertin (1925) made an analyti- 

 cal study of populations of threespine stickleback 

 throughout Europe and came to the conclusion 

 that the number of bony plates is directly corre- 

 lated with the salinity of the environment, and 

 that a continuous gradation occurs from water 

 of high salinity to fresh water. 



However, Heuts (1947, 1947a) presents strong 

 evidence that such a clinal gradation does not 

 exist. The European stickleback populations are, 

 instead, of two types, one of which shows a strong 

 mode at about 5 to 7 plates, the other having a 

 mode at around 32 plates. Within a given popu- 

 lation, individuals with an intermediate number 

 of plates may occur, but there are no populations 

 which consist only or largely of fish with inter- 

 mediate plate numbers. Therefore, Heuts con- 

 cludes that individuals with an intermediate num- 

 ber indicate only a skewed distribution curve of 

 one type or the other and not a genotypical inter- 

 mediacy. 



On the Atlantic coast of North America, both 

 the form with many plates and that with few are 

 present (Bigelow and Schroeder, 1953). As in 

 Europe, the fish with strong armature are to be 

 found in salt water, while the fresh-water popu- 

 lations generally have few plates. The same 



489027 O — 59 2 



situation occurs also along the Pacific coast of 

 North America. 



The geographic pattern of variability in Alas- 

 kan waters has not been worked out thoroughly. 

 Marine collections have been few, and in many 

 fresh waters the species has been recorded with- 

 out reference to which of the two forms was 

 present. It is likely that there are N many bodies 

 of fresh water, particularly those close to the 

 sea, in which both the many-plated and the par- 

 tially plated forms may occur, at least for part 

 of the year. An instance is Chignik Lake on 

 the south side of the Alaska Peninsula (private 

 communication from Fredrik Thorsteinson of the 

 Fisheries Research Institute, 1956). 



There is very little, if any, migration of the 

 Karluk and Bare Lakes sticklebacks to and from 

 the ocean. No substantial numbers of stickle- 

 backs have been observed in movement in Bare 

 Creek or upper Karluk River. There is no evi- 

 dence of a sea-run stock of sticklebacks in either 

 lake. The number of lateral plates on all speci- 

 mens examined does not exceed nine. 



Plates and fin rays were counted on a total of 

 200 fish from each of the 2 lakes. The total for 

 each lake was made up of 4 subsamples of 50 

 fish each, taken in different years (table 1). 



Each subsample probably included individuals 

 of at least 2 age groups; therefore, there is con- 

 siderable overlapping of year classes. However, 

 as there is little if any exchange of stock with 

 outside populations, the genetic makeup of the 

 population of the lake should be about the same 

 from one year to the next. 



Table 1. — Subsamples of fish taken in different years from 

 Karluk and Bare Lakes 



An additional sample consisted of 50 specimens 

 from O'Malley River, a tributary to Karluk 

 Lake. This sample was taken in 1951. 



The lateral plates were counted on the left side 

 of the fish under low-power magnification. The 

 point of a scalpel was used to lift the edge of 

 each plate, so that none would be missed in 

 counting. The first (most anterior) plate is 



