MANN: (iONADAI, DKVKLOPMKNT IN ARCTICA ISLANDICA 



nadal condition is subject to the following in- 

 herent inadequacies. First, a continuous game- 

 togenic process is being described in discrete 

 stages. Second, it is difficult to consistently ob- 

 tain a midventral section that is a representative 

 mean of the cline of gonadal developmental 

 stages within one animal. Third, sample collec- 

 tions were relatively infrequent considering the 

 small differential in bottom temperatures be- 

 tween the stations and the comparatively high 

 rate of change of bottom temperature during the 

 summer months (Fig. 1). It is probable that these 

 three factors effectively combined to mask any 

 depth-dependent difference in gonadal develop- 

 ment. 



Morphologically ripe specimens were present 

 from March through October, but predominated 

 from May through September. Although no mor- 

 phologically ripe specimens were found in De- 

 cember or January — in contrast to the data of 

 Loosanoff (1953) — the presence of partially 

 spawned animals at this time, followed by the 

 first appearance of early gametogenic stages in 

 February and March, supports Loosanoffs hy- 

 pothesis that no significant "resting" or "indif- 

 ferent" period occurs in the annual gametogenic 

 cycle. The sequential ripening of gonadal folli- 

 cles from the hinge (dorsally) towards the foot 

 (ventrally), in a manner similar to that described 

 for Mija arenaria by Coe and Turner (1938), was 

 not described by Loosanoff (1953). No examples 

 of the "atypical" sperm development described 

 by Loosanoff ( 1953) were observed in the present 

 study. The present data are, however, not in com- 

 plete agreement with the recent statement in 

 Thompson, Jones, and Dreibelbis (1980), with 

 respect to A. islandica, that "All individuals 

 spawned once and once only in each of the two 

 years studied." Jones (1980) is quoted as the 

 source of documentation substantiating this 

 statement. This is somewhat surprising in that, 

 like the present study, Jones (1981) found some 

 sequential gonadal development and the pres- 

 ence of a large proportion of partially spent indi- 

 viduals of both sexes in samples collected in the 

 late summer and fall months. Both of these obser- 

 vations support the conclusion that individual 

 specimens spawn at least once per annual repro- 

 ductive cycle. 



It is relevant to speculate on the nature of the 

 proximal stimuli (sensu Baker 1938) of gameto- 

 genesis and spawning in A. islandica, given the 

 present physical and biological data, based on 

 the extensive discussion of the subject by Baker 



(1938) and Giese and Pearse (1974). Arctica 

 islandica initiated gametogenesis in February 

 when water temperature is lowest. This is in con- 

 trast to the more intensively studied intertidal 

 species which either cease gametogenesis during 

 the period of lowest temperature, e.g., Mytilus 

 edulis in Northern Europe (Chipperfield 1953), 

 or initiate gametogenesis only with rising water 

 temperatures and at the time of the phytoplank- 

 ton spring bloom, e.g., Ostrea edulis and Crassos- 

 trea gigas (Walne and Mann 1975). Ansell (1974), 

 Gabbott (1975), and Mann (1979a, b) found that 

 the initiation of gametogenesis in bivalves is 

 often preceded by a period of accumulation of 

 carbohydrate reserves which are subsequently 

 used as a predominant respiratory substrate 

 during gametogenesis and that this period of 

 accumulation usually coincides with a period of 

 high primary productivity and food availability. 

 The author can find no data on seasonal phyto- 

 plankton productivity for the region immediately 

 east and south of Block Island; however, substan- 

 tial data are available for the lower Narragan- 

 sett Bay (Hitchcock and Smayda 1977; Pratt 

 1965; Smayda 1957), Block Island Sound (Riley 

 1952b), and Long Island South (Smayda 1976). 

 Lower Narragansett Bay is characterized by an 

 intense winter (January to March) diatom bloom 

 and a smaller, late summer to autumn (July to 

 October) bloom (Smayda 1976). Block Island and 

 Long Island Sounds exhibit a more classical 

 spring and autumn bloom (Hitchcock and 

 Smayda 1977; Smayda 1976). It is not unreason- 

 able to suggest that phytoplankton from the 

 autumn and winter blooms in Narragansett Bay 

 are washed into Block Island Sound (Riley 

 1952a), and that, because of the vertically well- 

 mixed nature of the water column at this time, 

 both they and the phytoplankton from the classi- 

 cal spring and autumn blooms become available 

 to A. islandica. Phytoplankton was probably 

 made available by wind and storm events similar 

 to those which effect the mixing and distribu- 

 tion of chlorophyll in the New York Bight and 

 Georges Bank, as described by Walsh et al. 

 (1978). In turn these blooms may be a potential 

 food source for storage metabolism in A. islan- 

 dica during the late fall, winter, and early spring 

 months prior to and coincident with the initial 

 stages of gametogenesis. Specimens collected 

 throughout this period had both bright green 

 digestive glands and a well-developed crystalline 

 style indicative of active feeding on phytoplank- 

 ton. 



323 



