672 



Fishery Bulletin 91(4). 1993 



continues through the next winter, when eclosion oc- 

 curs. Second, females separated into two groups on 

 the basis of brood development, shell condition, and 

 maturity of ovaries (Table 1; Figs. 5 and 6). In general, 

 females that carried eggs in ES IV and ES V were in 

 advanced molt stages (AIM and PRM) and had fully 

 developed ovaries; by contrast, females with eggs in 

 ES I, ES II, and ES III were in EIM or MIM stages 

 and had small ovaries. Additionally, the presence of 

 oocytes in early vitellogenesis in the recently spawned 

 ovary denotes that oogenesis started before spawning 

 and thus lasts more than two years. In Lithodes 

 santolla, oogenesis lasts 24 months (Vinuesa, unpubl. 

 data) while embryogenesis lasts 11 months (Vinuesa, 

 1984). 



The asynchronous embryonic development of P. 

 granulosa is a novel feature among lithodid crabs. Our 

 data suggest that some of the eggs within a single 

 clutch develop in 12-14 months and hatch in early 

 summer, while the remainder of the eggs com- 

 plete their development during the next 10 

 months. We discount the possibility of a second 

 mating and egg extrusion without molting be- 

 cause 1) there is no seminal receptacle, and chitin- 

 ous plates cover the gonopores during intermolt, 

 and 2) the ovaries of females with asynchronously 

 developing eggs were in the first year of their 

 cycle. There is clearly a regional effect involved 

 in asynchronous development since females from 

 Ushuaia had more uniformly developed broods. 

 However, with data presently available, we can- 

 not speculate on the causes of this phenomenon. 



Since gonadal maturity (presence of gametes) 

 does not necessarily imply morphometric matu- 

 rity (crabs with differentiated secondary sexual 



characters), these two terms should be used to define 

 different events in life history that may or may not 

 occur simultaneously. We consider that these concepts 

 are applicable to P. granulosa since different sizes at 

 maturity were calculated by using different features 

 (Figs. 2 and 3). Size at maturity for P. granulosa tends 

 to increase with increasing latitude (Table 3). This re- 

 lationship has been noted for L. santolla as well 

 (Vinuesa, 1985). By contrast, in the Northern Pacific, 

 size at maturity of lithodid crabs decreases with in- 

 creasing latitude (Jewett et al., 1985; Somerton and 

 Otto, 1986; Blau, 1990; Otto et al., 1990). Causes of 

 geographical variation in size at maturity are still un- 

 known, but environmental conditions such as bottom 

 temperature may be a factor. 



Traps may give biased samples because 1) they se- 

 lect for larger animals (Miller, 1990, but see Blackburn 

 et al., 1990), 2) small crabs and females are excluded 



