FISHERY BULLETIN: VOL. 85, NO. 3 



the female's ovaries mature (Hirano 1935), 

 leaving the swollen membrane in place. The 

 membrane eventually becomes flaccid, perhaps to 

 allow egg extrusion. As the spawning period 

 approaches, the spermatophores gradually break 

 up and release spermatozoa. Fertilization occurs, 

 either as the eggs are extruded (Kawakami 1934; 

 Yoshida 1940) or afterwards, perhaps by con- 

 traction of the sperm sac and expulsion of the 

 spermatozoa over the eggs (Sakurai et al. 

 1972). Our estimates of the number of external 

 embryos carried by EBS hair crabs are in accord- 

 ance with Sakurai et al. (1972, see Literature 

 Review). 



Molting and Growth 



Molting frequency varies with age and sex of 

 the crab. Because the percentage of recently 

 molted male crabs declined with increasing size, 

 the NMFS data appears to support Yamamoto 

 (1971), who indicated that molting frequency of 

 males decreases as size increases. Yoshida (1941) 

 claimed that female hair crab molt every other 

 year. Our data neither support nor refute this. 

 However, old and new shell conditions are diffi- 

 cult to distinguish in hair crab, because shell dis- 

 colorations and epibiota are uncommon. Thus, it 

 is difficult to determine when a particular crab 

 last molted. A higher frequency of molting in 

 February (based on the 1983 data, Table 38) for 

 EBS hair crab tends to support the conclusions of 

 Abe (19841'') and Sakurai et al.(1972) (Fig. 3) that 



hair crab molt and mate in winter, although the 

 1985 data did not. The second period of molting 

 exhibited by female crabs in July tends to support 

 the contention by Sakurai et al. that multiparous 

 females molt and breed from August to Novem- 

 ber; however, females exhibited a low frequency 

 of molting in August. Crabs of all sizes molted in 

 both periods. 



The effect of temperature on molting periodic- 

 ity of hair crab is unknown. Mean bottom temper- 

 atures that were 2.5°C warmer in February 1985 

 than February 1983 for stations sampled for hair 

 crab in both those years could have affected the 

 onset of molting in 1985. Whereas in 1983, 22% of 

 the hair crab caught were molting or in soft-shell 

 condition, in 1985 only 1% were. 



Abe (1982) determined the mean carapace 

 length of 14 male and 13 female hair crab postlar- 

 val instars (Table 7) through length-frequency 

 analysis of 10,547 individuals. For the first 9 in- 

 stars, there was no difference in mean carapace 

 length between males and females. However, 

 after sexual maturity (i.e., sizes greater than 

 about 55 mm RL or 50 mm CL) females showed 

 less gi"owth per molt than males, although growth 

 increments decreased with age for both sexes. 

 Abe plotted postmolt carapace lengths (L„^^i) 



I'^Abe, K. 1984. Reproductive cycle of hair crab. Unpubl. 

 manuscr. Chief, Fisheries Resources Division, Hokkaido Wak- 

 kanai Fisheries Experimental Station, 5-4 Horai, 4-chome, 

 Wakkanai, Hokkaido, 097, Japan. Presented at the hair crab 

 conference, Yoichicho, Japan, January 31, 1984. 



Table 7. — Mean carapace length (mm RL) for each postlarval instar of male 

 and female hair crab, Erimacrus isenbeckii, (from Abe 1982), and calculated 

 mean growth increment for the next molt. 



'Lengths of females were same as males for instars 1-9 



542 



