Ryer et al.: Depth distribution, habitat associations, and differential growth of Chionoecetes bciirdi 
267 
The observed difference in molt-stage frequency be- 
tween sites may be a product of differential emigra- 
tion or mortality. If predation is size-dependent, and 
predators are largely consuming smaller crabs, the 
size-frequency distribution of crabs at Womens could 
be skewed toward larger crabs by heavy predation. If 
this size-selective postsettlement process was in fact 
occurring, we would expect crab numbers to decline 
rapidly as a result of predation. Settlement was large- 
ly complete by July, but Womens was the only site to 
experience no population decline from July to August, 
indicating that predation was low at that site. Alterna- 
tively, differential migration may offer an explanation, 
if it is a natural progression for Tanner crabs to settle 
in shallow water and then migrate to deeper water. 
Unlike the other sites, Womens has a narrow entrance 
with a sill that rises up to a depth of approximately 11 
m. This narrow entrance could reduce offshore migra- 
tion of juveniles in Womens Bay. However, there is no 
structural hindrance to offshore migration at Kalsin, 
where crabs were also relatively large. If anything, we 
suspect that the shallow sill at Womens may block the 
offshore migration of larger crabs in the fall, winter, or 
spring because Womens was the only site that retained 
an appreciable number of age-1 crabs. 
Lastly, sites like Womens and Kalsin have finer sedi- 
ments, which are likely to accumulate organic carbon 
and support a denser infaunal community than the 
other sites. Therefore, crabs may have more or better 
food there. Results of preliminary lipid and essential 
fatty acid analysis that we performed on crabs from 
our study sites (Copeman and Ryer 5 ) indicate that 
crabs from Womens and Kalsin had higher levels of 
storage lipids and of diatom fatty acid markers than 
crabs from Pillar and Holiday. Overall higher levels 
of storage lipids and diatom-derived fatty acids have 
been associated with accelerated growth in larval Pa- 
cific cod (Copeman and Laurel, 2010) and juvenile red 
king crabs (Copeman et ah, 2012). 
We also observed that, for individual molt stages, 
crabs were larger at Womens and Kalsin than at Holi- 
day and Pillar. In a review of crustacean growth, Hart- 
noil (1982) concluded that, for many species, increases 
in the quantity or quality of food not only decreased the 
intermolt period but also increased the molt increment. 
However, this effect of food availability on growth may 
vary ontogenetically. Among larger juvenile blue crabs, 
growth has been correlated with higher food density 
(Seitz et ah, 2005), whereas, for smaller juveniles, this 
link has not been made (Long et al., 2011). In contrast, 
increases in temperature typically decrease both the 
intermolt period the molt increment, although Stoner 
et al. (2010) reported an increase in red king crab molt 
increments with increased temperature. A further un- 
derstanding of the relative role of temperature ver- 
sus food on growth of the Tanner crab will await con- 
5 Copeman, L. A., and C. H. Ryer. 2010. Unpubl. data. 
Alaska Fisheries Science Center, 2030 S. Marine Science Dr., 
Bldg. RSF951, Newport, OR 97365-5296. 
trolled laboratory experiments that manipulate these 
parameters. 
Whether juvenile Tanner crabs use habitats that 
are distinct from those occupied by adults remains 
unclear. The arguments presented here make the case 
that juveniles would fare better in shallow waters 
(depths <50 m). Further, we encountered few individ- 
uals larger than the C7 molt stage (carapace widths 
of 20-26 mm), indicating that crabs older than 2 or 
3 years of age are found in different, perhaps, deep- 
er habitats. In Glacier Bay, Alaska, Tanner crabs are 
segregated by ontogenetic stage; smaller, or juvenile, 
crabs are located at the heads of fjords, near glaciers, 
and adult, or larger crabs are more centrally locat- 
ed in fjords and in inlet areas (Nielsen et al., 2007). 
However, this segregation did not appear to be depth 
related, and the authors speculated that cannibalism, 
competition, predation, or differences in substrate 
preferences might be responsible. In Cook Inlet, Alas- 
ka, large individuals were found throughout the inlet, 
whereas juveniles with carapace widths <20 mm were 
concentrated in the inlet mouth at generally greater 
depths (Paul, 1982). This diversity of results indicates 
that the factors that control the distribution of age-0 
crabs can vary between areas. 
Although we do not yet know the full range of 
depths and habitats used by age-0 Tanner crabs, the 
settlement densities and patterns documented in this 
work indicate that relatively shallow waters ( <50 m) 
may constitute an important habitat for Tanner crabs 
around Kodiak Island, particularly because tempera- 
tures in these shallow waters can be expected to speed 
growth and shorten the number of years before crabs 
recruit to the fishable or reproductive population. 
Acknowledgments 
We thank M. Ottmar, S. Haines, and C. Sweitzer for as- 
sistance with logistics. R. Foy, E. Munk, P. Cummiski, 
and K. Swiney provided logistical support in Kodiak. 
We also wish to thank the captain of the FV Miss-O, T. 
Tripp, for valuable assistance on the water. Two anony- 
mous reviewers provided helpful comments on an early 
draft of this manuscript. This work was supported by 
2010 and 2011 Essential Fish Habitat funds from the 
NOAA Alaska Fisheries Science Center. 
Literature cited 
Baker, R., and M. Sheaves. 
2007. Shallow-water refuge paradigm: conflicting evi- 
dence from tethering experiments in a tropical estu- 
ary. Mar. Ecol. Prog. Ser. 349:13-22. 
Conover, W. J. 
1971. Practical nonparametric statistics, 462 p. John 
Wiley & Sons, Inc., New York. 
Copeman, L. A., and B. J. Laurel. 
2010. Experimental evidence of fatty acid limited growth 
