Laman et al . : Correlating environmental and biogenic factors with abundance and distribution of Sebastes alutus in Alaska 
275 
Figure 2 
Nominal morphogroups assigned to field identified sponges on the basis of mor- 
photypes from Boury-Esnault and Rtitzler (1997) and Bell and Barnes (2000): 
(A) arborescent, (B) clavate, (C) encrusting, (D) flabellate, (E) flagelliform, (F) 
globular, (G) globular-papillate (e.g., Weberella bursa , an image of which rep- 
resents this morphogroup in this figure), (H) massive (e.g., Halichondria spp.), 
(I) ovate, (J) papillate, (K) pedunculate, (L) repent, (M) stipitate, (N) tubular, 
and (O) vase. (P) The Porifera unidentified group is represented by an identifi- 
able sponge (e.g., Plakina tanaga ) that did not fit in to the predetermined mor- 
phogroups. Globular-papillate and vase morphogroups are modified from those 
of Boury-Esnault and Rutzler (1997) to incorporate local faunal variants; for 
example, the vase morphogroup is a composite of the turbinate, caliculate, and 
infundibuliform morphotypes from Boury-Esnault and Rutzler (1997). 
majority of these taxa (85) are 
primnoids, gorgonians, stony cor- 
als, and octocorals, all of which 
provide vertical relief over the 
trawlable habitats we sampled, 
and a few are among the tall- 
est biogenic structures that oc- 
cur in the Aleutian Islands (e.g., 
Primnoa willeyi can reach 3 m 
in height). Similarly, we included 
all bryozoans reported from our 
Aleutian Islands bottom trawl 
catches (~25 taxa) as a compos- 
ite predictor variable because of 
the biogenic structure they may 
provide. 
Our standard trawl net with 
its rubber bobbin roller gear 
(Stauffer, 2004) is not optimized 
for collecting attached epiben- 
thic invertebrates. Therefore, all 
of the biogenic structures used 
to parameterize our models were 
coded as presence or absence be- 
fore analyses. We also considered 
that trawl nets sampled sponges 
differentially (see Wassenberg et 
al., 2002), integrating their ac- 
tual distribution over trawlable 
bottom for the duration of the 
trawl haul. Therefore, we consid- 
er the coding of biogenic struc- 
tures in our trawl catches as 
presence-absence factors to be 
a conservative measure of their 
occurrence. 
Fish data 
Catch per unit of effort (CPUE) 
was measured in kilograms per 
hectare and calculated for ju- 
venile and adult Pacific ocean 
perch with the area-swept meth- 
od (Wakabayashi et al., 1985). 
Length composition, estimated 
for each trawl by measuring a 
selected subset of up to 200 fork 
lengths (FL) of Pacific ocean 
perch from the catch, was used 
to proportionally allocate the 
catch of Pacific ocean perch on 
the basis of length at first matu- 
rity (Paraketsov, 1963; Chikuni, 
1975) into juveniles (FL<250 
mm) and adults (FL>250 mm) 
for each trawl tow. Nonzero 
CPUE values (conditional abun- 
dance) were log-transformed and 
then placed on a common propor- 
