368 
Fishery Bulletin 108(4) 
seminal vesicles tightly coiled and testes lobed; and 
stage 4 = similar to stage 3, but with semen present 
in the distal portion of the seminal vesicle. Individuals 
with reproductive tracts at stages 1 and 2 are regarded 
as sexually immature, whereas those at stages 3 and 4 
and, in females, also at stage 5, can reproduce or have 
reproduced and are therefore considered mature. When 
we alternatively employed full clasper calcification as 
the indicator of maturity, we considered that males with 
noncalcified and partially calcified claspers were sexu- 
ally immature, and those with fully calcified claspers 
were mature because they have the ability to copulate. 
The probability (P) that an individual is mature was 
assumed to be a logistic function of its length (L): 
P = jl + exp[-(a + /?L)]} , (1) 
where a and /3 are parameters that determine the loca- 
tion and shape of the logistic curve. 
The parameters a and ft were transformed to the 
lengths L 50 (TL 50 or DL 50 ) and L 95 (TL 95 or DL 95 ) by 
which 50% and 95% of fish have attained maturity, 
respectively, using the equations 
^50 ~ Otl P, 
(2) 
L 95 = D°ge (19) -«]/£■ 
(3) 
The equation for the probability that a fish is mature 
thus becomes 
P = {l + exp[-log e (19)(L-L 50 )/(L 95 -L 50 )]} \ (4) 
Logistic relationships of the above form were derived 
for females and males of A. vincentiana, S. australis , 
and M. australis by using gonadal maturity status, i.e., 
>stage 3, as the criterion for maturity and, in addition, 
for males, employing full clasper calcification as the 
criterion for maturity. Logistic regression analysis was 
used to fit these logistic curves by using Solver in Excel 
(Microsoft, Redmond, WA) to maximize the log-likeli- 
hood. We used likelihood-ratio tests (Cerrato, 1990) to 
compare the L 50 of the females and males of each spe- 
cies at maturity using gonadal status as the criterion 
for maturity and to compare the L 50 derived for males 
using both gonadal and clasper calcification status as 
the criteria for maturity. For the likelihood-ratio tests, 
the hypothesis that the data for both the females and 
males of each species could be described by a common 
logistic curve was rejected at the a=0.05 level of signifi- 
cance if the test statistic, calculated as twice the dif- 
ference between the log-likelihoods obtained by fitting 
maturity curves with a common value of L 50 for both 
sexes and by fitting separate maturity curves for each 
sex, exceeded ^ 2 „(q), where q is the difference between 
the numbers of parameters in the two approaches. Note 
that the L 50 of females and males of H. portusjacksoni 
at maturity had been determined previously with this 
approach (Jones et al., 2008). WinBUGS (Lunn et al., 
2000) was also used to fit the logistic curves to the 
maturity-at-length data for the females and males of 
each species and to calculate the proportion that were 
mature at length, thereby enabling derivation of the 
upper and lower confidence intervals for the L 50 and 
L 95 values and for the proportion of mature males and 
females in each 50-mm length class (see Jones et al. 
[2008] for further details of this WinBUGS analysis). 
Results 
Species compositions by fishing method 
A total of 4820 individual elasmobranchs, representing 
10 families and 22 species of sharks and 7 families and 
11 species of rays, were recorded during regular onboard 
examinations of the catches of commercial trawl, gillnet, 
and longline vessels operating off the southwestern coast 
of Australia (Tables 1-3). 
The 2986 elasmobranchs caught by prawn and scal- 
lop trawling were dominated by rays, which comprised 
10 of the 14 species and contributed 87% to the total 
elasmobranch catch (Table 1). The species of a single 
family of rays, the Urolophidae, comprising four spe- 
cies and two genera, contributed as much as 67% to 
the total trawl catch of elasmobranchs. The two species 
of shark ( H . portusjacksoni and S. australis) and the 
two species of ray (A. vincentiana and M. australis), 
whose biological characteristics were determined (see 
later), each contributed between 4.5% and 8% to the 
total number of elasmobranchs caught by trawling and 
collectively as much as 25% (Table 1). Two species of 
shark, M. antarcticus and C. brevipinna, which were 
caught in very small numbers, were retained and thus 
constituted byproduct. 
Gillnet catches yielded 1260 elasmobranchs, repre- 
senting 19 species of shark and 6 species of ray, with 
sharks contributing 96% to the total catch of elasmo- 
branchs (Table 2). The most abundant species in the 
gillnet catches was a targeted shark, Carcharhinus 
obscurus, which contributed more than a third to the 
total elasmobranch catch. The other two targeted spe- 
cies, Mustelus antarcticus and Furgaleus macki, which 
were dominated by females, ranked third and fifth in 
terms of abundance, respectively, and contributed an 
additional 14.2% and 6.5%, respectively (Table 2). The 
second most numerous species, however, was the shark 
H. portusjacksoni, a bycatch species, which constituted 
one-fifth of the total catch. None of the six species of 
ray caught by gillnetting was abundant in the catches 
obtained by this method. The byproduct and bycatch 
species contributed 18.7% and 25.7%, respectively, to 
the total gillnet catch. Thirteen of these species, which 
contributed more than one third to the total number 
of individuals of elasmobranchs caught, were always 
discarded as bycatch. 
The 22 species of elasmobranch caught by longlining 
were dominated by one of the three targeted species, M. 
antarcticus, which made up 63% of the total catch of 574 
individual elasmobranchs (Table 3). The next four most 
