Clarke et al.: Elasmobranch bycatch from the shrimp trawl fishery along the Pacific coast of Costa Rica 
7 
o 
<u 
Q. 
3 - 
1 ■ 
i*. 
i 
JUS 
50 100 150 200 250 300 
Depth (M) 
Figure 2 
Spearman’s rank correlation coefficient for the rela- 
tionship between depth and species richness for elas- 
mobranchs caught during surveys conducted along the 
Pacific coast of Costa Rica during 2010-2012. Relation- 
ship is between 20-m depth categories and average 
number of species per trawl haul. Error bars indicate 
standard errors of the means. 
vian torpedo (Table 4). Conversely, depth was a factor 
that significantly influenced maturity ratios of brown 
smoothhound, rasptail skate, and giant electric ray. Al- 
though mature brown smoothhound were more abun- 
dant at all depths, the proportion of immature indi- 
viduals peaked at the depths of 50-100 m (Fig. 4). For 
the rasptail skate and giant electric ray, proportions of 
immature individuals were higher in the shallow lim- 
its of these species’ depth ranges (50-100 m and <50 
m, respectively) (Fig. 4). Maturity ratios in the catch 
of Panamic stingray and thorny stingray varied sig- 
nificantly among diel periods; proportionally more im- 
mature individuals were caught during the day than 
during the night (Fig. 5). A high proportion (>50%) 
of immature Peruvian torpedo was recorded across 
all depths and diel periods. For the witch guitarfish, 
a high proportion (73%) of immature individuals was 
found at shallow depths (<50 m), and a high proportion 
(57%) of mature individuals was found at depths of 50- 
100 m. A higher abundance of mature round stingrays 
( Ui'otrygon spp.) was observed at shallow depths than 
at other depths. 
Elasmobranch assemblage 
The elasmobranch assemblage varied significantly 
among depths (ANOSIM 77=0.710, P=0.001). The elas- 
mobranch assemblage in shallow waters differed from 
the assemblages in intermediate-depth (P=0.792, 
P=0.001) and deep (P=0.934, P=0.001) waters. The 
shallow-water assemblage was characterized by the 
dominance of Panamic stingray (0.00-138.70 individu- 
als/hour); the remaining species were less abundant 
(0.01-2.15 individuals/hour). The deepwater assem- 
blage was composed of rasptail skate, brown smooth- 
hound, witch guitarfish, Peruvian torpedo, prickly 
shark, and Pacific angel shark, of which the first 2 
species were most abundant. The elasmobranch assem- 
blage in intermediate-depth waters was characterized 
by a combination of both shallow-water and deepwater 
species. Differences in the elasmobranch assemblage 
between deep and intermediate-depth waters were 
the smallest (77=0.474, P=0.001). A SIMPER analysis 
revealed that Panamic stingray, rasptail skate, brown 
smoothhound, witch guitarfish, and Peruvian torpedo 
were also responsible for 49.8-75.8% of the differences 
in the elasmobranch assemblage between depths. 
In the RDA applied to deepwater survey data, depth, 
latitude, year, and season represented 13% of the vari- 
ance in species data (Fig. 6A). The biplot that resulted 
from this analysis displays 97.2% of this variability in 
its first 2 axes, and all canonical axes were significant 
(Fig. 6A; 77=4.224, P=0.002). This RDA biplot shows the 
brown smoothhound, rasptail skate, and Peruvian tor- 
pedo as a group. This group was negatively correlated 
with depth and strongly related to the years 2008 and 
2009. The species in this group were separated mainly 
by latitude: the brown smoothhound was slightly more 
associated with southern latitudes, and the Peruvian 
torpedo was slightly more associated with northern 
latitudes. The prickly shark was placed separately and 
presented a strong positive correlation with depth. 
The RDA for monitoring survey data explained 12% 
of the variability between environmental variables and 
species data (Fig. 6B). A biplot captures 91.4% of this 
variability, and all canonical axes were significant (Fig. 
6B; F= 3.222, P=0.002). Although all species are grouped 
quite closely together in this biplot, the strongest as- 
sociations were between the witch guitarfish and the 
rasptail skate. All species had negative correlations 
with depth and positive associations with the diel pe- 
riod of 0600-1800 and the rainy season. As in the deep- 
water RDA, the brown smoothhound was more common 
at the southern limit of the latitudinal range, and the 
Peruvian torpedo and witch guitarfish were associated 
with the northern limits of the latitudinal range. 
In the RDA applied to commercial survey data, 
depth, latitude, diel period, season, and year explained 
10% of the variance in species data (Fig. 60. The bi- 
plot that resulted from this analysis represents 76.9% 
of this variance, and all canonical axes were significant 
(Fig. 6C; F=2.772, P=0.002). This RDA biplot separates 
2 groups of species, mainly according to depth and lati- 
tude. The shallow-water assemblage was composed of 
golden cownose ray ( Rhinoptera steindachneri ), whites- 
nout guitarfish ( Rhinobatos leucorhynchus), giant elec- 
tric ray, bullseye electric ray, vermiculate electric ray, 
longtail stingray, equatorial skate ( Raja equatorialis), 
blotched stingray, thorny stingray, and Panamic sting- 
ray. The deepwater assemblage was composed of witch 
guitarfish, sicklefin smoothhound, brown smoothhound, 
