Lopez et al.: A model for estimating biomass of fish species associated with fish aggregating devices 
169 
Layer 1 (3-14.2 m) 
Layer 2 (14.2-25.4 m) 
Layer 3 (25.4-36.6 m) 
Layer 4 (36.6-47.8 m) 
nj 
2 Layer 5 (47.8-59 m) 
oo 
.c 
g- Layer 6 (59-70.2 m) 
Q 
Layer 7 (70.2-81 .4 m) 
Layer 8 (81 .4-92.6 m) 
Layer 9 (92.6-103.8 m) 
Layer 10 (103.8-1 15 m) 
0 20 40 60 80 
Relative acoustic density (%) 
Figure 2 
Average percentages of acoustic backscatter (black bars) and their standard deviations (lines) 
recorded at sunrise from about 2000 acoustic samples recorded by 25 Satlink echosounder 
buoys attached to drifting fish aggregating devices (DFADs) in the Indian Ocean between 2009 
and 2012. This information was used as complementary information to define the preliminary 
depth limits between the different fish groups present at DFADs. The figure indicates a poten- 
tial segregation of size increasing with depth and a greater likelihood of larger tuna occupying 
waters deeper than 80 m (individuals with swim bladders). The large amount of acoustic back- 
scatter recorded in the first 25 m could be due to the presence of nontuna species, which also 
have swim bladders. Skipjack tuna, Katsuwonus pelamis, the main target species of the fleet 
fishing around DFADs is usually known to occupy medium depths and has no swim bladder. 
This potential segregation of size increasing with depth has also been found in previous works 
with conventional scientific echosounders around DFADs (Moreno et al., 2007). 
of the average acoustic backscatter recorded by each 
echosounder buoy depth layer (Fig. 2), combined with 
the behavioral information in the references cited 
above, suggested a potential segregation of tuna size 
increasing with depth, indicating a greater likelihood 
of larger tuna occupying waters deeper than 80 m. 
This seems to be in agreement with previous findings 
obtained through conventional scientific echosounders 
around DFADs (Moreno et al., 2007). 
From this accumulated information, the vertical 
boundary between nontuna species and small tunas 
was set at 25 m. Similar depth limits were adopted 
in previous studies with the use of the same echo- 
sounder buoys to separate bycatch from tuna (Lopez et 
al. 7 ; Robert et al., 2013). Because vertical depth limits 
7 Lopez, J., G. Moreno, M. Soria, P. Cotel, and L. Dago- 
rn. 2010. Remote discrimination of by-catch in purse seine 
fishery using fisher’s echo-sounder buoys. Indian Ocean 
Tuna Commission (IOTC), Working Party Ecosystem Bycatch 
between small and large tunas may be vague, a pre- 
liminary limit was initially set at 80 m depth, and was 
then re-adjusted in agreement with a set of 21 purse 
seine fishing operations conducted on DFADs in the At- 
lantic Ocean (see Parameter optimization section). 
Assigning TS and weights values to each species 
group Selecting appropriate TS and weight values for 
different species is also crucial for adequately convert- 
ing acoustic backscatter into reliable biomass estima- 
tions, by species or fish groups. Because no specific TS/ 
fork length (FL) relationship was available in the lit- 
erature for nontarget species, we considered a TS value 
of -42 dB for the entire group based on previous field 
studies (Josse et al., 2000; Doray et al., 2006; Doray et 
ah, 2007; Lopez et al. 7 ). The mean weight used for the 
biomass characterization of this community was 1 kg/ 
ind, which was estimated from the mean length of most 
(WPEB) IOTC-2010-WPEB-03, 14 p. (Available at website.] 
