170 
Fishery Bulletin 1 14(2) 
Table t 
Target strength (TS), acoustic backscattering cross-section (Ob s , TS in linear scale) and weight (w) parameters 
used by the proposed method to conduct the echo-integration for each group of fish considered in this study. Below 
are the coefficients of correlation ( r ) and determination (r 2 ) between the uncorrected predicted biomass and real 
catches depending on the depth limit that was set to separate small and large tunas. 
Non- Small Large 
tunas tunas tunas 
TSldBrelm 2 ) -42 -35.1 -29.9 
o bs (m 2 ) 6.31 10- 5 3.09 10- 4 1.02 10“ 3 
W (kg) 1 2 21 
Depth limit (m) 
25 
36 
47 
59 
70 
80 
92 
104 
115 
r 
0.85 
0.84 
0.84 
0.84 
0.74 
0.74 
0.75 
0.76 
0.85 
r 2 
0.73 
0.70 
0.71 
0.71 
0.54 
0.54 
0.57 
0.58 
0.73 
represented nontuna species at DFADs, and their cor- 
responding weights (i.e., 72 cm FL for dolphinfish [Co- 
ryphaena hippurus], 100 cm FL for wahoo [Acanthocy- 
bium solandri], 30 cm FL for triggerfish [Canthidermis 
maculata], and 54 cm FL for rainbow runner [Elegatis 
bipinnulata] [Forget 8 ]). 
Moreno et al. (2007) analyzed the spatial distri- 
bution of the TS of fish aggregations around DFADs. 
Because the study of Moreno et al. (2007) is the only 
study conducted around DFADs with conventional 
scientific echosounders, to determine our tuna depth 
layers we used the TS values found by Moreno et al. 
(2007) (-35.1 dB for acoustic structures found between 
20 and 80 m; and -29.9 dB for acoustic structures 
found between 80 and 100 m), which fitted reason- 
ably well with the range of observations we initially 
proposed for small and large tunas. According to the 
most common tuna sizes caught at DFADs (Delgado de 
Molina et al. 9 ; IATTC, 2013; Chassot et al. 10 ; Floch et 
al. 11 ; Fonteneau et al., 2013; Harley et al. 12 ), the depth 
8 Forget, F. 2012. Personal commun. Institut de Recher- 
che pour le Developpement (IRD), UMR EME 212, Ave. Jean 
Monnet, CS 30171, 34203 Sete Cedex, France 
9 Delgado de Molina, A., J. C. Santana, J. Ariz, and I. Sabate. 
2012. Estadfsticas Espanolas de la pesquerfa atunera trop- 
ical, en el Oceano Atlantico, hasta 2010. Collect. Vol. Sci. 
Pap. ICCAT 68:1200-1220. [Available at website.] 
10 Chassot, E., A. Delgado de Molina, C. Assan, P. Dewals, 
P. Cauquil, J. J. Areso, D. M. Rahombanjanahary, and L. 
Floch. 2013. Statistics of the European Union and associ- 
ated flags purse seine fishing fleet targeting tropical tunas 
in the Indian Ocean 1981-2012. Indian Ocean Tuna Com- 
mission (IOTC), Working Party Tropical Tuna (WPTT) IOTC- 
2013-WPTT15-44, 28 p. [Available at website.] 
n Floch, L., A. Damiano, A. Tamegnon, P. Cauquil, P. Chavance, 
I. Terrier, and E. Chassot. 2014. Statistics of the French 
purse seine fishing fleet targeting tropical tunas in the At- 
lantic Ocean (1991-2012). Collect. Vol. Sci. Pap. ICCAT 
70:2669-2692. [Available at website.] 
12 Harley, S., P. Williams, S. Nicol, and J. Hampton. 2013. The 
western and central Pacific tuna fishery: 2011 overview and 
status of stocks, 31 p. Secretariat of the Pacific Community, 
Noumea, New Caledonia. [Available at website.] 
range between 25 and 80 m was considered to be popu- 
lated by skipjack, yellowfin and bigeye tuna of a mean 
mass of 2 kg/ind (about 50 cm FL), whereas the depth 
between 80 and 115 m was assumed to be populated 
by larger yellowfin and bigeye tuna individuals with a 
mean weight of 21 kg/ind (110-100 cm FL, for yellowfin 
and bigeye tuna, respectively). 
Echo-integration Area backscattering coefficients (s a 
m 2 /m 2 ; MacLennan et al., 2002) recorded in each of 
the ten echosounder buoy layers were converted into 
biomass following a depth layer echo-integration pro- 
cedure. A specific acoustic backscattering cross-section 
value (ab s , m 2 , TS in linear scale; MacLennan et al., 
2002) was used to obtain the number of individuals 
for each of the integrated layer (n=l, 2, ..., 10) accord- 
ing to the presence of each group initially assigned in 
this study to each depth layer (i.e., nontuna [3-25 m], 
small tuna [25-80 m], large tuna [80-115 m]). The Gbs 
and weight values used for the echo-integration of the 
acoustic backscatter are shown in Table 1. The number 
of fish per group and layer (N[n, gr]) were estimated 
as follows: 
N(n,gr) = Sa ~ n - A(n) (i) 
CT bs(gr) 
where s a (n) 
<7 (bs(gr)) 
Ain) 
the TVG-corrected (time-varied-gain, a cor- 
rection function to compensate the sig- 
nal for spreading and absorption losses; 
Simmonds and MacLennan, 2005) area 
backscattering coefficient (Maclennan et 
al., 2002) in each layer (n); 
the mean TS of a known group (i.e., of 
nontuna species, small or large tunas) 
in linear scale; and 
the mean cross sectional area sampled by 
the beam of the cone for each layer (n). 
Then, the total number of fish per group Mgr) were 
obtained by summing for all layers (2): 
