NOTE Koen Alonso et al. : Diet of Lagenorhynchus obscurus in waters off Patagonia 
367 
NF - LO + 0.5NO, if LO >RO, 
or as 
NF = RO + 0.5NO, if RO>LO, 
where LO, RO, and NO = the number of left, right, 
and not assigned otoliths respectively. 
Minimal number of cephalopods was obtained with 
maximum count of upper and lower beaks for each 
species. 
All complete and nondigested elements (otoliths 
and beaks) were measured with digital calipers. Some 
otoliths or beaks were digested, or broken, and could 
not be measured. Any otolith was considered too di- 
gested to be measured if it presented rounded bor- 
ders and rostrum, and an ill-defined sulcus. Around 
42% of the prey items were represented by broken or 
digested elements. Size values for these pieces were 
randomly assigned from sizes of items taken from 
the stomach and measured. When only broken or 
digested elements were found in the stomach, the 
average individual weight of the prey species in the 
whole sample was used to estimate the individual 
weight of the items. 
The length of prey items was estimated by using 
regressions between total length (TL) and otolith 
length (OL) for fishes, and between dorsal mantle 
length (DML) and lower rostral length (LRL), or 
lower hood length (LHL), for squids. Wet weight (W) 
was estimated from regression-estimated length of 
prey by using regressions between W and TL for 
fishes, and between W and DML for squids. Regres- 
sions for most prey species were developed during 
this study on the basis of materials collected from 
commercial hauls. All regressions were calculated as 
simple linear regressions. Those variables that did 
not have linear relationships were made linear with 
natural logarithms (Table 1). The TL and W of 
Notothenia sp. was estimated with Notothenia 
angustifrons regressions (Table 1). The W of Octo- 
pus tehuelchus was estimated with Octopus vulgaris 
regression between W and LHL (Table 1). The W of 
each Semirossia tenera was assumed to be 3 g, as 
was the average individual weight recorded for this 
species in a sample taken onboard a commercial 
trawler by one of the authors (N.A.G.). In one stom- 
ach (that of L022 dolphin), only fish eye lenses and 
a few beaks were found. In this case, eye lenses were 
similar in size to those of anchovy; therefore the an- 
Table 1 
Equation, sample size (n), and determination coefficient ( r 2 ) of the regressions used to estimate size and weight of prey of dusky 
dolphins in Patagonia. The variables used in the regressions were otolith length (OL), lower hood length (LHL), lower rostral 
length (LRL), total length (TL), dorsal mantle length (DML) and wet weight (W). OL, LHL, and LRL are in mm, TL and DML are 
in cm, and W is in g. 
Scientific name 
Common name 
Equation 
n 
r 2 
Source 
Engraulis anchoita 
Argentine anchovy 
TL = 2.368 + 3.56 LO 
79 
0.70 
This study 
W= 2.5 10- 3 TL 3 353 
81 
0.93 
Merluccius hubbsi 
Argentine hake 
TL = 1.823 LO 1 072 if OL<15 
447 
0.93 
W = 4.76 HP 3 TL 3 061 if OL< 15 
469 
0.92 
This study 
TL = 1.984 LO 1 05 if OL>15 
693 
0.91 
W = 9.72 HP 3 TL 2 886 if OL>15 
742 
0.96 
Stromateus brasiliensis 
“pampanito” 
TL = 3.042 LO 1 159 
51 
0.98 
This study 
W = 6.418 10- 4 TL 3 917 
63 
0.98 
Notothenia angustifrons 
Southern cod 
TL = 4.142 LO 0 768 
22 
0.91 
Hecht, 1987 
W = 3.73 10- 6 (10 TL) 3 16 
24 
0.98 
Illex argentinus 
Argentine shortfin squid 
DML = -3.178 + 5.617 LHL 
27 
0.93 
This study 
DML = 8.257 10~ 2 + 6.009 LRL 
63 
0.87 
W = 9.82 10- 3 LDM 3 238 
66 
0.98 
Loligo gahi 
Patagonian squid 
DML = -0.712 + 4.622 LHL 
98 
0.76 
This study 
W = 2.6 10- 2 LDM 2 753 
102 
0.93 
Octopus vulgaris 
octopus 
_ gl.82 + 3.03 Ln (LHL) 
108 
1 
Clarke, 1986 
' The determination coefficient of this regression was not available 
in the original source. 
