Wood et al.: Diet a of Isurus oxyrinchus in the northwest Atlantic Ocean 
79 
individuals with predictive equations (Wood, 2005). To 
explore predator-size-prey-size relationships a shortin 
mako size-bluefish size scatter plot was analyzed with 
least squares regression. Quantile regressions (5 th and 
95 th ) were used to determine changes in minimum and 
maximum prey size with increasing predator size. In 
addition, relative and cumulative frequency histograms 
were used to explore patterns in the size of prey con- 
sumed (Bethea et al., 2004). 
Daily ration 
Two methods were used to estimate daily ration of the 
shortfin mako: a bioenergetics approach and the use 
of the average weight of stomach contents (following 
Elliot and Persson, 1978). These approaches were both 
used so that a comparison between the resulting daily 
ration estimates could be made. In addition, both of these 
methods were previously used to calculate daily ration 
of shortfin makos (Stillwell and Kohler, 1982). 
The bioenergetics approach used by Stillwell and 
Kohler (1982) did not include growth information and 
was based on the volume of oxygen consumption (V0 2 ) 
extrapolated from four species of squaloid sharks. More 
recently, V0 2 has been measured directly for the short- 
fin mako at various swimming speeds (U) (Graham et 
al., 1990). Stillwell and Kohler’s (1982) estimate of V0 2 
(284.2 mg/kg/h) was much lower than the values of V0 2 
actually measured for the shortfin mako by Graham et 
al. (1990), who found an average V0 2 of 369 mg/kg/h 
for routine metabolism. 
The bioenergetics model for this study followed a 
form commonly used for teleost fishes which has been 
successfully applied to blue sharks (Prionace glauca) 
(Schindler et al., 2002). To calculate daily consumption, 
the model incorporates growth rates, metabolism, and 
other energy parameters in an energy balance equation: 
C = M + SDA + (F + Ur) + (G t + Rp), 
where C 
M 
SDA (specific dynamic action) 
F and Ur 
G t 
Rp 
consumption rate; 
metabolism; 
the amount of energy 
used for digestion; 
energy lost to waste; 
growth over time; and 
the amount of energy 
allocated towards 
reproduction. 
Metabolism (M) in the model was assumed to be ac- 
tive metabolism because shortfin makos are obligate 
ram ventilators (must continually swim in order to 
breathe). To generate a relationship between swimming 
speed (U) and mean V0 2 , a least squares regression was 
fitted mean V0 2 data at a variety of swimming speeds 
taken from Graham et al. (1990)’s data. The resulting 
regression equation, along with observed rates of travel 
determined from satellite telemetry tracking of shortfin 
makos, was used to calculate active metabolism. An 
energy equivalence of 13.6 J/mg 0 2 was used to convert 
the V0 2 consumed into energy (Schindler et al., 2002), 
and a Q 10 value for the bonnethead shark ( Sphyrna 
tiburo) of 2.3 (Carlson and Parsons, 1999) was used to 
adjust the final metabolic rate to 18.8°C (the preferred 
temperature of shortfin makos in the northwest Atlan- 
tic; Stillwell and Kohler, 1982). 
Specific dynamic action (SDA) was set at a fraction 
of consumption rate (C) equal to 0.10C (Schindler et 
al., 2002), and the amount of energy lost to waste 
(F + Ur) was fixed at 0.27C (Stillwell and Kohler, 1982; 
Schindler et al., 2002). For growth, sex-specific growth 
rates (G t ) were taken from Natanson et al. (2006) who 
found that growth in length was best modeled by a 
three-parameter von Bertalanffy growth curve for males 
and a three-parameter Gompertz growth curve for fe- 
males. Fork length (FL) was converted to weight with 
the relationship WT - 5.2432 x 10~ 6 FL 3 1407 , with weight 
in kg and FL in cm (Kohler et al., 1996). The energy 
density value used for the shortfin mako was 20.6 kJ/g 
dry weight, which was converted to wet weight energy 
by assuming a 73% water content for shortfin mako 
flesh (Steimle and Terranova, 1985). The resulting wet- 
weight energy assumed for all body sizes of the shortfin 
mako was 5562 kJ/kg which is very close to the average 
estimate calculated for all sharks of 5414 kJ/kg (Cortes 
and Gruber, 1990; Schindler et al., 2002). 
Energy allocation to reproduction (Rp) was only cal- 
culated for females and was assumed insignificant in 
male sharks. Reproductive growth for mature females 
(>18 years; Natanson et al., 2006) was calculated by 
assuming the following reproductive characteristics: 
mean litter size = 11.1, mean size at birth = 74 cm total 
length (TL), 24-month gestation period, and 3-year re- 
productive cycle (Mollet et al., 2000). This reproductive 
information coupled with the energy density (5562 kJ/ 
kg) for shortfin makos gave an estimated energy cost 
for reproductive growth. 
The overall energy content of the shortfin mako diet 
was determined with species-specific energy values from 
Steimle and Terranova (1985). The resulting value was 
used to calculate daily ration based on the overall en- 
ergy demand from the bioenergetics model. For compari- 
son of daily ration estimates based on the bioenergetics 
model, the method of Elliot and Persson (1978) was 
applied to the stomach contents data collected in the 
present study. Previously, for the shortfin mako, time 
for 90% evacuation of a meal was estimated at 36 to 
48 hours (Stillwell and Kohler, 1982). It is now known 
that the V0 2 of shortfin makos is in the same range as 
that of tunas (Graham et al., 1990; Korsmeyer et al., 
1996), which is unsurprising given the similarities that 
exist between these pelagic predators (i.e., body form, 
prey, endothermic capability). Studies have revealed 
that evacuation time for larger species of tuna, such as 
yellowfin tuna ( Thunnus albacares ), can range from 6 to 
20 hours for complete evacuation depending on the prey 
type (Olson and Boggs, 1986). Based on similarities 
with tunas, as well as on a markedly higher metabolic 
rate than that estimated in Stillwell and Kohler (1982), 
