Abstract The cost of locomotion is defined as the energetic cost of transporting one unit of body mass one unit distance. Previous experiments reported that the cost of locomotion in swimming fish of sizes .OO1 to 1 kg follows a common trend: as body mass increases, the cost of locomotion decreases. In this study, costs of locomotion were calculated for sevengill sharks, Notorynchus cepedianus, of weights 10.3 to 51 kg. The costs calculated followed the trend previously reported. The sharks used in this study represent the largest swimming fish for which the cost of locomotion has been calculated and demonstrate that the cost of locomotion trend holds for animals upward of 50 kg. Average swimming velocities were also calculated in order to compare them to a cost-optimization model based on the hypothesis that an unmolested fish will swim at a velocity that requires the least energy expenditure. In this study, swimming velocities varied from .31 to .44 m smi over a total body length range of 1.34 to 2.21 m. Ihese velocities are lower than those predicted by the cost¬ optimization model and those observed in two other shark species. Introduction An animal must move to compete for prey, to avoid predation, and to migrate. Movement requires energy as the animals must produce thrust, overcome drag, and maintain bouyancy (Alexander, 197781982). Since animals require movement for daily activities, the amount of energy put into the transport of body mass is of interest. Cost of locomotion was defined by Schmidt-Nielsen (1972) as the amount of fuel (in cal) necessary to transport one unit of body weight (in kg) over one unit of distance (one m). Schmidt¬ Nielsen (1984) and Beamish (1978) collated cost of locomotion data for fish of sizes 1 to 1000 g and found that the cost of locomotion decreases with increasing mass (Fig. 1). The largest animal studied, however, weighed only 1000g. It was of interest, therefore, to examine the cost of locomotion in larger fish. The sevengill shark makes an ideal subject for such a study for two reasons. First, individuals of this species can have a mass in excess of 50 kg. This mass allows for a test of the descending trend of Fig. 1 for an animal 50 times larger than the largest shown. Only one other study of cost of locomotion in sharks has been reported (Parsons 1990). The largest shark in that study, a bonnethead shark, weighed only 8 kg. Secondly, the sevengill shark is a ram ventilator. The shark must constantly swim in order to circulate water over its gills and breathe. Despite this ceaseless motion, a 51 kg captive sevengill shark eats less than .8 kg of salmon a week (Gilbert Van Dykhuizen, personal comment). This paper also compares swimming velocity of the sevengill shark to both theoretical predictions and other shark species. Weihs' (1977) hypothesized that an unmolested fish will swim at a yelocity that requires the least energy expenditure per unit distance travelled (Fig. 2). Parsons (1990) reported swimming velocities for bonnethead sharks and Weihs (1981) reported swimming velocities for bull and sandbar sharks. In both of these studies, the observed speeds matched Weihs' prediction. The velocities for these three sharks are compared to the average velocity of the sevengill sharks. Materials and Methods General Six sevengill sharks, Notorynchus cepedianus, were used for this study. The sharks were held at the Monterey Bay Aquarium, Monterey, California. Sex, weight, and length data for each shark are contained in Table 1. During the study, water temperature was 130c (+-10c). Cost of Locomotion Four pieces of information are necessary to compute an animal's cost of locomotion: 1) the amount of fuel ingested, 2) the resting metabolic rate, 3) the animal's mass, and 4) the distance travelled. For this study, the amount of fuel ingested by each animal was obtained from feeding records by averaging the number of grams of salmon eaten by each individual during the five months preceeding this study. Next, caloric intake was calculated by first taking the caloric value of the ingested salmon. (2,170 cal q-i, Altman 1968). Second, the number of calories absorbed from the food was estimated by multiplying the total caloric value of the salmon by an estimated assimilation rate. Wetherbee (1988) reported assimilation rates of 622 to 834 in the lemon shark at 25°C and these values are used here. Finally, resting metabolic costs were subtracted. Brett and Blackburn (1978) reported resting metabolic rates for elasmobranchs in general of 97 to 269 cal kg-i h-l at 150C, Scharold (198981990) and Bushnell (1982) reported metabolic rates within this range for the leopard and lemon sharks. Adjusting these values to 130C using a o 2.3 produced a resting metabolic range of 82 to 228 cal kgri hri. Each shark was weighed during the study. Distance travelled was measured as follows, A blueprint of the tank where the sharks were held was obtained and visual reference points such as anemones and rocks were marked on the blueprint. Each shark was observed for one hour sessions and its position in the tank was recorded on the blueprint every ten seconds. Each shark was watched for three sessions during daylight hours and one session at night. he beginning of the daytime sessions varied from 1000 to 1800 and the nighttime session took place at 0000. Swimming Velocity From the collected distance data, an average swimming velocity was calculated and compared with Weihs' (1977) cost¬ optimization model. Results Cost of Locomotion The cost of locomotion, calculated using different combinations of assimilation rates and resting metabolic rates, are shown in Fig. 3. Most of the values generated from the possible combination of assimilation and resting metabolic rate cluster around the regression line obtained for other fish. However some combinations, such as the lowest assimilation rate (624) and the highest resting metabolic rate (269 cal kg-i h-i), generate a cost of locomotion as low as O cal gri kmi. This cost is obviously too low since an animal cannot move its body mass without any energy expenditure. These zero results occur when the assimilation estimate is too low, the resting metabolic rate estimate is too high, or a combination of both. These points illustrate the problem of defining a resting metabolic rate for an animal that is in constant motion. Swimming Velocity The swimming velocities found in this study are lower than the optimal velocities predicted by Weihs (Table 1 and Fig. 2). Discussion Cost of Locomotion The total cost of locomotion estimates made here were slightly higher than the cost of locomotion predicted from the general trend for swimming organisms (Fig.4), but when possible resting rates are subtracted to yield net costs, the values cluster around the predicted result (Fig. 3). However, since assimilation rate and resting metabolic rate estimates were drawn from other shark species, these results must be viewed with caution. In conclusion, the results of this study support the cost of locomotion regression found for other fish (Figs. 1 and 3) and show that this relation correctly predicts the cost of transport for sharks weighing upward of 50 kg. This explains the somewhat surprising fact that a 50 kg shark can survive on less than 1 ko of food per week. Swimming Velocity Ihis study found swimming velocities of .31 to .44 m s-i for sharks of lengths 1.34 to 2.21 m (Table 1). Weihs (1981) reported a swimming velocity of .62 to .72 m smi for bull and sandbar sharks of lengths 2.0 to 2.3 m, and Parsons (1990) reported swimming speeds of .29 to .67 m s-* for bonnethead sharks of lengths .34 to .95 m. Both of these studies however were done at 25"C and, more importantly, measured these speeds over a fixed, relatively short, and straight section of the tank. This study, on the other hand, averages data from both the straight sections of the tank and the sections where the sharks had to turn to negotiate the walls. It was very clear from my observations that the sharks swam fastest during the straight sections of the tank. Since periods of wall negotiation were included in the calculation of the average speed, it may explain why the speeds observed in this study are lower than speeds of other species and lower than Weihs' predicted velocities (Fig. 2). The curved portions of the tank simply may prohibit the sharks from travelling at the optimal velocity that Weins predicts. Literature Cited: Alexander, R. McNeill and G. Goldspink (1977). Mechanics and energetics of animal locomotion. John Wiley & Sons, New York. Alexander, R. McNeill (1982). Locomotion of animals. Blackie & Son Limited, Bishopbriggs, Glasgow. Altman, Philip L. and Dorthy S. Dittmer (ed.) (1968). Metabolism. Biological Handbooks, Bethesda, Maryland, p. 13. Brett, J.R. and J.M. Blackburn (1978). Metabolic rate and energy expenditure of the spiny dogfish, Squalus acanthias. J. Fish Res Board Can 35: 816-821. Bushnell, P.G. (1982). Respiratory and circulatory adjustments to exercise in the lemon shark, Negaprion brevirostris. Master Thesis, University of Miami, Florida. 90pp. Parsons, G.R. (1990). Metabolism and swimming efficiency of the bonnethead shark, Sphyrna tiburo. Marine Biology 104: 363-367. Scharold, Jill, N. Chin Lai, William R. Lowell, and Jeffrey B. Graham (1989). Metabolic rate, heart rate, and tailbeat frequency during sustained swimming in the leopard shark, Triakis semifasciata. Exp. Biol. 48: 223-230. Scharold, Jill and Samuel H. Gruber (1990). Telemetered heart rate as a measure of metabolic rate in the lemon shark, Negaprion brevirostris. Unpublished paper. Schmidt-Nielsen, Knut (1972). Locomotion: energy cost of swimming, flying, and running. Science, New York 177: 222-228. Schmidt-Nielsen, Knut (1984). Scaling: why is animal size so important. Cambridge University Press, New York. Weihs, Daniel (1977). Effects of size on sustained swimming speeds of aquatic animals. In: Pedley, T.J. (ed.) Scale effects in animal locomotion. Academic Press, New York, p.333-338. Weihs, Daniel, Raymond S. Keyes and David M. Stalls (1981). Voluntary swimming speeds of two species of large carcharhinid sharks. Copeia 1981: 219-222. Wetherbee, Bradley M. (1988). Absorption efficiency of the juvenile lemon shark, Negaprion brevirostris, at varying rates of energy intake. Unpublished Masters Thesis. University of Miami, Florida. 78pp. Van Dykhuizen, Gilbert (1990). Senior Oquarist, Monterey Bay Aquarium. Telephone: (408) 648-4810. Table 1. General data on the sevengill sharks Weight Shark Sex Total Body (kg) Length (m) Male 51.0 29.1 Male 1.83 31.4 Female 1.74 Female 1.50 13.2 Female 1.47 14.5 1.34 Female 10.3 Avg. Swimming Velocity ms- (STD) .40 (0.03) 40 (o.02 .44 (0.01) 33 (0.01) .33 (0.01) .31 (0.01) Table 2. Caloric intake and total cost of locomotion with different estimated assimilation rates Total TOTAL COST OF LOCOMOTION (cal kg-i m-*) Shark caloric using 837 using 627 intake estimated estimated cals (STD) assimilation assimilation rate rate 2.66 (1.91) .0812 .1087 2.18 (1.62) .1558 .1164 4.70 (2.03) 2788 2083 1.30 (1.07) .2449 1829 1.66 (0.90) .2888 2158 .3333 1.27 (O.89) .2490 Table 3. Net cost of locomotion using different combinations of estimated assimilation rates and resting metabolic rates ET COST OF LOCOMOTION (cal kg-i m-i) Shark etimated assimilation rate: 837 627 627 82 228 resting met. rate: 82 228 -i h-1) (al k .0511 .0236 .0590 .0984 2274 .1589 .1366 .O661 .1765 .1145 .0556 o2 .2193 .0963 .1462 2589 1274 .0430 .1746 Figure Legend: Fig. 1: The regression for cost of locomotion for swimming animals The equation, y - 1.416w-o., relates the cost of locomotion y in cal gri km' to the weight of the animal w in g. Note that weight is expressed g. All other figures express the weight in kg. Fig. 2: The optimal swimming velocities predicted by Weins (1981) Weihs derived an equation which relates optimum velocity to total length: Ua = 0.45641°. where Up is optimum velocity in m sri and L is total length in m. The average swimming velocities found for each of the sharks in this study are also shown in this figure. Fig. 3: Cost of locomotion using combinations of estimated assimilation rates and resting metabolic rates The line in this figure is the regression from Figure 1. When weight w is measured in kg and cost of locomotion y is measured in cal kgr' mi, the equation of this line becomes y - O.252w-o.. Fig. 4: Total cost of locomotion compared to predicted net cost This diagram shows the relationship between estimated total costs, using assimilation rates of 624 to 834, and predicted net costs. Energy expenditure, cal q km —5 /3 0.8 0.4 0.2 + 1.0 Figure 2. 2.0 1.5 Length (m) 2.5 o. 5 01- Figure 3. Sumbol estimated estimated resting assimilation metabolic rate rate cal/(kg t) 83X 228 0 337 82 228 628 628 82 100 10 Weight (kg) 0 Figure 4. 337 estimated assimilation rate a 627 estimated assimilation rate 100 Weight (kg)