Spatial and temporal variation in diet and prey preference of nursery-bound juvenile lemon sharks (Negaprion brevirostris) at Bimini, Bahamas
Prey selection has never been determined in an elasmobranch, primarily because of the large home ranges possessed by adults making accurate quantification of prey in the environment problematic. Juvenile lemon sharks spend their first few years of life within protected nursery grounds, enabling the first quantification of prey selection due to the restricted area that they inhabit. Growth and survival of juvenile lemon sharks strongly influences adult fitness and recruitment, and therefore prey selection may play an important role in the life history of lemon sharks. The selection of a preferred species or size of prey by juvenile lemon sharks was determined by comparing the proportions of prey in the diet with proportions of prey in the environment at Bimini, Bahamas, between March 2000 and March 2003. The diet of lemon sharks was quantitatively described from the analysis of 642 shark stomachs (54.7 ± 0.3 cm precaudal length PCL, mean ± S.E., range 43.5 to 90.0 cm), of which 396 (62 %) contained food items. The main prey of juvenile lemon sharks at Bimini were mojarras (69% index of relative importance, IRI), parrotfish (5.5 % IRI), swimming crabs (5.1 %) and barracuda (3.1 % IRI). The yellowfin mojarra Gerres cinereus was the main prey of lemon sharks regardless of location, season, shark size or sex. Contingency table analysis revealed the diet of juvenile lemon sharks to be specific to location (χ² = 65.54, p < 0.0001), but homogeneous with season (χ² = 17.91, p = 0.118), shark size (χ² = 64.36, Ρ = 0.057) and shark sex (χ² = 13.21, Ρ = 0.354). Prey sizes were measured where possible, or calculated using least squares linear regression equations relating bone or carapace dimensions with original size. Original size was obtained for 350 dietary items, with 85 % calculated using bone regressions. Juvenile lemon sharks demonstrated no significant spatial or temporal variation in the size of prey consumed (Kolmogorov-Smirnov and Student's t-tests, Ρ > 0.05), but juveniles over 60.0 cm PCL consumed significantly larger prey than smaller sharks (ANOVA, Ρ < 0.001). Bone-length regressions also enabled a more accurate estimate of meal size (2.17 ± 0.17 % BW, mean± S.E., range 0.01 to 21.4 % BW, n = 407) and subsequently daily ration, 1.31 - 1.80% BW (depending on shark size), in comparison to traditional back-calculation techniques. Forty-three blocknets, 540 seine nets and 498 trawls were closed to sample mangrove and seagrass communities, resulting in the capture, identification and measurement of 216,150 fish and macro invertebrates. Catches were extrapolated over the entire study area providing an estimate of population sizes. Prey preference was estimated using chi-square residuals and a traditional electivity index. Values and rankings of selection varied with technique, but both revealed similar trends in prey preference. Proportions of prey families and prey sizes in the diet of lemon sharks from Bimini were significantly different to those found in the environment (χ², Ρ < 0.001 and Kolmogorov-Smirnov test, Ρ < 0.001 respectively). Lemon sharks demonstrated a preference for slower moving prey that were easier to capture (e.g. mojarras, toadfish, parrotfish and filefish), while avoiding larger, faster and harder to catch prey. Yellowfin mojarra were consumed in proportion to the distribution of fish lengths in the environment, suggesting that their importance in the diet may be due to preferred sizes in the environment as well as their ease of capture. Lemon shark diet was closely correlated with mangrove communities, demonstrating the importance of mangroves and the need for their protection in the Bahamas. The degree of selection exhibited by juvenile lemon sharks was greatest when prey were more abundant (off South Bimini and in the wet season), suggesting that lemon sharks conform to the optimal foraging theory.