NONLINEAR FORAGING RESPONSE OF A LARGE MARINE PREDATOR TO BENTHIC PREY - EAGLE RAY PITS AND BIVALVES IN A NEW-ZEALAND SANDFLAT

The density-dependent foraging response of eagle rays (Myliobatis tenu icaudatus) to infaunal bivalves (Macomona lilliana) was measured in a New Zealand sandflat. Disturbance pits provided unequivocal indicators of ray feeding activity, and pits were counted on a plot (250 m x 500 m) which had prey density mapped in a 200 cell (25 m x 25 m) grid. Al though foraging response increased significantly with prey density tre ated as a nominal (class, ANOVA-type) variable, treating bivalve densi ty as a ratio scale (continuous, regression-type) variable provided mo re information about characteristics of the response. Eagle rays exhib ited a nonlinear segmented response to prey density, in which ray fora ging activity was low and independent of prey density at low Macomona densities, while foraging increased sharply above a threshold density of prey but did not reach satiation at the highest prey densities in o ur site. By counting ray pits repeatedly over a 31 day period, we show ed that the levels and slope of the foraging response (no. of ray pits per 707 m(2) per 4 days) varied temporally during the season, but the nonlinear characteristic and the threshold of prey density were consi stent. Correlation analysis showed that the distribution of bivalve pr ey and ray foraging was spatially constant during the season. Comparis on of 3 estimators of prey density showed that a fitted polynomial den sity was the best predictor of ray foraging, and indicated that rays w ere responding to prey patches on a scale of 75-100 m. The temporal fe atures of the response to prey density were incorporated into a nonlin ear segmented model and integrated with respect to time for each cell of the study grid. The impact of ray foraging estimated from the integ ral indicated that only about 1.6% of the Macomona population was cons umed and 5.0% of the total plot was disturbed by rays during one month of study. However, the nonlinearity of response indicated that foragi ng impacts were concentrated disproportionately on high density patche s of prey, which suffered up to 4% mortality and 13% disturbance. Maco mona gained a refuge from predation and disturbance at low density, wh ich would stabilize prey populations and even out prey distribution. ( C) 1997 Elsevier Science B.V.