MULTISCALE EXPERIMENTAL-ANALYSIS OF AGGREGATIVE RESPONSES OF MOBILE PREDATORS TO INFAUNAL PREY

Aggregative responses of predatory birds and fish to their prey are st ronger at some spatial scales than others. In waders feeding on benthi c invertebrates, aggregative responses decrease in strength at scales less than 100 m. However, decoupling of predator density or consumptio n rate from prey density at smaller spatial scales has not been tested experimentally. Three wader species (an endemic species, South Island Pied Oystercatcher Haematopus ostralegus finschi Martens, and the mig ratory species Red Knot, Calidris canutus rogersi (Mathews) and Easter n Bar-tailed Godwit, Limosa lapponica baueri Naumann) were abundant on an intertidal sandflat in Manukau Harbour, New Zealand before, during and after an experiment in which densities of their bivalve prey (Mac omona liliana Iredale) were manipulated on a small scale (0.5-m x 0.5- m plots) within a 250-m x 500-m study site. Faecal dropping analysis w as used to confirm the diet of South Island Pied Oystercatchers and Re d Knot, and to determine the diet of Bar-tailed Godwits. Based on know ledge of the foraging behaviour of these waders, we predicted the foll owing responses to the experimentally-induced increases in bivalve den sity: 1. waders would discover and then focus their foraging on plots with experimentally elevated densities of Macomona liliana; 2. wader d ensity within the study site would increase during the course of the e xperiment; and 3. wader density would then decrease when experimental plots were removed. To test these predictions the density and rate of prey attack by waders were measured before, during and after the densi ty-manipulation experiment. Although the waders discovered the experim ental aggregations of prey, there was no response at the scale of the plot, or at the larger scale of the study site. This study is the firs t experimental verification of no response by waders to small-scale in creases in prey patchiness, where larger scale changes in patchiness w ere controlled. We hypothesize that a lower limit on the spatial scale of aggregative response is set by the mobility of prey relative to th e predator-the patch structure of a relatively immobile prey will chan ge on a much smaller scale than that of a highly mobile prey, and thei r respective predators become adapted accordingly. (C) 1997 Elsevier S cience B.V.