The energetics of foraging in wading birds (Charadrii)

Abstract

A model is presented which predicts the simultaneous searching strategy (walking speed) and diet choice of a terrestrial predator, assuming the behaviour is selected to maximize the net rate of energy gain The model predicts an inverse relationship between predator velocity and prey availability, independent of prey type, and that predators should stop foraging below a critical prey availability. It is predicted that diet choice should become more restricted with increases in the availability of highly profitable prey (ie net energy return per second spent handling) but broader with increases in the relative density of low profitability prey. Parameters of the model, prey availability, energy content and handling time were measured for the common prey of two estuarine wading birds - the Redshank (Tringa totanus L.) and the Oystercatcher (Heamatopus ostralegus L ) on the mid—estuarine Firth of Forth, Scotland. Predator energy expenditures whilst handling and searching for prey were estimated using radiotelemetry of the heart rate from six unrestrained Redshank in an outdoor aviary. At the lower critical temperature (16°C), the handling costs averaged l.9xBMR (Aschoff and Pohl 1971) and 2.OxBMR for pecking and probing respectively, whilst searching cost l.7xBMR (walking at 30cm.s-1). Observed walking speeds in both species were well matched with the model's predictions at medium and high encounter rates, but at low encounter rates (2 items m walked-1) were lower than predicted The critical low availability at which it is profitable to stop foraging did not occur in the field during the study period (February 1981-May 1982). In Redshank the observed diet was not consistent with the net energy maximization model in Autumn or Early and Late winter and instead fitted better a model of gross protein maximization. In spring the observed diet was best described by the maximization of net energy gain. Including costs had a significant effect on the diet predictions in the Redshank. Differences between predicted and observed diet choice in the Oystercatcher were a result of the underselection of very large, high profitability items and partial selection of low ranking prey. Including costs had no effect on the model's predictions for the Oystercatcher. Differences between model predictions and the observed behaviour are discussed In the 'prizing' Oystercatcher differences appeared to be a result of inaccuracy in collection of one of the model parameters (unsuccessful manipulation rates) and invalid assumptions concerning the discriminant abilities of the predator. Whilst conflicting selective pressures - protein requirements and the avoidance of bill damage, probably explain the deviations in Redshank and 'hammering' Oystercatchers respectively.