The relative roles of hunger and size asymmetry in sibling aggression between nestling ospreys, Pandion haliaetus Marlene M. Machmer and Ronald C. Ydenberg Abstract: The food-amount hypothesis, proposed to account for sibling aggression in brood-reducing bird species, predicts that broods will be aggressive when hungry, and that as a result, senior siblings will be able to take a greater share of the total available food. We performed a field experiment to test these predictions in ospreys, Pandion haliaetus. Broods were removed from their natal nests and subjected to a series of experimental trials. Each trial consisted of a 3.5-h starvation period followed by a feeding or sham-feeding treatment designed to manipulate the hunger of the nestlings. Following treatment, broods were placed in a viewing nest along with a fish. While the viewing-nest mother fed the fish to the brood, we recorded food allocation and aggresssion. Each of 10 broods experienced a number (one to six) of feeding and sham-feeding treatments over a period of 2 days and was then returned to its natal nest. Broods were slightly but significantly more aggressive after sham-feeding treatments, and senior siblings took a greater share of the meal. The tremendous variation in the level of sibling aggression observed between broods was inversely related to brood mass asymmetry but not to condition, suggesting that siblings adjusted the rate of aggression according to the level of competition within their broods. Résumé : L’hypothèse de la quantité de nourriture comme cause de l’agressivité des oisillons à l’égard de leurs frères et soeurs chez les espèces qui contrôlent le nombre d’oisillons dans les couvées suppose que les oisillons sont agressifs quand ils ont faim et que, en conséquence, les oi3sillons plus vieux devraient pouvoir s’approprier une plus grande partie de la nourriture disponible. Nous avons éprouvé cette hypothèse chez des Balbuzards pêcheurs, Pandion haliaetus, en nature. Les oisillons ont été enlevés de leurs nids d’origine et soumis à une série d’expériences au cours desquelles leur faim était manipulée : ils étaient privés de nourriture pendant 3,5 h, puis nourris ou soumis à une période d’alimentation simulée. Après l’expérience, les couvées étaient placées dans un nid bien en vue avec un poisson. Pendant que la mère distribuait le poisson à sa progéniture, nous avons noté la ration reçue par chaque oisillon et l’intensité des agressions. Chacune des 10 couvées a reçu un certain nombre de repas (un à six) ou de repas simulés au cours d’une période de 2 jours, puis a été retournée à son nid d’origine. Les oisillons étaient légèrement, mais significativement, plus agressifs après les repas simulés et les oisillons les plus âgés s’appropriaient une plus grande part du repas. L’énorme variation de l’intensité des agressions chez les oisillons des différentes couvées était inversement proportionnelle à l’asymétrie de la masse de la couvée, mais pas au coefficient d’embonpoint, ce qui semble indiquer que les oisillons ajustent l’intensité de leurs agressions selon le degré de compétition au sein de leur couvée. [Traduit par la Rédaction] Introduction The brood-reduction hypothesis (Lack 1968; see O’Connor 1978; Stinson 1979; Mock et al. 1990 for reviews) states that in birds, competitive asymmetries established through hatch- ing asynchrony and exaggerated through sibling aggression facilitate reduction of broods to a level matching prevailing food levels. The food-amount hypothesis (FAH; Mock et al. 1987) proposes that sibling aggression varies inversely with the quantity of food delivered by parents. By virtue of their larger size, senior siblings are able to use aggression to obtain a greater share of the delivered food, and they do so when hungry. The FAH generates two predictions: broods will be more aggressive when hungry than when satiated, and senior siblings will take a greater share of the total available food. Experimental tests of the FAH (Procter 1975; Mock et al. 1987; Drummond and Garcia Chavelas 1989) have revealed two patterns of siblicide, the distinction between them being the presence or absence of a proximate link betweeen hunger and aggression. When a link is present, nestlings are more aggressive when hungry, as predicted by the FAH. When a link is absent, aggression occurs regardless of food-intake level, and the ability of the junior sibling(s) to survive prolonged aggression depends on the amount of food provided by the parents (Mock et al. 1987). Sibling aggression and brood reduction are an integral part of the reproductive ecology of ospreys (Pandion haliaetus; Poole 1982, 1984; Jamieson et al. 1983; Hagan 1986; Forbes 1991; McLean and Byrd 1991; Steidl and Griffin 1991). Os- preys typically lay three eggs, and incubation generally begins with the second egg. Nestlings hatching last are at a competi- tive disadvantage because they are smaller (Poole 1984; Steeger 1989). The incidence of apparent brood reduction var- ies both among years (Stinson 1977; Poole 1989; McLean and Can. J. Zool., 76: 181–186 (1998) Received February 13, 1997. Accepted August 18, 1997. M.M. Machmer. Pandion Ecological Research Ltd., P.O. Box 26, Ymir, BC V0G 2KO, Canada. R.C. Ydenberg. 1 Behavioural Ecology Research Group, Department of Biological Sciences, Simon Fraser University, Burnaby, BC V5A 1S6, Canada. 1 Author to whom all correspondence should be addressed (e-mail: ydenberg@sfu.ca). 181 © 1998 NRC Canada