Effects of food and fire on the demography of a nectar- feeding marsupial: a field experiment A. I. Tulloch & C. R. Dickman Institute of Wildlife Research, School of Biological Sciences, University of Sydney, Sydney, NSW, Australia Keywords Cercartetus nanus; pygmy-possum; diet; energy; experimental supplementation; food specialist; nectar; resource use. Correspondence Chris R. Dickman, Institute of Wildlife Research, School of Biological Sciences, University of Sydney, Sydney, NSW 2006, Australia. Email: cdickman@bio.usyd.edu.au Received 4 January 2007; accepted 20 March 2007 doi:10.1111/j.1469-7998.2007.00339.x Abstract Animals that specialize on nectar can be expected to face shortages of energy at times or in places where floral resources fade. Using a nectar-feeding marsupial, the eastern pygmy-possum Cercartetus nanus as a model, we predicted that animals would respond rapidly to an artificial energy supplement by showing improved body condition and increased local abundance. We also tested the hypothesis that responses would be more pronounced in burnt than in unburnt habitats due to the expectation that food would be limiting after fire. Energy was added in the form of sugar solution to two burnt and two unburnt woodland sites on the New South Wales coast, but not to two sets of equivalent control sites. The responses of pygmy-possums were compared between sites a month before and a month after supplementation during early autumn when flowers were diminishing. The energy supplement increased on-site immigration and a tail-fat index of body condition, but had no overall effect on the abundance or mean body mass of pygmy-possums. There was no effect of habitat; floral abundance was equal across sites. The results show that C. nanus responded rapidly to the energy supplement, and indicate that animals monitor shifts in the resource base continuously. Similar resource tracking abilities have been demonstrated previously in more mobile nectar-feeders, such as honeyeaters and bats. We suggest that such abilities should be generally advantageous in any species that depend on ephemeral resources. Introduction Food is an important determinant of the demography and local distribution of many species of animals, and a key driver in the evolution of life histories. The extent to which animals are food limited can often be gauged by measuring the body condition of individuals, the rate of population increase or other parameters, but such assessments may be unreliable unless the food supply is known. Supplementing the available food, so that it is not limiting, can overcome this problem. In mammals, the provision of supplementary food frequently leads to increased body condition, repro- duction and population density (Dickman, 1989; Boutin, 1990; Banks & Dickman, 2000). Individuals may also shift their periods of activity and tolerate higher risks of competi- tion or predation if the amount or quality of added food is sufficiently high (Desy, Batzli & Liu, 1990; Kotler et al., 2005). Despite the findings of these studies, generalizations about the influence of food are premature. Some species of small mammals, for example, show no response to supple- mentary food or respond only under certain conditions (e.g. Cole & Batzli, 1978). Others exhibit unexpected responses to supplementary feeding, such as decreased body condition in resident animals (Law, 1995) or expansion of movements and home range areas (Broughton & Dickman, 1991). Such disparate responses may arise if the added food disrupts behaviour, if factors other than food are limiting or if the added food is not readily used by the study species (Leung, 1994). In many situations, however, responses may differ between habitats and times, or because particular compo- nents of food such as protein, energy or specific nutrients are in short supply (Banks & Dickman, 2000; Sutherland et al., 2004). Hypotheses about food limitation may thus be tested most readily in species that have known or specialist diets, such as nectar-feeders, and if supplementation experiments are replicated in space and time. Nectar-feeding, or nectarivory, occurs commonly in in- vertebrates, but much less so in vertebrates (Recher, 1981). Studies on nectarivorous birds have often demonstrated that their abundance, activity and nesting are linked to varia- tions in nectar-producing flowers or directly to energy returns from nectar (Pyke, 1985; cf. Armstrong & Pyke, 1991). Law (1994) also found that local abundance of the blossom bat Syconycteris australis was correlated positively with the density of its main food (nectar). Supplementary feeding of blossom bats supported models of food limita- tion, with numbers of immigrants and body conditions of non-resident bats in supplementary-fed areas increasing compared with those in control areas (Law, 1995). Journal of Zoology 273 (2007) 382–388 c  2007 The Authors. Journal compilation c  2007 The Zoological Society of London 382 Journal of Zoology. Print ISSN 0952-8369