Food selection in an herbivorous rodent: balancing nutrition with thermoregulation

Ecology, Oct, 1997 by Hugo Torres-Contreras, Francisco Bozinovic

INTRODUCTION

The selection of food is a hierarchical process. Animals actively select the feeding microhabitat and, within this, they select food items. Consequently, factors that determine an animal's diet depend not only on food quality and abundance, but also on the multiple factors that determine microhabitat selection. The avoidence of risky microhabitats may be costly if food of high quality is abundant in these stressful microhabitats. Theoretically, the strategy of food selection should be determined by the solution of several cost-benefit functions, or a multifactorial one-currency function (Stephens and Krebs 1986).

In this work, we investigate the effect of experimentally varying food quality and ambient temperature on diet selection, time budget, and nutritional ecology. Foraging ecology and microhabitat use in desert-dwelling rodents have been well studied as an attempt to explain mechanistically the coexistence of rodents in arid regions (Reichman and Price 1993). We present experiments that evaluate the interaction between microhabitat selection and food quality in diet selection by degus. The degu, Octodon degus (Octodontidae), is a diurnal, herbivorous, caviomorph rodent inhabiting the semiarid and mediterranean environments of northern and central Chile. Degus feed primarily on foliage of forbs and grasses (Meserve 1981). The habitat of the degus has an irregular topography, and the dominant evergreen vegetation commonly shows a dispersed pattern producing a mosaic of patches that vary on a broad spatial scale (Fuentes et al. 1984, 1986). In this habitat, degus could, in theory, specialize in the exploitation of different microhabitats. In fact, O. degus restricts its activities to the vicinity of potential protective cover such as bushes, burrows, and rock piles (Jaksic 1986). On the other hand, Bozinovic (1995) found that, independently of food abundance, degus minimize ingestion of food with high dietary fiber content and show pronounced preferences for low-fiber food. Nevertheless, environmental temperatures and predation risk (Jaksic 1986) may also reduce diet selection and, hence, niche breadth in degus. In fact, when foraging, diurnal endotherms must avoid predators and also dissipate excess heat load through behavioral mechanisms (Chappell and Bartholomew 1981). In central Chile, where the climate is mediterranean with warm, dry summers and cool, wet winters, some species of rodents appear to be limited in their use of foraging space by environmental temperature (Bozinovic et al. 1995). Lagos et al. (1995) found that degus are constrained to specific foraging areas by their limited thermal tolerance. Because degus have a low capability for evaporative water loss (Rosenmann 1977), intolerance to enviromental temperatures higher than 32 [degrees] C was proposed as a factor explaining micro habitat use in this species. Lagos et al. (1995) recorded degu field activity under shrubs and in open areas, in large plots with predators either present or excluded. They postulated that high ambient temperatures, coupled with the degu's thermoregulatory constraints, reduce activity patterns and foraging to sheltered microhabitats, even when predators are experimentally excluded.

Our experiments attempted to determine whether degus show a trade-off between diet selection and thermoregulatory risk, feeding selectively to minimize dietary fiber and avoiding high ambient temperature. That is, we measured the effect of experimentally varying food quality and ambient temperature on food selection and consumption, as well as on time budget. Also, we investigated the criteria by which an herbivorous rodent selects food and the relationship between the diet selected and the nutrients that become assimilated under different thermal loads. To evaluate the costs of microhabitat selection on the rate of energy assimilation, we measured assimilation efficiency and digesta transit time of high- and low-quality diets, both factors associated with food profitability (Bozinovic and Martinez del Rio 1996). We hypothesized that one effect of the presence of preferred food in nonpreferred microhabitats is a decrease in the energy assimilation rate and in the transit time of digesta. Consequently, intermicrohabitat variation in food quality and ambient temperature may influence rodents' decisions on microhabitat use and, in turn, may provide mechanistic explanation of species coexistence (Kotler 1984, Kotler and Brown 1988, Brown et al. 1994).

MATERIAL AND METHODS

Animals were nonreproductive, all captured in Quebrada de la Plata, central Chile (70 [degrees] 50 [minutes] W, 33 [degrees] 31 [minutes] S). We used a total of 30 individuals, 20 different animals for experiments of food selection and 10 for nutritional trials. They were maintained on rabbit food pellets and water, and were later randomly assigned to different experiments. The study was conducted during winter and summer of 1995. Diets were prepared by adding and homogenizing cellulose (Sigma Chemical Company, Saint Louis, Missouri, USA) with commercial rabbit food and then pelletizing the mixture. Diets were analyzed for neutral detergent fiber, NDF (Bjorndal and Bolten 1993); nitrogen content of the diets was also measured using the microKjeldahl method (AOAC 1980). Energy content was determined in a Parr 1261 computerized calorimeter (Molina, Illinois, USA). Two replicates were determined to be ash free and were considered reliable when the difference between two measurements was [less than]1% (see Table 1). The level of fiber incorporated into the diet was set according to Bozinovic (1995), who documented that the grass consumed by O. degus in central Chile contained 61.1% NDF during the dry season (summer), and 37.3% NDF during the wet season (winter-fall).