Nutrient-Specific Foraging in Invertebrate Predators

See allHide authors and affiliations

Science  07 Jan 2005:
Vol. 307, Issue 5706, pp. 111-113
DOI: 10.1126/science.1105493


Many herbivores and omnivores adjust their food selection behavior to regulate the intake of multiple nutrients. Carnivores, however, are generally assumed to optimize the rate of prey capture rather than select prey according to nutrient composition. We showed experimentally that invertebrate predators can forage selectively for protein and lipids to redress specific nutritional imbalances. This selection can take place at different stages of prey handling: The predator may select among foods of different nutritional composition, eat more of a prey if it is rich in nutrients that the predator is deficient in, or extract specific nutrients from a single prey item.

It is generally believed that the body composition of prey animals is nutritionally balanced for carnivores, which consequently should be limited by the availability rather than the nutritional balance of prey (13). Herbivores and omnivores, by contrast, are known to face a heterogeneous nutritional environment and therefore possess a well-developed ability to balance their intake of multiple nutrients (4, 5). There is, however, accumulating evidence to suggest that some carnivores too might benefit from having nutrient-specific regulatory abilities. For example, different insect species vary widely in their body composition (68), and such variation can affect the performance of their invertebrate predators (9, 10). One species of spider has been observed to mix its intake from several prey species, leading to the yet untested hypothesis that they do so to compose diets with optimal amino acid makeup (11). Moreover, some domesticated carnivorous fish can compose diets from pure macronutrient sources in proportions that differ from random selection (12, 13), but these fish were unable to compensate for enforced nutritional imbalance by differentially selecting among foods (12).

We tested whether invertebrate predators can forage selectively for nutrients in order to redress specific nutritional imbalances. We investigated three polyphagous predators that use very different hunting tactics for catching prey: a highly mobile ground beetle, an ambushing wolf spider of intermediate mobility, and a web-building spider. The general approach was, first, to manipulate the nutritional state of the predators by feeding them a pretreatment diet with either a high or low ratio of protein to lipid for 1 or 2 days. Thereafter, we tested their feeding responses to the pretreatment food and/or a nutritionally complementary alternative.

In the case of the ground beetle Agonum dorsale (Coleoptera: Carabidae), we offered both the pretreatment food and the alternative food simultaneously, to emulate the scenario of a roving predator that actively selects between different prey types. In this experiment, semisynthetic food (based on powdered locusts) was used to avoid treatment-specific predator avoidance behavior that might result from the use of live prey. The results showed that intake of the two test foods was strongly dependent on the nutrient composition of the pretreatment diet (Fig. 1). Beetles pre-fed a lipid-enriched diet subsequently consumed more protein-rich than lipid-rich food, and those pre-fed with a protein-enriched diet preferentially consumed food with high lipid content, as well as eating more food overall. Thus, the beetles were able to select among foods of different protein and lipid composition to compensate for previous imbalances of these nutrients.

Fig. 1.

Nutrient balancing by the predatory ground beetle A. dorsale, when given a free choice between lipid-rich or protein-rich foods. Beetles were pre-fed with one of the two food types for 48 hours and were subsequently tested on both foods at the same time for 48 hours. The two food types were dried and crushed locusts mixed with either protein (protein-rich food) or lipids (lipid-rich food) (15). Values shown are means + SEM; P values test for equal consumption of+ lipid-rich and protein-rich foods (paired t tests).

The wolf spider Pardosa prativaga (Araneae: Lycosidae) is a sit-and-wait predator that frequently moves between foraging patches. To emulate a scenario where foraging patches vary in the quality of available food, we offered these predators only one prey type (either the pretreatment or the complementary prey type) and recorded how much of each was eaten. Food was live Drosophila fruit flies that had been raised on different media, resulting in body compositions with either a high or low protein-to-lipid ratio (Fig. 2). The data show that wolf spiders consumed more prey mass if the prey contained a high concentration of the nutrient that was in low concentration in their previous meal (Fig. 3). When feeding on lipid-rich flies, wolf spiders ate more if their previous diet was protein-rich flies; but when test prey were protein-rich, the wolf spiders ate more if their previous diet was lipid-rich. Like the beetles, these predators redressed a nutrient-specific imbalance but did so by varying their consumption of only a single prey type.

Fig. 2.

Relative composition of lipid and crude protein in adult Drosophila melanogaster fruit flies. Variations in body compositions were produced by varying the nutrient composition of the fruit fly cultures or by starving flies for 24 hours at 20°C. Fruit fly media were produced from a basis of Carolina medium, to which additional casein or sucrose was added (15). Each point represents a sample of 100 ± 5 male and female flies from a single culture bottle.

Fig. 3.

Intake of protein-rich or lipid-rich flies (dry mass) over 72 hours in no-choice tests by wolf spiders, P. prativaga. Before testing, spiders were pre-fed one of the same two prey types for 24 hours. Protein-rich flies were reared on a mixture of 40% Carolina medium and 60% casein; lipid-rich flies were reared on pure Carolina medium (15). Values shown are means + SEM; P values test for equal consumption by spiders with different feeding history (t tests).

The desert spider Stegodyphus lineatus (Araneae: Eresidae) builds durable webs for prey capture and spends relatively long periods at a single hunting site. Prey items arrive sporadically, and their nutritional composition is beyond the control of the spider. We therefore tested whether these spiders were able to extract nutrients differentially from within a single prey item to redress a nutritional imbalance. The spiders' nutritional state was manipulated by feeding them either lipid-rich or protein-rich flies for 24 hours. We subsequently measured how efficiently the spiders extracted nitrogen (N) and carbon (C) from a single prey item during 30 min of feeding. We found that the percentage of N left in the fly remnants after feeding was lower if the spiders' previous prey was protein-poor than if it was protein-rich, and this effect appeared when spiders were tested on both protein-poor (Fig. 4A) and protein-rich (Fig. 4B) flies. Furthermore, the extent to which spiders extracted N relative to C (N extracted per unit of C extracted) varied with feeding history: More N was extracted per unit of C if the spiders were previously fed protein-poor prey, both when test prey were protein-poor (Fig. 4C) and when they were protein-rich (Fig. 4D). This shows that these spiders have the ability to compensate for nutritional imbalances by extracting nutrients selectively from within individual prey items.

Fig. 4.

Differential extraction of N from individual test prey items by the web-building spider S. lineatus. Spiders were pre-fed protein-rich or protein-poor flies for 24 hours before being provided with a single protein-poor (A and C) or protein-rich (B and D) test fly. [(A) and (B)] % of N to total dry weight of live flies and of test fly remnants after 30 min of feeding by the spiders. Values shown are means + SEM. P values refer to differences between prey remnants from spiders with different feeding histories (t tests on arcsine-transformed data). [(C) and (D)] Extraction of N relative to C from test flies by spiders pre-fed with protein-rich (dashed lines) or protein-poor (solid lines) flies. The regression lines were not significantly different from parallel, either when test prey was protein-poor (C) or when it was protein-rich (D) [P > 0.05, analysis of covariance (ANCOVA) test for homogeneity of slopes]. P values test for different elevations of regression lines (ANCOVA, test for different intercepts) by spiders with different feeding histories. Different elevation indicates differences in the extraction of N relative to C (16).

Our study shows that invertebrate predators can regulate their intake of protein and lipids, and they can use this ability to redress an existing imbalance of these nutrients. The fact that only relatively short periods of prefeeding (24 to 48 hours) were sufficient to generate nutrient-specific compensatory responses suggests that the observed regulatory mechanisms are involved in fine-tuned compensation for day-to-day variations in prey nutrient composition. The need for such a capacity in insectivorous predators is highlighted by the demonstration that prey composition varies markedly with rearing diet, even within a single species (Fig. 2). Our data show, furthermore, that nutrient selection may occur at several stages of prey handling: either by selecting among foods of different nutritional composition (ground beetles), by adjusting consumption of a single prey type depending on its nutrient composition (wolf spiders), or by extracting nutrients selectively from within individual prey items (web-builders). These results are highly relevant for optimal foraging models, where it is generally assumed that for carnivores, prey quantity rather than quality matters (2). They are also relevant to ecological stoichiometry, in which models usually assume that prey animals (such as the flies in our study) maintain a constant body composition in the face of variable diets (14).

Supporting Online Material

Materials and Methods


References and Notes

View Abstract

Stay Connected to Science

Navigate This Article