Experience Influences Diet Mixing by Herbivores: Implications for Plant Biochemical Diversity
Villalba, J. J., Provenza, F. D. and Han, G. 2004. Experience influences diet mixing by herbivores: implications for plant biochemical diversity. -- Oikos 107: 100-109.
We often assume the intrinsic value of a food or habitat is similar for individuals of a species and above a certain threshold density more profitable foods should always be preferred over less profitable foods. Nevertheless, individual herbivores differ in their preferences for foods due in part to experience, and experience in variable environments is variable. In this report, we show that how sheep learned about their foraging environment was crucial to the development of their dietary habits, and that experience with foods that contained plant secondary metabolites (PSM) markedly enhanced their use of PSM-containing foods, even when familiar, nutritious alternatives were available ad libitum. Lambs who learned to eat foods that contained either tannins, terpenes, or oxalates ate more when they could select two of the foods offered simultaneously (tannins-terpenes, tannins-oxalates, or terpenes-oxalates) than when they were offered only one food. Lambs offered foods containing all three toxins ate more than lambs offered two of the toxins, and their intake was comparable to lambs offered the food that contained no toxins. Experience and the availability of nutritious alternatives both influenced food choice when the preferences of lambs with 3 months’ experience mixing tannin, terpenes, and oxalates were compared with lambs naive to the toxin-containing foods. During these studies, all lambs were offered five foods, two of them familiar to all of the lambs (ground alfalfa and a 50:50 mix of ground alfalfa:ground barley) and three of them familiar only to experienced lambs (a ground ration containing either tannins, terpenes, or oxalates). Half of the lambs were offered the familiar foods ad libitum, while half of the lambs were offered only 200 g of each familiar food daily. Throughout the study, naive lambs ate much less of the foods with toxins if they had ad libitum as opposed to restricted access to the nutritious alternatives (66 vs 549 g d−1). Experienced lambs also ate less of the foods with toxins if they had ad libitum, as opposed to restricted, access to the nutritious alternatives (809 vs 1497 g d−1). In both cases, however, lambs with experience ate remarkably more than naive lambs of the foods containing the toxins, whether access to the alfalfa-barley alternatives was ad libitum (811 vs 71 g d−1) or restricted (1509 vs 607 g d−1). These differences in food preferences and intake persisted during trials 8 months later. Plant communities offer a diverse matrix of biochemicals to herbivores, which may produce an array of interactions not accounted for by the traditional approach of studying nutrients and plant secondary metabolites (PSM) in isolation. How herbivores experience nutrient-PSM interactions may influence defoliation patterns and the potential for plant survival within plant communities. Thus, learning to mix foods that differ in kinds and concentrations of nutrients and PSM can enhance diet breadth and promote more uniform use of all plants in a community, which can influence the structure and function of ecosystems. Conversely, lack of experience learning to eat a variety of foods can diminish diet breadth and result in less uniform use of plants in a community.
Herbivores are important agents of change in ecosystems and their influences vary across time and space (Hobbs 1996). Thus, ecologists have long been interested in developing rules to account for foraging behavior, but the generality of the proposed rules is lacking. According to optimal foraging theory, a consumer should either eat or ignore a food depending on the density of other more profitable foods – the “zero-one-rule” (Stephens and Krebs 1986), or the “none-or-all” rule (McNamara and Houston 1987). Nevertheless, herbivores typically exhibit partial preferences, such that a food is sometimes ingested and sometimes ignored (Kennedy and Gray 1993, Wilmshurst et al. 1995). Partial preferences have been explained in various ways, including limited perception (Berec and Krivan 2000), inability to discriminate among different plant species (Illius et al. 1999), and attempts to meet nutritional needs (Westoby 1978) and reduce intake of plant secondary metabolites (PSM) (Freeland and Janzen 1974). Most of these conclusions are based on the assumption that a single nutrient is sufficient to measure foraging success, with little consideration for the biochemical context where a plant is growing (Atsatt and O'Dowd 1976, Milchunas and Noy-Meir 2002).
Nutrients and PSM do not act in isolation (Villalba et al. 2002a). Plant secondary metabolites impose nutritional costs to herbivores and nutrient intake affects detoxification processes (Illius and Jessop 1995). Nutrient-nutrient interactions influence growth and reproduction efficiencies in animals (Westoby 1978), while PSM–PSM interactions may influence the total amount of food an herbivore can ingest. Animals may be able to eat more foods with different kinds of PSM if they produce different effects in the body and they are detoxified by different mechanisms (Freeland and Janzen 1974). These interactions can lead to complementary relationships such that eating a combination of foods may exceed the benefit of consuming any one food in isolation (Tilman 1982).
Emerging evidence suggests partial preferences occur as the body integrates the postingestive effects of nutrients and PSM with a food's flavor, which result in flavor–nutrient–PSM interactions manifest as transitory preferences for a variety of foods (Provenza 1995, Provenza et al. 1998, 2002). In the process, individuals with different experiences may learn to eat different combinations of foods. Experience improves performance by inducing neurological, morphological, and physiological changes in animals (Provenza 1995, 1996, Provenza et al. 1998). Social and physical environments, continually interacting with the genome during growth and development, influence gene expression and generate new behavioral responses (McCormick et al. 2000, Dufty et al. 2002). Thus, while the body influences the structure of experience, experience also influences the structuring and functioning of the body (Provenza et al. 2003). These ongoing developmental processes enable individuals to adapt to local conditions and imply that what constitutes a “high quality” diet or habitat will differ for animals with different experiences. Sheep, goats and cattle reared on low-quality foods and habitats eat more low-quality foods and perform better than do animals that lack experience with those foods and habitats (Distel and Provenza 1991, Distel et al. 1994, 1996, Wiedemeier et al. 2002).
Herein we show that sheep learn to mix diets as a function of: (1) their experiences eating a variety of foods that contain different kinds of PSM and (2) of the availability of nutritious alternatives that do not contain PSM. We show that how sheep learn about their foraging environment is crucial to the development of their dietary habits. Animals who learn to eat different combinations of foods exhibit greater dietary breadth than animals familiar with only a few foods. Thus, experience enhances diet breadth and may promote greater use of all plants, which can influence the structure and function of ecosystems.