Graminoid Responses to Grazing by Large Herbivores: Adaptations, Exaptations, and Interacting Processes

The problem of ascribing adaptive significance to traits that enable graminoids to tolerate or evade ungulate herbivory is examined. Some of these traits may have originally evolved in response to nongrazing selection pressures, thus constituting grazing exaptations rather than true adaptations. The fossil record indicates that semiarid habitats, extensive grasslands, and grazers appeared, interacted, and evolved together. However, several traits that are advantageous in semiarid habitats have beneficial effects for grazed plants. Other traits, such as developmental plasticity, enhance competitive ability in certain environments, but also increase grazing tolerance or resistance. Experiments and simulation modeling showed that defoliation responses are embedded in a network of interacting processes, including photosynthesis, transpiration, nutrient uptake, and resource allocation. Responses and adaptations to defoliation must be interpreted in this context. Although traits may have arisen due to non-herbivorous selection pressures, they may subsequently have been selected, combined, or amplified through grass-grazer coevolution to form species, polymorphic populations, phenotypically plastic individuals, or communities of species that evade, resist, or tolerate herbivory. Graminoid grazing tolerance and the nearly simultaneous increase of grasses and grazers in the fossil record (Stebbins, 1981) suggest that grasses are adapted to herbivory, perhaps as a result of coevolution. Grasses and herbivores may, consequently, be somewhat mutualistic (Owen & Wiegert, 1981). Graminoid tolerance of herbivory due to continued elongation from the base of the leaf and lateral growth through tillering following defoliation have long been known (Lisle, 1757; Arber, 1934). Interacting physiological and morphological characteristics of graminoids that contribute to such grazing tolerance have been identified (McNaughton, 1979a) and analyzed (Coughenour et al., 1984b, 1985a). However, the origin of present traits in organisms cannot always be attributed to selection pressures of their present environment (Gould & Lewontin, 1979). The current beneficial effect of a trait is actually incidental or secondary if the trait frequency increased in response to another previous selection pressure. Thus, we must distinguish between beneficial traits (aptations) that are only incidentally beneficial (exaptations), and traits that have actually resulted from selection pressures to confer their present benefits (adaptations) (Gould & Vrba, 1982). We should consider alternative non-adaptive hypotheses in any investigation of the presumed adaptive significance of traits (Eriksson et al., 1983). Evolutionary constraints reduce the variety of evolutionary options available to a species. A trait may be non-adaptive in the sense that it is a necessary consequence of another trait (Eriksson et al., 1983). More fundamentally, individual traits are not in themselves the targets of natural selection. It is the fitness of the whole organism, or interacting system of traits, that determines the outcome of natural selection (Mayr, 1983). This "cohesion of the genotype" results in inevitable compromises among competing demands and a residue of sub-optimal parts that are necessary for the working of the whole (Mayr, 1983). Most plants are in essence metapopulations (White, 1979) of repeating modular units (Harper, 1981). The effect of herbivory on modular organisms differs from the effect of predation on non-modular organisms in that modular organisms survive predation through vegetative reproduction, whereas non-modular organisms do not. Therefore, it is necessary to consider defense against herbivory in a context of other interacting physiological processes and morphometric traits. Factorial experimentation and simulation modI I thank Sam McNaughton for an entree into this field via an extended post-doctoral fellowship on his Serengeti project. He and Linda Wallace were responsible for most of the experimental results. Interactions with them and with Roger Ruess were invaluable. This work was funded by grants from the National Science Foundation DEB-77 20350 and DEB-79 2291 to S. J. McNaughton. 2 Biological Research Lab, Syracuse University, 130 College Place, Syracuse, New York 13210. Present address: Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colorado 80523. ANN. MISSOURI BOT. GARD. 72: 852-863. 1985. This content downloaded from 157.55.39.112 on Wed, 07 Sep 2016 04:51:41 UTC All use subject to http://about.jstor.org/terms 1985] COUGHENOUR-GRAMINOID RESPONSES TO GRAZING 853 eling are used here to examine interactive processes, causal networks, and possible competing demands that are relevant to herbivory response in perennial graminoids. EVOLUTIONARY HISTORY OF GRASSES