Feeding experiments were carried out on larvae of 22 species of Lepidoptera, belonging mainly to the Papilionidae and the Bombycoidea, with a range of feeding habits from specialists on one plant family to polyphagous on a wide range of families. Consumption rates, conversion rates and hence growth rates were lower in tree-feeders than herb-feeders and may be related to differences in leaf water content. Such "quantitative" plant defense mechanisms may be metabolically more costly for insects to overcome than "qualitative" chemical defenses. It is suggested that, although it is only one of several possible factors involved, leaf water content might provide a useful index of the range in larval gross production efficiences and relative growth rates to be expected from a range of acceptable food plants. Feeny (1976) and Rhoades & Cates (1976) have suggested that "apparent" or "predictable" plant resources tend to be defended from phytophagous herbivores by relatively high concentrations of "quantitative" (see Feeny, 1975) or "digestibility-reducing" types of chemicals such as tannins, silica, lignin, resins, and other factors contributing to leaf toughness. Such defences may be effective against both specialist and generalist herbivores. By contrast, "unapparent" or "unpredictable" plants, or plant parts, may be able to escape in space or time, or else deter potential herbivorous enemies with a divergent array of toxic or repellent compounds which are effective in relatively low concentrations (Feeny, 1975).In this paper data are presented on experiments with 22 species of phytophagous Lepidoptera fed various natural foodplant leaves. From these studies several general patterns seem to have emerged which may provide a useful framework within which many insect/plant interactions may be more clearly understood or interpreted, and the results are discussed with reference to the concepts of quantitative and qualitative defenses.
MATERIALS AND METHODSLarvae of 22 species of Lepidoptera were reared on mature leaves of natural food plants with an abundance of food. Leaf turgot and shape, and water availability for the plants during the feeding experiment was maintained by the use PLANT GROWTH FORM AND LARVAL FEEDING 695 of Aquapics® (see Scriber, 1977). All experiments were run with a photo: scotophase of 16:Shr with corresponding day: night temperatures of 23.5°: 19.5 °.The insect species used and their feeding habits are listed below. They are classified as "specialist" feeders if restricted to one plant family, "intermediate" if feeding on two to ten families, and polyphagous if feeding on plants from more than ten families. Slansky & Feeny, 1977).
Family PapilionidaeApproximately 800 larvae of these species were individually monitored in the penultimate and final larval instars, and the following parameters were calculated (Waldbauer, 1968;Kozlovsky, 1968), all on a dry weight basis: G.R.: Growth Rate = mg (dry wt.) biomass gained/day R.G.R.: Relative growth rate =mg biomass gained/rag larval biomass/day C.R.: Consu...