The relative importance of food quality vs. enemy‐free space remains an unresolved but central issue in the evolutionary ecology of host use by phytophagous insects. In this study, we investigate their relative importance in determining host‐plant use by a generalist caterpillar, Estigmene acrea (Lepidoptera: Arctiidae). In nature, E. acrea late‐instar caterpillars preferred Senecio longilobus (Asteraceae), which contains pyrrolizidine alkaloids that the caterpillars sequester, over Viguiera dentata (Asteraceae), a natal host, and typically suffered a 28% mortality risk from parasitoids. We hypothesized that the natural, mixed diet of caterpillars provides high‐quality food via hosts like Viguiera as well as antiparasitoid defense via sequestered toxins from Senecio. We found that a pure Viguiera diet provides superior growth performance over a pure Senecio or mixed diet in the absence of parasitism. However, when parasitism risk is at least moderate, the mixed diet provides a survival advantage over the pure diets of Viguiera or Senecio. We therefore conclude that the balance between benefits of growth (food quality) and defense (enemy‐free space) maintains the use of a mixed diet in nature. Furthermore, the value of enemy‐free space supercedes the value of food quality in determining the host‐plant preference of late‐instar caterpillars.

Abstract.  1. Seventy‐seven individual last‐instar caterpillars foraging in the field were examined for 6 h each. They represented four species of Arctiidae of similar size and habitat use. Two, Hypocrisias minima and Pygarctia roseicapitis, are specialists restricted to particular plant genera. The other two, Grammia geneura and Estigmene acrea, are extreme generalists that use many host plant species from multiple plant families.2. Parameters of behavioural efficiency were monitored. Generalists spent more time walking, rejected more potential host plants, took longer to decide to feed after inspecting a plant, and took relatively more small feeding bouts compared with specialists.3. This is the first test of differential foraging efficiency in the field in relation to diet breadth of insects and the data indicate that generalists are less efficient in their foraging activities and may suffer from divided attention. The need for attentiveness to enhance efficiency and thereby reduce ecological risk is discussed.

The underlying mechanisms mediating the use of multiple host plants were investigated in Parrhasius polibetes (Lycaenidae), a florivorous and facultative myrmecophilous butterfly. Plant traits such as presence of ant-treehopper associations as a source of enemy-free space, flower bud dimensions, toughness, thickness, trichomes, and the corresponding performance and wear of P. polibetes mandibles were examined for three natural hosts: Schefflera vinosa (Araliaceae), Pyrostegia venusta (Bignoniaceae) and Luehea grandiflora (Malvaceae). Parasitism levels of larvae found on the three hosts were also determined. Almost all Luehea had ant-treehopper associations, and all larvae found on this host were non-parasitized. Parasitism was low in larvae found on Schefflera, half of which hosted ant-treehopper associations. No ant-treehopper association was found on Pyrostegia, where parasitism was significantly higher compared to other hosts. In the laboratory, P. polibetes performed well on Schefflera, followed by Pyrostegia. No larvae survived when fed with Luehea. Flower buds of Luehea were thicker and tougher than those of Schefflera and Pyrostegia. Indeed, mandibles of larvae reared on Luehea showed substantial wear, whereas those reared either on Schefflera or Pyrostegia presented no significant damage. Additionally, we suggest that co-occurrence with ant-treehopper associations on a plant provides parasitoid-free space for P. polibetes larvae. Our results support the hypothesis that ecological trade-offs among host plants (i.e., food quality and enemy-free space) promote polyphagy in natural populations of P. polibetes. Host morphological traits seem to play a relevant role in P. polibetes performance. To our knowledge, this is the first report showing the costs of polyphagy in a myrmecophilous butterfly.

Adult body size, a key life history component, varies strongly within and between Heliconius erato phyllis (Lepidoptera: Nymphalidae) populations. In the present study, we determined whether seasonal variation in adult body size is temperature related and/or determined by seasonal changes of host plants (Passifloraceae) used by the larval stage. A population of H. erato phyllis located in a Eucalyptus plantation (Barba Negra Forest, Barra do Ribeiro County, Rio Grande do Sul State, Brazil) was sampled every 45 days from March 1997 to October 1998 to quantify seasonal variation in adult body size and use of larval host plants. In the laboratory, the effects of the following factors on adult body size were quantified: (i) host plant species ( Passiflora misera or Passiflora suberosa ); (ii) food quantity consumed by larvae (experimentally manipulated for each passion vine species); (iii) winter and summer temperatures (15 and 25 Њ C, respectively); and (iv) the interaction between host plant species and temperature. Adults emerging during summer were larger than those emerging in other seasons. Female butterflies oviposited selectively on P. misera even when the dominant passion vine was P. suberosa . They only switched from using P. misera to P. suberosa during later autumn and winter, when P. misera vines were completely defoliated. The laboratory feeding trials with both passion vines showed a strong positive association between food quantity consumed by larvae and adult size. They also confirmed that adults are larger when their larvae are reared on P. misera than on P. suberosa . Temperature during larval development had no effect on H. erato phyllis adult size. Thus, seasonal variation of H. erato phyllis adult size in a given place is primarily determined by the availability and quality of host plant species used by the larval stage.