M. Deane Bowers

EFFECTS OF PLANT AGE, GENOTYPE, AND HERBIVORY ON PLANTAGO PERFORMANCE AND CHEMISTRY'

Plant performance and chemistry may vary due to a variety of factors, such as plant genotype, environmental conditions, presence of herbivores, timing of herbivory, and species of herbivore. The relative importance of these factors, and how the plants respond to them, may affect the dynamics of the plant population, as well as the insect herbivores feeding on those plants. To understand the relative importance of some of these factors on plant performance and chemistry, we used Plantago lanceolata L. (Plantagina- ceae). In an experimental garden at Binghamton, New York, we examined the effects of plant age, plant genotype, and herbivory by generalist or specialist caterpillars on P. lan- ceolata. There were two parts to this experiment. In the first, we examined variation in nutritional quality (nitrogen) and defensive chemistry (the iridoid glycosides aucubin and catalpol) as a consequence of plant age, plant genotype, and leaf age class. We compared these parameters for a set of four pairs of plants of each of five genotypes that were not exposed to herbivores, and were harvested in early July (control-start plants) with another set harvested 6 wk later in mid-August (control-end plants). The older plants harvested in August (control-end plants) had concentrations of aucubin, catalpol, total iridoid glycosides, and nitrogen approximately one-half that of plants harvested 6 wk earlier. Leaf age affected all of these variables, and plant genotype influenced the iridoid glycoside variables but not the nitrogen concentration. Overall, leaf age explained twice as much of the variation in iridoid glycosides as did plant age, which accounted for twice as much variation as did plant genotype. The second part of the experiment compared the effects of herbivory, genotype, and leaf age on plant performance (leaf biomass, scape (flower stalk) biomass, and total biomass) and chemistry (nitrogen, protein, and iridoid glycoside concentration). We compared these measures for a set of four pairs of replicate plants of each of five genotypes, exposed to one of three herbivory treatments: no herbivory, herbivory by specialist caterpillars, and herbivory by generalist caterpillars. The results of this experiment showed that herbivory had little effect on plant performance, and there was no difference due to herbivory by the specialist Junonia coenia Hibner (Nymphalidae), compared to the generalist Spilosoma congrua Wlk. (Arctiidae). However, plant chemistry was significantly affected by herbivory. Herbivory by both caterpillar species induced iridoid glycosides and resulted in an increased concentration of catalpol and an increase in the proportion of catalpol relative to total iridoid glycosides in those plants exposed to herbivores compared to controls. In addition, plants exposed to specialist caterpillars had higher concentrations of catalpol and a higher proportion of total iridoids that was catalpol than those exposed to generalist caterpillars. Overall, in this experiment, leaf age explained three times as much of the variation in iridoid glycosides as did plant genotype and herbivory. The results of the two parts of this experiment show that iridoid glycoside concentration in P. lanceolata leaves decreases over time, and that exposure to herbivores induces increased concentrations of iridoid glycosides.

Response of generalist and specialist insects to qualitative allelochemical variation

We examined the effects of a set of four biosynthetically related iridoid glycosides, aucubin, catalpol, loganin, and asperuloside, on larvae of a generalist,Lymantria dispar (Lymantriidae), the gypsy moth, and an adapted specialist, the buckeye,Junonia coenia (Nymphalidae). In general,L. dispar grew and survived significantly less well on artificial diets containing iridoid glycoside, compared to a control diet without iridoid glycosides. In choice tests, previous exposure to a diet containing iridoid glycosides caused larvae subsequently to prefer iridoid glycoside-containing diets even though they were detrimental to growth and survival. In contrast,J coenia larvae grew and survived better on diets with aucubin and catalpol, the two iridoid glycosides found in the host plantPlantago lanceolata (Plantaginaceae), than on diets with no iridoid glycoside or with loganin and asperuloside. The results of choice tests of diets with and without iridoid glycosides and between diets with different iridoid glycosides reflected these differences as well. These results are discussed in terms of (1) differences between generalists and specialists in their response to qualitative variation in plant allelochemical content, (2) the induction of feeding preferences, and (3) the evolution of qualitative allelochemical variation as a plant defense.

VARIATION IN FOOD QUALITY AND TEMPERATURE CONSTRAIN FORAGING OF GREGARIOUS CATERPILLARS

We examined the separate and combined effects of changing food quality, thermal environment, and larval behavior on the foraging pattern of a gregarious caterpillar, Hemileuca lucina (Saturniidae), which specializes on the shrub Spiraea latifolia (Rosaceae). First-instar larvae, which encounter only new leaves under field conditions, showed no preference for new leaves over mature ones. In contrast, third-instar larvae, which normally encounter both leaf types, exhibited a preference for new leaves, in both 12-h and instar- length choice tests. In laboratory tests, larvae in groups were more likely to survive and exhibited less variation in biomass gained than solitary larvae. An experiment was con- ducted to determine the interactions of temperature (with daytime temperatures of 200, 250, or 300C), group size (solitary or in groups of 10), and diet (new or mature leaves) on third-instar larvae. The effect of diet and group size on relative growth rate depended on the daytime temperature. At a daytime temperature of 200, representing overcast and cool spring conditions, growth rates were similar for all group size-diet combinations. Thus, at such a cool temperature, there was no advantage in feeding on new leaves or in larval groups. In contrast, solitary larvae fed mature leaves gained less mass and developed more slowly than the other group size-diet combinations at the mean maximal day temperature (250) for field conditions. At 300, representing the average temperature reached through basking, larvae eating new leaves grew significantly faster than those on mature leaves, regardless of larval group size. These results indicate that in the absence of predators, in which case larvae eat new leaves, bask, and remain aggregated, the larvae can grow twice as fast and attain more biomass per instar than solitary larvae eating mature leaves in the shade, which is often the case when they avoid predators.

Iridoid glycosides and host-plant specificity in larvae of the buckeye butterfly, Junonia coenia (Nymphalidae)

Larvae of the buckeye,Junonia coenia (Nymphalidae) feed primarily on plants in four families: Scrophulariaceae, Plantaginaceae, Verbenaceae, and Acanthaceae. These plant families have in common the presence of a group of plant secondary compounds, the iridoid glycosides. Larvae were reared on three plant species and two artificial diets, one with and one without iridoid glycosides.Larvae grew poorly and had low survivorship on the artificial diet without iridoid glycosides, while growth and survival on the artificial diet with iridoid glycosides was comparable to that on plants. Choice tests using artificial diets with and without iridoid glycosides showed that larvae: (1) chose diets with iridoid glycosides (in the form of a crude extract or pure compound) over a diet without; (2) showed no preference between the diet with the crude extract and that with pure iridoid glycoside, and (3) preferred the artificial diet with ground leaves of the host plant,Plantago lanceolata, over the diet with pure iridoid glycosides. The artificial diet that larvae had been reared on prior to these tests had no effect on subsequent larval preferences in the choice tests.

Genetic variation in defensive chemistry in Plantago lanceolata (Plantaginaceae) and its effect on the specialist herbivore Junonia coenia (Nymphalidae)

To examine genetic variation in defensive chemistry within and between natural populations of Plantago lanceolata, we performed a greenhouse experiment using clonal replicates of 15 genotypes from each of two populations, from a mowed lawn and an abandoned hayfield. Replicates of each genotype were harvested for determinations of aboveground biomass and leaf chemical content either at the beginning of the experiment (initial controls), after exposure to herbivory by larvae of Junonia coenia, a specialist on P. lanceolata (herbivory treatment), or at the end of the experiment without exposure to herbivory (final controls). Allocation to the iridoid glycosides aucubin and catalpol and the phenylpropanoid glycoside verbascoside displayed significant genetic variation within and between populations, and differed with leaf age. Significant genotypextreatment interactions indicated genetic variation in response of leaf chemistry to the treatments. There was no evidence for a cost of allocation to chemical defense: genetic correlations within and between chemical pathways and between defensive chemicals and aboveground growth were positive or nonsignificant. Although iridoid glycosides are known to be qualitative feeding stimulants for J. coenia, multiple regression of larval survivorship on leaf chemical content and shoot biomass indicated that larvae had lower survivorship on P. lanceolata ge-notypes with higher concentrations of aucubin in the leaves. Larval survivorship was unaffected by levels of catalpol and verbascoside. Thus, although specialist herbivores may respond to defensive chemicals as qualitative feeding stimulants, they do not necessarily have higher fitness on plant genotypes containing higher concentrations of these chemicals.

Chemical variation within and between individuals ofPlantago lanceolata (Plantaginaceae).

Variation in concentrations of leaf nitrogen and iridoid glycosides was examined in replicate plants of five genotypes ofPlantago lanceolata (Plantaginaceae) grown in an experimental garden. Nitrogen concentration and iridoid glycoside concentration were affected by leaf age. New leaves had nitrogen concentrations 1.7 to 2.7 times higher than mature leaves. Catalpol concentration was highest in new and intermediate-aged leaves. The concentration of aucubin, the biosynthetic precursor to catalpol, was higher in intermediate-aged leaves than in mature leaves, in three of five genotypes. Consequently, the proportion of aucubin relative to total iridoid glycosides increased as leaves aged. Concentration of iridoid glycosides was not correlated with plant size. Plant genotype significantly affected concentration of nitrogen and iridoid glycosides, as well as plant size. Thus, major indicators of hostplant quality for insect herbivores varied considerably both within and among plant genotypes and individuals.

UNPALATABILITY AS A DEFENSE STRATEGY OF EUPHYDRYAS PHAETON (LEPIDOPTERA: NYMPHALIDAE)

The use of unpalatability as a defense strategy of butterflies has been studied since the time of Bates (1862). Most studies have focused primarily on the role of unpalatability in mimicry systems, rather than on the defensive function of unpalatability itself. Unpalatable butterflies have been found primarily in five major groups: (1) the tribe Troidini of the family Papilionidae, whose members feed on plants in the Aristolociaceae, and the following subfamilies of the Nymphalidae: (2) the Danainae, feeding on milkweeds (Asclepiadaceae); (3) the Ithomiinae, which feed on solanaceous plants; (4) the Acraeinae and (5) the Heliconiinae, both of which feed on plants in the family Passifloraceae. Some pierids (Pieridae) are unpalatable as well (Finn, 1895, 1897a, 1897b; Swynnerton, 1919; Pough and Brower, 1977), but unpalatability does not seem to be common in this group. Research on unpalatable North American butterflies has considered primarily three species: the Monarch, Danaus plexippus (Danainae), and the Queen, Danaus gilippus (Danainae), both feeding on milkweeds which contain cardiac glycosides (J. Brower, 1958a, 1958c; Roeske et al., 1975); and the Pipevine Swallowtail, Battus philenor (Papilionidae), which sequesters aristolochic acids from its foodplants in the Aristolochiaceae (J. Brower, 1958b; Rothschild et al., 1970; Platt et al., 1971). Other North American butterflies have been neglected in studies on unpalatabil-

Pattern of Leaf Damage Affects Fitness of the Annual Plant Raphanus Sativus (Brassicaceae)

We investigated how the pattern of leaf damage influences reproduction, growth, and allocation in the wild radish, Raphanus sativus (Brassicaceae). We removed an equivalent leaf area from plants with four leaves in five treatments ranging from con- centrated to dispersed damage: one entire mature leaf removed, one entire new leaf re- moved, 50% of two mature leaves removed, 50% of two new leaves removed, and 25% of all four leaves removed. Plants in a control group were undamaged. Reproduction, growth, and allocation were not affected by the age of the damaged leaf. However, the pattern of leaf damage significantly affected our three measures of plant fitness: the number of flowers produced, the reproductive biomass, and the total biomass. Plants in the treatment in which the damage was most dispersed had significantly higher flower number, reproductive biomass, and total biomass than an intermediate damage treatment and significantly more reproductive biomass than the concentrated damage treatment. There were no significant differences between the concentrated and intermediate damage treatments and no differ- ences between the dispersed damage treatment and the undamaged control. Our data indicate that more dispersed damage is less detrimental to the plant than more concentrated damage. Therefore, the pattern of leaf damage must be considered in determining the impact of herbivores on plant performance.

Immunological cost of chemical defence and the evolution of herbivore diet breadth

Selective pressures from host plant chemistry and natural enemies may contribute independently to driving insect herbivores towards narrow diet breadths. We used the specialist caterpillar, Junonia coenia (Nymphalidae), which sequesters defensive compounds, iridoid glycosides, from its host plants to assess the effects of plant chemistry and sequestration on the larval immune response. A series of experiments using implanted glass beads to challenge immune function showed that larvae feeding on diets with high concentrations of iridoid glycosides are more likely to have their immune response compromised than those feeding on diets low in these compounds. These results indicate that larvae feeding on plants with high concentrations of toxins might be more poorly defended against parasitoids, while at the same time being better defended against predators, suggesting that predators and parasitoids can exert different selective pressures on the evolution of herbivore diet breadth.

Fate of ingested iridoid glycosides in lepidopteran herbivores.

Thin-layer chromatography was used to follow the fates of iridoid glycosides ingested by four species of lepidopteran herbivores. These four species differed in their feeding strategy, ranging from generalist to monophagous specialist; and in their predator avoidance strategy, ranging from cryptic and palatable to aposematic and unpalatable. The fates of the iridoid glycosides ranged from sequestration by the unpalatable specialist,Euphydryas phaeton (Nymphalidae); to passage into the hemolymph and eventual elimination in the meconium by the specialistsJunonia coenia (Nymphalidae) andCeratomia catalpas (Sphingidae); to elimination of the intact compounds in the feces of the generalist feeder,Lymantria dispar (Lymantriidae).