PREFERENCE HIERARCHIES AND THE ORIGIN OF GEOGRAPHIC SPECIALIZATION IN HOST USE IN SWALLOWTAIL BUTTERFLIES

Four allopatric populations of the widely distributed western anise swallowtail butterfly, Papilio zelicaon, use different plant genera as hosts, but simultaneous choice experiments showed that these populations have diverged only slightly in oviposition preference. Of the four populations—two from southeastern Washington State, one from coastal southwestern Washington State, and one from central California—three use hosts that are not available to any of the others. Although variation for the degree of preference for particular plant species occurred within and among populations, all four populations ranked hosts in the same overall order. Monophagy on a local, low-ranking host outside the range of high-ranking hosts did not necessarily lead to the loss of preference for those high-ranking hosts, thereby indicating that the high-ranking hosts would still be accepted, and in some cases even preferred, if a population encountered them again. Hence, the overall preference hierarchy among P. zelicaon populations appears to be evolutionarily conservative. Analyses of differences among families within the California population indicated that increased preference for some hosts is inversely correlated, whereas preference for other hosts may be uncorrelated. Positive correlations may also occur but were not observed among the plant species tested. Overall, the results indicate local monophagy on different plant species in P. zelicaon has not involved major reorganizations in the preference hierarchy of ovipositing females, even in populations that may have fed on a low-ranking host for many generations. Instead, small increases in preference for local hosts have occurred within an evolutionarily conservative preference hierarchy.     

EVOLUTIONARY GENETICS OF OVIPOSITION PREFERENCE IN SWALLOWTAIL BUTTERFLIES

Models of the evolution of host shifts and speciation in phytophagous insects are often built upon the assumption that host selection is under simple genetic control, perhaps even a single locus. The genetic basis for differences in host-plant preference by ovipositing insects was investigated using two closely related species of swallowtail butterfly, Papilio oregonius and P. zelicaon, which differ in the plant families on which females oviposit. Both species had been shown previously to vary within populations in host selection. A combination of analyses using reciprocal interspecific crosses and isofemale strains within populations indicated that oviposition preference in these species is determined significantly by one or more loci on the X chromosome, which female Lepidoptera inherit only paternally. Hence, preferences in hybrid females tended toward the paternal species. This is the first insect group for which partial control of oviposition preference has been localized onto a particular chromosome. In addition, one or more loci on another chromosome(s) appear to contribute to preference, as indicated by the partially intermediate preferences of some hybrid crosses. The overall differences in preference in the reciprocal interspecific hybrids were restricted to differences in the distribution of eggs laid among the local host plants of these two Papilio species; the reciprocal crosses did not differ in the small percentage of eggs laid on a novel potential host species. The variation in host selection found among the isofemale strains reinforces earlier results for these strains, indicating that there is genetic variation in host selection within these populations. Overall, the results indicate that the evolution of oviposition preference in these species involves genetic changes at two or more chromosomes with the X chromosome playing an important role in determining preference.     

VARIATION IN PREFERENCE AND SPECIFICITY IN MONOPHAGOUS AND OLIGOPHAGOUS SWALLOWTAIL BUTTERFLIES

Although variation in oviposition preference and specificity for host plants has been demonstrated within populations of a variety of oligophagous insect species, it is unknown whether genetic variation in host choice is lost within populations of monophagous species. Analysis of a locally monophagous butterfly species, Papilio oregonius, and a locally oligophagous species, P. zelicaon, showed significant variation in oviposition preference within populations of both species. Females of both species chose primarily their native hosts. Nonetheless, the percentages of eggs laid by individual females among the plant species and the number of plant species on which individual females laid eggs differed significantly among isofemale strains within populations. Moreover, some females within all isofemale strains of both species laid a few eggs on Foeniculum vulgare, an umbelliferous species that does not occur in the native habitats of these populations but is a host for Papilio species in other geographic areas. The results suggest that local monophagy and oligophagy in these species reflect the relative ranking among potential plant species. Both populations harbor variation in oviposition choice that could allow for host shifts if these populations invaded new habitats.     

Oviposition preference and larval performance of North American monarch butterflies on four Asclepias species

Monarch butterflies, Danaus plexippus L. (Lepidoptera: Nymphalidae), occur world‐wide and are specialist herbivores of plants in the milkweed family (Asclepiadaceae). In North America, two monarch populations breed east and west of the continental divide in areas populated by different host plant species. To examine the population variation in monarch responses to different Asclepias species, we measured oviposition preference and larval performance among captive progeny reared from adult butterflies collected in eastern and western North America. Host plant use was evaluated using two milkweed species widely distributed in eastern North America (A. incarnata and A. syriaca), and two species common to western North America (A. fascicularis and A. speciosa). We predicted that exposure to different host plant species in their respective breeding ranges could select for divergent host use traits, so that monarchs should preferentially lay more eggs on, and larvae should perform better on, milkweed species common to their native habitats. Results showed that across all adult female butterflies, oviposition preferences were highest for A. incarnata and lowest for A. fascicularis, but mean preferences did not differ significantly between eastern and western monarch populations. Larvae from both populations experienced the highest survival and growth rates on A. incarnata and A. fascicularis, and we again found no significant interactions between monarch source population and milkweed species. Moreover, the average rank order of larval performance did not correspond directly to mean female oviposition preferences, suggesting that additional factors beyond larval performance influence monarch oviposition behavior. Finally, significant family level variation was observed for both preference and performance responses within populations, suggesting an underlying genetic variation or maternal effects governing these traits.     

Butterfly oviposition preference is not related to larval performance on a polyploid herb

The preference–performance hypothesis predicts that female insects maximize their fitness by utilizing host plants which are associated with high larval performance. Still, studies with several insect species have failed to find a positive correlation between oviposition preference and larval performance. In the present study, we experimentally investigated the relationship between oviposition preferences and larval performance in the butterfly Anthocharis cardamines. Preferences were assessed using both cage experiments and field data on the proportion of host plant individuals utilized in natural populations. Larval performance was experimentally investigated using larvae descending from 419 oviposition events by 21 females on plants from 51 populations of two ploidy types of the perennial herb Cardamine pratensis. Neither ploidy type nor population identity influenced egg survival or larval development, but increased plant inflorescence size resulted in a larger final larval size. There was no correlation between female oviposition preference and egg survival or larval development under controlled conditions. Moreover, variation in larval performance among populations under controlled conditions was not correlated with the proportion of host plants utilized in the field. Lastly, first instar larvae added to plants rejected for oviposition by butterfly females during the preference experiment performed equally well as larvae growing on plants chosen for oviposition. The lack of a correlation between larval performance and oviposition preference for A. cardamines under both experimental and natural settings suggests that female host choice does not maximize the fitness of the individual offspring.     

High-density populations of diamondback moth have broader host-plant diets

Although ecological specialists exploit a relatively limited resource base, it is unclear whether specialization limits local population density. Here, we focus on the relationship between diet specialization and local population density of a phytophagous insect, the diamondback moth, Plutella xylostella L. (Lepidoptera: Plutellidae). We compared key life history traits and food plant use among five pairs of high- and low-density populations from widely separate geographical areas. Moths from populations that persist at high densities were more generalized in food plant use than moths from low-density populations. In particular, the oviposition preference and larval performance of moths from some high-density populations were less variable across a suite of food plants, suggesting that moths from high- density populations had a broader diet. In addition, low- density populations were less similar to each other, exhibiting opposing preferences for particular plant species. Hence diet breadth may explain some of the persistent differences in the population density of diamondback moths in the field, consistent with the idea that ecological specialization may be generally associated with population density. Received: 19 December 1996 / Accepted: 2 May 1997     

The evolution of diet breadth: Monophagy and polyphagy in swallowtail butterflies

Much of the world's biodiversity has resulted from specialization of insect populations onto different plant species, partially through evolution of preference in ovipositing females. Here I report an experimental analysis of how an oviposition preference hierarchy has evolved during the evolutionary diversification of an insect group to produce taxa ranging from monophagous to polyphagous. Tests on the Papilio machaon group of swallowtail butterflies show that the preference hierarchy for plant species is evolutionarily dynamic within this species complex, yet constrained among most populations within species, creating a geographic mosaic of populations differing to various degrees in patterns of host preference. The results indicate that different diet breadths can evolve within a group of closely-related species through a combination of conservatism in preference hierarchy among some populations, occasional but rare rearrangements in preference among others, correlations in preference for some plant species, and availability of similarly ranked hosts.     

Phenological matching rather than genetic variation in host preference underlies geographical variation in host plants used by orange tip butterflies

An insect species that shows variation in host species association across its geographical range may do so either because of local adaptation in host plant preference of the insect or through environmentally or genetically induced differences in the plants, causing variation in host plant suitability between regions. In the present study, we experimentally investigate the host plant preference of Anthocharis cardamines (orange tip butterfly) in two populations from the UK and two from Sweden. Previous reports indicate that A. cardamines larvae are found on different host plant species in different regions of the UK, and some variation has been reported in Sweden. Host plant choice trials showed that females prefer to oviposit on plants in an earlier phenological stage, as well as on larger plants. When controlling for plant phenological stage and size, the host species had no statistically significant effect on the choice of the females. Moreover, there were no differences in host plant species preference among the four butterfly populations. Based on our experiment, the oviposition choice by A. cardamines mainly depends on the phenological stage and the size of the host plant. This finding supports the idea that the geographical patterns of host–plant association of A. cardamines in the UK and Sweden are consequences of the phenology and availability of the local hosts, rather than regional genetic differences in the host species preference of the butterfly.     

Reciprocal latitudinal clines in oviposition behavior ofPapilio glaucus andP. canadensis across the Great Lakes hybrid zone: possible sex-linkage of oviposition preferences

Summary The eastern tiger swallowtail butterfly,Papilio glaucus, is the most polyphagous of all Papilionidae species. While diverse larval detoxication abilities are known for bothPapilio glaucus and the closely relatedP. canadensis, the factors affecting oviposition preferences in adult females are unknown. These congeneric species were studied to determine the extent of oviposition mistakes on toxic plants. We were also interested in comparing the geographic patterns of variation in oviposition preferences and the genetic basis of these differences. We conducted oviposition three-choice studies with the Canadian tiger swallowtail butterfly,Papilio canadensis, and the Eastern tiger swallowtail,Papilio glaucus, giving them the choice of (1) tuliptree,Liriodendron tulipifera, which is toxic to virtually all populations (P. canadensis) north of the Great Lakes Region of North America, (2) quaking aspen,Populus tremuloides, which is toxic to essentially all populations (P. glaucus) south of the Great Lakes, and (3) black cherry,Prunus serotina, which is an excellent foodplant for all members of thePapilio glaucus group, but which does not occur at latitudinal extremes of North America (in Alaska and most of Canada or the southern half of Florida). Handpaired interspecific hybrids were tested under the same experimental design to evaluate the possibility of sexlinked oviposition behavior. There was considerable variability in the choice of plants by individual butterflies, but a general trend suggesting that the females of each species had a lower preference for the plant toxic to their larvae. More than 6000 oviposition bouts were counted from 37 differentp. canadensis and 54p. glaucus females along a latitudinal transect of approximately 5000 km from Alaska south through the Great Lakes hybrid zone region to southern Florida. While not exceptionally high anywhere, the preference for aspen (Salicaceae) declined precipitously in central Michigan (45° N latitude) and remained very low (5–12%) in all locations southward to Florida, whereas we observed a reciprocal trend in preference for tuliptree (Magnoliaceae) which was greatest in Florida (87% of all eggs) and steadily declined northward across the Great Lakes region. Cherry was selected in these 3-choice tests at a relatively consistent and low frequency at all latitudes. Fixed allele differences in sex-linked (LDH and PGD) and autosomal (HK) electromorphs are known forP. glaucus andP. canadensis. Our electrophoretic data suggest that the preference of an individual female for aspen is not simply a characteristic of the northern species (P. canadensis) but can occur inP. glaucus females. The reciprocal situation is also evident in Northern Michigan and Wisconsin females (scored electrophoretically and morphologically asP. canadensis) which sometimes exhibit a clear preference for the toxic tuliptree. In fact, Alaskan populations ofP. canadensis chose tuliptree for about 52% of their eggs, even though none of their offspring has ever survived on this plant species in laboratory studies. We conclude that even with distinctive latitudinal trends, a considerable amount of local variation in relative oviposition preference exists among individuals of these polyphagous species. BothP. glaucus andP. canadensis will lay eggs on toxic plants. It appears that factors selecting against oviposition on toxic tuliptrees have been minimal (relative to other factors) in Alaska and somewhat stronger in the Great Lakes hybrid zone. It is in this zone of contact with tuliptree where selection against theP. canadensis populations ovipositing on tuliptree may be strong due to high larval mortality when such natural mistakes are made. We do not know whether behavioral preference changes evolutionarily preceded or followed the development of specific physiological detoxication abilities for tuliptree or quaking aspen. However, for bothP. canadensis andP. glaucus the occurrence of oviposition mistakes on toxic plants by adults extends geographically well beyond the larval detoxication abilities of their offspring. Hybrid female offspring of pairings with Michigancanadensis females andglaucus males show distinct preferences for tuliptree, suggesting that oviposition may be controlled by a factor (or factors) on the sex chromosome. Unfortunately we were unable to obtain reciprocal hybrids to evaluate the possibility of sex-linked aspen preference.     

INCORPORATION OF A EUROPEAN WEED INTO THE DIET OF A NORTH AMERICAN HERBIVORE

Populations of the butterfly Euphydryas editha living within a 30 times 100–km region on the eastern slope of the Sierra Nevada range were compared for oviposition preference and ability of larvae to grow and survive on two host plants, Collinsia parviflora and Plantago lanceolata. Since its introduction approximately 100 years ago, P. lanceolata has been incorporated in the diet of E. editha in one of the study populations. The populations differed in oviposition preference; only the population that uses P. lanceolata contains some individuals that prefer P. lanceolata. Larvae from two populations, one using both P. lanceolata and C. parviflora, the other using only C. parviflora, were not found to differ in relative abilities to grow or survive on P. lanceolata. The potential for E. editha to use P. lanceolata appears in populations that have had no prior exposure to this plant, while oviposition preference for this plant has evolved in the population in which the plant now grows.     

Patterns and processes of nestedness in a Great Basin butterfly community

We examined nestedness and potential mechanisms causing that distributional pattern in resident butterfly communities of the Toiyabe Range, a mountain range in the central Great Basin of western North America. We tested whether life history characteristics, including habitat use and vagility, affected the relative degree of nestedness or mean species incidence. We also tested whether nestedness at the level of individual species was independent of life history. Relationships between distributional patterns and habitat use, particularly in ecologically sensitive riparian areas, are relevant to ongoing conservation planning in the Great Basin. The distributional pattern of the 68 resident butterfly species in 19 Toiyabe Range canyons was significantly nested, as was the distribution of all functional subgroups that we tested. Life history affected neither relative nestedness of species groups nor mean species incidence. More than 80% of the individual butterfly species that inhabit the Toiyabe Range had distributions that were more nested than expected. Colonization does not appear to have played an important role in determining the composition of butterfly communities in Toiyabe Range canyons. Likewise, selective dispersal has probably played a minor role in producing nested distributions of Toiyabe Range butterflies. Our results suggest either that impacts to riparian areas are not jeopardizing species viability, or that highly sensitive butterfly species have already been extirpated from the Toiyabe Range. Received: 15 February 1998 / Accepted: 19 December 1998     

Specialist pollinating seed predator exhibits oviposition strategy consistent with optimal oviposition theory

相似文献   

Conservative resource utilization in the common blue butterfly–evidence for low costs of accepting absent host plants?

We studied host plant preference of the common blue butterfly, Polyommatus icarus , and larval performance on two different host plants, Oxytropis campestris and Lotus corniculatus . The study species is a small lycaenid butterfly believed to be relatively sedentary. The study populations originated from two different and widely separated geographical areas. In one area both hosts are naturally occurring, with O. campestris being most abundant at the study sites, in the other area only one of the host plants, L. corniculatus , is present. There was no difference in oviposition preference or larval performance between populations from the two different areas. Hence, P. icarus from sites dominated by O. campestris has not evolved a higher preference for or better performance on this host plant. More surprisingly, P. icarus from the area were O. campestris is completely absent has retained not only good larval performance on this host plant but also high female preference for it. This conservatism at a large geographical scale is seen even though there seems to be genetic variation present in both populations, at least for preference but perhaps also for performance. We suggest that such lack of variation in resource utilization between populations may be evidence for weak selection against "preferences" for plants that are rare or absent. A combination of other constraining factors may also contribute to some degree, especially stepping-stone gene flow between populations.     

Salicaceae detoxification abilities in Florida tiger swallowtail butterflies (Papilio glaucus maynardi Gauthier): Novel ability or Pleistocene holdover?

Abstract Florida populations of the eastern tiger swallowtail butterfly, Papilio glaucus L., have unique morphological features and ecological adaptations that have contributed to their subspecies status (P. g. maynardi Gauthier). We describe geographically unique abilities for detoxification of Carolina willow, Salix caroliniana Michx. (Salicaceae), for several Florida populations of P. g. maynardi. Of all the approximately 570 worldwide species of the Papilionidae, such Salicaceae detoxification abilities exist only in the allopatric North American western and northernmost species (P. rutulus Lucas, P. eurymedon Lucas and P. canadensis Rothschild & Jordan). Females of P. glaucus collected from populations in southeastern USA were examined for oviposition preference in 5‐choice assays, and displayed a low preference for Salicaceae (<5%), but larvae from Florida populations exhibited a high survival (>60%) on these plants. Detoxification abilities have previously shown to be autosomally inherited, and can be transferred via natural or hand‐paired interspecific hybrid introgression. However, these Florida populations are at least 700–1 500 km from the nearest hybrids or the hybrid species, P. appalachiensis Pavulaan & Wright, which possess these detoxification abilities. In any case, the Z ( = X)‐linked oviposition preferences for Salicaceae are lacking in these Florida populations, illustrating genetic independence of oviposition preference determination and larval survival/performance abilities. The origins of detoxification abilities are unlikely to be due to recent climate‐driven introgression, and may represent ancestral trait carry‐overs from interglacial refugium populations of the Pleistocene epoch.     

Geographically multifarious phenotypic divergence during speciation

Speciation is an important evolutionary process that occurs when barriers to gene flow evolve between previously panmictic populations. Although individual barriers to gene flow have been studied extensively, we know relatively little regarding the number of barriers that isolate species or whether these barriers are polymorphic within species. Herein, we use a series of field and lab experiments to quantify phenotypic divergence and identify possible barriers to gene flow between the butterfly species Lycaeides idas and Lycaeides melissa. We found evidence that L. idas and L. melissa have diverged along multiple phenotypic axes. Specifically, we identified major phenotypic differences in female oviposition preference and diapause initiation, and more moderate divergence in mate preference. Multiple phenotypic differences might operate as barriers to gene flow, as shown by correlations between genetic distance and phenotypic divergence and patterns of phenotypic variation in admixed Lycaeides populations. Although some of these traits differed primarily between species (e.g., diapause initiation), several traits also varied among conspecific populations (e.g., male mate preference and oviposition preference).     

Preference and performance of Lepidoptera varies with tree age in juniper woodlands

1. As trees age, they undergo significant physiological and morphological changes. Nevertheless, tree ontogeny and its impacts on herbivores are often overlooked as determinants of plant–herbivore population dynamics and the strength of plant–herbivore interactions. 2. Juniperus (Cupressaceae) is a dominant, long‐lived conifer that serves as the sole host to a specialised assemblage of caterpillars. Over the past 150 years, several juniper species in western North America have expanded their geographic occupancy at local and regional scales, which has resulted in an increase in the number of immature trees on the landscape. Using assays in the laboratory, the effects of tree ontogeny on caterpillar performance and oviposition preference for two juniper specialist caterpillars, Callophrys gryneus (Lycaenidae) and Glena quinquelinearia (Geometridae), were examined. The study considered whether responses to tree ontogeny were consistent across caterpillar species and juniper host species. 3. Tree age was found to be a reliable predictor of caterpillar performance, with caterpillars developing more quickly and growing larger when fed foliage from young trees. Differences in the phytochemical diversity between foliage from trees of different ages might help to explain observed differences in caterpillar performance. Interestingly, the specialist butterfly, C. gryneus, displayed an oviposition preference for foliage from old‐growth Juniperus osteosperma trees, despite the fact that larvae of this species performed poorly on older trees. 4. It is concluded that young juniper trees are an important resource for the specialised Lepidopteran community and that tree ontogeny is an important component of intraspecific variation, which contributes to the structure of plant–herbivore communities.     

Patch dynamics ofGlaucopsyche lygdamus (Lycaenidae): correlations between butterfly density and host species diversity

Glaucopsyche lygdamus egg densities were surveyed over a 2000-m section of Gold Creek and at 30 different isolated patches in the Gold Basin drainage in Colorado. Host plant numbers and diversity were quantified, as well as other variables potentially influencing butterfly population size, such as patch size and isolation. Egg densities correlated significantly only with measures of host species diversity. Patches consisting of a single host species, no matter how large, did not support high butterfly densities, but patches of multiple, equitably distributed host species did. The most likely explantation, in light of oviposition preference and larval performance data accumulated for this butterfly species, is that host species diversity is necessary for the persistence ofG. lygdamus populations, because alternative host species buffer population losses during poor or unusual years. The dependence of both ovipositing butterflies and developing larvae on the ephemeral, young, host plant flowers make the butterfly especially vulnerable to year-to-year variation in host plant availability and quality.     

Diet divergence in two sympatric congeneric butterflies: Community or species level phenomenon?

Summary Two species ofEuphydryas butterflies were studied in California, USA, and showed considerable diet overlap at the species level. They utilize many of the same plant genera for oviposition. However,E. editha is less likely to use woody perennials than isE. chalcedona.Both butterfly species are known to specialize on different host plants in different populations, so species level divergence may not be a good predictor of community level divergence. Within five communities,E. editha andE. chalcedona showed no dietary overlap. A major component of the niche ofE. editha in one community was occupied byE. chalcedona in a second community, even though both butterfly species occupied both communities. These resource use patterns indicate that community level interactions may affect diet divergence. The degree to which divergence within communities is greater (or less) than expected from a species level comparison may be used to provide a measure of community organization. Equations are given in the Appendix for calculating overlap probabilities from presence/absence types of data; in this study, presence is oviposition on a particular plant species, absence is no oviposition on that plant species. Given the various assumptions of the model,E. editha andE. chalcedona show significant community level components of their dietary divergence in the areas studied. However, in some other communitiesE. editha andE. chalcedona do share host plant species. Therefore, we could not demonstrate community level divergence conclusively, nor has this been demonstrated for any other pair of insect herbivore species. We do not know whether this is because the phenomenon is truly rare or just very hard to demonstrate.     

Plant-mediated decisions by an herbivore affect oviposition pattern and subsequent egg parasitism

Natural enemies are important mortality factors for herbivores and thus may influence herbivore population dynamics. In response to natural enemy pressure, herbivores can alter life history decisions, such as oviposition behavior, so that offspring are protected from natural enemies. One such strategy is to deposit eggs into structures where vulnerability to natural enemies is reduced or eliminated, i.e., use enemy-free space. The plant bug, Lygus lineolaris (Palisot de Beauvois), is native to North America and has a broad host range (>350 plant species), including crops. This bug’s eggs are attacked by a native parasitoid, Anaphes iole Girault, and parasitism levels vary greatly among host plant species. Weed hosts are critical to contemporary L. lineolaris life history because they serve as an ecological bridge from one crop growing season to the next. We investigated the egg distribution pattern of L. lineolaris on 11 host plant species (nine weeds and two crops), and parasitism by A. iole, to determine whether oviposition choices by L. lineolaris females protect their eggs from parasitism and to demonstrate the mechanism of this protection. Our results indicate that the reproductive structures of Erigeron annuus, as well as those of several other host plant species, provide a refuge from parasitism for most L. lineolaris eggs. This refuge is due to the deposition of host eggs deeper in plant tissue than the length of the ovipositor of A. iole. Also, overall parasitism levels were greater on non-Asteraceae host plant species compared with host plant species belonging to Asteraceae. Oviposition site choice by female bugs appears to be a selective strategy to take advantage of enemy-free space.     

Timing of flowering and intensity of attack by a butterfly herbivore in a polyploid herb

Timing of plant development both determines the abiotic conditions that the plant experiences and strongly influences the intensity of interactions with other organisms. Plants and herbivores differ in their response to environmental cues, and spatial and temporal variation in environmental conditions might influence the synchrony between host plants and herbivores, and the intensity of their interactions. We investigated whether differences in first day of flowering among and within 21 populations of the polyploid herb Cardamine pratensis influenced the frequency of oviposition by the butterfly Anthocharis cardamines during four study years. The proportion of plants that became oviposited upon differed among populations, but these differences were not related to mean flowering phenology within the population in any of the four study years. Attack rates in the field were also not correlated with resistance to oviposition estimated under controlled conditions. Within populations, the frequency of butterfly attack was higher in early‐flowering individuals in two of the four study years, while there was no significant relationship in the other 2 years. Larger plants were more likely to become oviposited upon in all 4 years. The effects of first flowering day and size on the frequency of butterfly attack did not differ among populations. The results suggest that differences in attack intensities among populations are driven mainly by differences in the environmental context of populations while mean differences in plant traits play a minor role. The fact that within populations timing of flowering influenced the frequency of herbivore attack only in some years and suggests that herbivore‐mediated selection on plant phenology differs among years, possibly because plants and herbivores respond differently to environmental cues.