Changes in landscape and climate in Mexico and Texas reveal small effects on migratory habitat of monarch butterflies (Danaus plexippus).

The decline of the iconic monarch butterfly (Danaus plexippus) in North America has motivated research on the impacts of land use and land cover (LULC) change and climate variability on monarch habitat and population dynamics. We investigated spring and fall trends in LULC, milkweed and nectar resources over a 20-year period, and ~ 30 years of climate variables in Mexico and Texas, U.S. This region supports spring breeding, and spring and fall migration during the annual life cycle of the monarch. We estimated a - 2.9% decline in milkweed in Texas, but little to no change in Mexico. Fall and spring nectar resources declined < 1% in both study extents. Vegetation greenness increased in the fall and spring in Mexico while the other climate variables did not change in both Mexico and Texas. Monarch habitat in Mexico and Texas appears relatively more intact than in the midwestern, agricultural landscapes of the U.S. Given the relatively modest observed changes in nectar and milkweed, the relatively stable climate conditions, and increased vegetation greenness in Mexico, it seems unlikely that habitat loss (quantity or quality) in Mexico and Texas has caused large declines in population size or survival during migration.

Late-instar monarch caterpillars sabotage milkweed to acquire toxins, not to disarm plant defence.

Sabotaging milkweed by monarch caterpillars (Danaus plexippus) is a famous textbook example of disarming plant defence. By severing leaf veins, monarchs are thought to prevent the flow of toxic latex to their feeding site. Here, we show that sabotaging by monarch caterpillars is not only an avoidance strategy. While young caterpillars appear to avoid latex, late-instar caterpillars actively ingest exuding latex, presumably to increase sequestration of cardenolides used for defence against predators. Comparisons with caterpillars of the related but non-sequestering common crow butterfly (Euploea core) revealed three lines of evidence supporting our hypothesis. First, monarch caterpillars sabotage inconsistently and therefore the behaviour is not obligatory to feed on milkweed, whereas sabotaging precedes each feeding event in Euploea caterpillars. Second, monarch caterpillars shift their behaviour from latex avoidance in younger to eager drinking in later stages, whereas Euploea caterpillars consistently avoid latex and spit it out during sabotaging. Third, monarchs reared on detached leaves without latex sequestered more cardenolides when caterpillars imbibed latex offered with a pipette. Thus, we conclude that monarch caterpillars have transformed the ancestral 'sabotage to avoid' strategy into a 'sabotage to consume' strategy, implying a novel behavioural adaptation to increase sequestration of cardenolides for defence.

Testing the selective sequestration hypothesis: Monarch butterflies preferentially sequester plant defences that are less toxic to themselves while maintaining potency to others.

Herbivores that sequester toxins are thought to have cracked the code of plant defences. Nonetheless, coevolutionary theory predicts that plants should evolve toxic variants that also negatively impact specialists. We propose and test the selective sequestration hypothesis, that specialists preferentially sequester compounds that are less toxic to themselves while maintaining toxicity to enemies. Using chemically distinct plants, we show that monarch butterflies sequester only a subset of cardenolides from milkweed leaves that are less potent against their target enzyme (Na+ /K+ -ATPase) compared to several dominant cardenolides from leaves. However, sequestered compounds remain highly potent against sensitive Na+ /K+ -ATPases found in most predators. We confirmed this differential toxicity with mixtures of purified cardenolides from leaves and butterflies. The genetic basis of monarch adaptation to sequestered cardenolides was also confirmed with transgenic Drosophila that were CRISPR-edited with the monarch's Na+ /K+ -ATPase. Thus, the monarch's selective sequestration appears to reduce self-harm while maintaining protection from enemies.

Urbanization and a green corridor do not impact genetic divergence in common milkweed (Asclepias syriaca L.).

Urbanization is altering landscapes globally at an unprecedented rate. While ecological differences between urban and rural environments often promote phenotypic divergence among populations, it is unclear to what degree these trait differences arise from genetic divergence as opposed to phenotypic plasticity. Furthermore, little is known about how specific landscape elements, such as green corridors, impact genetic divergence in urban environments. We tested the hypotheses that: (1) urbanization, and (2) proximity to an urban green corridor influence genetic divergence in common milkweed (Asclepias syriaca) populations for phenotypic traits. Using seeds from 52 populations along three urban-to-rural subtransects in the Greater Toronto Area, Canada, one of which followed a green corridor, we grew ~ 1000 plants in a common garden setup and measured > 20 ecologically-important traits associated with plant defense/damage, reproduction, and growth over four years. We found significant heritable variation for nine traits within common milkweed populations and weak phenotypic divergence among populations. However, neither urbanization nor an urban green corridor influenced genetic divergence in individual traits or multivariate phenotype. These findings contrast with the expanding literature demonstrating that urbanization promotes rapid evolutionary change and offer preliminary insights into the eco-evolutionary role of green corridors in urban environments.

Tissue and toxin-specific divergent evolution in plant defense Evolución divergente específica de tejido y toxina en defensa de plantas.

A major predicted constraint on the evolution of anti-herbivore defense in plants is the nonindependent expression of traits mediating resistance. Since herbivore attack can be highly variable across plant tissues, we hypothesized that correlations in toxin expression within and between plant tissues may limit population differentiation and, thus, plant adaptation. Using full-sib families from two nearby (<1 km) common milkweed (Asclepias syriaca) populations, we investigated genetic correlations among 28 distinct cardenolide toxins within and between roots, leaves, and seeds and examined signatures of tissue-specific divergent selection between populations by QST-FST comparisons. The prevalence, direction, and strength of genetic correlations among cardenolides were tissue specific, and concentrations of individual cardenolides were moderately correlated between tissues; nonetheless, the direction and strength of correlations were population specific. Population divergence in the cardenolide chemistry was stronger in roots than in leaves and seeds. Divergent selection on individual cardenolides was tissue and toxin specific, except for a single highly toxic cardenolide (labriformin), that showed divergent selection across all plant tissues. Heterogeneous evolution of cardenolides within and between tissues across populations appears possible due to their highly independent expression. This independence may be common in nature, especially in specialized interactions in which distinct herbivores feed on different plant tissues.

Differences in δ13C and δ15N values for monarch butterflies raised on milkweeds from within and outside of crop fields: implications for inferences about natal origins.

Stable isotope (δ2H, δ13C) measurements of wing tissue have been used to determine the natal geographic origin of migrant monarch butterflies that overwinter in Mexico. This study examines the possibility of using δ13C and δ15N to identify the milkweed habitat used by monarchs in their natal region. Milkweeds were common in corn and soybean fields before herbicide use led to their extirpation around 2006, and the loss of those milkweeds has been proposed as a reason for the monarch population decline. If crop-field monarchs can be identified, then historical samples of monarchs could be examined to test that hypothesis. The δ15N and δ13C values of leaves from milkweeds growing in corn fields, soybean fields and non-agricultural habitats were examined as well as monarchs that were raised on those leaves. There were no δ15N values for leaves or monarchs that were distinctive for crop fields. Milkweeds in corn fields, and monarchs that were raised on those milkweeds, were found to have δ13C values distinctly lower than those of other habitats and unlike those of locations within the summer breeding range. Thus, it should be possible to identify monarchs that came from cornfields in samples of overwintering monarchs made before ca. 2006.

Plant water limitation and its impact on the oviposition preferences of the monarch butterfly (Lepidoptera: Nymphalidae).

Intensifying drought conditions across the western United States due to global climate change are altering plant-insect interactions. Specialist herbivores must find their host plants within a matrix of nonhosts, and thus often rely upon specific plant secondary chemistry for host location and oviposition cues. Climate-induced alterations to plant chemistry could thus affect female selection of larval food plants. Here, we investigated whether host-plant water limitation influenced oviposition preference in a threatened invertebrate: the monarch butterfly (Danaus plexippus). We found that females deposited more eggs on reduced-water than on well-watered narrowleaf milkweed plants (Asclepias fascicularis), but we could not attribute this change to any specific change in plant chemistry. Specialist herbivores, such as the monarch butterfly, which are tightly linked to specific plant cues, may experience shift in preferences under global-change conditions. Understanding oviposition preferences will be important to directing ongoing habitat restoration activities for this declining insect.

Temporal matches between monarch butterfly and milkweed population changes over the past 25,000 years.

In intimate ecological interactions, the interdependency of species may result in correlated demographic histories. For species of conservation concern, understanding the long-term dynamics of such interactions may shed light on the drivers of population decline. Here, we address the demographic history of the monarch butterfly, Danaus plexippus, and its dominant host plant, the common milkweed Asclepias syriaca (A. syriaca), using broad-scale sampling and genomic inference. Because genetic resources for milkweed have lagged behind those for monarchs, we first release a chromosome-level genome assembly and annotation for common milkweed. Next, we show that despite its enormous geographic range across eastern North America, A. syriaca is best characterized as a single, roughly panmictic population. Using approximate Bayesian computation with random forests (ABC-RF), a machine learning method for reconstructing demographic histories, we show that both monarchs and milkweed experienced population expansion during the most recent recession of North American glaciers 10,000-20,000 years ago. Our data also identify concurrent population expansions in both species during the large-scale clearing of eastern forests (∼200 years ago). Finally, we find no evidence that either species experienced a reduction in effective population size over the past 75 years. Thus, the well-documented decline of monarch abundance over the past 40 years is not visible in our genomic dataset, reflecting a possible mismatch of the overwintering census population to effective population size in this species.

A nutrition-defence trade-off drives diet choice in a toxic plant generalist.

Plant toxicity shapes the dietary choices of herbivores. Especially when herbivores sequester plant toxins, they may experience a trade-off between gaining protection from natural enemies and avoiding toxicity. The availability of toxins for sequestration may additionally trade off with the nutritional quality of a potential food source for sequestering herbivores. We hypothesized that diet mixing might allow a sequestering herbivore to balance nutrition and defence (via sequestration of plant toxins). Accordingly, here we address diet mixing and sequestration of large milkweed bugs (Oncopeltus fasciatus) when they have differential access to toxins (cardenolides) in their diet. In the absence of toxins from a preferred food (milkweed seeds), large milkweed bugs fed on nutritionally adequate non-toxic seeds, but supplemented their diet by feeding on nutritionally poor, but cardenolide-rich milkweed leaf and stem tissues. This dietary shift corresponded to reduced insect growth but facilitated sequestration of defensive toxins. Plant production of cardenolides was also substantially induced by bug feeding on leaf and stem tissues, perhaps benefitting this cardenolide-resistant herbivore. Thus, sequestration appears to drive diet mixing in this toxic plant generalist, even at the cost of feeding on nutritionally poor plant tissue.

Aphid infestations reduce monarch butterfly colonization, herbivory, and growth on ornamental milkweed.

Anthropogenic disturbance is driving global biodiversity loss, including the monarch butterfly (Danaus plexippus), a dietary specialist of milkweed. In response, ornamental milkweed plantings are increasingly common in urbanized landscapes, and recent evidence indicates they have conservation value for monarch butterflies. Unfortunately, sap-feeding insect herbivores, including the oleander aphid (Aphis nerii), frequently reach high densities on plants in nursery settings and urbanized landscapes. Aphid-infested milkweed may inhibit monarch conservation efforts by reducing host plant quality and inducing plant defenses. To test this, we evaluated the effects of oleander aphid infestation on monarch oviposition, larval performance, and plant traits using tropical milkweed (Asclepias curassavica), the most common commercially available milkweed species in the southern U.S. We quantified monarch oviposition preference, larval herbivory, larval weight, and plant characteristics on aphid-free and aphid-infested milkweed. Monarch butterflies deposited three times more eggs on aphid-free versus aphid-infested milkweed. Similarly, larvae fed aphid-free milkweed consumed and weighed twice as much as larvae fed aphid-infested milkweed. Aphid-free milkweed had higher total dry leaf biomass and nitrogen content than aphid-infested milkweed. Our results indicate that oleander aphid infestations can have indirect negative impacts on urban monarch conservation efforts and highlight the need for effective Lepidoptera-friendly integrated pest management tactics for ornamental plants.

Extent and impacts of winter breeding in the North American monarch butterfly.

Since the 1960s, scientists have observed the North American monarch butterfly (Danaus plexippus) continuing reproductive activities past the fall migration and into the winter months when the climate is mild. Recent work suggests that small populations of winter breeding monarchs are present in western and southeastern USA, as well as northwestern Mexico, with new winter breeding populations forming in areas where non-native milkweeds are planted. The year-round presence of milkweed plants and temperatures suitable for immature monarch development are vital factors allowing for winter breeding. Non-native milkweeds, in conjunction with novel barriers to migration, are likely contributing to the rise in winter breeding behavior. Warmer climates are already impacting milkweed phenology and range, possibly favoring winter breeding behavior. Similar pressures but different implications are expected for eastern and western winter breeding monarchs given the differences in the migration ecology, milkweed species, and climate changes in the two regions.

Overwintering and breeding patterns of monarch butterflies (Danaus plexippus) in coastal plain habitats of the southeastern USA.

Understanding variability in species' traits can inform our understanding of their ecology and aid in the development of management and conservation strategies. Monarch butterflies (Danaus plexippus) are native to the western hemisphere and are well-known for their long-distance migrations but have experienced significant population declines in recent decades. Here we use a 5-year capture-mark-recapture dataset to compare monarch distributions, mating activity, and larval host plant use between two coastal plain habitats in South Carolina, USA. We observed seasonally specific habitat use, with maritime habitats serving as overwintering areas while nearby inland swamps support significant breeding in spring, summer, and fall seasons due to an abundance of aquatic milkweed (Asclepias perennis). We also observed mating activity by fall migrating monarchs and their use of swallow-wort (Pattalias palustre) in the spring as an important larval host plant in maritime habitats. This phenology and habitat use of monarchs diverges from established paradigms and suggest that a distinct population segment of monarchs may exist, with significance for understanding the conservation status of monarch butterflies and associated habitats in eastern North America. Further research should explore how monarchs along the Atlantic coast of North America relate to other eastern monarch populations.

Tissue-specific plant toxins and adaptation in a specialist root herbivore.

In coevolution between plants and insects, reciprocal selection often leads to phenotype matching between chemical defense and herbivore offense. Nonetheless, it is not well understood whether distinct plant parts are differentially defended and how herbivores adapted to those parts cope with tissue-specific defense. Milkweed plants produce a diversity of cardenolide toxins and specialist herbivores have substitutions in their target enzyme (Na+/K+-ATPase), each playing a central role in milkweed-insect coevolution. The four-eyed milkweed beetle (Tetraopes tetrophthalmus) is an abundant toxin-sequestering herbivore that feeds exclusively on milkweed roots as larvae and less so on milkweed leaves as adults. Accordingly, we tested the tolerance of this beetle's Na+/K+-ATPase to cardenolide extracts from roots versus leaves of its main host (Asclepias syriaca), along with sequestered cardenolides from beetle tissues. We additionally purified and tested the inhibitory activity of dominant cardenolides from roots (syrioside) and leaves (glycosylated aspecioside). Tetraopes' enzyme was threefold more tolerant of root extracts and syrioside than leaf cardenolides. Nonetheless, beetle-sequestered cardenolides were more potent than those in roots, suggesting selective uptake or dependence on compartmentalization of toxins away from the beetle's enzymatic target. Because Tetraopes has two functionally validated amino acid substitutions in its Na+/K+-ATPase compared to the ancestral form in other insects, we compared its cardenolide tolerance to that of wild-type Drosophila and CRISPR-edited Drosophila with Tetraopes' Na+/K+-ATPase genotype. Those two amino acid substitutions accounted for >50% of Tetraopes' enhanced enzymatic tolerance of cardenolides. Thus, milkweed's tissue-specific expression of root toxins is matched by physiological adaptations in its specialist root herbivore.

No physiological costs of dual sequestration of chemically different plant toxins in the milkweed bug Spilostethus saxatilis (Heteroptera: Lygaeidae).

Many herbivorous insects not only cope with plant toxins but also sequester them as a defense against predators and parasitoids. Sequestration is a product of the evolutionary arms race between plants and herbivorous insects and has been hypothesized to incur physiological costs due to specific adaptations required. Contradictory evidence about these costs exists for insects sequestering only one class of toxin, but very little is known about the physiological implications for species sequestering structurally different classes of compounds. Spilostethus saxatilis is a milkweed bug belonging to the cardenolide-sequestering heteropteran subfamily Lygaeinae (Heteroptera: Lygaeidae) that has shifted to the colchicine-containing plant Colchicum autumnale, a resource of chemically unrelated alkaloids. Using feeding-assays on artificial diet and chemical analysis, we assessed whether S. saxatilis is still able to sequester cardenolides apart from colchicine and related metabolites (colchicoids), and tested the effect of (1) either a natural cardenolide concentration (using ouabain as a model compound) or a natural colchicine concentration, (2) an increased concentration of both toxins, and (3) seeds of either Asclepias syriaca (cardenolides) or C. autumnale (colchicoids) on a set of life-history traits. For comparison, we assessed the same life-history traits in the milkweed bug Oncopeltus fasciatus exposed to cardenolides only. Although cardenolides and colchicoids have different physiological targets (Na+/K+-ATPase vs tubulin) and thus require different resistance traits, chronic exposure and sequestration of both isolated toxins caused no physiological costs such as reduced growth, increased mortality, lower fertility, or shorter adult life span in S. saxatilis. Indeed, an increased performance was observed in O. fasciatus and an according trend was found in S. saxatilis when feeding on isolated ouabain and isolated colchicine, respectively. Positive effects were even more pronounced when insects were provided with natural toxic seeds (i.e. C. autumnale for S. saxatilis and A. syriaca for O. fasciatus), especially in O. fasciatus. Our findings suggest, that S. saxatilis can sequester two chemically unrelated classes of plant compounds at a cost-free level, and that colchicoids may even play a beneficial role in terms of fertility.

Spatial metabolomics reveal divergent cardenolide processing in the monarch (Danaus plexippus) and the common crow butterfly (Euploea core).

Although being famous for sequestering milkweed cardenolides, the mechanism of sequestration and where cardenolides are localized in caterpillars of the monarch butterfly (Danaus plexippus, Lepidoptera: Danaini) is still unknown. While monarchs tolerate cardenolides by a resistant Na+ /K+ -ATPase, it is unclear how closely related species such as the nonsequestering common crow butterfly (Euploea core, Lepidoptera: Danaini) cope with these toxins. Using novel atmospheric-pressure scanning microprobe matrix-assisted laser/desorption ionization mass spectrometry imaging, we compared the distribution of cardenolides in caterpillars of D. plexippus and E. core. Specifically, we tested at which physiological scale quantitative differences between both species are mediated and how cardenolides distribute across body tissues. Whereas D. plexippus sequestered most cardenolides from milkweed (Asclepias curassavica), no cardenolides were found in the tissues of E. core. Remarkably, quantitative differences already manifest in the gut lumen: while monarchs retain and accumulate cardenolides above plant concentrations, the toxins are degraded in the gut lumen of crows. We visualized cardenolide transport over the monarch midgut epithelium and identified integument cells as the final site of storage where defences might be perceived by predators. Our study provides molecular insight into cardenolide sequestration and highlights the great potential of mass spectrometry imaging for understanding the kinetics of multiple compounds including endogenous metabolites, plant toxins, or insecticides in insects.

Compound-Specific Behavioral and Enzymatic Resistance to Toxic Milkweed Cardenolides in a Generalist Bumblebee Pollinator.

Plant secondary metabolites that defend leaves from herbivores also occur in floral nectar. While specialist herbivores often have adaptations providing resistance to these compounds in leaves, many social insect pollinators are generalists, and therefore are not expected to be as resistant to such compounds. The milkweeds, Asclepias spp., contain toxic cardenolides in all tissues including floral nectar. We compared the concentrations and identities of cardenolides between tissues of the North American common milkweed Asclepias syriaca, and then studied the effect of the predominant cardenolide in nectar, glycosylated aspecioside, on an abundant pollinator. We show that a generalist bumblebee, Bombus impatiens, a common pollinator in eastern North America, consumes less nectar with experimental addition of ouabain (a standard cardenolide derived from Apocynacid plants native to east Africa) but not with addition of glycosylated aspecioside from milkweeds. At a concentration matching that of the maximum in the natural range, both cardenolides reduced activity levels of bees after four days of consumption, demonstrating toxicity despite variation in behavioral deterrence (i.e., consumption). In vitro enzymatic assays of Na+/K+-ATPase, the target site of cardenolides, showed lower toxicity of the milkweed cardenolide than ouabain for B. impatiens, indicating that the lower deterrence may be due to greater tolerance to glycosylated aspecioside. In contrast, there was no difference between the two cardenolides in toxicity to the Na+/K+-ATPase from a control insect, the fruit fly Drosophila melanogaster. Accordingly, this work reveals that even generalist pollinators such as B. impatiens may have adaptations to reduce the toxicity of specific plant secondary metabolites that occur in nectar, despite visiting flowers from a wide variety of plants over the colony's lifespan.

The timing of heat waves has multiyear effects on milkweed and its insect community.

Extreme heat events are becoming more frequent and intense as climate variability increases, and these events inherently vary in their timing. We predicted that the timing of a heat wave would determine its consequences for insect communities owing to temporal variation in the susceptibility of host plants to heat stress. We subjected common milkweed (Asclepias syriaca) plants to in-field experimental heat waves to investigate how the timing of heat waves, both seasonally and relative to a biotic stressor (experimental herbivory), affected their ecological consequences. We found that heat waves had multiyear, timing-specific effects on plant-insect communities. Early-season heat waves led to greater and more persistent effects on plants and herbivore communities than late-season heat waves. Heat waves following experimental herbivory had reduced consequences. Our results show that extreme climate events can have complex, lasting ecological effects beyond the year of the event-and that timing is key to understanding those effects.

The price of defence: toxins, visual signals and oxidative state in an aposematic butterfly.

In a variety of aposematic species, the conspicuousness of an individual's warning signal and the quantity of its chemical defence are positively correlated. This apparent honest signalling is predicted by resource competition models which assume that the production and maintenance of aposematic defences compete for access to antioxidant molecules that have dual functions as pigments and in protecting against oxidative damage. To test for such trade-offs, we raised monarch butterflies (Danaus plexippus) on different species of their milkweed host plants (Apocynaceae) that vary in quantities of cardenolides to test whether (i) the sequestration of cardenolides as a secondary defence is associated with costs in the form of oxidative lipid damage and reduced antioxidant defences; and (ii) lower oxidative state is associated with a reduced capacity to produce aposematic displays. In male monarchs conspicuousness was explained by an interaction between oxidative damage and sequestration: males with high levels of oxidative damage became less conspicuous with increased sequestration of cardenolides, whereas those with low oxidative damage became more conspicuous with increased levels of cardenolides. There was no significant effect of oxidative damage or concentration of sequestered cardenolides on female conspicuousness. Our results demonstrate a physiological linkage between the production of coloration and oxidative state, and differential costs of sequestration and signalling in monarch butterflies.

Timing of a plant-herbivore interaction alters plant growth and reproduction.

Phenological shifts have the potential to change species interactions, but relatively few studies have used experimental manipulations to examine the effects of variation in timing of an interspecific interaction across a series of life stages of a species. Although previous experimental studies have examined the consequences of phenological timing in plant-herbivore interactions for both plants and their herbivores, less is known about their effects on subsequent plant reproduction. Here, we conducted an experiment to determine how shifts in the phenological timing of monarch (Danaus plexippus) larval herbivory affected milkweed (Asclepias fascicularis) host plant performance, including effects on growth and subsequent effects on flower and seed pod phenology and production. We found that variation in the timing of herbivory affected both plant growth and reproduction, with measurable effects several weeks to several months after herbivory ended. The timing of herbivory had qualitatively different effects on vegetative and reproductive biomass: early-season herbivory had the strongest effects on plant size, whereas late-season herbivory had the strongest effects on the production of viable seeds. These results show that phenological shifts in herbivory can have persistent and qualitatively different effects on different life stages across the season.

Herbivory, plant traits and nectar chemistry interact to affect the community of insect visitors and pollination in common milkweed, Asclepias syriaca.

Herbivory can alter plant fitness directly through changing reproductive allocation and indirectly through changing pollinator identity or behavior. Common milkweed is a plant of conservation concern with an inducible chemical defense that is also an important nectar resource. In this study, we aim to understand how herbivory severity and plant traits, including morphology and nectar chemistry, interact to affect insect visitation and pod production in common milkweed. We conducted pollinator watches on plants with experimentally varied herbivory severity and quantified insect frequency and visit length as a response to nectar chemistry, ramet height, number of inflorescences, number of flowers per inflorescence and percent tissue removed. We also quantified pollinator effectiveness and importance. Increased herbivory severity reduced floral displays, including fewer inflorescences and fewer flowers per inflorescence. A reduced floral display was correlated with reduced sucrose, fructose and glucose and resulted in a reduced number and species richness of insect visitors. Fewer flowers per inflorescence reduced the frequency of bumble bee and fly visitors, which were two important pollinators. Although honeybees, flies, small bees, soldier beetles and bumble bees were equally effective pollinators, only bumble bee frequency was positively correlated with pod production. The differences in pollinator visitation have the potential to create diversifying selection on plant floral traits, many of which are also affected by herbivores. This research demonstrates potentially conflicting selection pressures between native and non-native pollinators as well as non-native herbivores.