Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Host plant specificity of the monarch butterfly Danaus plexippus: A systematic review and meta-analysis

Abstract

The preference-performance hypothesis explains host specificity in phytophagous insects, positing that host plants chosen by adults confer the greatest larval fitness. However, adults sometimes oviposit on plants supporting low larval success because the components of host specificity (adult preference, plant palatability, and larval survival) are non-binary and not necessarily correlated. Palatability (willingness to eat) is governed by chemical cues and physical barriers such as trichomes, while survival (ability to complete development) depends upon nutrition and toxicity. Absence of a correlation between the components of host specificity results in low-performance hosts supporting limited larval development. Most studies of specificity focus on oviposition behavior leaving the importance and basis of palatability and survival under-explored. We conducted a comprehensive review of 127 plant species that have been claimed or tested to be hosts for the monarch butterfly Danaus plexippus to classify them as non-hosts, low performance, or high performance. We performed a meta-analysis to test if performance status could be explained by properties of neurotoxic cardenolides or trichome density. We also conducted a no-choice larval feeding experiment to identify causes of low performance. We identified 34 high performance, 42 low performance, 33 non-hosts, and 18 species with unsubstantiated claims. Mean cardenolide concentration was greater in high- than low-performance hosts and a significant predictor of host status, suggesting possible evolutionary trade-offs in monarch specialization. Other cardenolide properties and trichome density were not significant predictors of host status. In the experiment, we found, of the 62% of larvae that attempted to eat low-performance hosts, only 3.5% survived to adult compared to 85% of those on the high-performance host, demonstrating that multiple factors affect larval host plant specificity. Our study is the first to classify all known host plants for monarchs and has conservation implications for this threatened species.

Introduction

Host-selection in phytophagous insects is a complex process mediated by many factors including oviposition stimulants [1, 2], feeding stimulants (phagostimulants) [3], vision [4], experience [5], and deterrents [6]. The preference-performance hypothesis explains the phenomenon of host specificity by positing that adult insects prefer to oviposit on plants that confer the greatest larval fitness [711] and most studies on host-specificity focus on adult oviposition behavior. For specialist insects, plant host status is often portrayed as binary (host or non-host) but, in reality, specificity is a multi-dimensional continuum including preference and performance. Performance can be further divided into palatability–larval willingness and ability to eat a plant–and survival–larval ability to develop into adulthood [5, 12, 13]. Palatability may be governed by chemical signals, i.e., phagostimulants and deterrents [14, 15], and physical barriers, such as trichomes and waxes [16, 17]. Survival is influenced by nutritional value and toxicity [1820]. It has been shown that adult insects do not always select the host plants conferring the greatest larval fitness, demonstrating the importance of larval performance to fully understand host specificity [810, 2125].

The components of host specificity are encountered sequentially. Survival is irrelevant if palatability is too low to facilitate feeding and palatability is meaningless for plants on which adults do not oviposit. However, because all three components are continuous, “mistakes” in host selection are made, advancing a plant to the next component of host status, which may or may not also be low [13, 26]. It is possible for a low-preference plant to have high palatability and for a low-palatability plant to support high survival (Fig 1). These plants can help explain the biological and chemical factors underlying host selection and may be important to understanding shifts or expansions to new host plants [4, 27].

thumbnail
Fig 1. Components and continua of host status.

Each of an insect’s potential host plants may land on different parts of preference, palatability, and survival continua, which are not necessarily correlated. A plant’s position along these continua determines its quality as a host.

https://doi.org/10.1371/journal.pone.0269701.g001

The monarch butterfly Danaus plexippus Linnaeus (Lepidoptera: Nymphalidae) is one of the best known examples of host specificity, as its larvae live and feed almost exclusively on milkweed plants in the genus Asclepias (Apocynaceae) [28]. Publications and online sources often mis-state true host status for monarchs, occasionally resulting in conflicting information. For example, the UK Natural History Museum’s HOSTS database [29] and the U.S. National Resources Conservation Service [30] list Cynanchum laeve (commonly climbing milkweed) as a monarch host plant but this species is absent from host lists published by the Xerces Society for Invertebrate Conservation [31], the U.S. Forest Service [32], and Monarch Joint Venture [33].

Even among Asclepias species, adult monarchs occasionally oviposit on plants that do not result in the greatest larval survival [5, 18, 34, 35]. This mismatch in oviposition preference and larval performance coupled with the inability of young larvae to move to better hosts necessitates research into the larval components of host specificity (palatability and survival). As with many other insects, considerable variation has been documented in monarch larval survival among host plants [5, 3638] with differences of up to 45% between plants [18]. While some studies have reported mortality rates of larvae reared on uncommonly utilized plant species [34, 39], few attempts have been made to determine the drivers of variation in larval survival on these plants [6].

Cardenolides, also called cardiac glycosides (CGs), are part of the potential explanation for variation in monarch preference and performance. CGs are a class of steroid-derived toxins produced by many milkweeds and other plants [40, 41]. The CG backbone (genin) varies slightly, and each can have many different sugar moieties, resulting in a wide range of chemical properties, such as polarity. As a result, individual plants can produce a profile of many CG compounds [4244]. Monarchs have evolved to contend with and even benefit from these chemicals in their host plants [4550], but high concentrations of CGs can still decrease performance [5155]. Additionally, less polar CGs and more diverse CG profiles tend to be more toxic [44, 5560].

Physical and mechanical defenses may also contribute to the low palatability and survival conferred by some plants. Waxes and trichomes (small, hair-like structures covering many plants) act as physical barriers, slowing or preventing feeding. Young monarch larvae must ‘mow’ trichomes covering milkweed plants before they can begin feeding: a time-consuming process shown to reduce fitness [15, 57, 6165]. Latex, the sticky sap exuded when milkweeds are damaged, reduces performance by miring larvae [52, 63] and also contains high levels of CGs [52, 56, 66].

Although monarchs have been extensively studied, the larval performance aspect of their host specificity is not well characterized, and there is conflicting information about which plants act as hosts. Monarch populations have declined in recent years, possibly due to scarcity of their milkweed hosts [67]. A deeper understanding of larval host specificity and identification of alternate host plants are therefore important to conservation efforts. To produce a comprehensive list of monarch host plants, we conducted a literature review of 127 potential host species and characterized what is known about their host status. We classified each plant as unsubstantiated, confirmed unpalatable, low performance, or high performance (Table 1). To investigate the importance of specific factors to palatability and survival in the differences between low- and high-performance hosts, we conducted a meta-analysis to test the hypotheses that plant trichome density, CG concentration, CG diversity, and CG polarity explain host status (A Table in S2 Appendix). Finally, we conducted a no-choice feeding experiment to evaluate the general, relative importance of palatability and survival to larval performance in a small number of species.

Materials and methods

Literature review and classification

We first compiled a preliminary list of 76 monarch host plants from five online sources providing advice for non-scientists interested in monarch conservation [2933]. Then, following the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) framework [68], we used Google Scholar to search the literature for evidence of an association between these species and monarchs (A Fig in S1 Appendix and A Fig in S2 Appendix). Synonymous scientific names were identified using the Global Biodiversity Information Facility (GBIF) database [69]. Searches were conducted with accepted scientific names, and publications using a different species name were noted. To reduce the likelihood of screening out relevant articles, relevant article citations were also assessed for inclusion. If the search returned no results, “Research-Grade” iNaturalist [70] observations of the plant were manually screened for co-observation with monarch larvae to account for potential publication bias. Sources containing information about associations between monarchs and plant species not on the preliminary list were added to the preliminary list to create a comprehensive list. The literature was further searched for associations with newly added plants using the previous methods. After duplicate studies were removed, all articles were manually screened for relevance based on publication and title. The full texts of remaining sources were manually screened to remove irrelevant sources (A Fig in S1 Appendix). Individual sources were assessed for bias, and notes may be found in A Table in S1 Appendix. We perceive the risk of bias in relevant, individual studies to be low and effectively managed because of our standardized protocol, broad search, inclusion of iNaturalist observations, and assessment of individual sources. This process in total produced a list of 127 possible host plant species.

Each plant on the list was classified as unsubstantiated (U), confirmed unpalatable (N), oviposit or some observations (L1), lethal attempt (L2), some attempt (L3), observed high performance (H1), implied high performance (H2) or confirmed high performance (H3) based on search results and iNaturalist observations as defined in Table 1. These classes were simplified into unsubstantiated (U), non-host (N), low performance (L1-L3), and high performance (H1-H3) for analysis (Table 1).

Complete methods, including PRISMA flow diagrams are available in S1 Appendix. A PRISMA Checklist is available in S1 File.

Meta-analysis

Following the PRISMA framework, we used Google Scholar to search for aboveground CG concentrations, diversities, and polarities (A Fig in S2 Appendix) in addition to trichome densities (A Fig in S2 Appendix) for all low and high performance host plants on the list. We screened CG results, leaving only studies containing aboveground CG concentration measurements in units of (digitoxin equivalent mass)/(dry mass), analyses of polar plant extracts not finding cardenolides, CG diversity expressed as the Shannon-Wiener index applied to HPLC data [71], and the CG polarity index developed by Rasmann and Agrawal [71]. All aboveground CG concentrations excluding those from seeds were converted to units of (mg digitoxin equivalent)/(g of dry mass) (mg/g) and averaged for each plant species (grouping sub-species). Trichome search results were also filtered, leaving only publications indicating trichome density (# of trichomes/area) either on both abaxial and adaxial leaf surfaces or the sum of the abaxial and adaxial surfaces. Trichome data were converted to units of (# sum of both surfaces)/mm2 and averaged together for each species. Individual sources were assessed for bias, and notes may be found in B and C Tables in S2 Appendix. For this meta-analysis we draw only raw data from included sources because of a dearth of research evaluating the role of plant characteristics on host status at an inter-specific scale.

Statistical analysis.

Average CG concentrations, CG polarities, CG diversities, and trichome densities were compared between high-performance and low-performance hosts using Wilcoxon rank sum tests because of the data’s non-normal distributions and inequal variances. We also calculate Hodges-Lehmann estimators (effect estimator) for significant Wilcoxon rank sum tests. To test the influence of plant characteristics (CG concentration, CG diversity, CG polarity, and trichome density) on plant host status, we constructed generalized linear models (GLMs) with binomial distributions using mean characteristic values for each plant species as predictor variables and low versus high performance status as a response. Because we were unable to find plant characteristic data for all species, we used three GLMs with different predictors to maximize sample sizes: CG concentration only, CG diversity and CG polarity, and trichome density only. Complete meta-analysis methods, including phylogenetic analyses and PRISMA resources, are included in S2 Appendix.

No-choice feeding experiments

We performed no-choice feeding experiments to quantitatively compare the relative importance of palatability and survival on plant species with different performance statuses. We selected six plant species (in Apocynaceae, Brassicaceae, and Solanaceae) of varying palatability: three confirmed unpalatable plants (Nerium oleander, Solanum dulacamara, and Brassica oleracea var. capitata; commonly oleander, bittersweet nightshade, and cabbage respectively) as negative controls, two low-performance hosts, Gonolobus suberosus (anglepod) and Araujia sericifera (moth plant; both classified as L3), and one high-performance host, Asclepias curassavica (commonly tropical milkweed; classified as H3). All plants were purchased online or from local garden centers except G. suberosus, which was collected wild in the New Orleans, Louisiana area. Plants were watered ad libitum under greenhouse conditions.

Monarch butterflies were caught wild in New Orleans, Louisiana and first bred in captivity during October 2020. The first generation progeny of a single mating pair of adults (referred to as a family line) was used for each trial and each trial employed larvae from a different pair (A Table in S3 Appendix). Only adults that tested negative for the protozoan parasite Ophryocystis elektroscirrha were used for mating to reduce the probability of infecting larvae [72]. O. elektroscirrha is a detrimental parasite for monarch butterflies that can reduce the reproductive success and survival of larvae [73, 74]. Neonatal larvae (<24 hours old) were randomly assigned a feeding treatment before being placed in a petri dish with a moistened paper towel and an isolated piece of leaf tissue. Two larvae were placed in each dish with abundant food to eliminate the risk of competition. All experiments were carried out in an indoor laboratory where air temperature was maintained between 18 and 24°C.

After reaching the third instar, each larva was moved to a larger, individual enclosure with petioles of whole leaves or cut stems placed in water filled test tubes. To reduce exposure to bacteria and parasites, plant tissue was soaked in a 2% bleach solution for 2 minutes and then thoroughly rinsed with tap water before being given to larvae and was replaced as necessary due to wilting or consumption. The larval instar and presence of new frass was recorded daily for each individual until the larva died or entered the pupal stage. Experiments were carried out in four trials, and larvae from all but one trial were raised to adulthood (A Table in S3 Appendix).

Survival analysis.

Larvae reared on non-host plants were pooled as negative controls for analysis. A Cox proportional hazard (CPH) model using species as a predictor variable in conjunction with a likelihood ratio χ2 test was used to determine if plant species was a significant predictor of survivorship curves.

To determine if host species was a significant predictor of differences in survival from days 1–5 and days 6–30, we used GLMs with larval survival during days 1–5 and from days 6–30 as response variables and plant species as a categorical predictor. All larvae reared on non-host plants (B. oleracea var. capitata, N. oleander, and S. dulcamara) died within 5 days of hatching and were excluded from both models. Individuals not reared to adulthood were excluded from the model of survival to day 30. Pairwise post-hoc tests (Tukey) were performed on the estimated marginal means for each species in these models to test for significant differences.

Development analysis.

To identify differences in larval development time between plant species, we created a linear mixed model of development curves [75] with average larval instar on each day after hatching used as a response variable for the model and predictor variables of day after hatching and species (with negative controls pooled as non-hosts because these larvae did not develop; no other data were pooled for analysis). We included random effects of trial number and individual by day. Pairwise comparisons (Tukey) on the estimated marginal means of each species were carried out to test for differences between rearing treatments.

Results

Literature review

The host status of 127 plant species from 54 genera were assessed (B Table in S1 Appendix). We classified 18 as unsubstantiated (U), 33 as confirmed unpalatable (N), 5 as oviposit or some observations (L1), 8 as lethal attempt (L2), 29 as some attempt (L3), 8 as observed high performance (H1), 15 as implied high performance (H2), and 11 as confirmed high performance (H3). After simplification, 42 plants were identified as low-performance and 34 as high-performance hosts (Table 2). Our classification of 28 plant species conflicted with other lists of monarch host plants (excluding unsubstantiated claims) (B Table in S1 Appendix). No iNaturalist co-observations between monarchs and plants classified as unsubstantiated were found.

Meta-analysis

Cardenolides.

CG concentrations in aboveground tissues (excluding seeds) were identified for 78 plant species including 32 low- and 33 high-performance hosts from 71 unique sources (Table 2 and D Table in S2 Appendix). Mean cardenolide concentrations of aboveground plant tissues ranged from 0 mg/g (multiple species) to 7.2 mg/g (Asclepias vestita). Of the 41 sources that reported a sample size, the average was 17 plants. CG concentration for each species was obtained by averaging reported mean values from 1 to 29 sources (mean 3.86 ± 4.86 (sd) sources). The mean of standard deviations calculated between values for each plant species was 1.25 mg/g with values ranging from 0.03 (A. pumila) to 6.8 (A. vestita). We found sufficient evidence to include 11 plant species lacking cardenolides in analyses.

High-performance hosts had a greater average cardenolide concentration than low-performance hosts (Z = -2.64, p = 0.008) (Fig 2). The Hodges-Lehmann estimator was calculated as -0.72 mg/g (mean cardenolide concentration across all species = 1.67 ± 1.70 (sd) mg/g). Correspondingly, cardenolide concentration was a significant predictor of host status (z(63) = 2.67, p = 0.023) (G Table in S2 Appendix).

thumbnail
Fig 2. Density plot of plant species mean cardenolide concentrations.

Vertical, dashed lines indicate mean values.

https://doi.org/10.1371/journal.pone.0269701.g002

CG polarity and diversity values were found for 49 plant species, of which 21 were low performance and 28 were high performance (Table 2 and E Table in S2 Appendix). We found no significant differences in polarity or diversity between high- and low-performance plants (S2 Appendix). Neither polarity nor diversity were significant predictors of host status (H Table in S2 Appendix).

Trichome densities.

We found trichome densities for 29 low-performance and 30 high-performance hosts (Table 2 and F Table in S2 Appendix). We found no significant differences in trichome densities between high- and low-performance host plants (S2 Appendix) and trichome density was not a significant predictor of host status (I Table in S2 Appendix).

When all non-peer reviewed sources (three conference papers and two academic theses; see A Table in S1 Appendix and B Table in S2 Appendix) were removed and the same statistical analyses were performed, our results remained identical (at the reported number of digits) to those presented above (S2 Appendix).

No-choice feeding experiments

Survival.

The Cox proportional hazard model was found to be globally significant when compared to the null model based on a likelihood ratio χ2 test (χ2(3) = 126.9, p = 2e-16) indicating that species or species group is a significant predictor of survivorship (B Table in S3 Appendix). The proportionality assumption for this model was upheld (χ2(3) = 1.38, p = 0.71). Examination of 95% confidence intervals showed that survivorship on the high-performance host, Asclepias curassavica, was significantly different than on non-hosts and low-performance hosts. Confidence intervals of the two low-performance hosts overlapped slightly, and therefore were not significantly different from each other (Fig 3A).

thumbnail
Fig 3. Kaplan Meyer curves from a CPH model and larval development over time.

(A) Survivorship curves calculated by a CPH model from ages 1–30 days for larvae reared on different plant species. Negative controls (non-host species) were pooled. Color shaded areas represent 95% CIs of each curve. (B) The development of monarch larvae reared on different plants as the mean instar by larval age. Negative controls are omitted. Error bars indicate the standard error of mean instar for each day excluding dead individuals.

https://doi.org/10.1371/journal.pone.0269701.g003

Two drops in survival probability (early die-off: days 1–5, late die-off: days 15–25) were observed in all plant treatments but varied in magnitude and cause (Fig 3A). Significantly larger early and late die-offs occurred when larvae were reared on A. sericifera than when they were reared on A. curassavica. The early die-off on G. suberosus was not significantly different from A. curassavica but was significantly larger than A. sericifera. However, the late die-off on G. suberosus was significantly larger than on A. curassavica and not significantly different from A. sericifera (Table 3).

thumbnail
Table 3. Pairwise Tukey post-hoc tests on the estimated marginal means of modeled larval survival from 1–5 and 6–30 days after hatching.

https://doi.org/10.1371/journal.pone.0269701.t003

All 22 of the larvae reared on non-host plants (B. oleracea var. capitata, N. oleander, and S. dulcamara) died within 5 days of hatching, and only one individual produced frass (Fig 4A). Of the 127 larvae reared on A. sericifera, 2 (1.6%) reached adulthood, but 79 (63%) produced frass. A. sericifera larvae succumbing in the early die-off either did not attempt to eat or ate and still died, while 81% of the remaining individuals ate and died during the second die-off (Fig 4). Only 1 of 9 larvae (11%) reared on G. suberosus survived to adult, but 6 (67%) attempted to eat. G. suberosus larvae starved during the early die off and died despite eating during the late die-off (Fig 4). In contrast, 17 of 20 larvae (85%) reared on A. curassavica reached adult, despite all 20 producing frass.

thumbnail
Fig 4. Outcomes of die-off events by plant treatment.

(A) Outcomes of the early die-off. (B) Outcomes of the late die-off. Non-hosts are omitted because no larvae survived past day 5.

https://doi.org/10.1371/journal.pone.0269701.g004

Larval development.

Larval development on A. curassavica was significantly faster than development on A. sericifera (t(134.8) = -10.551, p < 0.0001). The growth curve of larvae reared on G. suberosus was not significantly different from that of A. curassavica (t(90.7) = 2.089, p = 0.0979) or A. sericifera (t(101.4) = -2.343, p = 0.0544) (F Table in S3 Appendix).

Discussion

The larval components of host specificity–plant palatability and larval survival–are influenced by numerous factors and are not necessarily correlated with each other resulting in a multi-dimensional continuum of host status. We identified a wide range of claims regarding the palatability of plants to monarch larvae that provided evidence for a broad list of possible host categories, blurring the distinction between what is a host and what is not. Our classification revealed some conflicting claims. Excluding unsubstantiated claims, we classified 28 plants differently from other sources (B Table in S1 Appendix). For example, some sources list Euphorbia and Gossypium as monarch host plants [29, 76], but these claims may be the result of mistranslation and misinterpretation of vernacular plant names [37]. Citrus and Ipomea batatas have also been documented as host plants [29]: a claim that appears to originate from observations of monarch larvae on these plants [77], but survival of larvae on these species has not been validated.

Another dimension on which to assess host status is the preference of adults to oviposit on a species, termed “oviposition use” [26]. The factors affecting oviposition behavior have been studied extensively. Multiple chemical oviposition stimulants have been identified [1, 2]. Taller plants and those with new growth are especially preferred by adults [55, 78, 79]. Oviposition use has even been shown to change depending on O. elektroscirrha infection status [80]. However, there do not appear to be differences between Eastern and Western monarch populations [35]. Some lists of monarch host plants ranked by oviposition preference exist [16, 39, 78, 79, 8183], but these studies only include a few Asclepias species and, because adult oviposition behavior is affected by many environmental and contextual variables, wider comparisons of oviposition preference are difficult to make [5, 79, 81, 84, 85]. Further complicating the issue, oviposition use is a preference, meaning that pairwise comparisons between all plant species would be necessary to compile a comprehensive, ranked list. Future research is needed to further investigate the relationship between oviposition use and larval performance.

We found that cardenolide concentration was a significant and biologically relevant predictor of a plant species’ host class, suggesting a possible evolutionary tradeoff between high CG concentrations and more diverse cocktails of chemical defenses and its importance in the component of survival. Higher CG concentration was positively associated with performance despite previous research suggesting that milkweeds with higher cardenolide concentrations confer lower larval survival [52, 53, 55, 84]. Despite their toxicity, plants that produce cardenolides tend to produce fewer other defenses, potentially making these species better hosts [8688]. Since milkweed plants produce other defensive compounds in addition to CGs [87, 8992] and plants containing high CG concentrations tend to have lower CG inducibility and lower concentrations of non-CG defensive chemicals [8688], it is possible that the evolution of cardenolide sequestration in monarchs occurred in part to avoid larval exposure to a more diverse cocktail of host plant defensive chemicals [93, 94].

It is possible that our cardenolide concentration analysis was confounded by genetic relatedness of plant species since we found no significant differences in CG concentration between low- and high-performance hosts using phylogenetic ANOVAs on the subset of species for which we had phylogenetic data. However, we also did not find any significant phylogenetic signal in mean cardenolide concentration or performance, and because the sample size was small, standard ANOVAs also did not reveal significant differences in CG concentration (S2 Appendix). A more comprehensive phylogeny would be needed to confirm our findings. Another potential factor affecting our results is that cardenolide concentration varies depending on multiple factors [51, 9598]. However, we mitigated this risk by averaging together values from all available sources (3.86 on average) for each plant species.

Our experimental results support the roles of both palatability and survival in the classification of host status. During the early die-off, survival accounted for most larval mortality on A. sericifera as larvae attempted to eat this plant and produced frass, but low palatability nearly exclusively caused the mortality of larvae reared on G. suberosus and non-hosts. During the late die-off, mortality was almost entirely attributed to low survival (Fig 4). The different causes of mortality between die-offs in larvae reared on A. sericifera and G. suberosus demonstrate the potential for mismatch between the dimensions of performance. Pocius et al. [18] observed a similar, late die-off, which they partially attributed to variation in adult lipid content, supporting the importance of nutritional value to survival. Further experiments using full plants instead of excised leaves are necessary to further explore the role of latex in larval performance.

Despite controlling for larval genetics and environmental differences, we still observed considerable variation among individuals in feeding behavior and survival. There is no obvious explanation for why some larvae attempted to eat plants while others did not. It is possible that, because larval host status is determined by a combination of many factors, even full-sibling larvae have sufficient genetic variation across the many implicated loci to result in behavioral differences [99, 100] or are able to modulate gene expression differently [101]. Our results are unlikely to be affected by local adaptation. Freedman et al. [36] found that larval survival decreased from 79.7% to 75.7% when monarchs were reared on allopatric host plants, which is insufficient to justify geographic differences for host classes. Furthermore, even if the adults collected for experiments were part of the New Orleans resident population, they are unlikely to have adapted to local host plants (e.g., A. curassavica and G. suberosus) [102104] because the population is young and gene flow with migratory populations is common [105108].

Our classification may allow us to better characterize monarchs’ realized and fundamental host ranges [4, 109, 110]. For example, we found that A. sericifera supported development to adulthood albeit for a small number of larvae; however, adult monarchs rarely or never oviposit on this plant [78, 111]. This implies that A. sericifera is within the monarch fundamental host range, but outside of the realized host range. The identification of plant species that can support larval development but are not hosts in the wild introduces the possibility of host niche expansion in response to decreasing abundance of high-performance hosts, which could have important implications for conservation since the loss of Asclepias hosts has been blamed for declines in migratory monarch populations [67].

Conclusion

Through a systematic review, we produced a comprehensive list of 34 high-performance and 42 low-performance host plants for monarch butterflies. We found that low performance can result from low palatability or low larval survival. Our meta-analysis showed that plants containing higher concentrations of cardenolides were more likely to be high-performance hosts, but other cardenolide properties and trichome density were not significant predictors of performance status. Further analyses and experiments are necessary to understand the specific causes of early and late larval mortality in addition to the cause of variation in feeding behavior among closely related individuals. The results of this study suggest the possibility of host range expansion and could be useful in monarch conservation as a guide for selecting beneficial monarch host plants.

Supporting information

S1 Appendix. Literature review supplement.

Includes supplemental methods (including PRISMA resources), supplemental results (included studies and host classes for plant species), and PRISMA checklist. (A Fig and A and B Tables).

https://doi.org/10.1371/journal.pone.0269701.s001

(DOCX)

S2 Appendix. Meta-analysis supplement.

Includes hypotheses, supplemental methods (including PRISMA resources, and phylogenetic analysis), and supplemental results (included studies, mean character values for plant species, model details, and phylogenetic analysis results). (A and B Figs and A-I Tables).

https://doi.org/10.1371/journal.pone.0269701.s002

(DOCX)

S3 Appendix. Experimental supplement.

Includes supplemental methods and model details. (A-F Tables).

https://doi.org/10.1371/journal.pone.0269701.s003

(DOCX)

Acknowledgments

The authors wish to thank Dr. Mark Fishbein for allowing us to use his Asclepias phylogeny data. We are grateful to all the authors who made their data publicly available, thereby making this publication possible. We also thank Dr. Jelagat Cheruiyot and Dr. Robert Pascal for serving as readers on Lewis Greenstein’s honors thesis committee. Finally, we acknowledge and thank the many insects and plants that died to make this research possible.

References

  1. 1. Haribal M, Renwick JAA. Oviposition stimulants for the monarch butterfly: Flavonol glycosides from Asclepias curassavica. Phytochemistry. 1996;41: 139–144. pmid:8588865
  2. 2. Haribal M, Renwick JAA. Identification and Distribution of Oviposition Stimulants for Monarch Butterflies in Hosts and Nonhosts. J Chem Ecol. 1998;24: 891–904.
  3. 3. Bernays EA, Chapman RF. Host-Plant Selection by Phytophagous Insects. Springer Science & Business Media; 1994.
  4. 4. Park I, Eigenbrode SD, Cook SP, Harmon BL, Hinz HL, Schaffner U, et al. Examining olfactory and visual cues governing host-specificity of a weed biological control candidate species to refine pre-release risk assessment. BioControl. 2018;63: 377–389.
  5. 5. Jones PL, Agrawal AA. Beyond preference and performance: host plant selection by monarch butterflies, Danaus plexippus. Oikos. 2019;128: 1092–1102.
  6. 6. Vickerman DB, Boer G de. Maintenance of narrow diet breadth in the monarch butterfly caterpillar: response to various plant species and chemicals. Entomologia Experimentalis et Applicata. 2002;104: 255–269.
  7. 7. Gripenberg S, Mayhew PJ, Parnell M, Roslin T. A meta-analysis of preference–performance relationships in phytophagous insects. Ecology Letters. 2010;13: 383–393. pmid:20100245
  8. 8. Thompson JN. Evolutionary ecology of the relationship between oviposition preference and performance of offspring in phytophagous insects. Entomologia Experimentalis et Applicata. 1988;47: 3–14.
  9. 9. Mayhew PJ. Adaptive Patterns of Host-Plant Selection by Phytophagous Insects. Oikos. 1997;79: 417–428.
  10. 10. Courtney SP, Kibota TT. Mother Doesn’t Know Best: Selection of Hosts by Ovipositing Insects. In: Bernays EA, editor. Insect-Plant Interactions. Boca Raton, FL: CRC Press; 1990. p. 28.
  11. 11. Jaenike J. Host Specialization in Phytophagous Insects. Annual Review of Ecology and Systematics. 1990;21: 243–273.
  12. 12. Charlery de la Masselière M, Facon B, Hafsi A, Duyck P-F. Diet breadth modulates preference—performance relationships in a phytophagous insect community. Sci Rep. 2017;7: 16934. pmid:29208939
  13. 13. Nylin S, Janz N. Oviposition preference and larval performance in Polygonia c-album (Lepidoptera: Nymphalidae): the choice between bad and worse. Ecol Entomol. 1993;18: 394–398.
  14. 14. Robin AHK, Hossain MR, Park J-I, Kim HR, Nou I-S. Glucosinolate Profiles in Cabbage Genotypes Influence the Preferential Feeding of Diamondback Moth (Plutella xylostella). Front Plant Sci. 2017;8.
  15. 15. Robertson GF, Zalucki MP, Paine TD. Larval Host Choice of the Monarch Butterfly (Danaus plexippus L.) on Four Native California Desert Milkweed Species. J Insect Behav. 2015;28: 582–592.
  16. 16. Mattila HR, Otis GW. A comparison of the host preference of monarch butterflies (Danaus plexippus) for milkweed (Asclepias syriaca) over dog-strangler vine (Vincetoxicum rossicum). Entomologia Experimentalis et Applicata. 2003;107: 193–199.
  17. 17. Fürstenberg-Hägg J, Zagrobelny M, Bak S. Plant Defense against Insect Herbivores. International Journal of Molecular Sciences. 2013;14: 10242–10297. pmid:23681010
  18. 18. Pocius VM, Debinski DM, Pleasants JM, Bidne KG, Hellmich RL, Brower LP. Milkweed Matters: Monarch Butterfly (Lepidoptera: Nymphalidae) Survival and Development on Nine Midwestern Milkweed Species. Environ Entomol. 2017;46: 1098–1105. pmid:28961914
  19. 19. Pocius VM, Debinski DM, Bidne KG, Hellmich RL, Hunter FK. Performance of Early Instar Monarch Butterflies (Danaus plexippus L.) on Nine Milkweed Species Native to Iowa. lepi. 2017;71: 153–161.
  20. 20. Dobler S, Petschenka G, Pankoke H. Coping with toxic plant compounds–The insect’s perspective on iridoid glycosides and cardenolides. Phytochemistry. 2011;72: 1593–1604. pmid:21620425
  21. 21. Rausher MD. Larval Habitat Suitability and Oviposition Preference in Three Related Butterflies. Ecology. 1979;60: 503–511.
  22. 22. Valladares G, Lawton JH. Host-Plant Selection in the Holly Leaf-Miner: Does Mother Know Best? Journal of Animal Ecology. 1991;60: 227–240.
  23. 23. Underwood DLA. Intraspecific variability in host plant quality and ovipositionaI preferences in Eucheira socialis (Lepidoptera: Pieridae). Ecological Entomology. 1994;19: 245–256.
  24. 24. Fritz RS, Crabb BA, Hochwender CG. Preference and performance of a gall-inducing sawfly: a test of the plant vigor hypothesis. Oikos. 2000;89: 555–563.
  25. 25. Faria ML, Fernandes GW. Vigour of a dioecious shrub and attack by a galling herbivore. Ecological Entomology. 2001;26: 37–45.
  26. 26. Pocius VM, Pleasants JM, Debinski DM, Bidne KG, Hellmich RL, Bradbury SP, et al. Monarch Butterflies Show Differential Utilization of Nine Midwestern Milkweed Species. Frontiers in Ecology and Evolution. 2018;6: 169.
  27. 27. Jaenike J, Papaj D. Behavioral plasticity and patterns of host use by insects. In: Roitberg BD, Isman MB, editors. Insect Chemical Ecology: An Evolutionary Approach. New York: Springer Science & Business Media; 1992. pp. 245–264.
  28. 28. Ackery P, Vane-Wright R. Milkweed Butterflies, Their Cladistics and Biology: Being an Account of the Natural History of the Danainae, a Subfamily of the Lepidoptera, Nymphalidae. London: British Museum of Natural History; 1984.
  29. 29. Robinson GS, Ackery P, Kitching IJ, Beccaloni GW, Hernández LM. HOSTS—A Database of the World’s Lepidopteran Hostplants. London: Natural History Museum; 2010. Available: http://www.nhm.ac.uk/hosts
  30. 30. Conservation Cover (Monarch Habitat). Natural Resources Conservation Service; 2015.
  31. 31. Borders B, Shepherd M. A Guide to the Native Milkweeds of Oregon. The Xerces Society for Invertibrate Conservation; 2012.
  32. 32. Milkweed Species Beneficial to the Monarch Butterfly. In: United States Forest Service [Internet]. [cited 17 Sep 2021]. Available: https://www.fs.fed.us/wildflowers/pollinators/Monarch_Butterfly/habitat/milkweed_list.shtml
  33. 33. Plant Milkweed for Monarchs. Monarch Joint Venture; Available: https://monarchjointventure.org/images/uploads/documents/MilkweedFactSheetFINAL.pdf
  34. 34. DiTommaso A, Losey JE. Oviposition preference and larval performance of monarch butterflies (Danaus plexippus) on two invasive swallow-wort species. Entomologia Experimentalis et Applicata. 2003;108: 205–209.
  35. 35. Ladner DT, Altizer S. Oviposition preference and larval performance of North American monarch butterflies on four Asclepias species. Entomologia Experimentalis et Applicata. 2005;116: 9–20. https://doi.org/10.1111/j.1570-7458.2005.00308.x
  36. 36. Freedman MG, Jason C, Ramírez SR, Strauss SY. Host plant adaptation during contemporary range expansion in the monarch butterfly. Evolution. 2020;74: 377–391. pmid:31891187
  37. 37. Erickson JM. The Utilization of Various Asclepias Species by Larvae of the Monarch Butterfly, Danaus Plexippus. Psyche: A Journal of Entomology. 1973;80: 230–244.
  38. 38. Pegram KV, Melkonoff NA. Assessing preference and survival of Danaus plexippus on two western species of Asclepias. J Insect Conserv. 2020;24: 287–295.
  39. 39. Dixon CA, Erickson JM, Kellett DN, Rothschild M. Some adaptations between Danaus plexippus and its food plant, with notes on Danaus chrysippus and Euploea core (Insecta: Lepidoptera). Journal of Zoology. 1978;185: 437–467.
  40. 40. Bauer P, Munkert J, Brydziun M, Burda E, Müller-Uri F, Gröger H, et al. Highly conserved progesterone 5β-reductase genes (P5βR) from 5β-cardenolide-free and 5β-cardenolide-producing angiosperms. Phytochemistry. 2010;71: 1495–1505. pmid:20598327
  41. 41. Singh B, Rastogi RP. Cardenolides—glycosides and genins. Phytochemistry. 1970;9: 315–331.
  42. 42. Agrawal AA, Petschenka G, Bingham RA, Weber MG, Rasmann S. Toxic cardenolides: chemical ecology and coevolution of specialized plant–herbivore interactions. New Phytologist. 2012;194: 28–45. pmid:22292897
  43. 43. Roeske CN, Seiber JN, Brower LP, Moffitt CM. Milkweed Cardenolides and Their Comparative Processing by Monarch Butterflies (Danaus plexippus L.). In: Wallace JW, Mansell RL, editors. Biochemical Interaction Between Plants and Insects. Boston, MA: Springer US; 1976. pp. 93–167. https://doi.org/10.1007/978-1-4684-2646-5_3
  44. 44. Hoch HJ. A survey of cardiac glycosides and genins. Charleston: University of South Carolina Press; 1961. Available: https://catalog.hathitrust.org/Record/001573138
  45. 45. Brower LP, Brower JVZ, Corvino JM. Plant poisons in a terrestrial food chain. PNAS. 1967;57: 893–898. pmid:5231352
  46. 46. Brower LP. Ecological Chemistry. Sci Am. 1969;220: 22–29. pmid:5767170
  47. 47. Malcolm SB, Brower LP. Evolutionary and Ecological Implications of Cardenolide Sequestration in the Monarch Butterfly. Experientia. 1989;45: 284–295.
  48. 48. Reichstein T, Euw J von, Parsons JA, Rothschild M. Heart Poisons in the Monarch Butterfly. Science. 1968;161: 861–866. pmid:4875496
  49. 49. Karageorgi M, Groen SC, Sumbul F, Pelaez JN, Verster KI, Aguilar JM, et al. Genome editing retraces the evolution of toxin resistance in the monarch butterfly. Nature. 2019;574: 409–412. pmid:31578524
  50. 50. Petschenka G, Agrawal AA. Milkweed butterfly resistance to plant toxins is linked to sequestration, not coping with a toxic diet. Proc Biol Sci. 2015;282. pmid:26538594
  51. 51. Faldyn MJ, Hunter MD, Elderd BD. Climate change and an invasive, tropical milkweed: an ecological trap for monarch butterflies. Ecology. 2018;99: 1031–1038. pmid:29618170
  52. 52. Zalucki MP, Brower LP, Alonso M A. Detrimental effects of latex and cardiac glycosides on survival and growth of first-instar monarch butterfly larvae Danaus plexippus feeding on the sandhill milkweed Asclepias humistrata. Ecological Entomology. 2001;26: 212–224.
  53. 53. Zalucki MP, Brower LP. Survival of first instar larvae of Danaus plexippus (Lepidoptera: Danainae) in relation to cardiac glycoside and latex content of Asclepias humistrata (Asclepiadaceae). Chemoecology. 1992;3: 81–93.
  54. 54. Tao L, Berns AR, Hunter MD. Why does a good thing become too much? Interactions between foliar nutrients and toxins determine performance of an insect herbivore. Functional Ecology. 2014;28: 190–196. https://doi.org/10.1111/1365-2435.12163
  55. 55. Zalucki MP, Brower LP, Malcolm SB. Oviposition by Danaus plexippus in relation to cardenolide content of three Asclepias species in the southeastern U.S.A. Ecol Entomol. 1990;15: 231–240.
  56. 56. Nelson CJ, Seiber JN, Brower LP. Seasonal and intraplant variation of cardenolide content in the California milkweed, Asclepias eriocarpa, and implications for plant defense. J Chem Ecol. 1981;7: 981–1010. pmid:24420825
  57. 57. Malcolm SB. Cardenolide-Mediated Interactions between Plants and Herbivores. 2nd ed. In: Rosenthal G, Berenbaum M, editors. Herbivores: Their Interactions with Secondary Plant Metabolites. 2nd ed. Elsevier Inc.; 1991. pp. 251–296.
  58. 58. Rasmann S, Johnson MD, Agrawal AA. Induced Responses to Herbivory and Jasmonate in Three Milkweed Species. J Chem Ecol. 2009;35: 1326–1334. pmid:20012168
  59. 59. Detzel A, Wink M. Evidence for a Cardenolide Carrier in Oncopeltus fasciatus (Dallas) (Insecta: Hemiptera). Zeitschrift für Naturforschung C. 1995;50: 127–134.
  60. 60. Jones CG, Firn RD, Malcolm SB, Chaloner WG, Harper JL, Lawton JH. On the evolution of plant secondary chemical diversity. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences. 1991;333: 273–280.
  61. 61. Jones PL, Petschenka G, Flacht L, Agrawal AA. Cardenolide Intake, Sequestration, and Excretion by the Monarch Butterfly along Gradients of Plant Toxicity and Larval Ontogeny. J Chem Ecol. 2019;45: 264–277. pmid:30793231
  62. 62. Zalucki MP, Clarke AR, Malcolm SB. Ecology and Behavior of First Instar Larval Lepidoptera. Annual Review of Entomology. 2002;47: 361–393. pmid:11729079
  63. 63. Agrawal AA, Fishbein M, Jetter R, Salminen J-P, Goldstein JB, Freitag AE, et al. Phylogenetlc Ecology of Leaf Surface Traits in the Milkweeds (Asclepias spp.): Chemistry, Ecophysiology, and Insect Behavior. The New Phytologist. 2009;183: 848–867. pmid:19522840
  64. 64. Malcolm SB. Milkweeds, monarch butterflies and the ecological significance of cardenolides. Chemoecology. 1994;5: 101–117.
  65. 65. Hulley PE. Caterpillar attacks plant mechanical defence by mowing trichomes before feeding. Ecological Entomology. 1988;13: 239–241. https://doi.org/10.1111/j.1365-2311.1988.tb00351.x
  66. 66. Seiber JN, Nelson CJ, Lee SM. Cardenolides in the latex and leaves of seven Asclepias species and Calotropis procera. Phytochemistry. 1982;21: 2343–2348.
  67. 67. Pleasants JM, Oberhauser KS. Milkweed loss in agricultural fields because of herbicide use: effect on the monarch butterfly population. Insect Conservation and Diversity. 2013;6: 135–144. https://doi.org/10.1111/j.1752-4598.2012.00196.x
  68. 68. Moher D, Liberati A, Tetzlaff J, Altman DG, Group TP. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLOS Medicine. 2009;6: e1000097. pmid:19621072
  69. 69. GBIF: The Global Biodiversity Information Facility. What is GBIF? 2021 [cited 22 Apr 2021]. Available: https://www.gbif.org/what-is-gbif
  70. 70. iNaturalist. [cited 22 Apr 2021]. Available: https://www.inaturalist.org
  71. 71. Rasmann S, Agrawal AA. Latitudinal patterns in plant defense: evolution of cardenolides, their toxicity and induction following herbivory. Ecology Letters. 2011;14: 476–483. pmid:21371232
  72. 72. Altizer SM, Oberhauser KS, Brower LP. Associations between host migration and the prevalence of a protozoan parasite in natural populations of adult monarch butterflies. Ecological Entomology. 2000;25: 125–139.
  73. 73. Altizer SM, Oberhauser KS. Effects of the Protozoan Parasite Ophryocystis elektroscirrha on the Fitness of Monarch Butterflies (Danaus plexippus). Journal of Invertebrate Pathology. 1999;74: 76–88. pmid:10388550
  74. 74. Altizer SM. Migratory behaviour and host–parasite co-evolution in natural populations of monarch butterflies infected with a protozoan parasite. Evol Ecol Res. 2001;3: 567–581.
  75. 75. Mirman D. Growth Curve Analysis and Visualization Using R. London: Chapman and Hall/CRC; 2014.
  76. 76. Higgins LG, Riley ND. A field guide to the butterflies of Britain and Europe. [1st American ed.]. Boston, MA: Houghton Mifflin; 1970. Available: http://hdl.handle.net/2027/umn.31951000090635g
  77. 77. Kimball CP. The Lepidoptera of Florda: An Annotated Checklist. Gainesville, Florida: Florida Department of Agriculture; 1965.
  78. 78. Zalucki MP, Kitching RL. Dynamics of oviposition in Danaus plexippus (Insecta: Lepidoptera) on milkweed, Asclepias spp. Journal of Zoology. 1982;198: 103–116. https://doi.org/10.1111/j.1469-7998.1982.tb02063.x
  79. 79. Cohen JA, Brower LP. Oviposition and Larval Success of Wild Monarch Butterflies (Lepidoptera: Danaidae) in Relation to Host Plant Size and Cardenolide Concentration. Journal of the Kansas Entomological Society. 1982;55: 343–348.
  80. 80. Lefèvre T, Chiang A, Kelavkar M, Li H, Li J, Castillejo CLF de, et al. Behavioural resistance against a protozoan parasite in the monarch butterfly. Journal of Animal Ecology. 2012;81: 70–79. pmid:21939438
  81. 81. Pocius VM, Debinski DM, Pleasants JM, Bidne KG, Hellmich RL. Monarch butterflies do not place all of their eggs in one basket: oviposition on nine Midwestern milkweed species. Ecosphere. 2018;9: e02064.
  82. 82. Malcolm SB, Brower LP. Selective Oviposition by Monarch Butterflies (Danaus plexippus L.) in a Mixed Stand of Asclepias curassavica L. and A. incarnata L. in South Florida. Journal of the Lepidopterists’ Society. 1986;40: 255–263.
  83. 83. Zalucki MP, Oyeyele S, Vowles P. Selective Oviposition by Danaus Plexippus (l.) (lepidoptera: Nymphalidae) in a Mixed Stand of Asclepias Fruticosa and a. Curassavica in Southeast Queensland. Australian Journal of Entomology. 1989;28: 141–146. https://doi.org/10.1111/j.1440-6055.1989.tb01211.x
  84. 84. Oyeyele SO, Zalucki MP. Cardiac glycosides and oviposition by Danaus plexippus on Asclepias fruticosa in south-east Queensland (Australia), with notes on the effect of plant nitrogen content. Ecol Entomol. 1990;15: 177–185.
  85. 85. Van Hook T, Zalucki MP. Oviposition by Danaus Plexippus (Nymphalidae: Danainae) on Asclepias viridis in Northern Florida. Journal of the Lepidopterists’ Society. 1991;45: 7.
  86. 86. Agrawal AA, Salminen J-P, Fishbein M. Phylogenetic Trends in Phenolic Metabolism of Milkweeds (Asclepias): Evidence for Escalation. Evolution. 2009;63: 663–673. https://doi.org/10.1111/j.1558-5646.2008.00573.x pmid:19220456
  87. 87. Agrawal AA, Lajeunesse MJ, Fishbein M. Evolution of Latex and its Constituent Defensive Chemistry in Milkweeds (Asclepias): a Phylogenetic Test of Plant Defense Escalation. Entomologia Experimentalis et Applicata. 2008;128: 126–138.
  88. 88. Agrawal AA, Hastings AP. Trade-offs constrain the evolution of an inducible defense within but not between plant species. Ecology. 2019;100: e02857. pmid:31365759
  89. 89. Arribére MC, Cortadi AA, Gattuso MA, Bettiol MP, Priolo NS, Caffini NO. Comparison of Asclepiadaceae latex proteases and characterization of Morrenia brachystephana Griseb. cysteine peptidases. Phytochemical Analysis. 1998;9: 267–273.
  90. 90. Trejo SA, López LMI, Cimino CV, Caffini NO, Natalucci CL. Purification and Characterization of a New Plant Endopeptidase Isolated from Latex of Asclepias fruticosa L. (Asclepiadaceae). J Protein Chem. 2001;20: 469–477. pmid:11760121
  91. 91. Liggieri C, Arribére MC, Trejo SA, Canals F, Avilés FX, Priolo NS. Purification and Biochemical Characterization of Asclepain c I from the Latex of Asclepias curassavica L. J Protein Chem. 2004;23: 403–411.
  92. 92. Dubey V Kumar, Jagannadham MV. Procerain, a stable cysteine protease from the latex of Calotropis procera. Phytochemistry. 2003;62: 1057–1071. pmid:12591258
  93. 93. Cornell HV, Hawkins BA. Herbivore Responses to Plant Secondary Compounds: A Test of Phytochemical Coevolution Theory. The American Naturalist. 2003;161: 507–522. pmid:12776881
  94. 94. Agrawal AA, Conner JK, Rasmann S. Tradeoffs and Negative Correlations in Evolutionary Ecology. Evolution since Darwin: The First 150 Years. Sunderland, Mass.: Sinauer Associates; 2010. pp. 243–268.
  95. 95. Brower LP, Seiber JN, Nelson CJ, Lynch SP, Tuskes PM. Plant-determined variation in the cardenolide content, thin-layer chromatography profiles, and emetic potency of monarch butterflies, Danaus plexippus reared on the milkweed, Asclepias eriocarpa in California. J Chem Ecol. 1982;8: 579–633. pmid:24415043
  96. 96. Brower LP, Seiber JN, Nelson CJ, Lynch SP, Holland MM. Plant-determined variation in the cardenolide content, thin-layer chromatography profiles, and emetic potency of monarch butterflies, Danaus plexippus L. Reared on milkweed plants in California: 2. Asclepias speciosa. J Chem Ecol. 1984;10: 601–639. pmid:24318600
  97. 97. Brower LP, Seiber JN, Nelson CJ, Lynch SP, Hoggard MP, Cohen JA. Plant-determined variation in cardenolide content and thin-layer chromatography profiles of monarch butterflies, Danaus plexippus reared on milkweed plants in California: 3. Asclepias californica. J Chem Ecol. 1984;10: 1823–1857. pmid:24318436
  98. 98. Helmus MR, Dussourd DE. Glues or poisons: which triggers vein cutting by monarch caterpillars? Chemoecology. 2005;15: 45–49.
  99. 99. Niepoth N, Bendesky A. How Natural Genetic Variation Shapes Behavior. Annu Rev Genom Hum Genet. 2020;21: 437–463. pmid:32283949
  100. 100. Jeschke V, Kearney EE, Schramm K, Kunert G, Shekhov A, Gershenzon J, et al. How Glucosinolates Affect Generalist Lepidopteran Larvae: Growth, Development and Glucosinolate Metabolism. Front Plant Sci. 2017;8.
  101. 101. Tan W-H, Acevedo T, Harris EV, Alcaide TY, Walters JR, Hunter MD, et al. Transcriptomics of monarch butterflies (Danaus plexippus) reveals that toxic host plants alter expression of detoxification genes and down-regulate a small number of immune genes. Molecular Ecology. 2019;28: 4845–4863. pmid:31483077
  102. 102. Spellman DL, Gunn CR. Morrenia odorata and Araujia sericofera (Asclepiadaceae): Weeds in Citrus Groves. Castanea. 1976;41: 139–148.
  103. 103. Carvalho R, Pellissari LCO, Pace MR, Scremin-Dias E, De Oliveira Arruda R, Farinaccio MA. Leaf morphoanatomy of Araujia and Morrenia (Asclepiadoideae, Apocynaceae): phylogenetic implications and species key. Botanical Journal of the Linnean Society. 2017;183: 280–293.
  104. 104. Araujia sericifera (AJASE). In: European Plant Protection Organization (EPPO) [Internet]. 26 Nov 2020 [cited 15 Mar 2021]. Available: https://gd.eppo.int/taxon/AJASE/distribution
  105. 105. Howard E, Aschen H, Davis AK. Citizen Science Observations of Monarch Butterfly Overwintering in the Southern United States. Psyche: A Journal of Entomology. 2010;2010: 1–6.
  106. 106. Sternberg ED, Li H, Wang R, Gowler C, de Roode JC. Patterns of Host-Parasite Adaptation in Three Populations of Monarch Butterflies Infected with a Naturally Occurring Protozoan Disease: Virulence, Resistance, and Tolerance. The American Naturalist. 2013;182: E235–E248. pmid:24231547
  107. 107. Satterfield DA, Maerz JC, Altizer S. Loss of migratory behaviour increases infection risk for a butterfly host. Proc R Soc B. 2015;282: 20141734. pmid:25589600
  108. 108. Altizer S, Davis AK. Populations of Monarch Butterflies with Different Migratory Behaviors Show Divergence in Wing Morphology. Evolution. 2010;64: 1018–1028. pmid:20067519
  109. 109. Cao Z, Wang H, Meng L, Li B. Risk to nontarget plants from Ophraella communa (Coleoptera: Chrysomelidae), a potential biological control agent of alien invasive weed Ambrosia artemisiifolia (Asteraceae) in China. Appl Entomol Zool. 2011;46: 375–381.
  110. 110. Schaffner U. Host Range Testing of Insects for Biological Weed Control: How Can It Be Better Interpreted?: Data on the host range of biocontrol candidates are particularly relevant in assessing potential detrimental effects to nontarget organisms. BioScience. 2001;51: 951–959.
  111. 111. Singer MC. Evolution of Food-Plant Preference in the Butterfly Euphydryas editha. Evolution. 1971;25: 383–389. pmid:28563107