Temporal matches and mismatches between monarch butterfly and milkweed population changes over the past 12,000 years

Abstract

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, 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 via Random Forests (ABC-RF), a machine learning method for reconstructing demographic histories, we show that both monarchs and milkweed experienced concurrent range expansion during the most recent recession of North American glaciers ∼12,000 years ago. Our data identify an expansion of milkweed during the large-scale clearing of eastern forests (∼200 years ago) but was inconclusive as to expansion or contraction of the monarch butterfly population during this time. Finally, our results indicate that neither species experienced a population contraction 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.

Population genetics of a recent range expansion and subsequent loss of migration in monarch butterflies

“…This is due to the conflicting inferences provided by the two best‐performing model structures and the wide range of parameter estimates in the Three Epoch models. Although we present the results of both the simpler Found and Grow and Two Epoch models, and the more complicated Three Epoch model, we are inclined to place more confidence in the estimates produced by the Three Epoch model for two reasons: (i) the demographic scenario that it specifies—recent demographic expansion in the ancestral North American population prior to geographical expansion—has empirical support from other studies (Boyle et al, 2022; Pfeiler et al, 2017; Zhan et al, 2014) and accords with our understanding of past changes in climate, and (ii) this model structure produces parameter estimates that match our prior understanding for how and when monarch range expansion may have occurred. The latter point is related to the former: because a North American population expansion is not allowed until after the founding of Hawaii in the Found and Grow model, this model forces an ancient founding of the Hawaiian population in order to allow for the ancient growth of the North American population.…”

Functional evidence supports adaptive plant chemical defense along a geographical cline

“…S4 and S5 and Table S4 ). We assessed F ST using 925 established single nucleotide polymorphisms (SNPs) ( 23 ) in each of 12 of the 24 sampled populations. Phenotypic differences between populations ( P ST ) were substantially greater than F ST for labriformin and syrioside B (but not for the other dominant cardenolides), indicating spatially divergent selection on these compounds ( SI Appendix ).…”

Genome and tissue-specific transcriptome of the tropical milkweed (Asclepias curassavica)

“…Although smaller in total length, the 11 pseudomolecules of the A. curassavica genome assembly are highly colinear with the A. syriaca genome assembly (Boyle et al, 2022) (Figure 3). The A. syriaca genome was found to have a greater expansion of LTR retroelements relative to A. curassavica , contributing to its larger size (Figure 2).…”

“…OrthoFinder v2.5.2 (Emms & Kelly, 2015) was used to identify gene families using protein sequences from genome-sequenced close relatives to A. curassavica with genome annotations. The following species were included: A. curassavica v1.0 , Asclepias syriaca v1.0 (Boyle et al, 2022), Calotropis gigantea (Hoopes et al, 2018), Catharanthus roseus (Kellner et al, 2015), Rhazya stricta (Sabir et al, 2016), Eustoma grandiflorum (Liang et al, 2022), Gelsemium sempervirens (Frank et al, 2015), Neolamarckia cadamba (Zhao et al, 2022), Chiococca alba (Lau et al, 2020), Coffea canephora v1.0 (Denoeud et al, 2014). These results were used to create an Upset plot using the UpSetR package v1.4.0 (Conway et al, 2017) depicting shared and unique gene clusters.…”