Getting western: biogeographical analysis of morphological variation, mitochondrial haplotypes and nuclear markers reveals cryptic species and hybrid zones in the Junonia butterflies of the American southwest and Mexico

Abstract

The American southwest and northern Mexico has a great degree of endemic diversity compared with the rest of North America. The Pleistocene glaciations and the dispersal of species from glacial refuges in this region have been important engines for the production of biodiversity in the region. The New World Junonia are a recent radiation of butterflies that are thought to have spent time in these refuges during periods of glacial advancement. We have reconstructed the plausible movements and the contemporary geographic distributions of the five taxa (J. coenia, J. grisea, J. litoralis, J. nigrosuffusa and J. zonalis) that occur in the American southwest and northern Mexico using phenotypic and genotypic information primarily from specimens preserved in museum collections, supplemented with additional contemporary collections. Evidence of cryptic species and hybridization events were observed using mitochondrial haplotypes, genotypic variation at the nuclear wingless locus, multilocus DNA fingerprinting, patterns of larval host plant use, variation in life-history traits and morphological characteristics. Based on these lines of evidence, and in spite of low levels of hybridization between them, we argue that all five Junonia taxa are independent evolutionary lineages. Junonia grisea and J. coenia are morphologically very similar, but differences in morphology, life-history traits, nuclear genotypes and mitochondrial haplotypes suggest that they are a cryptic species pair, thus elevating J. grisea comb.n. to a full species when it had previously been considered to be a subspecies of J. coenia.

Genomics of a complete butterfly continent

“…Sex chromosome-linked genes are shown to resist introgression in multiple species (42)(43)(44), and thus the Z chromosome tree may better reflect the history of speciation, not the averaged history of introgression. For instance, in accord with morphology, the Z chromosome suggests a sister relationship between morphologically similar but allopatric species Junonia coenia and Junonia grisea (45,46). In contrast, the autosomal tree groups morphologically different but sympatric Junonia grisea and Junonia nigrosuffusa, who experience frequent hybridization and introgression.…”

How old can we go? Evaluating the age limit for effective DNA recovery from historical insect specimens

“…To evaluate the maximum insect specimen age that still permits mitochondrial genotyping, we selected 48 New World Junonia butterfly specimens collected prior to 1910 (ranging in age from pre‐1813 to 1909) (Table S1) for comparison with specimens from reference populations that we had previously genotyped for other projects (Table S2) (Lalonde et al ., ; Lalonde & Marcus, , b). To reduce possible biases due to differences in storage techniques, we sampled specimens from five natural history museum collections [American Museum of Natural History (AMNH), Illinois Natural History Survey (INHS), Natural History Museum London (NHMUK), Smithsonian National Museum of Natural History (USMNH), Yale Peabody Museum of Natural History (YPM)] and two private collections (Thomas W. Turner and specimens maintained as part of the research collection within the Marcus laboratory) for inclusion in our dataset.…”

“…In the New World Junonia , due to retained ancestral polymorphism and/or lateral transfer between species, mitochondrial haplotypes are not species‐specific (Borchers & Marcus, ; Gemmell et al ., ; Gemmell & Marcus, ). However, regardless of species, all of the Junonia from any one locality typically share mitochondrial haplotype frequencies for the two most common haplotype groups (A and B) found in most of the western hemisphere (Pfeiler et al ., ; Gemmell & Marcus, ; Peters & Marcus, ; Lalonde et al ., ; Lalonde & Marcus, ). It is noteworthy that haplotype group A is virtually absent from all species of North American Junonia except in south Florida (Gemmell & Marcus, ; Lalonde et al ., ; Lalonde & Marcus, , b).…”

“…However, regardless of species, all of the Junonia from any one locality typically share mitochondrial haplotype frequencies for the two most common haplotype groups (A and B) found in most of the western hemisphere (Pfeiler et al ., ; Gemmell & Marcus, ; Peters & Marcus, ; Lalonde et al ., ; Lalonde & Marcus, ). It is noteworthy that haplotype group A is virtually absent from all species of North American Junonia except in south Florida (Gemmell & Marcus, ; Lalonde et al ., ; Lalonde & Marcus, , b). Two additional, rarer Junonia haplotypes are geographically restricted, with haplotype C known only from the Peruvian Andes, and a distinct variant of haplotype group B, haplotype B CA , is commonly found in the American southwest and Mexico (Gemmell & Marcus, ; Lalonde & Marcus, ).…”

“…It should be noted that this new genotyping method does not involve sequencing DNA, but instead assigns mitochondrial genotypes based on the presence or absence of sequence‐ and haplotype‐specific restriction enzyme cut sites in the DNA fragments being evaluated [which were already known for each haplotype based on prior exploratory work that included conventional DNA sequencing (Borchers & Marcus, ; Gemmell et al ., ; Peters & Marcus, ; Lalonde et al ., )]. In our previous work, the fragment‐based genotyping technique worked with 100% efficiency for > 1800 specimens collected in the 20th century and contemporary specimens (1910–2016) from the butterfly genus Junonia (Lepidoptera: Nymphalidae)(Table ), as well as for a handful of 19th‐century specimens, with the oldest genotyped museum specimen dating from 1866 (Lalonde & Marcus, , b).…”