Host-plant adaptation as a driver of incipient speciation in the fall armyworm (Spodoptera frugiperda)

Abstract

Background Divergent selection on host-plants is one of the main evolutionary forces driving ecological speciation in phytophagous insects. The ecological speciation might be challenging in the presence of gene flow and assortative mating because the direction of divergence is not necessarily the same between ecological selection (through host-plant adaptation) and assortative mating. The fall armyworm (FAW), a major lepidopteran pest species, is composed of two sympatric strains, corn and rice strains, named after two of their preferred host-plants. These two strains have been hypothesized to undergo incipient speciation, based on (i) several lines of evidence encompassing both pre- and post-zygotic reproductive isolation, and (ii) the presence of a substantial level of genetic differentiation. Even though the status of these two strains has been established a long time ago, it is still yet to be found whether these two strains indeed exhibit a marked level of genetic differentiation from a large number of genomic loci. Here, we analyzed whole genome sequences from 56 FAW individuals either collected from pasture grasses (a part of the favored host range of the rice strain) or corn to assess the role of host-plant adaptation in incipient speciation. Results Principal component analysis of whole genome data shows that the pattern of divergence in the fall armyworm is predominantly explained by the genetic differentiation associated with host-plants. The level of genetic differentiation between corn and rice strains is particularly marked in the Z chromosome. We identified one autosomal locus and two Z chromosome loci targeted by selective sweeps specific to rice strain and corn strain, respectively. The autosomal locus has both increased DXY and FST while the Z chromosome loci had decreased DXY and increased FST. Conclusion These results show that the FAW population structure is dominated by the genetic differentiation between corn and rice strains. This differentiation involves divergent selection targeting at least three loci, which include a locus potentially causing reproductive isolation. Taken together, these results suggest the evolutionary scenario that host-plant speciation is a driver of incipient speciation in the fall armyworm.

Global genomic signature reveals the evolution of fall armyworm in the Eastern hemisphere

“…Our f-branch results also indicate gene flow between the two genetic clusters (Figure 4b), which aligns with the field conditions in the Americas (Dumas, 2015). Our phylogenetic results, along with previous studies, suggest that the rice strain may represent an earlier divergent population (Fiteni et al, 2022;Yainna et al, 2022). Previous studies have reported detection of rice-strain Tpi in the Eastern hemisphere based on a single site, TpiE-183 (Jing et al, 2020;Tay et al, 2022).…”

The evolutionary process of invasion in the fall armyworm (Spodoptera frugiperda)

“…2 B should be considered to be sfC, even though a proportion of this group was classified as sfR according to the mitochondrial COX1 marker because these samples were sampled only from corns. We reported that the TPI marker can be reliably used to identify host plant strains (sfC or sfR) whereas the COX1 marker should be used to identify two phylogenetic entities within sfC 48 . In this study, according to the TPI marker, almost all invasive FAWs were classified as sfC (Fig.…”

“…Nucleotides with a Phred score less than 20 and adapter sequences were removed from the reads using AdapterRemoval v2.1.7 72 . Reads were mapped against the ver7 reference genome ( https://bipaa.genouest.org/sp/spodoptera_frugiperda_pub/download ) 48 , which had chromosome-sized scaffolds, using bowtie2 v2.3.4.1 with the -very-sensitive-local preset 73 . The read depth of mapping of resequencing data was 17.72X in the median.…”

“…We performed whole genome sequencing from FAW samples collected in Benin (39 individuals), India (14), Mexico (26), Florida (24), French Guiana (3), and Guadeloupe (4) using the NovaSeq 6000 System with 20X coverage on average with 150 bp paired-end and 300 bp insert length. Then, we combined resequencing data from Santa Isabel in Puerto Rico (15) (NCBI SRA: PRJNA577869), Stoneville in Mississippi (17) (PRJNA494340), and from Citra and Jacksonville in Florida (USA) (24) (PRJNA639296), which were generated for our previous studies 45,47,48 . The resequencing data from Kunming in China (2) 44 was also included in the analysis.…”

“…genou est. org/ sp/ spodo ptera_ frugi perda_ pub/ downl oad) 48 , which had chromosome-sized scaffolds, using bow-tie2 v2.3.4.1 with the -very-sensitive-local preset 73 . The read depth of mapping of resequencing data was 17.72X in the median.…”