The worldwide invasion history of a pest ambrosia beetle inferred using population genomics.

Xylosandrus crassiusculus, a fungus-farming wood borer native to Southeastern Asia, is the most rapidly spreading invasive ambrosia species worldwide. Previous studies focusing on its genetic structure suggested the existence of cryptic genetic variation in this species. Yet, these studies used different genetic markers, focused on different geographical areas and did not include Europe. Our first goal was to determine the worldwide genetic structure of this species based on both mitochondrial and genomic markers. Our second goal was to study X. crassiusculus' invasion history on a global level and identify the origins of the invasion in Europe. We used a COI and RAD sequencing design to characterize 188 and 206 specimens worldwide, building the most comprehensive genetic data set for any ambrosia beetle to date. The results were largely consistent between markers. Two differentiated genetic clusters were invasive, albeit in different regions of the world. The markers were inconsistent only for a few specimens found exclusively in Japan. Mainland USA could have acted as a source for further expansion to Canada and Argentina through stepping stone expansion and bridgehead events. We showed that Europe was only colonized by Cluster II through a complex invasion history including several arrivals from multiple origins in the native area, and possibly including bridgehead from the United States. Our results also suggested that Spain was colonized directly from Italy through intracontinental dispersion. It is unclear whether the mutually exclusive allopatric distribution of the two clusters is due to neutral effects or due to different ecological requirements.

Deciphering the demographic history of allochronic differentiation in the pine processionary moth Thaumetopoea pityocampa.

Understanding the processes of adaptive divergence, which may ultimately lead to speciation, is a major question in evolutionary biology. Allochronic differentiation refers to a particular situation where gene flow is primarily impeded by temporal isolation between early and late reproducers. This process has been suggested to occur in a large array of organisms, even though it is still overlooked in the literature. We here focused on a well-documented case of incipient allochronic speciation in the winter pine processionary moth Thaumetopoea pityocampa. This species typically reproduces in summer and larval development occurs throughout autumn and winter. A unique, phenologically shifted population (SP) was discovered in 1997 in Portugal. It was proved to be strongly differentiated from the sympatric "winter population" (WP), but its evolutionary history could only now be explored. We took advantage of the recent assembly of a draft genome and of the development of pan-genomic RAD-seq markers to decipher the demographic history of the differentiating populations and develop genome scans of adaptive differentiation. We showed that the SP diverged relatively recently, that is, few hundred years ago, and went through two successive bottlenecks followed by population size expansions, while the sympatric WP is currently experiencing a population decline. We identified outlier SNPs that were mapped onto the genome, but none were associated with the phenological shift or with subsequent adaptations. The strong genetic drift that occurred along the SP lineage certainly challenged our capacity to reveal functionally important loci.

Draft genome and reference transcriptomic resources for the urticating pine defoliator Thaumetopoea pityocampa (Lepidoptera: Notodontidae).

The pine processionary moth Thaumetopoea pityocampa (Lepidoptera: Notodontidae) is the main pine defoliator in the Mediterranean region. Its urticating larvae cause severe human and animal health concerns in the invaded areas. This species shows a high phenotypic variability for various traits, such as phenology, fecundity and tolerance to extreme temperatures. This study presents the construction and analysis of extensive genomic and transcriptomic resources, which are an obligate prerequisite to understand their underlying genetic architecture. Using a well-studied population from Portugal with peculiar phenological characteristics, the karyotype was first determined and a first draft genome of 537 Mb total length was assembled into 68,292 scaffolds (N50 = 164 kb). From this genome assembly, 29,415 coding genes were predicted. To circumvent some limitations for fine-scale physical mapping of genomic regions of interest, a 3X coverage BAC library was also developed. In particular, 11 BACs from this library were individually sequenced to assess the assembly quality. Additionally, de novo transcriptomic resources were generated from various developmental stages sequenced with HiSeq and MiSeq Illumina technologies. The reads were de novo assembled into 62,376 and 63,175 transcripts, respectively. Then, a robust subset of the genome-predicted coding genes, the de novo transcriptome assemblies and previously published 454/Sanger data were clustered to obtain a high-quality and comprehensive reference transcriptome consisting of 29,701 bona fide unigenes. These sequences covered 99% of the cegma and 88% of the busco highly conserved eukaryotic genes and 84% of the busco arthropod gene set. Moreover, 90% of these transcripts could be localized on the draft genome. The described information is available via a genome annotation portal (http://bipaa.genouest.org/sp/thaumetopoea_pityocampa/).