Aim

To investigate the drivers of intra-specific genetic diversity in centipedes, a group of ancient predatory soil arthropods.

Location

Asia, Australasia and Europe

Time period

Present

Major taxa studied

Centipedes (Class: Chilopoda)

Methods

We assembled a database of 1245 mitochondrial cytochrome c oxidase subunit I sequences representing 128 centipede species from all five orders of Chilopoda. This sequence dataset was used to estimate genetic diversity for centipede species and compare its distribution with estimates from other arthropod groups. We studied the variation in centipede genetic diversity with species traits and biogeography using a beta regression framework, controlling for the effect of shared evolutionary history within a family.

Results

A wide variation in genetic diversity across centipede species (0 to 0.1713) falls towards the higher end of values among arthropods. Overall, 27.57% of the variation in mitochondrial COI genetic diversity in centipedes was explained by a combination of predictors related to life history and biogeography. Genetic diversity decreased with body size and latitudinal position of sampled localities, was greater in species showing maternal care and increased with geographic distance among conspecifics.

Main conclusions

Centipedes fall towards the higher end of genetic diversity among arthropods, which may be related to their long evolutionary history and low dispersal ability. In centipedes, the negative association of body size with genetic diversity may be mediated by its influence on local abundance or the influence of ecological strategy on long-term population history. Species with maternal care had higher genetic diversity, which goes against expectations and needs further scrutiny. Hemispheric differences in genetic diversity can be due to historic climatic stability and lower seasonality in the southern hemisphere. Overall, we find that despite the differences in mean genetic diversity among animals, similar processes related to life history strategy and biogeography are associated with the variation within them.

Centipede mitochondrial COI accession numbers were obtained from published literature and GenBank sequence datasets. These were supplemented with sequences from additional centipede species using the phylogatR database (https://phylogatr.org/). Information on geographic coordinates associated with sequences was obtained from source literature and by querying specimen voucher numbers against museum databases. For a small proportion of sequences, coordinates were obtained by geocoding locality descriptions. Information on species traits was obtained from species descriptions in taxonomic studies, and species range was obtained from Chilobase 2.0 (https://chilobase.biologia.unipd.it/), GBIF (https://www.gbif.org/), species descriptions and regional atlases. Species with at least three sequence representatives and complete trait information were retained for analysis. Sequence alignments were carried out for each species and sequence statistics, including genetic diversity, were calculated using R scripts. Important correlates of intra-specific genetic diversity were identified using a beta regression model, where average pairwise difference was the response variable, and, body size, blindness, maternal care, species latitudinal range, mean latitude of sequences, average geographic distance between sequences and number of sequences were the predictor variables. Spatial autocorrelation in model residuals was corrected using additional spatial eigenvectors as predictors, and sensitivity of results to sample size was tested. The analysis described above was carried out using custom R scripts.

The analysis code can be run in the R programming language.