Recovery and analysis of ancient beetle DNA from subfossil packrat middens using high-throughput sequencing

Abstract

The study of ancient DNA is revolutionizing our understanding of paleo-ecology and the evolutionary history of species. Insects are essential components in many ecosystems and constitute the most diverse group of animals. Yet they are largely neglected in ancient DNA studies. We report the results of the first targeted investigation of insect ancient DNA to positively identify subfossil insects to species, which includes the recovery of endogenous content from samples as old as ~ 34,355 ybp. Potential inhibitors currently limiting widespread research on insect ancient DNA are discussed, including the lack of closely related genomic reference sequences (decreased mapping efficiency) and the need for more extensive collaborations with insect taxonomists. The advantages of insect-based studies are also highlighted, especially in the context of understanding past climate change. In this regard, insect remains from ancient packrat middens are a rich and largely uninvestigated resource for exploring paleo-ecology and species dynamics over time.

Figure 1

Gene arrangement and read depth along reconstructed ancient mitochondrial genomes. All aDNA reads were mapped in BWA v.0.7.17 to the Philolithus actuosus TB20958 (Tenebrionidae: Asidini) mitogenome. Asterisk indicates samples containing > 1% ambiguous sequences from lack of coverage. Red lines mark average coverage. White fields represent control (CR) and other non-coding regions (nc). Color codes: Red: C to T substitutions; Blue: G to A substitutions; Grey: All other substitutions; Orange: Soft-clipped bases; Green: Deletions relative to the reference; Purple: Insertions relative to the reference. Subfossil fragments (Philolithus actuosus) used in the extraction process were illustrated for 34,355 ybp sample.

Figure 2

Comparison of the number of successfully recovered mitochondrial (a) and nuclear (b) reads from the analyzed libraries representing ancient samples mapped to different reference sequences. Subsequent reference sequences represent species increasingly less related to the analyzed ancient species based on morphological traits.

Figure 3

Phylogenetic placement of recovered ancient samples. Maximum likelihood tree generated by analysis of a concatenated dataset of 13,926 bp (16 s, 28 s, 12 s, atp6, atp8, cob, COI-III, NAD1-6) in IQ-Tree with ancient midden samples assembled de novo or using Philolithus actuosus (TB20958) as the reference. Total number of bp used for each taxa is listed in Supplemental Table 1. Posterior probabilities (PP) from Bayesian analyses are displayed on the right, bootstrap values from IQ-Tree on the left. An asterisk indicates either PP of 1.0 or bootstrap value of 100%.