Leveraging adaptive sampling of environmental DNA for monitoring the critically endangered kākāpō

Lara Urban (University of Otago, New Zealand) kicks off with an overview of eDNA (environmental DNA) and conveys the power of eDNA as a tool to monitor biodiversity.

Lara explains that wild populations are decreasing at unprecedented rates resulting in cultural, social, and economic problems. Biodiversity loss is identified as a driver of exacerbated food insecurity as well as increased pandemic risk; threats which Lara identifies as being driven by a lack of quantitative measurements to determine biodiversity loss. Lara explains how adaptive sampling of eDNA with nanopore sequencing provides a scalable cost-efficient way of monitoring biodiversity.

Before going into the specifics of the research a range of eDNA monitoring projects are shared including bacterial diversity rivers in Cambridge, asses marine biodiversity around New Zealand, and terrestrial ecosystem monitoring to protect endangered flightless bird species.

Monitoring the critically endangered kākāpō species with eDNA

The critically endangered kākāpō is a flightless, nocturnal parrot. Native to New Zealand, the kākāpō has suffered since arrival of human settlers due to loss of habitat and the introduction of predators. These factors have led to the extinction of other species. Kākāpōs remain critically endangered with only 204 alive today, resulting in low genetic diversity, infertility, and increased disease susceptibility in the population.

The kākāpō recovery programme was set up to save the kākāpō from extinction; however, its success has depended on local tribes, the guardians of the species, and the Department of Conservation in New Zealand. All kākāpōs now live on remote predator-free islands except for one male kākāpō found on the mainland.

Her team set up an eDNA-based approach for monitoring the kākāpō population, and to seek evidence of a remnant population on the mainland. This involved DNA extraction from soil samples (soil was chosen as the kākāpō is a ground-dwelling, flightless bird), prior to amplifying the 12S-ribosomal RNA region and short-read amplicon sequencing. Sets of samples were taken directly from feeding stations, bowls, and abandoned nests, and at 12 metres and 24 metres away from these sites. Amplicon sequence variants from these samples were then used to successfully determine whether kākāpōs were present, confirming a high degree of localisation, with high abundance at the feeding stations and bowls. Samples taken from a botanic garden also confirmed that the methodology used to detect eDNA did not confuse kākāpōs for other avian species.

To go beyond the ability to accurately detect species and identify individual kākāpōs from eDNA, a whole-genome sequencing approach was needed. Using adaptive sampling on the GridION, Lara aimed to enrich the 1 Gbp kakapo genome.  For this, the DNA processing steps included bead clean up and library prep with the Ligation Sequencing Kit (SQK-LSK109), prior to flow cell loading. Data analysis was conducted with Guppy (basecalling), Porechop (trimming), and Minimap2 (mapping to reference).

Lara explained that a mixture of short reads and long reads were found in the data, as was expected, due to the samples being quite fragmented, which included fragments that mapped to the reference kākāpō genome as well as contaminants from the environment.

Variant calling and phasing with Medaka and WhatsHap identified 26.4 thousand SNPs and 476 haplotypes.

Lara then explained that this data, in combination with 1.6 million high-quality SNPs from kākāpō125+ consortium data, resulted in the identification of 34 haplotypes with overlapping SNPs (position and allele). The SNPs and haplotypes identified were then used in combination with statistical inferences (log-likelihood) to identify individual kākāpōs. Lara shared how this was a ‘eureka moment’ as it was the first reported use of eDNA to identify individuals within a population.

Examples of identification of individual kākāpō alongside contamination detection

A stunning series of slides followed, showing the names of kākāpōs identified, as well as instances where the methodology accurately distinguished contaminants. Lara then shared plans to establish the genetic structure, identify inbreeding, and even assess fitness-related phenotypes within populations, using the same methodology. Lara highlighted the challenges of the obtaining the coverage required to establish polygenic traits in these settings, explaining that DNA extraction methods are improving so rapidly that the possibility of doing so is on the horizon if not here already.

In the final part of this breakout session Lara explained how biodiversity monitoring projects with nanopore sequencing are straightforward to set up and shared details of the Takahē recovery project which she also set up and received funding for. In this project, long nanopore reads from blood samples from wild specimens were used to develop a reference genome, and then establish the genetic diversity of the population. A proportion of the Takahē population were located on a remote rugged island, making capturing, and taking blood samples directly next to impossible – faecal samples were therefore used instead.

Finally, Lara mentioned her team members Jo Stanton, University of Otago, and Miles Benton, ESR, and shared their collective ambition to further develop eDNA applications with nanopore sequencing in both biodiversity monitoring and pathogen identification in New Zealand.