Searching the dirt for missing species

Lire en français

The rediscovery of a long-lost, sand-dwelling desert mole through soil forensics has conservationists in South Africa hoping to use the same method to track other rare and elusive species, such as riverine rabbits and tortoises.

A search for De Winton’s golden mole was the first time environmental DNA (eDNA) traces in soil were analysed to confirm the presence of a specific vertebrate.

All animals shed material into the environment, such as dead skin cells, hair, and traces of body fluids. Like crime scene investigators, researchers increasingly use non-invasive eDNA techniques to, for instance, survey water samples for the presence of fish and other aquatic species, analyse scat to determine animal diets, and to check the microbial life contained in soil and faeces.

“eDNA is a viable and useful approach where species are elusive and difficult to study. It can play a valuable role in refining species distribution, which informs conservation action and allows us to monitor endangered and elusive species,” says Cobus Theron of the Drylands Conservation Programme of the Endangered Wildlife Trust (EWT), a South African NGO.

Since 2019 he has led the search for the critically endangered De Winton’s golden mole (Cryptochloris wintoni), last collected in 1937 along the Northern Cape coastline. The effort formed part of a Re:wild funded project to locate species that had not been seen for a decade or more.

The blind animal with an iridescent coat is the size of a child’s hand. It feeds underground on insects as it ‘swims’ through soft coastal sand. Previous searches were complicated by the fact that it looks remarkably like its sister species, the endangered Van Zyl’s golden mole (Cryptochloris zyli), and that two other species, the Cape golden mole (Chrysochloris asiatica) and Grant’s golden mole (Eremitalpa granti), also inhabit the same coastline.

Trying soil-based eDNA techniques to find De Winton’s made sense, as these are permanent soil-dwellers that would leave DNA traces behind, says EWT conservation geneticist, Samantha Mynhardt, who is based at Stellenbosch University, South Africa. She was lead author of a paper in Biodiversity and Conservation on the case.

Mynhardt says it is more difficult to analyse eDNA from soil samples than from water or scat, as soil typically contains organic inhibitors, such as humic acids and silicas, that could interfere with the reaction process used to analyse genetic samples. Samples must first be cleared of such inhibitors, without damaging DNA traces. Testing for vertebrate DNA is also more effective when more advanced polymerase chain reaction (or PCR) techniques are used to amplify or highlight a particular gene or its fragments.

Mynhardt found it frustratingly difficult to collect enough golden mole genetic samples. South Africa’s coastline is home to 17 of the 21 currently described golden mole species worldwide. Sixteen are endemic, and 10 are listed as either vulnerable, endangered, or critically endangered, according to the IUCN Red List. Any options that could kill one of the rare species were off the table. So too were non-lethal yet potentially costly and labour-intensive options such as camera traps, because of the sandy underground environments in which the animals live. Trapping one is very difficult, as the tell-tale long ridges of sand that golden moles typically push up while burrowing 5 to 10 centimetres underground collapse in dry sand or easily blow away. They are only readily detectable in damp sand after rains.

In 2020, Mynhardt teamed up with the EWT for a pilot study around the coastal town of Lambert’s Bay to find golden moles using eDNA techniques. Full-scale fieldwork at six sites over two weeks followed in 2021.

In both cases, a border collie named Jessie was taken to spots where fieldworkers found fresh signs of burrowing moles. The dog was trained to lie down whenever she scented the more common Grant’s golden mole. Those burrows in which Jessie showed no interest, indicated that they were made by something other than a Grant’s, says Jessie’s handler, Esther Matthew of the EWT.

The team managed to catch, photograph and swab an elusive mole after waiting at a spot ignored by Jessie until a mole again made its presence known by moving sand.

At the six field sites, Mynhardt scraped some 5 millilitres of soil from inside various tunnel systems that Jessie ignored. The subsequent lab-based DNA extraction and next-generation sequencing of more than 100 samples detected DNA footprints of between 20 and 50 vertebrate species per soil sample. This included birds and lizards.

The DNA traces of unidentified mole species were confirmed to be De Winton’s only after University of Pretoria researchers published a more informative reference DNA sequence for the species on the Genbank online database in January 2022. DNA from a specimen housed at the Ditsong National Museum of Natural History was used.

“Our sequences were a perfect match to theirs. We could confidently and definitively say that we had found a De Winton’s,” says Mynhardt. “We also demonstrated that it may be widespread along South Africa’s western coastline, albeit in low abundance.”

Their DNA sampling further shows that there might even be a wholly undescribed golden mole species to be found around one of the fieldwork sites.

The team hopes that the De Winton’s golden mole will become a flagship species for conservation efforts on the coastline, where alluvial diamond mining, coastal housing developments and farming practices are on the increase.

eDNA techniques are further being developed and deployed where necessary or appropriate, for instance to survey endangered and endemic species of South African dwarf tortoises. One of Mynhardt’s honours students at Stellenbosch University, Jeanelize van den Heever, has shown, in unpublished research, that one can determine whether at least one such species, the common angulate tortoises (Chersina angulata), passed through an area at least 24 hours prior to sampling.

“It is the first study globally to look at extracting tortoise DNA from soil,” says Mynhardt.

She is sequencing DNA from rabbit pellets collected in key areas of South Africa’s Karoo region to understand how widespread critically endangered riverine rabbits (Bunolagus monticularis) are.