Yilin’s PhD, based at the Australian Wine Research Institute in Adelaide through the ARC Training Centre for Biosecurity, is focused on scale insects in Australian vineyards, a widespread pest family that feeds on almost every part of the vine, spreads persistent grapevine viruses, and leaves behind a sugary residue that promotes sooty mould, damaging both fruit and, eventually, wine quality.

For a long time these insects sat in the background as a manageable, low-level nuisance, kept in check by natural predators. That’s changed. A 2023 metagenomic study by Yilin’s supervisors uncovered something unexpected: a previously unrecorded species, referred to as Parthenolecanium corni, wasn’t just present in Australian vineyards, it was the most dominant scale species across every sample taken. As Yilin explained, “the species we assume wasn’t there turns out to be the main one that we are dealing with.”

That matters for biosecurity because P. corni behaves very differently to the species current control strategies were designed around. Where the previously known species complete a single generation a year, P. corni can run through three or four in the same period overseas, and current controls, tuned for one generation, are losing effectiveness against it. Australia’s only prior record of the species was a single specimen from 1976, so almost nothing is known about how it arrived or spread. Yilin’s project aims to answer that using genetic data, combining Cox1 DNA barcoding to confirm species identity with whole genome sequencing to reconstruct the invasion’s history.

All of which depends on a step that sounds simple but turned out to be a real bottleneck of the project: getting enough DNA out of the insects in the first place. Scale insects move through several life stages, and the two available to Yilin are second instar nymphs, tiny, almost transparent, but available for a wide window of the year, and adult females, which yield more DNA but are only around for a narrow seasonal window centred on September and October. Since the nymphs are the more practical option for most of the year, Yilin needed a method that could pull usable DNA out of them.

It didn’t come easily. She tested a silica column kit, a magnetic bead kit, and a traditional phenol-chloroform extraction, and for the nymphs, two of the three failed outright, recovering no detectable DNA at all. The bead kit did slightly better, but only in the range of one to seven nanograms, well short of what’s needed. Even adult females, where plenty of DNA should have been available, came back lower than typically reported. “This difficulty isn’t only about body size,” Yilin told the seminar, “there’s something else, something about the samples or this protocol, that is limiting the yield, even when there is plenty of tissue.”

From there it became a genuine optimisation puzzle. Adding Proteinase K to break down proteins and free up trapped DNA lifted adult yields by ten to twenty times, but did almost nothing for the nymphs, telling Yilin that their bottleneck wasn’t poor digestion but simply too little DNA to begin with. Switching to a blood and tissue kit built around Proteinase K helped further, and reducing the elution buffer volume improved recovery from around 64% to 83%, though that concentration came with its own cost: larger samples started carrying sediment contamination, including waxes, lipids and honeydew residue, into the final extraction, which meant reverting to a larger buffer volume to dilute it back out.

Even after all that optimisation, the second instar nymphs still fall short of the roughly five nanograms per microlitre needed for whole genome library preparation, a hard floor set by how little tissue the insects simply contain. Yilin’s formal sampling has now shifted to adult females instead, trading a much tighter collection window for DNA yields that actually clear the bar.

Why it matters: while the nymphs won’t make it into the formal experiment, the troubleshooting behind them produced useful, transferable findings, including Proteinase K in the lysis step, reducing elution volume to concentrate DNA, and matching kit chemistry to sample type, tips that should help anyone working with similarly low-input material.

As Yilin summed it up, it’s better to have found the limits of the method now, during a pilot, “rather than once we’ve committed to a full sampling season.” It’s a good example of how a methods question ends up shaping the entire direction of a biosecurity project.

Thank you to Yilin, and to her supervisors at the Australian Wine Research Institute and The Australian National University.