Dieback's bill comes due in soil, not just in dead grasstrees

A Murdoch University team spent a year counting foraging pits across five heavily dieback-infested bushland reserves in the Shire of Mundaring, east of Perth, and found quenda digging activity 33% lower in infested patches than in healthy forest ^[1][2][3]. That number, published in Austral Ecology, is the useful one, because it prices the introduced pathogen Phytophthora dieback not in dead plants but in a soil process that stops happening ^[4][5].

The structural losses were larger than the behavioural one. Compared with healthy forest, infested areas had 46% less low shrub cover, 45% fewer habitable grasstrees and 40% more bare ground ^[6]. The grasstree is one of the plants most affected by dieback, and its value to fauna is mechanical: dead leaves accumulate around the trunk and form shelter ^[7]. Lead author Dr Tom Mansfield said quendas and related bandicoots prefer dense shrubby vegetation like grasstrees for refuge from predators and for nesting ^[8].

The team then checked whether hunger or fear explained the drop. Evidence that truffle-like fungi were fewer was inconclusive, and cover of graminoids and herbs, which can produce bulbs and tubers quendas eat, was not affected ^[9]. Mansfield's reading is that quendas spend less time in degraded areas because they are more exposed to predators, not because food has gone ^[10]. That distinction matters operationally: a food problem needs a different intervention from a cover problem, and cover can be built.

What is actually lost when the digging stops is a set of functions. Mansfield described quendas as ecosystem engineers whose foraging turns over soil, helps water infiltrate, buries seeds and spreads beneficial fungi ^[11], and noted they are among the few remaining digging mammals still performing those functions in southwest Western Australia ^[12]. On that account, a dieback front is also a slow withdrawal of soil turnover and seed burial from the landscape it passes through, with cascading effects if shrub and grasstree loss continues ^[13].

The encouraging detail is that the animals have not written the infestations off. Quendas still entered infested areas to dig, possibly using surviving grasstrees and shrubs, and Mansfield suggested they may enter more if habitat is supplied through restoration or installed refuges ^[14]. The reported declines imply infested areas still retain roughly half their low shrub cover and habitable grasstrees, which is what is left to protect ^[15]. Mansfield's stated priorities are protecting surviving grasstrees, preventing further spread, identifying dieback-resistant plants for restoration, and testing artificial refuges in degraded areas ^[16], with revegetation groups and fauna-friendly gardens as small-scale contributors ^[17].

Two things worth watching. First, whether artificial refuges or resistant plantings measurably move digging rates back up, since the 33% figure is a baseline someone can now be held to ^[3]. Second, whether the inconclusive fungal result resolves in either direction, because a confirmed food effect would change the intervention list ^[9]. The study compared healthy and infested areas rather than tracking a front through time, so causal ordering rests on the authors' interpretation ^[2][10].