Diana Fisher

The speed of animal extinction

The three main categories of past extinction causes are invasive species, habitat loss, and over-exploitation. The project would test if mean lag times between invasive predator arrival and extinction is more rapid than mean lag times between habitat loss and extinction, for species with one main identified cause of extinction (an invasive predator vs habitat loss). Adult individuals can sometimes persist for long periods in poor habitat for recruitment (a mechanism of 'extinction debt'). Extinctions due to invasive predators can be very rapid. For example, the red fox Vulpes vulpes is likely to have been responsible for many of the >30 mammal species extinctions on mainland Australia between the 1860s and 1950s, and the range decline (population extinction) of about 40% of surviving mammal species. Medium-sized mammals declined the most and are the preferred prey size of foxes. Evidence from museum collectors and expeditioners of European descent, and First Nations accounts suggest rapid decline and local extinction of susceptible native mammal species as the fox encountered them in its spread from its introduction site in Victoria northwards and westwards. For example, in early 20th century administrative districts in New South Wales, a detailed study of bettong and fox records showed that bettongs disappeared within five years of fox arrival in each district (Short 1998). A set of mammals on the Nullabor plain including a bettong, a bandicoot and a species of mulgara went extinct in the 1920s when the fox reached this region of WA (Newman-Martin et al. 2025). Studies in 2018 and 2025 provide detailed coordinates and dated records of the trajectory of fox invasion. I have a database of coordinates and dated records of mammal decline trajectories. This project would analyse the mean and variability of the lag time between fox arrival and population loss of Australian mainland mammals that have declined, and test correlates of this lag time (body size, climate zone, phylogeny etc). Other data sources for this project include the many worldwide island, lake / spring, and mountain-top species eradicated by habitat loss or invasive species with known dates of threat arrival and extinction.

Another aspect of this project would investigate the speed of decline of non-threatened species.

Species with indicators such as small populations or existing range or population decline are more likely to become extinct, so they are placed in a threatened category by the IUCN Red List and national lists- NT, VU, EN, CR, or EX (rather than the non-threatened category LC). However, there are many examples of common species (LC) suddenly declining to extinction or near-extinction without an intervening period passing through progressively higher categories of risk. This project would test if the incidence of species jumping from least concern to critically endangered or extinct is increasing over time, by analysing past Red List categories and the years when species moved from LC to CR or EX and any intervening categories, to test if there is a bias towards later years in species jumping from non threatened to nearly extinct or extinct. It would also test if rates of common species becoming threatened are similar to rates of species moving between threat categories. These rates can tell us if extinction rates are likely to accelerate, improve accuracy of estimates of extinction rate in different taxa, and how much the cost of addressing the extinction crisis is escalating.

We can also compare the rate of movement between categories for different taxonomic groups and regions, and the frequency of skipping categories, which is an indicator of poor knowledge of the status of these species.

Projects on life history evolution including - sexual selection in animals with male-only semelparity

An analysis in this paper https://www.pnas.org/doi/10.1073/pnas.1310691110 tested if semelparous marsupials (that reproduce once in a lifetime) have larger testes and longer mating duration than iteroparous marsupials in the same families (that reproduce repeatedly). They did- this is evidence that intense post-mating male competition is a component of the evolution of life history in these taxa. All other animals with male-only semelparity are arthropods. This project would analyse species in Families of ants, bees, mantids and spiders that contain some species with male-only semelparity, to test if greater semelparity (fewer males surviving after mating) is associated with indices of higher male reproductive investment in sperm volume and mate guarding, that indicate intense sperm competition. This project would involve 1) measuring body parts of museum specimens (pedipalp size and body size) as well as specimens of insects and spiders that you collect in the field (pedipalp size, body size, testes size of mature males / spermatophore volume where available), 2) measuring body parts of digital images of insects and spiders from published sources, 3) using published data on testes size in mature males and the loss of body parts during mating, to calculate male investment in mating (testes size accounting for body size, proportion of body mass lost, time). The main analyses are phylogenetic regressions. This project is part of an ARC grant, there is a PhD student and a postdoc in my group working on other aspects of the evolution and implications of animal semelparity. There would be a co-supervisor who is a specialist in behavioural and evolutionary ecology of spiders and insects.

Identifying causes of initial bias to improve the Threatened Species Index

Overall Threatened Species Index (TSX) curves and individual time series and taxa show a glitch in the first two or three years- the curve nearly always drops or rises before it settles into a more consistent trajectory (https://tsx.org.au/). This project would test the potential reasons for this. These glitches might be due to a sampling bias artefact related to the choice of initial sampling site. Many studies have addressed spatial bias, such as surveying mainly near roads. Few studies have addressed bias that comes from sampling at certain times in the trajectory of a fluctuating population. Over-sampling of one phase of a fluctuating population (e.g. at a time and place of maximum abundance) is demographic sampling bias. Demographic sampling bias comes from researchers typically choosing thriving populations to study (so that their projects are feasible, or because these are most likely to be detected). In these cases most short term studies will report declines, because fluctuating populations with stable long-term means tend to decline in the short term following a peak (this is 'regression to the mean'). Overestimated declines are also expected if researchers preferentially monitor populations in steady decline and ignore stable ones- this might also be likely in the TSX time series. However, regression to the mean implies that underestimates of decline will ensue if studies focus on recently-declined populations, as these are likely to be in a short term increase phase. This project can use aggregated data to find which curves have which pattern, and understand why this initial glitch occurs. We have access to metadata for the TSX curves and can see which types of studies increase or decrease initially. We might expect that single-species studies might be more likely to initially decline from a peak, and community studies that have not chosen a site based on single species abundance might not show this pattern, as the site was not chosen based on that species. Some researchers provided data later than the start of a long-term study, and some from the start. Birds and mammals would be good data sets to test this.

Conserving threatened species in the Gondwana cloud forests facing drought and climate change

Gondwana cloud forest animal species include the small-range high elevation insectivorous endemics the black-tailed dusky antechinus Antechinus arktos and the rainforest cool-skink Harrisoniascincus zia. These threatened species rely on cool, moist montane Antarctic beech forest in northern NSW and SEQ, and they are vulnerable to extinction if their whole range is affected by the imminent El Nino drought. Government action plans identify climate change and droughts as their major threats, but the only suggested solution is ex situ conservation. However, these species are too scarce to procure founders for captive breeding or translocations, and Antechinus, in particular the dusky antechinus complex does not breed successfully in captivity over multiple generations. In-situ actions are needed.

This project would use a BACI experiment to test if watering leaf litter at key sites for threatened cloud forest specialist vertebrates maintains prey abundance at critical times and prevents decline of leaf litter arthropod-dependent vertebrates, and loss of plants. The methods would involve watering the leaf litter (e.g. using a water truck suitable for off-road tracks with a long hose) and misting the understorey during the drought. It would establish regularly watered plots at locations where the species have been detected and control unwatered plots. It would monitor vertebrates including non-threatened small mammals and reptiles at these drought refuges and control plots using camera trapping and live capture. The project would test the magnitude of prey depletion and soil structure change as drought progresses, quantifying the effect of climate change on insectivorous creatures and arthropods in cloud forest. It would use these data to project broader effects of climate change on cloud forests.

The scale of supplementary watering would be similar to the scale of supplementary feeding and competitor removal that Chris Dickman achieved in manipulative experiments with the related species A swainsonii and A agilis, which had a huge effect on reproduction and offspring survival (Dickman 1986 doi:10.2307/2937075; Dickman doi:10.1111/j.1442-9993.1989.tb01449.x.).