Zach Carter is a PhD student at the University of Auckland School of Biological Sciences, focusing on invasion biology, island ecosystems, and physical geography. He is currently studying landscape-scale eradication of invasive mammals of the New Zealand archipelago, including rats, mustelids, and the common brushtail possum. Zach shares the motivations and challenges behind his recent publication in the Journal of Biogeography, which aims to improve the understanding of reinvasion potential and to identify the greatest threats to the long-term eradication of invasive species.
(left) Zach sexing a kiore (Rattus exulans) on Slipper Island off of New Zealand’s Coromandel Peninsula. (right) Zach reluctantly holding a deceased eel (Anguilla sp.) on Motukawanui, an island off of New Zealand’s Northland coast that was featured in his recent article.
Links: Institutional webpage | Lab group
Institution: University of Auckland, School of Biological Sciences
Current academic life stage: PhD
Research interests: I am interested in prioritising landscape-scale eradication projects of invasive species to achieve desirable conservation outcomes. My work has a strong statistical modelling and analytical component, focusing primarily on invasion biology, island ecosystems, and physical geography.
Current study system: I am currently studying invasive mammals of the New Zealand archipelago, including rats (Rattus rattus, R. norvegicus, R. exulans), mustelids (Mustela erminea, M. furo, M. nivalis) and common brushtail possum (Trichosurus vulpecula). New Zealand’s endemic biota evolved in the absence of terrestrial mammals, making them highly susceptible to mammalian predation. As a result, the species I study constitute a grave biological threat even though they are relatively unassuming (and even cute). Understanding how rats, mustelids and possum affect their surroundings, and how their negative impacts can be reversed, is particularly interesting.
Recent paper in Journal of Biogeography: Carter ZT, Perry GLW, Russell JC. (2020) Determining the underlying structure of insular isolation measures. Journal of Biogeography 47:955–967. DOI: 10.1111/jbi.13778
Motivation for the paper: Our motivation for this study was to better understand reinvasion potential; reinvasion represents the greatest threat to long-term eradication success. Rodents and stoats readily hitchhike to islands as seafaring stowaways or disperse naturally by swimming between islands. As a result, an island’s reinvasion risk is directly correlated with its level of geographical isolation. However, quantifying insular isolation is not necessarily a straightforward process: there are many different metrics describing isolation and it can be difficult to ascertain which metrics are necessary for quantification. We thought resolving this issue would make for a terrific study. Because isolation is such an important aspect of biology, we made a concerted effort to make our findings applicable and accessible across disciplines.
(left) Common brushtail possum (Trichosurus vulpecula) and (right) the Polynesian rat (kiore, Rattus exulans). Both species have had a significant impact on New Zealand’s biota through predation and resource competition. New Zealand’s endemic species evolved in the absence of terrestrial mammals and, as such, are highly susceptible to their threats.
Key methodologies: We conducted a principal components analysis (PCA) on 16 different measures frequently used to describe insular isolation, including Euclidean-based distance metrics, landscape connectivity measures derived from least-cost and circuit theory modelling, stepping stones, and buffer metrics, among others. PCA simply accounts for redundancy within a set of input variables; variables that describe a similar thing will cluster (or ‘load’) together. We used these clustered components (the ‘principal components’ or PCs) to understand what variables may be necessary for measuring insular isolation.
Unexpected challenges: Interestingly, we found that a simple Euclidean distance measure describes insular isolation comparably to graph-theoretic and least-cost connectivity metrics. Connectivity metrics are incredibly useful for understanding aspects of a heterogeneous environment, but are apparently limited in areas that are homogenous and inhospitable (such as an ocean body to a terrestrial mammal). Moreover, calculating a Euclidean distance is straightforward because it is simply the straight-line distance between two locations. Connectivity metrics, on the other hand, are often computationally intensive. We were required to calculate our least-cost paths, resistance distances, and their derived metrics using a high performance cluster computer. These calculations took months of continuous computing due to the scale of our analysis. In the end, we found that a simple 5-minute calculation was comparable to one taking hundreds of hours to complete. I had to take a small vacation after I completed this paper in order to keep my sanity in check.
Major result and contribution to the field: The basis of a robust description of insular isolation comes from three independent PCA components, including distance from the mainland source to the focal island (PC1), stepping stones available along the dispersal pathway (PC2), and the focal island’s position within the landscape (PC3). Importantly, these PCs are relatively easy to calculate. PC1 can be measured using a simple Euclidean distance measurement, PC2 by counting the number of stepping stone islands available along a dispersal pathway, and PC3 by measuring the focal island’s size. The parsimony and simplicity afforded by these measures should facilitate application across disciplines.
What are the next steps? We make no claims regarding the ecological importance of the identified principal components. The PCA cannot tell us which measures are most important, only what measures are necessary for describing isolation. As such, we place all three identified PCs on equal footing. A significant next step in this research would be to identify the relative importance of each component. Doing so would allow for creation of an equation describing insular isolation with a single metric. Our paper only focused on small mammalian dispersers, though. It is very likely the components describing isolation would look entirely different for other dispersers (particularly those that are well-equipped for long-distance dispersal).
If you could study any organism on Earth, what would it be and why? I feel incredibly lucky to be studying New Zealand’s invasive species. In 2012, the New Zealand government championed the Predator Free 2050 programme, an initiative to extirpate rats, stoats, and possum from the entirety of the country by 2050. This programme is a first of its kind and I couldn’t be happier to be studying the organisms and systems that I am. I hope to continue my work on these invasive species even after completion of my PhD.
Any other little gems you would like to share? I have taken quite a meandering path to get to where I am today: my BSc and MSc are both in engineering. Studying complex biological systems was initially difficult to grapple with at the beginning of my PhD but I am finding my engineering background has been exceedingly helpful. The time I’ve spent working as an engineer has given me a fresh perspective to problem solving that I wouldn’t have had otherwise. I am proof that having an initial lack of direction can turn out to be a beneficial career move!
Islands at the northern end of the Cavalli Islands. Panaki Island (identified under the arrow) was used as the example focal island in our recent paper. The Cavalli’s are a great example of how stepping stone islands facilitate dispersal of terrestrial biota.