Curtis Champion is a PhD candidate at the Institute for Marine and Antarctic Studies – University of Tasmania. He stopped by the Fish Thinkers blog to give us a run down on some of the research he is working on.
Species on the move and a quick explainer about how range-shifts are commonly identified
Climate. Change. No doubt you’ve heard of the phenomenon. And while a small number of our political reps sporadically break into the headlines for criticising its reality, the global scientific community has been busy forging novel territory to understand its ecological consequences. This emergent field is most-commonly referred to as “species redistribution science” because plants and animals shifting where they live (generally towards the poles or up mountains) in response to changes in temperature is perhaps the most perceptible ecological effect of climate change. An important milestone for this area of study was recently acknowledge in the scientific literature, when Bonebrake and co (2017) asserted that species redistribution science is now “a field in its own right.” This sentiment was already well-established in February 2016 when 277 researchers gathered in Hobart to attend the inaugural Species on the Move conference to discuss the topic, and interest has only grown, with planning presently underway for the next international gathering (Species on the Move 2.0) in South Africa, 2019.
The overarching objective of the field is to understand the ecological effects of human-caused climate change to enhance our capacity to predict and adapt to future change (yes, we’re currently committed to future climate change, even if we were to stop polluting now). It’s a pretty damn important area of research too, having already informed society that contemporary climate change is driving a rapid, global redistribution of biodiversity (Pecl et al., 2017), and that these changes are occurring fastest in the ocean (Poloczanska et al., 2013). Implications of these findings include negative effects to human health and well-being, particularly in the poorest regions of the world, and a redistribution of valuable and delicious natural resources, like coffee, chocolate and wine crops (scary stuff!).
But when it comes to the nitty-gritty of attempting to measure climate-driven changes in species distributions over time, there is often devil in the detail (or in fact lack of detail). Detecting changes in the species distributions, and attributing these changes to human-caused climate change, is hard because decent historical datasets of species occurrence locations are elusive. For example, observations of animals commonly come from locations that are easy for people to access, and may not actually represent the true distribution of a species, meaning the data are usually confounded by observer bias. In the absence of high-quality data sources, we can instead turn to another approach for keeping tabs on the how species distributions are likely being effected by climate change. As species redistribution science progresses, it’ll likely become increasingly important for interested persons to have some understanding of, or have been exposed to, the most common of approaches for identifying species on the move… so here’s a quick explainer.
The approach does not involve directly assessing changes in locations where plants and animals have been observed over time, but instead relies of assessing shifts in species’ preferred environmental conditions – e.g. the range of temperatures an animal is best suited to. By focusing on species environmental habitat preferences, we rely less on high quality, long-term datasets of species observations and can begin to infer likely range-shifts based on changes in the environmental conditions that species respond to. It’s a pragmatic approach that adopts the “let’s do the best with what we’ve got” mentality. This technique is commonly referred to as ‘species distribution modelling’, which is misleading because we’re actually modelling species’ environmental habitat preferences and not their distributions directly, so I prefer the alternative term of ‘habitat suitability modelling.’
The advent of satellite technology for remotely sensing environmental data (such as sea surface temperature, current speed, salinity and dissolved oxygen) has been instrumental in our ability to thoroughly understand the suite of environmental factors that combine to form the overall environmental habitat preference of different marine species. This is achieved by matching the known location and time of, for example, fishes caught by recreational anglers with multiple variables that likely explain why that species occurred in that location, such as temperature. In this way, it’s possible to define the combination of multiple environmental variables that best describe where a species has been observed, and thus estimate the likelihood of a species occurring in other locations, given some information about the environmental conditions of those locations (again, easily available from satellites). It is these predictions of the habitat suitability, particularly when they are made regularly and over a long period of time (e.g. each month for 20 years), that are commonly mapped and used to infer climate-driven changes to species distributions.
As species redistribution science becomes increasingly important on a planet whose climate is rapidly changing, it’s interesting to consider how reported changes in species distributions are actually measured. While direct approaches for measuring range-shifts are usually preferable when long-term datasets permit, I think it’s important to consider the strong links and transfer of information between habitat suitability modelling and climate change ecology. This point was highlighted by Bonebrake and others (2017), in the same paper that announced species redistribution science and a field in its own right, by showing that ‘species distribution modelling’ (or habitat suitability modelling) is the top trending phase within this emergent field, suggesting this method is one worth knowing about.