It is well established that many species of fish are threatened by warming oceans, acidification and deoxygenation. But which species are most vulnerable to these changes, and how vulnerable are they? A recent paper highlights a mathematical tool that might answer those questions.
Researchers at the University of British Columbia (UBC), Vancouver, used a type of algorithm known as “fuzzy logic” to rank how vulnerable 1,074 species of fish are to climate change. Fuzzy logic measures gradients of uncertainty in an answer. It is used to look at answers to questions that do not necessarily have one correct response or another but fall somewhere in between; not true or false, but 67 percent true. Not A, B or C, but 32 percent A, 58 percent B and 10 percent C. Researchers might not be 100 percent sure that a fish is vulnerable to climate change, but 72 percent sure.
Every ranking by the UBC researchers was based on publicly available data of attributes such as average body size, rate of reproduction and how likely specific ocean changes were to impact that fish. Using fuzzy logic, the researchers boiled all of this down to a risk score for each species. They found that 157 species were most at risk – fish such as Eastern Australian salmon and spotted grouper, a fish whose market demand may reach 100,000 metrics tons annually by 2020. Another 294 species were at high risk.
William Cheung, an associate professor at the Institute for the Oceans and Fisheries at the University of British Columbia, Vancouver, and corresponding author on the paper, said that this fuzzy-data-method can be applied to other data to give researchers a better idea of which marine organisms are most at risk. Oceans Deeply spoke with Cheung about fuzzy logic and how it can be used to evaluate fish species around the world.
Oceans Deeply: Why did you decide to use fuzzy logic for this research?
William Cheung: The reason that we used fuzzy logic is that it can account for uncertainty and gaps in our knowledge about marine species and how they will be impacted by climate change. When we [talk about] the vulnerability of a fish to climate change, one indication is the body size of fish. Bigger fish may be more sensitive to ocean warming. However, what do you mean by big or small? It’s pretty fuzzy. So, in this case fuzzy logic allows us to prevent that kind of uncertainty – we can classify fish as both large and small with different degrees of knowledge.
The way that fuzzy logic works represents how experts think and make conclusions. Imagine that you go to a medical doctor – the doctor would do lots of different tests to collect lots of information about your temperature, heart rate, blood pressure. But ultimately, it is the doctor that assesses this information to come up with an overall diagnosis.
Fuzzy logic is designed to mimic that – where you have solid information about different metrics based on robust science. Measuring, for example, the body size of fish, the growth rate of fish. The fuzzy logic then accounts for all of this information to come up with an overall conclusion.
The purpose of [fuzzy logic] is not to come up with really precise predictions of how fish will exactly respond to climate change. The idea is to highlight the kinds of species, the groups of species, that could be more susceptible to climate change, so we can pay more attention to the conservation and management of those species.
Oceans Deeply: You identified some of the most vulnerable species, like the East Australian salmon and the spotted grouper. What aspects of those species make them most vulnerable to climate change?
Cheung: I’ll take the Australian salmon as an example: It only occurs in the southern range of Australia and New Zealand, so their distribution is really limited. That is one aspect that makes them highly vulnerable to climate change, because it means that they do not have lots of room to adapt by changing their distribution. Secondly, the species has a very narrow range of temperatures that they can survive, and that makes them sensitive to ocean warming.
All together these provide a general conclusion that the Australian salmon is one of the most vulnerable species.
Oceans Deeply: Do you have an idea of things that you are not able to measure with fuzzy logic?
Cheung: One thing is that we cannot incorporate anything outside of our current knowledge base. So, if there is a something that will impact a species’ vulnerability to climate change, then it is very difficult for us to incorporate that component into the algorithm. But one aspect of fuzzy logic is that it is adaptive, so we can calculate more rules and relationships into the fuzzy logic system to help improve it. If there were a specific set of species where we know that additional considerations will be added in order to better understand the vulnerability of those species, then we can add those considerations fairly easily into the algorithm.