#34: Part 2, Fish distribution shifts from climate change: Empirical evidence for range contractions
Dan Isaak
Biological evidence of fish skies falling is almost nonexistent…
Hi Everyone,
So as we saw last time, the fisheries community has definitely recognized the threat posed by climate change to aquatic critters this century. More than 20 bioclimatic models have been developed and published in the last two decades (graphic 1; Blog #33). Most of these predict large distributional shifts and range contractions in high profile species, so there’s much on the line in terms of conserving biodiversity, maintaining important fisheries, and spending limited resources wisely (graphic 2). In contrast to the extensive empirical evidence available to confirm distribution shifts in many terrestrial plant and animal taxa (Blog #31), the evidence for aquatic biotas is very limited. This hurts us in two ways—by contributing to an “inertia of inaction” regarding tough choices that are better made sooner rather than later, and by running the risk of making poor decisions when we do act if the model predictions turn out to be wrong.
With this in mind, I’m always keeping an eye out for studies that provide good evidence of range contractions in fish populations related to recent climate trends, but as far as I can tell, so far there are only 2 (if I’m missing some, I’d love to have them brought to my attention). The first study is by Hari and colleagues on brown trout in Switzerland, and so far, represents the seminal piece of work on the topic. This study is so comprehensive, in fact, that it was highlighted in an earlier blog (#10) on stream temperatures for the insights it provided about regionally consistent stream warming trends related to climate change, climate cycles, and glaciers in Swiss streams.
Today, to quote Paul Harvey, we’ll tell “the rest of the story” from Hari et al. and what was done with the 24 years of consistent stream temperature monitoring data at their disposal. In the second phase of the study, thermal bioenergetic criteria for growth and survival of brown trout were compared to thermal conditions in 87 river sections across the country for which temperature data were available. This allowed the authors to make much more precise local predictions of potential effects on brown trout during a warming period from 1978-2001 than would have been the case from broad-scale bioclimatic models (& as we’ll see in future blogs, this precision is often needed to tease out the gradual changes caused by climate change). In the third phase of the study, these predictions were compared to trends in population abundance over the same period.
Many useful details are in Hari et al, but the take home is that over the 24 year period encompassed by the study, catch rates and trout densities declined dramatically in the warmest, lowest elevation river sections that were close to being too warm for the species at the outset (graphic 3). Populations declined less rapidly in sections with intermediate thermal conditions, and showed no trends in the coldest river sections. That pretty well matches what the most basic predictions of climate change and warming effects should be for an ectothermic species at thermally mediated boundaries during periods of warming. However, it was also never the case that dead fish just floated up everywhere one day when a critical temperature was exceeded to make the temperature linkage obvious. Instead, the pattern was subtle and protracted, exacerbated by increased incidence of disease, slower growth and survival, etc., etc. And that’s often going to be typical of how climate effects manifest in many populations, it’ll be a very slow process that interacts with other stressors to chip away at populations (although these gradual trends may also help set up populations for catastrophic events in certain contexts as we’ll see in a future blog), which means long-term monitoring datasets are often going to be mandatory for discerning the climate signal from the noise.
OK, so the second study is by Winfield & colleagues and looks at long-term population trends in lake populations of Arctic char across the United Kingdom. And that right there should tell you we’ve got to be at the southern extent of the species’ range (& thermal tolerances) because “Arctic” generally isn’t the first thing that pops to mind when thinking about the landmass inhabited by our trans-Atlantic friends. Regardless, the premise of the study is much the same as Hari’s—look at many populations over a long period of time spread across a range of thermal conditions (graphic 4). What Winfield finds is that the southern-most populations, located several hundred kilometers away from the northern population, have all undergone significant population declines over the recent 20 – 40 year period. The northern population, in contrast, appeared to also decline, but by nowhere near as much as those to the south. A variety of factors could play roles in these declines, but the authors do a good job of addressing them, so here again, we see a pattern that matches the basic predictions derived from the bioclimatic models & that other investigators have observed for other taxa (Blog #31).
And for fish, that’s it. Those are the 2 best studies I’ve seen documenting the process of range contraction related to climate change in previous decades. Now we just need a few dozen additional studies from other continents to get a better sense of how widespread this phenomenon may ultimately be for aquatic biotas & how the biological rates of change relate to environmental rates (Blogs 10, 11, 13, 16, 17, 18, and 23). Ideally, some of those studies would be done within the same areas, & for the same species, that bioclimatic models have been previously developed so that we can test, refine, and improve the models to the point that they begin to provide sound scientific information at resolutions useful for prioritizing local conservation efforts. We’re not there yet, but I suspect we’ll be surprised at how much farther down this path we are in the next 5 years. The current state of knowledge within the fisheries community has strong parallels to where things were within the climate science community 30 – 40 years ago when lots of models were being developed and a consensus of predictions emerged that global warming was likely (graphic 5). Subsequent to those model predictions, we’ve accumulated decades of empirical evidence from global temperature monitoring networks indicating that the Earth has indeed become warmer and that the climate models were more or less right (graphic 6). I suspect a similar dynamic will play out for fish in future years as we start to look more closely in more places for the biological fingerprint of climate change.
Next time out, we’ll take a quick look at the evidence for range expansions in fish populations where temperatures have previously been too cold to support certain species, & thereafter, we’ll spend some blog time thinking in detail about how to accurately discern whether the fish skies are falling in your local streams and the rates at which they may be doing so.
Until next time, best regards,
Dan