Humans are good at altering ecosystems, even unintentionally. Human actions can also alter selection pressures on organisms, driving new evolution. One example of such evolution that demands urgent attention is fisheries-induced evolution, in which harvest drives trait change in fish. The details of fisheries-induced evolution are fairly straightforward: larger, bolder fish are more likely to be caught by fishing gear, leaving smaller, shier individuals behind. The result is an evolved fishery, which may interact differently with its environment and have lower long-term stability.
 |
| Harvest drives changes in fish behavior and morphology |
Growing evidence for trait change in harvested species has led to calls for fishery evolutionary impact assessments, which consider the effect of harvest-induced evolution on fishery yield, sustainability, and recovery. However, little work has considered the potential for evolution in lower, non-target trophic levels to impact fisheries.
Our recent modeling experiments reveal a harvest-driven eco-evolutionary trophic cascade: Declines in harvested species can drive cascading changes down food webs, altering selection pressures on lower trophic levels. Evolution in these lower trophic levels can cause cascading changes back up the food web, ultimately altering the abundance of the harvested species.
 |
| Harvest-induced evolution in lower trophic levels predictably feeds-back to affect the harvested species. From Wood et al 2018, Scientific Reports. Used with a CC license. |
The direction and impact of evolution in lower trophic levels is predictable, according to our models. Odd numbered trophic levels (referring to position below the harvested species, e.g. secondary consumers in our table) have fewer prey and predators during harvest, and thus evolve increased competitive ability, at the cost of lower defense. These numerous, vulnerable individuals give a bottom-up boost to the food web, which can bolster the harvested species. Even numbered trophic levels (e.g. primary consumers in our table) have more predators and prey, and thus evolve increased defenses, at the cost of competitive ability. These less numerous, more defended individuals undermine the trophic levels above them, thus ultimately undercutting the harvested species. The net effect of evolution on the harvested species therefore depends on which trophic levels evolve, with potential for evolution at multiple trophic levels to be cancelling or synergistic.
Our models provide some insights on where effects of evolution in lower trophic levels might have the strongest effect on harvested species. Species with abundant variation, even tradeoffs between competition and defense, and strong interactions with predators and prey are most likely to drive the eco-evolutionary dynamics we observed. We suggest that metrics of trophic cascade strength or interaction strength are a good first clue towards finding systems where eco-evolutionary feedbacks are of particular concern.
Our model does not merely indicate another source of doom for fisheries. Indeed, it suggests that some evolution in non-target trophic levels might even help offset some more direct effects of harvest. Rather, our model highlights the importance of understanding evolution and eco-evolutionary feedbacks in communities, rather than just a single species, when approaching conservation problems.
The major next step of this work—one which we hope will be investigated by evolutionary biologists and ecologists alike—is to collect the data necessary to detect eco-evolutionary feedbacks from non-target species. Monitoring phenotypes of lower trophic levels in harvested ecosystems, as well as mining current datasets for evidence of trait change in non-target species, are two possible ways forward. If you have any ideas, feel free to get in touch!
Making a modeler
I started my PhD at the University of Maine in fall 2015. Naively intending to immediately begin some intrepid and fulfilling experiments with mosquitofish, I quickly languished under long waits at the hands of state government, institutional animal care, and facilities. It seemed I would have an endless PhD, devoid of any fishy work.
 |
| Advisor Mike Kinnison demonstrating his fishing skills. |
As a side project to preserve my dwindling patience and sanity, I began working on a short R script to model two-species eco-evolutionary dynamics. After some encouragement from my advisors Mike Kinnison and Eric Palkovacs, I expanded my model to include three, then four evolving trophic levels. Running against the limits of processing power on my laptop, I translated the entire script to Matlab (though not without several angry weekends of debugging), and used some limited funding to purchase a lab computer. Mike eventually intervened, encouraging me to channel my newfound obsession into a productive problem: fisheries induced evolution. One year, several lost weekends, and two failed attempts at beards later, we had a new theory on fisheries induced evolution and trophic cascades. And I finally started working on some real fish.
 |
| Looking for eco-evolutionary trophic cascades with Eric Palkovacs' UCSC lab. |
Three years into my degree, I now have more real-world fishy data than ever needed to occupy my time. In retrospect, I think starting my degree program with some modelling was a great opportunity. It’s hard to know if I would have had the same bandwidth to launch into a similar model at this point in my program, and the models have helped me focus my field and lab work. Based on my experience, my tip to anyone contemplating some modelling as part of their dissertation is to start early when everything else is tied up in limbo. The nice thing about models is that most can be adapted as your dissertation morphs and grows.
Zach Wood is a PhD Candidate in Mike Kinnison's Evolutionary Applications Lab at the University of Maine, and definitely not a morning person.
No comments:
Post a Comment