Parasite-mediated food web dynamics

The ways that parasites affect host ecological interactions can influence the net impact of parasties on host survival. This is especially true for predation because predators almost always kill their prey. If parasites make you more vulnerable to predators, that might be pretty important relative to other costs of parasitism! Plenty of theoretical studies have shown that host-parasite dynamics are affected by whether parasites make hosts more, or less, susceptible to predation. But empirical studies are rare, particularly for multi-host systems where generalist predators may further complicate things!

Previous research on the effects of sea lice on juvenile salmon found that predators prefer a lousy lunch, but our paper recently published in Ecosphere suggests that the effect of predation on infested juvenile pink and chum salmon might not be the same. We looked at how predation by coho salmon mediated the impact of sea lice on sympatric juvenile pink and chum salmon. The results of two years of experiments (at the Shenty, right) showed that coho preferred to eat pink salmon over chum salmon! Coho also seemed to pick off prey that had the most sea lice, consistent with previous work. Together, these findings suggest that lousy pink salmon are the most at risk of being eaten. In fact, if there are enough pink salmon around to satiate predators, predation on chum salmon may even decline with sea lice because predators are so focused on those lousy pinks. This might explain why mortality in pink salmon populations seems to be higher in years of high sea lice, but that trend doesn’t hold for chum salmon.

This study emphasizes that there’s only so much we can learn in a lab; field-based studies that investigate the broader ecosystem effects of parasites are needed to assess their impact on hosts.

Footnote: This was my first time trying to publish "non-significant" results. This paper was rejected from two "higher impact" journals despite very positive reviews. One reviewer even stated, "This is good, honest science done well - but unfortunately any signal in the data is not strong..." Fair enough. I ended up trying to focus on biological significance rather than statistical signifciance, by using a mathematical model to investigate some "what if" scenarios. Pretty happy to have it finally published in Ecosphere!