Control of the common carp through species-specific toxin delivery systems and biocontrol by bluegill sunfish

Abstract

The Common carp (Cyprinus carpio, or ‘carp’) is an invasive fish native to Eastern Europe and Asia and is one of the world’s most ecologically harmful species. It is known to cause issues with water clarity, increase nutrient levels, reduce aquatic vegetation, and impact waterfowl. There are many existing strategies to control carp, however, each have various issues. This has led to the search for more practical, sustainable, or broadly applicable control strategies. I investigated two emerging control strategies for carp: biocontrol of carp by bluegill sunfish, and selective toxin delivery systems. Biocontrol has been used successfully for decades in the agriculture industry; however, biocontrol for aquatic pest control is still rudimentary. Previous evidence has suggested that bluegill sunfish (Lepomis macrochirus) may be capable of controlling carp recruitment by consuming eggs and larvae of carp, however, this has never been tested in natural, whole-lakes. To test if bluegill were capable of limiting carp recruitment in natural lakes, I conducted a two-year experiment where carp were stocked into natural lakes, and bluegill were stocked in half of the lakes. The recruitment success of carp was assessed at various stages in first growing season of development: (1) the egg stage, (2) the post-larval juvenile stage, and (3) the end-of-season juvenile stage. The results indicate that bluegill predation had a major effect on the egg stage and the abundance of post-larval carp, but the abundance of end-of-season carp were affected by other processes (such as density-dependence). This is the first experiment to my knowledge that has conducted an experiment on biocontrol of a fish using multiple whole-lake manipulations. This may provide the most robust evidence to date that bicontrol of a fish is possible in natural systems. Next, I performed an analysis to examine lake characteristics that cause bluegill abundance to be low, and thus define conditions where bluegill density is too low to control carp through biocontrol. My model incorporated 12 lake and watershed variables that were used to explain variation in bluegill sunfish catch-per-unit-effort (CPUE). Of those variables, depth and water clarity had the largest effect. Specifically, bluegill abundance declined rapidly in lakes with maximum depths of less than 7 m, and a secchi disk depth less than 0.7 m. These conditions are indicative of lakes that winterkill, thus aeration may be a feasible way to stabilize bluegill populations in these systems. Last, I incorporated antimycin-a (ANT-A), a known fish toxicant, into a corn-based bait and conducted a series of experiments to determine its toxicity, leaching rate, and species-specificity. My results showed that ANT-A was lethal to carp at doses ≥ 4 mg/kg and that the amount of ANT-A that leached out of the bait in 72 h was not lethal to carp or bluegill (Lepomis macrochirus). Species-specificity trials were conducted in 227 L tanks, in which carp were stocked with three native species representing families that occur sympatrically with carp in my study region: the fathead minnow (Pimephales promelas), yellow perch (Perca flavescens) and bluegill. These trials showed high mortality of carp (46%) and fathead minnows (76%) but no significant mortality of perch or bluegill. Finally, a pond study, which used the same species composition except for fathead minnows, resulted in 37% morality among adult carp and no mortality among perch or bluegill. My results suggest that corn-based bait that contains ANT-A could be used to selectively control carp in ecosystems dominated by percids or centrarchids, such as lakes across the Great Plains ecoregion of North America, where carp are especially problematic.