Abstract
Measurement of environmental DNA (eDNA) is becoming a common technique to survey for rare and invasive fish due to its sensitivity and specificity. However, its utility is limited by an incomplete understanding of factors governing its sources and fates. Failure to detect eDNA is especially difficult to interpret so surveillance techniques often collect large numbers of samples across broad regions. If, however, fish could be reliably attracted to a single location where their eDNA could be easily measured that would be useful. We conducted a proof-of-concept study of this idea using invasive Common Carp. We monitored the distribution of radio-tagged Carp and their eDNA across a 67 ha lake focusing at the bait site while a pheromone (Prostaglandin F2α; PGF 2α) was also measured to determine their reproductive condition. Prior to baiting, Carp were patchily distributed and while eDNA was occasionally detectable, it was patchy and only loosely associated with moderately dense groups of fish. Further, neither Carp, nor their eDNA were consistently measurable at the bait site and surrounding region, and the pheromone was not measurable at all. However, once baiting commenced, Carp started visiting the bait site and feeding, especially at night, where eDNA levels increased 500-fold as fish densities doubled and PGF 2α became detectable. Fish presence, eDNA and pheromone concentrations peaked at night after 6 days, strongly suggesting feeding activity was the main driver. While the presence of eDNA precisely coincided with this aggregation, levels had dropped dramatically within 5 m. PGF 2α levels dropped less rapidly and demonstrated the presence of live mature fish. We suggest that food could be used to train fish to come to locations where they otherwise are too scarce to be reliably measured, increasing their eDNA release, making them measurable, and their reproductive condition also discernable by measuring pheromones.