Jessica J. Eichmiller

The Relationship between the Distribution of Common Carp and Their Environmental DNA in a Small Lake

Although environmental DNA (eDNA) has been used to infer the presence of rare aquatic species, many facets of this technique remain unresolved. In particular, the relationship between eDNA and fish distribution is not known. We examined the relationship between the distribution of fish and their eDNA (detection rate and concentration) in a lake. A quantitative PCR (qPCR) assay for a region within the cytochrome b gene of the common carp (Cyprinus carpio or ‘carp’), an ubiquitous invasive fish, was developed and used to measure eDNA in Lake Staring (MN, USA), in which both the density of carp and their distribution have been closely monitored for several years. Surface water, sub-surface water, and sediment were sampled from 22 locations in the lake, including areas frequently used by carp. In water, areas of high carp use had a higher rate of detection and concentration of eDNA, but there was no effect of fish use on sediment eDNA. The detection rate and concentration of eDNA in surface and sub-surface water were not significantly different (p≥0.5), indicating that eDNA did not accumulate in surface water. The detection rate followed the trend: high-use water > low-use water > sediment. The concentration of eDNA in sediment samples that were above the limit of detection were several orders of magnitude greater than water on a per mass basis, but a poor limit of detection led to low detection rates. The patchy distribution of eDNA in the water of our study lake suggests that the mechanisms that remove eDNA from the water column, such as decay and sedimentation, are rapid. Taken together, these results indicate that effective eDNA sampling methods should be informed by fish distribution, as eDNA concentration was shown to vary dramatically between samples taken less than 100 m apart.

Optimizing techniques to capture and extract environmental DNA for detection and quantification of fish

Few studies have examined capture and extraction methods for environmental DNA (eDNA) to identify techniques optimal for detection and quantification. In this study, precipitation, centrifugation and filtration eDNA capture methods and six commercially available DNA extraction kits were evaluated for their ability to detect and quantify common carp (Cyprinus carpio) mitochondrial DNA using quantitative PCR in a series of laboratory experiments. Filtration methods yielded the most carp eDNA, and a glass fibre (GF) filter performed better than a similar pore size polycarbonate (PC) filter. Smaller pore sized filters had higher regression slopes of biomass to eDNA, indicating that they were potentially more sensitive to changes in biomass. Comparison of DNA extraction kits showed that the MP Biomedicals FastDNA SPIN Kit yielded the most carp eDNA and was the most sensitive for detection purposes, despite minor inhibition. The MoBio PowerSoil DNA Isolation Kit had the lowest coefficient of variation in extraction efficiency between lake and well water and had no detectable inhibition, making it most suitable for comparisons across aquatic environments. Of the methods tested, we recommend using a 1.5 μm GF filter, followed by extraction with the MP Biomedicals FastDNA SPIN Kit for detection. For quantification of eDNA, filtration through a 0.2–0.6 μm pore size PC filter, followed by extraction with MoBio PowerSoil DNA Isolation Kit was optimal. These results are broadly applicable for laboratory studies on carps and potentially other cyprinids. The recommendations can also be used to inform choice of methodology for field studies.

Effects of Temperature and Trophic State on Degradation of Environmental DNA in Lake Water

Degradation of environmental DNA (eDNA) in aquatic habitats can affect the interpretation of eDNA data and the ability to detect aquatic organisms. The effect of temperature and trophic state on the decay of Common Carp (Cyprinus carpio) eDNA was evaluated using lake water microcosms and quantitative PCR for a Common Carp-specific genetic marker in two experiments. The first experiment tested the effect of temperature on Common Carp eDNA decay. Common Carp eDNA exhibited exponential decay that increased with temperature. The slowest decay rate was observed at 5 °C, with a T90 value (time to 90% reduction from initial concentration) of 6.6 days, as opposed to ∼1 day at higher temperatures. In a second experiment, decay was compared across waters from lakes of different trophic states. In this experiment, Common Carp eDNA exhibited biphasic exponential decay, characterized by rapid decay for 3-8 days followed by slow decay. Decay rate was slowest in dystrophic water and fastest in oligotrophic water, and decay rate was negatively correlated to dissolved organic carbon concentration. The overall rapid decay of eDNA and the effects of temperature and water quality should be considered in protocols for water sample storage and field sampling design.

Distribution of genetic markers of fecal pollution on a freshwater sandy shoreline in proximity to wastewater effluent

Water, sand, and sediment from a Lake Superior harbor site continuously receiving wastewater effluent was sampled monthly for June to October 2010 and from May to September 2011. Understanding the dynamics of genetic markers of fecal bacteria in these matrices is essential to accurately characterizing health risks. Genetic markers for enterococci, total Bacteroides, and human-associated Bacteroides were measured in site-water, sand, and sediment and in final effluent by quantitative PCR. The similarity between the quantity of molecular markers in the water column and effluent indicated that the abundance of genetic markers in the water column was likely controlled by effluent inputs. Effluent turbidity was positively correlated (p ≤ 0.05) with AllBac and HF183 in final effluent and AllBac in the water column. In sand and sediment, Entero1 and AllBac were most abundant in the upper 1-3 cm depths, whereas HF183 was most abundant in the upper 1 cm of sand and at 7 cm in sediment. The AllBac and Entero1 markers were 1- and 2-orders of magnitude more abundant in sand and sediment relative to the water column per unit mass. These results indicate that sand and sediment may act as reservoirs for genetic markers of fecal pollution at some freshwater sites.

Correlations between pathogen concentration and fecal indicator marker genes in beach environments.

The concentrations of several potential pathogens were measured using a microfluidic quantitative PCR (MFQPCR) platform in beach water, sand, and sediment samples collected in Duluth, MN. Among the 19 pathogen marker genes examined, eaeA from Escherichia coli and plc from Clostridium perfringens were most frequently detected in all samples. In beach water and wastewater samples, positive correlations were observed between quantities of potential pathogens and most of the fecal indicator genetic markers. Such correlations, however, were not frequently observed in sand and sediment samples. Our results suggest that the behavior of potential pathogens and FIB may vary by sample type and source of contamination. Consequently, appropriate FIB marker genes need to be chosen for reliable water/sand quality monitoring.

Attracting Common Carp to a bait site with food reveals strong positive relationships between fish density, feeding activity, environmental DNA, and sex pheromone release that could be used in invasive fish management

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.