Temperature Controls eDNA Persistence across Physicochemical Conditions in Seawater

Abstract

Environmental DNA (eDNA) quantification and sequencing are emerging techniques for assessing biodiversity in marine ecosystems. Environmental DNA can be transported by ocean currents and may remain at detectable concentrations far from its source depending on how long it persist. Thus, predicting the persistence time of eDNA is crucial to defining the spatial context of the information derived from it. To investigate the physicochemical controls of eDNA persistence, we performed degradation experiments at temperature, pH, and oxygen conditions relevant to the open ocean and the deep sea. The eDNA degradation process was best explained by a model with two phases with different decay rate constants. During the initial phase, eDNA degraded rapidly, and the rate was independent of physicochemical factors. During the second phase, eDNA degraded slowly, and the rate was strongly controlled by temperature, weakly controlled by pH, and not controlled by dissolved oxygen concentration. We demonstrate that marine eDNA can persist at quantifiable concentrations for over 2 weeks at low temperatures (≤10 °C) but for a week or less at ≥20 °C. The relationship between temperature and eDNA persistence is independent of the source species. We propose a general temperature-dependent model to predict the maximum persistence time of eDNA detectable through single-species eDNA quantification methods.

Keywords: coral; deep sea; environmental DNA; marine; persistence.

Figure 1

(A) Matrix of target conditions for 11 combinations of temperature, pH, and [DO] investigated to determine the persistence of eDNA from the coral Lophelia among a range of marine physicochemical states. Two replicate experiments were conducted for each combination of temperature, pH, and [DO]. At 20 °C temperature, half- and fully saturated oxygen experiments (shaded light blue) represent conditions characteristic of subtropical near-surface environments. At 4 °C temperature, half-saturated experiments (shaded dark blue) represent conditions characteristic of the deep-sea environment. The remaining cells (shaded blue) represent other conditions characteristic of the global open and deep ocean. (B) Schematic example of experimental conditions. Compressed air and nitrogen gas flow rates were adjusted to reach the target [DO]. Small doses of 0.5 M HCl were automatically administered to the tanks when target pH values were exceeded. Physicochemical measurements were monitored by suspending probes in the tanks. Caps were secured with O-rings to control the oxygen concentration in the above headspace. All tubing and probes were placed through small openings in these caps. A photograph of the experimental setup is presented in Figure S1. (C) Measurements of temperature, pH, and [DO] over the 22 eDNA degradation experiments. Points indicate individual measurements, and paths are drawn through daily averages. All measurements were recorded daily and at each sampling time point. Temperature measurements were made using a laser thermometer. pH and [DO] measurements were made with probes.

Figure 2

Degradation of Lophelia eDNA in 22 experiments among a range of marine physicochemical states. eDNA concentration was measured as the concentration of a 154 base pair fragment of the Lophelia mitochondrial COI gene. The concentration of eDNA (y-axis) is plotted against time (x-axis). The y-axis is natural-log scaled. Different colors represent the two experimental replicates at each experimental condition. Points represent the average of three qPCR replicate measurements for two samples at a given time point. Lines represent the fit to a biphasic model. Thin lines connecting points to the line of best fit represent the distance from the observed to the fitted values at each time point (the residuals). Panels are arranged by temperature (descending top to bottom) and [DO] (increasing left to right). Values shown on top of each panel indicate the target experimental conditions. Points below the limit of quantification of the qPCR assay (77.8 copies/reaction or 13 copies/mL seawater filtered) are not plotted. The dashed line indicates the limit of quantification.

Figure 3

Decay rate constant (k) estimates for the initial (A) and second (B) phases of eDNA degradation for 22 experiments conducted across 11 combinations of temperature, pH, and [DO]. Decay rate constants were estimated by fitting an exponential decay equation with initial and second degradation phases with different rates (biphasic). Decay rate constants are arranged on the x axis by the average pH over the course of each experiment. Vertical error bars represent 95% confidence intervals for the decay rate constants. The color of each point represents the average [DO] over each experiment.

Figure 4

eDNA persistence time (time until degradation of 99.9% of starting eDNA concentration), estimated using decay rate constants from our study and other published marine studies, as a function of temperature. The linear model was only fit to decay rate constants calculated from simple exponential models. The best fit line is indicated by the dashed line, and the shaded region represents the 95% confidence interval for the slope. Points are dodged slightly from actual recorded temperatures to improve visualization. (Inset) Comparison of calculated persistence times when fitting a single exponential versus a biphasic model to the data in this study. Data are reported in Table S7.