The abundance of environmental DNA (eDNA) in water samples has been proposed as a sensitive, cost-efficient, and non-invasive alternative to infer population abundance and biomass, based on the unstated assumption that all organisms within a population shed the same amount of eDNA. We tested how metabolic, nutritional, and life history processes shape intraspecific eDNA shedding rates in the freshwater invertebrate Daphnia magna. We extracted water sample from each D. magna individual that was raised in a glass vial under a 2x2 longitudinal factorial manipulation of temperature (15oC vs 25oC) and food levels (high and low concentrations of green algae Chlorella vulgaris) over their entire lifespan, and quantified eDNA daily shedding rates using digital droplet PCR (ddPCR). We measured individual D. magna body mass and recorded any observed life history characteristics including molting, reproductive status, and mortality at daily intervals. These repeated individual measurements in combination with eDNA shedding rates curated a quantitative framework that allowed us to mechanistically evaluate the following suite of hypotheses: eDNA shedding rates depend on (1) individual metabolic rate as determined by body mass and temperature, (2) nutritional conditions of varying food levels, and (3) life history events such as growth, reproduction, and mortality.
Results/Conclusions
Analyzed using a nested mixed-effect modelling framework, we showed that D. magna eDNA shedding rates varied by an order of magnitude over the course of each individual’s lifespan as a result of simultaneous processes. The total eDNA (log-transformed mean ± sd, 3.236 ± 2.41) ranged from 0.3 to 15184.65 copies uL-1 Day-1. Changes in body mass (β = 0.05, CI = 0.04 – 0.06, P < 0.001) and reproductive status ((β = 1.69, CI = 00.93 – 2.45, P < .001) were the two most significant factors driving eDNA shedding rates in D. magna. Residual variation in eDNA shedding rate was enhanced under conditions of high ambient temperature and abundant food conditions (β = 1.95, CI = 1.04 - 2.87, P < 0.001) and was higher for recently deceased individuals (β = 0.84, CI = 0.03 – 1.65, p = 0.042). Our work supplies a more nuanced understanding about myriad factors that shape eDNA production, suggesting new and more useful ways to interpret eDNA monitoring data. We suggest that future work using eDNA to estimate population abundance or biomass should account for both ecological and energetic conditions facing their target organism, especially when working with size-structured populations.
