A statistical model for calibration and computation of detection and quantification limits for low copy number environmental DNA samples

Environmental DNA (eDNA) has been increasingly utilized by academic, industry, and government groups for environmental monitoring due to its efficiency in regards to both time and cost, as well as non‐invasiveness to target organisms, and reduced dependency on trained biologists for sample collection. The methods typically employ quantitative real‐time polymerase chain reaction (qPCR) to detect the presence of a given organism's DNA in a sample. Currently, there is a drive to use qPCR data to infer biomass or abundance by quantitating the copy number or concentration of a given target gene fragment in a sample, which is often very dilute. Before eDNA can be fully accepted as an environmental decision‐making tool, however, certain aspects of the methods require standardization, including the quantification of target DNA in low copy number samples. Models that are not able to properly make use of data from highly dilute samples are severely hampered in their definitions of the limits of detection and quantification at the lower end of the detection curve. We propose a statistical model for a standard curve that relates the number of qPCR‐detected technical replicates to the copy number in the case of low copy number samples. Likelihood methods are used to estimate the parameters of the model and we derive inverse regression estimates together with their standard errors. Limits of copy number detection and quantification, and their confidence intervals are derived using a well‐accepted statistical approach thus providing a more broadly applicable and robust method for reporting eDNA abundance into the low copy number range. The method is illustrated using experimental results from multiple laboratories.