Simulating data and fitting models for eDNA studies can be accomplished using the artemis package. In particular, artemis implements a Bayesian latent-variable model in which the predictors affect a latent variable, eDNA concentration. This latent variable is related to an observed variable, CQ cycles via a standard curve calibration. CQ cycles, are often truncated at a certain limit, where more than X cycles is considered a non-detection.
Additional details on the model implemented by the artemis package can be found in the "Getting Started" vignette.
Stan Development Team (2018). RStan: the R interface to Stan. R package version 2.18.2. http://mc-stan.org