Tracer addition experiments, particularly using isotopic tracers, are becoming increasingly important in a variety of studies aiming at characterizing the flows of molecules or nutrients at different levels of biological organization, from the cellular and tissue levels, to the organismal and ecosystem levels. However, performing rigorous statistical analyses to gain reliable quantitative insights from these experiments often remains challenging.

We present an approach based on Hidden Markov Models (HMM) to estimate nutrient flow parameters across a network, and its implementation in the R package isotracer. The isotracer package is capable of handling a variety of tracer study designs, including continuous tracer drips, pulse experiments, and pulse-chase experiments. It can also take into account tracer decay when radioactive isotopes are used.

To illustrate its use, we present three case studies based on published data and spanning different levels of biological organization: a molecular-level study of protein synthesis and degradation in Arabidopsis thaliana, an organismal-level study of phosphorus incorporation in the eelgrass Zostera marina, and an ecosystem-level study of nitrogen dynamics in Trinidadian montane streams. With these case studies, we illustrate how isotracer can be used to estimate uptake rates, turnover rates, and total flows, as well as their uncertainty. We also show how to perform model selection to compare alternative hypotheses.

The isotracer package allows researchers from a broad range of disciplines to fully take advantage of their datasets through rigorous statistical analyses. We conclude by discussing isotracer's further applications, limitations, and possible future improvements and expansions.