Analyzing time activity curves from spatio-temporal tracer data to determine tracer transport velocity in plants

Abstract

Non-invasive methods utilizing tracers have a great potential to investigate carbon allocation in plants. Specifically, radioactive tracers, such as ${}^{11}\text{C}$, enable the monitoring of spatially localized transport processes on short time scales in living plants. Typically, such tracer transport experiments yield time activity curves (TACs) of tracer activity over time at various locations along a transport pathway. These TACs can exhibit different characteristic shapes that strongly depend on tracer transport dynamics, reflecting properties such as transport velocity, exchange with surrounding tissue, and tracer storage along the pathway. Various methods, either data-driven or model-based, exist to determine transport velocities from TACs. However, for some TAC shapes, the inferred carbon tracer velocity values can be inconsistent and greatly vary between analysis methods. In the present study, we review and evaluate different analysis methods for their suitability to reliably determine tracer transport velocities from typical TAC shapes. For this evaluation, we use both in silico generated and experimentally acquired TACs from positron emission tomography measurements on tomato, barley, and bean. We demonstrate that each of the compared methods can be suitable for specific TAC shapes while being less or not appropriate for others. In conclusion, we present a case-specific evaluation of methods as a reference for analyzing TACs from tracer transport experiments, which allows to ensure a robust and globally comparable determination of transport velocities.