An innovative monitoring approach to measure spatio-temporal throughfall patterns in forests

Precipitation is partitioned and redistributed by vegetation when it passes through a forest canopy resulting in interception, stemflow and throughfall. Throughfall is known to be spatially highly heterogeneous beeing lower than precipitation in certain areas and higher in others. As these emerging spatial throughfall patterns infiltrate, they may propagate into soil moisture patterns and influence root water uptake, percolation, runoff generation and ultimately the entire forest water balance. Despite the relevancy of throughfall in water resources research, there is a scarcity on experimentally-derived high-quality datasets on its spatio-temporal dynamics. Sampling procedures changed little over the past decades and are often not optimal for systems under study. Large sampling efforts especially for complex vegetation structures limit most studies to investigate on either high temporal or spatial resolution of throughfall.

We present an innovative throughfall sampling approach for continuous measurement of throughfall at high spatial resolution. The sampling scheme allows to quantify the spatio-temporal throughfall variability at both intra-event and intra-stand levels and assess spatial throughfall patterns and their temporal persistence across precipitation events of varying size during leafed and non-leafed periods. 60 self-built throughfall sampler boxes featuring a cost-efficient design with four troughfall collectors and tipping bucket units each, were distributed in a stratified sampling design in forest plots of pure and mixed Beech, Douglas and Silver fir (total area 0,4 ha). The tipping buckets are controlled with newly developed micro boards connected to data loggers so that the network measures continuously, automatically and requires minimal maintenance during precipitation events. The sampler boxes operate with an inlay of litter material on top of a grid and mesh, which allows to include forest floor interception as part of the overall throughfall process and reduces throughfall splash. The placement of the boxes at the forest floor boxes is minimally invasive as quantified throughfall can percolate into the soil below allowing to monitor soil moisture patterns in the same plot. The network of sampler boxes is supplemented with classical throughfall samplers, stemflow- and lysimeter measurements at every plot. Throughfall data were analysed from the network at a Beech, Douglas fir, Silver fir and mixed plot for the first three months of measurement. Using spatial correlation and temporal stability analysis of the observed data, we show spatio-temporal throughfall dynamics of different tree species, during and among precipitation events of varying size and for a stand that develops from dormancy to vegetation period.