Linking sap flow dynamics of birch tree to atmospheric and rainfall conditions

Sap flow is one of the best direct indicators of transpiration at the tree level, which is a major component of the water balance. Understanding daily sap flow variability is essential for interpreting tree water consumption and its response to environmental conditions. This study aimed to classify days according to the behaviour of daily sap flow and determine whether the resulting categories differ in terms of meteorological and precipitation conditions. We used sap flow measurements from a birch tree located in a small urban park in Ljubljana, Slovenia. Along with meteorological variables such as solar radiation (SR), air temperature (T), vapor pressure deficit (VPD), wind speed (WS), and amount of precipitation, we analyzed the shape of daily sap flow curves for the fully leafed birch tree in 2025. For each day, we smoothed the daily sap flow curves and analyzed them using principal component analysis (PCA) and obtained daily PC1-PC3 values to compactly describe the curve shape. Based on these values, we grouped the curves and selected the number of clusters using information criteria (BIC). We tested for differences between clusters using nonparametric tests and interpreted the clusters in terms of meteorological conditions and precipitation amount. Five characteristic clusters of daily sap flow curves were defined, differing in curve shape, peak time and magnitude, and midday depression intensity. The clusters differed most significantly in terms of SR, while cluster 4 stood out from the others in terms of precipitation. This is also reflected in the greater variability of daily curve shapes within cluster 4. Despite the known positive correlation between VPD and sap flow, the days in cluster 5 showed a weak negative correlation, which is consistent with their pronounced midday depression. In contrast, cluster 1, which is characterized by the lowest values of SR, T, and VPD, exhibits the flattest average daily sap flow curve and is the only cluster without a pronounced midday depression. This is consistent with the known physiological regulation of transpiration at higher VPD values, when plants limit water flow by closing their stomata.

Acknowledgment: This work was supported by the research program P2-0180 through the Ph.D. grant of the first author that is financed by the Slovenian Research and Innovation Agency (ARIS). It is also part of the ongoing research project entitled “Evaluation of the impact of rainfall interception on soil erosion” supported by the Slovenian Research and Innovation Agency (J2-4489) and the Austrian Science Fund (FWF) I 6254-N.