Intra-annual tree-ring cellulose δ2H as an indicator of drought

Temporal variability of tree-ring cellulose δ²H (δ²H_ring-cel) can be a unique tool for understanding tree physiology and climate. However, we do not fully understand the drivers of temporal variability in δ²H_ring-cel. Investigating seasonal δ²H_ring-cel in boreal forests is particularly challenging. Previous studies on intra-annual tree-ring δ¹⁸Ohave shown that tree-ring isotope variability can result from the combined but opposing effects of source water and leaf assimilates, a dynamic likely relevant for δ²H_ring-cel as well. To be able to use δ²H_ring-cel as a standalone and reliable bioindicator, it is important to understand the variable hydrogen isotope fractionation between source water and tree rings. Our study aimed to provide context to this variability in a natural forest by being the first study to trace intra-annual δ²H_ring-cel to the δ²H of its sources and drought indices.

The δ²H of source water, leaf water and carbohydrate pools (i.e. water-soluble carbohydrates, starch) were analysed from five pine (Pinus sylvestris) trees during 2019 at Hyytiälä forest, Finland. Their δ²H were used to model continuous δ²H of source water (δ²H_source) and leaf sugars (δ²H_leaf-sug). Modelled and measured δ²H_leaf-sug matched moderately for 2019, and the model was applied to predict δ²H_leaf-sug during 2018. Intra-annual δ²H_ring-cel were analysed in these two years at a resolution of 5-10 timepoints per year, and they were allocated to xylogenetic timepoints. They were then compared to time-integrated δ²H_source, δ²H_leaf-sug, net assimilation rate, evapotranspiration and drought indicators.

Carbohydrate δ²H was significantly different among leaves, branches and stems. δ²H_ring-cel had strong time-integrated relationships to modelled δ²H_source, net leaf assimilation rate and evapotranspiration, but the direction of their relationships was different between years. At monthly resolution, water-soluble carbohydrate δ²H measured from one year-old needles had a strong, positive relationship to δ²H_ring-cel. Similarly, the modelled δ²H_leaf-sug, had strong positive relationships to δ²H_ring-cel, which were robust between years. δ²H_ring-cel also had strong relationships to Standardized Soil Moisture Index (SSMI). 

We show that the role of δ²H_leaf-sug superseded the role of δ²H_source in intra-annual δ²H_ring-cel, because δ²H_leaf-sug had a consistent relationship to intra-annual δ²H_ring-cel in both years while δ²H_source did not. This clearly supports the growing body of evidence that δ²H_ring-cel is strongly mediated by physiological processes. Our results show promise for δ²H_ring-cel functioning as a bioindicator of soil drought related physiological stress signals in long-term tree ring chronologies.