The unknown third - exploring the climatic and non-climatic signals of hydrogen isotopes in tree-ring cellulose across Europe

Stable carbon (δ¹³C_c) and oxygen (δ¹⁸O_c)isotopes measured in tree-ring cellulose, together with tree-ring width (TRW), have been used extensively to investigate the effects of past climatic conditions on tree growth. By contrast, the information recorded by the third major chemical component of tree-ring cellulose, the non-exchangeable carbon-bound hydrogen, has been explored far less due to methodological drawbacks and lack of understanding of ²H-specific fractionations. In this first Europe-wide assessment we investigate the signals stored in the hydrogen isotope ratios in tree-ring cellulose (δ²H_c), from a unique collection of 100-years records, from two major genera (Pinus and Quercus) across 17 sites (36°N to 68°N).

The climate correlation analyses showed weak climate signals in the δ²H_c high-frequency chronologies, compared to those recorded by δ¹³C_c and δ¹⁸O_c, but similar to the TRW ones. The δ²H_c climate signal strength varied across the continent and was stronger and more consistent for Pinus than for Quercus. The δ²H_c inter-annual variability was strongly site-specific. Focusing on the effect of extreme climatic conditions during years with extremely dry summers, we observed a significant ²H-enrichment in tree-ring cellulose for both genera. Our findings clearly indicate that δ²H_c registers information about hydrology and climate, but it also records non-climatic signals such as physiological mechanisms associated with carbohydrates storage remobilization ²H-specific fractionations and growth.

To disentangle the climatic and non-climatic signals in δ²H_c, we investigated its relationships with δ¹⁸O_c and TRW. We found significant relationships negative between δ²H_c and TRW at 7 sites and positive between δ²H_c and δ¹⁸O_c at 10 sites, while the rest of the sites did not show any significant relationships. The agreement with the TRW chronologies confirms the relationship between growth and δ²H_c, while the divergencefrom δ¹⁸O suggests a loss of the hydrological signal in δ²H_c. These highlights, once again,a stronger physiological component in the δ²H signature independent from climate. Advancements in the understanding of ²H-fractionations and their relationships with climate, physiology, and species-specific traits are therefore needed to improve the mechanistic modeling and interpretation accuracy of δ²H_c in plant physiology and paleoclimatology. Such advancements could lead to new insights into trees’ carbon allocation mechanisms, and responses to abiotic and biotic stressors.