Tree-ring δ18O climate signals vary among tree functional types in South Asian tropical moist forests.

We present the first annually resolved and statistically reliable tree-ring δ18O (δ18OT) chronologies for the three South Asian tropical moist forest tree species (Chukrasia tabularis A. Juss., Toona ciliata M. Roem., and Lagerstroemia speciosa Roxb.) which differ in their shade tolerance and resistance to water stress. We found significantly higher mean δ18OT values in light-demanding T. ciliata than in intermediate shade tolerant C. tabularis and shade tolerant L. speciosa (p < 0.001). δ18OT in C. tabularis was mainly influenced by pre-monsoon vapor pressure deficit (VPD; r = -0.54, p < 0.01) and post monsoon maximum temperature (Tmax) (r = 0.52, p < 0.01). δ18OT in T. ciliata was strongly negatively correlated with a dry season drought index PDSI (r = -0.65, p < 0.001) and VPD (r = -0.58, p < 0.001). Pre-monsoon Tmax was strongly positively linked with δ18OT in L. speciosa (r = 0.65, p < 0.001), indicating that climatic influences on δ18OT are species-specific and vary among tree functional types. Although there was a week correlation between local precipitation and δ18OT in our studied species, we found a strong correlation between δ18OT and precipitation at a larger spatial scale. Linear mixed effect models revealed that multiple factors improved model performance only in C. tabularis, yielding the best model, which combined VPD and Tmax. The top models in T. ciliata and L. speciosa included only the single factors PDSI and Tmax, highlighting that the way C. tabularis interacts with climate is more complex when compared with other two species. Our analyses suggest that stable oxygen isotope composition in tree rings of South Asian tropical moist forest trees are a suitable proxy of local and regional climate variability and are an important tool for understanding the physiological mechanisms associated with the global hydrological cycle.