Meltwater is the dominant water source controlling α-cellulose δ18O in a vascular-plant-dominated alpine peatland in the Altai Mountains, Central Asia

Abstract

Few studies have fully considered how hydrological inputs and climate influence the linkage between the oxygen isotopic composition of α-cellulose (δ18Ocell) and their source waters in vascular-plant-dominated alpine peatlands. In this study, in order to test whether δ18Ocell can be used as a hydroclimatic indicator, we investigated the relationship between the δ18O values of cellulose extracted from the dominant plant species (Carex pamirensis) and those of potential source waters during two growing seasons (from May to September) in the Sahara sand peatland in the southern Altai Mountains. Concurrently, the Roden–Lin–Ehleringer (RLE) mechanistic model was applied to better understand oxygen isotopic fractionation during cellulose synthesis. We found that the meltwater from snow/ice cover and seasonally-frozen soil played a critical role in controlling the hydrological processes in this alpine peatland. The δ18O values of both swale water and soil water responded more sensitively to variations in inflow meltwater than did summer precipitation. IsoSource modelling demonstrated that the mean 76–24% split between meltwater and summer precipitation inputs determined the overall isotopic composition of the peatland water. The slopes of the peatland water line (6.38) and meltwater line (6.87) were slightly lower than that of the local meteoric water line (LMWL) (7.72), suggesting limited evaporation from these external water sources due to the relatively high humidity (73–80%) in the studied fen. However, the slopes of the stem water line (4.53) and leaf water line (3.57) were generally lower than that of the LMWL, indicating that the internal plant waters have experienced isotopic enrichment during transpiration. Interestingly, the variations of δ18Ocell (of shoots, stems and leaves) fall within a relatively narrow range (<±3‰), which is mainly attributed to a nearly constant monthly net 18O enrichment factor (11.13 ± 1.07‰ for ΔModelled), together with a smoothing effect. Furthermore, empirical correlation and RLE modeling both showed the co-variation of external source water and cellulose synthesis, demonstrating that the cellulose oxygen isotope composition of C. pamirensis can faithfully reflect variations in soil/swale water. These findings strongly suggest that inflow meltwater is the dominant water source in determining δ18Ocell, implying that the δ18Op signals of precipitation in the winter half-year (October–April) may be inherited by the δ18Ocell values of C. pamirensis at the study site. Therefore, our results provide important insights for the hydroclimatic interpretation of past variations in δ18Ocell values in vascular plants in alpine peatlands.