Characterization of the plant-derived isotopic proxies for precipitation reconstruction in the tropics

Abstract

Proxy-based paleoclimate reconstructions provide climate data extending back to thousands to millions of years. These reconstructions not only contribute to our understanding of the past climate but also play a crucial role in validating climate models that can be used to project future climate changes. For past precipitation reconstructions, a variety of climate proxies are used. However, the reconstructions are often qualitative, limiting their applicability. To quantify past precipitation variability, it is essential to apply a transfer function to the climate proxies. This requires characterization of the response of the climate proxies and the establishment of an empirical relationship between their response and concurrent precipitation. This thesis work focuses on characterizing various plant-derived isotope proxies widely utilized for past precipitation reconstructions, viz. (i) the hydrogen isotopic composition (D) of leaf wax n-alkanes and n-alkanoic acids, (ii) plant-derived carbon (13C) and D isotopes in soil and plants. The emphasis is on characterization of the plant-derived proxies from the tropical climate. Quantifying the hydrogen isotope fractionation between leaf wax compounds and source water (ε_app) is a prerequisite for leaf wax-based paleo-hydrological studies. However, the characterization of the ε_app values, mostly done in field-based studies, are predominantly carried out in northern mid-latitude regions as compared to that in the tropics. Further, the ε_app values estimated in field-based studies are often associated with inherent uncertainties that could stem from (i) incorrect source water D values, (ii) species-effect, and (iii) varying climatic conditions (as in transect studies). Hence, to characterize the ε_app values in the tropics and to decouple the factors affecting the variability of ε_app, an outdoor experiment was conducted wherein four evergreen and three deciduous angiosperm trees were grown under similar climatic conditions for 85 days with water of known δD value (−2‰). The ε_app values in the studied species were −119 ± 23‰ (n = 14) for n-alkanes and −126 ± 27‰ (n = 12) for n-alkanoic acids of chain lengths C31 and C30, respectively. Inter-species variabilities in ε_app values that are consistent with previous field and transect studies were observed. As the plants were grown under similar climatic conditions and irrigated with water of the same δD value, the variability in ε_app values observed here suggested that the species-specific hydrogen isotopic fractionation likely has a dominant control over the uncertainty in the community-averaged ε_appvalues. Further, the ε_app values of deciduous and evergreen species showed no systematic differences, suggesting that changes in the relative proportion of these taxa may not affect the community-averaged ε_app and the reconstructed D values of paleo-precipitation in angiosperm tree-dominated catchments. The climatic parameters controlling the hydrogen isotope composition (D) of leaf-wax n-alkanes and n-alkanoic acids, used widely in paleohydrologic research, are fairly known. However, the climatic parameters of exactly which period (i.e. early or entire period of the leaf’s lifespan) these biomarkers represent is unclear. A longer duration experiment was conducted, wherein, tropical deciduous and evergreen species were grown with the normal (δD = −2‰) water followed by the isotopically-labeled water (δD = 1000‰) during a growing season. The observed differences between the δD values of n-alkanes (and n-alkanoic acids) of the leaves collected later and prior to the application of the isotopically-labeled water were 38 ± 38‰ (and 32 ± 42‰). These differences were much smaller than the corresponding expected differences that varied from 272 ± 118 to 547 ± 127 for n-alkanes and 272 ± 118 to 547 ± 133‰ for n-alkanoic acids, indicating insignificant (7-9%) production of these biomarkers in the mature leaves of deciduous and evergreen species. The δD analysis of n-alkanes and n-alkanoic acids of new leaves emerging during the next growing season in deciduous species indicated minor incorporation of the previous year’s photosynthates in the leaf wax pool of the current year’s mature leaves. The overall positive correlation between the δD values of n-alkanes and n-alkanoic acids (r2= 0.7, n = 60, p < 0.05) indicated synchronous synthesis of these biomarkers. Unlike inferences from extra-tropical studies, this work suggests the δD records of both biomarkers are biased towards the climatic conditions prevailing during the early stages of the leaf’s growth. The climate models aimed at reproducing the leaf wax δD-based hydroclimatic reconstructions should consider this bias. Precipitation plays a crucial role in controlling the δ13C and δD values of plants. The organic matter derived from plants, when integrated into the soil and sedimentary archives, carries valuable information about past precipitation conditions. In this study, the possibility of establishing a relationship between various plant-derived isotopic proxies and precipitation amount was explored along a precipitation gradient ~500 to 3700 mm/yr in western peninsular India. It was observed that the δ13C values of soil organic matter (δ13CSOM) and n-alkanes (δ13Calk) in soil exhibited a negative correlation with the precipitation amount. The calibration equation derived from δ13CSOM values and precipitation amount provides a means for past precipitation reconstruction from δ13CSOM values of paleosols in western peninsular India. The δ13Calk values of different n-alkane homologues in soil correlated negatively with the precipitation amount, as did the δ13CSOM; measurement of δ13Calk values in soil (which is analytically more challenging) may not be necessary for paleo precipitation reconstruction, as their interpretations would yield outcomes similar to that derived from the δ13CSOM studies. The δD values of n-alkanes (δDalk) in soil showed a positive correlation with precipitation amount, contrary to the expected negative correlation arising due to the ‘amount effect’ in tropical precipitation. This discrepancy indicated the influence of local vegetation composition (i.e. tree vs grasses) on the measured δDalk values in soil. Therefore, a vegetation composition correction might be necessary before reconstructing the δD values of precipitation from the measured δDalk values in various archives. The δ13C values of C4 grasses were positively correlated with precipitation amount indicating an influence of precipitation on the end-member δ13C values of C4 grasses. However, the observed sensitivity for this dependence was low: 0.07 ± 0.03 ‰ per 100 mm of precipitation. The low sensitivity implies that minor fluctuations in precipitation amount may not significantly impact the δ13C values of C4 grasses. In such cases, the use of uncorrected (i.e. modern) δ13C values of C4 grass end-member may not introduce significant error in the vegetation reconstruction studies from western peninsular India.