Influence of Dual Monsoonal Precipitation on Vegetation in the Indian Subcontinent During the Late Quaternary: Clues from Modern Analogues

Basu, Sayak (2019) Influence of Dual Monsoonal Precipitation on Vegetation in the Indian Subcontinent During the Late Quaternary: Clues from Modern Analogues. PhD thesis, Indian Institute of Science Education and Research Kolkata.

PDF (PhD thesis of Sayak Basu (11RS036)) 11RS036.pdf - Submitted Version Restricted to Repository staff only Download (10MB)

Abstract

The agricultural output, industrial development and hydroelectric production across the Indian subcontinent are dependent on the precipitation during Indian summer monsoon (ISM, June-September) and winter monsoon (October-December). Understanding the correlation between forcing factors and monsoonal precipitation is a fundamental challenge because both natural and anthropogenic variables govern variability, intensity and duration of monsoon. The major aim of the thesis is to understand the roles of controlling factors on monsoonal precipitation and associated changes in the vegetation composition. The temporal scale of this study is extended from the present-day to the Late Quaternary. Isotopic compositions of precipitation, vegetation, carbonates and organic matter have been measured for the climatic reconstruction. The modern precipitation trend has been used as a premise to select the sites for the paleoclimatic reconstructions. Mann-Kendall test statistic of the precipitation dataset for the years 1901 to 2016 exhibits a weak yet significant declining trend of ISM precipitation (τ = 0.3, p < 0.05) in central part of the Indian subcontinent (studied region: 20-30 °N, 80-95 °E). Backward moisture trajectory analysis shows that precipitation in this region is fed from adjacent Bay of Bengal (BoB). A weakening of the land-sea thermal gradient (at troposphere) and subsequent decrease of ISM precipitation are plausibly related to the monotonic warming over BoB, decreased convection over continental landmass and El-Nino Southern Oscillation (ENSO). It is important to account that Arabian Sea (AS) is the dominant moisture contributor of ISM precipitation in western India. The distinct meteorological settings in central part of the Indian subcontinent and western India intrigues to compare the long-term (> 50 ka) contemporaneous hydrological changes in these regions but such records are yet to be available from western India. Carbonates at different stratigraphic levels of cliff sections from western India provide an opportunity to resolve this issue. Gridded dataset also shows that the trend of Northeast monsoon (NEM) precipitation during winter in southern peninsular India (studied region: 8-16 °N, 75-80 °E) is in anti-phase with ISM precipitation in central part of the Indian subcontinent. Commencement and causal factor(s) of this inverse relationship remains elusive due to the absence of precipitation records from southern peninsular India. With the advent of analytical techniques, hydrogen isotopic measurements of the lacustrine sediments at the molecular level could help to understand the paleoclimatic conditions in southern peninsular India. Comparison of this time-series with the published records from ISM domain would be useful to investigate the inverse relationship between ISM and NEM precipitation. Dataset collected by the International Atomic Energy Agency from tropical island stations showed that the average rate of change in oxygen isotopic composition of precipitation (δ¹⁸Oprec) value is 1.5 ‰ for every 100 mm increase in the monthly precipitation and this negative correlation (i.e., amount effect) commonly used as a framework to interpret the past hydroclimatic fluctuations. However, several records from different parts of India found weak or absence of amount effect and warranted for other explanatory factors. Therefore, interpretation of isotope-based climate records requires regional dataset as the precise mechanism responsible for the variability of isotopic composition of precipitation (δprec) is not consistent across the Indian subcontinent. In this study, δprec values were measured from three locations; (a) Falakata (foothills of eastern Himalaya), (b) Vadodara (western India) and (c) Chennai (southern India) and subsequently used as analogues to interpret isotopic values of geological archives. Sparse investigations from eastern Himalaya restrict the interpretation of isotopebased proxy records from this high precipitation region (annual precipitation > 1500 mm). In this thesis, the role of climate variables during moisture transport from BoB to foothills of eastern Himalaya has been assessed through an isotope-equipped box-model coupled to climate data and the model observations have been compared with measured δ¹⁸Oprec values from Falakata (26.5 °N, 88.9 °E). The daily δ¹⁸Oprec values at the studied site are weakly related to local precipitation amount and instead controlled primarily by 5-days integrated upstream rainout and regional convective activity. Prominent phase-lag between the timing of maximum rainout fraction and precipitation amount diminishes the amount effect in the foothills of eastern Himalaya. A systematic trend of the residuals between measured δ¹⁸Oprec values and box-model observations is observed; positive residuals occur during pre-ISM (April-May) or early ISM (June) while negative residuals are more frequent during late ISM (September). Comparison of the measured δ¹⁸Oprec values with an isotope-enabled general circulation model (IsoGSM2) simulations also show a similar trend in the residual values. The positive residuals are attributed to secondary evaporation of falling raindrops and IsoGSM2 model possibly underestimated the extent of evaporation in foothills of eastern Himalaya. The negative residuals between measured and model observations reflect moisture transport time increases (> 5 days) during late ISM due to a southward shift of the monsoonal disturbances over BoB. The above-mentioned observations indicate that future isotope-based climate records from eastern Himalaya should be interpreted for rainout, instead of amount effect. The monthly amount-weighted δ¹⁸Oprec and δDprec values at Chennai vary from 1.5 ‰ to 7.4 ‰ and 12.0 ‰ to 52.0 ‰, respectively. The seasonal variations of δprec values in southern peninsular India are linked to the seasonal shifts of the moisture source. More positive δDprec values ( 19 9.7 ‰) are associated with precipitation sourced from AS during ISM while more negative δDprec values ( 37.4 14.7 ‰) are related to precipitation derived from BoB during NEM. Therefore, more negative isotopic excursions in geological archives in southern peninsular India would correspond to higher NEM precipitation and vice-versa. This isotopic dataset of modern precipitation helped to interpret the hydrogen isotopic composition (δD) of n-alkanes retrieved from the Lake Ennamangalam, southern peninsular India and has been used to reconstruct precipitation variability for the last 5000 cal yr BP. More negative δD values of n-alkane of the lake sediments after 3000 cal yr BP advocate for an increase in NEM precipitation over southern peninsular India. Intensification of NEM precipitation after 3000 cal yr BP is related to the southward migration of Intertropical Convergence Zone and warm phases of El-Nino Southern Oscillation. Comparison with the existing records suggests that the anti-correlation between NEM and ISM precipitation is mainly caused by the more frequent El-Nino events after 3000 cal yr BP. The monthly amount-weighted δ¹⁸Oprec values at Vadodara range from +3.0 ‰ to 5.2 ‰ from 2012 to 2013. The linear regression analysis between precipitation amount and δ¹⁸Oprec values exhibits a negative relationship with a correlation coefficient of 0.6, and thereby isotope-based proxy records from Gujarat alluvial plain could be interpreted via amount effect. Reconstruction of past hydrological conditions in western India is based on oxygen isotopic composition of carbonates (δ¹⁸Ocarbonate) collected from two cliff sections exposed on the flanks of Mahi (Rayka section) and Sabarmati (Mahudi section) rivers. The amount effect based monsoon reconstruction shows that two distinct dry phases at 75 ka to 60 ka and 30 ka to 10 ka were punctuated by a wet phase from 60 ka to 30 ka. These hydroclimatic variations in western India during the Late Quaternary shaped the regional fluvial architecture. Comparison with published climate records shows that fluctuations of the precipitation were more pronounced in western India compared to central India. The influence of precipitation variability on the vegetation composition has been assessed through the carbon isotopic (δ¹³C) measurements of modern plants, carbonates and organic matters. Based on the photosynthetic pathways, non-desertic plants can broadly be divided into two categories: C₃ (Calvin cycle) and C₄ (Hatch-Slack cycle). Most of the previous studies have used global mean δ¹³C values of C₃ plants (δ¹³Cc₃) at 26 ‰ or 27 ‰ to calculate the changes in the relative abundance of C3 plants during geological past. The accuracy of such vegetation construction depends on end-member δ¹³Cc₃ values which vary from 22 ‰ to 36 ‰ on a global scale. Understanding the causes of such variations in the δ¹³Cc₃ values is important before the vegetation reconstruction. For this purpose, δ¹³Cc₃ values have been measured from different parts of Indian subcontinent along a precipitation gradient of 300 mm to 1400 mm. The δ¹³Cc₃ values from India decrease 0.2‰ for every 100 mm increase in the precipitation amount. The newly measured δ¹³Cc₃ values from India have been combined with the published results from 723 locations across the world (n = 2621) to understand the global variations of δ¹³Cc₃ values and their causal factors. The carbon isotopic discrimination between air and leaf (Δair-leaf) have been calculated from all the data points; depending upon the regional δ¹³C value of atmospheric CO2. Prior to the meta-analysis of the compiled data set, individual results for a given location have been averaged to provide community-level Δair-leaf values. The average Δair-leaf values in equatorial or low-latitude regions of the Northern Hemisphere is ca. 2 to 3 ‰ higher compared to that in the areas of mid-latitude or Southern Hemisphere. Sampling biasness in the compiled Δair-leaf values lowered the global mean air-leaf value (19.3 1.9 ‰) as 46 % of the total data points in the present compilation have been taken from mid-latitude regions. On a global scale, Δair-leaf values exhibit a strong hyperbolic correlation (R = 0.69, p < 0.05) with mean annual precipitation (MAP). The positive relationship between MAP and Δair-leaf values is observed for different latitudinal bins though the slope of the correlation changed with regional precipitation amount. It is noticed that MAP exerts higher control on the Δair-leaf values in the dry ecosystems (< 500 mm) while MAP has little impact on the Δair-leaf values beyond the threshold value of 1500 mm. In this work, moisture-stress factors have been calculated for different latitudinal bins and subsequently have been used to understand the evolution of C₄ plant. For this purpose, available records of δ¹³C values of soil carbonate and tooth enamel have been compiled, and our analyses show an early appearance of C₄ plants in Siwalik Group sediments of the Indian subcontinent than previously reported timing. The carbon isotopic measurements of the n-alkane from the Lake Ennamangalam shows that the increased cold season NEM precipitation after 3000 cal yr BP favored the C₃ plants in southern peninsular India. The long-term vegetation reconstruction from western India using carbon isotopic ratio of carbonate (δ¹³Ccarbonate) illustrates fluctuations in the relative abundance of C₃ plants. Significant and strong positive correlation between δ¹⁸Ocarbonate and δ¹³Ccarbonate values indicates that changes in precipitation amount controlled the relative abundance of C₃-C₄ plants in western India during the Late Quaternary period.

Actions (login required)

View Item