Description
TitleHydrological controls on ecosystem structure and function
Date Created2023
Other Date2023-05 (degree)
Extent135 pages : illustrations
DescriptionUnderstanding the patterns of plant distribution is a long-standing question in Ecology. Since the early explorers set foot on the New World, accounts of lush and exotic vegetation occurring in the warm, humid equatorial areas, so different from the conifer and oak/beech forests cold northern latitudes, have paved the way for our understanding of climatic controls on vegetation distribution. While climate is the ultimate source of water, the failure of precipitation in explaining the high local variability of plant structure and function points to the importance of considering local drivers. At the local scale, climate is blurred by topographic relief, and lateral drainage from hills provides valleys with a more stable supply of water in the form of groundwater. In this dissertation, I focus on describing the mechanisms by which groundwater controls vegetation structure and function based on remote sensing, modeling, and field data. First, I look at forest-savanna mosaics in the Tropical South America using a hydrological framework of two stresses: drought stress, when the water table is deep and plants are exposed to climatic droughts; and waterlogging stress, when the water table is shallow and creates anoxia in the rooting zone. I show that forests can adapt to both types of stresses, but the combination of both (“double stress”) creates a physiological challenge to which trees cannot survive. In areas of double stress, savanna is the dominant vegetation type, with low tree cover and predominance of herbaceous vegetation, which can enter dormancy and wait for optimal hydrological conditions to resprout. Second, I look at the benefit of having groundwater as a source. Using a high- resolution hydrological model in Amazônia, I map the amount of extra water supply plants in river valleys receive by groundwater convergence from surrounding hills. By integrating this basic hydrologic concept with a forest resilience framework, in which forests are more likely to exist in more humid conditions (i.e., higher resilience), I show that forest resilience across Amazônia is highly heterogeneous, following topographic patterns. Forests in river valleys are buffered from strong climatic droughts such as the 2005 and 2010 droughts, with little to no change in groundwater supply, and thus have higher resilience than their counterparts on hills. Third, I compile fieldwork data of vegetation resistance to drought in three locations spanning a rainfall-vegetation gradient, from the humid Amazônia rainforest to the dry Cerrado savannas. By comparing vegetation communities on hills and valleys in each location, I show that vegetation in valleys – which can access groundwater – is consistently more vulnerable than on hills. The former is “spoiled”, having access to water year-around even in the five-month dry season of the Cerrado, while the latter has to withstand long periods of no rain and low atmospheric humidity. I show that the hydrologic gradient consistently shapes the functional composition of vegetation communities in regard to resistance to embolism, modulated by rainfall amounts and seasonality, affecting both the overall composition as well as diversity. The work in this dissertation shows that the hydrologic gradient is a consistent, predictable environmental driver acting as an architect of vegetation and ecosystems. The redistribution of rainfall water across the terrain and the temporal dynamics of groundwater establishes the hydrological niche on landscapes, from dry hills to wet valleys. This in turn is expressed in a spatial turnover of vegetation type, composition, and function. Accounting for the systematic patterns of groundwater-vegetation influences in Earth System Models, which currently have little to no representation of groundwater dynamics and feedbacks, is paramount for the correct representation of ecosystems and thus more accurate predictions of climate change impacts on them.
NotePh.D.
NoteIncludes bibliographical references
Genretheses
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Statistical Profile