Academic Commons

Theses Doctoral

Crops, Canopies and Waiting for Rain Water for Small-Plot Agricultural Production in the Tropics

Clark, Benjamin D

Water will become increasingly scarce in the 21st century. Agriculture dominates anthropogenic water use and accounts for about 70% of water withdrawals globally. Unique challenges face tropical small-plot agricultural water management that differs from region to region. This dissertation examines two challenges facing tropical small-plot agriculture. Chapter 2 uses an experimental trial in Western Tanzania to create a unique longitudinal dataset of crop water stress measured over the growing season. The trial tests the effect of seed variety and fertilizer treatment on crop water stress over the growing season and during dry spells. Results demonstrate that hybrid varieties yield significantly more than the locally adapted traditional variety because they are better able to access nutrients and have better stomatal regulation over dry spells. Chapters 3 and 4 shift the focus to India. Chapter 3 characterizes the inter-annual dynamics of anthropogenic water stress across the Central Indian Highlands (CIH), while Chapter 4 examines the hydrological impacts of increasing forest cover on regional water supply and its implications for sustainable irrigation as well as food production. Within Chapter 3, I use extensive data sourced from the Indian government to spatially characterize water demand over the past decade by spatially mapping multiple waves of the Minor Irrigation Scheme Census and Livestock Census collected at the household level, along with monthly power generation datasets. The patio-temporal water demand data is coupled with remotely sensed precipitation and evapotranspiration data to force a customized Sacramento Soil Moisture Accounting Model that computes water supply. Finally, I developed a Groundwater Supply Stress Index to account for the impact of irrigation groundwater withdrawals over the course of the year. Chapter 3 finds that 70% of CIH is water-stressed during some portion of the year and that irrigation makes up approximately 95% of anthropogenic water withdrawals. Chapter 4 extends the findings of chapter 3 in utilizing the infiltration-evapotranspiration trade-off hypothesis to understand the impact of converting croplands to forest on groundwater recharge within the CIH. In this Chapter, I collected and analyzed field data on field-saturated hydraulic conductivity to find that forested land has significantly higher infiltration rates than croplands. These finding are then included in a Spatial Processes in Hydrology model to simulate intra-annual hydrological dynamics of current forest cover versus a forest cover increased to 30% within the river basins of the CIH. Increased forest cover is one of India’s Nationally Determined Commitments at COP21 within the Mission to Green India with a stated aim of improving landscape hydrological functioning. I demonstrate that forest cover increase has the potential to increase groundwater recharge, which could be used to irrigate a second growing season and help offset the loss of cropland through conversion to forest. Collectively, these three chapters harness multiple sources of data and leverage a wide array of innovative methods at multiple scales to shed light on important water management issues faced by small-plot agriculture in the tropics and on opportunities for better agricultural water resource management across two continents.

Geographic Areas


  • thumnail for Clark_columbia_0054D_15079.pdf Clark_columbia_0054D_15079.pdf application/pdf 7.24 MB Download File

More About This Work

Academic Units
Ecology, Evolution, and Environmental Biology
Thesis Advisors
DeFries, Ruth S.
Ph.D., Columbia University
Published Here
February 7, 2019