Pedogenesis, soil fertility, and indigenous adaptation: The formation of Hawaiian volcanic soils across climate, time, and parent material

Abstract

Soils exist at the interface of the critical zone, hosting many of the chemical reactions that regulate the Earth’s biogeochemical cycles. The formation of soils, or pedogenesis, is a key process in understanding soil functioning and is controlled by environmental drivers (climate, parent material, topography, and biota) over time. The Hawaiian Islands are considered a “model system” for ecological and pedological studies because they contain broad gradients in substrate age and climate while simultaneously having relatively consistent parent material and biotic components. Many pedological studies have utilized Hawai‘i as a study system to see how soils evolve across climate and time, and have observed that soil properties, including soil nutrients, behave in response to a series of pedogenic thresholds and soil process domains. Pedogenic thresholds and soil process domains have been observed and quantified in Hawaiian soils along three different-aged climosequences (20, 150, and 4,100 ky). However, the age component in understanding how pedogenic thresholds and soil process domains occur over time is limited. A broadened chronosequence of climosequences would support a better understanding of the evolution of pedogenic thresholds in Hawaiian soils and could potentially allow for modeling pedogenic thresholds across climate and time. Modeling the development and evolution of pedogenic thresholds over time could have implications for understanding how soil nutrients accumulate and deplete in Hawaiian soils, which could support soil fertility spatial models. But this type of modeling assumes that the parent material component in Hawaiian soil formation is consistent, which it may not be.Through this dissertation, I sought to understand the connections between soil fertility, pedogenesis, and the adaptation of traditional Hawaiian agroecology by modeling the evolution of pedogenic thresholds and soil process domains of Hawaiian volcanic soils across a broadened chronosequence of climosequence and implementing this work into generating geospatial models for traditional Hawaiian rainfed agroecological systems. I then questioned the pedogenic consistency of Hawai‘i’s parent material by examining the influence of parent material geochemistry on soil heterogeneity along a lithosequence. In Chapter 2, I sampled and analyzed hundreds of soils across the Hawaiian archipelago to construct a broadened chronosequence of five climosequences (5, 15, 120, 450, and 1,400 ky), which I then used to observe the evolution of pedogenic thresholds and soil process domains across time. I was able to quantify a pedogenic threshold along each substrate climosequence that represents the exhaustion of primary mineral inputs and found that this threshold shifts to drier climates as soil substrates get older. I then modeled the occurrence of this pedogenic threshold with an empirical equation, which predicts the exhaustion of primary mineral inputs as a function of climate and substrate age. In Chapter 3, I used this empirical soil threshold equation to predict soil fertility and generate geospatial models for traditional Hawaiian rainfed agroecology. These models predicted the extent of Intensive Rainfed, Marginal Rainfed, and Agroforestry agriculture in ancient Hawai‘i, which were ancient rainfed agroecological systems characterized by unique crops and cultivation methods. I also validated these empirically-constructed models with archaeological remnants and historical accounts, which suggested that my models were more accurate at predicting dryland rainfed agriculture but were less accurate at predicting agroforestry than previously constructed rainfed geospatial models. In Chapter 4, I looked at soil heterogeneity across a geochemical lithosequence of Hawaiian lava flows in order to address the consistency of Hawai‘i’s parent material. Across the lithosequence, soil geochemical composition and nutrient concentrations were significantly different. Through correlation analysis, I also observed that soil geochemical concentrations were inherited from the parent material, suggesting that Hawai‘i’s parent material is not consistent from a pedogenic perspective and needs to be taken into account in future studies.