UC Berkeley

Carbon dioxide (CO2) emissions, the primary driver of climate change, are increasing and may soon pass a threshold of 1.5 ºC warming that is effectively irreversible on ecological time scales. While progress is being made towards emissions reduction, emissions reduction alone will not alleviate climate change without removal of CO2 from the atmosphere (deemed negative emissions). Soil carbon (C) sequestration in managed and natural systems has been proposed as an immediately deployable technology to contribute to negative emissions, but the degree to which soil C sequestration could directly mitigate climate change, the duration for which a sequestration practice remains effective, and the time it takes for added C to cycle through differently stabilized soil pools remains uncertain. This dissertation was motivated by three questions: 1) To what extent can soil C sequestration contribute to the mitigation of global warming? 2) For how long will a specific C sequestration technique be effective, given possible future climate change scenarios? 3) How might more frequent extreme weather events affect how quickly C cycles through soil pools? Chapter 1 translated the technical potential of soil C sequestration due to improved agricultural land management to the measurable impact on future climate. Using only existing technologies, improved agricultural land management could reduce future warming of global surface temperatures by up to 0.26 oC under a low emissions trajectory (RCP2.6), or up to 0.14 oC under a high emissions trajectory (RCP8.5) by the end of the century. Results highlighted the sensitivity of future warming to both the rate of anthropogenic greenhouse gas emissions and the duration of effective C sequestration rates. Chapter 2 used experimental data with the DayCent model to analyze the biogeochemical response of grasslands across a California precipitation gradient to a one-time compost amendment, and the influence of future climate change on this response. Compost amendments increased NPP and soil C stocks at a much greater rate than it increased nitrous oxide emissions when compared as CO2 equivalents. Net soil C sequestration peaked 22 ± 1 years after amendment, and soil C stocks remained higher in compost amended soil relative to control soil until 2100 in every climate scenario. Chapter 3 used natural abundance and radiocarbon isotope and C stock measurements from two time points, three physical fractions of soils, and at three depths to examine the relative dominance of old and young organic C in each fraction and depth. These data constrained the soilR model to numerically solve for the distribution of C ages and transit times. I found that between 1988 and 2018 (a period including three major hurricanes), soil C stocks increased in the occluded light fraction but not the free light or mineral associated fractions. Soil organic C traveled relatively quickly through all fractions and depths, with isotopic signals dominated by year to decades old C in every case. Comparison of soil transit times constrained by the same C data but modeled with and without hurricane inputs revealed a shift towards both younger mean C ages and faster transit times in the occluded and mineral associated C pools. The increased loss of older C with hurricane inputs supports the hypothesis that hurricane events may cause priming of mineral-associated soil C in this tropical forest.