Regional and national-scale analysis of cropland carbon cycling

Abstract

Increased global greenhouse gas (GHG) emissions, including carbon dioxide (CO2), are known to contribute to global warming. Previous research has found carbon (C) sequestration in agricultural soils as a potential way of mitigating atmospheric CO2 emissions. In the first part of this study, I evaluated the methods used by Annex 1 (developed) countries in inventorying the sources and sinks of agricultural soil GHG emissions. In the second part, I assessed cropland soil C balance and C storage, considering residue C inputs and CO2 output from soils, at regional and national scale. One of the main components in this study was estimating the crop residue C inputs, using available county-level yield and area data for major US crops during the period 1982-1997. Since the existing annual data reported by the National Agricultural Statistics Service (NASS) have a large number of gaps (missing data), I filled those gaps by using regression analyses with the data from the Census of Agriculture, and a suite of linear mixed effect models that incorporated county level environmental and economic variables. These comprehensive crop databases were then used to estimate residue C inputs and cropland Net Primary Production (NPP). Interannual and spatial variability of residue C inputs were analyzed in relation to changes in production, weather and climate. I also evaluated the potential use of Advanced Very High Resolution Radiometer (AVHRR) Normalized Difference Vegetation Index (NDVI) data in estimating crop aboveground biomass and residue C inputs. Finally I estimated the soil C stocks and annual stock changes in cropland soils (and CO2 loss due to decomposition), by using the Introductory C Balance Model (ICBM), and evaluated the C storage in soil and overall C balance over the US cropland. During the 16-year period, I found an average cropland NPP and residue C input rate of 504 and 312 Tg C yr-1, respectively. The interannual variability of C stocks incurred within a much narrower range (within 20 Tg) compared to the variation of residue C inputs (2 to 80 Tg), indicating the scale of contribution from cropland productivity towards soil C sequestration and overall short-term C dynamics.