The Response of Mineral-Associated Organic Matter to Drought Conditions And The Signatures of Land Use in Sediment Loading in Central Oklahoma

Abstract

Soil organic carbon (SOC) acts as a significant terrestrial carbon sink, and its turnover is a factor of the different pools of SOC in a soil. Mineral-associated organic matter (MAOM) is considered to be a particularly stable SOC pool. MAOM is protected from microbial respiration primarily through strong chemical bonds between the OC and mineral, promoting soil carbon storage. Despite this inherent stability, MAOM turnover has been found to depend upon climatic conditions and the physiochemical properties of a given soil. Thus, there is a need to evaluate MAOM stability across different climates while simultaneously considering the mineralogical controls on MAOM storage. In Central Oklahoma, changes in the amount and frequency of rainfall are anticipated due to climate change. To assess the impact of a changing climate on SOC, we characterized total SOC and MAOM-C in soil samples from a field-manipulated rainfall gradient in an existing temperate prairie. We hypothesized that there would be alterations to total SOC and MAOM-C concentrations in response to a simulated drought experiment. Soils were analyzed following the 4-year DroughtNet experiment at the Kessler Atmospheric and Ecological Field Station in Central Oklahoma. Here, shelters were constructed so that soil plots received a controlled amount of precipitation ranging from 50% addition to 100% exclusion of the natural precipitation. Changes in C concentrations were evaluated following a series of extractions that separate MAOM-C based off the crystallinity of the iron or aluminum oxide to which it was bound. We further hypothesized that the response to precipitation treatments would be evident within these pools of increasingly stable metal species with more crystalline species increasingly holding C in drought conditions. However, no statistically significant relationships were identified between total SOC or MAOM-C pools and precipitation. Still, the most abundant pools of Fe and Al species were identified as well as to what species most of the MAOM-C was likely bound. We found that soil Fe was mostly in the form of crystalline iron oxides. However, the extractable MAOM-C was preferentially bound to less crystalline Fe species such as short-range order (SRO) Fe-(hydr)oxides or Fe found in the form of organo-metallic associations. Al species were more abundant than Fe species in these less crystalline states. Interestingly, much of the MAOM-C remained following the removal of these reducible metal-oxide species. Using X-ray diffraction, we confirmed the presence mixed layered 2:1 phyllosilicate clays that likely hold the remainder of the C in the < 53 µm MAOM fraction. The preferential adsorption of C to less crystalline, more reactive metal species and clay could result in MAOM vulnerability over a greater timescale. However, the whole soil SOC as well as the MAOM-C fractions appear to be stable across the precipitation gradient and are governed by soil texture rather than moisture. Excessive loading of sediment into waterbodies has been shown to severely diminish water quality. In Norman, Oklahoma, this issue impacts a source of drinking water for the residents, Lake Thunderbird. Here, we evaluated potential sources of soil erosion that could cause high suspended sediment within this watershed. Specifically, we conducted geochemical analysis on suspended sediment collected using a passive sediment trap deployed in Dave Blue Creek, a stream which feeds into Lake Thunderbird. We compared the suspended sediment to soil in surrounding areas that were representative of potential source groups. These source areas were ultimately grouped by land use: grasslands, tree cover, and developed land as classified by the National Land Cover Database (NLCD). We hypothesized that most of the suspended sediment in Dave Blue Creek would originate from upland urban soils. Using the Sediment Source Assessment Tool (Sed_SAT), we obtained unique geochemical fingerprints for each source group and were able to characterize changes in sediment input contributions. We successfully identified contributing sources to the suspended sediment in Dave Blue Creek based off land use. However, we found that we were unable to differentiate between bank and upland sources of the same land use. Still, clear contributions from developed land were identified across all suspend sediment samples. Additionally, we found that during the sampling periods with more precipitation and higher discharge in the stream, there was a strong grassland signature. Only in the month will little precipitation and low discharge were tree-covered land inputs detected. Overall, our results indicated that erosion control strategies in the Lake Thunderbird watershed should focus on ensuring bank stability in developed areas and mitigating run-off in grasslands.