Constraining uncertainties in long-term particulate organic matter exports by headwater streams

Printer-friendly version

Headwaters are considered important components of the global carbon cycle. However, basic questions about the carbon that flows associated to particulate organic matter (POM) in these ecosystems remains largely unanswered. Since headwaters are highly sensitive to climatic and land-use perturbations, long-term POM fluxes analysis are required to better understand the role of these ecosystems in the C cycle. Our overall objective is to calculate annual POM exports in headwater streams that have undergone perturbations associated with climate and forestry practices in the last 30 years. We analyzed OM on suspended sediments collected on glass fiber filters, at H. J. Andrews watersheds 9 (Control) and 10 (logged in 1975) since 1980. Our POM analyses included elemental (C, N) and isotopic composition (¹³C, ¹⁵N). However, due to elemental, ¹³C, and ¹⁵N detectability thresholds, only ~ 30% of the available filter samples could be analyzed. This non-ignorable source of missing values contributes with an amount of uncertainty that cannot be estimated directly from the data (e.g. as variance). Here, we propose an approach to correlate annual hydrologic uncertainties, measured as the information entropy of the daily discharge, with the uncertainty of the estimates of annual POM exports. Our preliminary results support the hypothesis that hydrologic uncertainties could help to explain the number of filter samples that meet our analytical thresholds. Further analysis will consider the relationship between hydrologic uncertainties and the estimates of sediment export and the chemical properties of POM. We expect that by constraining the contribution of those uncertainty sources to our annual POM estimates, we can better analyze the long-term impacts of climate and land use changes on the C exported from headwaters.