CSU Thurgoona:Notebook/Dissolved Iorganic Carbon in headwaters/2009/10/14
|Project name||Main project page|
Previous entry Next entry
This is the DIC Page!
Literature Review Rivers represent the major pathway for lateral transport of carbon (Particulate/Dissolved Organic Carbon POC/DOC and Particulate/Dissolved Inorganic Carbon PIC/DIC) from terrestrial environments to the ocean, with approximately 1 Gt of carbon transported annually (Gaillardet et al., 1999; Amiotte-Suchet et al., 2003; Aumont et al., 2001; Cole et al., 2007). A further 0.8 Gt enters fluvial systems but is later degassed back to the atmosphere during transport (Cole et al., 2007). DIC is derived from weathering of silicate and carbonate rocks (with CO2 consumption rates highest in steep wet tropical regions) and from terrestrially-sourced respiration transferred from land to rivers by runoff, soil and groundwater discharge. Despite the close linkage between the terrestrial (plant/soil) and riverine carbon cycles, few studies have attempted to link the processes driving terrestrial and riverine fluxes (e.g Billett et al., 2004). In addition, the fact that weathering also removes CO2 from the atmosphere is usually neglected in carbon cycle studies driven by the biological science community. The importance of these lateral fluxes to the global carbon cycle has been recently underscored by studies such as Mayorga et al. (2005) who demonstrate that substantial downstream efflux of vegetation-derived ‘young’ carbon from the Amazon River, while Raymond and Cole (2003) have demonstrated significant long-term anthropogenic changes in alkalinity export by the Mississippi River. In part, the disjunction between ‘terrestrial’ and ‘riverine’ carbon cycle studies occurs because, while high temporal resolution techniques such as eddy covariance have enabled the development of long high resolution records of carbon cycle dynamics in the terrestrial environment, no comparable techniques for monitoring carbon fluxes are currently readily available for deployment in riverine environments (although many other parameters of importance in fluvial systems can be logged in situ). Jones and Mulholland (1998) laboriously compiled a five year record of terrestrial and riverine carbon fluxes at weekly resolution, and correlation between the terrestrial and riverine fluxes led them to speculate that riverine DIC flux represents an integrative measure of whole catchment soil respiration, and hence is at least a relative measure of changes in net ecostystem production at the catchment scale. Hope et al. (2004) also concluded from a fortnightly sampling program over two years that river pCO2 was a useful measure of whole catchment respiration, if soil-stream leakage was strong. Likewise, Kardjilov et al. (2006), using monthly sampling, found that DIC flux was correlated with net ecosystem exchange in small streams in Iceland. Billett et al. (2004) have examined the potential significance of lateral fluxes to estimenting terrestrial net ecosystem exchange, showing that lateral losses of carbon to streams from a lowland peat system were equal to, or greater than, net annual carbon uptake by photosynthesis/respiration at the land surface. All the above studies we based on low-resolution (weekly or more) samplings. Only one study has directly examined diurnal variations in pCO2 by direct in situ measurement (over a single day) and found diurnal variations (Parker et al., 2007), while other studies (e.g. Jarvie et al., 2001) have inferred diurnal variability from logged pH records. While the existence of such variations, resulting from diurnal variations in in-stream photosynthesis/respiration and possibly also in soil respiration, is not surprising (eg Dawson et al., 2001), it follows that no study based on daily or weekly sampling during the day, could accurately measure DIC flux (Jarvie et a., 2001). Even longer timescales, Douglas (2006) has demonstrated that weathering fluxes of ions and CO2 from small streams vary seasonally, and therefore short-term or low-resolution sampling cannot adequately characterize annual fluxes. While pCO2 may be the variable of interest for studies of in-river biology/chemistry, the parameter of interest in many, if not most, studies of landscape-scale biological processes and geochemistry is DIC. Commercial instruments are available that can log in situ measurements of pCO2, generally for marine deployment (SAMI-CO2, oxy-logger, PSI-CO2 pro) but there is no instrument that can log in situ measurements of DIC. DIC is currently inferred from alkalinity measurements or determined by acidification in the laboratory, with quantification of the partial pressure of CO2 liberated into a headspace (e.g. St-Jean, 2003; Doctor et al., 2007), spectrophotometrically (Twe-Nguen et al., 2005), or by ion chromatography (Polesello et al., 2006). In all cases, the sampling is manual with measurement usually undertaken in a laboratory environment, greatly limiting the number of samples collected and the length and temporal resolution of the time series acquired. Attempts at producing higher resolution records of pCO2/DIC in rivers have generally relied on attempting to calculate pCO2/DIC from logged records of temperature, pH and conductivity, calibrated via spot measurements of alkalinity (e.g. Neal et al., 1998), with attendant additional uncertainty. It therefore seems likely that the development of an instrument capable of in situ of rapid unattended in situ direct measurement and logging of pCO2 and DIC would assist in catalysing a better and more integrated understanding of both the terrestrial and fluvial carbon cycles and the links between them. To enable the development of a new generation of long, high resolution records of inorganic carbon in terrestrial waters, an optimal instrument should be capable of unattended, sub-hourly, field measurement and logging of (i) DIC, (ii) ambient pCO2 and (iii) CO2 flux across the water-air interface over at least yearly periods. This proposal seeks to develop and deploy such an instrument. The key to the development of the instrument is a system for in-line acidification of sequential aliquots of water, with DIC converted to CO2 which rapidly diffuses into a headspace through proprietary expanded Teflon (ePTFE) tubing that is not permeable to liquid water. The partial pressure of CO2 in the headspace is measured by a field-portable infra-red gas analyser (IRGA) and is proportional to the DIC concentration in the original water. ePTFE tubing has been used for DIC determination in a flow-injection system with photometric detection (e.g. Motomizu et al., 1987) but, although this demonstrates the feasibility of the gas diffusion approach the analytical methodology is not amenable to field deployment. While in situ measurements of pCO2 of river water (e.g Johnson et al., 2006) and lake water (Uusimaa et al., 2006) have been made using gas permeable (generally silicone) tubing, this approach measures only a small component of DIC unless the water is very acidic. DIC is made up of several carbonate species and the relative proportion of each is highly pH dependent. Hence inference of DIC from pCO2 is problematic as the proportion of DIC that is present as CO2(aq) decreases rapidly as pH increases. For example, many rivers have a pH value between 6 and 8, and over this pH range, the proportion of DIC that is present as CO2(aq) changes by two orders of magnitude. As a result, the calculation of DIC from pCO2 (assuming equilibrium) rapidly becomes very imprecise due to the errors associated with the measurement of pH and extrapolation of a (large) calculated DIC concentration from a (small) measured pCO2.