« Previous | Contents | Next »
Listen
Module 4 The effects of land-use change and climate change on the release of Dissolved Organic Matter from organic soils
4.1 Introduction
4.1.1. Background
Dissolved organic carbon ( DOC) is the carbon contained within organic matter in solution, and represents a potentially important pathway for carbon loss from organic soils. Previous studies have shown that riverine DOC fluxes are closely related to the extent of peat within the catchment (e.g. Hope et al., 1997). There is growing evidence that DOC concentrations in UK rivers have increased in recent decades (Freeman et al., 2001a; Worrall et al., 2004; Evans et al., 2005), and that rates of increase are greatest in the most organic-rich catchments (Freeman et al., 2001a). The recent paper by Bellamy et al. (2005) suggests that these same organic soils have experienced a loss of carbon over the same period. This desk study provides a review of existing literature, supported by some unpublished CEH data, to assess 1) the significance and sources of the DOC flux term within the overall C budget of organic soils; 2) the external environmental drivers potentially responsible for observed increases in DOC; and 3) the extent to which land management may accelerate or ameliorate rates of DOC loss.
4.1.2. The importance of DOC fluxes in the C budget
Many studies of the carbon balance of organic soils focus on the exchange of C between the atmosphere and the land surface. Net ecosystem CO 2 exchange represents the balance between two large gross fluxes (primary productivity and ecosystem respiration), and may be positive, negative or zero. Riverine DOC fluxes are typically one to two orders of magnitude smaller than gross land-atmosphere C fluxes (Hope et al., 1997), and have therefore been neglected in some studies of ecosystem carbon balance. However, DOC fluxes are unidirectional (i.e. always represent a C loss), and comparisons between riverine DOC export and net ecosystem CO 2 exchange, both globally (Hope et al., 1994; Schlesinger, 1997) and in the UK (Cannell et al., 1999) suggest that they are actually of similar magnitude. A peatland catchment carbon flux study by Worrall et al. (2003a) recorded a DOC flux at Moor House, Northern England, of 94 g C/ha/yr, compared to a (literature-based) estimated net CO 2 exchange of 400-700 g C/m 2/yr. For Auchencorth Moss in Scotland, Billett et al. (2004) measured a DOC flux of 283 g C/m 2/yr, and a net CO 2 exchange of 278 g C/m 2/yr. At Auchencorth Moss, the inclusion of the DOC (and other riverine) loss terms was sufficient to turn the catchment from an apparent C sink to a substantial (83 g C/m 2/yr) C source. In general, it is clear that the omission of DOC from studies of carbon in organic soil ecosystems will lead, at best, to substantial errors in the estimate, and at worst to the misattribution of systems that are either stable, or losing carbon, as carbon sinks.
4.1.3. Sources of DOC
DOC may derive from plant litter, root exudates, microbial biomass or soil organic matter. Identifying the origin of (rising) DOC losses from organic soils is important in determining whether they are indicative of the destabilisation of 'old' soil carbon (as suggested, for example, by the results of Bellamy et al., 2005), or of an accelerated throughput of 'new' carbon which has been assimilated relatively recently from the atmosphere. Measurements of radiocarbon, 14C, provide a means to estimate DOC age. Previous studies show mixed results; some suggest a predominantly recent origin for DOC (Palmer et al., 2001; Benner et al., 2004), others a substantial older component (Raymond and Bauer, 2001), or a shift from older to newer material with increasing flow (Schiff et al., 1997). In a study linked to the current project, DO 14C was measured in streams and soil solutions draining a range of soil and land-use types in North Wales. Results suggest that most of the DOC exported from organic soils is 'new' (post-1950s), especially at high flow, when the majority of DOC flux export occurs (Figure 4.1). These data are difficult to reconcile with possible large-scale peatland destabilisation (e.g. Freeman et al., 2001a,b; Bellamy et al., 2005), at least if riverine DOC represents a significant loss pathway for soil C. Instead, they suggest that rising DOC in surface waters must be explained by either 1) rising production of 'new' DOC from recent plant material, or 2) decreasing retention of this DOC within the soil profile. It is worth noting that 14C data for mineral (upland and improved grassland) soils do suggest an older soil origin, and may therefore be more consistent with suggested long-term soil organic matter depletion from these systems.
Figure 4.1 14C Isotopic composition and concentration of stream DOC at a) summer low flow, and b) autumn high flow for five major soil/land-use types in North Wales. Error bars show range of observed values. Samples in upper shaded area have 14C composition indicative of a probable post-1950s source, lower shaded area indicates pre-1950s source. From Evans et al. (in review).
4.1.4. Causes of rising DOC concentrations
Figure 4.2. DOC concentrations at 10 lakes in the UK Acid Waters Monitoring Network. Circles and bars represent the median and interquartile range of DOC concentrations across all sites at each sampling interval. (From Evans et al., 2006)
DOC concentrations have been rising in the 22 lakes and streams of the UK Acid Waters Monitoring Network, including all sites in Scotland and Wales, since 1988 (Freeman et al., 2001; Evans et al., 2005; Figure 4.2). These increases have been replicated across a wider range of UK surface waters (Worrall et al., 2004), in some water colour records going back to the 1970s (Worrall et al., 2003b); and across large areas of Scandinavia and Northeast North America (Skjelkvåle et al., 2005). The geographical extent of observed trends, from heathland, grassland and plantation forest catchments in the UK, to unmanaged mixed forests in North America, argues against local management factors as the fundamental cause of these increases, and this section therefore focuses on large scale environmental drivers that might account for observed changes. However, there is good evidence that management and land-use can have a significant impact on DOC export, either contributing to observed increase in some areas, or providing the means (through management or land-use change) to ameliorate changes; these issues are discussed in the following section.
Possible drivers of DOC change operating at a large scale include climate (e.g. temperature, rainfall, drought), changes in atmospheric chemistry (CO 2), and deposition (of sulphur and nitrogen compounds). Each of these mechanisms has been reviewed in Evans et al. (2005), a paper supported by the project. The following summarises and updates the published review:
Decreasing acid deposition: Sulphur (S) deposition has more than halved across the UK since the mid-1980s, triggering widespread increases in soil solution and runoff pH, and decreases in ionic strength ( IS). Low pH and high IS reduce the solubility of weakly acidic humic substances (Tipping and Hurley, 1988; Kalbitz et al., 2000). Although buffering of mineral acid changes by organic acids in peaty soils was proposed as early as the 1980s (Krug and Frink, 1983), it has only relatively recently been considered as a likely driver of rising DOC (e.g. Stoddard et al., 2003; Evans et al., 2005). Experimental studies appear to confirm that soil solution DOC is highly sensitive to changes in acidity and IS (Mulder et al., 2001; Clark et al., 2005). The DOC model of Lumsdon et al. (2005), and the Dy DOC model (Michalzik et al., 2003) both incorporate pH effects on DOC solubility. In a second paper supported by the project (Evans et al., 2006), decreasing S deposition was argued to be the dominant cause of rising DOC, in the UK and elsewhere.
Rising temperature: Due to the temperature-dependence of organic matter decomposition rates, when DOC increases were first noted in UK waters it was suggested that this could be a response to long-term warming (Freeman et al., 2001a). A large number of studies have observed an increase in DOC production with rising temperature (see e.g. Kalbitz et al., 2000; Evans et al., 2005, and references therein). In addition, seasonal DOC variations are closely related to temperature (e.g. Scott et al., 1998; Clark et al., 2005; Billett et al., 2006), and the DOC model developed by Lumsdon et al. (2005) required a temperature-dependent term describing biological activity, in addition to geochemical controls, to reproduce seasonal variations in soil solution DOC. However, the near-doubling of UK surface water annual average DOC concentrations since the 1980s, for an approximate 0.6 °C average warming over the same period, would require a Q 10 for DOC production (the amount by which the rate or a process changes over a 10°C change in temperature - a doubling of a rate over 10 °C would be a Q 10 of 2) well in excess of those recorded in peat warming experiments (1.33 to 2.13 under anaerobic conditions; Freeman et al., 2004; Evans et al., 2006). In addition, both decomposition rates, and the partitioning of decomposition products between CO 2 and DOC, are sensitive to factors other than temperature. In peatlands especially, the maintenance of anaerobic conditions due to waterlogging may restrict decomposition (e.g. Freeman et al., 2001b; Davidson and Janssens, 2006). In aerated peat, decomposition rates are higher and temperature sensitivity increases (Q 10 for DOC production 3.66, Evans et al., 2006). Therefore a combination of warming and drying may have made a significant (but probably partial) contribution to observed DOC increases (e.g. Worrall et al., 2004; Evans et al., 2006).
Figure 4.3. Increasing DOC under all flow conditions at Dargall Lane AWMN site (from Evans et al., 2006)
Hydrological changes: Several possible mechanisms exist by which hydrological factors may affect DOC export. These include shifts in water flowpath from (low- DOC) mineral horizons into (high- DOC) organic horizons during rain or snowmelt events (e.g. McDowell and Likens, 1988; Hongve et al., 2004), and increased DOC production rates under aerated conditions during droughts. The former is more important in organic-mineral soils, the latter in peats. The effects of drought in peats may be complicated by constraints on enzyme diffusion at low soil moisture levels (see e.g. Davidson and Janssens, 2006); more complete mineralisation of organic matter (i.e. to CO 2 rather than DOC; Scott et al., 1998) and concurrent chemical changes (e.g. oxidation of reduced sulphur compounds) which reduce DOC solubility in soil water (Clark et al., 2005). Although there is no question that hydrological factors have a major influence on DOC export in the short-term, there is less consensus as to whether long-term changes can explain DOC trends. Consistent UK-wide hydrological trends are difficult to identify, and there have certainly not been consistent hydrological changes across the wider European and North American area throughout which rising DOC trends have been recorded. Analysis of UK monitoring data suggest that DOC increases have occurred under all flow conditions rather than due to any long-term change in the dominance of low- or high-flows (Figure 4.3). Results from an ongoing CEH experiment on an organo-mineral soil in North Wales (Figure 4.4) provide clear evidence of the short-term impact of annual induced droughts, but evidence of a long-term response is equivocal.
Figure 4.4 The effect of experimental drought on heathland organic horizon DOC concentrations at the Clocaenog site, North Wales. DOC concentrations are reduced during experimental summer drought, and increased during winter ( CEH Bangor, unpublished data).
Productivity changes: As noted above, much of the DOC exported from organic soils appears to comprise carbon assimilated from the atmosphere relatively recently, rather than from decomposition of 'old' soil organic matter. Changes in ecosystem productivity have the potential to increase DOC export through accelerated production of labile litter or root exudates. Freeman et al. (2004) found that vegetated peat cores exposed to elevated atmospheric CO 2 underwent an increase in productivity (in part due to vegetation changes) and that the proportion of recently-assimilated carbon in the DOC pool increased greatly. The resulting increase in the bulk DOC pool, relative to an experimental doubling of atmospheric CO 2, appears insufficient to explain the full magnitude of surface water DOC increases (Evans et al., 2006), but may be a partial contributor to observed trends. In addition, productivity of most organic soil ecosystems is N-limited, and productivity increases due to elevated anthropogenic N deposition have been proposed as a cause of rising DOC (Findlay, 2005). While evidence for recent productivity changes on the required scale in UK upland ecosystems has not been identified, this is considered to be an area requiring further investigation. A review of existing N manipulation experiments was carried out, with inconsistent results: some experiments show a substantial DOC increase (e.g. Pregitzer et al., 2004) but many others show no change. Because experimental N addition can potentially change many factors other than productivity (most notably soil acidity), considerable caution is required in the interpretation of these results.
4.1.5. Influence of land use and management on DOC loss
At the scale of Northern Europe and Northeast North America across which DOC increases have been recorded, no single explanatory land-use change factor can be identified. Within the UK itself, DOC increases have been observed in waters draining grassland, heathland and plantation forest catchments, subjected to varying types and intensities of management (Evans et al., 2005). However, at a local and regional level, changes in land use and management have the potential to significantly alter carbon cycling in general, and DOC export in particular. This section therefore summarises both the extent to which past changes may have influenced DOC losses from managed upland systems, and the extent to which future management might be adapted in order to minimise DOC losses.
Drainage: Peatland drainage has been widely practiced in the UK since the 1940s, e.g. to improve grazing quality or to permit afforestation. By lowering water tables and providing a more intensive drainage network, this may both increase decomposition rates (due to increased aeration), and reduce the potential for DOC retention within the soil (e.g. Holden et al., 2004; Worrall et al., 2004). However as a mechanism to explain recent DOC increases this is unlikely to be important, because since the 1990s there has been an increasing tendency towards peatland restoration through drain blocking. A recent study by Wallage et al. (2006) confirms that DOC losses were higher from drained peats, and suggests that drain blocking will reduce long-term DOC losses to levels below those of undrained sites. Worrall et al. (in review) found that drain blocking actually led to a short-term (1 year) DOC flush, but a modelling study (Worrall et al., submitted) does suggest that, in the longer term, DOC losses will be reduced. Drain-blocking may therefore represent an effective amelioration strategy for reducing DOC loss in drained peatlands.
Burning: Heathland burning is practiced in many areas, both for habitat management, and (more intensively) for grouse shooting. For England, Yallop et al. (2006) found evidence of burning in around 70% of the heathland area, with a near-doubling in the extent of new burns since the 1970s. In the Pennines, a strong correlation has been shown between surface water DOC concentrations and extent of burning within the catchment (A. Yallop, pers. comm.), suggesting that increased burning could be a significant contributor to increased DOC loss. For a long-term burn manipulation at Moor House, however, Worrall et al. (in press) recorded reduced DOC concentrations in soil solution beneath burnt plots. Although the reduction or cessation of burning in meanaged heathlands might be expected to reduce catchment C losses, then, further work is required to confirm this. A paired-catchment experiment in the Peak District (O'Brien et al., 2005), in which burning, drainage and grazing have all been manipulated, may provide new information, but results are not yet available.
Afforestation: Establishment of plantation forestry on grazed or semi-natural organic soils could alter DOC export in several ways. Drainage will increase soil aeration, as decribed above, and enhanced transpiration will have the same effect. Changes in carbon cycling, including the development of a forest litter layer, might be expected to increase the production of DOC. However, the evidence that afforestation alters DOC export is surprisingly weak; although Grieve and Marsden (2001) recorded higher DOC in forest versus moorland soil solutions, Hope et al. (1994) actually found slightly lower export rates for temperate forests versus moorland and grasslands. Comparing CEH data from moorland and forest catchments on peaty podzols, no effect of land-use on DOC loss can be identified for either streams (Reynolds, in press) or soil solutions ( CEH unpublished data - see 2004 report). The strongest forest impact appears to accompany clearcutting, when large DOC pulses may occur (Hughes et al., 1990).
Grazing: Grazing of the uplands, primarily by sheep, can affect vegetation cover, compaction and drainage (Milne, 1996), and grazing intensity has in increased widely in recent decades. Available data, however, show little evidence that grazing changes affect DOC export. Worrall et al. (in press) found no significant difference in soil solution DOC between grazed and ungrazed heathland plots at Moor House. Data from the CEH Pwllpeiran experiment (B. Emmett, unpublished data) also showed similar soil solution DOC in a podzolic grassland under low and high grazing intensity. However, complete cessation of grazing in grasslands may lead to a relatively rapid transition to heathland (e.g. Hill et al., 1992), with potentially greater consequences for DOC losses, and C cycling in general.
Liming: Liming of acid soils may increase DOC production, by accelerating microbial activity, or reduce DOC retention by raising pH, a process analagous to that described earlier with regard to falling S deposition. Increased DOC loss due to liming has been shown in studies of moorland podzols (Hornung et al., 1986; Reynolds et al., 1994), peats (Soulsby and Reynolds, 1995) and forests (Andersson and Nilsson, 2001). Liming was widely practiced in the UK to improve the productivity of acid grassland, but has become less widespread since the 1980s due to the removal of subsidies. However, recent restoration work to revegetate degraded peats in the Pennines has included application of lime. It is suggested that this will lead to elevated DOC losses, potentially over a sustained period.
In summary, if future climate change results in warming and drying, DOC losses would be expected to increase. Further, climate induced changes in hydrology and plant productivity could increase C losses (at least in the short term) though the longer term impacts are uncertain. A number of land management practices, including drainage, burning, afforestation, grazing and liming, might also influence DOC losses. This is significant also from a water quality perspective as there are significant costs associated with removing colour from drinking water.
« Previous | Contents | Next »