The role of plants and land management in sequestering soil carbon in temperate arable and grassland ecosystems

Abstract

Global climate change and concerns about soil quality have led to a widespread interest in the opportunities that are available to sequester carbon in soils. To achieve a better understanding of the changes in C storage, we need to be able to accurately measure and model inputs and losses of C from soils. This in turn requires a thorough understanding of the biological processes involved and the way in which they are influenced by the soil's physical and chemical environment. The amount of C present in a soil is determined by the difference between C addition and C loss. Because these fluxes are large relative to changes in C storage, net storage can be very difficult to measure, particularly in the short term. Carbon is added to soil from plant and animal materials deposited on the soil surface. It is known that approximately 50% of C assimilated by young plants can be transferred below ground; some is used for root construction and maintenance as well as root respiration; some organic C is lost to the soil through exudation and root turnover. A comparison of eight studies has shown that the input to the soil of root derived organic C during a growing season can range between 0.1 and 2.8 t C ha⁻¹. Quantifying inputs from different processes has proved difficult and the relative importance of exudation and root death under field conditions remains uncertain. The chemical composition of substrates released by exudation and root death is known to be very different. Exudates contain high concentrations of soluble organic substrates and as a consequence are highly labile, whereas additions of C from root death have structural organic substrates with lower potential decomposition rates. Losses of C from soil occur as a consequence of plant and microbial respiration. However, identifying the source of evolved CO₂, whether it be from root or microbial respiration, is much more difficult. Some new methods using isotopic labelling and pool dilution have been developed to separate plant and microbial respiration, and despite difficulties, these promise to provide valuable information on the processes of C input and loss from soils.

At a field scale measurements and models would suggest that soil and crop management can play a significant role in determining the extent of C sequestration by soils and the proportion of labile C present. A comparison of 11 field studies showed that soil respiration varies between 4 and 26 t C ha⁻¹ year⁻¹, with management such as tillage, drainage, grazing and manure application exerting a strong influence on the magnitude of fluxes. Net ecosystem exchange of C has been shown to be at least an order of magnitude lower than respiratory losses in comparable studies, but land management is important in determining the direction and magnitude of the C flux. Recent studies have suggested that although the overall quantity of C stored in European soils is increasing, this increase is confined largely to forested areas and that many cropped soils are losing soil organic matter. It is has been suggested that that the biological potential for C storage in European cropland lies between 9 and 120 Mt C year⁻¹. In order to take advantage of this potential and to develop management systems that promote C storage we need to achieve a better understanding of the processes of C input and loss, and develop improved models using pools that coincide with measurable soil C fractions.

Introduction

There is an urgent requirement to improve our understanding of the processes contributing to C storage in soils. This has arisen because of the need to sequester C to overcome global climate change (Paustian et al., 1998a) and to improve soil quality, as we develop more sustainable and land management practices (Carter, 2002). Protecting and enhancing stocks of soil organic matter will remain a global challenge that needs to involve environmental scientists, socio-economists and policy makers for the next decade and beyond. This exercise will require an improved understanding of the mechanisms underlying C sequestration and involve a re-evaluation of modern and traditional farming practices, as the principles of environmental protection and sustainable management of non renewable resources gain a higher priority. Reflecting these concerns, the European Union's common agricultural policy will place increased importance on soil quality and land management, as financial support is shifted from production towards environmental protection (European Commission, 2002).

A range of new techniques has become available in recent years that have advanced our understanding of the processes underlying organic matter transformations in soil. Through the use of isotopic tracers we now have estimates of the quantities of plant derived C that are added to soil. Both continuous labelling and pulse labelling techniques have indicated that the soil environment is a major sink for C derived from exudation and root turnover (Meharg, 1994). This C contributes to root construction and maintenance, but is also added to the soil through plant respiration and rhizodeposition. The addition of C to soil stimulates microbial activity, which is the driving force behind transformations of organic matter. Molecular techniques are allowing us to examine the relationship between microbial activity, biodiversity and function in soils. It has become clear that the diversity of microbial communities is greater than had been previously thought (Tiedje et al., 2001), however, such diversity may not always be a pre-requisite for soils to undertake a wide range of functions. Many soil functions have been shown to be unaffected by experimentally manipulating the biodiversity of microbial communities (Griffiths et al., 2001), indicating that there is widespread functional redundancy within the system. Where soil functions can be linked to specific biological populations an opportunity exists to manipulate soil processes through alterations of the soil microbial community (Atkinson and Watson, 2000). Experimental techniques allowing physical and chemical characterisation of soils are becoming increasingly sophisticated and are allowing us to explore the relationships between living organisms and the environment within which they function. Young et al. (1998) emphasise the importance of the soil's structure in maintaining biological diversity through the provision of habitat and by influencing patterns of resource storage and transport. Modern tools in analytical chemistry are forcing us to re-examine long held theories of soil organic matter formation. For example Hatcher (2003) points to evidence from recent NMR studies that suggest that soil organic matter forms as a consequence of the transformation of biopolymers by microbial action and not as a result of random condensation reactions. The physical fractionation of soils is also helping us to understand how organic matter pools can become chemically and physically isolated. Hassink et al. (1997) have shown that the amount of organic matter that a soil can store is largely regulated by its silt and clay content although management can also influence storage, particularly of larger macro-organic matter fractions (Balesdent et al., 2000, Denef et al., 2001). Spatial variation of soil organic matter content particularly at national and regional scales are also strongly influenced by climate and land use (Joos et al., 2001, Paustian et al., 1998b).

The atmospheric concentration of CO₂ is currently higher that it has been for at least 420 000 years, and may well double in this century. This is leading to a very complex series of interactions between the biosphere, hydrosphere and atmosphere as the equilibration between soils, plants and the atmosphere breaks down. Human activity has had a major impact on the C cycle. Pre-industrial concentrations of CO₂ were below 300 μL L⁻¹, but today's atmospheric CO₂ concentrations are around 370 μL L⁻¹ and rising. Measurements of the isotopic signature of CO₂ in the atmosphere have established that fossil fuel consumption is principally responsible for the concentration rises (Quay et al., 2003). Land use change also contributes to an increase in atmospheric CO₂ as a consequence of deforestation and cultivation of new arable land (Schimel et al., 2001). Most recent estimates suggest that during the period between 1989 and 1998 land use changes resulted in the addition of 1.6 Pg C per year to the atmosphere (Houghton, 2000). However, not all of the anthropogenically produced CO₂ remains each year in the atmosphere. There is a significant removal of atmospheric C by the oceans (2.3 Pg C year⁻¹) and an increasing sink within the terrestrial biosphere (Bremer et al., 1998). Studies of the C cycle during the 1990s suggest that as a consequence of both fossil fuel consumption and land use change the atmospheric pool of C is currently increasing by around 3.2 Pg per year (Frank et al., 2002). The regional distribution of sink strength in the terrestrial biosphere is more difficult to determine but it is suggested that the mid to high latitudes of the northern hemisphere are particularly important (Janssens et al., 2003). The Kyoto protocol commits signatories to reducing C emissions to 0.3 Pg below 1990 levels. Although this could be achieved partly by reduced dependence on reserves of fossil fuels, it is likely that land use change will play a vital role in contributing to atmospheric CO₂ removal. It has been estimated that agricultural soils offer the potential to sequester between 0.4 and 0.9 Pg C year⁻¹, through improved management of existing agricultural soils, restoration of degraded land, the more extensive use of setaside, and the restoration of wetlands (Paustian et al., 1998a). These opportunities to sequester C will probably only be available until the middle of the 21st century after which it is anticipated that soils may return to being net source, as a consequence of respiratory losses increasing more rapidly than inputs by photosynthesis (Prentice, 2001).

To understand why the soil has a particular organic matter content, it is necessary to quantify both inputs and outputs of C to the system. Unfortunately these are both difficult to measure and the difference between input and output is normally small, further complicating estimates of change in storage. To predict how the organic matter content responds to changing environmental and management pressures at local, regional and global scales, we need robust models that take account of the complex physical, chemical, and biological controls influencing organic matter turnover. Models of soil organic matter turnover allow us to encapsulate our understanding of the mechanisms regulating organic matter dynamics in soil and then to investigate the effects of environmental and management change on rates of turnover. Although short-term (less that 1 year) environmental and management changes remain difficult to predict, a number of models can successfully predict longer-term change in a range of diverse environments (Smith et al., 1997, Parton et al., 1998, Paustian et al., 1997b). The integration of soil organic matter models within whole ecosystem simulations allows a more holistic assessment of ecosystem responses to environmental change and is highlighting management strategies that can be used to optimise C sequestration through targeted management of soils and vegetation (Paustian et al., 2000, Eve et al., 2002). Models have also highlighted the complexity of interactions that exist within the soil environment, and have indicated areas where our understanding is still incomplete. This review aims to summarise our understanding of the mechanisms by which plants influence the quantity of C contained in soils, through both a consideration of the processes of C addition and loss. We then go on to discuss how land management can be used to influence that content in agro-ecosystems.