The quantity, distribution and stability of root carbon inputs from pasture species into soil

Abstract

<p>Levels of global atmospheric carbon dioxide (CO₂) concentrations have risen in the last 150 years due to the burning of fossil fuels, land clearing, intensification of agriculture and increasing urbanisation. The increasing levels of greenhouse gases (GHG), including atmospheric CO₂, methane and nitrous oxide, have led to considerable international concern regarding the climate change implications of GHG emissions and considerable research investment has taken place to explore mitigation/adaptation options. Soils globally have been depleted of their carbon as a result of land clearing, soil disturbance and agricultural activity, and are estimated to have lost approximately 50-75% of their original carbon content. This is alarming, but equally offers the potential to recover at least some of this lost carbon through modified management. Indeed, soil is one of the main carbon reservoirs that humans can influence, and soils as a potential carbon sink therefore, represent a key area for research.</p> <p>In order to fully understand the carbon sequestration potential of our soils, we require information on all aspects of the soil ecosystem, including the influence of roots on the soil organic carbon (SOC) pool. This includes the quantity of root carbon contributions, the extent to which roots influence the soil matrix, both close-to and further-from the root, and the stability of these root carbon inputs. Puget and Drinkwater (2001) reported that nearly 50% of the root derived carbon remained present in the soil, while only 13% of the shoot-derived carbon remained. Despite the recognition of the role of roots in building soil organic matter, evidence regarding the nature and magnitude of their influence on SOC is currently lacking. The overall aims of this research were: 1) to quantify root carbon inputs and identify the root parameters that might be used as indicators to determine the potential for roots/plants to contribute to soil carbon sequestration" 2) to determine the zones of influence of the root system on soil carbon and the diffusion distances of carbon-containing root exudates" 3) to investigate how much of the soil volume is explored by a root system and how sequential plant growth cycles change the volume explored" 4) to assess the stability of root carbon inputs over multiple phases of growth and decomposition" and 5) to investigate the influence that soil type has on root carbon inputs, diffusion distances and stability.</p> <p>To achieve these aims, a series of pot trials was established using C⁴ pasture species (Chloris gayana and Sorghum bicolor) grown in soils with a history of C³ vegetation. This allowed changes in total organic carbon (TOC) and δ¹³C due to root inputs to be measured using an isotope ratio mass spectrometer (IRMS) across a series of experiments with various temporal and spatial scales. The 'clay soil' was sourced from the 0-20 cm depth of a Ferrosol and had a clay content of 51.1%. The 'sand soil' was sourced from the 0-20 cm depth of a Chromosol and had a clay content of 10.0%.</p> <p>Root dry matter and root length density were proven to be useful indicators in determining the potential for root carbon inputs. Lower root dry matter or root length density resulted in lower additional carbon inputs. Regression analysis demonstrated that both indicators were useful for determining additional organic carbon inputs from a root system, with the strongest relationship evident in the clay soil compared to a sand. With evidence for soil type as a factor that influences the quantity of root-derived organic carbon present in the soil, it is vital that a variety of soil types and textures are utilised in SOC research.</p> <p>The addition of root exuded organic carbon was greatest in the areas closest to the roots and decreased with distance away from the root. The clay and sand soil types contrasted in their behaviour with regard to TOC and δ¹³C. In the clay there was a loss of TOC but an enriched δ¹³C value (less negative δ¹³C), whereas in the sand the TOC increased but the δ¹³C became depleted (more negative δ¹³C). This was attributed to differences in the existing organic carbon levels of the two soil types (clay 4.65% TOC, sand 0.93% TOC), soil particle structure and bioavailability of the root derived carbon. With the clay having a larger existing TOC (4.65%) and the short time period of root carbon inputs, the decrease in TOC is likely due to a rhizosphere priming effect. With the loss of existing TOC that is C³ in origin and the input of some C⁴ carbon, the ratio of ¹³C/¹²C can alter without an increase in the TOC, hence the relative change in δ¹³C values.</p> <p>A significant difference in root distribution was observed with soil depth, with the quantity of C⁴ plant root biomass in the 0-10 cm soil depth ranging between 42 – 46%, which was significantly greater than each of the lower depths (14 – 27%). Despite the addition of C⁴ plant roots at all depths, in particular a greater quantity of C⁴ root material in the 0-10 cm depth, no significant change was observed in TOC or δ¹³C after two periods each of plant growth and decomposition. Gas analysis indicated that the rate of carbon loss was greatest during the first seven days of the decomposition phase, with the C⁴-derived carbon the main source of microbial respiration, indicating that much of the newly added organic carbon was rapidly decomposed and lost from the soil.</p> <p>X-ray computed tomography scans of the soil matrix using 'region of interest' on a 50 mm pot at a resolution of 20 µm were inadequate to separate the roots from the soil and thereby determine the successive utilisation of pore spaces and potential for organic carbon hotspots. This was due to similar contrast to noise ratios of the two variables and the threshold analysis being unable to separate the roots and soil. Nevertheless, these x-ray computed tomography scans were used to assess, soil porosity, pore thickness and pore connectivity, and indicated the development of soil structure associated with root growth following a series of Rhodes grass and Sorghum growth and decomposition cycles. The resulting soil structure was more homogenous with Rhodes grass and heterogenous with Sorghum. The contrasting soil structures and root morphology of the two plant species potentially leads to carbon 'hotspots' within the soil matrix, which prompts the question about whether small quantities of root carbon inputs evenly distributed throughout the soil matrix by a fine root system, is superior to a concentrated 'hot spots' of carbon inputs from a coarse root system, for carbon sequestration potential.</p> <p>The root influence on the TOC varied with time and space and the level of influence was determined by the scale at which the factors, relationships and processes were investigated. In this research it has been demonstrated that soil type had a major influence on the capacity of soil to capture and store root derived carbon. Root dry matter and root length density measures were proven to be useful indicators in determining the potential for root carbon inputs. Changes in root exuded organic carbon were greatest in the areas closest to the roots and decreased with distance away from the root, with newly added organic carbon rapidly decomposed and lost from the soil. Furthermore, the potential for carbon 'hotspots' within in soil matrix, and how this is dependent on the root morphology of the plant species and the structure of the soil was demonstrated. Overall, this research has highlighted how the effectiveness of pasture roots in contributing additional organic carbon to the soil for carbon sequestration involves more complex mechanisms than is often reported in the literature. Continued research into how roots influence the soil environment will advance our understanding of and help address the demonstrated complexities in determining the quantity, spatial distribution and stability of organic carbon inputs from a root system.</p>