Storage and stability of soil organic carbon down the profiles under native woodland, native pastures and cultivation

Abstract

Soils are widely recognized as having potential to sequester significant amounts of carbon (C). There is much speculation that deeper soil layers can store significant amounts of C in a relatively stable form and that land use can influence soil organic carbon (SOC) stocks at depth. Our understanding of the potential for subsoils to store C in the long term is however limited here in Australia, as little work has been done to quantify C stocks, and the available studies on C dynamics have been mainly focused in surface soils (30 cm or less), and conducted from a narrow range of management options. The main aim of this research was to examine SOC stocks and stability down texturally contrasted soil layers (up to 0.80 m) under three major land uses, namely native woodland, native pastures and cultivation, in the Northern Tablelands of New South Wales, Australia. Specifically, this study examined the effects of land use on a) the quantity of total SOC stocks down the profiles; b) dissolved organic carbon (DOC) concentrations down the profiles; c) the amounts and composition of dissolved organic matter (DOM) extracted from different litter types; b) quantity and quality of SOC among soil particle size fractions; c) SOC mineralization dynamics including C pools and turnover kinetics; and d) C and nitrogen (N) mineralization dynamics of decomposing plant litter and interactions with initial biochemical composition of litter. Understanding C stocks and organic matter stability in soils taking into account deeper soil layers as affected by land use will identify the most effective land use for carbon sequestration and thereby inform land use decisions in the northern region of NSW Australia.<br/> Strong differences in SOC stocks between land uses were only apparent in the surface 20 cm, with native pastures and cultivation which were statistically similar to each other containing on average 9 and 11 t/ha less C than native woodland. Significantly larger SOC stocks in surface soils under native woodland may be partly attributed to more lignin-rich recalcitrant above ground litter inputs compared with the other two land uses. The combined subsoil (20 to 80 cm) layers contained 40 % of the total profile SOC stocks across all land uses, demonstrating that substantial amounts of SOC stocks reside in deeper layers and so the importance of preserving it.<br/> The relative proportion of aromatic C in dissolved organic matter (DOM) extracted from litter was highest under native woodland, followed by native pasture then cultivation indicating qualitative differences in DOM fractions which may in turn influence biodegradability of DOM among litter types. DOC represented between 0.01 and 0.1 % of total SOC down the soil profiles and across all land uses showing that the contribution of DOC to C stocks was relatively small.<br/> Native woodland soils were associated with consistently wider C:N ratios in the particulate organic matter (POM) fraction throughout the soil profile compared with native pastures and cultivation which differed between each other in the top 50 cm. The result indicated differences in the quality of organic matter inputs (litter) entering the POM fraction with native woodland soils associated with less easily decomposed inputs due to their inherent chemical composition. Consequently, native woodland showed the least decline in particulate organic carbon (POC) with soil depth compared with the other two land uses. The proportion of MOC to the total SOC increased with soil depth indicating that subsurface C was more protected than surface C probably due to mineral association with clayey subsoils. Compared with native pastures and cultivation which were largely similar, native woodland soils contained significantly larger amounts of MOC in all soil depths suggesting that C was more physically protected from microbial attack.<br/> SOC mineralization kinetics over 419 days was well described by decomposition of a single pool. Compared with native pastures and cultivation, native woodland had larger amounts of the active C pools in all soil depths which were mainly related to larger amounts of labile substrates mainly DOC and POC down the profiles. The decomposition rate of the active C pool, measured by laboratory incubations, was strongly dependent on soil depth with turnover of 66 and 47 days in surface and subsoils respectively. Shorter turnover of active C pool in subsoils compared with surface soils may be linked to destabilization of active C stores when environmental constraints on decomposition which might inhibit decomposition in the undisturbed profile are removed following incubation under similar conditions in the laboratory. Consequently, it is important that the current C stores remain undisturbed.<br/> For all litter types, the active C pool whose decay rate constants ranged from 0.072 d-1 to 0.805 d-1, initially constituted < 20 % of the litter mass, while the slow C pool with decay rates of between 0.002 d-1 to 0.019 d-1 comprised > 80 % of the litter mass. The decomposition rate of the slow C pool in litter was strongly and negatively correlated with the initial lignin:N ratio of plant litter suggesting that the interaction between these two litter quality variables had important controls over litter decomposition. Compared with other litter types, above and below ground litter from native woodland had higher initial lignin:N ratio and were associated with more stable slow C pools with longer half lives of 109 and 446 days respectively.<br/> The above and below ground litter components had distinct N mineralization patterns during the early stages of incubation as influenced by the initial biochemical composition of plant litter namely C:N ratio, % lignin and % water soluble carbon (WSC). Our results suggest that the biochemically recalcitrant lignin influenced the susceptibility of substrates to microbial attack and thereafter a demand for N by microbial decomposers. However, the subsequent release of N from substrates depended on the C:N ratio.<br/> The results of this thesis are relevant to landholders, natural resource managers and policy makers as they inform that native woodland soil has an important role in storing a) larger SOC stocks in surface (20 cm) layers compared with native pastures and cultivation which were generally similar, b) relatively less decomposable C down the soil profiles as larger amounts of C are associated with the mineral fraction, and c) more slowly decomposing organic matter inputs which were characterized by relatively stable slow C pools in litter compared with native pastures and cultivation in the northern region of NSW Australia.<br/> Key future research areas arising from this thesis include to: a) investigate land use effects on total SOC stocks in soil profiles using increased intensity of soil sampling in order to represent spatial variation and increase capacity to detect differences in total SOC especially where differences in C might be small such as in non-wooded land uses, b) compare land use effect on DOC storage at various depths during crop/pasture growing season as the living plant biomass may influence DOC concentrations, c) determine the stability of DOC derived from both soils and litter in order to determine its long term dynamics in soils, d) determine the long term stability of soil slow C pool and investigate the mechanisms by which C might be stabilized including the contribution of char C at various depths under the three land uses, e) determine the mechanism by which lignin and N interact to influence decomposability of litter, and f) understand the mechanism of SOC and N stabilization.