Soil carbon density can increase when Australian savanna is converted to pasture, but may not change under intense cropping systems

https://doi.org/10.1016/j.agee.2021.107527Get rights and content

Highlights

  • Soil C was measured to 1.0 m under savanna woodland, pasture and cropping systems.

  • Converting savanna to pasture significantly increased soil C density in the upper 0.3 m.

  • Melon-sorghum rotation cropped soil had significantly greater bulk density than savanna in the upper 0.5 m.

  • Converting savanna to melon-sorghum rotation cropping did not significantly change soil C density.

  • Estimating soil C density to 0.3, 0.6 or 1.0 m changed statistical significant differences.

Abstract

Savanna regions are increasingly developed for agriculture to support population growth, food demand and export economies. This is driving interest in the conversion of natural savanna to cattle grazed pastures or horticultural crops in Northern Australia. Savanna clearing leads to aboveground carbon (C) loss but impacts below-ground are less certain as studies have focused on high rainfall regions, low reactive clay soils and shallow depths of 0.3 m or less.

To examine impact of land use change (LUC) in Northern Australia, we sampled the upper 1.0 m soil profile of: 1) savanna woodlands, 2) old and young C4 perennial cattle grazed pastures and 3) old and young, melon-sorghum rotation fields. Soil C concentrations and soil bulk density in the upper 0.3 m were significantly greater in cattle grazed pastures than savanna woodland, such that savanna to pasture LUC significantly increased soil C density from 30.2 ± 1.5–43.4 ± 3.0 Mg C ha-1 over 28 years. Soil bulk density in the upper 0.7 m of savanna woodland was significantly less than that in the melon-sorghum rotations. In contrast, soil C concentrations below 0.5 m were significantly greater in the savanna woodland than in the 3 y.o. melon-sorghum rotation fields. There was no significant difference in soil C density between savanna woodland and 12 y.o. melon-sorghum rotations. This may relate to the beneficial effects of ploughing in sorghum as a green manure, or simply that the decrease in soil C concentration was offset by increased soil C density in the upper layers of soil where most C is located. In contrast, conversion of savanna to cattle grazed pasture sites is equivalent to a soil sequestration rate of 0.34 Mg C ha-1 y-1, approximately 50% of the observed net ecosystem productivity estimated using eddy covariance methods. Changing the depth to which soil C density was estimated changed significant differences amongst land-use systems. Whereas, estimating soil C density using a ‘fixed depth’ or ‘equivalent mass’ method did not change the statistical significance.

There is an increasing focus on the potential of soil carbon sequestration especially in north Australia and understanding potential change and impact of accounting methods is essential for meaningful policy development. Policy settings to maintain, or improve, soil C storage and soil health may contradict agricultural development goals.

Introduction

The tropical savanna is an extensive tropical ecosystem defined by the coexistence of both woody (trees, shrubs) and grass lifeforms, and are transitional between grasslands and forests (Scholes and Archer, 1997, Hutley and Setterfield, 2019). They occur as major biomes in South America (Cerrado), Sub-Saharan Africa and Northern Australia and are increasingly under pressure from land use change (LUC) (Eriksen and Watson, 2009, Powers et al., 2011, Adams and Pressey, 2014). Rates of savanna clearing in South America now exceed tropical rainforest clearing (Carvalho et al., 2010, Grecchi et al., 2014). Rates of savanna clearing in Australia will increase if proposed agricultural development of the northern regions proceeds (Bradshaw, 2012, Lawes et al., 2015). The global importance of above-ground C stock loss from savanna LUC to agriculture has been well researched (Law and Garnett, 2011, Buitenwerf et al., 2012, Liu et al., 2013, Poulter et al., 2014, Bristow et al., 2016) but impacts of savanna LUC on soil C storage has received less attention. This is despite the fact that soils provide the greatest terrestrial C store globally (Lal, 2004) and are the target of C sequestration initiatives in Australia (Australian Government, 2018) as well as globally (Minasny et al., 2017).

Within the tropical bioclimatic region, spatial variability in soil C is such that large prediction errors of unknown direction and magnitude can be expected if soil C change is extrapolated from one region to others (Powers et al., 2011). Previous savanna soil C studies (Buitenwerf et al., 2012) have: i) focused on high rainfall (>1500 mm y-1) regions with low reactive clay soils, ii) have often measured to a depth of 0.3 m or less, and iii) have often not measured concurrent changes in soil bulk density (BD) (Powers et al., 2011). Strategic research to sample under-represented LUC transitions and bioclimatic regions (such as tropical savannas) is necessary. There have been few studies of soil C change in lower rainfall (<1200 mm y-1) and in sandy to sandy loam soils that typify many savanna ecosystems (Lehmann et al. 2014). Furthermore, future global LUC is expected to continue in tropical savanna regions, such that savannas may be a net C source to the atmosphere (Eglin et al., 2010, Beringer et al., 2015). Over the last decade, the rate of Amazonian forest deforestation has more than halved (Lapola et al., 2014), but clearing rates in the vast Brazilian Cerrado (2.04 million km2) have been maintained (Spera et al., 2016) and are likely to become the principal region of LUC in Brazil (Lapola et al., 2014, Barros and Fearnside, 2016). In Australia, savanna clearing rates in the State of Queensland have increased during the last decade and Australia is now a top ten country for rates of deforestation; the only developed county with this recognition (Taylor, 2015, Queensland Department of Science, Information Technology and Innovation, 2017).

The impact of any LUC event upon soil properties is driven by changes in both soil BD and soil C concentration (Paustian et al., 2019, Smith et al., 2020). Detecting the magnitude and significance of any change in soil C density depends upon i) the rate of change in soil properties, ii) the period of time between measurement events, and iii) the spatial coefficients of variation for those soil properties. There can be an apparent change in soil C density simply from an increase or decrease in soil BD following a LUC or management change (Goidts et al., 2009, Lee et al., 2009). It is therefore essential to concurrently measure soil BD and soil C concentration; however, many studies have relied on pedo-transfer functions to estimate soil BD because soil C inventories do not always provide this data (Schrumpf et al., 2011). In addition, there are several methods by which soil C density may be estimated from measures of soil C concentration and soil BD.

Soil C density is often estimated by a ‘fixed depth’ method, for example 0.3 m, regardless of horizon changes or differences in soil BD. The ‘equivalent mass’ method is encouraged to better account for concurrent changes in soil BD from management or LUC (Ellert et al., 2007, Schrumpf et al., 2011). The importance of these different methods for detecting change in soil C density under LUC have been recently highlighted by von Haden et al. (2020) who illustrate significant differences in these approaches when quantifying massed based soil carbon estimates, C:N mass ratio and δ13C.

Understanding how LUC impacts different soil horizons can greatly assist in designing appropriate protocols for sampling soil C change (Ellert and Bettany, 1995). However, many soil types in Australian savanna ecosystems exhibit possess weakly defined A and B horizons (Isbell, 2002). A further issue is the depth to which soil C density is estimated (Jobbágy and Jackson, 2000, Rumpel and Kögel-Knabner, 2011). It has been demonstrated that the conclusions of soil C change studies can change once deeper soil layers (> 30 cm) are included (Degryze et al., 2004, Harrison et al., 2011). Soil C density deeper in the soil profile can change because of disequilibrium between continued subsoil C mineralisation and C input from root turnover, root exudates, dissolved organic matter, bioturbation, and downward translocation of colloidal C (Osher et al., 2003, Rumpel and Kögel-Knabner, 2011). In general, the greatest changes in soil C density occur in the topsoil and subsoil layers experience smaller changes in the same direction (Guo and Gifford, 2002, Poeplau and Don, 2013). However, the proportional importance of lower soil depths to soil C density of a whole profile (e.g. 1.0 m) is greatly influenced by climate, vegetation type and soil type (Jobbágy and Jackson, 2000). LUC invariably leads to a change in vegetation type, which in turn can lead to significant changes in bulk density or soil C concentration deep in the soil profile (Osher et al., 2003, Chen et al., 2005, Rumpel and Kögel-Knabner, 2011).

The savanna biome of Northern Australia is one of the largest and most intact tracts of savanna globally and provides an opportunity to investigate C dynamics when intact savanna woodland undergoes LUC to agricultural systems. The Douglas-Daly River catchment has only experienced LUC since the early 1970’s with less than 10% of the catchment cleared to date. Dominant land use includes pasture and meat livestock grazing as well as some irrigated and fertilised, high-value, horticultural crops and hay production (Bowman et al., 2010, Grover et al., 2012). Previous rates of LUC are unlikely to reflect future patterns given the renewed political focus towards agricultural development of savanna regions as a climate change adaptation response given the high likelihood of southern agricultural basins experiencing drying and warming trends (Adams and Pressey, 2014).

This study aims to assess the impact of two LUC transitions upon soil properties and soil C density, in the Douglas-Daly River catchment of northern Australia. The contributory importance of changes in soil C concentration and soil BD was investigated as well as the impact of measurement depth. The specific research questions were:

  • 1)

    How do changes in soil BD and carbon concentration contribute to change in soil carbon density when savanna is converted to a) cattle grazed pasture or b) melon-sorghum rotation cropping?

  • 2)

    Are differences in soil C density associated with LUC, the time since conversion and their levels of subsequent management intensity?

  • 3)

    Are our estimates of soil carbon change significantly influenced by whether ‘fixed depth’ or ‘equivalent mass’ methods of estimation are used?

Section snippets

Materials and methods

Savanna LUC and the impacts upon soil properties were investigated in two regions of the Douglas-Daly River catchment: 1) Daly River and 2) south Katherine (Fig. 1). These two regions are approximately 125 km apart. The natural savanna woodland in both regions is dominated by an overstorey of Eucalyptus tetrodonta (F. Muell.), Corymbia capricornia, and Erythrophleum chlorostachys (F. Muell.) trees with smaller deciduous trees and shrubs and an understorey dominated by native annual and

Savanna to pasture LUC impact upon soil BD and %C

Soil BD in the Daly River increased significantly in the upper 0.3 m after clearing and conversion to cattle grazed pastures (p ≤ 0.05; Table 1). Soil BD increased 5 years after LUC and continued to increase between 5 and 28 years after LUC, but only significantly so at 0.2–0.3 m (Table 1). There was no significant change in soil BD below 0.3 m. Similarly, soil %C increased significantly (p ≤ 0.05) with savanna conversion to cattle grazed pasture, but only in the upper 0.2 m. Soil C

The impact of savanna LUC upon soil properties

With LUC of savanna to cattle grazed pasture there was an increase in soil BD in the upper 0.3 m and an increase in soil %C in the upper 0.2 m. Whereas, soil BD was significantly greater in the upper 0.5 m of melon-sorghum cropping rotation as compared to savanna, and soil %C concentration was significantly less in the upper 0.1 m and significantly greater at 0.2–0.3 m of melon-sorghum cropping in 3 and 12 year-old fields, respectively. Soil BD can increase for several reasons. Grazed savanna

Conclusions

In our study, both soil BD and soil C concentration changed significantly with the conversion of savanna woodland to either cattle grazed pasture or melon-sorghum rotation cropping. There were some significant changes in soil C density in the upper 0.3 m of the cattle grazed pasture systems, and some significant differences in soil C density between savanna and melon-sorghum systems down to a depth of 0.8 m. The changes in soil BD and C concentration in combination led to a significant increase

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research was funded by the Australian Research Council grants LP0774812, LP100100073 and the Discovery Project, grant DP0772981. Support was provided by our partner organisations the Northern Territory Government and the Department of Environmental and Energy, Canberra. We would like to acknowledge the field and laboratory assistance of Bianca Baldiserra, Luke Wiley, Hizbullah Jamali, Susanna Venn, Mick Brand, Amanda Lilleyman and Emma Lupin. We also thank the many land holders in the Daly

References (94)

  • C.P. Jantalia et al.

    Tillage effect on C stocks of a clayey Oxisol under a soybean-based crop rotation in the Brazilian Cerrado region

    Soil Tillage Res.

    (2007)
  • R. Lal

    Soil carbon sequestration to mitigate climate change

    Geoderma

    (2004)
  • J. Lee et al.

    Determining soil carbon stock changes: Simple bulk density corrections fail

    Agric. Ecosyst. Environ.

    (2009)
  • S.J. Livesley et al.

    Seasonal variation and fire effects on CH4, N2O and CO2 exchange in savanna soils of northern Australia

    Agric. For. Meteorol.

    (2011)
  • F. Mapanda et al.

    Short-term responses of selected soil properties to clearing and cropping of miombo woodlands in central Zimbabwe

    Soil Tillage Res.

    (2013)
  • S.A. Materechera

    Tillage and tractor traffic effects on soil compaction in horticultural fields used for peri-urban agriculture in a semi-arid environment of the North West Province

    S. Afr. Soil Tillage Res.

    (2009)
  • A. Metay et al.

    Storage and forms of organic carbon in a no-tillage under cover crops system on clayey Oxisol in dryland rice production (Cerrados, Brazil)

    Soil Tillage Res.

    (2007)
  • B. Minasny et al.

    Soil carbon 4 per mille

    Geoderma

    (2017)
  • K.I. Paul et al.

    Improved models for estimating temporal changes in carbon sequestration in above-ground biomass of mixed-species environmental plantings

    For. Ecol. Manag.

    (2015)
  • C. Poeplau et al.

    Sensitivity of soil organic carbon stocks and fractions to different land-use changes across Europe

    Geoderma

    (2013)
  • M.J. Pringle et al.

    Soil carbon stock in the tropical rangelands of Australia: effects of soil type and grazing pressure, and determination of sampling requirement

    Geoderma

    (2011)
  • R. Roscoe et al.

    Tillage effects on soil organic matter in density fractions of a Cerrado Oxisol

    Soil Tillage Res.

    (2003)
  • W.C. Tim Chamen et al.

    Mitigating arable soil compaction: a review and analysis of available cost and benefit data

    Soil Tillage Res.

    (2015)
  • V.M. Adams et al.

    Uncertainties around the implementation of a clearing-control policy in a unique catchment in northern Australia: exploring equity issues and balancing competing objectives

    PLoS One

    (2014)
  • D.E. Allen et al.

    Impacts of fire on soil organic carbon stocks in a grazed semi-arid tropical Australian savanna: accounting for landscape variability

    Rangel. J.

    (2014)
  • D.E. Allen et al.

    Effects of land-use change and management on soil carbon and nitrogen in the Brigalow Belt, Australia: I. Overview and inventory

    Rangel. J.

    (2016)
  • D.E. Allen et al.

    A review of sampling designs for the measurement of soil organic carbon in Australian grazing lands

    Rangel. J.

    (2010)
  • A.J. Ash et al.

    How season of grazing and herbivore selectivity influence monsoon tall-grass communities of northern Atlanta

    J. Veg. Sci.

    (1998)
  • A.J. Ash et al.

    Building grass castles: integrating ecology and management of Australia’s tropical tallgrass rangelands

    Rangel. J.

    (1997)
  • E.D. Assad et al.

    Changes in soil carbon stocks in Brazil due to land use: paired site comparisons and a regional pasture soil survey

    Biogeosciences

    (2013)
  • Australian Government

    The Supplement to the Carbon Credits (Carbon Farming Initiative—measurement of Soil Carbon Sequestration in Agricultural Systems) Methodology Determination

    (2018)
  • J. Beringer et al.

    Fire in Australian savannas: from leaf to landscape

    Glob. Change Biol.

    (2015)
  • J. Beringer et al.

    An introduction to the Australian and New Zealand flux tower network – OzFlux

    Biogeosciences

    (2016)
  • J. Beringer et al.

    Technical note: dynamic INtegrated Gap-filling and partitioning for OzFlux (DINGO)

    Biogeosciences

    (2017)
  • D.M.J.S. Bowman et al.

    Biogeography of the Australian monsoon tropics

    J. Biogeogr.

    (2010)
  • C.J.A. Bradshaw

    Little left to lose: deforestation and forest degradation in Australia since European colonization

    J. Plant Ecol.

    (2012)
  • M. Bristow et al.

    Quantifying the relative importance of greenhouse gas emissions from current and future savanna land use change across northern Australia

    Biogeosciences

    (2016)
  • R. Buitenwerf et al.

    Increased tree densities in South African savannas: >50 years of data suggests CO 2 as a driver

    Glob. Change Biol.

    (2012)
  • R. Buitenwerf et al.

    Interactions between large herbivores and litter removal by termites across a rainfall gradient in a South African savanna

    J. Trop. Ecol.

    (2011)
  • X. Chen et al.

    Soil organic carbon content at a range of north Australian tropical savannas with contrasting site histories

    Plant Soil

    (2005)
  • X. Chen et al.

    Carbon balance of a tropical savanna of northern Australia

    Oecologia

    (2003)
  • F. Conen et al.

    Potential for detecting changes in soil organic carbon concentrations resulting from climate change

    Glob. Change Biol.

    (2003)
  • F. Conen et al.

    The carbon balance of forest soils: detectability of changes in soil carbon stocks in temperate and Boreal forests

  • B. De Vos et al.

    Predictive quality of pedotransfer functions for estimating bulk density of forest soils

    Soil Sci. Soc. Am. J.

    (2005)
  • S. Degryze et al.

    Soil organic carbon pool changes following land-use conversions

    Glob. Change Biol.

    (2004)
  • M. Dilshad et al.

    Surface runoff, soil and nutrient losses from farming systems in the Australian semi-arid tropics

    Aust. J. Exp. Agric.

    (1996)
  • R.P. Eclesia et al.

    Shifts in soil organic carbon for plantation and pasture establishment in native forests and grasslands of South America

    Glob. Change Biol.

    (2012)
  • View full text