Mapping Soil Organic Carbon in SA Modelling the

Mapping Soil Organic Carbon in SA & Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes Stefanie Schütte (Ph. D Candidate) & Prof Emeritus Roland Schulze SANCIAHS, 18. 9. 2018 INSPIRING GREATNESS

Outline: 2 related projects 1. ) Mapping Soil Organic Carbon in SA • • Soil Organic Carbon SA’s variable soils Land Types and Terrain Unit Databases ARC Soils Database: carbon, clay, sand content Calculating median carbon % per soil series Calculating and mapping average carbon content per terrain unit Summary 2. ) Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes

Soil carbon • A typical soil consists of inorganic minerals, water, air and organic matter • Physical soil properties include texture, structure, density and porosity (Shaxson, 2006) • Soil organic carbon (SOC) is the main component of soil organic matter (50 -58%) • (Howard, 1965; Pribyl, 2010) • 3

There is a Wide Spectrum of Soils with associated Properties in South Africa

OBJECTIVE DETAILED MAPPING OF SOIL CARBON STOCKS OF SA SOILS TOOLS ARC / Beukes Terrain Unit Database 27 471 Lines of Attributes Soils Land Type Maps & Descriptions 6 700 Units Multiple Soil Series ACRU Model Hydrological Attributes 1 779 Lines of Attributes ARC Soil Carbon Database 19 138 Lines of Attributes

SOIL MAPS OF THE RSA WITH ASSCIATED DATABASES Soils LAND TYPES from Binomial Soil Classification # Based on relatively uniform climate, terrain, soil patterns # With detailed soils inventory on soil series, clay %, texture class, profile thickness…

Terrain Units A Land Type is made up of several Terrain Units, each with several soil series (crest, scarp, midslope, footslope, valley bottom) Delineated with a 90 m DEM (Beukes, ARC)

Locations of ARC Soil Samples Locations of > 11. 000 sample points from which soil carbon data, as well as clay and sand content were used in this study

OBJECTIVE DETAILED MAPPING OF SOIL CARBON STOCKS OF SA SOILS TOOLS ARC / Beukes Terrain Unit Database Soils Land Type Maps & Descriptions Multiple Soil Series SA SOIL CLASSIFICATIONS ANALYSES Binomial Carbon content (%) f(C%; BD; TH) = C density (t C/m 3) Per Soil Series, per Topsoil, Subsoil, Natural Vegetation, Agricultural Land Uses Analyses Per “ 501 Soil Series” from 19 138 Soil Samples Binomial 5 356 Locations, 10 559 Samples Taxonomic 5 740 Locations, 8 569 Samples Quality Control Reconcile Convert ARC Soil Carbon Database

Median Carbon Content per Soil Series were calculated Case Study Example: Median Carbon Percentages for Hu 16, grouped by topsoil, subsoil, as well as natural vegetation and agricultural land uses • Median Carbon %, Hu 16 • 2, 0 • 1, 8 • 1, 6 • Carbon % • 1, 4 • 1, 2 • 1, 0 • 0, 8 • 0, 6 • 0, 4 • 0, 2 • 0, 0 • Topsoil All Land Uses • Subsoil All Land Uses • Topsoil Nat Vegetation • Subsoil Nat Veg • Topsoil Agriculture • Subsoil Agriculture • Topsoil All Land Uses • Subsoil All Land Uses • Topsoil Nat Vegetation • Subsoil Nat Veg • Topsoil Agriculture • Subsoil Agriculture

OBJECTIVE DETAILED MAPPING OF SOIL CARBON STOCKS OF SA SOILS TOOLS ARC / Beukes Terrain Unit Database Soils Land Type Maps & Descriptions Multiple Soil Series ANALYSES Terrain Unit Analyses: up to 15 soil series per TU Averages per TU, per Topsoil, Subsoil, Natural Vegetation, Agricultural Land Uses C%, Thickness Carbon content (%) Per Soil Series, per Topsoil, Subsoil, Natural Vegetation, Agricultural Land Uses Analyses Per “ 501 Soil Series” from 19 138 Soil Samples ARC Soil Carbon Database

OBJECTIVE DETAILED MAPPING OF SOIL CARBON STOCKS OF SA TOOLS ARC / Beukes Terrain Unit Database Soils Land Type Maps & Descriptions Multiple Soil Series ANALYSES Terrain Unit Analyses: up to 15 soil series per TU Averages per TU, per Topsoil, Subsoil, Natural Vegetation, Agricultural Land Uses Thickness; Area; C%; Carbon content (%) Per Soil Series, per Topsoil, Subsoil, Natural Vegetation, Agricultural Land Uses Analyses Per “ 501 Soil Series” from 19 138 Soil Samples ARC Soil Carbon Database PRODUCTS DETAILED MAPPING PER TERRAIN UNIT: CARBON CONTENT, THICKNESS, FOR TOP & SUBSOILS

Example 1: Median % SOC in the Topsoil, Natural Vegetation. . and zooming in

Example 2: Median % SOC in the Subsoil Natural Vegetation…Zooming In

Example 3: Median % SOC, Top- and Subsoil, Agric LU

Mapping at TU Level: Soil Horizon Thicknesses

• • • In Summary: Median values of soil carbon content (%), as well as soil thicknesses per soil series were calculated A method for upscaling from point measurements to terrain unit areas was developed, by using a terrain unit database and calculating mean values based on up to 15 area-weighted soil series per TU Soil carbon content and soil thickness were mapped on a TU level, for topsoil, subsoil, natural vegetation or agricultural land uses

Project 2: Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes Overview: Aim Literature Review • Soil water properties • Soil carbon impacts on soil water properties • Pedotransfer Functions • Knowledge Gaps Methods Some Preliminary Results Next Steps Expected further results

Project: Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes The Aim: To quantify the impact of soil carbon concentrations on hydrological responses in South Africa

Project: Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes • Soil water properties are one of the regulators of water flows Soil water storage is making water available for plant- and other ecological uses, thereby extending water availability into short-term dry periods •

Project: Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes Soil water storage is influenced by • inherent soil texture properties, as expressed by sand, silt and clay fractions • other properties such as soil cracking, fault lines or impermeable layers • soil management factors, e. g. tilling practices in agriculture or use of heavy machinery which might lead to soil compaction

Soil profile characteristics affect hydrological attributes

Soil water storage is also influenced by soil carbon content

• Soil carbon content: is important to soil fertility and soil life • impacting soil water storage While soil texture components are invariants at a location, soil carbon content can be influenced. E. g. : • Some agricultural practices such as ongoing mulching and/or planting of cover crops usually leads to an increase in soil carbon • While the ongoing removal of all plant matter and/or burning practices which leaves soil bare for extended periods usually lead to a marked reduction of soil carbon content

Literature Review: Soil carbon concentrations – Hydrology link (1) • Soil carbon affects soil water retention properties, including • Soil porosity • Field capacity (drained upper limit), at 33 k. Pa suction • Permanent wilting point (1500 k. Pa suction), and hence • Plant available water • (Franzluebbers, 2002; Rawls et al. , 2003; Olness and Archer, 2005; Saxton and Rawls, 2006; Resurreccion et al. , 2011; da Costa et al. , 2013; Ankenbauer and Loheide, 2017) • The above are key inputs into Hydrological Models • 25

Literature Review: [Soil carbon] – Hydrology link (2) The impacts of [soil carbon] changes (assuming biological soil activity, favourable soil p. H and T) depend on • soil texture • original carbon content • soil water content, e. g. being more pronounced at field capacity than at permanent wilting point (Franzluebbers, 2002; Rawls et al. , 2003; Olness and Archer, 2005; Saxton and Rawls, 2006; Resurreccion et al. , 2011; da Costa et al. , 2013; Ankenbauer and Loheide, 2017) • 26

• Literature Review: [Soil carbon] – Hydrology link (3) • Soil water retention curves coloured according to their organic matter content (g/g), based on measurements in soils with similar texture in the USA (Ankenbauer and Loheide, 2017) • 27

Pedotransfer Functions Rawls, Pachepsky, Ritchie, Sobecki and Bloodworth (2003): Equation to estimate water retention at – 33 k. Pa ( Θ 33) and -1500 k. Pa (Θ 1500), both in vol. %, using Clay, Sand Soil Organic Carbon (all in %) as inputs

Project: Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes Literature Review: Knowledge Gaps • What are the effects of reduced (or increased) soil carbon concentrations in soils on hydrological processes and responses in South Africa?

• • • Methodology: Soil carbon - Hydrology Identify selected areas with known soil carbon, clay and sand content, using the ARC soil carbon database Calculate new soil water properties based on clay, sand soil carbon concentrations, using the formulas by Rawls et al. (2003) Calculate soil water properties based on assumed carbon content changes Using ACRU, model the impacts of these changed soil water properties on hydrological responses under natural vegetation and selected crops for selected areas Analyse and map results, e. g. changes in soil moisture, transpiration, runoff, stormflows, baseflows • 30

Results (1) Calculated Change in Drained Upper Limit (mm/m) from 1 to 2% Carbon content change from 2 to 4%

Results (2) Calculated Change in Drained Upper Limit (%) from 1 to 2% Carbon content change from 2 to 4%

Results (3) Calculated Change in Drained Upper Limit (mm/m) to half Carbon content change to double

Results (4) Calculated Change in Drained Upper Limit (%) to half Carbon content change to double

Next steps: Modelling scenarios using • the variables of field capacities and wilting points calculated for the various carbon contents • but leaving other inputs the same, e. g. climate, vegetation and flow routing • 35

Using the physicalconceptual daily time-step ACRU Model (after Schulze, 1995) Quinary Catchments Database, where South Africa, Lesotho and Swaziland have been delineated into 5 838 relatively homogeneous agro-hydrological response units (Schulze and Horan, 2010) • 36

Expected Results It is expected that an increase in soil carbon content, in generally would lead to: • an increase in soil moisture, transpiration, crop yield and flow attenuation • a reduction in runoff, especially stormflows • a reduction in soil loss and sediment yield

New Knowledge Creation Envisaged • A better understanding of soil carbon effects on hydrological processes under natural vegetation and selected agricultural crops • Maps of modelled hydrological responses of effects of reduced and increased soil carbon concentrations for selected areas in South Africa for natural vegetation and selected agricultural land uses • 38

Project: Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes In Summary (1): • • Sampled areas with measured soil carbon, clay and sand content from the ARC soils database were used (from Project 1) New soil water properties (Field Capacity and Wilting Point) based on clay, sand soil carbon concentrations were calculated, using the formulas by Rawls et al. (2003) Soil water properties based on assumed carbon content changes were calculated These were mapped on TU level, using the upscaling methods from Project 1

Modelling the Effects of Soil Carbon Concentrations on Hydrological Processes In Summary (2): • Expected Results were presented Next Steps: • Using ACRU, the impacts of these changed soil water properties are to be modelled • The results are to be analysed and mapped, e. g. changes in soil moisture, transpiration, runoff, stormflows, baseflows

Acknowledgements Our colleagues from the CWRR Project 1: • GIZ for supporting the soil carbon mapping project • ARC (Agricultural Research Council) for the use of their soil carbon database • Hein Beukes for the use of the Terrain Unit Database Project 2: • SASAC (Southern African Systems Analysis Centre) for Stef’s Doctoral Scholarship • Prof Mary Scholes, University of Witwatersrand (Stef’s Co-Supervisor) • WRC, results will be reported on in the K 5/2833 Project • • 41

Thank you for listening Stefanie Schütte E-Mail: Schuttes@ukzn. ac. za Prof Roland Schulze E-Mail: Schulzer@ukzn. ac. za SANCIAHS 18. 9. 2018 INSPIRING GREATNESS