The influence of land use on humic substances in three semi-arid agro-ecosystems in the Free State

English: This study was initiated to complement earlier investigations into soil organic matter degradation and restoration on account of agricultural land use in the Free State Province of South Africa. In these studies no attention was given to the response of humic substances which represent the most active fraction of organic matter. The aim with this study was therefore to quantify the influence of agricultural land use on humic substances in soils of semi-arid regions. Topsoil (0-200 mm) samples from distinctive agro-ecosystems at Harrismith (Mean annual rainfall, MAR = 624 mm and Mean annual temperature, Ta = 13.8°C), Tweespruit (MAR = 544 mm and Ta = 14.8°C) and Kroonstad (MAR = 566 mm and Ta = 16.6°C) were selected for use in this study. An agroecosystem implies a region where the three environmental factors affecting yield, namely climate, slope and soil are for practical purposes homogeneous. The selected samples represent a virgin (grassland soil never cultivated before), cultivated (formerly grassland soil cultivated for at least 20 years) and restored (formerly cultivated soil converted to perennial pasture for at least 15 years) Plinthustalfs (10.6 to 13.5% clay) at every agro-ecosystem. Parameters quantified comprise crude soil, extractable soil, humic acid and fulvic acid C contents, N contents and C/N ratios. Concerning these parameters, cultivated soil was compared with virgin soil and restored soil with cultivated soil. The crude soil C content of the virgin soils varied from 7.3 g C kg-1 soil in the warmer, drier Kroonstad agro-ecosystem to 21.6 g C kg-1 soil in the cooler, wetter Harrismith agro-ecosystem. Across agro-ecosystems the contribution of extractable to crude soil C was almost constant, namely 47.1 to 48.4%. The contribution of humic acid C to extractable soil C decreased and that of fulvic acid C to extractable soil C increased from the Kroonstad to Harrismith agroecosystem. Cultivation reduced crude soil C in the three agro-ecosystems with 50.2 to 51.8%. This is equivalent to absolute losses of 3.8, 8.2 and 10.8 g C kg-1 soil at Kroonstad, Tweespruit and Harrismith agro-ecosystems respectively. Loss of extractable soil C was more variable ranging from 36.7% or 1.3 g C kg-1 soil in the warmer, drier Kroonstad agro-ecosystem to 48.2% or 5.1 g C kg-1 soil in the cooler, wetter Harrismith agro-ecosystem. Trends of this nature were nonexistent for either humic or fulvic acid C losses. Gains of crude soil C ranged from 5.4 g C kg-1 soil in the warmer, drier Kroonstad agro-ecosystem to 8.0 g C kg-1 soil in the cooler, wetter Harrismith agroecosystem. This trend manifested also in extractable soil C gains which were lowest at Kroonstad (1.5 g C kg-1 soil) and highest at Harrismith (2.8 g C kg-1 soil). Neither humic acid C nor fulvic acid C showed trends of this nature. The N contents although more variable than the C contents are to a large extent supportive concerning organic matter in the virgin, cultivated and restored soils of the three agro-ecosystems. Further elaboration on the N contents is therefore not justified here. Based on both C and N indices, it can be stated that humic substances did not show explicit trends on account of land use as was the case with organic matter per se. This phenomenon warrants further investigation since humic substances are regarded as the most reactive fraction of organic matter.