RECLAMATION OF A SALINE SODlC SOIL IN THE NKWALINI VALLEY By M. A. JOHNSTON South African Sugar Association Experiment Station, Mount Edgecombe

An area of saline sodic soil was intensively drained and treatments of gypsum (31 t/ha) or sulphur (6 t/ha) were applied, while control plots received no ameliorants. Physical and chemical soil analyses showed gypsum to have an ameliorative effect slightly superior to that of sulphur. Both treatments were more beneficial than the control but differences were not always as great as expected. For the plant and first ratoon sugarcane crops grown on the experiment, average yields were 100, 99 and 82 tc/ha for the gypsum, sulphur and control, respectively. Introduction Excessive salt levels in the soil have for some years been recognised as a problem in the dry, irrigated regions of the South African sugar industry (Maud3, von der Meden7). The saline sodic (saline alkali) condition is by far the most common, but saline non-sodic and non-saline sodic soils are also found. Salinisation of soils in the sugarcane areas is generally caused by improper water management, ie inadequate drainage is provided for irrigation excesses and this results in a rise in the level of the water table. Salts dissolved in the groundwater reach the soil surface by capillary movement and then accumulate there. It is usually soils of naturally poor drainage that are affected in this way, and in many instances these are marginally affected by salts in the virgin state. In reclaiming sodic soils, improvement or maintenance of soil permeability is the prime consideration. This is usually achieved by supplying calcium ions to the soil, either directly by adding gypsum (calcium sulphate) or indirectly by acidifying the soil with sulphur or sulphuric acid to dissolve free calcium carbonate. Techniques of reclamation have been thoroughly discussed in the literature (US Salinity Lab Staff6; Szabolcs5; Kovda, van den Berg and Hagen2) but, because of a general lack of local practical experience, it was decided that it would be worthwhile, as a first step, to test commonly used ameliorants in a small field trial. Selecting a site for reclamation work presented difficulties since, in such soils, marked variations in salt level frequently occur over very short distances. Nevertheless, a site was selected for an experiment in the Nkwalini Valley (Zululand) which represented a typical example of a saline sodic condition. The climate in this area is described as semi-arid, the mean annual rainfall being approximately 700 mm, so that intensive irrigation is required to achieve satisfactory sugarcane yields. Experimental procedure The soil is of the Nyoka series (Swartland form) and consists of a very dark grey brown porous sandy clay loam (f 25% clay) about 36 cm in depth, overlying a B horizon of dark to yellowish brown, slowly permeable sandy clay (f 40% clay). The transition between A and B horizons is gradual and the profile depth generally exceeds 0,9 metres. The underlying material consists of water worn gravel and weathering Middle Ecca sediments, the latter being the dominant parent material in this area. The highly saline sodic condition that exists (see Table 1) is believed to have arisen primarily from a high water table which, in turn, was caused by seepage from an unlined dam up the slope from the area. At the time when the experiment was laid down, much of the soil was in a highly dispersed state, with little plant growth before t,reatment (Fig. 1). TABLE 1 Average chemical status of the soil prior to treatment Exch. SARse Na (mes %) FIGURE I Appearance of the experiment site prior to application of treatments. Three replications of the following treatments were tested: Drainage alone as control ( c ) Drainage + gypsum at 3 1 t/ha (G) Drainage + sulphur at 6 t/ha (S) The gypsum (CaS0,.2H,O) requirement was determined from the exchangeable Na level of the worst affected areas of the site. The aim was to displace the exchangeable Na in excess of 2,O meq/100 g with Ca. An amount of sulphur that was chemically equivalent to the gypsum was used. The ameliorants were of fine texture, 90 % and 100 % of the gypsum and sulphur, respectively, passing a 100 mesh sieve, and both were at least 98 % pure. As the salt status of the soil varied quite markedly across the site, it was considered necessary to group together the three worst, the three intermediate and the three least affected plots. One of each of the three treatments was allocated at random to a plot in each group to minimise bias in the results towards any specific treatment. In December 1970, flexible plastic pipe drains (50 mm diam.) were installed at a spacing of 8,5 mebres and an average depth of about 0,85 metres. Ten parallel drains formed the boundaries of the nine plots and these extended down the 36 metre