SOIL USE AND MANAGEMENT zyxwvu kolume 2, Number 1, March 1986 3 Soil acidification during more than 100 years under permanent grassland and woodland at Rothamsted A. E. Johnston, K. W. T. Goulding zyxw & P. R. Poulton* Abstract Soil samples have been taken periodically from unlimed plots of’ the 130-year-old Park Grass Experiment and from the 100-year-old Geescroft Wilderness at Rothamsted. Changes in the pH of the samples show how acidification has progressed. The soils are now at, or are approaching, equilibrium pH values which depend on the acidiflmg inputs and on the buffering capacities ofthe soils. We have calculated the contributions to soil acidification of natural sources of acidity in the soil, aimospheric deposition, crop growth and nutrient removal, and, where applicable, additions of fertilizers. The relative importance of each source of acidification has changed as the soils have become more acid. Acid rain (wet deposited acidity) is a negligible source, but total atmospheric deposition may comprise up to 30% ofacidifylng inputs at near neutral soil pH values and more as soil pH decreases. Excepting fertilizers, the greatest causes ofsoil acidification at or near neutral pH values are the natural inputs ofH’ from the dissolution ofCO, and subsequent dissociation of carbonic acid, and the mineralization ot‘organic matter. Under grassland, single superphosphate and small amounts of sodium and magnesium sulphates have had no effect on soil pH, whilst potassium sulphate increased soil acidity slightly. All of these effects are greatly outweighed under grassland, however, by those of nitrogen fertilizers. Against a background of acidification from atmospheric, crop and natural inputs, nitrogen applied as ammonium sulphate decreased soil pH up to a maximum of 1.2 zyxwvuts units at a rate in direct proportion to the amount added, and nitrogen applied as sodium nitrate increased soil pH by between 0.5 and 1 unit. HE Park Grass Experiment at Rothamsted, started in T 1856 by Lawes and Gilbert, examines the effect of various manurial treatments on permanent grassland more than 300 years old. Results were summarized by Thurston zyxwvu et zyxwvutsrqp al. (1976) who also listed previously published data. The experimental site is flat and is never ploughed. Botanical composition of the sward has been considerably altered by both soil pH and manurial treatment. The remarkable sharpness of visible differences at the boundaries between adjacent plots suggests that there has been little sideways movement of nutrients in drainage. The Geescroft Wilderness was part of an arable field on the Rothamsted estate map on 1623. Its total N content in 1883 suggests that it had continued mainly in arable crops. Fenced off and left unattended in 1886, it developed into deciduous woodland. Jenkinson (1 97 1) lists the references which document this change, and also gives some soil analysis data. This site is also flat. When under arable crops it was probably in ridge-and-furrow and, in 1849, was hle drained in the bottom of the furrows to aid autumn culti- vations, but the drains no longer run. T h e soils from both sites have been sampled periodically. From the changes in their pH with time, and from soil, plant and rainfall analyses, the contributions to soil acidification have been estimated of (i) soil-derived natural sources of acidity, (ii) atmospheric deposition, (iii) nutrient uptake by plants, and (iv), in the case of the Park Grass, fertilizer input. *Soils and Plant hutrition Uepamnent, Rothamsted lixperimental Station, Harpenden, Herts h15 2JQ Whilst much of (ii) might be considered ‘natural’ relative to (iv), we have separated it from soil-derived natural sources (i) for ease of presentation. zyx All four inputs are defined below. METHODS zyx Soils The samples of Park Grass soils were taken from plots which have been unlimed except for small dressings of chalk and slaked lime during 1881-96. Fertilizers are applied annually, and manuring on the plots considered here has remained largely unchanged since 1856, or 1858 in the case of those receiving sodium nitrate. Relevant details are given in Table 1; zyxw full details were given by Warren 8i Johnston (1964). One crop of hay made in June, and one aftermath, cut in October, are removed each year. Geescroft soils have been unlimed and unfertilized and no crop has been removed since 1885. Soil pH On Park Grass, the p H values of the surface soils (0-23 cm) and subsoils (23-46 cm) were first measured by Crowther in 1923, using the hydrogen electrode method and a 1:5 soil :water suspension. The next measurements were made on samples taken in 1959, using a glass electrode and a 1 : 2.5 soil :water suspension (Warren & Johnston, 1964). The different soi1:water ratios were shown to have little effect on measured pH. Further samples were taken in 1976