Recovery of nitrogen from four fertilizers applied to Rhodes grass in small plots

Abstract

The response of Rhodes grass to nitrogen applied as sodium nitrate, ammonium sulphate, urea, or ammonium nitrate-limestone at rates of 56, 112, 224, and 448 kg N ha-1 year-1 was measured in small plots over seven years at the Samford Pasture Research Station. The ranking order of mean yields of dry matter and nitrogen at the fertilizer rates where Rhodes grass remained dominant (the 224 and 448 kg rates) was sodium nitrate> ammonium nitrate-limestone> urea. At these rates ammonium sulphate was about as effective as sodium nitrate, until the comparison was confounded by changes in soil pH. The comparative efficiency of urea varied widely between harvests. Taking the weighted mean for the 224 and 448 kg rates over seven years, urea gave 87 per cent of the nitrogen yield that sodium nitrate gave. On plots receiving 0 to 112 kg N, Rhodes grass was replaced progressively by other species, chiefly Digitaria spp., ,Axonopus affinis, and Glycine tabacina. Negative apparent nitrogen recoveries recorded from the 56 and 112 kg N treatments in the seventh year were attributed to a bias caused by nitrogen fixation by native legumes. Nitrogen fertilization caused a significant increase in the nitrogen content of soil and roots (0-15 cm) at the end of the experiment. The increase was significantly smaller with ammonium sulphate than with other forms. About 20-40 per cent of the fertilizer nitrogen added over seven years could not be accounted for in top growth or in soil plus roots. A microplot study with 15N-labelled ammonium- and nitrate-N in the seventh year showed an apparent loss of added 15N, which increased with fertilizer rate, of from 5 to 43 per cent.