Improvement of dry tropical rangelands on Hainen Island, China. 3. Legume response to initial fertilizers.

Abstract

During 1981-83, we identified nutrient requirements and fertilizer strategies for the development of legume-based pastures in dry tropical rangelands of Hainan Island, China. Nutrient requirements for pastures grown on sandy and red loam soils were established using soil tests and missing element experiments with 3 test legumes. An acute P deficiency (< 5 mg kg-1 Bray P) was detected by soil tests on sandy soils, but no deficiencies were found on loam soil. The exclusion of P from a total fertilizer treatment reduced relative yield to 20% of potential on sandy soil. A factorial P rates experiment replicated in space and time showed a strong curvilinear response for stylo (Stylosanthes hamata L. Taub. cv. Verano) and siratro (Macroptilium atropurpureum (DC.) Urban) based pastures on sandy soil. No response was observed on loam soil. Given limited superphosphate supplies available for range improvement in south China, results from these studies showed no benefit from applying P fertilizer to stylo pastures grown on soils with available P > 20 mg kg-1. In contrast, application of P fertilizer proved both profitable and essential for development of legume-based pastures on P-deficient sandy soils. It is recommended that a rate of about 40 kg of P ha-1 should be applied at establishment of stylo pastures on sandy soil. At this rate both legume yield and P content of tops should be maintained at levels needed to sustain a beef production enterprise and return about $1.5 per dollar invested in fertilizer.