A 3-year field experiment beginning in 1995–96 was undertaken on a Gangetic alluvial soil (Typic Ustochrept) at Modipuram, India, to examine how phosphorus (P) demands of rice–wheat cropping systems might be met with heavy initial dressings of phosphate rock. Treatments were: (i) 1 rice–wheat cycle where P was applied as phosphate rock to each successive rice crop (54 kg P/ha i.e. 27 kg P/ha to rice and 27 kg P/ha to wheat); (ii) 2 rice–wheat cycles where P was applied as phosphate rock to alternate rice crops (108 kg P/ha); and (iii) 3 rice–wheat cycles where P was applied as phosphate rock to the initial rice crop (1995–96) (162 kg P/ha). The performance of these treatments was compared with those that received fertiliser P at the recommended rate of 27 kg P/ha for each crop as diammonium phosphate or phosphate rock, and with a nil-P (control). In phosphate rock treatments, rice and wheat were established with or without an inoculation of the P-solubilising microorganism Aspergillus awamori. Compared with nil-P, fertiliser P applied as diammonium phosphate increased rice and wheat yields significantly in different years. Equivalent P levels, applied as phosphate rock, did not increase the yields of either crop significantly over nil-P plots. Application of 54 kg P/ha as phosphate rock to every rice crop, however, resulted in significantly greater yields of rice compared with those that received nil-P or 27 kg P per crop as phosphate rock, but the subsequent wheat yields suffered due to P stress. A further increase in P application rate as phosphate rock to the initial rice crop (1995–96) increased the magnitude of yield response compared with lower rates of initial application. The residual effect in subsequent crops varied in accordance with the amount of P applied. Pre-plant inoculation of rice seedling-roots or wheat seeds with P-solubilising microorganism led to a yield increase over non-inoculated treatments of 0.09–0.22 t/ha in rice and 0.15–0.45 t/ha in wheat, in different years. The agronomic efficiency and recovery efficiency of fertiliser P in the rice–wheat system were highest (57.2 kg grain/kg P and 40.4%, respectively) under diammonium phosphate-fertilised treatments. The agronomic efficiency ranged from 14.3 to 44.4 kg grain/kg P and the recovery efficiency ranged from 7.7 to 26.4% for phosphate rock treatments. The agronomic efficiency and recovery efficiency increased with increasing initial phosphate rock application rate and with P-solubilising microorganism inoculation. Economic returns (per rupee invested in fertiliser P) were greater (Rs.9.46–12.95) under heavy initial phosphate rock application (108 or 162 kg P/ha), compared with those (Rs.8.60) under diammonium phosphate-fertilised treatments. Overall, we recommend that, in rice–wheat cropping systems on neutral soils when phosphate rock is the available fertiliser, heavier application every 2–3 years should be considered over application to every crop. Outcomes would be marginally improved with P-solubilising microorganism inoculation.
Does the inverse farm size productivity hypothesis hold for perennial monocrop systems in developing countries? Evidence from Kenya
