26 Better Crops/Vol. 96 (2012, No. 2) Common abbreviations and notes: N = nitrogen; P = phosphorus; K = potassium; C = carbon; RDF = recommended dose of fertilizer; GR = Gross returns; NR = net returns; B:C ratio = beneft:cost ratio; M = million. South ASiA I ndia has the largest area in the world under cotton at 10.1 M ha and is the second largest producer in the world at 31 M bales. However, India’s average cotton productivity is 478 kg lint yield /ha, combining both irrigated and rainfed felds, and this is low compared to other countries like China (1,311 kg lint/ha), Brazil (2,027 kg lint/ha), U.S. (945 kg lint/ha) as well as the world average yield of 763 kg lint/ha (ICAC, 2011). Rainfed cotton occupies 7 M ha (70%) with an average productivity of 230 kg/ha in India. The majority (90%) of cotton in the State of Maharashtra is rainfed, and this area is expected to increase in the coming years. A system of high density planting (HDP) leading to more rapid canopy closure and decreased soil water evaporation, is becoming popular to address water scarcity challenges. In many countries, narrow row plantings have been adopted after showing improvement in cotton productivity (Ali et al., 2010). The adoption of HDP, along with good fertilizer management and better genotypes, is a viable approach to break the current trend of stagnating yields under primarily rainfed hirsutum (upland) cotton growing areas. Experimental Sites and Design Four sites (Nagpur, Akola, Parbhani, and Nandyal) were selected during the kharif (rainy) season of 2010-11. The com- mercial varieties grown at these sites included NH 615, NH 452, and PKV 081 at Nagpur, AKH 081, NH 615, and NH 630 at Akola, NH 545, NH 452, and AKH 081 at Parbhani, and NDLH 1938 at Nandyal. The varieties were sown at the onset of the monsoons during the last week of June at Nagpur and Akola, and in the second week of July at Parbhani and Nandyal. Each sub-plot contained ten rows of cotton, 5.4 m long. A plant spacing of 0.45 m x 0.15 m produced a high density population of 14.9 plants/m 2 . The traditional plant population is 5.5 plants/m 2 . In Maharashtra and Andhra Pradesh, the climate is hot, dry, and sub-humid. Soils are dominated by Vertisols and Ver- tic intergrades. Soils at the experimental sites were medium deep black, slightly alkaline (pH 7.8 to 8.4), low-to-medium in available N and P, and high in K status (Table 1). A split-plot design with three replications was used with the main plots being three upland varieties and the sub- plots being four levels (75, 100, 125, and 150% RDF) of NPK fertilizers. Data were collected for seed cotton yield, yield components, and biological yield. Total heat unit dur- ing the crop period was estimated on the basis of weekly maximum and minimum temperatures using this equation: Total Heat Unit=(Max.Temp+Min.Temp)÷2 - Base Temp. (15 °C) Results and Discussion Seed cotton yield varied greatly among locations (Table 2). Mean seed cotton yield recorded at Nandyal was 13% more than the yield obtained at Akola and >70% higher than yields obtained at Nagpur and Parbhani. From the last week of June through the last week of December (the growing period of cotton), the observed average maximum and minimum tem- peratures were 30.2 and 21.1 °C at Nagpur, 32.2 and 21.2 °C at Akola, 31.1 and 20.3 °C at Parbhani, and 31.6 and 22.8 °C at Nandyal, respectively. Precipitation was high in the months of July and August at Nagpur and Parbhani, which reduced the growth and yield components. Precipitation during the period from fowering to early boll development (65 to 100 days after planting) is crucial for rainfed cotton. This amount was 339, 225, 222 and 198 mm at Parbhani, Akola, Nagpur, and Nandyal, respectively. During the crop period, Nandyal recorded higher heat units (360.5 °C) as compared to Akola (350.5 °C) and Parbhani (324.1 °C), which might have led to the differences in yields among locations. No signifcant difference in seed cotton yield was found between the three varieties used at Nagpur, Akola, and Parb- hani. However, variety NDLH 1938 recorded highest yield with the highest bolls per square meter (139.8 BPM) compared to NH 630 (125.5 BPM) under HDP. Seed cotton yield and yield components like plant height, bolls, sympodia, and biomass of upland cotton were increased by 125% NPK fertilization at all four locations. Seed cotton yield recorded with 125 and 150% RDF were signifcantly higher than that with 75% and 100% RDF. This indicates that fertilizer requirement is most likely to be higher under HDP (Jost and Cothren, 2000; Ali et al., 2007). Interaction effects between main plot and sub plot treatments were not signifcant for yield and BPM at all the locations. Total nutrient (N, P, and K) uptake per hectare was higher By Jagvir Singh, Shilpa Babar, Shalu Abraham, M.V. Venugopalan, and G. Majumdar Despite large tracts of irrigated cotton, rainfed systems remain the most important op- tion for improving cotton production in India. Within rainfed fields, the potential effects of adopting high plant population with adequate NPK fertilizer management offer a good opportunity to increase crop productivity. Fertilization of High Density, Rainfed Cotton Grown on Vertisols of India Table 1. Physiochemical properties of selected soils and their respective fertilizer recommendations (100% RDF). Particulars Nagpur Akola Parbhani Nandyal Clay, % 153 148 154 150 Available N, kg/ha 111 218 116 127 Available P 2 O 5 , kg/ha 118 123 126 187 Available K 2 O, kg/ha 481 365 675 591 Organic C, g/kg 13.5 13.7 14.1 14.0 100% RDF, N:P:K 60:13:26 50:20:0 84:18:36 40:9:18 Available N, P 2 O 5 , and K 2 O values reported were determined using the Kjeldahl, Olsen, and 1 N ammonium acetate extraction methods, respectively.