Sodium leaching efficiencies (moles of Na removed per unit leachate volume) were measured and compared from four noncropped and four cropped treatments applied in duplicate to 1.0 m deep sodic calcareous silt loam in lysimeters. Treatments included a check, gypsum, chopped alfalfa (Medicago sativa L.), fresh manure, alfalfa, sorghum (Sorghum bicolor), Sudan grass (Sorghum Sudanese) hybrid (which will be called sorghum hybrid for simplicity), sorghum hybrid + leaching, and cotton (Gossypium hirsutum L.). The sorghum hybrid + leaching treatment soil was leached with tap water until 0.5 pore volume of leachate was collected from lysimeter bottoms, and then sorghum hybrid was planted. Sorghum hybrid was the most efficient treatment in reclaiming Na-affected soil. All four non-cropped soils eventually became dispersed in the lower part of the profile and hydraulic conductivity became very low. Cropped treatments continued to conduct water at a satisfactory rate for reclamation; however, due to low water use, cotton treatment produced a low total Na removal. Sorghum hybrid shows promise as a crop that could be used to speed reclamation of sodic calcareous soils. The treatments producing the highest sodium removal efficiencies also produced the highest soil atmosphere CO² concentrations. By selecting crops, amendments, and water application rates and timing, calcareous sodic soil reclamation can very likely be accomplished faster and more economically than in the past.

Nitrogen transfer between intercropped legumes and nonlegumes has not been consistently documented. However, N transfer between mixtures of nodulating (Rj₁) and nonnodulating (rj₁) soybeans (Glycine max (L.) Merr.) has been shown. A 2-yr field study was conducted to examine the effects of intercropped nodulating and nonnodulating soybeans on sorghum [Sorghum bicolor (L.) Moench] yield and N yield, and N transfer. In both years, the soybean cultivar (cv.) Clark 63 (Rj₁) and its rj₁ isoline, L73-1054, were planted at different distances between sorghum (‘DeKalb BR45+’) rows. Nitrogen rate was either 0 to 100 kg N/ha in 1979 and either 0 or 175 kg N/ha in 1980. The soil was a Cisne silt loam (mesic Mollic Albaqw in both years. Monocultures of both crops were also planted in both years. Sorghum N yield was 9% greater with Rj₁ than with rj₁ in 1979. This was due to increased N yield of sorghum associated with Rj₁ in 0.4 m rows; soybean cv. had no effect on sorghum N yield in 0.8 m rows. Sorghum yield was not affected by soybean CV. Yields and competition of rj₁ soybeans in 1980 were severely reduced by the hotdry growing season confounding N-accumulation differences between rj₁ and Rj₁. However, from the 1979 results, we conclude that sorghum N accumulation was increased by association with Rj₁ soybeans through N transfer, and that closer proximity of sorghum and soybeans enhanced this beneficial effect.

Nitrogen requirements and utilization of mineral nitrogen (N) by sorghum and groundnut were compared. At the maximum N use level, sorghum genotypes showed greater N use efficiency (120 kg biomass/kg N harvested) than groundnut genotypes (36 kg biomass/kg N harvested). Using a non-nodulating groundnut genotype (Non-nod) or sorghum as controls for soil N uptake, the amounts of N2 fixed by the nodulated groundnut genotypes were estimated to be 183–190 kg N/ha. Nitrogen fertilization increased harvest index and percentage N translocated to seeds in sorghum genotypes, but decreased harvest index and had variable effects on percentage N translocated to seed in groundnut genotypes. Leaf nitrate reductase activity (NRA) and nitrate content in the leaves of two sorghum genotypes, one nodulating, and ‘Non-nod’ groundnut genotypes were also compared. The concentration of nitrate was lower in sorghum than in groundnut leaves, but NRA was higher in sorghum. It is suggested that either NRA in the groundnut leaves has relatively lower affinity for the substrate (higher Km, the Michaelis-Menton constant) or higher nitrate is required for the induction of nitrate reductase in groundnut than in sorghum. This implies that groundnut is a poor utilizer of fertilizer nitrogen.

High ratio cakes baked from composite flours of hard red winter wheat and grain sorghum were inferior to those baked from wheat flour alone. The lipid, water-soluble and starch fractions of grain sorghum were interchanged with those of wheat to identify the responsible component(s). Sorghum lipids did not display functionality, perhaps due to their low concentration of glycolipids. The effect of exchanging water-solubles on cake quality was negligible. Both volume and texture were inferior when sorghum starch replaced wheat starch. Replacement of sucrose with dextrose greatly improved cake volume and texture, apparently by lowering the high gelatinization temperature of the sorghum starch.

The analysis of cost and return from sorghum production were divided into two parts: the first part attempted to compare cost and return from the production of white and red sorghum varieties. Cost of production of white sorghum was 406.50 baht per rai with the gross return of 546.22 baht, and a net profit of 139.22 baht per rai. For red sorghum, on the other hand, cost of production was 621.07 baht per rai with the gross return of 721.44 baht with a slightly low net profit of 100.37 baht per rai. The main difference in net profit was mainly due to lower seed cost and higher grain price of white sorghum. The second part dealt with the comparison of cost and return of red sorghum and cassava production under 'Cassava Acreage Reduction Program'. Cost of hybrid sorghum production was 628.42 baht per rai with the gross return of only 155.54 baht per rai; a loss of 428.88 baht per rai. Cost of cassava production was 833.74 baht per rai with the gross return of 767.00 baht or a net loss of 66.74 baht per rai. Red sorghum gave more risk in this instance and it was doubtful to use it in the substitution program unless other better solutions or measures are found and wisely applied.

The revised sorghum simulation model, SORGF, was used to demonstrate the applications of crop simulation models in the semi-arid tropics. Comparison between simulated and observed grain yield data showed that the SORGF model can be used to estimate sorghum yields with reasonable accuracy before harvest. Responses of sorghum to drought-stress and to changes in plant density were simulated. The correlation coefficient between observed and simulated sorghum grain yield data pooled over 5 levels of plant density and 2 cultivars was 0.91. The correlation coefficient between observed and simulated sorghum grain yield data pooled over 2 water treatments, 2 cultivars, 2 seasons was 0.81. The model was used to compute the probabilities of simulated available soil water at the end of rainy season, runoff amounts, sorghum grain yield and the requirements of N-fertilizers based on 30 years of climatic data for 4 locations in India