Sorghum is a very adaptable crop. It is quite tolerant to drought stress and can still potentially give high grain yields, but the actual average grain yields are often low because the farmers grow sorghum in the late rainy season when the rainfall is inadequate. To increase the value of sorghum, its stalk may be used as forage for ruminant animals. In 1992 the variety Suphanburi 1 was recommended as a dual purpose sorghum. Suphanburi 1 has a good grain and forage yield in the regular season but it gave a lower forage yield in summer. The objective of this program is to improve new sorghum varieties which have higher grain yield and stem weights with better forage quality than Suphanburi 1 and methods of increasing forage yield in all seasons. The evaluation of forage and grain yield in 1993 and 1994 showed that there were some lines which had 12-71 % greater fresh stem weight and 2-64 % greater grain yields than Suphanburi 1. The lines which had the highest stem weight, were not the same lines that gave the highest grain yield.

The objective of this study was to evaluate stem yield in plant crop and grain yield of 9 sorghum lines. The selected sorghum lines were grown at 6 locations in 4 row plots, six meter long and 60 cm wide with 4 replications from May to December 1995. The results showed that all the tall plant sorghum lines gave more stem weight than the short plant lines. The line IS 23585 gave significantly higher stem weight than Suphanburi 1 in all locations. The grain yield in ratoon crop was highly variable.

In 1996, 22 sorghum lines were evaluated in RCBD with 3 replications. Eight lines produced greater grain yields than the check variety, Suphanburi 1 but they did not differ statistically. Yields of ICSV 93046, A 2267-2 and SPLB 94022 were 6.81, 6.74 and 5.41 t/ha respectively which were significantly greater than Suphanburi 1. Also, their height and diameter were significantly greater than Suphanburi 1. Seed germination of all lines were lower than Suphanburi 1.

A number of ICRISAT multicut forage sorghum (39 lines) were grown in 6 m long rows at 50 x 20 cm and evaluated for their fresh stalk yield. The sorghum was harvested 56 days after flowering. Among the lines tested, there were 12 lines that produced yields 5-61 % greater than Suphanburi 1 named as GD 47832, GD 47831, GD 47809, GD 47841, GD 47828, GD 47816, GD 47842, GD 47815, GD 47803, SSG 59-3 (N), GD 47843, GD 7838. Tillering of these lines was similar to that of Suphanburi 1.

Two dual purpose lines of sorghum were compared to Suphanburi 1 when under different fertilizer regimes. Three rates of nitrogen 0, 62.5 and 125 kg N/ha were used and managed by means of split applications at different proportions for planted and ratoon sorghum compared with unsplit applications. The results showed that there were no significant differences among sorghum lines either between the fresh stem weight or grain yield. It was also found that there were no significant differences between rates of nitrogen fertilizer on fresh stem weight or grain yield. The methods of nitrogen fertilizer application also, did not show any effects on them. However it was noted that the percentage of nitrogen in the leaves of planted sorghum increased when nitrogen fertilizer was applied to them.

The amount and distribution of rain was of most benefit to sorghum planting in Southern Region. Planting time was from the beginning of May to the beginning of September. Sorghum planted between the beginning of May to the beginning of June gave more fresh stalk weight (for both plant and ratoon) than the others and planting from mid-June to the beginning of July gave the lowest fresh weight. Suphanburi 1 tended to give 14.0 %, more yield, than UT 23585 but this was not significantly different.

Variable rates of Nitrogen at 62.5 kg and fixed rates of K2O5 and P2O5 (62.5 kg/ha) per hectare were applied to plant sorghum. Two rates of 62.5 and 125.0 kg of N per hectare were applied to plant and ratoon with (0, 0) (62.5, 0) (31.25, 31.25) (125.0, 0) (62.5, 62.5) (93.75, 31.25) (31.25, 93.75) kg/ha systems. In two crops, plant and ratoon showed non-significant differences in fresh stalk and grain weight. Higher N rates could increase % N by weight in the leaf blade, leaf sheath and stem. N by weight in leaf blade was more than the % N by weight in stem and leaf sheath and showed significantly different. Total % N by weight of UT408-1 and Suphanburi 1 were not different but they were significant different from IS 23585.