EVAPOTRANSPIRATION ETC., WITH IRRIGATED GRAIN SORGHUM 



11 



The soil moisture extraction pattern was evalu- 

 ated during sampling periods when little or no 

 rainfall occurred. The results indicated that as 

 the season progresses from July 15 to September 

 15, the percentage of soil moisture obtained from 

 the top foot during a sampling period decreased 

 from about 50 to 35 percent. The percentage of 

 extraction from the other depths to 4 feet in- 

 creased during this period. 



Delaying irrigations until small amounts of 

 available water remained in the upper layers of 

 soil decreased total seasonal E t , but yields de- 

 creased by a greater proportion. This relation was 

 evaluated by considering the average seasonal 

 E t on the F 4 and F 5 fertilizer plots of the M 4 

 moisture level as E t with optimum soil moisture 

 (E tu ). The yield on the F 5 fertilizer treatment 

 on the M 4 moisture level was used as the maximum 

 (7 mai ). The data presented in figure 5 are mean 

 relative yields (F/F max ) on F 4 and F 5 fertilizer 

 plots and mean relative E t (E t /E t0 ) where adequate 

 nitrogen appeared to have been provided. Since 

 the time of occurrence of low moisture conditions 

 on different treatments was not always at the same 

 stage of growth even within one year, some scatter 

 of points is to be expected. However, the general 

 trend indicated yields decreased more rapidly 

 than E t . For example, if irrigations were delayed, 

 causing a 10-percent reduction in seasonal E t , 

 yields were reduced about 20 percent. Likewise, 

 a 20-percent reduction in E t reduced yields about 

 35 percent. The same type of relation would not 

 necessarily occur if yields included total dry matter 

 produced. 



RELATIVE E T (E T /E TMAX ) 



Figure 5. — Relative yield decreased more rapidly than 

 relative seasonal evapotranspiration when irrigations 

 were delayed. 



Larger yields on the nitrogen-fertilized subplots 

 in 1957-59 resulted in lower average soil moisture 

 than on the 0-nitrogen subplots. The mean soil 

 moisture percentages for sampling periods from 

 July 15 to September 15 for each depth are pre- 

 sented in table 15 in the appendix. These 

 values are averages of 64 to 112 soil samples. 

 Average values cannot be used to compare mois- 

 ture levels directly, because different sampling 

 dates were involved. Differences in mean sou 

 moisture in the 0- to 4-foot depth between the F 2 

 plot and the average of F 4 and F 5 plots are plotted 

 in figure 6 against differences in yield between the 

 two. 



These results indicate that if large yield or 

 plant-growth differences are expected between 

 treatments and if soil moisture level is an impor- 

 tant factor, experiments should not be designed 

 with subplot treatments having large yield or 

 plant-growth differences. The mean soil moisture 

 percentage on the nitrogen-fertilized plots aver- 

 aged 0.9 percent (on an oven-dry weight basis) 

 less than the 0-nitrogen plots when yields on 

 the nitrogen-fertilized plots were 4,000 lb./acre 

 greater. This difference represents about 10 per- 

 cent of the total available soil moisture. The 

 difference in soil moisture tension in the upper 

 soil layers just before irrigating may have reached 

 several atmospheres, especially on treatments 

 where the soil moisture tension was allowed to 

 reach 4 or 9 atmospheres before irrigating. 



Grain Quality 



Soil moisture and nitrogen fertilizer affected the 

 quality of the grain as well as the yield. Test 

 weight and protein content were used to evaluate 

 quality. 



Jest Weight 



A summary of test weights for all soil moisture 

 and fertilizer treatments and an analysis of vari- 

 ance are presented in table 16 in the appendix. 

 The low 4-year average test weights on the Mi 

 moisture level were due primarily to greater 

 lodging in 1958 as rate of applied nitrogen 

 increased. 



The low values on the M 2 treatment were due 

 primarily to lodging in 1956. Lodging and low 

 test weights appeared to have been more severe 

 when soil moisture was adequate early in the 

 season but deficient during the latter part of the 

 season. 



Protein 



Protein content of grain was inversely related 

 to the level of production and directly to the 

 amount of nitrogen fertilizer applied (appendix 



