44 MISCELLANEOUS PUBLICATION 95 2, U.S. DEPT. OF AGRICULTURE 



All plots showed approximately 

 80 percent loss of aggregation in the 

 surface-tilled zone since the area was 

 broken from sod about 1902. The 

 lower half of the top foot of soil and 

 the middle of the second foot 

 showed only slightly lower aggrega- 

 tion than native pasture (64). 

 There were practically no water- 

 stable aggregates at 30 inches. 



Field examination indicated that 

 tilth was better on the small-grain 

 plots than on the row-crop plots 

 (table 28). Small-grain plots were 

 loose, well granulated, and had a 

 high infiltration capacity. Row- 

 crop plots were compact and mas- 

 sive with high volume weights. 

 However, the well-granulated wheat 

 plots did not have a higher content 

 of water-stable aggregates than the 

 row-crop plots. Data show that 

 barnyard manure and green ma- 

 nures (rotations 560-A and 92-A) 

 had not significantly improved 

 aggregation. 



Items 9, 10, and 11 are of special 

 interest. Plot 591-A had been con- 

 tinuously planted to wheat without 

 tillage for 29 years. Plots C-15-7 

 and C-15-8 had been planted to 

 continuous wheat without tillage for 

 18 years. These three plots showed 

 high aggregation. Aggregation must 

 be attributed to noncultivation on 

 these plots. Average wheat yields 

 were extremely low — less than 6 

 bushels per acre. 



Grasses that had been seeded 7 

 years (items 15 to 21) showed defi- 

 nite improvement in aggregation 

 compared to nongrass plots. West- 

 ern wheatgrass was least effective. 

 Buffalograss was most effective, 

 followed by blue grama. Four 

 years of buffalograss doubled aggre- 

 gation compared to nongrass plots. 



Sweetclover was not effective as 

 a soil-aggregating crop. Fourteen 

 years of sweetclover, sweetclover, 

 fallow produced only 12.4 percent 

 water-stable aggregates in the 

 plowed zone. 



Myers and Myers (60) studied 

 soil aggregation as a factor in yields 

 after alfalfa at Manhattan. They 

 made aggregate analyses of soil 

 from legume and nonlegume ro- 

 tations. Legume rotation was 2 

 years of alfalfa, row crop, oats, 

 and wheat. Nonlegume rotation 

 was row crop, oats, and wheat. 

 They found approximately 50 per- 

 cent higher total aggregation in 

 samples taken before row crop and 

 before oats from the legume rota- 

 tions than in those taken from the 

 nonlegume rotations. Improved 

 aggregation lasted for about 2 

 years after the legumes were plowed. 

 The breakdown of water-stable ag- 

 gregates in the legume rotation 

 coincided with small-grain \ields 

 in the rotations. The legume ro- 

 tations increased oats yields an 

 average of 6.7 bushels over yields 

 from the nonlegume rotation, but in- 

 creased wheat yields only 1 . 1 bush- 

 els. Row-crop yields were erratic 

 because summer drought reduced 

 yields and appeared to mask any 

 favorable effect of the legumes. 



The authors believed that in- 

 creased small-grain yields were 

 caused by favorable soil structure 

 created by the legumes rather than 

 from their influence on soil nitrogen. 



Chepil (9) mixed dry wheat 

 straw and freshly cut alfalfa hay 

 with five Kansas soils and exposed 

 them under field conditions. After 

 partial decomposition, he found 

 that straw and alfalfa hay had 

 increased the proportion of water- 

 stable particles greater than 0.84 

 mm. in diameter, decreased the 

 proportion of water-stable parti- 

 cles smaller than 0.02 mm., in- 

 creased the proportion of dry soil 

 clods greater than 0.84 mm., and 

 slightly decreased erodibility of 

 soil by wind. Effects were more 

 pronounced with large amounts of 

 added vegetative matter. 



After 2 to 5 years, his results 

 show that the decomposed vege- 



