During the past century major improvements have been made in field crop yields. A review of the plant types grown over the years indicates dry weight production has remained nearly constant and yield increases have come from a higher harvest index. The same has been true for cotton and studies by others have demonstrated modern cotton varieties have a higher harvest index, have less leaf area, develop an intercepting leaf area sooner, and begin to fruit earlier and faster.

These changes in plant type have been made with little change in an accompanying cropping system. If more determinate plant types have been the result of breeders selecting on the basis of yield, it might be possible to further take advantage of higher harvest index plant types by changing the production system to favor them. This paper is a brief summary of 19 replicated field experiments conducted in the San Joaquin Valley of California from 1984-1987 where phenotypes and a cultural system to fit them has been under evaluation.

Biomass partitioning data indicates all plant types have similar plant height, number of main stem nodes, leaf area, and total dry weight until early square. At early squaring high harvest index phenotypes partition a greater percentage of energy to fruit and less to vegetation. This begins to be expressed by a lower leaf area index, less leaf weight, and plant height for high harvest index phenotypes. Number of main stem nodes is not altered by phenotype until September. By early September Acala varieties range between 47 to 52 percent of the above ground weight in bolls compared to 65 percent for SS2086.

This indicates a yield increase of about 30 percent would be possible if the same total dry weight could be achieved. Thus far SS2086 has only produced about 90 percent of the total dry weight of commercial varieties. The key to achieving this potential yield increase lies in developing cultural practices that are suited to more determinate varieties which already have some measure of growth control.

Our tests have shown a strong phenotype by row spacing interaction. Acala varieties (SJ-2, SJC-1, and GC-510) showed no change in yield on very productive soils by going from 40 to 30-inch rows (average of 1878 lbs lint/A). The three most compact plant types under evaluation averaged 1938 lbs/A in 40-inch row spacings but 2121 in 30-inch rows. This difference of 243 lbs/A lint between the two systems represents an increase that would be important to the long term competitiveness of U.S. cotton production.