Soil management and landscape variability affects field-scale cotton productivity
Terra, J.A., Shaw, J.N. , Reeves, D.W., Raper, R.L., Van Santen, E., Schwab, E.B., Mask, P.L. Soil management and landscape variability affects field-scale cotton productivity Soil Science Society of America Journal Volume 70, Issue 1, January 2006, Pages 98-107
Terra, J.A., Shaw, J.N. , Reeves, D.W., Raper, R.L., Van Santen,
E., Schwab, E.B., Mask, P.L. Soil management and landscape
variability affects field-scale cotton productivity Soil Science
Society of America Journal Volume 70, Issue 1, January 2006, Pages
98-107
Abstract - A better understanding of interactions between soil
management and landscape variability and their effects on cotton
(Gossypium hirsutum L.) productivity is needed for precision
management. We assessed management practices and landscape
variability effects on seed cotton yield in a 9-ha, Alabama field
(Typic and Aquic Paleudults) during 2001-2003. We hypothesize that
landscapes have major effects on cotton productivity, but these
effects vary based on management and climate. Treatments were
established in replicated strips traversing the landscape in a corn
(Zea mays L.)-cotton rotation. Treatments included a conventional
system with or without 10 Mg ha-1 yr-1 dairy manure (CTmanure or
CT), and a conservation system with and without manure (NTmanure or
NT). Conventional systems consisted of chisel plowing/disking +
in-row subsoiling without cover crops. Conservation systems
combined no surface tillage with in-row subsoiling and winter cover
crops. A soil survey, topographic survey, and interpolated surfaces
of soil electrical conductivity (EC), soil organic carbon (SOC),
and surface soil texture were used to delineate five zones using
fuzzy k-means clustering. Overall (2001-2003), conservation systems
improved cotton yield compared with conventional systems (2710 vs.
2380 kg ha-1); neither manure nor treatment X year interactions
were significant. The conservation system was more productive than
the conventional system in 87% of the cluster X year combinations.
Slope, EC, SOC, and clay content were correlated with yield in all
treatments. Soil and terrain attributes explained 16 to 64% of
yield variation, however, their significance fluctuated between
years and treatments. In dry years, factor analyses suggested
variables related with soil quality and field-scale water dynamics
had greater impacts on CT yields than NT yields. Our results
indicate that management zones developed using relatively static
soil-landscape data are relatively more suitable for conservation
systems, and these zones are affected by soil management. In
addition, the impact of NT on yields is most apparent on degraded
soils in dry years.



