(P) fertilization of cotton is one of the oldest fertilizer
practices used in the southeastern United States. The discovery of rock
phosphate in South Carolina in 1867 and in Florida in 1887 culminated
in a superphosphate industry in the heart of the old Cotton Belt. Many
of the early experiments with P application to cotton in the southeastern
U.S. showed yield responses up to about 40 pounds superphosphate (P2O5)
per acre. A 1965 survey found that most agronomists from Alabama to Virginia
commonly recommended 60 to 80 pounds P2O5 per acre
for upland cotton.
Alabamas early P recommendations
were based on P-rate experiments maintained by the AAES on research stations
in the major land resource areas of the state and on smaller experiment
fields in minor soil regions. As a result of standard P recommendations
through the early 1950s, many fields in long-term cotton production reached
high or very high soil P levels and were no longer
responsive to additional P applications.
University Soil Testing Laboratory began service in 1953 to help cotton
and corn producers identify those fields that no longer needed additional
P application but could benefit from other soil amendments. However, soil
test summaries for cotton samples from the AU Soil Testing Laboratory
since the early 1960s seem to indicate a gradual trend toward lower soil
test P values (figure
1). More samples are testing medium and low
in P while fewer samples are testing high and very high.
Possible explanations for this trend include (1) use of less P fertilizers
for cotton, (2) deeper tillage which dilutes residual soil P, (3) soil
erosion, (4) P fixation by soil minerals, and (5) failure of growers/consultants
to routinely sample high-testing fields.
A cotton soil fertility survey of
312 Alabama cotton fields in 1991 pointed out declining plow layer soil
test P levels but identified only 10% of the fields that had leaf samples
below the established sufficiency range for P; only 9% of the plow layers
tested low or very low in P.
Despite a strong research basis for soil
testing in Alabama, producers and their consultants often question the
validity of soil test interpretations. No doubt part of this is due to
a wide range of interpretations available from different public and private
soil testing services, opinions of consultants, and competition for fertilizer
sales. Emphasis on precision agriculture may have created expectations
from soil testing beyond what it is capable of delivering. Nevertheless,
as producers adopt new technologies, genetically improved varieties, and
new production practices, they expect and deserve periodic verification
of soil testing interpretations from their public laboratories.
Phosphorus variables on the Two-Year
Rotation and Rates of N-P-K experiments were summarized in 1992 in order
to validate or update Alabamas current soil test calibration for
P using modern varieties with higher yield potentials.
Since 1982, the P variable treatments
in both experiments have been in a residual mode (i.e. they have not received
any additional P applications). However, they continued to be fertilized
with other nutrients as in the fertilized control. Having the experiments
in a residual P mode allows true soil test P calibration because responses
are a result of residual soil P and are not complicated by additional
P fertilization. Only the fertilized control treatments continued to receive
P fertilization. These treatments served as a standard for yield comparison.
Plow-layer soil samples were collected from each plot every other year
and analyzed by the Auburn University Soil Testing Laboratory using the
dilute, double acid extractant (Mehlich-1 extraction procedure) that is
used by most southern soil testing laboratories.
To remove the large year-to-year
variation in yield, yields are reported as a percent yield relative to
the fertilized control treatment in each test for each year. The control
treatments received a standard N rate of 90 pounds N per acre in split
applications, 60 and 100 pounds P2O5 per acre in
the Two-year Rotation and Rates of N-P-K experiments, respectively, and
60 and 100 pounds K2O per acre in the Two-year Rotation and
Rates of N-P-K experiments, respectively.
results for P are calibrated with relative cotton yields for each of the
experiments at each of the five locations, several locations and tests
fail to demonstrate a response to increasing levels of residual soil P.
Treatments on all soils except the Lucedale sandy clay loam at Prattville
would be rated low by the southern public soil testing laboratories
(see table). Failure to get
dramatic responses all the time demonstrates the inherent difficulties
of trying to make soil testing a definitive and infallible tool. However,
when yield and soil data from both tests at all Coastal Plain locations
over the entire seven-year period are pooled (figure
2), a reasonable critical value for P can be estimated.
Most southern soil testing laboratories,
including the Auburn University laboratory, uses a critical value
in their soil test report. The Soil Science Society of America (1997)
defines critical soil test concentration as . . . that concentration
at which 95% of maximum relative yield is achieved. Above
this value, no fertilizer P is recommended because the probability of
a yield response is extremely low. Below this value, P is recommended.
The lower the soil test value, the higher the P recommendation.
The critical values used for P on cotton by the Auburn University Soil Testing Laboratory were established in 1968 and verified and updated in numerous AAES reports since then. Alabama's critical values appear to be in line with other state laboratories in the southern U.S. that use the same extraction procedure for Coastal Plain soils. Alabama has a different critical value for the red, clayey soils of the Tennessee Valley (figure 2). Recent yields from the Two-year Rotation and Rates of N-P-K experiment verify these values. The values in use still appear as accurate as possible.