-N
Surface Applications of Potassium
Best for Cotton
When growing cotton on a soil
with "low" subsoil potassium 
(K),
what rate of K fertilizer should be
used and how should it be applied?
These questions were addressed
through an Alabama Agricultural
Experiment Station field study at the
Tennessee Valley Substation, Belle
Mina.
Often in Alabama, cotton with K
deficiency has been found to have ad-
equate surface soil K, but subsoil K
levels are "low." A recent survey of
cotton fields across Alabama found
that 29 percent of the subsoils tested
"low" or "very low" in K, while only 19
percent of the subsoils tested "high."
The research site rated "medium"
in surface K, but "low" in subsoil K,
according to Soil Testing Laboratory
results. Potassium fertilizer, at rates
shown in the table, was either broad-
cast in the spring or 
split-ap- A
plied with fall and spring appli-
cations. The fall application was 
F
turned under with a moldboard
plow.
Cotton was grown in the
test area in 1989 and 1991 and
exhibited severe K deficiency
symptoms during both years. 1
Surface application of K rates in
the spring produced yields similar to
cotton treated with K split-applied in
the spring and fall. Using higher rates
of K fertilizer than normally recom-
mended (60 pounds K
2
0) for a soil
rated "medium" in K did result in some
yield response. However, cotton yield
response to K fertilizer indicates little
benefit from the deep turning under of
K fertilizer rates in the fall.
This research would indicate 
that,
for cotton grown on soils testing "low"
EFFECTS OF RATES AND TIMING OF POTASSIUM FERTILIZER ON
OTTON YIELDS, TENNESSEE VALLEY SUBSTATION, BELLE MINA
nnual K
2
0 rates,
yield/acre 
Seed cotton, yield/acre
all Spring 1989 1991 
Avg.
0 0 1,777 1,538 1,658
0 60 2,649 1,807 2,228
0 120 2,814 1,794 2,304
0 180 2,942 1,924 2,433
60 60 2,860 1,872 2,366
90 90 2,840 1,800 2,320
20 0 2,740 1,598 2,169
in subsoil K, increasing K
2
0 applica-
tions 30 to 60 pounds per acre above
surface soil recommendations may be
necessary. Applying this rate of K
fertilizer to the soil surface before plant-
ing seems the best method to correct a
K deficiency in research plots. How-
ever, these rates may have to be applied
for several years before subsoil K levels
are increased.
C.H. Burmester and G.L. Mullins
Microbial Insecticides Compare Favorably to Conventional
Insect Control in Cotton
Research in the United States and
in other countries, such as Australia,
has shown that various microbial prod-
ucts, primarily based on the bacterium
Bacillus thuringi'ensis (B.t.), can result in
yields similar to that of conventional
insecticides, such as Karate?, against
cotton bollworm and tobacco bud-
worm. Additionally, B.t. products
should reduce the rate of increase of
insecticide resistance to conventional
insecticides because of their novel
mode of action and their low toxicity
towards beneficial insects.
Alabama Agricultural Experiment
Station field studies were initiated in
1991 to see if similar results could be
obtained in Alabama and to determine
which microbial products were the
most similar to that of conventional
insecticides.
Cotton (Deltapine 90) was planted
May 23, 1991, at the Wiregrass Substa-
continued on page 2
A IC AUBURN UNIVERSIT
LOWEL T.F~ossm, IRCO AuUR UNVRIY ALBM
RESEARCH UPDATE
1992
COTTON
Reducing Cotton Production Inputs
Significant tillage and chemical in-
puts are currently used to produce
cotton in Alabama's Tennessee Valley
region. Reducing these inputs without
affecting yield would provide substan-
tial savings to growers.
An Alabama Agricultural Ex-
periment Station study was initi-
ated in 1991 at the Tennessee Valley
Substation, Belle Mina, to investigate
the potential for reducing cotton pro-
duction costs. The test area was
divided into no-till and conventional
till strips. Prowl 4E@ at 1 pint per
acre was incorporated on all con-
ventional tillage plots, and Prowl at 2
pints per acre plus Roundup@ at 2
pints per acre was sprayed on no-
till plots prior to planting. Soil fungi-
cide treatments consisted of in-fur-
row (TSX6) or hopperbox (Apron@).
Soil insecticide treatments consisted
of Temik 15G9 at 3 or 5 pounds per
acre. Preemergence herbicide treat-
ments consisted of Cotoran 4L
broadcast at 2 or 4 pints per acre.
All possible combinations of
chemical inputs were used in each
tillage system. Conventional tillage
plots were cultivated and direct
sprayed with Cotoran plus MSMA.
Control treatments (no fungicide,
Temik, or Cotoran applied) were
included for each tillage system.
REDUCED COTTON PRODUCTION INPUTS TRIALS, 1991,
TENNESSEE VALLEY SUBSTATION, BELLE MINA
Tillage
type
Stand count Seed cotton
no./6 ft. row yield
Lb./A
No-till
In-furrow ...............
Hopper box ..........
Temik, 3 lb. ...........
Temik, 5 lb ...........
Cotoran, 2 pt. .......
Cotoran, 4 pt. .......
Average ..............
Conventional
In-furrow ...............
Hopper box ..........
Temik, 3 lb ...........
Temik, 5 Ib ...........
Cotoran, 2 pt. .......
Cotoran, 4 pt .......
Average ................
No-till control ..........
Conventional control
1,573
1,546
1,488
1,632
1,568
1,551
1,560
1,647
1,638
1,577
1,708
1,611
1,673
1,642
1,514
760
Cotton stand counts were equal for
both tillage systems and were not af-
fected by soil fungicide treatments, see
table. Seed cotton yield was slightly
higher for conventional than no-till
systems when averaged over chemical
inputs. Temik at the higher labeled
rate provided numerically higher
yields in both tillage systems. Annual
morningglory control was equal for
both the low and high Cotoran rates in
both tillage systems, averaging 94
percent overall. Prickly sida control
was lower in conventional (81 percent)
than in no-till (92 percent) for the low
Cotoran rate.
Yields were not affected by
Cotoran rate in either tillage system.
Seed cotton yield was higher for the
no-till control than for the conventional
tillage control, primarily due to weed
germination suppression in the non-
tilled soil. First-year results indicate
some production inputs might be re-
duced without adversely affecting
yield.
M.G. Patterson. B.E. Norris, and B.L. Freeman
Microbial Insecticides, continued
tion, Headland. Each treatment plot
was eight rows wide by 75 feet long
and was replicated four times. In addi-
tion to an untreated control, the follow-
ing insecticides were tested: pyre-
throid (Karate); ovicide(Larvin?and
Ovasyn); B.t. (Dipel?, Javelin@, EXP
60516A, Bactec I, and Bactec IIl);
Thuringiensin (Di-Beta); virus
(Elcar@); gut poison (potassium car-
bonate); and a feeding stimulant
(strawberry gelatin). Foliar treatments,
including Prime Oil as a spreader/
sticker, were applied seven times (July
3, 9, 22, and 30 and Aug. 6, 9, and 16)
followed by one blanket coverage
spray of Karate on Aug. 22 to preserve
damage levels obtained from the mi-
crobial treatments.
Treatments were initiated when at
least five bollworm/budworm eggs
were found per 25 terminals. Yields
were taken by mech-anically harvest-
ing the middle two rows per
plot on November 6.
There were significant dif-
ferences among cotton 
yields,
with Karate-treated plots
yielding notably more than the
untreated control or the EXP
60516A, Di-Beta, and Bactec III
plus gelatin treatments. There-
fore, all treatments with a cotton
yield greater than 30 pounds per
plot were not significantly dif-
ferent from Karate, see table.
Also, the addition of an ovicide
always resulted in at least a
slight increase in control.
The summer of 1991 was a
relatively light year for boll-
worm/budworm infestations in
south Alabama. Continued re-
search in upcoming summers will
needed to determine if microb
prod ucts can be as effective as conv
MICROBIAL CONTROL OF BOLLWORM/BUDWORM ON COTTON IN
HEADLAND, ALABAMA, 1991
Treatments'
K arate, 0.04 lb. a.i. ....................................
Javelin + Larvin + Elcar, .125 lb. ..............
Dipel ES + Larvin ....................................
EXP60516A + Larvin .................................
Larvin, 0.25 Ibs. a.i. .................................
Dipel ES + Ovasyn, .125 lb. a.i. ..................
Di-Beta + Dipel ES ................... .............
Javelin, .25 lb., + Larvin ...........................
Dipel ES, 1.5 pt .....................................
Bactec III1, .75 lb., + Larvin 4
potassium carbonate, 150 g ....................
Bactec I, .75 lb ............... .. ...............
Untreated control .......................................
Di-Beta, 10 g a.i ........................
Bactec III, .75 lb., + strawberry
gelatin ..............................................
Yield
,3
37.05
35.53
35.15
34.35
33.20
32.30
31.90
31.60
31.55
31.10
30.90
28.25
27.55
25.00
1 Treatment rates per acre.
2 Mean of four replicates 
(25 plants sampled/
replicate) season long (six sampling dates).
3 Yield in pounds seed cotton/plot.
tional controls when bollworm/bud-
be worm 
pressure is 
greater.
ial 
W.J. Moar and R.H. 
Smith
Cotton Responds to Foliar Feeding of Potassium 
Nitrate
The occurrence of late season po-
tassium (K) deficiency in cotton is be-
coming more and more common in the
Southeast. Visual symptoms of K defi-
ciency develop late in the season due to
the high demand for K during boll de-
velopment. Preliminary work in Ar-
kansas has indicated possible yield
and quality increases by foliar applica-
tions of potassium nitrate (KNO
3
) to
cotton.
In 1991, an Alabama Agricultural
Experiment Station field test at the
Tennessee Valley Substation, Belle
Mina, and an on-farm test at Sam
Spruell's farm in Lawrence County
were established to evaluate foliar
feeding of KNO
3 
in Alabama. The soil
at the Tennessee Valley Substation was
rated "high" in K, while the farm site
tested "low" in 
K. At both sites 
KNO
3
(10 pounds KNO
3 
per 10 gallons of
water per acre) was applied at 10- to
14-day intervals beginning 
2 RE
weeks after first white bloom.
The Substation cotton received 
Pota
four applications 
of KNO
3
,
while the on-farm cotton re-
ceived three applications.
Severe late season K defi-
ciency symptoms were noted in
the on-farm field, while only
slight K deficiency symptoms
were noted in the Substation
field. The on-farm cotton com-
pletely defoliated when the cotton
was only 30 percent open due to the K
deficiency.
Cotton at the Substation site was
irrigated and yields were excellent, see
table. A positive yield response to both
soil and foliar applied K was observed
at this site, even though the soil had a
"high" soil test rating for K. The on-
farm site also produced good yields,
but no yield response to either the soil
SPONSE OF COTTON TO SOIL AND FOLIAR APPLIED POTASSIUM
ssium treatments/acre Seed-cotton yield/acre
Soil Foliar TVS On-farm
Lb. Lb. Lb. Lb.
0 None 2,648 3,002
0 KNO
3  
3,034 2,733
30 None 2,896 
-
30 KNO
3  
3,011
60 None 3,034 
-
60 KNO
3  
3,126 2,761
or foliar applied K was observed.
Initial first-year results appear to
be conflicting. However, these results
suggest that cotton yields can be in-
creased by foliar applications of
KNO
3 
under conditions of slight K
deficiency and high yields. In contrast,
when cotton K deficiency was severe,
foliar K was not sufficient to correct the
deficiency.
G.L. Mullins and C.H. Burmester
Effects of Planting Date, Row Spacing, 
Variety, and Plant
Growth Regulator on Cotton
Results of Alabama Agricultural
Experiment Station research from 1988
through 1990 at the Gulf Coast Substa-
tion, Fairhope, indicated an April plant-
ing date for Deltapine (DPL) 90 cotton
on a solid-36-inch rowproduced greater
yields than May or June dates in 2 of 3
years. New research was initiated at
Fairhope in 1991 to study the interaction
between row spacing, planting date,
cotton variety, and PIX? plant growth
regulator.
Cotton was planted April 15 and
May 15. DPL 20, a short sea-
son variety, was 
compared to 
PLANTIN
DPL 90, a long season variety,
on both planting dates. A Planti
solid 36-inch pattern was
compared to a skip pattern
where pairs of rows 
36 inches April 
15
apart were separated by 60-
inch skips. Lastly, PIX was
either applied or not applied.
All possible 
combinations 
of May 
15
planting date, row spacing,
variety, and PIX treatments
were represented.
Seed cotton yields were
approximately 900 pounds per acre
greater for the May planting date when
averaged over row spacing and variety,
see table. Within planting dates, the
solid 36-inch pattern produced slightly
higher yields than the skip pattern.
Except for DPL 20 planted on May 15,
neither variety demonstrated a clear
yield advantage over the other.
PIX did not significantly affect
yield in any case; however, cotton
height was reduced 8 to 12 inches in
PIX treated plots. More bolls per plant
NG DATE, Row SPACING, VARIETY, AND GROWTH REGULATOR
EFFECTS ON COTTON, FAIRHOPE, 1991
ng date Row 
spacing Variety 
Seed cotton
yield
In. Lb.
Solid-36 DPL 90 3,100
Solid-36 DPL 20 3,279
Skip-36 DPL 90 2,968
Skip-36 DPL 20 2,858
Average 3,052
Solid-36 DPL 90 3,836
Solid-36 DPL 20 4,341
Skip-36 DPL 90 3,799
Skip-36 DPL 20 3,802
Average 3,944
were counted for the April planting
date (average 26) compared to the May
planting (average 21). Rotten bolls per
plant were numerically higher for the
April planting (average 6) than the May
planting (average 4). No differences in
the number of rotten bolls were found
among row patterns within each plant-
ing date.
M.G. Patterson, M.D. Pegues, and K.L. Edmisten
Chlorophyll Meter
Has Potential for
Determining Cotton
N Status
Cotton producers, consultants, and
researchers have long sought for a
quick, reliable method to determine
cotton nitrogen (N) needs. Improper N
application can affect production and
the environment.
Petiole nitrate tests are used in sev-
continued on page 4
Chlorophyll Meter, continued
eral southern states to monitor cotton N
status, but research has not confirmed
the reliability of these tests in Ala-
bama. Leaf-blade total N analysis has
been useful in predicting cotton N re-
quirements, and is less affected by cli-
mate and seasonal changes than petiole
nitrate.
Because leaf N status is directly
related to leaf chlorophyll content, a
newly developed hand-held chloro-
phyll meter could offer a substitute for
leaf-blade total N analysis. Chlorophyll
meter readings are measured in soil
plant analysis development (SPAD)
units and are instantaneous. If chloro-
phyll meters 
could predict 
cotton N
needs, farmers could realize substantial
savings of money, labor, and time asso-
ciated with leaf-tissue collection and
analysis.
Research was conducted in 1991
to determine the feasibility of using
chlorophyll meter readings for deter-
mining cotton N status. The experi-
ment was conducted at the E.V. Smith
Research Center, Shorter. Several N
rates were used to establish a range of
cotton (Deltapine 50) yield, tissue N
concentration, and chlorophyll con-
centration. Chlorophyll meter read-
ings were taken at first square, first
bloom, and mid-bloom on the upper-
most fully developed leaves. For com-
parison, leaf-blade and petiole samples
were collected and analyzed for total N
and nitrate, respectively.
Abundant, well distributed rainfall
during the 1991 growing season pro-
moted high seed cotton yields and re-
sponse to N fertilizer. Seed cotton
yields ranged from about 1,000 pounds
per acre with no N fertilizer to 3,500
EDITOR S NOTE
Menitioni of compaliiy or trade naitis. does not
indicate cteorscimcnit 
b
y the Alabama Agricul
ftural Experiimcit Station or Auburii Lhliversity of
one branl over ianother Aiiny menition of inoimabe
uses or applications in excess of labetled rates of
pesticide o crothcr chmical;l doiios ot constitute a
r'coplptdalttioi1. Sucli iuse iil risetarchl is simply
part of the scientific ivlstigation nicessary to fiully
tvalliittt' llatrials la d tr atieits.
Info rmationi contained hereini is available' to aill
pcrsoils wtithlout regard to race, 
color , sx, or lilatiorlill
ortiii.
4,000
- 3,000 
<
36  
A A
A A
A34 A 2,000 :
A A
32 A A A 
A 1,000
30 I 
I I I I I
3 4 5 6 32 34 36 38 40
Leaf N, % SPAD Reading
Relationship between (A) chlorophyll meter readings (SPAD) and cotton leaf-blade 
N at first
square, and (B) between seed cotton yield and chlorophyll meter readings 
at first square.
pounds per acre with 160 pounds N per
acre.
Chlorophyll meter readings on
leaves at first square were highly re-
lated to leaf-blade total N, see figure A.
Relationships between leaf-blade total
N and chlorophyll meter readings at
first square also were highly related to
cotton yield, figure B. Similar relation-
ships were found at first bloom and
mid-bloom.
These relationships suggest that
cotton N requirements could be pre-
dicted with the chlorophyll meter. In
addition, the chlorophyll meter pre-
dicted seed cotton yields as well as or
better than leaf-blade N and petiole ni-
trate analyses at all three stages of
growth. Especially promising is the
connection between chlorophyll meter
readings at first square and seed cotton
yield, because supplemental N fertilizer
could easily be applied at that stage of
growth.
Further calibration of the meter is
needed before this technology can be
useful to producers.
C.W. Wood. D.W. Reeves, and K.L. Edmisten
Editor's Note: Please use the form below to send the name and address of any neighbor
or friend who should receive the report. If you do not wish to receive future issues, please
indicate that fact on the form and we will remove voour name from the mailing list.
J Add the following name to receive the AAES Cotton Update.
U Remove the following name from the mailing list for the AAES Cotton
Update.
Name
Street, Box, or Route No.
City State 
Zip
L . . . . . . . . .
. . .
'4,
Alabama Agricultural Experiment Station
Auburn University
Auburn University, Alabama 36849-0520
NON-PROFIT ORG.
POSTAGE & FEES
PAID PERMIT NO.9
AUBURN, ALA.
Address Correction Requested
March 1992 4 M
3uA B
38 L? A