Comparison of Potassium Fertilizer Products and
Amounts on DPL555BR Cotton, 2003
Michael D. Rethwisch, Mark Reay, Aron Quist, Tim Cox, Jessica Grudovich, and Jessica Wellman
Abstract
Introduction
Potassium fertility of cotton is of great interest to area growers and consultants, especially in years of heavy fruit
set/high retentions early when boll carpels serve as sinks for potassium (K). These early and mid season high retention
rates on cotton varieties that have high yield potential can result in what appears to be a potassium deficiency of new
cotton growth, especially on sandier soils that may not have as much available potassium as heavier soils.
Silvertooth and Norton (1989) had noted that an NH4-acetate extractable K level of greater than 150 ppm is usually
considered sufficient for cotton production, and a level less than 150 ppm is indicative of possible K deficiency in low
desert soils. Fertilizer K is often applied under conditions where soils are coarse in texture where levels of K might be
higher than 150 ppm.
Unruh et al. (1993) noted that previous experimentation in California which had resulted in notable improvements in
yield and lint quality had primarily been conducted on vermiculitic soils, and vermiculitic soils are well known for their
K fixation capabilites (Essington et al., 2000). Unruh et al. (1993) also noted that the majority of soils used for
agricultural production in the low desert contain K-bearing mica and very little of the vermiculite known to have a
tendency to fix K.
Past field experiments on low desert cotton have shown no significant response in lint yield due to the addition of a K
fertilizer (Silvertooth and Galadima 2002, and references therein). A five year study of potassium on low desert Pima
and upland cottons (Galadima et al., 1998) did not find positive increases for lint yields when using K, although quality
data was not included. A 2001 study conducted on a Toltec La Palma sandy loam soil and involving two fields with
high levels of soil test K (>400 ppm) found that additional applications of K (4.6 lb K20/acre) resulted in an average
reduction of 39.5 lbs lint/acre as well as increasing micronaire by 0.1 (Silvertooth and Galadima, 2002).
This experiment was initiated to compare differing potassium levels applied both at planting and at first bloom to help
ascertain the effects of potassium levels on cotton yields and quality under local conditions.
Arizona Cotton Report (P-138) May 2004
159
Methods and Materials
A field with soil types consisting primarily of Meloland fine sandy loam, with some Imperial fine sandy loam and
Ripley silty clay loam, was chosen for this experiment. The previous crop had been alfalfa, and several cuttings of alfalfa
had been removed from the field prior to experiment initiation in April 2003. Soil samples were obtained prior to
planting to document nutrient levels and availability.
Treatments were applied both at planting (April 19) and again at first flower (June 25). Each treatment had five
replications in a randomized complete block design. Plots were 8 rows wide by field length (1,215 ft), resulting in plots
that were 0.744 acres in size. DPL555BR cotton was planted April 19, 2003 with an 8 row planter equipped to apply
liquid fertilizer. Treatments were applied as liquid solutions, and the application equipment was calibrated to apply 23.6
gal/acre at planting.
Treatments consisted of : A) KTS = potassium thiosulfate (0-0-25, 17% sulfur; 12.2 lbs./gal) applied at 5 gallons per
acre at both planting and first bloom. B,C,D) Tracite7 0-0-21 (21% soluble potash, 17% sulfur; (Helena Chemical, 12
lbs./gal)) applied at both 2.5 and 5 gallons/acre at both planting and first bloom, resulting in three treatments (2.5 + 2.5,
2.5 + 5, and 5 + 5). E) HMBL (Helena Multiproduct Blend) which consisted of 20 gallons of Tracite7 0-0-21, 8 gallons
of Hydra-HumeTM (2-0-0.1, also contains 8% humic acid; 9 lbs/gal, Helena Chemical Company) and 48 oz of Asset7
(2% Mg, derived from magnesium ammonium carboxylates, Helena Chemical Company; 9.9 lbs./gal) for the planting
application, but did not contain Asset for the first flower application. This resulted in a solution weighing 11.15 lbs./gal.
Three gallons/acre of the solution were applied at each application. F) Grower modified starter solution consisting of
3.125 gallons of 10-34-0, 1.25 gal of UN32, 0.3125 gallons of Tracite7 0-0-21, and 0.156 pts of Trafix7 Zn/acre (4.5% S,
10% Zn; Helena Chemical Company, 10.51 lbs./gal). Total amounts of each nutrient applied are listed in Table 1.
Second application of materials was made on June 25 via sidedress application, and cotton had first open flower on
this date. Injector was calibrated to deliver 49.5 gpa of liquid. Although the grower solution treatment was applied only
at planting, these plots were also injected with water at this time in the event that sidedressing may interact with roots.
All plots were plant mapped twice during the season, with first plant mapping occuring on July 14-15, and the second
on August 14th. Plant mapping was completed using five plants per plot; data were obtained for plant height, total nodes,
first fruiting node, and fruiting site retention at site 1-3 for each fruiting node as applicable. For each plot the following
averages were calculated: height:node ratio; plant height, first fruiting node, nodes, average retention for positions
(fruiting sites) 1, 2, 3; and retention for positions 1-2 and positions 1-3.
Petioles for potassium analyses were obtained on August 12 and September 24 by collecting 15-25 petioles of 5th
leaves from top of plant in each plot. Potassium analyses were determined through laboratory analyses (Stanworth Crop
Consulting, Blythe, CA).
Plots were harvested December 22 with a John Deere 9976 four row cotton picker. All eight rows of each plot were
harvested. After each plot was harvested, seed cotton was emptied into trailer which was set up on scales so that seed
cotton weights for each individual plot could be obtained. Cotton from plots receiving the same treatment were kept
together and not allowed to be mixed with cotton from other treatments. Seed cotton for all five plots of each treatment
were combined for ginning (Modern Gin, Blythe, CA), thus allowing quality data to be collected for each treatment.
Cotton lint yields for each plot were calculated by multiplying average turn-out (36.22%) from the entire treated area
for each treatment. Quality data for each bale from the respective plots were obtained and average quality factors
calculated for each treatment. Values for cotton lint were also calculated using data from each individual bale using the
average premium or discount applied to the lint based on the CCC loan schedule.
Means and separations for all treatments were calculated using Fisher's least significant difference (Statgraphics Plus
for Windows, version 3; Manuguistics, Inc.). Quality data used a completely randomized design as cotton from all plots
of a single treatment was co-mingled, however, each treatment had a minimum of nine ginned bales for data analyses.
Arizona Cotton Report (P-138) May 2004
160
Results
The summer of 2003 was very hot year, and was not conducive to high retention of developing fruiting structures.
The lower than anticipated retention rates during the peak of bloom coupled with an excellent fall (September-October)
production weather resulted in cotton yields that included a great deal of top crop rather than than from the summer. This
may have affected resulted noted for yields and quality.
Plant mapping
No significant differences were noted for height, nodes, or height:node ratio at either sampling date (Tables 4-5). The
grower solution applied at planting set fruit lowest (node 4.92), however, while fruiting set noted for the the KTS 5 gal +
5 gal treatment was significantly higher (node 6.36) on July 14-15. These differences were not noted at the next plant
mapping (August 14) as lowest fruiting structures were now at node 7.3+1 for all treatments except for the grower
solution, which was now at node 7.6.
Retention
Retention rates were also very similar for all positions (Tables 6-7), although the KTS 5 gal + 5 gal and the HMBL 3
gal + 3 gal treatments did have significantly higher retention rates (73.4% and 73.6% respectively) at position 1 than the
Tracite 0-0-21 5 gal+ 5 gal treatment on July 14-15. No differences were seen for any other position, nor were
treatments significantly different when retentions for positions 1+2 were calculated. No differences were noted on
August 14 for retention.
K% in petioles
The grower solution applied only at planting resulted in the least amount of K in cotton petioles (5.75%) from samples
obtained on August 12 (Table 3). This level was significantly less than that noted from petioles treated with KTS 5 gal +
5 gal (6.18%), or the Tracite 2.5 + 5 gal/acre treatment (6.09%).
No statistical differences were noted between treatments for the September 24 sampling for K% in cotton petioles. The
grower solution continued to have the lowest levels of petiole K% (3.24). Treatments with highest levels of K (3.863.98%) in cotton petioles on this date were the three treatments that had received between 10-20 lbs/acre of potassium
during the study (Tracite 2.5 + 2.5, HMBL 3 + 3, Tracite 2.5 + 5).
Yields
No statistical differences existed for lint per acre (Table 8), and cotton lint/acre ranged from 1,233 lbs (2.5 gal + 2.5
gal of Tracite 0-0-21) to 1,289 lbs (2.5 +5 gal/acre of Tracite 0-0-21).
Quality
Treatments affected some aspects of cotton quality, but not others. No differences were noted for leaf or for the +b
aspect of color. Significant differences existed for all other aspects of cotton quality that were analyzed (Table 9).
Micronaire
Micronaire of cotton receiving the 2.5 gal of Tracite 0-0-21 at both planting and first bloom (4.9) was significantly
lower than cotton from any other treatment. Highest micronaire (5.1) was noted from cotton treated with 5 gal of Tracite
0-0-21 (both 2.5 + 5, and 5 + 5) at first bloom, as well as the HMBL 3 +3 treatment. The grower solution and KTS 5+5
treatments resulted in cotton with a micronaire of 5.0. The reason(s) for these differences are unclear.
Staple/length
Highest rates of potassium (KTS 5+5, Tracite 0-0-21 %+5) in this experiment resulted in a significantly shorter fibers
(35 staple) than other treatments receiving less K. The grower solution resulted in cotton fibers that were significantly
longer than the high K treatments, but slighltly numerically less than cotton receiving less than 20 lbs./acre of K. This
indicates that treatments containing high amounts of K may be contributing to a slight reduction in fiber length under the
conditions of this experiment.
Arizona Cotton Report (P-138) May 2004
161
Strength
All treatments that included 10+ lbs./acre of K significantly resulted in increased fiber strength (mimimum of 30.4
g/tex) when compared with the grower solution (29.8 g /tex). All three Tracite 0-0-21 treatments resulted in cotton lint
that was significantly stronger (30.8-30.9 g/tex) than other treatments.
Uniformity
Treatments had little effect upon fiber uniformity, although the Tracite 2.5 +5 gal/acre treatment resulted in cotton
fibers with slightly higher uniformity (81) than all other treatments (80 to 80.1)
Red
Two treatments (Tracite 2.5 + 2.5, Tracite 5 + 5) resulted in cotton fibers having slightly higher readings (82.6, 82.7
respectively) than all other treatments (81.8-81.9).
Economic Values
All treatments resulted in at least a slight increase in value based on CCC loan programs (Table 8). Treatments
containing the highest amounts of K as well as S (KTS 5 gal + 5 gal, Tracite 5 gal + 5 gal) resulted in slight value
increases of 0.53 and 0.64/lb respectively. Both of these treatments had slightly shorter fiber lengths than other
treatments as well.
Highest value/lb. of lint (+4.864/lb.) was noted for the Tracite 2.5 + 2.5 treatment, due primarily to the micronaire
value of 4.9 and combined with strength and length characteristics. This value was also significantly higher than any
other treatment. This was followed by the Tracite 2.5 + 5 treatment (+1.214/lb.), which was also significantly higher
either cotton from the HMBL treatment (0.92 4/lb.) or grower solution (0.86 4/lb.).
When values per acres were calculated, highest valued cotton/acre ($701.08) was noted from the Tracite 2.5 + 2.5
treatment although this had the lowest yields of cotton lint. This treatment also had a gross value increase of $28.18/acre
compared with the grower solution. The other two Tracite 0-0-21 treatments had lower values than the 2.5 + 2.5 gal
Tracite 0-0-21 treatment, with values/acre decreasing as amount of Tracite 0-0-21 increased ($685.88 for 2.5 + 5;
$658.55 for 5 gal + 5 gal) in this experiment.
The HMBL treatment resulted in cotton with the lowest value per acre ($655.09) while the KTS 5+5 and grower
solution treatments were very similar ($674.49 and $672.91respectively). As costs for treatments have not been
calculated, it is unknown at this time which treatment (including application costs) has provided the highest net return.
Conclusions
Increasing levels of K did not result in significant increases in lint yields, however, heat stress during the summer of
2003 caused high levels of fruiting structure losses on late planted cotton which may have greatly obscurred results as
boll carpels serve as potassium sinks. Treatments were documented to have no lasting effect upon fruit loss amounts,
although differences in petiole K were noted in mid August.
Cotton treated with fertilizers resulting in the highest levels of K and S affected cotton uniformity and staple/length in
that both parameters were noted to be slightly decreased when compared with results from other treeatments. Increasing
K did result in stronger fiber strength.
The 2.5 gal + 2.5 gal Tracite 0-0-21 treatment had the lowest micronaire as well as lowest yields, but also had the
highest value/acre. There does not seem to be a clear pattern or reason for this occurrence at this time and further testing
is necessary to confirm this result.
Arizona Cotton Report (P-138) May 2004
162
Literature Cited
Essington, M.E., D.D. Howard, H.J. Savoy, and G.M. Lessman. 2002. Potassium fertilization of cotton produced on
loess-derived soils. Better Crops Wiith Plant Food. 86(4): 13-15.
Silvertooth, J.C., and A. Galadima. 2002. Evaluation of potassium fertility in a common agricultural soil of
Arizona. Pp. 128-132. In University of Arizona 2002 College of Agriculture Cotton Report, Series
P-130. J. C. Silvertooth, ed. 237 pp.
Galadima, A., J.C. Silvertooth, and E.R. Norton. 1998. Potassium fertilization of upland and Pima
cotton (19911995, a five year project review). Pp. 490-503. In University
of Arizona 1998 College of Agriculture Cotton Report, Series P-112. J. C.
Silvertooth, ed. 604 pp.
Unruh, B.L., J.C. Silvertooth, D.M. Hendricks, and J.E. Malcuit. 1993. Potassium fertility of several Arizona
soils.
Pp. 316-318. In University of Arizona 1993 College of Agriculture Cotton Report, Series P94. J.C. Silvertooth, ed.
Silvertooth, J. C., and E. R. Norton. 1998. Cotton monitoring and management system. Publication AZ
1049, University of Arizona College of Agriculture Cooperative Extension.
Acknowledgements
We thank Stanworth Crop Consulting for providing potassium and soil nutrient levels, Tara Farms (Bob Hull) for use
of injection equipment, Desert Cotton (Roger Murphey, Bob Suffle) for use of equipment for planting and for injection of
fertilizers at first bloom, and Berger Harvesting for assistance with harvesting plots. This study was made possible
through the financial support of Helena Chemical Company.
Arizona Cotton Report (P-138) May 2004
163
Table 1. Fertilizer treatments, timings and amounts (lbs) of nutrients/acre
At planting (lbs./acre)
Treatment
N
P
K20
8
12
0.15
Tracite 0-0-21
(2.5 gal + 2.5 gal)
S
Zn
0.78
0.65
0.02
0
5.3
4.3
0
0
0
6.3
5.1
0
Tracite 0-0-21
(2.5 gal + 5 gal)
0
0
6.3
5.1
Tracite 0-0-21
(5 gal + 5 gal)
0
0
12.6
KTS 0-0-25
(5 gal + 5 gal)
0
0
15.25
Grower solution
HMBL
(3 gal + 3 gal)
Arizona Cotton Report (P-138) May 2004
At first flower (lbs./acre)
Mg
Total lbs/acre
N
P
K20
S
Zn
Mg
N
P
K20
S
Zn
Mg
0
0
0
0
0
0
8
12
0.78
0.65
0.02
0
0.15
0
5.4
4.4
0
0
0.3
0
10.7
8.7
0
0.008
0
0
0
6.3
5.1
0
0
0
0
12.6
10.2
0
0
0
0
0
0
12.6
10.2
0
0
0
0
18.9
15.3
0
0
10.2
0
0
0
0
12.6
10.2
0
0
0
0
25.2
20.4
0
0
10.4
0
0
0
0
15.25
10.4
0
0
0
0
30.5
20.7
0
0
0
0.008
164
Table 2. Soil analyses from samples taken prior to planting after alfalfa disked into soil.
Depth
pH
NO3
PO4
K
SO4
Fe
Zn
Cu
Mn
Mg
0-6"
8.0
3
3
139
47
12
0.7
1.0
4
362
6-12"
8.0
1
1
110
60
10
0.7
0.8
4
353
12-24"
7.9
1
1
109
83
10
0.6
0.8
4
352
D
D
S
VH
H
L
H
H
H
D = Deficient
L = Low
S = Sufficient
H = High
VH = Very High
E = Excessive
Table 3. Potassium levels (%) in petioles at various points during the growing season.
Treatment and rate/acre
Grower solution
HMBL 3 +3
Tracite 2.5 + 2.5
Tracite 2.5 + 5
Tracite 5 + 5
KTS 5 + 5
Aug. 12
5.75 c
5.87abc
6.06abc
6.09ab
Sept. 24
3.24a
3.88a
3.86a
3.98a
3.54a
5.85 bc
6.18a
3.72a
Means in columns followed by the same letter are not significantly different the p<0.05 level (Fisher=s LSD).
Table 4. Plant mapping data July 14-15
Treatment and rate/acre
Grower solution
36.9a
HMBL 3 +3
Tracite 2.5 + 2.5
Tracite 2.5 + 5
KTS 5 + 5
Tracite 5 + 5
Ht (in)
Nodes
21.0a
36.8a
37.8a
37.2a
36.8a
37.0a
Ht:Node ratio
1.75a
21.2a
21.0a
20.4a
20.8a
20.9a
1.73a
1.80a
1.82a
1.76a
1.77a
First Fruiting Node
4.92a
5.76ab
5.76ab
5.60ab
6.36 b
5.40ab
Means in columns followed by the same letter are not significantly different the p<0.05 level (Fisher=s LSD).
Arizona Cotton Report (P-138) May 2004
165
Table 5. Plant mapping data August 14, 2003.
Ht (in)
Treatment and rate/acre
Grower solution
48.7a
Nodes
26.9a
Ht:Node ratio
1.81a
First Fruiting Node
7.60a
HMBL 3 +3
47.8a
27.7a
1.72a
7.32a
Tracite 2.5 + 2.5
50.2a
27.6a
1.82a
7.28a
Tracite 2.5 + 5
49.9a
27.8a
1.81a
7.40a
Tracite 5 + 5
49.8a
27.6a
1.81a
7.40a
KTS 5 + 5
48.7a
27.3a
1.78a
7.32a
Means in columns followed by the same letter are not significantly different the p<0.05 level (Fisher=s LSD).
Table 6. Retention percentages of fruiting structures on July 14-15, 2003.
Position 1
Grower solution
67.6ab
Pos. 2
66.6a
Pos. 3
44.8a
Pos. 1-2
67.1a
Pos. 1-3
59.6a
HMBL 3 +3
73.6a
64.2a
47.1a
68.9a
61.6a
Tracite 2.5 + 2.5
68.3ab
68.9a
42.3a
68.6a
59.8a
Tracite 2.5 + 5
68.8ab
68.4a
47.5a
68.6a
61.5a
Tracite 5 + 5
64.9 b
65.9a
44.9a
65.4a
58.6a
KTS 5 + 5
73.4a
63.6a
44.8a
68.5a
60.5a
Means in columns followed by the same letter are not significantly different the p<0.05 level (Fisher=s LSD).
Table 7. Retention percentages of fruiting structures on August 14, 2003.
Position 1
Grower solution
33.8a
Pos. 2
34.9a
Pos. 3
27.0a
Pos. 1-2
34.3a
Pos. 1-3
31.9a
HMBL 3 +3
37.1a
34.3a
28.8a
35.7a
33.4a
Tracite 2.5 + 2.5
39.3a
31.7a
30.0a
35.5a
33.7a
Tracite 2.5 + 5
36.7a
32.3a
27.8a
34.5a
32.3a
Tracite 5 + 5
34.3a
32.1a
26.7a
33.2a
31.0a
KTS 5 + 5
36.8a
34.8a
30.2a
35.8a
33.9a
Means in columns followed by the same letter are not significantly different the p<0.05 level (Fisher=s LSD).
Arizona Cotton Report (P-138) May 2004
166
Table 8. Yields and value per acre
Treatment
Lbs lint/acre
CCC (+/-)1
Grower solution
HMBL 3 +3
Tracite 2.5 + 2.5
Tracite 2.5 + 5
Tracite 5 + 5
KTS 5 + 5
1,273a
1,238a
1,233a
0.86 c
0.92 c
4.86a
Compared with
$/acre2 grower solution
$672.91
----$655.09
- $17.81
$701.08
+ $28.18
1,289a
1.21 b
$685.88
+ $12.97
1,252a
0.60 d
$658.55
- $14.36
1,284a
0.53 d
$674.49
+ $ 1.58
Means in columns followed by the same letter are not significantly different the p<0.05 level (Fisher=s LSD).
1
Value of cotton lint based on CCC loan schedule of discounts and premiums (listed as cents/lb of lint)
2
Assumes a base value of 52.00 cent/lb.
Table 9. Cotton lint quality parameters of DPL555BR cotton.
Treatment and rate/acre Bales
Grower solution
HMBL 3 +3
Tracite 2.5 + 2.5
Tracite 2.5 + 5
Tracite 5 + 5
KTS 5 + 5
Leaf
Mic
Staple Length
9
9
9
10
10
10
2.0a
2.0a
2.0a
2.0a
2.0a
2.0a
5.0 b
5.1 c
4.9a
5.1 c
5.1 c
5.0 b
35.9a
36.0a
36.0a
36.0a
35.0 b
35.0 b
Strength
110.9a
111.0a
111.0a
111.0a
110.0 b
110.0 b
Uniform
29.8 d
30.4 c
30.8ab
30.9a
30.8ab
30.7 c
+b
80.1 b
80.0 b
80.0 b
81.0a
80.0 b
80.0 b
Color
Red
7.4a
7.5a
7.5a
7.5a
7.5a
7.3a
Means in columns followed by the same letter are not significantly different the p<0.05 level (Fisher=s LSD).
Arizona Cotton Report (P-138) May 2004
167
81.9a
81.8a
82.6 b
81.9a
82.7 b
81.9a