See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/221988992

Yield and fiber quality of Upland cotton as influenced by nitrogen and

potassium nutrition

Article  in  European Journal of Agronomy · April 2006

DOI: 10.1016/j.eja.2005.10.004 · Source: OAI

CITATIONS READS

189 694

3 authors:

John J. Read K. Raja Reddy

United States Department of Agriculture Mississippi State University
81 PUBLICATIONS   2,531 CITATIONS    294 PUBLICATIONS   12,122 CITATIONS   

SEE PROFILE SEE PROFILE

Johnie Jenkins
United States Department of Agriculture
426 PUBLICATIONS   9,499 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Temperature effects soybeans growth and development View project

Abiotic Stress Tolerance in plants View project

All content following this page was uploaded by K. Raja Reddy on 19 February 2021.

The user has requested enhancement of the downloaded file.

 Europ. J. Agronomy 24 (2006) 282–290

Yield and fiber quality of Upland cotton as influenced

by nitrogen and potassium nutrition
John J. Read a,∗ , K. Raja Reddy b , Johnie N. Jenkins a
a USDA-Agricultural Research Service, Crop Science Research Laboratory, 810 Highway 12 East, Mississippi State, MS 39762-5367, USA
b Department of Plant and Soil Sciences, Box 9555, Mississippi State University, Mississippi State, MS 39762-9555, USA

Received 6 January 2005; received in revised form 30 September 2005; accepted 24 October 2005

Abstract
Nutrient stress in Upland cotton (Gossypium hirsutum L.) depresses lint yield, particularly of late-season fruit (bolls), and may disrupt fiber
development. A 2-year (1999 and 2000) study was conducted outdoors in large pots to determine individual effects of nitrogen (N) and potassium
(K) stress at flowering stage on lint yield and fiber quality. Treatments were half-strength nutrient solution from emergence to crop maturity
(control), 20% and 0% of control N from first flower onward, and 20% and 0% of control K from first flower onward in 1999 and first square
onward in 2000. Leaf N and K were determined every 2–3 days from an uppermost, fully expanded leaf on the main-stem of five plants selected
at random. Mature bolls were harvested from sympodial (fruiting) branches only and grouped according to week of anthesis across a 35-day
flowering period, providing five flowering groups, from which fiber length, strength, and micronaire were determined. Fiber length was not con-
sistently altered by stress, suggesting early stages of fiber development were indirectly affected by plant N and K status. Nitrogen deficiency
decreased yield through early termination of reproductive growth. In 1999, although flowering group four of N-deficient cotton had low length,
strength, and micronaire, values for weighted-sum micronaire (whole-plant micronaire) increased under N stress by about 12% in 0% N treatment
and about 18% in 20% N treatment. In general, N and K stress had opposite effects on weighted-sum micronaire. The year by N treatment
interaction was significant for weighted-sum strength, due to weak fibers in N-deficient cotton in 1999, but no treatment difference in 2000.
Apparently, crop response to N stress was influenced by environment, as flowering groups with low quality fiber also comprised a large fraction
of total lint, and thus placed heavy demands on plant N and carbohydrate reserves. Severe K deficiency in 2000 decreased yield and lint weight
boll−1 , and micronaire values of 3.7 or less were evident in flowering groups two, three and four. Results support evidence that N stress indi-
rectly affects cotton growth, as N deficiency decreased fiber length, strength and micronaire primarily in flowering groups with large percentage
of bolls. Results from 2000 support evidence that K deficiency adversely affects reproductive growth, boll weight, and sugar translocation in
cotton.
Published by Elsevier B.V.

Keywords: Cotton; Fiber development; Micronaire; Nitrogen deficiency; Potassium deficiency; 2.5% span length; Yarn strength

1. Introduction cotton will exhibit a more determinate growth and flowering

pattern (Gerik et al., 1998). The developing boll is also a major
Yield and quality in Upland cotton (Gossypium hirsutum sink for K, especially the seeds (Usherwood, 2000). Because K
L.) are influenced by genetics and environmental conditions is involved in plant water relations and carbohydrate transloca-
(Ramey, 1986; Reddy et al., 1999). Because nitrogen (N) or tion, K deficiency, unlike N deficiency, restricts fruit production
potassium (K) deficiency in cotton limit yield similarly through to a greater extent than vegetative growth (Kerby and Adams,
decreased leaf area expansion and CO2 assimilation capac- 1985; Pettigrew, 1997). Pettigrew et al. (1996) reported K defi-
ity, low productivity is often associated with low fiber quality ciency decreased lint yield, fiber elongation, 50% span length,
(Bradow and Davidonis, 2000; Reddy et al., 2004). Fruiting and micronaire in eight genotypes of differing relative earliness
structures (bolls) have a high requirement for N, and N-deficient and regional adaptation. In that study, varying N fertilization did
not affect yield or fiber quality.
Reports of fiber property trends in studies of cotton nutri-
∗ Corresponding author. Tel.: +1 662 320 7420; fax: +1 662 320 7544. tion are sometimes contradictory due to the interactive effects
E-mail address: jjread@msa-msstate.ars.usda.gov (J.J. Read). of genotype, weather, and soil (Minton and Ebelhar, 1991;

1161-0301/$ – see front matter. Published by Elsevier B.V.

doi:10.1016/j.eja.2005.10.004
 J.J. Read et al. / Europ. J. Agronomy 24 (2006) 282–290 283

Pettigrew, 2003; Reddy et al., 2004). Additionally, the indeter- Few studies have addressed the effects of N and K deficiency on
minate growth habit of cotton, and cultivar variation in develop- cotton quality (Bradow and Davidonis, 2000), although Heitholt
ment rate, may cause fiber properties to vary in different studies (1994) examined effects of nutrient stress on percent boll reten-
(Jenkins et al., 1990; Jones and Wells, 1998). On the same day, tion and fiber properties in cotton. Cotton producers may be able
individual bolls may be just starting fiber elongation, others to enhance overall profitability through utilization of agronomic
starting fiber thickening and others may be completely mature practices that optimize lint quality without sacrificing yields. In
(Davidonis et al., 2004). Because cotton plants continuously pro- precision farming research, Johnson et al. (2002) demonstrated
duce bolls that are strong sinks for plant nutrients, the onset of cotton yield and fiber quality are spatially correlated; however,
N or K deficiency that disrupts fiber development can be greatly Bradow et al. (1999) found no meaningful correlation between
influenced by plant growth and stage of development (Bradow fiber strength and spatial variations in levels of K or percent
and Davidonis, 2000; Boquet and Moser, 2003). Consequently, organic matter in the soil. Still, there is opportunity to direct
it is difficult to predict the effects of N or K deficiency on fiber cultural-input strategies during the production season and min-
development and lint quality without knowledge of the timing imize effects of nutrient stress on seed cotton yield and fiber
and intensity of stress (Ramey, 1986). quality. The objective of this study was to determine the inde-
The blooming period in cotton is about 6 weeks and is asso- pendent effects of N or K nutrition on cotton yield and fiber
ciated with increased uptake of soil-applied nutrients (Boquet quality in the cohort of bolls from five fruiting zones, grouped
and Breitenbeck, 2000). Under ideal growing conditions (e.g., according to week of anthesis, and hence, when fibers are elon-
average air temperature of 30 ◦ C), flowers are produced at 3-day gating. Results should guide further research on the impacts of
intervals on the first nodes of successive fruiting (sympodial) N and K deficiency and timing of plant nutrient stress on cotton
branches up the plant and at 3–4-day intervals at successive fiber development.
nodes out a sympodial branch (Gerik et al., 1998). Fibers orig-
inate from the outer seed coat of the developing seed and 2. Materials and methods
their development occurs in three distinct processes of elonga-
tion, secondary wall thickening or maturation and then drying The experiment was conducted at the R.R. Foil Plant Science
(Davidonis et al., 2004). Fiber elongation begins around anthe- Research Center at Mississippi State University, Mississippi
sis, with maximum length occurring at approximately 20–25 State, MS, USA (Lat. 33.416; Long. 88.782 W) with NuCotn
days after anthesis (DeLanghe, 1986). Potassium malate is used 33B, a mid-season Upland Bt (Bacillus thuringensis) cultivar.
to increase turgor pressure for growth and elongation (Ramey, Plants were grown outdoors in large, free-draining polyvinyl
1986). The fiber begins to thicken 15–20 days after anthesis, chloride (PVC) pots (15 cm diameter × 65 cm depth) filled with
as rings of cellulose are deposited in secondary wall forma- sand and supplied water and nutrients via plastic pipe and drip-
tion until about 50 days after anthesis. Cellulose is deposited at per system (Netafim,1 Fresno CA). Pots were set side-by-side
slightly different angles during this thickening process, a fea- in 3-m long rows spaced 1 m apart in a wooden rack that was
ture that ultimately has a role in giving strength to that fiber arranged in an east-west direction. Seeds were sown 17 May
(Davidonis et al., 2004). The degree of secondary wall deposition 1999 and 15 May 2000. Seedlings were emerged on 24 May
determines fiber maturity. Micronaire is a composite measure of 1999 and 21 May 2000, and thinned to one plant pot−1 on 11
maturity and fiber fineness since fiber cells with the same wall June 1999 and 6 June 2000. Each treatment (control, N stress
width can have different micronaire values (Davidonis et al., and K stress) was comprised of three, 1-m wide rows with plants
2004). Micronaire tends to increase when there is ample sup- spaced 0.15 m on center, providing 20 plants per row (replicate).
ply of carbohydrate to mature bolls set on the plant (Pettigrew, This arrangement of potted plants was equivalent to a popula-
2001). Micronaire was linearly related to the amount of canopy tion of 65,605 plants ha−1 . Insects were controlled as necessary
photosynthesis that occurred from 15 to 45 days after flower- using conventional practices.
ing (Bauer et al., 2000). Thus, seasonal shifts in plant growth Daily thermal units (TU) were calculated each year by the
and metabolism are manifest in higher levels of fiber matura- following equation:
tion in bolls from July flowers, as compared to fibers in bolls

from August flowers (Jenkins et al., 1990; Davidonis et al., Tmax + Tmin
TU = − 15.5 ◦ C (1)
2004). 2
Fiber length has always been important to cotton manufac-
turing and since the introduction of rotor spinning technology where Tmax and Tmin were the maximal and minimum daily
to cotton manufacturing in 1970, micronaire and strength both temperatures in degrees Celsius, respectively. Daily tempera-
have increased in importance relative to other quality character- tures were recorded from a weather station adjacent to the field
istics (Deussen, 1986). Studies of N and K nutrition in Upland site. Thermal units are sometimes used to predict the attainment
cotton have usually emphasized increased yield and fruiting effi- of different crop growth and development events, because cot-
ciency (Boquet and Breitenbeck, 2000; Pettigrew and Meredith,
1997), though several have been extended to include fiber quality 1 Mention of a trademark, proprietary product, or vendor does not constitute
(Minton and Ebelhar, 1991; Reddy et al., 2004; Reddy and Zhao, a guarantee or warranty of the product by the U.S. Department of Agriculture
2005). Pettigrew et al. (1996) found N and K supplements did or Mississippi State University, and does not imply its approval to the exclusion
not affect fiber strength, but added K increased fiber elongation. of other products or vendors that may also be suitable.
284 J.J. Read et al. / Europ. J. Agronomy 24 (2006) 282–290

ton growth increases linearly as temperature increases between sample standard deviation for leaf N was 2.0 g kg−1 for
15.5 and 31.1 ◦ C (Jones and Wells, 1998; Reddy et al., apple and 4.2 g kg−1 for spinach. The sample standard devi-
1999). ation for leaf K was 2.2 g kg−1 for apple and 2.5 g kg−1 for
All control and border rows were fed a favorable supply of spinach.
water and nutrients using half-strength nutrient solution (Hewitt, Daily during the reproductive period, sympodial-branch posi-
1952), containing 85.7 mg N L−1 and 138.1 mg K L−1 , via one tion white blooms (flowers at anthesis) were tagged with a
Netafim dripper per pot rated at 1 L liquid h−1 . Plants were jeweler tag bearing the current date. Bolls from monopodial
irrigated three times each day to provide 120% of daily pan branches were also tagged, but were not used for fiber anal-
evaporation, measured at the nearby weather station, in order ysis. Tagging white blooms for all treatments on the same day
to maintain optimum water conditions throughout the experi- insured that the tagged blooms were of equivalent metabolic and
ment. The different N and K treatment solutions were contained developmental ages for each treatment (Ramey, 1986; Pettigrew,
separately in mixing tanks and pumped through plastic lines to 2001). Lint from only sympodial branches was grouped accord-
each 3-row treatment plot. All timing and duration of flows was ing to week of anthesis across a 35-day flowering period, giving
under computer-controlled switches and solenoid valves. At var- five flowering groups, from which lint yield and fiber properties
ious phenological stages, the plants in some three-row plots were were measured.
supplied nutrients with an osmotically balanced nutrient solu- Irrigations were terminated on 29 September 1999 and on
tion (Hewitt, 1952). The three N treatments were: (1) control, 22 September 2000 when more than 60% of bolls were opened.
a half-strength nutrient solution supplied from plant emergence At harvest, the remaining tagged bolls (those not lost by injury
to maturity; (2) 20% N at first flower stage, a moderate stress or natural fruit abortion) were harvested by hand and their spe-
imposed from first flower stage (52 DAE) to maturity; and (3) cific node and sympodial-branch position recorded. Lint yield
0% N at first flower, a severe stress imposed by completely with- was determined from mature bolls that were each ginned indi-
holding N from first flower (52 DAE) to maturity. The three K vidually using a roller gin. Lint from each flowering group
treatments were: (i) control from plant emergence to maturity; (week of flowering for sympodial-branch fruit) was pooled
(ii) 20% of control K commenced either 52 DAE in 1999 or across 20 plants in each replicate and a 10-g sample was sent
31 DAE in 2000; and (iii) 0% of control K commenced either to Starlab (Knoxville, TN, USA). Individual, traditional instru-
52 DAE in 1999 or 31 DAE in 2000. The K-stress treatments ments (digital fibrograph, stelometer, and micronaire instru-
were initiated at an earlier growth stage in 2000 in order to elicit ment) were used to measure length (2.5% span length), strength
K deficiency sooner in plant development, as visible symptoms (tenacity) and micronaire. Length and strength were deter-
were observed late in the season in 1999 when K was withheld at mined twice and the values averaged for statistical analysis. The
flowering stage. Reducing the amount of N or K available would 2.5% span length is the length spanned by the longest 2.5% of
dilute the amount of nutrient in plant tissues due to continued the fibers in the test sample and was measured with a digital
growth. fibrograph. Fiber bundle strength was measured by stelometer.
Plant nutrient status was assessed every 2–3 days by excis- While additional factors of length uniformity, maturity index,
ing an uppermost, fully expanded leaf on the main-stem of five color, trash and sample preparation also establish quality, this
plants selected at random. Total N in these leaves provides an paper will deal only with 2.5% span length, fiber strength and
estimate of the N accumulated by the plant prior to sampling micronaire.
(Gerik et al., 1998). Although plant K is typically assessed All data were subjected to analysis of variance using PROC
from petiole K (Kerby and Adams, 1985), we used leaf blades GLM procedures in SAS (SAS Institute, 1999). Year and the
because of their importance to light interception and dry mat- interactions with year were designated as random effects in the
ter production and to minimize environmental effects on water various models. In terms of lint yield, data for each flowering
and nutrient uptake. The five leaves were combined, dried at group were summed across 20 plants in each replicate in order
70 ◦ C and ground to pass a 0.5-mm (40 mesh) screen in a small to determine whole-plant responses to N and K deficiency. Lint
Wiley mill. Leaf N was determined according to standard micro- weight boll−1 was expressed as the ratio of total lint produced to
Kjeldahl methods (Nelson and Sommers, 1972), and K was total number of bolls in each replicate. In regards to fiber quality
measured on nitric-perchloric acid digests using inductively cou- on a whole-plant basis, values for length, strength and micron-
pled plasma optical emission spectrometry. Because leaves were aire in each flowering group were weighted by the fraction of lint
pooled prior to nutrient analysis, the number of observations on produced, based on the summation of lint in each 20-plant repli-
each sampling date is equivalent to the number of treatments. cate, year, treatment combination. Then these weighted fiber
While pooling across reps precluded any statistical analysis of quality values were averaged across the five flowering groups.
treatment differences, one of the primary objectives of this study Thus, fiber quality expressed as a weighted-sum average was
was to determine if temporal changes in leaf N and K under nutri- weighted toward where (or more precisely when) on an average
ent stress are related to yield and quality of a cohort of bolls in plant the most amount of lint was produced. Analysis of vari-
different fruiting zones, based on week of anthesis. Several sam- ance was performed within each flowering group to determine
ples of NIST-certified standards of dried apple (N = 22.5 g kg−1 ; the effects of treatment and year by treatment interaction on yield
K = 16.1 g kg−1 ) and spinach (N = 59.0 g kg−1 ; K = 29.0 g kg−1 ) fraction, boll number and fiber quality. Treatment means were
leaves (U.S. Department of Commerce, NIST, Gaithersburg, separated using Fisher’s protected least significant difference
MD, USA) were included during laboratory analyses. The (LSD) test at P < 0.05.
 J.J. Read et al. / Europ. J. Agronomy 24 (2006) 282–290 285

Table 1
Week of anthesis, total number of mature bolls (number of plants harvested is in parentheses), and average lint weight boll−1 in five flowering groups of cotton grown
outdoors in large pots supplied half-strength Hoagland’s solution (control) in irrigation water throughout the experimental period
Flowering group Week of anthesis Bolls (no. three reps−1 ) Lint weight (g boll−1 )

1999 2000 1999 2000 1999 2000

One 11–17 July 9–16 July 135 (52) 231 (53) 1.44 1.71
Two 18–24 July 17–23 July 219 (55) 308 (57) 1.45 1.61
Three 25–31 July 24–30 July 188 (56) 229 (54) 1.92 1.59
Four 1–7 August 1–6 August 253 (56) 125 (47) 2.00 1.47
Five 8–14 August 7–15 August 48 (30) 42 (29) 1.62 1.22

Note: Data are for sympodial (fruiting) branches only from 60 control plants. Open bolls were first observed 22 August 1999 and 17 August 2000. Irrigations stopped
at 29 September 1999 and 22 September 2000. Including vegetative branches, the total number of bolls harvested was 924 in 1999 and 976 in 2000.

3. Results yield in both years (Table 2). Analysis within each treatment
found a significant (P < 0.01) effect of year by flowering group
3.1. Plant and environment effects interaction for yield fraction. This interaction was influenced
by boll number, because a large number was harvested from
Among control plants, the anthesis dates of each flowering flowering groups two, three and four in 1999; whereas, groups
group were similar in both years, although flowers occurred one, two and three were generally most productive in 2000
about 1 day sooner in 2000 than 1999 (Table 1). A large num- (Table 3).
ber of bolls was harvested from flowering groups two and While temperature conditions were not excessive, differences
three (mainstem nodes 6–12 and 8–14, respectively), which in environmental conditions between years had potential to alter
corresponded to nodes that are typically the most productive plant development (Fig. 1). Thermal units (Eq. (1)) accumulated
fruiting branches for cotton in the mid-south (Jenkins et al., between planting and harvest, 17 May to 31 October, were sim-
1990). Among all treatments, flowering groups two and three ilar in 1999 and 2000 (1515 versus 1543), but the distribution
made up a large fraction, from about 50–65%, of the total lint differed between years. In 1999, warm temperature conditions

Table 2
Effects of nitrogen (N) and potassium (K) stress at flowering stage on percentage of final lint weight in mature cotton bolls harvested in five flowering groups from
sympodial branches only of cotton grown outdoors in large pots in 1999 and 2000
Stress treatment Group one Group two Group three Group four Group five

1999 2000 1999 2000 1999 2000 1999 2000 1999 2000

Control 13.2 26.5 21.9 33.2 24.6 24.5 34.8 12.3 5.4 3.4
20% N FF 16.2 26.4 36.9 34.9 28.2 29.7 18.3 7.9 0.4 1.1
0% N FF 17.3 33.7 40.9 44.4 23.5 19.5 18.4 2.5 nd nd
20% K FF/FS 8.0 27.6 22.3 23.6 25.4 26.6 37.8 15.0 6.6 7.2
0% K FF/FS 13.7 40.6 24.3 33.0 24.4 21.6 34.5 3.6 3.1 1.9
Average S.E. 1.1 2.1 1.7 2.0 2.6 1.8 2.5 1.2 0.7 1.0

Note: Values represent the mean of three replicate rows with 20 plants in each row. Control, half-strength nutrient solution at plant emergence onward; 20% N FF,
20% of control N at first flower onward; 0% N FF, 0% of control N at first flower onward; 20% K FF/FS, 20% of control K at first flower (1999) or first square (2000)
onward; 0% K FF/FS, 0% of control K at first flower (1999) or first square (2000) onward. Average S.E., experimental standard error of the mean. nd, no data.

Table 3
Effects of nitrogen (N) and potassium (K) stress at flowering stage on number of mature cotton bolls harvested in five flowering groups from sympodial branches
only of cotton grown outdoors in large pots in 1999 and 2000
Stress treatment Group one Group two Group three Group four Group five

1999 2000 1999 2000 1999 2000 1999 2000 1999 2000

Control 45 77 73 103 63 76 84 42 16 14
20% N FF 55 71 105 96 74 91 50 26 2 3
0% N FF 47 70 96 98 56 55 46 7 nd nd
20% K FF/FS 42 80 78 79 76 88 105 48 26 24
0% K FF/FS 38 65 71 70 55 52 83 11 11 5
Average S.E. 4.0 6.0 4.9 7.4 6.8 6.1 5.3 3.4 2.1 3.0

Note: Values represent the mean of three replicate rows with 20 plants in each row. Control, half-strength nutrient solution at plant emergence onward; 20% N FF,
20% of control N at first flower onward; 0% N FF, 0% of control N at first flower onward; 20% K FF/FS, 20% of control K at first flower (1999) or first square (2000)
onward; 0% K FF/FS, 0% of control K at first flower (1999) or first square (2000) onward. Average S.E., experimental standard error of the mean. nd, no data.
286 J.J. Read et al. / Europ. J. Agronomy 24 (2006) 282–290

yield fraction and less retention of bolls in flowering group four

(Tables 2 and 3).
Weighted-sum fiber length averaged significantly less in 1999
than 2000 (28.2 mm versus 29.0 mm, Tables 4 and 5). Fiber
length of flowering group one was consistently less in 1999 than
2000, averaging from 1.0–2.2 mm shorter in the different treat-
ments. Low fiber length in 1999 was associated with somewhat
warmer conditions in July and somewhat lower leaf N under
N stress in 1999, as compared to 2000 (Fig. 2). Among con-
trol plants in 1999, micronaire values increased linearly across
flowering groups one to four (Table 4), and was associated with
increased lint weight boll−1 (Table 1). The opposite trend was
observed in 2000, as micronaire and lint weight boll−1 decreased
as flowering group increased.

3.2. Nitrogen stress effects

In 1999, the decline in leaf N in 20% N treatment resulted in

plants considered low in N (25–30 g kg−1 ; Gerik et al., 1998) on
most sampling dates in August. Leaf N levels in 0% N treatment
in 1999 ranged from 25 to 30 g kg−1 following 23 July, and were
below 25 g kg−1 on most sampling dates in August (Fig. 2). Boll
demand for N was evident from changes in leaf N in 1999, which
increased in the first week of boll-filling (about 7–14 August)
Fig. 1. Total weekly thermal units and average weekly solar radiation in July, before declining to about 25 g N kg−1 on 16 August 1999. Leaf
August and September 1999 and 2000 based on daily weather data collected at N declined rapidly under stress in 2000, but only 0% N treatment
Mississippi State, MS, USA.
was considered low or deficient in N by mid-August (Gerik et
al., 1998).
and high solar radiation from about 17 July to 7 August, the Lint yield was significantly lower in 0% N treatment than
peak-flowering period, may have contributed to less boll reten- control and 20% N treatments, and averaged about 9% and 20%
tion of flowering group one (Table 3), as well as group two in lower than control in 1999 and 2000, respectively (Fig. 3). Nitro-
control (Table 1). In 2000, warm temperature conditions from gen deficiency (leaf N < 25 g kg−1 ; Gerik et al., 1998) generally
about 21 August to 4 September may have contributed to low had an opposite effect on average lint weight boll−1 , which

Table 4
Effects of nitrogen (N) stress treatments on selected fiber quality characteristics in five flowering groups and for weighted-sum treatment means of cotton grown
outdoors in large pots in 1999 and 2000
N-stress treatment Group one Group two Group three Group four Group five Weighted sum

1999 2000 1999 2000 1999 2000 1999 2000 1999 2000 1999 2000

2.5% span length (mm)

Control 28.2 29.7 28.8 29.4 29.1 28.5 29.6 a 27.6 31.6 28.1 28.5 28.9
20% FF 27.9 29.6 28.4 29.4 29.1 28.4 28.4 a 28.3 28.7 28.2 28.4 29.0
0% FF 27.8 29.9 28.4 29.5 29.3 28.0 26.2 b 27.8 nd nd 28.1 29.3
Year × treatment 0.853 0.833 0.798 0.002 0.073 0.682
Strength (kN m kg−1 )
Control 220.2 205.5 224.2 a 199.9 230.3 a 198.6 221.8 a 186.0 232.4 191.4 b 219.3 a 199.2
20% FF 206.3 218.2 211.3 a 210.5 220.7 a 199.2 207.2 a 200.4 211.8 219.4 a 212.1 b 207.7
0% FF 210.4 210.5 193.2 b 203.8 207.4 b 197.8 189.4 b 201.7 nd nd 198.8 c 204.6
Year × treatment 0.105 0.177 0.233 0.031 0.045 0.009
Micronaire
Control 3.4 b 4.5 4.2 b 4.3 4.4 4.1 4.8 ab 4.0 4.2 3.4 b 4.3 b 4.2
20% FF 4.8 a 4.5 5.3 a 4.2 5.2 4.5 4.9 a 4.1 4.3 4.7 a 5.1 a 4.4
0% FF 4.8 a 4.8 5.2 a 4.8 4.8 3.9 4.2 b 3.7 nd nd 4.8 a 4.6
Year × treatment 0.034 0.022 0.530 0.689 0.024 0.069

Note: Weighted sums were calculated by weighting each fiber property by the amount of lint produced in each flowering group. Control, half-strength nutrient
solution at plant emergence onward; 20% FF, 20% of control N at first flower onward; and 0% FF, 0% of control N at first flower onward. Within a flowering group
and year, means followed by a different letter are significantly different by Fischer’s protected LSD test at P = 0.05. Year × treatment, probability of F-statistic for
year by N treatment interaction. nd, no data.
 J.J. Read et al. / Europ. J. Agronomy 24 (2006) 282–290 287

Table 5
Effects of potassium (K) stress treatments on selected fiber quality characteristics in five flowering groups and for weighted-sum treatment means of cotton grown
outdoors in large pots in 1999 and 2000
K-stress treatment Group one Group two Group three Group four Group five Weighted sum

1999 2000 1999 2000 1999 2000 1999 2000 1999 2000 1999 2000

2.5% span length (mm)

Control 28.2 29.7 28.8 29.4 29.1 28.5 29.6 27.6 31.6 28.1 28.5 29.0
20% FF/FS 28.4 29.4 28.9 28.7 30.1 28.4 29.6 28.0 31.8 29.2 28.3 28.7
0% FF/FS 27.7 29.9 29.1 29.2 30.1 28.0 27.9 27.7 29.2 26.7 28.4 29.1
Year × treatment 0.224 0.452 0.387 0.220 0.787 0.970
Strength (kN m kg−1 )
Control 220.2 205.5 224.2 199.9 230.3 198.6 221.8 186.0 b 232.4 191.4 b 219.3 199.2
20% FF/FS 201.5 223.4 217.2 216.6 221.2 220.6 216.1 210.0 ab 249.1 230.6 a 207.9 219.6
0% FF/FS 221.4 225.9 226.0 219.0 228.5 184.9 218.9 221.6 a 210.4 194.2 b 221.6 212.5
Year × treatment 0.182 0.540 0.132 0.063 0.280 0.075
Micronaire
Control 3.4 b 4.5 4.2 4.3 a 4.4 4.1 4.8 4.0 ab 4.1 3.4 4.3 4.2 ab
20% FF/FS 4.4 a 4.7 4.3 4.3 a 4.1 4.6 4.7 4.7 a 3.2 4.5 4.2 4.6 a
0% FF/FS 4.4 a 4.0 4.3 3.2 b 3.5 3.7 4.4 3.6 b 2.8 4.0 4.0 3.7 b
Year × treatment 0.049 0.021 0.496 0.149 0.089 0.126

Note: Weighted sums were calculated by weighting each fiber property by the amount of lint produced in each flowering group. Control, half-strength nutrient
solution at plant emergence onward; 20% FF/FS, 20% of control K at first flower (1999) or first square (2000) onward; 0% FF/FS, 0% of control K at first flower
(1999) or first square (2000) onward. Within a flowering group and year, means followed by a different letter are significantly different by Fischer’s protected LSD
test at P = 0.05. Year × treatment, probability of F-statistic for year by K treatment interaction. nd, no data.

was significantly greater than that of controls in both years, and greater in 20% N treatment (Table 4). Flowering groups one
greater than 20% N treatment in 2000. In three out of four com- and two, and to a lesser extent group three, contributed to high
parisons between control and 20% N treatments, both lint yield micronaire observed under N stress. In 20% N treatment, values
and weight boll−1 were significantly greater in 20% N treatment. for weighted-sum micronaire and the micronaire of flowering
As compared to control plants, weighted-sum micronaire groups two, three and four exceeded 4.9 (in other words, were
in 1999 was about 12% greater in 0% N treatment and 18%

Fig. 2. Nitrogen concentration in uppermost, fully expanded leaves of cotton Fig. 3. Effects of N stress on lint yield and lint weight boll−1 of mature bolls
grown outdoors in large pots under three N treatments in 1999 and 2000. Values harvested from sympodial branches of cotton grown outdoors in large pots.
represent a single observation on each sampling date, based on a pooled sample Values represent the mean of three observations, and were obtained by first
of five leaves. summing across five flowering groups in each 20-plant replicate.
288 J.J. Read et al. / Europ. J. Agronomy 24 (2006) 282–290

penalty grade). Nitrogen deficiency, however, did not consis-

tently increase fiber micronaire. In 1999, flowering group four
in 0% N treatment had significantly lower micronaire, as well
as fiber length and strength, than 20% N treatment. A similar,
but nonsignificant trend was observed in 2000, when micron-
aire of flowering group four averaged 3.7 in 0% N treatment.
The year by treatment interaction was significant for micronaire
of flowering groups one, two, and five.
The 0% N treatment in 1999 significantly decreased
weighted-sum fiber strength and fiber strength of flowering
groups two, three and four, as compared to control or 20% N
treatment (Table 4). Analysis of variance found a highly sig-
nificant (P < 0.01) effect of year by N treatment interaction for
weighted-sum strength, due to weak fibers under N stress in
1999, but similar fiber strength among treatments in 2000. The
year by treatment interaction was significant for fiber strength
of flowering groups four and five.

3.3. Potassium stress effects

As expected, symptoms of K deficiency, including yellowing

and premature leaf drop, were more pronounced in 2000 than
1999 because K was withheld sooner. In general, cotton is K
deficient when its leaf concentration falls below 15 g kg−1 at
early bloom (Kerby and Adams, 1985). The 20% K treatment Fig. 5. Effects of K stress on lint yield and lint weight boll−1 of mature bolls
lowered leaf K to about 10 g kg−1 in 1999 and 12.5 g kg−1 in harvested from sympodial branches of cotton grown outdoors in large pots.
Values represent the mean of three observations, and were obtained by first
2000 (Fig. 4). Severe K stress was possibly manifested in 0% summing across five flowering groups in each 20-plant replicate.
K treatment in both years, as K in uppermost leaves dropped
below 5 g kg−1 in mid August 1999 and late July 2000.

In 1999, lint yield and lint weight boll−1 did not differ among
treatments, but yields were least in 0% K treatment (Fig. 5).
This trend was highly significant in 2000, when K deficiency
decreased lint yield by about 55% and weight boll−1 by about
20%, as compared to control. Decreased yield of K-deficient
cotton in 2000 was associated with smaller number of bolls in
all flowering groups (Table 3). The year by treatment interaction
was significant for lint yield due to a large K-induced decrease
in 2000 when K was withheld sooner.
Fiber length was not significantly affected by K stress
(Table 5), although 0% K treatment in 1999 decreased fiber
length of flowering group four by about 1.7 mm (P < 0.20), as
compared to control or 20% K treatment. In flowering group
five, K stress decreased fiber length by about 2.5 mm in 1999
(P > 0.38) and by about 1.4 mm in 2000 (P > 0.11), as compared
to controls. Fiber strength of flowering group five decreased
slightly in 1999, and was significantly lower in 0% K than
20% K treatment in 2000. Micronaire decreased significantly in
2000, and values nearly 1.0 unit less than those in 20% N treat-
ment were evident in weighted-sum micronaire and in flowering
groups two and four (Table 4). In 2000, weighted-sum micron-
aire and micronaire of flowering groups two, three, and four
was 3.7 or less. Similarly, micronaire values below 3.7 were
evident in 1999 in flowering groups three and five, but no signif-
Fig. 4. Potassium concentration in uppermost, fully expanded leaves of cotton
grown outdoors in large pots under three K treatments in 1999 and 2000. Values icant treatment differences were detected. These results suggest
represent a single observation on each sampling date, based on a pooled sample severe K deficiency may decrease micronaire when leaf K falls
of five leaves. below 10 g kg−1 for an extended period.
 J.J. Read et al. / Europ. J. Agronomy 24 (2006) 282–290 289

4. Discussion either no change (Minton and Ebelhar, 1991) or decrease (Petti-

grew, 1999) in micronaire under low K fertility. It also disagrees
The present study determined changes in lint yield and fiber with leaf K values that were generally below 10 g kg−1 , a level
quality, due solely to N or K stress, in mature bolls grouped associated with decreased growth and photosynthesis (Kerby
according to week of anthesis. Cotton was grown outdoors in and Adams, 1985; Reddy and Zhao, 2005). Jones and Wells
large pots and provided different levels of N or K in nutri- (1998) reported a positive correlation between micronaire and
ent solution from flowering stage to crop maturity in order to dry weight boll−1 across various sympodial-branch positions
determine the dynamics of stress development and if changes in and plant populations in the field. Similarly, low micronaire was
leaf N or K are related to fiber quality. In agreement with other evident in control in 1999, particularly in flowering groups one
studies (Jenkins et al., 1990; Davidonis et al., 2004), difference and two, and was associated with low average lint weight boll−1 .
between years in boll distribution within the crop had a marked Favorable N or water status appears to result in low micronaire
effect on where (or more specifically when) a significant effect due to shading of lower bolls and leaves, which may reduce the
of nutrient stress was detected in the five flowering groups. The amount of carbohydrate available to mature bolls (Gerik et al.,
inconsistent effects of N and K stress on fiber quality across 1998; Pettigrew, 2001).
years, suggests that plants experienced different levels of stress Similar to results of Reddy et al. (1999) with temperature
because of what happened to boll development in the 2 years and Reddy et al. (2004) with N nutrition, our results of signif-
(Jones and Wells, 1998). Consequently, we were unable to deter- icant year by N treatment interactions indicate fiber strength is
mine quantitatively the relationships between changes in leaf N influenced by environment. Nevertheless, the 0% N treatment
or K and the fiber quality of five flowering groups. The influ- in 1999 often led to significant reductions in fiber strength. Cot-
ence of environment was most evident under N stress in 1999, ton was low or deficient in N following 23 July 1999, but was
as the cohort of bolls with low length, strength and micronaire generally not N-deficient in 2000. Nutrient stress 20–40 days
(flowering group four) also comprised a large fraction of total post-anthesis is expected to impact fiber strength development
lint, and thus placed a large demand on the plant for nutrient (Ramey, 1986; Bradow and Davidonis, 2000). Because fiber
and carbohydrate reserves. Because the N-stress treatments did development became increasingly N-limited in 1999 as nutri-
not differ in yield fraction (Table 2) or boll number (Table 3), ent reserves were utilized, the 0% N treatment produced low
a possible explanation is 0% N treatment depressed photosyn- strength fibers in flowering groups two, three, and four, that also
thesis enough to impact fiber quality (Reddy et al., 2004). Fiber supported a high percentage of the total lint yield. The 2.5% span
quality of flowering group four was not affected by N stress in length was influenced by both N stress and environment. Consis-
2000, when comparatively more lint was produced from earlier tent with evidence that fiber length is associated positively with
flowering groups that presumably escaped the adverse effects of leaf N during boll maturation period (Reddy et al., 2004), N-
N deficiency later in the season. deficient cotton in 1999 (leaf N < 25 g kg−1 ; Gerik et al., 1998)
Similar to Reddy et al. (2004), lint weight boll−1 increased had low fiber length in flowering group four. Similar to Reddy
in cotton provided either 0% or 20% of control N, but lint et al. (1999), the warm temperature conditions evident during
yield was significantly less than control in 0% N treatment the blooming period in 1999 were associated with significantly
only. These results support evidence that decreased yield under lower weighted-sum length of cotton in 1999 than 2000.
N stress is related more to decreased boll number than boll In contrast to N stress, the 0% K treatment from squaring-
weight (Wullschleger and Oosterhuis, 1990). The mechanism stage onward in 2000 led to significant reductions in lint weight
by which the crop partitions its C and N is complex (Boquet boll−1 , which adversely impacted lint yield. The observed reduc-
and Breitenbeck, 2000). Because N-stress limits several growth tions in cotton yield and micronaire in K-deficient cotton in 2000
processes (Reddy et al., 1997; Gerik et al., 1998), increased is consistent with reports that K deficiency causes premature
lint weight boll−1 under stress likely resulted from better light termination of reproductive growth (Pettigrew, 2003), low boll
distribution within the canopy due to reduced leaf area index weight (Kerby and Adams, 1985) and decreased translocation
(Wullschleger and Oosterhuis, 1990; Reddy et al., 2004). of sugars out of the leaf (Pettigrew, 1999). Recently, Reddy and
In 1999, values for weighted-sum micronaire increased to Zhao (2005) reported the critical leaf K for cotton photosynthe-
about 5.1 (in the discount range) when plants were sufficient sis, biomass and stem growth was 12 g kg−1 , and for leaf area
to low in N (20% N treatment) and to about 4.8 when plants expansion the critical value was 17 g kg−1 . In that study, less
were low to deficient in N (0% N treatment). This agrees with biomass partitioning to bolls was due to increased fruit abscis-
reports of increased micronaire as leaf N decreased due to lim- sion under K stress. In the present study, leaf K of 5–10 g kg−1
ited availability of N either in nutrient solution (Reddy et al., observed in the 0% K treatment would have certainly depressed
2004) or field soil (Bauer and Roof, 2004). An overall increase in leaf growth and photosynthesis, and altered biomass partitioning
assimilate supply would help to explain the increase in micron- to various plant components (Reddy and Zhao, 2005). Because
aire of flowering group one in N-deficient plants, particularly fiber length, strength, and micronaire were not adversely affected
if canopy photosynthesis was favored by the warm conditions until K deficiency was severe, there may be little concern that
evident early in the season 1999 (Reddy et al., 1999; Bauer et K deficiency will adversely impact fiber quality in commercial
al., 2000). A large (1.0 unit) increase in micronaire of flower- situations.
ing group one was also observed in both K-stress treatments in An apparent direct effect of K deficiency on lint quality was
1999. This response is inconsistent with field studies showing observed for micronaire (Table 5), with values sometimes at
 290 J.J. Read et al. / Europ. J. Agronomy 24 (2006) 282–290

or below the discount of 3.5 and not strictly related to fibers DeLanghe, E.A.L., 1986. Lint development. In: Mauney, J.R., Stewart,
that developed in the most productive flowering groups. No J.McD. (Eds.), Cotton Physiology. The Cotton Foundation, Memphis, TN,
other fiber property traits were consistently altered by K defi- pp. 325–350.
Deussen, H., 1986. Stressing high strength, low micronaire may require a
ciency. Low micronaire cotton (<3.5) will have a thin cell wall rethinking of breeding and marketing methods. In: Spencer, W. (Ed.),
with a smaller amount of cellulose in the fiber cell. Added K Cotton International, 53rd ed. Meister Publishing Co., Memphis, TN, pp.
appears to increase metabolic processes related to secondary- 32–36.
wall thickening (Bradow and Davidonis, 2000). Pettigrew (1999, Gerik, T.T., Oosterhuis, D.M., Tolbert, H.A., 1998. Managing cotton nitrogen
2003) found plants grown at 0 kg K ha−1 produced lint with low supply. Adv. Agron. 64, 115–147.
Heitholt, J.J., 1994. Supplemental boron, boll retention, ovary carbohy-
micronaire, but values were not less than 3.8. Reductions in over- drates, and lint yield in modern cotton genotypes. Agron. J. 86, 492–
all plant assimilate levels and partitioning to bolls would help 497.
explain the lower micronaire in some flowering groups in the Hewitt, E.J., 1952. Sand and water culture methods used in the study of plant
present study. nutrition, vol. 22. CAB Commonwealth Agricultural Bureau Technology
Despite decades of research on crop N and K management Communication; Farnham Royal, UK, p. 189.
Jenkins, J.N., McCarty Jr., J.C., Parrott, W.L., 1990. Fruiting efficiency in
(Kerby and Adams, 1985; Gerik et al., 1998), few studies have cotton: boll size and boll set percentage. Crop Sci. 30, 857–860.
determined the consequences of limited nutrient availability on Johnson, R.M., Downer, R., Bradow, J.M., Bauer, P.J., Sadler, E.J., 2002.
cotton fiber development in different fruiting zones (Bradow and Variability in cotton fiber yield, fiber quality and soil properties in a
Davidonis, 2000; Reddy et al., 2004). Our results indicated N south eastern coastal plain. Agron. J. 94, 1305–1316.
stress indirectly influenced fiber quality, because N deficiency Jones, M.A., Wells, R., 1998. Fiber yield and quality of cotton grown at two
divergent population densities. Crop Sci. 38, 1190–1195.
in cotton (leaf N < 25 g kg−1 ; Gerik et al., 1998) and a relatively Kerby, T.A., Adams, F., 1985. Potassium nutrition of cotton. In: Robert,
high boll load combined to produce low quality fiber. Nitro- P.C., Rust, R.H., Larson, W.E. (Eds.), Potassium in Agriculture.
gen stress in 1999 produced cotton with low fiber strength and ASA–CSSA–SSSA, Madison, WI, USA, pp. 843–860.
high micronaire. A lack of consistent (N stress) or significant (K Minton, E.B., Ebelhar, M.W., 1991. Potassium and aldicarb-disulfoton effects
on verticillium wilt, yield, and quality of cotton. Crop Sci. 31, 209–
stress) treatment difference in fiber length suggested that early
212.
stages of fiber development were indirectly affected by nutrient Nelson, D.W., Sommers, L.E., 1972. A simple digestion procedure for esti-
stress. Results also indicated that N and K stress in cotton will mation of total nitrogen in soils and sediments. J. Environ. Qual. 1,
likely have opposite effects on fiber micronaire. 423–425.
Pettigrew, W.T., 1999. Potassium deficiency increase specific leaf weights
Acknowledgements and leaf glucose levels in field-grown cotton. Agron. J. 91, 962–
968.
Pettigrew, W.T., 2001. Environmental effects on cotton fiber carbohydrate
The authors would like to thank P.J. Bauer, J.M. Bradow, and concentration and quality. Crop Sci. 41, 1108–1113.
W.T. Pettigrew for their helpful comments on the manuscript, Pettigrew, W.T., 2003. Relationships between insufficient potassium and crop
and Mr. D. Brand, Mr. K. Gourley, and Mr. W. Ladner for their maturity in cotton. Agron. J. 95, 1323–1329.
excellent technical assistance. This study was in part supported Pettigrew, W.T., Heitholt, J.J., Meredith Jr., W.R., 1996. Genotypic interac-
by The National Aeronautical and Space Administration-funded tions with potassium and nitrogen in cotton of varied maturity. Agron. J.
88, 89–93.
Remote Sensing Technology Center at Mississippi State Univer- Pettigrew, W.T., Meredith Jr., W.R., 1997. Dry matter production, nutrient
sity (NASA grant number NCC13-99001). uptake, and growth of cotton as affected by potassium fertilization. J.
Plant Nutr. 20, 531–548.
References Ramey Jr., H.H., 1986. Stress influences on fiber development. In: Mauney,
J.R., Stewart, J.McD. (Eds.), Cotton Physiology. The Cotton Foundation,
Bauer, P.J., Frederick, J.R., Bradow, J.M., Sadler, E.J., Evans, D.E., 2000. Memphis, TN, USA, pp. 315–359.
Canopy photosynthesis and fiber properties of normal- and late-planted Reddy, K.R., Hodges, H.F., McKinion, J.M., 1997. Crop modeling and appli-
cotton. Agron. J. 92, 518–523. cations: a cotton example. Adv. Agron. 59, 225–290.
Bauer, P.J., Roof, M.E., 2004. Nitrogen, aldicarb, and cover crop effects on Reddy, K.R., Davidonis, G.H., Johnson, A.S., Vinyard, B.T., 1999. Tempera-
cotton yield and fiber properties. Agron. J. 96, 369–376. ture regime and carbon dioxide enrichment alter cotton boll development
Boquet, D.J., Breitenbeck, G.A., 2000. Nitrogen rate effect on partitioning and fiber properties. Agron. J. 91, 851–858.
of nitrogen and dry matter by cotton. Crop Sci. 40, 1685–1693. Reddy, K.R., Koti, S., Davidonis, G.H., Reddy, V.R., 2004. Interactive effects
Boquet, D.J., Moser, E.B., 2003. Boll retention and boll size among intrasym- of carbon dioxide and nitrogen nutrition on cotton growth, development,
podial fruiting sites in cotton. Crop Sci. 43, 195–201. yield and fiber quality. Agron. J. 96, 1148–1157.
Bradow, J.M., Johnson, R.M., Bauer, P.J., Sadler, E.J., 1999. Site-specific Reddy, K.R., Zhao, D., 2005. Interactive effects of elevated CO2 and potas-
management of cotton fiber quality. In: Stafford, J.V. (Ed.), Precision sium deficiency on photosynthesis, growth and biomass partitioning of
Agriculture ’99, Proceedings of the Second European Conference on cotton. Field Crops Res. 94, 201–213.
Precision Agriculture. Odense, Denmark, 11–15 July 1999. Sheffield Aca- SAS Institute, 1999. SAS/STAT User’s Guide, Version 8.0, vol. 3. SAS Insti-
demic Press, Sheffield, UK, pp. 677–686. tute, Inc., Cary, NC, USA.
Bradow, J.M., Davidonis, G.H., 2000. Quantitation of fiber quality and the Usherwood, N.R., 2000. The influence of potassium on cotton quality. In:
cotton production-processing interface: a physiologist’s perspective. J. Agri-Briefs, Agronomic News, No. 8., Spring 2000. Potash and Phosphate
Cotton Sci. 4, 34–64. Institute, Norcross, GA, USA.
Davidonis, G.H., Johnson, A.S., Landivar, J.A., Fernandez, C.J., 2004. Cotton Wullschleger, S.D., Oosterhuis, D.M., 1990. Canopy development and pho-
fiber quality is related to boll location and planting date. Crop Sci. 96, tosynthesis of cotton as influenced by nitrogen nutrition. J. Plant Nutr.
42–47. 13, 1141–1154.

View publication stats