Agronomy Journal 95:709-714 (2003)
© 2003 American Society of Agronomy
COTTON
Sprinkler-Induced Flower Losses and Yield Reductions in Cotton (Gossypium hirsutum L.)
John J. Burke*
USDA-ARS, SPA, Plant Stress and Water Conserv. Lab., 3810 4th St., Lubbock, TX 79415
* Corresponding author (jburke{at}lbk.ars.usda.gov)
Received for publication May 8, 2002.
 |
ABSTRACT
|
|---|
Cotton (Gossypium hirsutum L.) pollen is highly sensitive to water, rupturing within 1 to 2 min of contact. Flowers sprayed with known volumes of water showed that a single spray with 1 mL of water reduced seed set by 55%. Additional spray applications resulted in further losses and ultimately flower shedding. Field studies in 2000 and 2001 used a center pivot equipped with sprinklers and drag socks to determine the effect of timing and water application methods on fruit losses. Treatments included different application times (0800 to 1000 h, 1000 to 1200 h, 1200 to 1400 h, and 1400 to 1600 h) with 77 m3 of water. The plots were irrigated eight times from 1 Aug. to 1 Sept. 2000 and 11 times from 10 July to 21 Aug. 2001. Flowers were tagged before irrigation and tracked throughout the season. No significant differences in fruit losses were observed in the 0800- to 1000-h treatment; however, fruit losses progressively increased under the sprinkler treatments compared with the drag sock treatments as the day progressed. Maximum fruit losses occurred in the 1400- to 1600-h treatments. The time course of sprinkler-induced fruit losses tracked the opening of the flower and dehiscence of the pollen. Evaluation of seasonal lint yields showed a 21% reduction in 2000 and 11% yield reduction in 2001 under sprinkler irrigation compared with drag sock irrigation. These results emphasize the need to use production practices that limit water contact with open cotton flowers.
|
INTRODUCTION
|
|---|
ORTON AND DUNCAN (unpublished, 1908) first reported studies evaluating water effects on flower shedding in cotton (G. hirsutum L.). Following spraying freshly opened flowers with an atomizer and comparing boll set with nonsprayed flowers, they observed only a 7% greater loss of treated flowers. These results, in combination with the observation that a 8.6-mm rainfall on 4 Aug. 1906 was not followed by any increase in shedding until 6 August when an additional light rain fell, led them to discount the importance of water on shedding of cotton flowers. Lloyd (1920) showed that the application of water to 20 cotton (G. herbaceum L) flowers at 1100 h on 21 August resulted in the shedding of 13 of the fruiting structures between 26 and 29 August. Lloyd (1920) reported that pollen grains burst not only in water, but also in 50% cane sugar solution, and he concluded that the wetting of the pollen by rain would destroy the pollen. Analyses of the timing of rainfall on flower shedding led Lloyd to conclude that "in interpreting our records of boll shedding therefore, we are compelled to ascribe importance, in some instances a very great deal of importance, to rainfall, but less to its amount than to the time at which it occurs."
Pearson (1949) analyzed daily fluctuations in humidity, rainfall, and temperature from successive days of flowering in relation to flower set and mote formation. She concluded that rain during the time of pollination, even that falling shortly thereafter, affected the number of ovules receiving pollen tubes by interfering either with pollen deposition on the stigma or with pollen tube growth. Rains occurring on 5 Aug. 1937 and 3 Aug. 1939 came at a critical time and in sufficient amounts to impact boll set. Thirteen bolls were set on 5 Aug. 1937 compared with 182 for 4 August and 233 for 6 August. For 3 Aug. 1939, 13 bolls were set compared with 80 on 2 August and 48 on 4 August (Pearson, 1949). Pearson stated that the amount of shedding for days in which heavy rains occurred before noon would usually be so extensive that a large mote production for the relatively few bolls matured would have little influence on the total mote production of the entire crop. Additional reports of heavy dropping of young bolls occurring 4 to 5 d after raining was reported by King et al. (1956), presumably because the processes of pollination were prevented by rain after the opening of the flowers. The destructive effects of rain continued from morning to 1600 h of the rainy day. In addition, it was found that boll shedding induced by spraying of liquid insecticide before noon was more serious than that from afternoon sprays and the untreated control. Clearly, these findings suggest an extreme sensitivity of cotton flowers to moisture at the time of pollination. In a review of the physiological aspects of flower and fruit set in cotton, Stewart (1986) described the problems researchers faced when studying in vitro germination and tube growth of cotton pollen. He reported that studies of cotton pollen had been limited by the extreme sensitivity of cotton pollen to moisture. Whenever the grains or pollen tubes contacted available water, they ruptured.
Pennington and Pringle (1987) reported that morning rainfall or sprinkler water applications on open cotton flowers in the Mississippi Delta would cause 65% of the wetted flowers to shed. Open flowers wetted by sprinkler irrigation in the morning experienced a substantial reduction in boll retention compared with flowers watered predawn or very late in the afternoon. They reported that starting in late morning, boll retention gradually increased through the afternoon. Based on yield data from fields following hand removal of flowers one, two, or three times a week for approximately 5 wk, Pennington and Pringle (1987) predicted yield losses in production fields of only 1 or 2% by watering twice a week as more energy would be available for other bolls remaining on the plant.
The present study re-evaluated the effects of sprinkler irrigation on flower retention and yield reductions because of concerns that the manual removal of open flowers described in the Pennington and Pringle (1987) study failed to adequately mimic what happens with water emasculation of cotton flowers. Water-emasculated flowers take 1 to 2 wk to fall from the plant (Lloyd, 1920; Pearson, 1949); therefore little excess energy would be available for remaining bolls to compensate for the removed flowers. Second, we wanted to evaluate sprinkler-induced yield reductions on the High Plains of Texas where the growing season is shorter and night temperatures cooler than in the Mississippi Delta where the study by Pennington and Pringle (1987) was performed.
|
MATERIALS AND METHODS
|
|---|
Greenhouse Plant Culture Protocols
Seeds of cotton cultivar Gregg 65 were planted into hydroponic rock wool slabs (15 by 90 by 8 cm, width by length by diameter) [Fibrgro Horticultural Rock Wool, Fiberx Insulations, Sarina, ON, Canada] that had been saturated with Peters Professional water soluble fertilizer [0.95 g L-1 5–11–26 HYDRO-SOL (Scotts-Sierra Hortic. Products Co., Maryville, OH), supplemented with 0.475 g L-1 calcium nitrate (Hydro Agri North America, Tampa, FL) and 0.238 g L-1 magnesium sulfate (Scotts-Sierra Hortic. Products Co., Maryville, OH)]. Three seeds were planted per pad, a total of 20 hydroponic rock wool slabs were placed on benches in a greenhouse, and nutrients were maintained using a nonrecycling hydroponic watering system. Plants were grown under greenhouse conditions (28 ± 5°C air temperature) for 120 d under natural-light conditions. Additional plantings occurred monthly to ensure a constant supply of flowers.
Determination of the Quantity of Water Spray that Reduces Seed Set
Small plastic hand-pump spray bottles were obtained from a local store, and experiments were performed to determine the quantity of water released during a single fine spray. The nozzle was placed inside the opening of a graduated cylinder and the amount of water released per spray determined. A bottle that emitted 1.0667 mL per spray was identified and used in subsequent experiments. In the winter and spring of 2000, three open flowers of greenhouse-grown cotton were sprayed one time, a second set of three flowers sprayed twice, and so on up to 15 repetitive sprays with the spray bottle on 26 February; 4 and 25 March; 29 April; 10, 16, and 30 May; and 10 and 26 June. The flowers were tagged and the fate of the fruiting structure followed until either the boll matured or fell from the plant. Bolls were harvested on maturity and changes in seed set determined.
2000 Field Experiment
Cotton cultivar PM 2326 RR was planted on 1-m centers on 19 May 2000 in Lubbock, TX. The field had been treated with 0.287 L ha-1 Treflan (
,,-Trifluoro-2,6-dinitro-N,N-dipropyl-p-toluidine) on the same day. Following germination and emergence of the cotton, alleyways were cut to provide paired 15-m test plots for each time-of-day analysis. The first 15 m of plot was used to tag flowers throughout the growing season while the second 15-m plot was used for yield studies at the end of the growing season (Fig. 1)
. Nine kilograms per hectare of 28–0–0 N–P–K fertilizer was applied on 3 July 2000. Pix (1,1-dimethylpiperdinium chloride) was applied at a rate of 138 g ha-1 on 7 August to control plant height. The field was sprayed for insects on 6, 12, and 24 July and on 21 August. Initially, all plots were irrigated with sprinkler drops from a center pivot until first flowering. After the appearance of the first flower, irrigations were applied on 1, 4, 8, 11, 15, 23, and 29 August and on 1 September via a center pivot equipped with drag socks and sprinkler drops randomly separated into three replications for each application method (Fig. 1).
View larger version (66K):
[in this window]
[in a new window]
|
Fig. 1. Aerial photograph showing the field layout and timing of the irrigation as the center pivot moved around the field (A) and sketch of the experimental design (B). The alleys separating treatments can be seen in A, and labels have been added identifying the location of the sprinkler (S) and drag sock (D) treatments in B.
|
|
Three sets of 25 flowers were tagged for both the sprinkler and drag sock irrigation protocols before each of the irrigation treatments (0800 to 1000 h, 1000 to 1200 h, 1200 to 1400 h, and 1400 to 1600 h). Tagged flowers were tracked throughout the experiment, and tags that fell when fruiting forms were shed were picked up on a weekly basis. Tagged bolls remaining on the plants at the end of the season were harvested and analyzed for lint and seed production. Yield measurements were made from 6 m of row harvested from each treatment replication. Lint quality was analyzed according to procedures used in cotton classing offices.
2001 Field Experiment
Cotton cultivar PM 2326 BG/RR was planted on 1-m centers on 11 May 2001 in Lubbock, TX. The field had been pretreated with Prowl [N-(1-ethylpropyl)-3,4-dimethyl-2,6-dinitrobenzenamine] at a rate of 10.287 L ha-1 on 10 April. Nine kilograms per hectare of 32–0–0 N–P–K fertilizer was applied on 25 May. Pix was applied at a rate of 106 g ha-1 on 23 July to control plant height. The field was sprayed for insects on 18 and 21 June, 11 July, and 2 and 13 August. Following germination and emergence of the cotton, alleyways were cut to provide paired 15-m test plots as described for the previous year. After the appearance of the first flower, irrigations were applied on 10, 13, 17, 20, 24, 27, and 31 July and on 7, 10, 17, and 21 August. Irrigation (77 m3 per irrigation) was applied from the time of planting via a center pivot equipped with drag socks and sprinkler drops randomly separated into three replications for each application method. Three sets of 25 flowers each were tagged for the sprinkler and for the drag sock irrigation protocols before each of the 0800- to 1000-h, 1000- to 1200-h, 1200- to 1400-h, and 1400- to 1600-h irrigations. Tagged flowers were evaluated throughout the experiment, and tags that fell when flowers were shed were picked up on a weekly basis. Tagged bolls remaining on the plants at the end of the season were harvested and analyzed for lint and seed production. Yield measurements were made from 2- to 3-m rows harvested from each replication and irrigation time treatment. Plant numbers and heights were obtained during the yield harvest.
|
RESULTS AND DISCUSSION
|
|---|
This study evaluated the fate of cotton flowers saturated with water from sprinklers located at the top of the canopy and the impact of the timing of the irrigation on fruit losses and yield reductions compared with irrigation using drag socks to apply the water at the soil surface.
Greenhouse Study
Initial studies evaluated water emasculation of cotton flowers from greenhouse-grown cotton. Flowers were sprayed with 1 mL of water per spray from a hand-held plastic spray bottle. Figure 2A shows the decline in seed production per boll as the number of sprays increased. A single spray resulted in a loss of over 50% of the seed set. Increasing the number of sprays resulted in further declines in seed set and ultimately fruit shedding. These results are similar to those of King et al. (1956), who reported that insecticide treatment of cotton when flowers were open and pollen had dehisced resulted in fruit losses. We have observed that flowers in the greenhouse open more fully and hold water for a shorter period of time than flowers in cotton fields on the High Plains of Texas. This observation is most likely temperature related and suggests that even greater losses in seed set and fruit shedding would occur with a small amount of water in the flower in the field as it would remain in the cupped flower for a longer period of time. The photograph shown in Fig. 2B illustrates the cupped nature of the flowers in the field and their ability to hold enough water to completely immerse the anthers (Fig. 2C).
View larger version (68K):
[in this window]
[in a new window]
|
Fig. 2. Graph showing the decline in seed-set following the spraying of open cotton flowers one or more times with water from a spray bottle (A). Photograph of cotton flower in the field (B) and a flower after passage of sprinkler irrigation (C).
|
|
Field Studies
Having identified water-induced fruit losses in the greenhouse, we next evaluated water-induced fruit losses and yield reductions in plots watered with either sprinkler or drag sock nozzles on a center pivot. Figure 3
shows the changes in boll retention determined from the tagged flowers for the sprinkler and drag sock treatments in 2000 and 2001. No significant differences between irrigation methods were observed in the 0800- to 1000-h treatments in either 2000 or 2001. A decline in boll retention was observed in the 1000- to 1200-h sprinkler treatment, with further declines continuing in the 1200- to 1400-h and 1400- to 1600-h sprinkler treatments. Boll retention was relatively constant across all times in the drag sock treatments for both 2000 and 2001 experiments. Pennington and Pringle (1987) also observed fruit shedding in the Mississippi River Delta following irrigation with a center-pivot sprinkler. Maximum loss (35% retention) was induced by a 0900-h irrigation, and then fruit retention increased with progressively later irrigation to an 80% retention following irrigation at 1600 h. One explanation for the differences between our results and those of Pennington and Pringle (1987) is that the warm night temperatures in the Mississippi Delta enhance flower opening and pollen dehiscence earlier in the day compared with the High Plains of Texas. The photographs shown in Fig. 4
provide representative examples of the time course of flower opening on the High Plains of Texas. Flowers begin to open between 0800 and 1000 h, with only a small opening of the petals detectable. Between 1000 and 1200 h, the petals slowly open, becoming fully open by 1200 h. The pattern of flower opening coincides with the relative sensitivity of the flower to water (Fig. 3). These data further suggest that the time of day that cotton flowers are susceptible to water coincides with opening of the petals and dehiscence of the pollen.
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 3. Graph showing changes in boll set following sprinkler irrigation (black bar) at one of 4 times of day in 2000 and 2001 compared to drag sock irrigation (gray bar).
|
|
View larger version (190K):
[in this window]
[in a new window]
|
Fig. 4. Photograph of representative flowers showing the time course of opening on the High Plains of Texas.
|
|
Cotton flowers open at multiple angles to the water stream from the sprinkler and, therefore, may be fully saturated with water or may have only a portion of the anthers contacted by water. If only a portion of the anthers contacts the water, then partial fertilization of the flower may occur and misshapen bolls would be evident. Figure 5
shows photographs of an array of misshapen bolls occurring within a 15-m length of row in the middle of one of the 1200- to 1400-h sprinkler-irrigated yield plots. Figure 5A shows a normal (left) and misshapen (right) boll located on a single plant. Figure 5B through 5E shows a range of misshapen bolls ranging from one to two unfertilized loculi (Fig. 5D) to mummified bolls (Fig. 5E). When the partially fertilized bolls mature, the lack of seed and lint development within individual loculi is apparent (Fig. 5F–5I). Each photograph in Fig. 5F through 5I shows a normal (left) and partially unfertilized (right) boll located on a single plant. The photographs illustrate the trend for one or more loculi to be devoid of seed and fiber, further suggesting a lack of fertilization as opposed to inhibition of development of fertilized ovules. To further evaluate this observation, the number of seeds per boll and the amount of lint per seed were analyzed for all time and irrigation treatments. Figure 6A
illustrates the number of seeds per boll, and Fig. 6B shows the amount of lint per seed for the two water application methods at the four times of day in 2001. Seed number per boll declined in the sprinkler treatments as irrigation timing progressed from the morning to afternoon. Although seed number per boll declined in those bolls remaining on the plant following sprinkler irrigation, lint per seed across all treatment and times was essentially equal. This finding supports the suggestion that the misshapen bolls observed in the sprinkler treatments were the result of a lack of fertilization and not a stress-induced reduction in lint development per seed.
View larger version (154K):
[in this window]
[in a new window]
|
Fig. 5. Photographs of misshapen bolls related to sprinkler irrigation of open flowers (A-E) and open bolls (F-I) showing reduced seed number related to sprinkler irrigation of open flowers.
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 6. Graph showing the number of seeds per boll (A) and the amount of lint per seed in bolls (B) from plants irrigated with either a sprinkler (black bar) or drag sock (open bar) at one of 4 times of day in 2001.
|
|
To further evaluate possible water stress–related responses occurring in the sprinkler treatment, plant heights were measured at harvest. Figure 7
shows the plant height for the two water application methods at the four times of day in 2001. Plant height decreased as the time of irrigation was later in the day. The greatest plant height (90 cm) occurred in the 0800- to 1000-h treatment and declined to approximately 55 cm in the 1400- to 1600-h treatment. Although plant height decreased with the time of irrigation, plant heights were similar for each irrigation method within a given time treatment. One possible explanation for the decline in plant height as the timing of irrigation progressed into the day is a water stress induced by the application of cold irrigation water at a time of high vapor pressure deficit. Earlier reports by Bolger et al. (1992) showed that as cotton root temperature declined, hydraulic conductivity also declined. Their results suggest that decline in root hydraulic conductivity resulting from the irrigation water lowering the soil temperature could result in inadequate movement of water from the soil to the plant. This would result in a temporary water stress condition until the root temperature increased. Essentially identical plant heights were observed in the 1000- to 1200-h treatment, yet significant loss of boll retention was observed in the sprinkler treatment. These data further support the conjecture that the decline in boll retention is associated with water emasculation and is not a result of water stress.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 7. Graph showing plant height of plants irrigated with either a sprinkler (black bar) or drag sock (open bar) at one of 4 times of day in 2001.
|
|
Yield measurements were evaluated for each irrigation treatment and time of irrigation application. Seasonal lint yields showed an overall reduction of 21% in 2000 and 11% reduction in 2001 under sprinkler irrigation compared with drag sock irrigation. Similar yields were observed in the sprinkler and drag sock treatments in the 0800- to 1000-h plots in both 2000 and 2001 (Fig. 8)
. Yield reductions were observed for sprinkler irrigation compared with drag sock irrigation in the 1000- to 1200-, 1200- to 1400-, and 1400- to 1600-h plots in 2000. Yield reductions also were observed for sprinkler irrigation compared with drag sock irrigation in the 1000- to 1200- and 1200- to 1400-h treatments in 2001. Similar yields were observed for the sprinkler and drag sock irrigations in the 1400- to 1600-h plots in 2001. The decline in lint production in the 1400- to 1600-h drag sock plot did not follow the pattern of boll loss seen in the tagged plots. The reason for this decline is not apparent and may reflect spatial variability within the field or a heat stress response associated with continued elevated temperatures until the irrigation arrived late in the afternoon. Clearly, the yield reductions in the 1000- to 1200- and 1200- to 1400-h sprinkler plots alone warrant a change in irrigation practices to avoid getting water in the flowers.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 8. Graph showing lint yields for plants irrigated with either a sprinkler (black bar) or drag sock (open bar) at one of 4 times of day in 2000 and 2001.
|
|
In summary, the greenhouse experiment of this study showed that small amounts of water can have deleterious effects on boll retention if applied to cotton flowers following pollen dehiscence. The field studies supported the observation that open cotton flowers were sensitive to water resulting in subsequent fruit shedding or development of partially fertilized bolls. Yield measurements showed 21 and 13% reductions under sprinkler irrigation compared with the drag sock treatment. The observed yield losses can be overcome with minimal modification of production practices. Possible modifications include (i) sprinkler irrigation when cotton flowers are not open, (ii) reducing the frequency of sprinkler irrigation by applying more water per irrigation, or (iii) irrigation using methods that do not expose the cotton flowers to direct contact with water.
|
ACKNOWLEDGMENTS
|
|---|
The author thanks Thomas D. Valco, Kent Wood, Alan Brashears, Jacob Sanchez, Will Frederick, Halee Hughes, Chris Huff, Trent Easter, DeeDee Laumbach, Michelle Marks, Felisia Picon, Tammi Sheikh, and J.R. Quilantan for their excellent assistance. Mention of a commercial or proprietary product does not constitute an endorsement by the USDA. The USDA offers its programs to all eligible persons regardless of race, color, age, sex, or national origin. This research was sponsored in part by funding from Cotton Incorporated (TX805-00).
|
REFERENCES
|
|---|
- Bolger, T.P., D.R. Upchurch, and B.L. McMichael. 1992. Temperature effects on cotton root hydraulic conductance. Environ. Exp. Bot. 32:49–54.
- King, C.C., Y.W. Tang, and T.S. Ni. 1956. Studies on the boll shedding of unfertilized ovaries in cotton plant. Acta Bot. Sin. 5:77–101.
- Lloyd, F.E. 1920. Environmental changes and their effect upon boll-shedding in cotton (Gossypium herbaceum). Ann. NY Acad. Sci. 29:1–131.
- Pearson, N.L. 1949. Mote types in cotton and their occurrence as related to variety, environment, position in lock, lock size and number of locks per boll. USDA Tech. Bull. 1000.
- Pennington, D.A., and H.C. Pringle III. 1987. Effect of sprinkler irrigation on open cotton flowers. p. 69–71. In Proc. Beltwide Cotton Prod. Res. Conf., Dallas, TX. 4–8 Jan. 1987. Natl. Cotton Counc. of America, Memphis, TN.
- Stewart, J. McD. 1986. Integrated events in the flower and fruit. p. 261–300. In J.R. Mauney and J. McD. Stewart (ed.) Cotton physiology. The Cotton Foundation, Memphis, TN.
This article has been cited by other articles:
|
|
|
|
 
J. R. Whitaker, G. L. Ritchie, C. W. Bednarz, and C. I. Mills
Cotton Subsurface Drip and Overhead Irrigation Efficiency, Maturity, Yield, and Quality
Agron. J.,
November 7, 2008;
100(6):
1763 - 1768.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
V. G. KAKANI, K. R. REDDY, S. KOTI, T. P. WALLACE, P. V. V. PRASAD, V. R. REDDY, and D. ZHAO
Differences in in vitro Pollen Germination and Pollen Tube Growth of Cotton Cultivars in Response to High Temperature
Ann. Bot.,
July 1, 2005;
96(1):
59 - 67.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
