Effect of irrigation termination times on cotton cultivars with contrasting maturities

Abstract

Optimizing irrigation termination time can save water and preserve lint yield and fiber quality in cotton. Although there is currently an irrigation termination recommendation in Georgia, the ideal time for termination could be influenced by cultivar differences in maturity. The hypotheses of this study were that differential irrigation termination times will affect lint yield, fiber quality, and incidence of hardlock and boll rot, and responses will be dependent on cultivar differences in maturity. In 2021 and 2022, a study was conducted in Camilla, GA, using two cotton cultivars with contrasting maturities under four irrigation termination treatments. Irrigation was terminated at cutout, first open boll, 2 weeks after first open boll, and 4 weeks after first open boll. Measurements included plant growth, cutout date, lint yield, irrigation water use efficiency (IWUE), fiber quality, and incidence of boll diseases. Gas exchange measurements and percent open boll estimates at each irrigation termination time were also conducted. Terminating irrigation at cutout did not significantly affect yield in either year; however, IWUE increased 12.6% relative to current recommendations and 13.2% relative to the latest termination time in the 2022 season. Cultivars differed significantly in cutout date, agronomic maturity, gas exchange rates, yield, hardlock/boll rot incidence, and fiber quality. However, there was no interaction between cultivar and irrigation termination time for any parameter. We conclude that irrigation can be terminated at cutout to maximize IWUE, for early and late-maturing cultivars, without limiting yield or fiber quality, assuming a water-replete soil profile at termination.

Abbreviations

A_N

net photosynthesis

DAP

days after planting

ET

evapotranspiration

ET₀

reference evapotranspiration

FFB

first fruiting branch

g_s

stomatal conductance

HU

heat units

IWUE

irrigation water use efficiency

NAWF

nodes above the uppermost first position white flower

PAR

photosynthetically active radiation

RZSWD

root zone water deficit

UHB

uppermost harvestable boll

1 INTRODUCTION

Cotton (Gossypium hirsutum L.) is an important row crop in the southern United States. Specifically, the United States produces approximately 4.5 billion kg (20 million bales) of cotton annually, equivalent to a $7 billion value. Furthermore, Georgia generally accounts for approximately 10% of the national total, making it the second largest cotton-producing state in the country (Raper et al., 2019). Although irrigation is commonly used in the southeastern United States to supplement rainfall and prevent exposure to yield-limiting drought, over-irrigation can lead to unnecessary costs for producers and negative impacts on the environment (Chalise et al., 2022; Ermanis et al., 2021; Liu et al., 2022; Plumblee et al., 2021). Consequently, implementing efficient irrigation management practices is an important way to conserve freshwater resources while maximizing yields and fiber quality (Sadler et al., 2005; Schaefer et al., 2018; Snowden et al., 2013).

Apart from the resource conservation benefits of maximizing water use efficiency through efficient irrigation management, there has been evidence that excessive irrigation can negatively affect yield and fiber quality in field-grown cotton (Chalise et al., 2022; Chohan et al., 2020; Ermanis et al., 2021; Himanshu et al., 2021). Over-watering causes an increase in vegetative growth, which results in low fruit retention and declines in harvest index, meaning lint yield is low relative to the total biomass of the plant (Ermanis et al., 2021). Losses in yield from overwatering can also be related to increased disease pressures (Chohan et al., 2020). Importantly, later in the growing season, dense canopies limit airflow, creating humid microenvironments, which favor pathogen growth (Bowen et al., 2018; Hu, 2018). Boll rot and hardlock have been shown to increase when excessive rain or irrigation occurs as the bolls are opening and throughout the late season (Dodds, 2018; Plumblee et al., 2021). In order to maximize returns from irrigation inputs, different irrigation scheduling tools are utilized to define the timing of irrigation events. These tools include methods to estimate crop and environmental factors, such as evapotranspiration, soil water status, and temperature of the canopy (Chastain, Snider, and Collins et al., 2016b; Francesca et al., 2010; Subedi & Chávez, 2015).

Accurately defining the time during the growing season at which irrigation is terminated presents opportunities for increased water savings near the end of the season (Plumblee et al., 2021). To determine the optimal time to terminate irrigation in cotton, one could utilize the developmental stages of the crop as a guide. Strategies utilized to track development include using calendar dates, heat unit accumulation, and visual staging of crop development (Himanshu et al., 2021; Lascano et al., 2017; Plumblee et al., 2021). In the High Plains Region of Texas, Ale et al. (2020) investigated the optimum irrigation termination periods for full and deficit irrigation strategies in dry, normal, and wet years using the DSSAT CSM CROPGRO-Cotton model. Their criteria for what constituted an optimum time to terminate irrigation was when irrigation water use efficiency (IWUE) and yield were simultaneously at the highest point. Based on the model simulations, to maximize yield and IWUE, they suggested the last week of August (111 days after planting [DAP]), first week of September (118 DAP), and second week of September (125 DAP) in wet, normal, and dry years, respectively. Similarly, Himanshu et al. (2021) used the CROPGRO—cotton model in the same region to re-evaluate irrigation termination times using evapotranspiration (ET)-replacement irrigation strategies. With an 85% ET-replacement strategy, they concluded the optimum irrigation termination time was during the last week of August (111 DAP) for wet and normal years and the first week of September (118 DAP) for dry years (Himanshu et al., 2021). Also, in West Texas, a 4-year, drip irrigation study reported no yield benefit from irrigating after 400 accumulated heat units (HU) following cutout (Multer & Sansone, 2007). Cutout is defined as the cessation of new vegetative growth and the end of the effective flowering period (Bange & Milroy, 2000; Chaudhry & Guitchounts, 2003; Guinn, 1985; Gwathmey et al., 2016). Lascano et al. (2017) found that lint yield was highest when irrigation was stopped at 890°C HU after planting, which was 84 DAP (Lascano et al., 2017). In Mississippi, Plumblee et al. (2021) showed that there were no negative effects on lint yield or fiber quality if the last irrigation occurred at cutout when using furrow irrigation. From the literature above, the two studies that defined termination dates from DAP found similar results that would indicate the optimal time to end irrigation would be from 111 DAP to 125 DAP (Ale et al., 2020; Himanshu et al., 2021). Differences in results between each of the studies mentioned could be a result of site-specific differences in rainfall, soil type, cultivar selection, or other management practices that affect growth and maturity.

Core Ideas

• Defining early irrigation termination times that do not penalize yield in cotton can increase water use efficiency.
• Terminating irrigation at cutout maximized water use efficiency in an early and a late maturing cultivar.
• Cultivar affected yield, fiber quality, water use efficiency, and disease incidence.

The current recommendation in Georgia for terminating irrigation in cotton is when there are approximately 10% of all bolls open across a field (Hand et al., 2022). However, cotton varieties can vary substantially in the timing of crop maturity, which could affect the optimal time for irrigation termination in cotton. For example, early maturing cotton varieties are characterized by early flowering, early cutout, and a more compact fruit distribution with bolls concentrated on lower mainstem nodes; these varieties also reach agronomic maturity earlier in the year than “full season” varieties (Bange & Milroy, 2000). Because of differences in maturity, one may expect to see differences in response to late-season irrigation. With more bolls open, the early maturing variety may experience more boll rot or hardlock if irrigation is continued longer than necessary (Chohan et al., 2020; Srivastava et al., 2010). However, if irrigation is terminated too soon, later maturing cultivars may experience yield or fiber quality-limiting stress during critical stages of boll development at higher mainstem nodes (McConnell et al., 1999; Reeves, 2012).

Although, the University of Georgia has clearly defined thresholds for determining when to terminate irrigation (Hand et al., 2022), the extent to which cultivar differences in maturity affect cotton yield and fiber quality responses to end-of-season irrigation has received minimal attention in Georgia. We hypothesized that differential irrigation termination times will affect lint yield, fiber quality, and incidence of hardlock and boll rot, and responses will be dependent on cultivar differences in maturity. The objective of this research was to evaluate maturity, lint yield, fiber quality, and plant stress for two cotton cultivars with contrasting maturities exposed to four end-of-season irrigation treatments. Results of this experiment will aid Georgia cotton producers with late-season irrigation decisions.

2 MATERIALS AND METHODS

2.1 Study site and plant material

This study was conducted during the 2021 and 2022 growing seasons at the University of Georgia's C.M. Stripling Irrigation Research Park (SIRP) near Camilla, GA (31°16′55.5″ N, 84°17′39.9″ W). The soil at this site is a Lucy loamy sand (loamy, kaolinitic, and thermic Arenic kandiudults; NRCS, 2023), and before planting, the fields were strip-tilled both years following the termination of a winter rye cover crop. Cotton seeds (cv. PHY 332 W3FE and PHY 580 W3FE) were sown on May 6,2021 and April 28, 2022. These two cultivars were selected from the same seed supplier because PHY 332 W3FE exhibits early-mid maturity and PHY 580 W3FE exhibits true full-season maturity (Corteva, Inc., 2023).

2.2 Design and treatment structure

The experiment was arranged as a split-plot design with six replications. Irrigation termination time represented the whole-plot factor, and there were four different irrigation termination treatments (timings): Treatment 1 (T1) = irrigation terminated at cutout (physiological maturity); Treatment 2 (T2) = irrigation terminated at first open boll (when open bolls were first observed in all plots); Treatment 3 (T3) = irrigation terminated at 2 weeks after first open boll; and Treatment 4 (T4) = irrigation terminated at 4 weeks after first open boll. The sub-plot factor was cultivar, and there were two different cultivars chosen for this experiment. These cultivars were specifically chosen to determine the broad applicability of irrigation termination times to cultivars with known differences in maturity dates.

Each plot was four-rows wide and 12.8-m long, with 0.91 m of inter-row spacing. The seeds were planted at a rate of 9.8 seeds per linear meter and a depth of 2.5 cm. Average plant densities for the two growing seasons were 8.98 plants m⁻² in 2021 and 7.37 plants m⁻² in 2022, which is above the minimum plant density (6.56 plants m⁻²) recommended by the University of Georgia (Hand et al., 2022). Other than irrigation, all management decisions, including pesticide applications, fertilization, and plant growth regulation, were carried out for all plots according to recommendations from the University of Georgia Cooperative Extension Service (Hand et al., 2022). All treatments were maintained well-watered via overhead sprinkler irrigation until irrigation termination treatments were imposed. Irrigation events were triggered according to recommendations from the Smart Irrigation Cotton App (App; Vellidis et al., 2016). This method calculated daily reference evapotranspiration (ET₀) from available meteorological data according to the FAO 56 method as described in Vellidis et al. (2014). Daily crop evapotranspiration (ET_c) was calculated by multiplying ET₀ by a crop coefficient (k_c) extracted from an empirically derived k_c curve that is based on accumulated heat units rather than DAP (Vellidis et al., 2016). The App maintains a daily soil water balance by subtracting daily ET_c and adding precipitation or irrigation. The difference between the soil profile's field capacity and current plant available soil water is the root zone water deficit (RZSWD). At planting, the assumed effective root zone depth is 15 cm. The depth increases in daily increment until it reaches a maximum of 61 cm. Irrigation was triggered when RZSWD = 16.2 mm. This soil water deficit was chosen because the maximum amount of water that could be provided at one time by the irrigation system without causing runoff was 19.0 mm, and the overhead sprinkler irrigation system is known to have an application efficiency of 85% (Chalise et al., 2022). Weather data (maximum daily temperature, minimum temperature, and precipitation) for the 2021 and 2022 growing seasons are provided in Table 1.

TABLE 1. Total irrigation, season-long rainfall, and average minimum (T_min) and maximum (T_max) temperature during the growing season.

Year	Treatment	Irrigation (mm)	Rainfall (mm)	Average Tmax (°C)	Average Tmin (°C)	PTR (mm)	Consecutive rain-free days	Time interval (days)
2021	T1	188.0	720.6	31.67	20.00	41	5	8
	T2	188.0	720.6	31.67	20.00	66	4	14
	T3	188.0	720.6	31.67	20.00	32	6	14
	T4	197.0	720.6	31.67	20.00	0	0	0
2022	T1	294.5	541.2	31.98	19.67	80	9	27
	T2	379.7	541.2	31.98	19.67	62	4	14
	T3	411.5	541.2	31.98	19.67	13	9	14
	T4	430.5	541.2	31.98	19.67	0	0	0

• Note: Additionally, post-termination rainfall (PTR), the maximum number of consecutive rain-free days, and time interval between each irrigation termination time and the next for four irrigation treatments [terminating irrigation at cutout (T1), first open boll (T2), 2 weeks after first open boll (T3), and 4 weeks after first open boll (T4)] during the 2021 and 2022 growing seasons at C.M. Stripling Irrigation Research Park near Camilla, GA.

2.3 In-season data collection

Beginning at the first flower stage of crop development (when approximately 50% of the crop has started flowering), bi-weekly measurements of plant growth and development were collected. Measurements included the number of mainstem leaf nodes above the uppermost, first position white flower (NAWF). Data were collected from five plants that were randomly selected in each plot, and average values were calculated for each plot prior to analysis. These measurements allowed us to quantify the number of days required to reach cutout. Specifically, NAWF was plotted versus DAP, and a linear function was fit to the resulting data (Figure 1). This function was used to estimate the DAP at which NAWF = 3. A NAWF value of 3 was used to define the timing of cutout based on previous research conducted in Georgia (Bednarz & Nichols, 2005). Cutout represents the cessation of new vegetative growth and the end of the effective flowering period in cotton (Bange & Milroy, 2000; Chaudhry & Guitchounts, 2003; Guinn, 1985; Gwathmey et al., 2016). During the 2022 season, at each termination time, five plants were also randomly chosen from each plot, and the total number of harvestable bolls (bolls greater than 2.54 cm in diameter) and open bolls were counted for each plant. Percent open boll (also a measure of agronomic maturity; Gwathmey et al., 2016) was calculated by dividing open bolls by the total number of bolls and multiplying this fraction by 100.

FIGURE 1

Visual example of the method utilized to derive cutout date. Nodes above the uppermost, first position white flower (NAWF) is represented on the Y-axis and days after planting (DAP) is represented on the X-axis. In the example given above, cutout date, where NAWF = 3, is 92 DAP.

2.4 Gas exchange measurements

Because single-leaf measures of net photosynthesis and stomatal conductance are commonly used to assess drought stress in cotton (Chastain et al., 2014; Chastain, Snider, and Choinski et al., 2016; Meeks et al., 2019; Snider et al., 2014, 2015), measurements of net photosynthesis (A_N) and stomatal conductance (g_s) were taken using the LI-6800 portable photosynthesis system at each irrigation termination time. Measurements were conducted on uppermost, fully expanded leaves at the fourth node below the plant terminal between 1200 and 1500 h. The aforementioned time period was selected because this is the time during the day at which maximum drought stress is observed and leaf water potential is normally at a minimum (Chastain, Snider, and Collins et al., 2016; Grimes & Yamada, 1982). Environmental conditions inside the leaf chamber were set as follows: flow rate = 600 μmol s⁻¹, reference [CO₂] = 400 μmol mol⁻¹, photosynthetically active radiation (PAR) = 1500 μmol m⁻² s⁻¹, chamber air temperature = ambient air temperature at the time of measurement, and relative humidity = 65% (Parkash et al., 2021). Leaves were enclosed in the chamber until steady-state A_N and g_s were observed (60–120 s).

2.5 End-of-season measurements

Near the end of the season, when approximately 60% of the bolls had opened in the late maturing variety, defoliants were applied to ensure leaf drop and boll opening before harvest (Hand et al., 2022). Immediately following defoliation and prior to harvest, the following were measured from five plants in each plot: plant height, node at which the first fruiting branch was observed, and the mainstem node at which the uppermost harvestable boll occurred. On the same day, five plants from each plot were chosen to determine the number of bolls on each plant showing symptoms of boll rot and hardlock. Boll rot and hardlock can affect cotton fiber quality and yield (Blasingame & Patel, 2013; Lawrence et al., 2015). The most notable symptoms of boll rot are dark lesions and areas of necrosis on the bolls that eventually spreads to deteriorate the fiber and seeds (Baga, 1970). Hardlock in cotton results in compacted fibers that are not easily harvested with conventional pickers (Srivastava et al., 2010). Often, there are overlaps in symptomology between boll rot and hardlock, so the number of bolls exhibiting hardlock and/or boll rot were determined in one count and reported in bolls m⁻² (Stewart, 2007). For a boll to be included in the count, disease symptoms had to be severe enough to prevent seed cotton from being harvestable. Cotton was mechanically harvested using a two-row spindle picker (John Deere 9930), and seed cotton was weighed in the field. Thereafter, harvested seed cotton was ginned at the University of Georgia MicroGin in Tifton, GA (Li et al., 2011) to obtain a realistic estimate of gin turnout (lint weight divided by seed cotton weight). Seed cotton yield was multiplied by gin turnout to estimate lint yield (in kg ha⁻¹). To obtain HVI fiber quality parameters, such as fiber length, strength, micronaire, and uniformity, a 0.5 kg sample of the ginned fiber was sent to the USDA classing office in Memphis, TN. IWUE was determined by dividing lint yield (kg ha⁻¹) of each plot by total irrigation (mm) applied in the corresponding plot (Vories et al., 2005).

2.6 Statistical analysis

Initially, a three-way mixed effect analysis of variance (ANOVA) was utilized to evaluate the effect of year, irrigation treatment, cultivar, and their interactions on lint yield, IWUE, fiber quality, end-of-season growth patterns, cutout date, and prevalence of hardlock and boll rot. Since there was a significant interaction between year and cultivar for multiple parameters of interest (lint yield, IWUE, cutout date, and disease prevalence), the effect of irrigation treatment and cultivar on all parameters of interest was evaluated within each year separately using a mixed-effect ANOVA. Specifically, irrigation treatment, cultivar, and the irrigation × cultivar interaction were treated as fixed effects, while replicate and irrigation treatment × replicate were considered random effects. Fisher's protected least significant difference (LSD) test was used to separate means at an alpha level of 0.05. All statistical analyses were conducted using JMP 16 pro software (SAS Institute), and graphs were constructed using SigmaPlot 14.0 (Systat Software Inc.).

3 RESULTS AND DISCUSSION

3.1 Weather conditions and irrigation

The two growing seasons at the study site experienced similar average daily temperatures but contrasting rainfall patterns (Table 1). The average maximum daily temperature for 2022 was 0.31°C higher than in 2021; however, the average minimum daily temperature for the second year was 0.33°C lower than the first year. Overall, the temperatures were similar for both years (Table 1). The first growing season of this study, which lasted from May 6, 2021 to October 19, 2021, was a relatively wet year, receiving a total of 720.6 mm of rainfall (Table 1). However, during the next growing season, April 28, 2022 to October 21, 2022, only 541.2 mm of rainfall was received (Table 1). The difference in rainfall between the 2 years was 179.4 mm. The time interval between cutout and first open boll was only 8 days during the 2021 season, but it was 27 days during the 2022 season. Although T1 in 2022 received approximately twice the rainfall between cutout and first open boll than was received during the 2021 season, the time interval between these two stages in 2022 was more than three times what it was in 2021. There were also fewer consecutive days between rainfall events in the 2021 season than in the 2022 season. The maximum number of rain-free days between successive termination times ranged from four (T2) to six (T3) during 2021 and from four (T2) to nine (T1 and T3) during 2022. Since irrigation termination for the latest irrigation termination time coincided with agronomic maturity in both years, the maximum number of consecutive days without rainfall for T4 was zero in both years.

Based on historical weather data (www.georgiaweather.net/?variable = HI&site = CAMILLA), we can conclude that 2022 was an exceptionally dry year from cutout to first open boll. For example, when historical rainfall data are evaluated for the same 27-day period for the last 10 years, there is only 1 year for which the number of consecutive days without rainfall exceeded those in the current study (2014 with 12 consecutive days without rainfall). This resulted in fewer irrigation events in 2021 compared to the 2022 season, as well as all treatments in 2021 receiving equal amounts of total irrigation (188.0 mm), except for treatment four, which received only 9.0 mm more irrigation water by the end of the season. In contrast, each of the four treatments in 2022 received different amounts of total irrigation, with irrigation amount increasing as irrigation termination was delayed. For example, terminating irrigation at cutout resulted in 85.2 mm less irrigation applied than terminating irrigation at first open boll and 136 mm less than the latest irrigation termination time.

3.2 Irrigation treatment effects

None of the parameters evaluated in the current study were significantly affected by irrigation termination time, except IWUE in 2022, where substantial differences in end-of-season irrigation water were applied among the four treatments. Furthermore, no interactions between cultivar and irrigation treatment were observed for any measurement of interest in both years (Table 2). Lint yield was not significantly affected by irrigation termination time in 2021 or 2022 (Figure 2). During the 2022 growing season, where substantial differences in the amount of water applied to each treatment were observed, lint yield ranged from 1058 kg ha⁻¹ for T1 (irrigation terminated at cutout) to 934 kg ha⁻¹ for T4 (Figure 2). IWUE was not statistically different among irrigation treatments in 2021, but as noted above, in 2022, T1 (termination at cutout) had a significantly higher IWUE than the other three treatments (Figure 2). The average IWUE for terminating irrigation at cutout in 2022 was 3.59 kg ha⁻¹ mm⁻¹, which surpassed the next highest IWUE treatment (terminating irrigation at first open boll) by 1.12 kg ha⁻¹ mm⁻¹ (Figure 2). To consider total amount of water received by the crop, water productivity was calculated. Similar trends to lint yield and IWUE in both years were observed (Figure 2). These data indicate that irrigation could be terminated as early as cutout, resulting in substantial increases in IWUE without incurring significant yield penalties. Since there was no interaction between cultivar and irrigation, despite the selection of these cultivars for their differences in maturity, the results should be equally applicable to early and later maturing cultivars. Other studies examining irrigation termination times in cotton have mostly been conducted in western Texas (Ale et al., 2020; Himanshu et al., 2021; Lascano et al., 2017; Multer & Sansone, 2007). Ale et al. (2020) and Himanshu et al. (2021) concluded the optimal time for terminating irrigation in western Texas was between 111 DAP and 125 DAP. In some cases, this window would occur soon after cutout, depending on environmental factors. Lascano et al. (2017) suggested a very early termination time (84 DAP) that may occur prior to cutout in some regions such as the southeastern United States. Also, in western Texas, a 4-year drip irrigation study showed the earliest termination time (400 heat units after cutout) maintained the same level of yield as the later termination dates (Multer & Sansone, 2007). In Mississippi, a furrow-irrigated study that focused on irrigation termination time concluded there were no negative effects on lint yield and fiber quality if the last irrigation event occurred at cutout (Plumblee et al., 2021). More recently, Himanshu et al. (2023) investigated the growth stages that are most susceptible to water-deficit stress and concluded that water deficits during cutout, and the growth stages following cutout, had no significant effect on yield. Terminating irrigation at cutout might be an effective strategy to maximize yield and water use efficiency in cotton for the humid southeastern United States. It should be noted that the cotton crop in the current study was irrigated using a 100% evapotranspiration replacement strategy. Thus, the crop is assumed to not be under any level of water deficit at each irrigation termination time. Therefore, these results are primarily applicable when there is initially a water-replete soil profile at cutout.

TABLE 2. Mixed-effects analysis of variance (ANOVA) results for irrigation, cultivar, and irrigation × cultivar on lint yield, irrigation water use efficiency (IWUE), fiber quality characteristics, cutout date, hardlock + boll rot occurrence, final plant height, node of first fruiting branch (FFB), and node of uppermost harvestable boll (UHB) for cotton grown in the Stripling Irrigation Research Park (SIRP) during the 2021 and 2022 growing seasons.

Yield	Parameters	p-value
		Irrigation	Cultivar	Irrigation × cultivar
2021	Lint yield	0.547	0.201	0.278
	IWUE	0.462	0.215	0.301
	Fiber strength	0.928	<0.001^*	0.192
	Micronaire	0.713	0.638	0.401
	Uniformity	0.374	0.407	0.517
	Fiber length	0.525	<0.001^*	0.619
	Cutout date	0.412	<0.001^*	0.888
	Hardlock + Boll rot	0.914	0.126	0.400
	Height	0.824	0.064	0.118
	FFB	0.138	0.012^*	0.883
	UHB	0.309	0.185	0.133
2022	Lint yield	0.076	<0.001^*	0.855
	IWUE	<0.001^*	<0.001^*	0.840
	Fiber strength	0.920	0.039^*	0.650
	Micronaire	0.403	0.001^*	0.955
	Uniformity	0.532	<0.001^*	0.722
	Fiber length	0.384	<0.001^*	0.444
	Cutout date	0.430	<0.001^*	0.205
	Hardlock + Boll rot	0.405	<0.001^*	0.247
	Height	0.333	<0.001^*	0.422
	FFB	0.352	<0.001^*	0.850
	UHB	0.950	<0.001^*	0.447

• * significant (p ≤ 0.05) main effects or interactions.

FIGURE 2

Yield (a), irrigation water use efficiency (IWUE, b), and water productivity (c) for four end-of-season irrigation treatments (terminating irrigation at cutout [T1], first open boll [T2], 2 weeks after first open boll [T3], and 4 weeks after first open boll [T4]) during the 2021 and 2022 growing season at the Stripling Irrigation Research Park (SIRP). Bars represent means ± standard error (n = 12), and bars not sharing a common letter within a given year are considered significantly different (p ≤ 0.05).

3.3 Cultivar effects

There were no significant effects of cultivar on lint yield or IWUE in 2021. However, in the 2022 growing season, significantly higher lint yield and IWUE were observed in the full-season PHY 580 W3FE cultivar than in the short to mid-season PHY 332 W3FE cultivar (Figure 3). In 2022, PHY 580 W3FE produced an average of 302 kg ha⁻¹ more lint yield than PHY 332 W3FE. Furthermore, the IWUE in 2022 for PHY 580 W3FE was higher than PHY 332 W3FE by 0.81 kg ha⁻¹ mm⁻¹ (Figure 3). The 2021 growing season was characterized by high rainfall and lower yields overall than the 2022 season, where the highest mean yield observed in 2021 was 944.3 kg ha⁻¹ and the highest in 2022 was 1115 kg ha⁻¹. Lower yielding environments, like in 2021, tend to produce less yield variation among different cotton genotypes, whereas higher yielding environments result in a greater ability to distinguish lint yield variation among cultivars (Campbell & Jones, 2005). Similar to the 2022 results, in 2020, a cotton variety performance review in Alabama showed that PHY 580 W3FE was a higher yielding variety than PHY 332 W3FE, regardless of irrigation management (Jordan Jr, 2021). Because water use efficiency is directly tied to lint yield (Chastain et al., 2016; Liu et al., 2022; Meeks et al.,2017), the results from the current study further illustrate how appropriate cultivar selection for specific environments can also increase IWUE. The fiber quality parameters significantly affected by cultivar in both 2021 and 2022 were fiber length and fiber strength. In 2022, uniformity and micronaire were also affected by cultivar. In both years, PHY 332 W3FE (2.96 cm in 2021 and 2.99 cm in 2022) had longer average fibers than PHY 580 W3FE (2.84 cm in 2021 and 2.89 cm in 2022). In contrast, PHY 332 W3FE had stronger fibers than PHY 580 W3FE in 2021 (5.0% stronger), but in the next year, PHY 580 W3FE produced stronger fibers than PHY 332 W3FE (2.6% stronger). In 2022, PHY 580 W3FE had higher fiber uniformity and a higher micronaire value than PHY 332 W3FE. In that year, PHY 580 W3FE produced a uniformity of 81.9% and a micronaire value of 4.16, while PHY 332 W3FE had a uniformity of 80.6% and a micronaire value of 3.97 (Figure 4). In the variety trial program mentioned earlier (Jordan Jr, 2021), PHY 580 W3FE and PHY 332 W3FE preformed similarly to the 2022 results of this irrigation study with regard to fiber quality parameters. In that study, PHY 580 W3FE had an average fiber length of 2.85 cm, average fiber strength of 30.2 g tex⁻¹, average uniformity of 82.4%, and an average micronaire value of 4.53, while PHY 332 W3FE produced an average fiber length of 2.86 cm, average fiber strength of 31.5 g tex⁻¹, average uniformity of 83.8%, and an average micronaire value of 4.61. While fiber quality parameters can be significantly affected by cultivar (Hu et al., 2018), multiple studies have highlighted the importance of environment or environment × genotype interactions in driving HVI-determined fiber properties (Campbell & Jones, 2005; Paterson et al., 2003; Snider et al., 2013; Zeng et al., 2014). Thus, differences in fiber quality results between the two seasons in the current study could be largely due to the substantial differences in rainfall or other environmental conditions between these two seasons.

FIGURE 3

Yield (a) and irrigation water use efficiency (IWUE, b) for two cotton cultivars (PHY 580 W3FE and PHY 332 W3FE) at a field site in the Stripling Irrigation Research Park (SIRP) during 2021 and 2022. Bars represent means ± standard error (n = 24), and bars not sharing a common letter within a given year are considered significantly different (p ≤ 0.05).

FIGURE 4

Fiber length (a), fiber strength (b), uniformity (c), and micronaire (d) of two cotton cultivars (PHY 580 W3FE and PHY 332 W3FE) at the end of the 2021 and 2022 growing seasons for cotton grown at the Stripling Irrigation Research Park (SIRP). Bars represent means ± standard error (n = 24), and bars not sharing a common letter within a given year are considered significantly different (p ≤ 0.05).

As expected, the cultivar selected based on its characterization as an early maturing cultivar, PHY 332 W3FE, reached cutout sooner than the late maturing cultivar, PHY 580 W3FE, in both years. In 2021, PHY 332 W3FE reached cutout (94 DAP) approximately 6 days before PHY 580 W3FE (100 DAP); however, in 2022, there was almost a 2-week (13 days) difference in cutout date between PHY 332 W3FE (93 DAP) and PHY 580 W3FE (106 DAP; Figure 5). Consequently, the inherent divergence in maturation between the two selected cultivars provide adequate representations of early-mid maturing cultivars and full-season cultivars to meet the objectives of this study. Cutout is defined as the cessation of new vegetative growth and the end of the effective flowering period (Bange & Milroy, 2000; Chaudhry & Guitchounts, 2003; Guinn, 1985; Gwathmey et al., 2016). The two growing seasons differed substantially in the incidence of hardlock and boll rot. Specifically, when considered across all irrigation treatments and cultivars, the density of diseased bolls in 2021 (32.9 bolls m⁻²) was approximately half of what it was in 2022 (63.5 bolls m⁻²). During the 2021 growing season, there were no cultivar differences in the incidence of hardlock and boll rot present. In contrast, during the 2022 season, PHY 332 W3FE had significantly higher rates of hardlock and boll rot in the field. There were approximately 52% more bolls m⁻² that were affected by these diseases in PHY 332 W3FE than PHY 580 W3FE (Figure 6). If we consider prevalence of hardlock and boll rot as a percent of all bolls produced by the plant for the 2022 season (the only season for which we have total boll counts), percentages ranged from a low of 40% of all bolls produced by the plant in PHY 580 W3FE to a high of 55% for PHY 332 W3FE (unpublished data). Therefore, hardlock and boll rot were likely significant constraints to achieving high yields in the current study. Unfortunately, reliable, chemical control options for managing boll rot are currently nonexistent (Kemerait, 2022). Hardlock and boll rot in cotton are highest when conducive environments, wet and warm conditions, initiate the growth and reproduction of a number of different pathogens. Injury from boll rot and hardlock increases when excessive rain or irrigation occurs as bolls are opening and throughout the late season (Dodds, 2018; Plumblee et al., 2021). Because PHY 332 W3FE reached cutout earlier, the bolls were exposed to environmental elements earlier than PHY 580 W3FE, which could cause the increase in disease incidence. In addition to earlier open bolls, early maturing cultivars have also been found to hold a large portion of their bolls at lower main stem nodes, where humid microenvironments are created due to limited airflow and dense canopies, which also favors pathogen growth (Bednarz & Nichols, 2005; Bowen et al., 2018; Hu, 2018). Greater disease presence in this variety could be one reason why PHY 332 W3FE produced less lint than PHY 580 W3FE, as discussed in the previous section.

FIGURE 5

Cutout date (DAP) for two cotton cultivars (PHY 580 W3FE and PHY 332 W3FE) grown at the Stripling Irrigation Research Park (SIRP) during the 2021 and 2022 growing seasons. Bars represent means ± standard error (n = 24), and bars not sharing a common letter within a given year are considered significantly different (p ≤ 0.05).

FIGURE 6

Total hardlock and boll rot (m⁻²) for two cotton cultivars (PHY 580 W3FE and PHY 332 W3FE) grown in 2021 and 2022 at the Stripling Irrigation Research Park (SIRP). Bars represent means ± standard error (n = 24), and bars not sharing a common letter within a given year are considered significantly different (p ≤ 0.05).

At the end of the first growing season, there was no significant difference between the two cultivars in final plant height (Table 2). However, in 2022, there was a significant cultivar effect for plant height, with PHY 580 W3FE producing taller plants than PHY 332 W3FE and the average difference being approximately 10.3 cm. In both years, cultivar had a significant effect on the mainstem node at which the first fruiting branch (FFB) was observed. In 2021, the FFB for PHY 580 W3FE averaged 7.20, whereas the mean FFB for PHY 332 W3FE was 6.73 (a difference of 0.47). Similarly, the next year showed a difference in average FFB for PHY 580 W3FE and PHY 332 W3FE of 0.48, where PHY 580 W3FE had the highest FFB values. Like final plant height, cultivar had a significant effect on the location of the mainstem node with the uppermost harvestable boll (UHB) in 2022 but not 2021. The difference in UHB means for the two cultivars was 1.56, with PHY 580 W3FE having the higher UHB (Figure 7). These findings correspond to the typical growth morphology exhibited by early and late-maturing cultivars. In general, cultivars that reach maturity at a quicker rate than full-season varieties are shorter and produce the first fruiting site at a lower node on the plant (Godoy & Palomo, 1999; Gwathmey et al., 2016; Jatoi et al., 2012; Ray & Richmond, 1966). They also typically have more fruiting sites lower in the canopy (Bednarz & Nichols, 2005). Physical determinants, such as the ones listed, give insight into the differences in rate of development between the two cultivars (Gwathmey et al., 2016).

FIGURE 7

Plant height (a), node of first fruiting branch (b), and the mainstem node with the uppermost harvestable boll (c) for two cotton cultivars (PHY 580 W3FE and PHY 332 W3FE) at the end of the 2021 and 2022 growing seasons at the Stripling Irrigation Research Park (SIRP). Bars represent means ± standard error (n = 24), and bars not sharing a common letter within a given year are considered significantly different (p ≤ 0.05).

3.4 End-of-season trends in gas exchange and percent open boll

A_N, g_s, and percent open boll data were collected on four dates throughout the 2022 growing season that corresponded to the four irrigation termination treatments: August 5, August 30, September 8, and September 19. A mixed-effects ANOVA revealed there were significant effects of cultivar on A_N, g_s, and percent open boll on some of the days that the data collection took place. However, irrigation treatment had no effect on gas exchange rates or percent open boll assessments, and there was no interaction between cultivar and irrigation on any date. On August 30, PHY 580 W3FE had a significantly higher A_N value than PHY 332 W3FE (by 22.9%). Similarly, PHY 580 W3FE had a statistically greater g_s value than PHY 332 W3FE (by 25.5%) on August 30 but also exhibited higher stomatal conductance on September 19 (by 15.3%).

Stomatal conductance is often considered a reference indicator of drought stress across multiple plant species (Medrano et al., 2002). Since there was no effect of irrigation on net photosynthesis or stomatal conductance, plants at each irrigation termination date did not appear to be under water-limiting stress and are comparable to plants receiving supplemental irrigation, meaning terminating irrigation at cutout in a humid region such as the southeastern United States would not result in drought-stressed plants later in the season (Ennahli & Earl, 2005; McMichael & Hesketh, 1982; Pettigrew, 2004). Because PHY 332 W3FE reached cutout sooner than the other cultivar, leaves measured on PHY 332 W3FE were likely older than those measured on PHY 580 W3FE at the end of the season. Cotton leaves peak in photosynthetic performance approximately 2–3 weeks after unfurling, and begin to decline thereafter until senescing at approximately 60 or 65 days after unfurling (Constable & Rawson, 1980; Wullschleger & Oosterhuis, 1990). Thus, an early-maturing cultivar would also be expected to show declines in photosynthetic rates of the uppermost, fully expanded leaves sooner than a later-maturing cultivar.

There were significant effects of cultivar on percent open bolls on each date, except for the first data collection date, where all plots had 0% open bolls. The early maturing cultivar, PHY 332 W3FE, had more open bolls in each case: 82.7% more on August 30, 47.6% more on September 8, and 27.9% more on September 19 (Figure 8). As seen in previous results, PHY 332 W3FE displayed the characteristics of an early-mid maturing cultivar, which includes developing mature bolls at a faster rate than full-maturity varieties (Gwathmey et al., 2016). The first collection date, where both cultivars had 0% open bolls, corresponds to T1 (irrigation termination at cutout). As stated previously, T1 had significantly higher IWUE than the other three treatments, meaning producers could terminate irrigation at cutout while maintaining relatively high IWUE and not decreasing yield when compared with the current recommendation of 10% open boll (Hand et al., 2022). Importantly, these observations do not appear to be affected by cultivar differences in maturity since the two cultivars in the current study exhibited notable differences in cutout date and end-of-season rates of boll opening.

FIGURE 8

Net photosynthesis (a), stomatal conductance (b), and open bolls (c) for two cotton cultivars (PHY 580 W3FE and PHY 332 W3FE) at four dates (August 5, August 30, September 8, and September 19) during the 2022 growing season at the Stripling Irrigation Research Park (SIRP). Bars represent means ± standard error (n = 24), and bars not sharing a common letter within a given date are considered significantly different (p ≤ 0.05).

4 CONCLUSIONS

The current study tested the hypothesis that differential irrigation termination times would affect lint yield, fiber quality, and plant stress, and that responses would be dependent on cultivar differences in maturity. The earliest irrigation termination time (at cutout) had significantly higher IWUE than the other three treatments. Also, there was no yield penalty associated with terminating irrigation at cutout, which agrees with the findings of Lascano et al. (2017) and Plumblee et al. (2021) for other production regions. Along with yield and IWUE, net photosynthesis and stomatal conductance were not affected by end-of-season irrigation termination time. Thus, plants did not appear to be under any water-deficit stress at each end-of-season irrigation termination date. Fiber quality and disease incidence were unaffected by irrigation treatment. The cultivars chosen for this study exhibited significant differences in yield (and by extension, IWUE), cutout date, boll disease incidence, growth, and end-of-season percent boll open estimates. Despite the substantial differences between the two cultivars, there was no interaction between irrigation and cultivar for any variable of interest. Thus, we can conclude that the early irrigation termination recommendation (terminate at cutout) can be applicable to cotton cultivars with contrasting maturities.

AUTHOR CONTRIBUTIONS

Bailey Lawson: Conceptualization; data curation; formal analysis; investigation; writing—original draft. John L Snider: Conceptualization; formal analysis; investigation; methodology; project administration; resources; supervision; writing—review and editing. Lavesta C Hand: Conceptualization; supervision; writing—review and editing. Wesley Porter: Conceptualization; writing—review and editing. George Vellidis: Supervision; writing—review and editing. Joshua M Lee, Devendra Prasad Chalise, Ved Parkash, Amrit Pokhrel, and Navneet Kaur: Investigation; writing—review and editing.

ACKNOWLEDGMENTS

The authors would like to thank the Georgia Cotton Commission and the University of Georgia for financial or material support of this research. The authors would also like to thank Calvin Perry, B. J. Washington, Lola Sexton, and Will Vance for their help and technical support.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.