© S. Jalilian et al., Published by EDP Sciences, 2023

1 Introduction

The cotton (Gossypium hirsutum L.) is the only fibrous crop primarily cultivated for fiber production in the world. However, cotton cultivation is currently facing rising production costs, leading producers to explore new and high yield potential cultivars. Crop management practices are another way to improve fiber quality and quantity while maintaining lint yield. It is worth pointing out that lint yield is negatively related to fiber quality and accessing methods to increase fiber quality while preserving cotton yield have become the focus of studies (Luo et al., 2019; Keshavarz, 2020).

Cotton growth and development is highly affected by genetic and environmental factors. One of these factors affecting cotton quality and yield is optimum plant density, and the determination of a suitable plant population differs by environment, cultural practices and cultivar (Dai et al., 2015; Keshavarz et al., 2021a). Usually, producers and farmers choose plant density based on tradition rather than variety requirement, potentially resulting in yield losses.

Plant population directly affects moisture availability, light interception and wind movement, which in turn influences the branching design, plant height, canopy architecture, boll shedding, plant maturity and balance between vegetative and reproductive growth (Romano et al., 2011). Establishing an appropriate plant density optimized the resource use, promote the water and fertilizer use efficiency and finally increase crop productivity and higher radiation interception due to early canopy closure. High plant density can prolong the vegetative phase, delay plant maturity and decrease net photosynthesis due to lower chlorophyll content and RUBP carboxylase enzymes activity (Li et al., 2019; Keshavarz et al., 2021a, b, c). Dense planting can promote auxin content and polar transport by enhancing the expression of the auxin polar transport gene (GhP1N1) and auxin biosynthesis gene (GhYUC5) (Khan et al., 2019). It can also prevent branching and limit plant structure by enhanced strigolactone content due to gene expression in the tips of vegetative branches followed by reducing gibberellic acid, indole acetic acid, cytokinin and brassinosteroid contents (Jiang et al., 2012; Keshavarz et al., 2018).

However, the relationship between lint yield and yield components of cotton are complicated. Darawsheh et al. (2022) documented that lint yield was mostly affected by bolls plant⁻¹, followed by seeds boll⁻¹ and lint mass seed⁻¹. Zhi et al. (2016) revealed high plant density decreased seed surface area, the number of fibers per unit and lint mass. Therefore, selection of smaller bolls with higher number of seeds might be useful because of the higher surface area for lint production (Keshavarz et al., 2016; Ul-Allah et al., 2021). Gao et al. (2012a, b) noted that seed size influenced lint production because cotton fiber is an extension of the seed’s epidermal cell. Khan et al. (2017a) discovered that boll weight, number of bolls plant⁻¹ as well as seed weight were significantly affected by plant density. On the other hand, low plant density prevents the expression of all-genetic potential, changes lint characteristics and reduces lint strength (Romano et al., 2011) while increasing lint micronaire (Suminarti et al., 2016). It is known that lint quality and maturity have an important role on yarn characteristics and spinning properties. Yang et al. (2014) resulted that a rational plant density provided a better canopy microenvironment to achieve higher lint percentage. In addition, the yield traits are a function of the cultivar (Hosseini et al., 2023). Lint fineness is acknowledged as ratio of lint cell-wall thickening to diameter of the lint and it depends on deposition of cellulose in lint cell, which is sensitive to growth conditions (Gao et al., 2012a, b). Synthesis of cellulose needs many enzymes and organic molecules that among them sucrose synthase and sucrose phosphate synthase are two essential enzymes in the carbon partitioning and sucrose synthesis (Khan et al., 2017b). Sucrose synthase provides initial carbon supply and converts to other macromolecular substances, such as cellulose for fiber synthesis and lipid in embryos. In the sucrose synthesis cycle, sucrose is catalyzed by sucrose synthase to provide fructose, while UDP-glucose provides substrates for the biosynthesis of cellulose (Yadav et al., 2013). The fructose released by sucrose synthase is reprocessed into another sucrose cycle to reproduction further cellulose accumulation, a pathway that regulates cellulose production under stress conditions (Liu et al., 2013). All these processes and enzymes are influenced by environmental conditions, but their response varies in different plants and/or how they are cultivated.

Producers of fiber continuously seeking for proper plant density new cotton genotypes aiming for higher lint yields. In presented study, we focused on cotton yield and lint quantities among ten cotton cultivars and hybrids and how these may influence by plant density. Therefore, the objective of this experiment was to determine the optimum plant density for hybrids and cultivars of cotton and what is the effect of high and low plant density on yield components and quality traits.

2 Materials and method

2.1 Study site and soil

A two years field study was carried out at Varamin Cotton and Fiber Crop Research Department, Tehran, Iran from 2017 to 2018. Varamin is located in the western zone of Tehran located at latitude 35.21°E and longitude: 51.38°N with an altitude of 927 m above mean sea level. Figure 1 is provided the total monthly precipitation and monthly averaged temperatures during the cotton-growing season. The soil in experimental site had loamy-sandy texture, neutral pH, 0.6% organic matter, low total nitrogen content, 15.9 mg · kg⁻¹ available P (Olsen) and 510 mg · kg⁻¹ available K in the top 30 cm of the soil profile.

Before sowing, the experimental field was ploughed and disking with a reel and harrow conditioner to smoothing seedbeds. The experimental design was a split plot with 30 treatment combinations in four replications. The main plot consists of three-intra row spacing of 20 cm (T1), 30 cm (T2) and 40 cm (T3) (with plant densities viz. 6.25, 4.16 and 3.125 plants · m⁻², respectively) with fixed row spacing of 80 cm. Each subplot size was 3.3 m wide and 6 m in length with four rows. Subplot comprised of ten G. hirsutum genotypes including six cultivars of Varamin, Bakhtegan, Sajedi, Sahel, T2 and T14 and four hybrids Sahel × Sajedi, Bakhtegan × T2, Varamin × T14 and Sajedi × T14 (referred hereafter as G1, G2, G3, G4, G5, G6, G7, G8, G9 and G10, respectively). All seeds got from the Varamin Cotton and Fiber Crop Research Department, Tehran. Before sowing, a heavy pre-sowing irrigation was applied, after which the cotton genotypes were seeded on April 20th by using the manual hill-drop planting method. Thinning was done (at the third leaf stage) to achieve uniform stands of the desired plant densities. All the potassium (in the form of potassium sulfate, 41.5% K) and half of the nitrogen (in the form of urea, CH₄N₂O, 46% N) were applied before planting. The other half of the nitrogen was top-dressed at flowering stage. All insect-pests and weeds were controlled and other cultural practices were in coordination with Varamin Cotton and Fiber Crop Research Department (Keshavarz et al., 2021a; Hassanzadehdelouei et al., 2022).

Fig. 1 Average monthly air temperature and precipitation during the period of April–October in 2017 and 2018.

2.2 Plant data collection and analysis

Data were collected for number of boll plant⁻¹, boll weight, fiber yield, seed cotton yield (kg · ha⁻¹), lint and cotton seed oil content (%) at maturity stage (when approximately 60% of the bolls were open) for each fruiting position. In each year, observations were taken from 10 randomly-guarded plants from the two central rows of each plot (except plants in the first and last 50 cm of each row) and the average was calculated for the number of boll plant⁻¹ and boll weight. Seed cotton (including seed and lint) was manually collected from 2 m⁻² and air dried for three days under the sun until the water content reached 12%). Seed and lint were separated using a roller-ginning machine (made in Anyang, China). Seed cotton yield (kg · ha⁻¹), lint yield (kg · ha⁻¹) and lint percentage (as lint, divided by total seed cotton yield) was calculated after ginning. The seed oil content was measured by using the Soxhlet extraction apparatus (Sadasivam and Manickam, 1995).

The lint properties were measured on the lint picked from the 10 boll samples of each plot and then combined. The lint from each plot were sent to laboratory with high volume instrument lint analysis (HVI model HVI Classic 900) for lint quality measurement such as lint length (as 2.5% of the span length and expressed as mm), lint uniformity (as the ratio of the mean length to the upper-half mean length and expressed as a percentage), lint strength (as the force cN · tex⁻¹ necessary to break the fiber bundle), lint fineness (as the fineness and maturity of the lint, expressed in standard and expressed as μg · in⁻¹), lint elasticity, and fiber quality index (Papastylianou and Argyrokastritis, 2014).

2.3 Statistical analysis

The variance of two years of data was analyzed using ANOVA and when the Ftest indicated statistical significance, the LSD’s mean wise comparison test (honest significant difference, at P < 0.01 or P < 0.05) was used to verify the significance of differences among treatment means of main effect and the significant interaction effects were separated by slicing method. SAS 9.1 and S-PLUS ver. 6.1 software, USA were used for Principal Component Analysis (PCA) based on biplot and clustering analysis.

3 Results

3.1 Cluster and PCA analysis

Cluster analysis divided the all-30 treatments into three groups, composed of eighteen, one and eleven treatments respectively (Fig. 2A). Based on the bi-plot results (Figs. 2B and 2C), T1G8, T2G1, T2G8, T3G1, T3G2, T3G8, T3G10 and T3G7 formed one group, which was strongly associated with lint strength, lint elasticity and lint quality indices. Treatments T1G1, T1G2, T1G7, T1G9, T1G10, T2G2, T2G7, T2G9, T2G10 and T3G9 were classified in another group, which was related to lint yield and lint length. Treatments T1G4, T1G6, T2G3, T2G5, T2G6, T3G3, T3G4 and T3G6 formed an adjacent group, and were related with number of bolls plant⁻¹, boll weight, lint uniformity and lint fineness.

Fig. 2 The results of dendrogram based on cluster analysis (A) and biplot of first and second components based on principal component analysis (B and C). T1: 6.26 plant · m−2; T2: 4.16 plant · m−2; T3: 3.125 plant · m−2; G1: Varamin; G2: Bakhtegan; G3: Sajedi; G4: Sahel; G5: T2; G6: T14; G7: Sahel × Sajedi; G8: Bakhtegan × T2; G9: Varamin × T14; G10: Sajedi × T14. Y1: bolls number plant−1; Y2: boll weight; Y3: lint yield; Y4: seed cotton yield; Y5: lint percentage; Y6: lint length; Y7: lint uniformity; Y8: lint fineness; Y9: lint strength; Y10: lint elasticity; Y11: lint quality indices; Y12: oil seed content; Y13: cotton seed.

3.2 Number of bolls plant⁻¹

The results showed that the main effect of year, plant densities, cotton genotypes and their interaction had significant effect on bolls plant⁻¹ (Tab. 1). For both years, remarkable differences in boll plant⁻¹ averaged across the cotton genotypes were found among the plant densities. In the first year, the bolls plant⁻¹ was the highest under moderate to low plant density (4.16 and 3.12 plant · m⁻²) for G3 (24.3) and G4 (25.5), respectively while under high plant density conditions, G6 (23.1) had the higher boll plant⁻¹ (Tab. 2). However, in high plant density and low plant density, there was no significant difference (P < 0.05) between G3, G4, G5, G6 and G9 genotypes. The maximum bolls plant⁻¹ in the first year (averaged by plant density treatments), was obtained by G3 (24 boll plant⁻¹). Quite different results were obtained in the second year, in which high, moderate and low plant densities produced the highest number of bolls plant⁻¹ for G4 (31.1), G7 (30.9) and G6 (29.8), respectively. However, in moderate and low plant density, almost all genotype was statistically at the same level (Tab. 2). It should be noted that the maximum boll plant⁻¹ in second year (averaged by plant density) was found in G3 (26.7 boll plant⁻¹) and G4 (27.2 boll plant⁻¹). Averaging all genotype treatments, boll plant⁻¹ in low plant density was 12.5% and 13.7% (in first year), and 17.5% and 0.6% (in second year) higher than high and moderate plant density, respectively (Tab. 2). Averaging across years and genotypes, the boll plant⁻¹ was 19.3 per plant at high plant density and increased by 9.8% and 15.7% at moderate and low plant density.

3.3 Boll weight

Boll weight was significantly (P < 0.05) affected throughout the two years by different genotypes and plant densities (Tab. 1). At high plant density and in first year, G1 and G10 recorded significantly higher boll weight than other genotypes (Tab. 2). Different results were observed at moderate and low plant density. Cotton genotypes G1 (6.6 g), G8 (5.7 g) and G10 (6.0 g) at moderate plant density and cotton genotype G1 (6.3 g), G7 (5.7 g), G8 (6.1 g) and G10 (6.4 g) at low plant density significantly had higher boll weight than other genotypes (Tab. 2). In term of boll weight affected by plant density × cotton genotype in second year, high, moderate and low plant density produced the maximum boll weight for G1 (4.3 g), G1 (4.06 g) and G2 (4.5 g), respectively (Tab. 2). In the first year, the boll weight at low plant density (averaged by cotton genotypes) was 5.4% and 2.5% higher than high and moderate plant density, respectively (Tab. 2). In second year, the boll weight at low plant density (averaged by cotton genotypes) was 4.4% lower and 4.9% higher than high and moderated plant density, respectively (Tab. 2). When averaged across the cotton genotypes and two years, the boll weight at high plant density (4.13 g) was slightly higher than moderate plant density (4.05 g) and low plant density (4.19 g), with an average of 1.9% and 1.4%, respectively (Tab. 2).

3.4 Lint yield

Lint yield was significantly (P < 0.05) affected by different year, plant densities and genotypes (Tab. 1). A significant difference in lint yield existed between year and lint yield in 2018 (1169 kg · ha⁻¹) was significantly higher than that in 2017 (1035 kg · ha⁻¹) because of higher air temperature in last two month growing season (Fig. 1). Cotton sowing with dense planting (6.26 plant · m⁻²) had the highest lint yield (1386 kg · ha⁻¹) while cotton sowing with moderate and low plant population resulted in lower lint yield with average of 25.7% and 35.7% rather than high plant density, respectively (Tab. 3). Cotton genotype G8 (1269 kg · ha⁻¹), had significantly higher lint yield than other genotypes and followed by G4 (1263 kg · ha⁻¹), G1 (1239 kg · ha⁻¹), G2 (1223 kg · ha⁻¹) and G6 (1093 kg · ha⁻¹) (Tab. 3). Correlation was achieved positive and significant for the lint yield with seed cotton yield and lint percentage (Tab. 3).

3.5 Seed cotton yield

The analysis of variance results (Tab. 1) showed that effects of year, planting density, cotton genotypes, and interaction of year × genotypes were significant on seed cotton yield (Tab. 1). In 2017, maximum seed cotton yield was obtained in G4 (4736 kg · ha⁻¹) and G5 (4726 kg · ha⁻¹) followed by G3 (4524 kg · ha⁻¹) (Tab. 4). However, different results were obtained in 2018, in which cotton genotypes G2 (4053 kg · ha⁻¹), G3 (3960 kg · ha⁻¹), G7 (3965 kg · ha⁻¹) and G1 (3692 kg · ha⁻¹) produced significantly the highest seed cotton yield (Tab. 4). Averaged by cotton genotypes, seed cotton yield was about 15% higher in 2017 (4205 kg · ha⁻¹) than in 2018 (3560 kg · ha⁻¹). Averaged over years, seed cotton yield of G3 (4242 kg · ha⁻¹) was higher than other genotypes, which followed G2 (4038 kg · ha⁻¹), G4 (4081 kg · ha⁻¹) and G5 (4097 kg · ha⁻¹) (Tab. 4). When compared with high plant density, the moderate and low plant densities caused decreased 32.9% and 42.2% in 2017, 28.9% and 44.1% in 2018 in seed cotton yield, respectively (Fig. 3). When averaged across the plant densities, the seed cotton yield at second year (3560 kg · ha⁻¹) was less than first year (4206 kg · ha⁻¹) by average of 15.3% (Fig. 3). The seed cotton yield at low plant densities (averaged over the two years), was less than moderate and high plant densities by averaged of 17.3% and 43% (Fig. 3). Dai et al. (2015), Luo et al. (2019) and Romano et al. (2011) derived information on seed boll⁻¹ and significant relation between seed boll⁻¹ with seed cotton yield, and other yield characters in G. hirsutum and stated that seed boll⁻¹ is an important yield factor and manage boll weight, lint percentage, seed cotton yield, lint yield and seed oil content.

Fig. 3 Mean comparison of seed cotton yield influenced by plant density in two years. Means within each year followed by the same letter are not significantly different (P < 0.05).

3.6 Lint percentage

Lint percentage was significantly (P < 0.05) influenced by year, planting density, cotton genotypes and interaction of year × cotton genotypes and year × plant density × cotton genotypes (Tab. 1). In the case of mean comparison (Tab. 2), some differences were clear between two experimental years. In 2017, the highest lint percentage of high, moderate and low plant density was obtained in cotton genotypes of G8 (26.4%), G1 (35.2%) and G2 (33.1%), respectively. However, diverging results were obtained in the second year, when the highest lint percentage of high, moderate and low plant density was obtained in cotton genotypes of G6 (40.8%), G8 (45.9%) and G10 (45.4%), respectively. Averaged over 2017 and 2018, regardless of planting date, G8 (34.1%) and G7 (23.2%) had the highest and lowest lint percentage (Tab. 2). On average across genotypes, lint percentage under high plant density conditions was 8% and 11.7% in 2017, 5.6% and 12.2% in 2018 lower than under moderate and low density, respectively (Tab. 2). Averaged across plant density treatments, G1 (30.8%) and G10 (42.5%) had the highest lint percentage of the first and second year, respectively. The highest lint percentage of high, moderate and low plant density (on average across years) was found in cotton genotypes G6 (33.05%), G8 (39.2%) and G1 (37.15%), respectively.

3.7 Lint length

Significant variation between year, planting densities, cotton genotypes, and interaction of year × genotypes was found for lint length (Tab. 1). Under high density (6.26 plant · m⁻²), lint length was reduced by 1.25% and 2.17% compared to moderate and lower density treatments, respectively (Tab. 3). Averaged by cotton genotypes, lint length was about 4.09% higher in 2018 than 2017 (Tab. 4). In 2017, cotton genotypes G6 (33.9 mm), G4 (33.4 mm) and G3 (32.9 mm) had higher lint length rather than other genotypes. Similar results were also achieved in 2018, in which cotton genotypes G3 (37.5 mm) and G4 (36.9 mm) produced the highest lint length. Averaged over both years, cotton genotypes G3 (35.2 mm) and G4 (35.15 mm) had the highest lint length (Tab. 4).

3.8 Lint uniformity

A significant difference (P < 0.05) in lint uniformity existed between years and lint uniformity in 2018 (85.1%) was significantly higher than that in 2017 (84%) by 1.29% on average, because of higher air temperature in last two months of the growing season (Fig. 1).

3.9 Lint fineness

The results revealed that the lint fineness was significantly affected by main effect of plant densities and cotton genotypes (Tab. 1). The means comparison indicated that the lint fineness was 4.29 μg · in⁻¹ under low plant density, and decreases of 3.72% and 1.65% were recorded when cotton genotypes were planted under moderate and high plant densities, respectively (Tab. 3). Cotton genotypes G6 (4.66 μg · in⁻¹) and G8 (4.54 μg · in⁻¹) had significantly higher lint fineness throughout the growing period than other genotypes during both years (Tab. 3).

3.10 Lint strength

The results showed that the main effect of year, plant densities, cotton genotypes and their interaction had significant effect on lint strength (Tab. 1). For both years, remarkable differences in lint strength averaged across the cotton genotypes were found among the plant densities. In the first year, the lint strength was the highest under low plant density for G4 (38.2 cN · tex⁻¹), G6 (37.6 cN · tex⁻¹) and G3 (36.9 cN · tex⁻¹) while under moderate and high plant density conditions, G6 (40.9 cN · tex⁻¹ and 39.5 cN · tex⁻¹, respectively) had the higher lint strength (Tab. 2). Similar results were obtained in second year, in which the highest lint strength of high, moderate and low plant density was obtained in cotton genotypes of G3 (36.2 cN · tex⁻¹), G4 (37.9 cN · tex⁻¹) and G4 (36.3 cN · tex⁻¹), respectively. However, in low plant density, G3 (35.8 cN · tex⁻¹), G6 (35 cN · tex⁻¹) and high plant density, G4 (35.6 cN · tex⁻¹) and G6 (34.4 cN · tex⁻¹) were statistically at the same level (Tab. 2). Averaged across plant density treatments, G6 (39.3 cN · tex⁻¹) had the highest lint strength of first year. It should be noted that the maximum lint strength in second year (averaged by plant density) was found in G3 (36 cN · tex⁻¹) and G4 (36 cN · tex⁻¹). On average genotype treatments, lint strength in high plant density was 3.37% and 3.17% in first year, 0.8% and 0.6% lower than moderate and low plant density, respectively (Tab. 2). Averaging across years and genotype, the lint strength was 31.84 cN · tex⁻¹ at high plant density and increased by 1.33% and 1.94% at moderate and low plant density.

3.11 Lint elasticity

The analysis of variance showed that effects of year, planting density, cotton genotypes, and interaction of planting density and cotton genotypes was significant on lint elasticity (Tab. 1). A significant difference in lint elasticity existed between years, and lint yield in 2018 (6.6%) was significantly higher than that in 2017 (6.5%) (Tab. 3). Cotton genotypes G6 and G9 had the lowest and highest lint elasticity compared to other genotypes under high (5.63% and 7.42%, respectively), moderate (5.46% and 7.29%, respectively) and low plant densities (5.58% and 7.28%, respectively) (Tab. 5). Under dense crops (averaged by cotton genotypes), lint elasticity was increased by 2.1% and 2.8% compared with medium (5.55%) and lower (6.51%) density crops, respectively (Tab. 5). Likewise, on averaged by plant density treatments, the highest lint elasticity was obtained in cotton genotypes G2 (6.98%) followed by G7 (6.81%) and G8 (6.84%).

3.12 LQI: lint quality indices

Different cotton genotypes had significant, while plant densities had non-significant effect on lint quality indices during both years (Tab. 1). However, significant variation between cotton genotypes × plant densities and year × cotton genotypes was found for lint quality indices. Averaged by cotton genotypes, lint quality indices were about 0.83% higher in 2018 than 2017 (Tab. 4). In 2017, cotton genotypes G4 (25124) and G6 (24040) had higher lint quality indices rather than other genotypes. Similar results were also achieved in 2018, in which cotton genotypes G4 (29210) produced higher lint quality indices rather than other genotypes. Averaged over both years, cotton genotypes G4 (27167) had the highest lint length (Tab. 4). The highest and lowest lint quality indices were achieved in G4 (26814) and G9 (16437) under high plant density, G4 (27214) and G2 (16693) under moderate plant density, G4 (27474) and G9 (15874) under low plant density (Tab. 5). Under high plant density (averaged by cotton genotypes), lint quality indices were reduced by 2.8% and 2.1% compared to medium (20941.9) and lower (20792.6) plant density, respectively (Tab. 5). Likewise, on averaged by plant density treatments, the highest lint quality indices were obtained in cotton genotype G4 (27167.3).

3.13 Seed oil content

Different plant spacing had significant effect on seed oil content, while cotton genotypes or their interaction had non-significant effect on seed oil content during both years (Tab. 1). All cotton genotypes sown with 4.16 plant · m⁻² resulted in higher seed oil content (16.4%) compared to 6.26 and 3.12 plant · m⁻² during both years. However, there was no significant difference (P < 0.05) between high (14.8%) and low (15.2%) plant density (Tab. 3).

4 Discussion

In this two-year field experiment, we studied the influence of three plant densities on lint yield and yield components of ten cotton genotypes. It has been suggested that cotton genotypes should be selected for high number of bolls per unit (Ajayakumar et al., 2017). Venugopalan et al. (2013) indicated selection for high lint yield resulted in genotypes with more boll plant⁻¹ and seed boll⁻¹. Our results showed that boll plant⁻¹ and seed boll⁻¹ (data not shown) increased with decreasing plant density in all cultivars (Tab. 2). A likely reason for the lower boll plant⁻¹ of high plant density was the reduction of carbohydrate availability caused by plant competition. At low plant density, growth resources like water and nutrients were completely utilized and plants produced more branches, resulting in a higher number of bolls plant⁻¹. Moreover, the high plant density treatment reduced the growth period to crop maturity, resulting in fewer bolls plant⁻¹ initiation compared to lower plant density treatments. Indeed, excessive competition for sunlight, water and nutrients force the plants to early maturity. Presumably, high plant density lessened the sunlight penetration to canopy due to high foliage and early row closure, which causes boll abscission and lower lint quality.

Increased boll plant⁻¹ in the current study with decreased plant density was offset by enhanced boll plant⁻¹ resulting in increased seed number plant⁻¹ (data not shown) which resulted in higher lint yield. Khan et al. (2019) reported that cotton planted in a 19 cm row spacing produced fewer boll plant⁻¹ than cotton planted in a 38 cm row spacing. Different results for boll plant⁻¹ among cotton genotypes can be primarily led different genetic diversity of plants, difference growth habits and variation in partitioning of assimilates to reproductive parts and consequently a difference in boll plant⁻¹ of each genotype. Results pertaining to individual boll weight showed that high plant density negatively affected the individual boll weight related to low plant density (averaged by genotypes and two years). This corroborates the findings of Tang et al. (2014), who observed that the enhance of plant density reduced the boll weight plant⁻¹ due to lower lint and seed mass. Darawsheh et al. (2009) indicated that narrow row system (high plant density) affected negatively and constantly seed weight, and the individual boll weight and its components (seed and lint).

The lint yield increase observed for cotton genotypes planted in high plant density could be attributed to more rapid canopy closure and improved soil conditions of high plant density beds in early season (Khan et al., 2019; Keshavarz et al., 2021b) which provided an opportunity for improved seedling percentage and probably better root system development and higher light interception, which decrease weed competition. Higher seedling percentage led to higher dry matter accumulation due to longer photosynthesis period. Luo et al. (2019) stated that cotton genotypes planted under low plant density did not perform well for lint yield because of high foliage and boll shedding. In fact, partitioning of assimilates into excess vegetative structures can significantly reduce cotton lint yield. Although not significant, but same results and positive correlation obtained with respect to lint yield and individual boll weight. In all plant densities, higher lint yield was positively related to the higher lint percentage (Tab. 6). The greater cotton seed yield in the plants cultivated in high population was more affected by high plant density rather than medium and low plant densities, regardless of the last two presenting higher boll plant⁻¹. Based on our results, the higher plant density compensated the fewer bolls plant⁻¹, while averaged boll weight reduced or even remained unchanged. Arunvenkatesh and Rajendran (2013) showed that highest seed cotton yield was achieved with row spacing 15 cm followed by 30 cm than 45 cm in India.

Lint quality, which depends on accumulation of cellulose in lint cell, is more sensitive to environmental factors such as humidity (Ul-Allah et al., 2021), temperature (Darawsheh et al., 2022), nutrient status (Gao et al., 2012a, b; Wang et al., 2014), plant population (as shown in our experiment) and so on. Moreover, agronomic practices and environmental factors could change cotton lint properties and carbohydrate content (Tang et al., 2017; Sabourifard et al., 2023). It has been reported that sucrose synthase lessened with increasing cotton population in area (Khan et al., 2019). Lower sucrose synthase activity decreased availability of UDP-glucose for cellulose accumulation due to lower sucrose synthesis metabolism, which could be the reason for the lower lint formation under high plant density rather than low plant density (Ruan, 2014). It could therefore be argued that difference in plant density change the carbon translocation from source to sink, which influenced the leaf carbohydrates content and quality of lint. It has also stated that the content of polysaccharide such as fructose, glucose, sucrose and starch decreased in leaves of plant growth in high plant density (Luo et al., 2019). Reduction in leaf carbohydrate content could be explained in a sense of its transposition to reproductive parts. Therefore, our results indicated that lint quality was higher in lower plant density maybe due to variation of non-structural carbohydrate under different plant density. The values of lint uniformity, lint strength and lint length are mainly affected by genotypes and environmental conditions (Tab. 6). Khan et al. (2019) asserted that highest lint strength and length is produced with larger cellulose molecules, lower break point and more linkages between lint. Siebert and Stewart (2006), Luo et al. (2019), and Jiang et al. (2012) also showed that plant density and lint quality are inversely related. In the present study, the increased in the lint quality under low-density treatments could be explained due to more carbohydrate sources (sucrose) to induce synthesis of free fructose and glucose. Moreover, a negative association was observed between lint percentage and lint elongation, which means that there were physiologic adjustments and down regulation partly compensate for the declined lint synthesis.

Very low or high plant density of cotton genotypes reduce seed oil percentage, since the inter-plant competition for resources (such as the availability of moisture and light) enhances with increased in dense populations, while the total number of plants per unit area decreases with wider spacing. Based on our results, seed oil content has a negative association with lint percentage (Tab. 6). The content of seed oil percentage depends on carbohydrates partitioning to two related sinks in the bolls: seeds and lint (Khan et al., 2017b; Keshavarz and Khodabin, 2019). More photo-assimilates flow to seeds in combination with less photo-assimilates flow to lint could be the reason for negative correlation between lint percentage and seed oil content.

5 Conclusions

In conclusion, higher plant density ensured higher seed cotton yield, and lint elasticity in all cotton genotypes although the number of boll plant⁻¹, lint percentage and lint quality indices were reduced. However, cotton genotypes in low plant density consistently resulted less lint yield than cotton planted in the high plant density. Cotton genotypes G3, G4 and G5 resulted in higher seed cotton yield by producing higher number of bolls plant⁻¹. Yield difference among two years was derived from plant genetics, which responded differently to weather conditions. Despite the variances in the boll plant⁻¹, seed cotton yield and lint percentage among the years and genotypes, there were similarities in trends of plant density and the lint percentage, lint strength and lint quality indices increased when the plant density decreased. Among the cotton hybrids, G8 and G9 had the highest boll number and lint percentage, respectively at all plant densities. In case of lint yield, G8 had the highest lint yield, which was higher than conventional cultivars. It may be concluded from this study that high plant density systems limited yield quantity potential but enhance yield quality capacity. Thus, the results demonstrate that low plant density in Varamin may have lower seed cotton yield than cotton planted on high plant density.

Declaration of interest statement

The authors declare that they have no conflict of interest in relation to this article.

Acknowledgements

The authors thank Henrique Rasera Raniro (Department of Crop Sciences, University of Natural Resources and Life Sciences Vienna (BOKU), Tulln, Austria) for critical reading and editing of the manuscript.

References

All Tables

All Figures

	Fig. 1 Average monthly air temperature and precipitation during the period of April–October in 2017 and 2018.
In the text

Fig. 2 The results of dendrogram based on cluster analysis (A) and biplot of first and second components based on principal component analysis (B and C). T1: 6.26 plant · m−2; T2: 4.16 plant · m−2; T3: 3.125 plant · m−2; G1: Varamin; G2: Bakhtegan; G3: Sajedi; G4: Sahel; G5: T2; G6: T14; G7: Sahel × Sajedi; G8: Bakhtegan × T2; G9: Varamin × T14; G10: Sajedi × T14. Y1: bolls number plant−1; Y2: boll weight; Y3: lint yield; Y4: seed cotton yield; Y5: lint percentage; Y6: lint length; Y7: lint uniformity; Y8: lint fineness; Y9: lint strength; Y10: lint elasticity; Y11: lint quality indices; Y12: oil seed content; Y13: cotton seed.

In the text

	Fig. 3 Mean comparison of seed cotton yield influenced by plant density in two years. Means within each year followed by the same letter are not significantly different (P < 0.05).
In the text