The Correlation of Machine-Picked Cotton Defoliant in Different Gossypium hirsutum Varieties

Abstract

Cotton mechanized harvesting is the development direction of cotton production. The rapid development of mechanized cotton harvesting in Xinjiang has significantly increased the efficiency of cotton harvesting and reduced the harvesting cost. However, in the rapid development of mechanized cotton harvesting, there are also the problems of net yield, recovery rate and poor harvesting quality, which lead to the poor quality competitiveness of mechanized cotton harvesting. In order to solve the problem of mechanized cotton loss, the key is to reduce the problem of cotton miscellaneous, and breed cotton varieties suitable for mechanized harvesting. The purpose of this study was to clarify the key trait correlation of defoliation through the establishment and association analysis of foliation and deciduous phenotype indicators in cotton. In this study, the phenotypic indexes of defoliation and deciduous traits were established through the comprehensive analysis of the defoliation rate of 273 cotton germplasm resources and other 11 related phenotypic traits in the field, in order to provide guidance for the breeding and production of cotton varieties collected by breeders. In addition to peeling rate, an analysis of the association between 11 agronomic parameters and peeling rate and hanging rate revealed that the number of effective branches, chlorophyll SPAD value, fruit branch angle, and hanging rate have substantial correlation in 3 years. Hence The hanging rate, fruit branch angle, effective branch number, and chlorophyll SPAD value can be used as the evaluation indicators of varieties for peeling ability trait index.

1. Introduction

China is a large cotton consumer, but it is also a mega cotton producer. Cotton is a labor-intensive field cash crop, with various field production processes, from planting to the collection of more than 20 workers per. The situation of high cost, low efficiency and poor quality makes China’s cotton production facing severe challenges, and it is urgent and significant to accelerate the realization of the full mechanization of cotton production. Cotton harvesting through mechanization is a systematic engineering process based mainly on the combination of agricultural machinery and agronomy, research of cotton varieties breeding, planting, plant protection, chemical control, top leaf ripening, and mechanical harvest. The second major technology has been clear for the mechanised cotton planting mode, to raise the level of cotton production and mechanisation. In recent years, due to the transfer of rural labor force, the sharp rise of labor prices and the requirement for high-quality of cotton, the cotton planting in Xinjiang has encountered many difficulties. At the same time, the CPC (Communist Party of China) Central Committee and the government of the autonomous region have actively guided new technologies, new models and new methods in Xinjiang for production practice and exploration, and strive to break through the existing bottlenecks and realize the full mechanization, high efficiency, standardization and green cotton production in Xinjiang. At present, mechanical cotton mining has been widely promoted in the cotton area of Xinjiang, China. By 2020, the area of Xinjiang cotton machine mining had increased to 28.176 million mu, while the area of northern Xinjiang machine mining had reached to 15 million mu, representing a 60% increase. However, there are two main problems restricting the promotion progress of mechanized cotton mining. First, China has not introduced any technical measures such as peeling and ripening for the mutual application of cotton varieties [1]. Second, although blind introduction leads to high yield, the germplasm resources themselves are not sensitive to defoliation agent, with low defoliation rate, poor quality, high miscellaneous rate and low net recovery rate, and the number of ginning cleaning process also increases greatly, which damages the fiber quality of cotton and greatly reduces the grade and quality of cotton [2]. Mechanical harvesting may reduce the fiber density [3]. Due to its unique ecological and climatic conditions, high large-scale planting and high-level mechanization operation, Xinjiang is by far the most suitable agricultural production area to promote and implement remote sensing technology and precision agriculture in China [4]. Lv Xin [5]. It is proposed that the GIS decision support system can provide fertilization decision and consultation for cotton, and can better cooperate with mechanized production. With Chen Xuegeng [6], the promotion of the proposed precision planting technology in Xinjiang cotton planting level to a new level, will bring greater economic and social benefits, and will solve the key problems in the cotton precision engineering system.

The influence of cotton varieties on the effect of mechanical cotton picking and removing leaves is mainly manifested in three aspects. The advantage of the short reproductive period variety is its early maturity, high spitting rate, good leaf effect, good boll weight, yield, and maturity despite having a minor impact is beneficial to machines. The varieties with long fertility period, thick leaf blade mature late and the leaf effect is poor, while the boll weight, yield, quality, and maturity influences are not machine friendly. On the other hand, mechanized cotton picking demands compact cotton plant type, short fruit branches, “fried sticks” fewer with concentrated bolls, middle and lower leaves suitable for dosing, strong peeling effect, and ultimately this plant type is ideal for peeling agent spraying and mechanical harvesting operations. The third factor is the sensitivity of cotton varieties. Cotton cultivars that are defoliation sensitive grow quickly, have a high cotton boll spitting rate, a good defoliation effect resulting a high net production rate, and suffer minimal cotton quality loss throughout the mechanical cotton picking process. Many factors determining the effect of defoliation have been discovered by domestic and international researchers in recent years. For example, Wang Xiaojing et al. found that the defoliation agent interfered with the balance of hormones in the plant to achieve the purpose of defoliation [7]. Similarly, Hu Junxia sprayed the cotton and she found that the peeling rate of cotton was significantly improved, and the yield was slightly increased [8]. Song Min and Qu Yanying analyzed the correlation between the machine-picked cotton plant type and the defoliation rate, and found that when the fruit branches were decreased significantly, but the effect of the leaves on the defoliation effect was not obvious [9]. Zhang Qiang, Zhao Bingmei and others evaluated the mechanical cotton varieties with good defoliation effect through cluster analysis [10]. In the field research results, Zhu Yijie and the Hong-xia Zhao team discovered that after leaf rate effect difference is significantly different in different crop varieties. Leaf soluble sugar concentration and soluble protein content levels change dramatically between varieties, and there is a significant negative association between leaf rate and soluble sugar content in each variety [11]. Zhou Tingting et al. found that leaves of some cotton varieties withered after the spraying of foliation agent, but did not fall, resulting in high mixed rate [12]. Defoliation agent is absorbed through cotton leaves, and it can finally act on the petiole off the layer, affecting the hormone balance, soluble sugar content and a series of physiological reactions to achieve the purpose of defoliation. The density of fruit branches will affect the absorption and utilization efficiency of defoliation agent in cotton [13].

Different cotton varieties have different sensitivity to cotton foliage agent, and the indicators to measure the sensitivity of mechanized cotton varieties are unclear. In this study, we comprehensively analyzed the defoliation rate, pruning rate and agronomic traits of 273 Gossypium hirsutum varieties around the world, and evaluated the key traits of defoliation and foliation, and selected the cotton varieties with good defoliation traits and suitable for mechanical picking; and provided material basis and theoretical reference for the genetic improvement and breeding of new varieties (Appendix A).

2. Materials and Methods

2.1. Experimental Materials

The experiment was carried out in 2019–2021 in the sixth group of Xinjiang Academy of Agricultural Sciences (80°50′31″ E, 40°30′13″ N, for 2021), and the experimental varieties were provided by the Cash Crops Research Institute of Xinjiang University of Agricultural Science and Technology. There are 273 Gossypium hirsutum varieties evaluated comprising both local and foreign germplasm (Figure 1). Mechanical cotton mining with submembrane drip irrigation was employed to plant all 273 accessions of Gossypium hirsutum materials used in the test study in 6 rows with plot design (66 cm length × 10 cm width). Each variety was sown in triplicate, with random areas of, 3 m long, and comprising of 150,000 plants/mu as per theoretical basis. The field management was the same as the field production. The formulation of leaf agent provided by China Agricultural University, is thiabene-ethylene oxide (50% total active ingredients, 10%, 40%), 150 mL/667 m², and was sprayed through the UAV spray injection, so that each blade is attached to the drug liquid. A total of two injections were applied through Specific time in 11 September and 21 September.

Figure 1. Defoliant spraying effect. Harvest map after spraying deciduous leaves.

2.2. Investigation of Traits and Methods

According to the cotton DUS test index, 10 consecutive evenly balanced cotton plants were selected in each replicate to collect the agronomic trait data, and the average value of the three replicates was finally calculated as the phenotypic value of the trait. The specific survey traits and survey methods used in the present study were i.e.

Fertility period (BD): growth period includes seedling stage, bud stage, flowering stage and batting stage. Growth period The date when the index of each period of the survey reached 50%.

Plant height (PH): On 8 September, the distance between the cotyledon node of the cotton plant and the top of the main stem (after topping) was measured using a tape measure.

Number of effective branches (EFB): On 8 September, after topping, the number of all effective fruit branches on a cotton plant is liquidated.

Number of fruit knots (FN): On 8 September, the distance between the cotyledon node of the cotton plant and the tip of the main stem (after topping) was measured using a tape measure.

Total leaf number (NB): The total number of main stem leaves and fruit branches of cotton plant was calculated before defoliating agent was applied on 10 September.

Determination of chlorophyll SPAD value (Chl): On 8 September, SpAD-502 chlorophyll meter (Minolta, JPN) was used to determine the SPAD value of chlorophyll content in functional leaves of cotton (measured two leaves of the main stem after topping). The SPAD value was measured once in the main vein and on both sides of the leaves, and the average value was taken three times.

Leaf area (LA): On 8 September, LA-S leaf area measuring instrument was used to measure the leaf area of upper, middle and lower leaves respectively.

Number of beginning nodes (HFNFB): On 8 September, from the number of cotyledon nodes (cotyledon nodes count 0) to the first fruit branch, between the number of nodes is the number of beginning nodes.

Initial node height (HFNFH): On 8 September, the distance between the cotyledon node and the first fruit branch was measured with a tape measure.

Fruit branch Angle (FBA): On 8 September, the Angle between the first fruit node and the main stem of the upper four branches was measured with a protractor.

Leaf inclination (LIA): On 8 September, a protractor was used to measure the included Angle between the normal direction and the Z-axis direction of the leaf surface, that is, the included Angle between the leaf (main stem inverted 4 functional leaves) and the main stem.

Number of leaves remaining after first application (NFS): Number of leaves remaining investigated on 15 September.

Number of remaining leaves after second application (NSS): Number of remaining leaves investigated on 25 September.

Total number of hanging branches and leaves: The number of “dead but not falling” ground leaves and the number of “falling and hanging” leaves of cotton plants were investigated on 25 September.

2.3. Data Statistics and Analysis

For data summary and statistical calculation, the data including maximum, minimum, mean, standard deviation, skewness, kurtosis, coefficient of variation and gray correlation were analyzed using SPSS26.0 and Excel 2010, in reference to Tan Hehe and others [14], Niu Yisong et al. [15], Yin Guangting et al. [16] Methods. Correlations of the phenotypic traits were analyzed using the SAS 9.3 statistical software, and the generalized heritability was analyzed according to the ANOVA results by formula

$$H^2=\frac{{\delta }_G^2}{\left({\delta }_G^2+{\delta }_{GT}^2+{\delta }^2\right)}\times 100$$

was calculated where ${\delta }_G^2$ genetic variance, ${\delta }^2$ error variance, and ${\delta }_{GT}^2$ gene-environment interaction variance. The calculation formula of defoliation rate and hanging rate is as follows:

$$\begin{array}{r}\mathrm{Defoliation} \mathrm{rate}=\frac{\left(\mathrm{Total} \mathrm{number} \mathrm{of} \mathrm{lesves} \mathrm{before} \mathrm{application}-\mathrm{Total} \mathrm{number} \mathrm{of} \mathrm{leaves} \mathrm{after} \mathrm{application}\right)}{\mathrm{Total} \mathrm{number} \mathrm{of} \mathrm{leaves} \mathrm{before} \mathrm{application}}\times 100\%\\ \mathrm{Percentage} \mathrm{of} \mathrm{hanging} \mathrm{branches} =\frac{\mathrm{Total} \mathrm{number} \mathrm{of} \mathrm{hanging} \mathrm{branches} \mathrm{and} \mathrm{leaves}}{\mathrm{Total} \mathrm{number} \mathrm{of} \mathrm{blades}}\times 100\%\end{array}$$

3. Results and Analysis

3.1. Statistical Analysis of Phenotypic Traits

The findings of a statistical analysis using Excel 2010 and SPSS 26.0 for 273 accessions and three years of phenotypic data are reported in Table 1. In 2019, the yearly mean difference, the overall coefficient of variation is less than 15%, and the overall kurtosis and deviation of 3-year phenotypic features fit the normal distribution, compared to the previous two years.

Table 1. Descriptive statistics of phenotypic traits in 273 materials.

Trait	Year	Mean	SD	Kurtosis	Skewness	Max	Min	CV
BD	2019	124.10	15.90	0.08	0.09	137.5	111.5	7.81
	2020	123.85	16.58	0.17	0.38	137.4	115.4	7.47
	2021	123.97	14.87	0.18	0.36	136.5	113.3	8.34
PH	2019	72.50	10.73	0.54	0.29	105.00	46.00	6.76
	2020	72.27	10.37	0.51	0.39	109.67	44.67	6.97
	2021	72.18	9.97	0.58	0.28	102.25	48.25	7.24
EFB	2019	8.53	1.31	−0.27	0.37	12.00	5.00	6.51
	2020	8.67	1.00	0.42	0.46	11.67	5.67	8.67
	2021	8.59	1.13	0.67	−0.13	12.25	5.50	7.60
FN	2019	8.67	1.85	−0.10	0.37	17.67	6.33	4.69
	2020	13.32	1.88	0.10	0.35	19.67	6.52	7.09
	2021	13.75	2.12	0.38	0.52	18.00	6.25	6.49
NB	2019	24.09	4.53	0.65	1.01	45.33	13.00	5.32
	2020	23.97	3.92	0.09	1.31	36.33	16.00	6.11
	2021	23.78	4.28	0.80	1.07	37.25	13.00	5.56
SPAD	2019	62.33	8.10	0.43	0.55	77.00	47.73	7.68
	2020	65.23	8.45	0.57	0.56	79.13	52.87	7.72
	2021	62.88	12.12	−0.08	0.58	85.53	50.13	5.19
LA	2019	94.16	19.27	−0.37	0.23	160.21	41.81	4.89
	2020	96.92	19.21	−0.06	0.38	174.29	42.45	5.05
	2021	96.08	19.42	0.14	0.27	173.76	45.12	4.95
H FNFB	2019	5.33	0.75	−0.49	0.31	8.00	3.00	7.11
	2020	5.38	0.72	−0.18	−0.10	7.66	3.67	7.47
	2021	5.40	0.71	−0.10	0.22	8.00	3.25	7.61
H FNFH	2019	25.26	4.34	−0.66	0.33	41.67	15.67	5.82
	2020	25.33	4.44	1.23	0.22	43.00	15.33	5.70
	2021	24.87	4.00	0.80	0.66	42.25	16.50	6.22
FBA	2019	32.47	5.36	0.87	0.43	72.80	32.47	6.06
	2020	37.33	4.78	−0.95	0.18	85.67	37.33	7.81
	2021	39.60	4.48	0.38	0.22	68.93	45.03	11.07
LIA	2019	56.76	9.37	−0.46	0.22	80.87	35.67	6.06
	2020	56.50	8.71	0.02	0.31	79.97	36.40	6.49
	2021	55.35	8.72	0.32	0.26	82.10	35.25	6.35
HBR	2019	0.111	0.05	0.15	0.17	0.315	0.012	2.22
	2020	0.115	0.06	0.21	0.56	0.367	0.024	1.92
	2021	0.114	0.03	0.18	0.12	0.272	0.015	3.80
LR	2019	0.887	0.09	0.23	0.35	1.00	0.551	9.86
	2020	0.887	0.08	0.22	0.18	1.00	0.614	11.09
	2021	0.874	0.09	0.24	0.22	1.00	0.559	9.71

3.2. Generalized Heritability Analysis

Comprehensive analysis of generalized heritability of different traits in consective three years is shown in Table 2. The mean generalized heritability of the 13 traits was 55.95%. EFB has the highest generalized heritability (66.6%) followed by NB (64.9%), BD (64.8%), FBA (59.3%), and LIA (58.9%) respectively. Lowest generalized heritability value was observed for FN (44.6%). It indicates that the number of effective branches, the total number of leaves and the reproductive period have great genetic effect. Lowest generalized heritability values were noted for FN (44.6%), HFNFB (45.4%), and HFNFH (49.1%) respectively. It shows that the genetic effect of fruit node number, beginning number and initial height is relatively small, mainly affected by the environment while other traits are in the middle. Hence, both the genetic and environmental influences are comparable.

Table 2. Generalized heritability assessment of 3-year phenotypic traits in 273 materials.

Trait	σ2G	σ2E	σ2G × E	σ2e	H2
BD	3.74	1.28	2.46	4.80	64.8%
PH	3.15	3.41	2.87	8.22	52.9%
EFB	25.476	113.01	30.56	61.81	66.6%
FN	3.067	0.617	11.566	11.02	44.6%
NB	4.303	2.306	3.010	4.919	64.9%
H FNFH	2.73	0.02	3.82	5.50	49.1%
H FNFB	27.68	-0.40	46.61	59.80	45.4%
LA	0.078	0.735	0.094	0.180	50.5%
SPAD	0.29	0.04	0.23	0.55	58.2%
FBA	3.881	11.392	3.684	4.914	59.3%
LIA	0.019	0.003	−0.002	0.08	58.9%
HBR	6.74	6.41	5.80	12.21	57.7%
LR	0.02	0.03	0.00	0.08	54.4%

3.3. Analysis of Foliation Rate, Hanging Rate and Agronomic Traits in 2019

After differential analysis of defoliation and pruning rates of 273 materials in 2019 the significance p-value of one-variate ANOVA test was less than 0.05 indicating a significant difference in defoliation and pruning rates of the 273 varieties in 2019 (Table 3).

Table 3. One-factor ANOVA test.

	S S	DF	MSE	F	p
Hang branch rate Interblock	0.388	272	0.0014	2.839	0.02
Hang branch rate Within the group	0.424	810	0.0005
Hang branch rate Total	0.812	1082
Take off the leaf rate Interblock	1.660	272	0.0061	4.067	0.01
Take off the leaf rate Within the group	1.284	815	0.0015
Take off the leaf rate Total	2.944	1087

Correlation analysis between reproductive period, fruit branch angle and leaf inclination angle, branch hanging rate, defoliation rate and plant height, number of effective branches, beginning number, initial height, total leaf number, leaf area, number of fruit nodes, and chlorophyll SPAD value was carried out in data collected during 2019. The results showed that the effective branch number, total leaf number, reproductive period and defoliation rate were significantly negatively correlated. There was a very significant negative correlation between chlorophyll SPAD value, hanging rate, fruit branch angle and defoliation rate. The largest correlation coefficient with the defoliation rate is the pruning rate. The second is the reproductive period, the chlorophyll SPAD value (Chl), the number of effective branches (EFB), the total number of leaves (NB), and the fruit branch angle (FBA). Start height had the lowest correlation (HFNFH) with defoliation rate. There was a very significant negative correlation between leaf removal rate and hanging rate, and between leaf inclination angle and hanging rate, and a significant positive correlation between the number of effective branches, chlorophyll SPAD value, fruit branch angle and branch angle and hanging rate. The rate of defoliation and hanging branches were significantly correlated with the number of effective branches, the total number of leaves, chlorophyll SPAD value, and fruit branch angle (Table 4).

Table 4. Correlation analysis of leaf shedding rate, branch hanging rate and agronomic traits in 2019.

Trait	DR	PH	EFB	H FNFB	H FNFH	NB	LA	FN	SPAD	HGY	BD	FBA	LIA
DR	1
PH	0.129	1
EFB	−0.510 *	0.218	1
H FNFB	0.091	0.190	0.300	1
H FNFH	0.030	0.721 **	0.236	0.261	1
NB	−0.528 *	0.335	0.171	0.026	0.384 *	1
LA	0.279	0.309	0.523 **	0.234	0.511 **	0.431 *	1
FN	−0.32	0.35	0.35	−0.23	−0.32	0.28	0.03	1
SPAD	−0.542 **	−0.092	−0.049	0.242	0.054	0.008	0.051	−0.032	1
HGY	−0.830 **	−0.122	0.244 *	0.025	−0.055	0.518 *	−0.023	−0.332	0.414 *	1
BD	−0.620 *	0.306	0.031	0.350	0.109	0.26	−0.033	0.367	0.059	−0.215	1
FBA	−0.522 **	0.315	0.290	0.132	0.149	0.382	0.136	0.668*	0.006	0.528 *	0.227	1
LIA	−0.225	0.027	0.124	−0.001	0.074	0.073	−0.034	0.036	0.059	−0.512 *	0.124	0.225	1

Note: * and ** indicate significant (p < 0.05) or extremely significant (p < 0.01) correlation.

3.4. Analysis of Foliation Rate, Hanging Rate and Agronomic Traits in 2020

Differential analysis of the defoliation and pruning rates of the 273 accessions in 2020 showed that the significance p-value of the one-factor ANOVA test was less than 0.05, indicating a significant difference in the defoliation and pruning rates of the 273 accessions in 2020 (Table 5).

Table 5. Univariate ANOVA test.

	Sum of Squares of Deviations SS	Free Degree DF	Mean Square MSE	F	Conspicuousness p
Hang branch rate Interblock	0.368	272	0.0014	2.794	0.03
Hang branch rate Within the group	0.428	810	0.0005
Hang branch rate Tote	0.692	1082
Take off the leaf rate Interblock	1.351	272	0.0050	2.778	0.01
Take off the leaf rate Within the group	1.484	813	0.0018
Take off the leaf rate Tote	2.831	1085

Through the 2020 branch hanging rate, defoliate rate and plant height, number of effective branches, beginning number, initial height, total leaf number, leaf area, number of fruit nodes, and chlorophyll SPAD value, Correlation analysis between reproductive period, fruit branch angle and leaf inclination angle, The results show that the total leaf number and foliation rate, There was a significant positive correlation between leaf area and defoliation rate, There was a very significant negative correlation between the number of effective branches, the chlorophyll SPAD value, the branch hanging rate, the fruit branch sandwich angle and the defoliation rate, The largest correlation coefficient with the defoliation rate is the pruning rate, Secondly, the number of effective branches, chlorophyll SPAD value, fruit branch Angle; The beginning number had the lowest correlation with defoliation rate. There was a very significant negative correlation between leaf removal rate and hanging rate, and between leaf inclination angle and hanging rate, and a significant positive correlation between leaf area, number of effective branches, chlorophyll SPAD value, and fruit branch angle and hanging rate. Leaves removal rate and branch hanging rate were significantly correlated with the number of effective branches, chlorophyll SPAD value, fruit branch clip angle, and leaf area (Table 6).

Table 6. Correlation analysis of leaf shedding rate, branch hanging rate and agronomic traits in 2020.

Trait	DR	PH	EFB	H FNFB	H FNFH	NB	LA	FN	SPAD	HGY	BD	FBA	LIA
DR	1
PH	0.059	1
EFB	−0.710 **	0.018	1
H FNFB	0.031	0.194	0.298	1
H FNFH	0.035	0.752 **	0.247	0.251	1
NB	−0.528 *	0.135	0.161	0.126	0.386	1
LA	0.579 *	0.323	0.623 **	0.244	0.311	0.431 *	1
FN	−0.225	0.251	0.354	−0.233	−0.322	0.281	0.03	1
SPAD	−0.642 **	−0.042	−0.029	0.142	0.154	0.008	0.051	−0.036	1
HGY	−0.855 **	−0.132	0.444 *	0.035	−0.155	0.318	0.623 *	−0.132	0.414 *	1
BD	−0.420	0.326	0.035	0.253	0.129	0.261	−0.079	0.367	0.059	−0.215	1
FBA	−0.622 **	0.218	0.292	0.232	0.145	0.382	0.136	0.642*	0.006	0.522 *	0.264	1
LIA	−0.280	0.022	0.124	−0.001	0.054	0.073	−0.125	0.036	0.059	−0.512 *	0.124	0.225	1

Note: * and ** indicate significant (p < 0.05) or extremely significant (p < 0.01) correlation.

3.5. Analysis of Foliation Rate, Hanging Rate and Agronomic Traits in 2021

Differential analysis of defoliation and pruning rates of 273 samples in 2021 and the significance p-value of one-factor ANOVA test was less than 0.05, indicating significant differences in defoliation and pruning rates of 273 samples in 2020 (Table 7).

Table 7. Univariate ANOVA test.

	SS	DF	MSE	F	P
Hang branch rate Interblock	0.348	272	0.0013	2.552	0.02
Hang branch rate Within the group	0.425	810	0.0005
Hang branch rate Tote	0.673	1082
Take off the leaf rate Interblock	1.352	272	0.0050	2.778	0.02
Take off the leaf rate Within the group	1.484	813	0.0018
Take off the leaf rate Tote	2.846	1085

Through the analysis of the correlation between the incidence, defoliate rate and the plant height, the effect number, the incidence number, the total leaf number, the incidence, the incidence and the defoliate rate are significantly negative correlation, the chlorophyll SPAD value, the incidence rate, the incidence and the incidence are the lowest. There was a very significant negative correlation between leaf removal rate and hanging rate, and between leaf inclination angle and hanging rate, and a significant positive correlation between the number of effective branches, chlorophyll SPAD value, fruit branch angle and branch angle and hanging rate. The rate of defoliation and hanging branches were significantly correlated with the number of effective branches, the total number of leaves, the chlorophyll SPAD value, and the Angle of fruit branches (Table 8).

Table 8. Correlation analysis of leaf shedding rate, branch hanging rate and agronomic traits in 2021.

Trait	DR	PH	EFB	HFNFB	HFNFH	NB	LA	FN	SPAD	HGY	BD	FBA	LIA
DR	1
PH	0.259	1
EFB	−0.612 *	0.022	1
HFNFB	0.026	0.195	0.312	1
H FNFH	0.033	0.765 **	0.255	0.255	1
NB	−0.433	0.135	0.168	0.146	0.342	1
LA	0.369	0.269	0.323	0.254	0.318	0.631 *	1
FN	−0.245	0.256	0.364	−0.123	−0.222	0.001	0.162	1
SPAD	−0.668 **	−0.142	−0.029	0.142	0.276	0.008	0.051	−0.116	1
HGY	−0.732 **	−0.225	0.565 *	0.135	−0.165	0.276	0.814 *	−0.132	0.578 *	1
BD	−0.620 *	0.374	0.122	0.363	0.159	0.273	−0.179	0.367	0.009	−0.215	1
FBA	−0.673 **	0.246	0.345	0.252	0.145	0.286	0.134	0.756 *	0.006	0.548 *	0.275	1
LIA	−0.185	0.009	0.034	−0.001	0.007	0.178	−0.126	0.007	0.108	−0.512 *	0.11	0.205	1

Note: * and ** indicate significant (p < 0.05) or extremely significant (p < 0.01) correlation.

3.6. Correlation Analysis of Foliation Rate, Hanging Rate and Number of Leaves during Application

This study also found a significant positive correlation between the number of fol leaves after the first application during the application, and a significant negative correlation with the hanging rate. There was a very significant negative correlation with the peeling rate after the second application, and with the branch hanging rate. There was a significant negative correlation between the number of detached leaves after the first application and that after the second application (Table 9).

Table 9. Correlation analysis of foliation and hanging rate and number of leaves after application.

	Number of Leaves after the First Application NFS	Number of Leaves after the Second Application NSS	DR	HGY
Number of leaves leaves after the first application NFS	1
Number of leaves leaves after the second application NSS	−0.53 **	1
DR	0.591 **	−0.873 **	1
HGY	−0.315 *	0.612 **	−0.830 **	1

Note: * and ** indicate significant (p < 0.05) or extremely significant (p < 0.01) correlation.

3.7. Establishment of Phenotypic Indicators for Defoliating and Deciduous Traits

Based on the correlation of defoliation and pruning rates from 3 years data since 2019 to 2021 with other agronomic traits, observed the number of effective branches. The Chlorophyll SPAD values, Fruit branch angles both significantly correlated with defoliation and hanging rates over 3 years. ANOVA was done for these parameters. The results showed that 273 varieties had defoliation, hanging and effective branches in different years. The Chlorophyll SPAD values were found Non significant (Table 10). The number of effect branches is available, The Chlorophyll SPAD values, The Angle of fruit branches can be used as a reference index for evaluating the leaves and leaves of mechanized cotton varieties. Defoliation and hanging rates can be stable indicators for screening of deciduous cotton varieties. According to the requirements of machine mining cotton picking standard, refer to the “Technical Specification for evaluation of main agronomic traits of machine cotton picking by machine mining”, the defoliation rate is above 95%, and the hanging rate is below 8%. Combined 3-year field phenotypic traits were selected from 273 varieties for C6524, New Luzhong 62, Xiangcotton 11, Hubei-resistant cotton 33, E-Cotton 6, Dongting 1, Dai 45A, Cloth # # 3,363, Sparse catkins, H10, Source Cotton 11, China cotton 41 and other 11 varieties. Its chlorophyll SPAD values range from 49.2 to 62.9. The number of effective branches ranges from 6.4 to 8.3. The angle of fruit branches ranges from 43.5° to 61.2°. According to the actual demand for cotton production, the SPAD value of chlorophyll is 50 to 65, the number of effective branches is 7 to 8.5, and the angle of fruit branches is 45° to 60°, which can be used as a technical parameter to evaluate the characteristics of defoliation and defoliation of cotton. The results show that most of the materials are relatively general or poor, and the varieties need to be improved.

Table 10. Inter-annual univariate ANOVA test.

	DF	F	p
The rate of leaf removal between years Interblock	818	1.319	0.925
Hang branch rate between years Interblock	818	2.28	0.816
Effective branches exist several years apart Interblock	818	3.56	0.724
Inter-annual chlorophyll SPAD values Interblock	818	2.74	0.965
Fruit branches between the years Interblock	818	3.88	0.942

The gray correlation analysis of leaf shedding rate and other agronomic traits (hanging percentage, number of effective fruit branches, chlorophyll SPAD value, and Angle between fruit branches) showed that the correlation coefficient between leaf shedding rate and hanging percentage was the highest of 0.997, and the correlation degree between leaf shedding rate and Angle between fruit branches was the lowest of 0.947. The results showed that the hanging rate of branches had the greatest effect on the defoliation rate, while the Angle of branches had little effect (Table 11).

Table 11. Grey correlation degree of dew rate with other significantly related agronomic traits.

Evaluation Item	Correlation Degree	Ranking
HGY	0.996	2
EFB	0.995	3
SPAD	0.997	1
FBA	0.947	4

4. Discussion

Cotton leaves are mainly determined by the hormone level in the crop body, after the use of peeling agent which can quickly enhance the plant synthesis of ethylene and abscisic acid, and produce high content of ripening hormone ethylene and abscisic acid, and inhibit prohormone auxin. GA, cytokinin transport in the plant, promote plant nutrition from “source” to “library”, and ultimately accelerate the process of plant aging and maturity. The relevant research results show that spraying the defoliation agent can promote the generation of cotton leaf abscisic acid and ethylene, which promote the formation of cotton petiole obilayer, and resulted in falling off the cotton leaves and finally achieve the effect of defoliation and ripening [17,18]. The factors affecting cotton defoliation are more complex. Du Gangfeng et al. found that the peeling rate of chemical jacking cotton at 8 d after spraying agent was more than 90%, and could improve the quality of peeling and reduce the hanging rate [19]. Zhang Wen et al. found that the peeling rate of 20 d after spraying the different peeling agent reached more than 70%, and the peeling effect of spraying the peeling agent twice was slightly better than that of one spraying [20]. Du et al. found that Champion element (COR), as a non-host-specific phytotoxin, causes defoliation and fruit shedding, and induced cotton defoliation through ethylene signaling and regulation of hydrolase activity [21]. Fan qinglu’s study found that the effect of defoliation was significantly different in the defoliation rate due to different varieties [22]. This may be due to differences in susceptibility to folispecies or the presence of agronomic traits unfavorable to plant absorption of the agent. The defoliation effect of 273 terrestrial cotton germplasm resources selected in this experiment was also significantly different under the action of the same defoliation agent, which was consistent with Fan Qingfu’s study. Li Jianwei, Wu Penghao and others found that the dense number of effective branches would lead to uneven spraying of foliation agent and poor medicinal effect of the leaves, resulting in the reduction of foliation rate and the increase of branch hanging rate [23]. Zhu Xiefei and others found a significant positive correlation between the number of effective branches and the yield. However, this study found that there was a significant negative correlation between the size of the number of effective branches and the defoliation rate. Too many effective branches may be too dense, and the defoliation agent could not be evenly sprayed in the middle and lower parts of the plant, resulting in a decrease in the defoliation rate. But reducing the number of effective branches can also affect the yield. It is speculated that the number of effective branches of cotton varieties should be moderate, and too many effective branches will affect the defoliation rate, and too few effective branches will affect the yield [24]. Song Min analyzed from the fertility period that the rate of defoliation in different periods showed a negative relationship with the fertility period of the varieties. The longer the variety growth period, the weaker the sensitivity to the defoliation agent, and the variety selection must be controlled within a reasonable reproductive period range [25], This is consistent with the results of the present study, where fertility period is significantly negatively associated with defoliation rate, and cultivar ripening will be an important measure of the sensitivity of cotton varieties to foliation agents.

Gao Lili et al. found that leaf senescence by reducing the rate of photosynthesis, thus increasing the rate of defoliation [7]. Courban, Xia Dong et al. found that by appropriately adjusting the frequency of drip irrigation to adjust the water content of soil 20–40 cm underground, the fluorescence parameters of chlorophyll can finally achieve better defoliation effect [26]. Wang Xiaojing, Li Sijia et al. found that the net photosynthetic rate of the medicated leaves and the net photosynthetic rate of the adjacent unmedicated leaves were also decreased, and the net photosynthetic rate of the adjacent leaves recovered quickly after the medicated leaves fall off [12]. In this study, chlorophyll value has a very significant negative correlation with defoliation rate, and a significant positive correlation with hanging rate, probably because high chlorophyll SPAD value and high photosynthetic rate will affect the process of peeling agent reducing the photosynthetic rate, delay leaf senescence, thus reducing the peeling rate and increasing hanging rate [8]. Therefore, this study believes that the frequency of drip irrigation should be appropriately adjusted before spraying the defoliation agent, so that the reduction of chlorophyll in the plant leaves can effectively improve the comprehensive defoliation effect. In addition, machine cotton requires compact cotton strains, short fruit branches, “fried dough sticks” less, bell concentration. This kind of strain type is convenient for defoliation agent spraying and mechanical harvesting operation, the strain type is more compact, good ventilation and light transmission, is conducive to agent spraying and cotton leaf medicine, play the effect of defoliation agent.

Zhu Jijie et al. found that different varieties had different speed and effect after spraying the same concentration and amount of foliation agent [27]. In this study, the defoliation rate found a significant correlation with the number of defoliation after application, and a very significant positive correlation with the number after the first application and the number after the second application. In this study, the result may be related to the sensitivity of the plant to the peeling agent. After the first application, the varieties with high peeling rate are high and low. After the second application, the varieties with less sensitive to the peeling agent began, but the final peeling rate could not reach a good level.

The defoliation ability of varieties is influenced by a combination of multiple agronomic traits, and it is difficult to assess its direct contribution to defoliation and pruning rates because these variables are difficult to control. However, varieties with high defoliation rates often have more favorable traits, and varieties with low rates are often influenced by other inferior trait factors [12,22]. That is, varieties with high defoliation rate and low hanging rate must have many advantageous traits, but some advantageous traits may not obtain high peeling rate and low hanging rate or high yield and high quality, which is a sufficient and unnecessary condition. For example, research has also found that the. Too high plant (for example, 17N10 plant high 93 cm) or too low plant high (for example, kk-351 plant high 47 cm) can lead to low peeling rate and high hanging rate, which indicates that some traits are too extreme can also affect the peeling rate and hanging rate.

Because of the selection of this experiment has certain limitations, and defoliation rate to a certain extent is also affected by climate temperature, spray time, so the experimental results need further verification, and defoliation traits should be further combined with yield traits, quality traits, screening defoliation, high yield, good quality varieties, lay a foundation for cotton varieties breeding.

5. Conclusions

Scientific and reasonable spraying technology of cotton defoliation ripening agent can improve the quality of cotton defoliation and reduce the impurities of broken leaves in seed cotton, which is of great significance to solve the quality problem of cotton. However, in the actual production, the application of defoliating ripening agent is often affected by many factors, among which the application equipment and application technology play an important role. From the earliest manual spraying, to the spray rod spraying machine, and then to the rapid development of plant protection drone spraying in recent years, the application efficiency of cotton defoliation ripening agent is constantly improving, which plays an important role in the high yield and stable yield of cotton. In this study, through the analysis of defoliation rate, hanging rate and 11 correlation traits of germplasm resources, C6524, 62 in Xinzhong 62, Xiangcotton 11, Hubei cotton 33, Hubei cotton 6, Dongting 1, Dai 45A, cloth 3363, sparse wool H10, source cotton 11, medium cotton 11 and 41; field peeling rate of 95.1–100%, the hanging rate is 0~4.1%, can be used as mechanical cotton breeding materials. And through the analysis of the correlation between 11 agronomic traits and leaf rate and branch rate, found that effective number, chlorophyll SPAD value, fruit branch angle and leaf rate and branch rate in 3 years, have significant correlation, the correlation analysis results showed that besides leaf rate, fruit branches, effect number, chlorophyll SPAD value can be used as an index of leaf ability.