*3.3. Relationship of Temperature and Solar Radiation Parameters with Relative LI and* 4*BM*

The relative LI increased significantly with the *Tmean* in each year (Figure 4A). Contrastingly, the relative 4BM decreased significantly with the *Tmean* in each year (Figure 4E). No significant correlations between the EAT and the relative LI or 4BM were observed in each year (Figure 4B,F). Significant linear correlations between the *Rmean* and the relative LI and 4BM were found in each year except in 2015 (Figure 4C,G). However, the linear correlations were reversed in 2017 and 2018. The relative LI showed a significantly negative correlation with the CSR in 2015 and 2017 (Figure 4D). The relative 4BM displayed the opposite result, as was expected (Figure 4H). However, both the relative LI and 4BM showed no significant correlation with the CSR in 2018.

showed no significant correlation with the CSR in 2018.

R379 5442.30 a 3728.34 b 3329.22 b 1836.23 a 1144.77 b 847.96 c 2181.02 a 1483.45 b 1454.91 b 62.88 9311 3678.55 a 2560.75 b 1839.95 c 727.80 a 407.82 b 461.61 ab 1043.13 a 663.80 b 488.15 b 86.82 Jiangan 5089.65 a 4192.39 ab 3563.15 b 2838.87 a 2033.76 b 1280.59 c 3062.08 a 2090.83 b 1609.43 c 43.87

5% probability level.

a CV, coefficient of variation of the △BM across the nine sowing dates. b Within a row for each year, means followed by the same letters are not significantly different determined by the Tukey test at

The relative LI increased significantly with the *Tmean* in each year (Figure 4A). Contrastingly, the relative △BM decreased significantly with the *Tmean* in each year (Figure 4E). No significant correlations between the EAT and the relative LI or △BM were observed in each year (Figure 4B,F). Significant linear correlations between the *Rmean* and the relative LI and △BM were found in each year except in 2015 (Figure 4C,G). However, the linear correlations were reversed in 2017 and 2018. The relative LI showed a significantly negative correlation with the CSR in 2015 and 2017 (Figure 4D). The relative △BM displayed the opposite result, as was expected (Figure 4H). However, both the relative LI and △BM

△*BM* 

*3.3. Relationship of Temperature and Solar Radiation Parameters with Relative LI and* 

**Figure 4.** Correlations between the relative LI of the twelve *indica* rice cultivars and average daily temperature (*Tmean*) (**A**) and solar radiation (*Rmean*) (**B**), effective accumulated temperature (EAT) (**C**) and cumulative solar radiation (CSR) (**D**) on the three sowing dates of each year. Correlations between the relative △BM of the twelve *indica* rice cultivars and *Tmean* (**E**), *Rmean* (**F**), EAT (**G**), and CSR (**H**) on the three sowing dates of each year. The green, blue, and red points represent the cultivars measured in 2015 at Xindu, and in 2017 and 2018 at Ezhou, respectively. *n* = 36 in each year. \* and \*\*, significant differences at *p* < 0.05 and *p* < 0.01, respectively. **Figure 4.** Correlations between the relative LI of the twelve *indica* rice cultivars and average daily temperature (*Tmean*) (**A**) and solar radiation (*Rmean*) (**B**), effective accumulated temperature (EAT) (**C**) and cumulative solar radiation (CSR) (**D**) on the three sowing dates of each year. Correlations between the relative 4BM of the twelve *indica* rice cultivars and *Tmean* (**E**), *Rmean* (**F**), EAT (**G**), and CSR (**H**) on the three sowing dates of each year. The green, blue, and red points represent the cultivars measured in 2015 at Xindu, and in 2017 and 2018 at Ezhou, respectively. *n* = 36 in each year. \* and \*\*, significant differences at *p* < 0.05 and *p* < 0.01, respectively.

When the nine temperature and solar radiation treatments were combined, significant correlations were also found between the temperature and solar radiation parameters and the relative LI and △BM (Figure 5). There were significant positive linear correlations between the relative LI and *Tmean* and EAT (Figure 5A,B). However, a significant quadratic relationship between the relative LI and *Rmean* was observed. The relative LI increased significantly with the *Rmean* from 2015 at Xindu to 2017 at Ezhou, and then decreased slightly from 2017 to 2018 at Ezhou (Figure 5C). In addition, the relative LI and △BM showed no significant correlation with the CSR (Figure 5D,H). When referred to the correlations between the relative △BM and *Tmean*, EAT and *Rmean*, they displayed the reversed results of the relative LI (Figure 5E–G). When the nine temperature and solar radiation treatments were combined, significant correlations were also found between the temperature and solar radiation parameters and the relative LI and 4BM (Figure 5). There were significant positive linear correlations between the relative LI and *Tmean* and EAT (Figure 5A,B). However, a significant quadratic relationship between the relative LI and *Rmean* was observed. The relative LI increased significantly with the *Rmean* from 2015 at Xindu to 2017 at Ezhou, and then decreased slightly from 2017 to 2018 at Ezhou (Figure 5C). In addition, the relative LI and 4BM showed no significant correlation with the CSR (Figure 5D,H). When referred to the correlations between the relative 4BM and *Tmean*, EAT and *Rmean*, they displayed the reversed results of the relative LI (Figure 5E–G). *Agronomy* **2022**, *12*, 2603 11 of 19

**Figure 5.** Correlations between the relative LI of the twelve *indica* rice cultivars and *Tmean* (**A**), *Rmean* (**B**), EAT (**C**), and CSR (**D**) on all the sowing dates across two locations and three years. Correlations between the relative △BM of the twelve *indica* rice cultivars and *Tmean* (**E**), *Rmean* (**F**), EAT (**G**), and CSR (**H**) on all the sowing dates across two locations and three years. The green, blue, and red points represent the cultivars measured in 2015 at Xindu, and in 2017 and 2018 at Ezhou, respectively. *n* = 36 in each year. \*\*, significant differences at *p* < 0.01. **Figure 5.** Correlations between the relative LI of the twelve *indica* rice cultivars and *Tmean* (**A**), *Rmean* (**B**), EAT (**C**), and CSR (**D**) on all the sowing dates across two locations and three years. Correlations between the relative 4BM of the twelve *indica* rice cultivars and *Tmean* (**E**), *Rmean* (**F**), EAT (**G**), and CSR (**H**) on all the sowing dates across two locations and three years. The green, blue, and red points represent the cultivars measured in 2015 at Xindu, and in 2017 and 2018 at Ezhou, respectively. *n* = 36 in each year. \*\*, significant differences at *p* < 0.01.

In order to determine the main temperature and solar radiation parameters that affected the lodging resistance of 12 *indica* rice cultivars, a stepwise regression analysis was

the *Tmean* and *Rmean* had significant effects on the lodging resistance within the temperature and solar radiation parameters. The relative LI increased with the *Tmean* but decreased with the *Rmean* (Figure 6A). The relative △BM decreased with the *Tmean* but increased with the

**Figure 6.** Stepwise regression analysis of the temperature and solar radiation parameters with the relative LI (**A**) and △BM (**B**) of twelve *indica* rice cultivars. *Tmean*, average daily temperature. *Rmean*,

Following this, the direct and indirect effects of the *Tmean* and *Rmean* on the relative LI and △BM were investigated using a path analysis to reveal the relative importance of the *Tmean* and *Rmean* on the lodging resistance of 12 *indica* rice cultivars. The results of the path

average daily solar radiation. *n* = 108. \*\*, significant differences at *p* < 0.01.

*Rmean* (Figure 6B).

In order to determine the main temperature and solar radiation parameters that affected the lodging resistance of 12 *indica* rice cultivars, a stepwise regression analysis was conducted (Figure 6). The results of the stepwise regression analysis indicated that only the *Tmean* and *Rmean* had significant effects on the lodging resistance within the temperature and solar radiation parameters. The relative LI increased with the *Tmean* but decreased with the *Rmean* (Figure 6A). The relative 4BM decreased with the *Tmean* but increased with the *Rmean* (Figure 6B). fected the lodging resistance of 12 *indica* rice cultivars, a stepwise regression analysis was conducted (Figure 6). The results of the stepwise regression analysis indicated that only the *Tmean* and *Rmean* had significant effects on the lodging resistance within the temperature and solar radiation parameters. The relative LI increased with the *Tmean* but decreased with the *Rmean* (Figure 6A). The relative △BM decreased with the *Tmean* but increased with the *Rmean* (Figure 6B).

36 in each year. \*\*, significant differences at *p* < 0.01.

**Figure 5.** Correlations between the relative LI of the twelve *indica* rice cultivars and *Tmean* (**A**), *Rmean* (**B**), EAT (**C**), and CSR (**D**) on all the sowing dates across two locations and three years. Correlations between the relative △BM of the twelve *indica* rice cultivars and *Tmean* (**E**), *Rmean* (**F**), EAT (**G**), and CSR (**H**) on all the sowing dates across two locations and three years. The green, blue, and red points represent the cultivars measured in 2015 at Xindu, and in 2017 and 2018 at Ezhou, respectively. *n* =

In order to determine the main temperature and solar radiation parameters that af-

*Agronomy* **2022**, *12*, 2603 11 of 19

**Figure 6.** Stepwise regression analysis of the temperature and solar radiation parameters with the relative LI (**A**) and △BM (**B**) of twelve *indica* rice cultivars. *Tmean*, average daily temperature. *Rmean*, **Figure 6.** Stepwise regression analysis of the temperature and solar radiation parameters with the relative LI (**A**) and 4BM (**B**) of twelve *indica* rice cultivars. *Tmean*, average daily temperature. *Rmean*, average daily solar radiation. *n* = 108. \*\*, significant differences at *p* < 0.01.

average daily solar radiation. *n* = 108. \*\*, significant differences at *p* < 0.01. Following this, the direct and indirect effects of the *Tmean* and *Rmean* on the relative LI and △BM were investigated using a path analysis to reveal the relative importance of the *Tmean* and *Rmean* on the lodging resistance of 12 *indica* rice cultivars. The results of the path Following this, the direct and indirect effects of the *Tmean* and *Rmean* on the relative LI and 4BM were investigated using a path analysis to reveal the relative importance of the *Tmean* and *Rmean* on the lodging resistance of 12 *indica* rice cultivars. The results of the path analysis are shown in Table 4. The direct effect of the *Tmean* on the relative LI was 1.556, and the indirect effect of *Tmean* on the relative LI through the *Rmean* was −0.673. The direct effect of the *Rmean* on the relative LI was –0.759, and the indirect effect of the *Rmean* on the relative LI through the *Tmean* was 1.380. These results indicated that the *Tmean* and *Rmean* independently had a positive and negative effect on the relative LI, respectively. Additionally, the absolute value of the direct effect of the *Tmean* (1.556) was higher than that of the *Rmean* (0.759), and the correlation coefficients of the relative LI with the *Tmean* and *Rmean* were all positive values (*r* = 0.883 \*\* and 0. 621 \*\*, respectively), which implied that the *Tmean* had a greater effect on the lodging resistance than the *Rmean*. The same results could also be obtained from the direct and indirect effects of the *Tmean* and *Rmean* on the relative 4BM. Thus, the correlations between the *Rmean* and lodging-related traits were not analyzed in the following part.

**Table 4.** The direct and indirect effects of the *Tmean* and *Rmean* on the relative LI and 4BM of twelve *indica* rice cultivars.


<sup>a</sup> *Tmean* and *Rmean* indicate the average daily temperature and average daily solar radiation, respectively. b \*\* indicate significant differences at *p* < 0.01.

#### *3.4. Correlation analysis of the Tmean and Relative Lodging-Related Traits*

In order to clarify the lodging-related traits affected by the *Tmean*, the correlations between them were analyzed (Table 5). The relative values of the breaking resistance (BR) and bending moment at breaking (M) were mostly affected by the *Tmean* with the highest coefficient *r* on the three sowing dates of each year and on all the sowing dates across two locations and three years. Furthermore, the relative value of the stem length (SL) had highly significant positive correlations with the *Tmean* on the three sowing dates of each year and on all the sowing dates across the two locations and three years. However, the rest lodging-related traits showed inconsistent correlations with the *Tmean*. These results implied that the BR, M and SL were the major lodging-related traits affected by the *Tmean*.

**Table 5.** Analysis of the correlations between the relative value of lodging-related traits of twelve *indica* rice cultivars and the *Tmean* during the determined growth durations on the three sowing dates of different years and on all the sowing dates across two locations and three years.


<sup>a</sup> CL, culm length; SL, stem length; PL, panicle length; BR, breaking resistance; FW, fresh weight; BM, bending moment of the whole plant; M, bending moment at breaking. <sup>b</sup> \* and \*\* indicate significant differences at *p* < 0.05 and *p* < 0.01, respectively.

The M value, which is used to determine the physical strength of the culm, can be further divided into two parts: the section modulus (SM), which is directly influenced by the culm diameter (CD) and culm wall thickness (CT); the bending stress (BS), which is an indicator of culm stiffness. In order to determine the reason behind the significant variation in the M value affected by the *Tmean*, the SM, BS, and other lodging-related traits were measured in 2018 at Ezhou. Additionally, correlation analysis was also conducted between the measured lodging-related traits and the *Tmean* (Table 6). The results of the correlation analysis showed that the *Tmean* had significant negative correlations with the relative value of CD, CT, SM and cellulose content (CC), but displayed significant positive correlations with the relative value of lignin content (LC), total content of lignin and cellulose (TC), and bending stress (BS). In addition, the absolute value of *r* between the *Tmean* and the relative value of SM was 0.837, which was higher than that between the *Tmean* and the relative value of BS (0.437).

**Table 6.** Analysis of the correlations between the relative value of lodging-related traits of twelve *indica* rice cultivars and the *Tmean* during the determined growth durations on the three sowing dates in 2018 at Ezhou.


<sup>a</sup> CD, culm diameter; CT, culm wall thickness; SM, section modulus; LC, lignin content; CC, cellulose content; TC, total content of lignin and cellulose; BS, bending stress. <sup>b</sup> \* and \*\* indicate significant differences at *p* < 0.05 and *p* < 0.01, respectively.

#### **4. Discussions**

As both of the experimental sites have a subtropical monsoon type of climate, the temperatures gradually increase from spring to summer and decrease in autumn (Figure 1). However, Xindu is in the basin area, while Ezhou is in the plain area. Xindu is situated at a higher altitude than Ezhou. The differences in topography and altitude generate different temperature and solar radiation conditions between the two locations (Figure 2). Rice is often sown in the late spring at Xindu and Ezhou, and delayed sowing will make rice suffer

from high temperatures sooner. Since high temperatures accelerated the growth of rice and advanced the rice maturity [78,79], the determined growth durations reduced with the delay in the sowing dates (Table 1). Hence, the average daily temperatures (*Tmean*) during the determined growth durations increased with the delayed sowing dates (Figure 2). As a high temperature tended to be accompanied by higher solar radiation [69], the average daily solar radiation (*Rmean*) also increased with the *Tmean* from SD7 to SD9. However, the *Rmean* varied little from SD1 to SD3 and even decreased from SD4 to SD6. This was due to the increase in cloudy and rainy days, which resulted in low sunshine hours and solar radiation. Delayed sowing can enhance the lodging resistance in winter wheat [80]. However, the lodging resistance of *indica* rice cultivars weakened with the delayed sowing dates in our research (Figure 3). Therefore, optimizing sowing dates could be a useful strategy for improving lodging resistance and growing crops to adapt to climate change.

In this study, we found that whether the range of the *Tmean* was less than 1 ◦C across the three sowing dates in each year or more than 4 ◦C across the two locations (Figure 2), the increased *Tmean* significantly decreased the lodging resistance of *indica* rice cultivars (Figures 4A,E and 5A,E), which indicated that the lodging resistance was more sensitive to the *Tmean* than the other temperature and solar radiation parameters. This could also be demonstrated by the results of stepwise regression analysis and path analysis. However, the response of the lodging resistance to other crops could be different. Previous research found that high temperatures showed an inconsistent effect on the stem-lodging resistance in canola plants, but reduced the root-lodging resistance significantly [81]. The results of stepwise regression analysis and path analysis also showed that the *Rmean* had a positive effect on the lodging resistance (Figure 5, Table 4). That is to say, an increase in the *Rmean* would improve the lodging resistance. However, this effect was not detectable under the greater negative influence of increased *Tmean* on the lodging resistance across the three sowing dates in 2018 and across the two locations (Figures 4C,G and 5C,G). However, the lodging resistance displayed a slightly increasing trend from 2017 to 2018 at Ezhou (Figure 5C,G). We inferred that the positive effect of the increased *Rmean* on the lodging resistance outweighed the negative effect of the increased *Tmean* because of the larger differences (4*Rmean* = 2.1 MJ m−<sup>2</sup> <sup>d</sup> −1 ) in the *Rmean* than the differences (4*Tmean* = 0.8 ◦C) in the *Tmean* between 2017 and 2018 at Ezhou (Figure 2). Several researchers found that low solar radiation could promote the vertical elongation of plants, such as an increase in the internode length and plant height, and restrain the lateral growth, such as a decrease in the culm diameter and wall thickness, and reduce cell wall carbohydrates; thus, low solar radiation could cause crop lodging [82–87]. Our results were similar to the previous findings, i.e., the *Rmean* was positively correlated with the lodging resistance. However, under the conditions of our present study, since the *Tmean* had a greater effect than the *Rmean* (Table 4), we could not establish the relationship between solar radiation and lodging-related traits.

According to the correlation analysis between the *Tmean* and lodging-related traits (Table 5), we found that with the increased *Tmean*, it was mainly the reduction in breaking resistance (BR) and the bending moment at breaking (M) of the basal second internode that weakened the lodging resistance of *indica* rice cultivars. The decreased stem-bending strength due to short periods of high temperature stress could also be found in canola [88]. In addition, the stem length (SL) of the basal second internode increased significantly with the increased *Tmean*, which could be a reason for the decline in the BR [89].

The culm stiffness of rice plants, which is expressed by the bending stress (BS), is a product of the structural carbohydrates (primarily lignin and cellulose) and non-structural carbohydrates contents at the lower internode. Carbohydrates in rice culms are accumulated before heading and transported to the ears after heading and it is mainly the non-structural carbohydrates which are transported to grains for grain filling [90,91]. Thus, the culm stiffness increases with more structural carbohydrates in the basal stems [33,41]. Li et al. [92] observed that the cellulose content (CC) of rice *brittle culm1* (*bc1*) was lower than that of the wild type, but the lignin content (LC) was higher. However, the total

content of lignin and cellulose (TC) of *bc1* was lower than that of the wild type, which led to the weaker culm strength of *bc1*. In our study, the increase in LC was more than the decrease in CC as the *Tmean* increased. As a result, the TC increased with the *Tmean*, which caused the increased BS (Table 6). The physical strength of the culm, which was represented by the M value, was significantly and positively correlated with the BS and section modulus (SM) [20,32]. Despite the BS increasing with the *Tmean*, the decrease in the SM was even more important with the higher correlation coefficient, which was responsible for the significant reduction in the M value (Table 6). Additionally, the decrease in the SM was attributed to the reduced culm diameter (CD) and culm wall thickness (CT) of the basal second internode with the increased *Tmean*. As a consequence, CD, CT, and SL were the major factors influencing the physical strength of the culm under the effect of temperature.

The soil properties of the experimental fields at Xindu and Ezhou were different, which might affect the lodging resistance of tested rice cultivars. However, rice had been cultivated in the experimental fields for many years, which indicated that the soil properties were suitable for rice growth and development. In addition, the fertilizer applications between the experimental fields at Xindu and Ezhou were identical, resulting in a similar uptake of nitrogen, potassium and silicon from the soil, which caused the soil properties to have little influence on the lodging resistance [93]. In the research of Niu, Feng, Ding and Li [47], strong wind and heavy rain were the two most important causes of lodging. Even though there was no strong wind and heavy rain during the determined growth durations in our study, we selected 12 *indica* rice cultivars which did not lodge in the fields for analysis in order to exclude the influence of wind and rain on the plant lodging. Therefore, ordinary wind and rain had little effect on the lodging resistance of *indica* rice cultivars.

Although the lodging resistance of 12 *indica* rice cultivars significantly weakened with the increased temperature, each of the cultivars responded differently to the temperature (Table 3). The lodging resistance of lodging-moderate cultivar Chuanxiang 29B was most affected by the temperature with the highest coefficient of variation, and that of lodgingresistant cultivar Jiangan responded least to the temperature with the lowest coefficient of variation. This implied that the lodging-resistant *indica* cultivar had the potential to adapt to a higher temperature.

#### **5. Conclusions**

The average daily temperature gradually increased by about 1 ◦C across the three sowing dates in each year and significantly increased by about 5 ◦C across the two locations. The average solar radiation varied inconsistently across the three years but prominently increased by about 4 MJ m−<sup>2</sup> day−<sup>1</sup> across the two locations. Under this condition, the temperature had greater negative effects on the lodging resistance of *indica* rice cultivars than the positive effects of solar radiation. That is to say, temperature was the main factor that affected the lodging resistance of *indica* rice cultivars. With the delay in the sowing dates, the lodging resistance of *indica* rice cultivars showed a significant decrease (LI increased by 22.65%, 19.82% and 17.58% in 2015, 2017 and 2018, respectively; 4BM decreased by 35.52%, 57.62% and 47.93% in 2015, 2017 and 2018, respectively) along with an increase in the average daily temperature. Among the morphological traits, the culm diameter and culm wall thickness decreased, and the stem length increased significantly with the increased average daily temperature, which led to the slender basal second internode. As the biochemical trait of the lignin content increased more than the cellulose content decreased, the total content of the lignin and cellulose increased with the increased average daily temperature, which resulted in the increased culm stiffness. In summary, owing to the slender basal second internode, the mechanical trait of the bending moment at breaking decreased significantly with the increased average daily temperature, increasing the potential risk of lodging in *indica* rice cultivars. The slender basal internodes would become a critical reason for the *indica* rice lodging with rising temperatures due to global warming.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/agronomy12112603/s1, Table S1: The genetic background and the origin of the experimental rice cultivars classified into three different lodging resistance groups; Table S2: Analysis of variance of temperature and solar radiation parameters, LI, and 4BM.

**Author Contributions:** Conceptualization, R.Z. and X.W.; Methodology, X.L., Z.W., L.F. and X.W.; Software, X.L.; Formal Analysis, X.L. and Z.W.; Investigation, X.L., Z.W., L.F., Z.Y. and T.L.; Resources, R.Z.; Data Curation, X.L. and Z.W.; Writing—Original Draft Preparation, X.L.; Writing—Review & Editing, Z.D., W.L. and X.W.; Visualization, X.L., Z.D. and X.W.; Supervision, R.Z. and Z.H.; Project Administration, X.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Key Research and Development Project (2021YFE0101000), the China Agriculture Research System (CARS-01-08), and the Key R&D Program of Hubei Province in China (2021BBA079).

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to their relevance to an ongoing Ph.D. thesis.

**Acknowledgments:** We thank the staff of the State Key Laboratory of Hybrid Rice, Wuhan University and the Crop Research Institute, Sichuan Academy of Agricultural Science for their experimental support.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

### **References**

