Growth Performance and Adaptability of Sugarcane Combinations Under Different Ecological Environments

Shiyun TANG　Rongzhong YANG　Hui ZHOU　Yan JING　Hongwei TAN

Abstract The experiments of sugarcane combination evaluation were conducted on 119 combinations during 2017-2018 crossing season in Nanning， Chongzuo and Laibin of Guangxi. The conjoint analysis of variance and estimation of genetic parameters were performed based on brix weight. Furthermore， combination stability was analyzed by regression analysis model and AMMI model. The results showed that differences in brix weight among combinations， environments and interaction between environments and combinations were all extremely significant （P<0.01）， and the broad-sense heritability of brix weight belonged to medium level or slightly lower at the three sites. In both Chongzuo and Laibin， the variation coefficients of brix weight were large， but that in Nanning was small. Combinations 643， 404， 575， 972， 636， 144， YC95， 1470， 755， 409， 701， 832， YC37 and 579 performed high yield， high genetic stability and stronger adaptability. The results of comprehensive analysis in both brix weight and selection rate of combination showed that 449， YC127， 796， YC44， 533， 570， YC123， 391， 546， 403， YC90 and 252 performed high brix weight and high selection rate of combination in Nanning; combinations 643， 212， YC61， 432， 903， YC95， YC44， 368， YC83， YC127， YC112， 701， 411， YC90 and YC123 demonstrated high brix weight and high selection rate of combination in Chongzuo; and combinations 643， 404， 449， 144， 403 and YC48 had high brix weight and high selection rate of combination in Laibin.

Key words Sugarcane; Combination; Brix weight; Stability; Adaptability

Received： July 25， 2020  Accepted： September 29， 2020

Supported by Guangxi Innovation-driven Development Project （GK AA17202042-4）; Special Fund for Sugar Crop Research System Construction （CARS_17015）; Guangxi Sugarcane Innovation Team Construction Project of National Modern Agriculture Industrial Technology System （gjnytxgxcxtd-03-01）.

Shiyun TANG （1978-）， male， P. R. China， associate researcher， PhD， devoted to research about breeding of new sugarcane varieties.

*Corresponding author. E-mail： tangshiyunok@163.com.

Sugarcane is Chinas most important sugar crop and plays an important role in ensuring Chinas sugar safety. The selection and promotion of new sugarcane varieties provide an important material basis for sugarcane planting and production. At present， through a large number of sugarcane family tests at home and abroad， the commonly used parents and combinations of sugarcane have been evaluated， and a large number of excellent parents and combinations have been screened， which has significantly improved the efficiency of sugarcane breeding[1-4]. High-yield and high-sugar pyramiding breeding is the main goal of sugarcane breeding[5]. Brix weight integrates the two factors of sugarcane yield and brix weight. It is one of the comprehensive characteristics that reflect the sugar content of sugarcane， as well as an important indicator to measure the comprehensive performance of sugarcane sexual generations[6]. It is often used as an important target trait for analyzing sugarcane seedling population[5] and family evaluation[7].

The use of crop genotype and environment interaction to carry out breeding for variety stability and extensive adaptability has become one of the important ways to improve crop productivity[8]. Researchers have conducted a lot of research on the interaction between new sugarcane varieties and environments[9-12]， and it is generally believed that the interaction between sugarcane varieties and environments is widespread. Deng et al.[13] used stability analysis methods to study the response of sugarcane yield and quality traits to different ecological environments， and predicted the production capacity and promotion scope of genotypes. However， under different ecological environmental conditions in Guangxi， there are still few reports on the genetic performance of sugarcane combinations， the interaction between combinations and environments， and the adaptability of combinations. In this study， a joint experiment of sugarcane combinations was carried out in three main sugarcane production areas in Guangxi to study the interaction between combinations and environments， aiming to provide a basis for screening good combinations suitable for different sugarcane production areas in Guangxi and for sugarcane regional breeding.

Materials and Methods

Materials

The test materials were seedlings prepared by Sugarcane Research Institute， Guangxi Academy of Agricultural Sciences in 2018， and 119 combinations were selected. The combinations with combination codes which were pure Arabic numerals were prepared by the Hainan Sugarcane Breeding Base of Sugarcane Research Institute， Guangxi Academy of Agricultural Sciences; and the combinations with the combination codes starting with YC were prepared by Hainan Sugarcane Breeding Farm which was commissioned. The parents used were mainly new varieties （lines） of Guitang series， and also covered series of parents from Guangdong， Yunnan， Fujian， Taiwan and abroad. The parental information of the combinations is shown in Table 1.

Experimental locations

There were three experimental locations： ① Dingdang Base （E1） of Sugarcane Research Institute， Guangxi Academy of Agricultural Sciences， Dingdang Town， Longan County， Nanning City， Guangxi， ② Fusui Agricultural Science Research Institute Base （E2）， Changping Township， Fusui County， Chongzuo City， Guangxi， and ③ Xingbin Agricultural Science Institute Base （E3）， Qiaogong Township， Xingbin District， Laibin City， Guangxi. The basic conditions of the experimental locations are shown in Table 2. The annual precipitation and annual average temperature in the table are provided by the local meteorological department.

Experimental design

Seeds were sown on March 9， 2018， and seedlings were subjected to heeling and transplanting on April 27. 119 combinations of seedlings were transported to Chongzuo， Laibin and Nanning test sites for field transplantation on June 14-20. Randomized block design was adopted. Each treatment had three replicates. One row was planted in each treatment with a row length of 7 m and a row spacing of 1.2 m， and 23 individual plants were planted in each row. After the seedlings survived， fertilization， soil cultivation and other field management were carried out according to conventional methods.

The combination investigation was conducted at Laibin site on January 23-25， 2019， Fusui site on February 1-3， 2019， and Nanning site on February 14-17， 2019. The plant height， stem diameter， millable stalk number per clump and brix weight of 15 clumps were investigated in each plot， and the brix weight of each clump of sugarcane was calculated by referring to the method of Xu et al.[14]. The brix weight of each combination at each site was expressed as the average of the three replicates. After the character investigation of the sugarcane combinations， the seed selection of sugarcane was carried out according to the conventional method at each test site. The combination selection rate was calculated with the percentage obtained by dividing the number of selected single plants by the total number of single plants. The combination selection rates were statistically analyzed with plot as a unit， and the result was expressed as the average of the three replicates.

Statistical Analysis

For the brix weights of the combinations at each test site， the analysis of variance and LSD multiple comparison were performed for single test points. According to the method of Liu et al.[15]， the phenotypic and genetic variation coefficients and broad-sense heritability of brix weight were estimated. Then， DPS 17.10 was used to perform multi-site conjoint analysis of variance， regression analysis and AMMI model analysis on brix weight at the three test sites[16]. The stability of the combinations was evaluated by the regression coefficient bi and the Dg value of the relative stability of the genotype in the AMMI model.

The brix weight effects of combinations at each test site were sorted from high to low， and combined with the combination selection rate of each test site， the combinations with higher brix weight and combination selection rate at each test site were selected.

Results and Analysis

Brix weight performance and genetic parameter estimation of combinations at each test site

The analysis of variance with a one-way randomized block design and genetic parameter estimation were performed on the brix weight of combinations at each test site. It can be seen from Table 3 that the brix weights at the three test sites had extremely significant differences among the combinations. The average brix weights from high to low were E1， E2， and E3， indicating that the sugarcane brix weight at the Nanning site was the highest， followed by the Chongzuo site， and the Laibin site had the lowest sugarcane brix weight. From the point of view of coefficient of variation， the phenotypic variation coefficient and genetic variation coefficient of sugarcane brix weight at E1 were both the smallest， and the genetic variation coefficient and the phenotypic variation coefficient of sugarcane brix weight were the largest at E3 and E2， respectively. The Laibin site and Chongzuo site had relatively larger brix weight variation coefficients， which is conducive to the selection of excellent combinations and clones. From the perspective of broad-sense heritability， the three test sites all showed heritability at a moderately low level， with E3 having the highest heritability and E2 having the lowest.

Multi-site conjoint analysis of variance， regression analysis and AMMI analysis of brix weight for sugarcane combinations

It can be seen from Table 4 that in the multi-site conjoint analysis of variance， the differences in brix weight among different combinations， environments， and interaction of combinations and environments reached a very significant level （P<0.01）. It can be seen from the variance component that the variance component of the environment was 8.1%， the variance component of the combination was 36.3%， and the variance component of the combination×environment interaction was 16.2%. It can be seen that the factor of combination×environment interaction had a great influence on the variation of combination phenotype， and it is particularly important to carry out regional breeding in different areas.

It can be seen from the analysis of variance of the regression model that the significance of the combination regression was P=0.054 1， which was relatively close to the significant level， and the sum of squares of combination regression accounted for 49.5% of the sum of squares of interaction. From the analysis of variance of the AMMI model， it can be seen that the significance of the first principal component axis （IPCA1） of genotype and environment reached an extremely significant level， and the sum of squares of IPCA1 accounted for 57.9% of the sum of squares of interaction.

Stability analysis of combination brix weight

The combinations with a positive brix weight effect were arranged according to the average phenotype value of brix weight and the effect value （the difference between the average brix weight of each combination and the average brix weight of all combinations） from high to low （Table 5）. The higher the combination ranked， the better its yielding ability was.

In regression analysis， a bi value greater than 1 indicates that the stability of the combination is lower than the average stability; and a bi value less than 1 indicates that the stability of the combination is higher than the average stability. In this study， the combinations 174， 643， 404， 575， 240， 972， 636， 144， YC95， 1470， 755， 409， 701， 832， YC37， 579 and 316 showed higher average brix weights and effect values and bi values less than 1， and had the advantages of strong yielding ability， high yield-increasing effect， and good stability. These combinations are suitable for planting at the three test sites.

In the AMMI model， the smaller the genotype stability parameter Dg， the more stable the genotype. It can be seen from the analysis results of the AMMI model that in this study， the combinations 643， 404， 575， 449， 972， 636， YC127， 144， YC95， 1470， 373， YC112， 755， 533， 838， 253， 409， 411， 701， 51， 570， YC123， 832， 1074， 291， 600， YC37 and 579 showed higher average brix weight and brix weight effect values， and the values of the genotype stability parameter Dg were lower. These combinations had high yielding ability and good yield stability.

It can be seen from the regression analysis and AMMI model analysis that the combinations 643， 404， 575， 972， 636， 144， YC95， 1470， 755， 409， 701， 832， YC37 and 579 had high yielding ability， good yield stability and wide adaptability.

Analysis of brix weight and selection rate of combination at the three test sites

The brix weight effects （the brix weight effect of combination here refers to the difference between the brix weight value of each combination and the average value of all the combinations at the test site） and the selection rates of combination at the three test sites were combined and comprehensively analyzed. It can be seen from Table 6 that there were 57 combinations with a positive weight effect at the Nanning site， among which the combinations 449， YC127， 796， YC44， 533， 570， YC123， 391， 546， 403， YC90 and 252 had a combination selection rate more than 10%， these combinations had a higher brix weight at the Nanning test site， and there were more selected good combination offspring. At the Chongzuo test site， there were a total of 62 combinations with a positive brix weight effect， among which the combinations 643， 212， YC61， 432， 903， YC95， YC44， 368， YC83， YC127， YC112， 701， 411， YC90 and YC123 had higher brix weights and higher selection rates of combination. At the Laibin site， a total of 59 combinations showed a positive brix weight effect， of which the combinations 643， 404， 449， 144， 403， and YC48 had higher brix weights and selection rates of combination.

Conclusions and Discussion

In this study， a multi-site joint analysis of variance was performed on the brix weights of combinations at the three different test sites， and the results showed that the differences in brix weight between the combinations reached a significant level. The phenotypic variation and genetic variation of the seedling brix weights at the Chongzuo and Laibin test sites were relatively large， while the phenotypic variation and genetic variation coefficients of the seedling brix weights at the Nanning test site were relatively small. The decrease of the population coefficient of variation increased the difficulty of selection.

There are many methods for analyzing the stability of crop genotypes， among which regression analysis and AMMI model are the more commonly used methods， but they both have their own advantages and disadvantages[11]. In this study， the regression analysis and AMMI model explained 49.5% and 57.9% of the sum of squares of interaction， respectively. Therefore， regression analysis and AMMI model had roughly the same effect on the combination stability evaluation in this study， but the effects were both slightly lower. In order to ensure the reliability of the combination stability evaluation in this study， we further integrated the two kinds of stability analysis results， and considered the combinations 643， 404， 575， 972， 636， 144， YC95， 1470， 755， 409， 701， 832， YC37 and 579 were high and stable-yielding and widely adaptable combinations， which can be adapted to the ecological environment of the three major production areas in Guangxi. They can be planted and used in Nanning， Chongzuo， and Laibin， and more new sugarcane varieties with excellent traits and wide adaptability are expected to be bred.

Breeding high and stable-yielding and high-sugar varieties is an important goal of sugarcane breeding[10]. The phenotype of crop trait is not a simple addition of genotype effects and environmental effects， but also involves the interaction between genotypes and environments[17]. Many traits of crops are affected by genotype and environment interaction， which is the root cause of genotype stability[18]. When promoting good combinations， the interaction between genotypes and environments should be fully utilized to select specific hybrid combinations for specific environments[19]. In this study， the combinations 449， YC127， 796， YC44， 533， 570， YC123， 391， 546， 403， YC90 and 252 showed higher brix weights and higher selection rates of combination in the Nanning test site， as well as high sugar yields， so they are more suitable for the environment at the Nanning test site and can be used in Nanning. The combinations 643， 212， YC61， 432， 903， YC95， YC44， 368， YC83， YC127， YC112， 701， 411， YC90 and YC123 showed higher brix weights and higher selection rates of combination at the Chongzuo test site， and are thus more suitable for the environment at the Chongzuo test site and can be used in Chongzuo. The combinations 643， 404， 449， 144， 403 and YC48 exhibited higher brix weights and higher selection rates of combination at the Laibin test site， and are thus more suitable for the environment of the Laibin test site and can be used in Laibin.

References

[1] STRINGER JK， COX MC， ATKIN FC， et al. Family selection improves the efficiency and effectiveness of selecting original seedlings and parents[J]. Sugar Tech， 2011， 13（1）： 36-41.

[2] ZHU JR， TIAO LA， DONG LH， et al. Heritability and combining ability of introduced sugarcane varieties and Yunrui innovative parents[J]. Journal of Hunan Agricultural University（Natural Sciences）， 2014， 40（6）： 571-577. （in Chinese）

[3] TANG SY， YANG RZ， WANG LW， et al. Genetic determination and selection of sugarcane families[J]. Journal of Hunan Agricultural University： Natural Sciences， 2016， 42（5）： 472-479. （in Chinese）

[4] LIU SM， WANG QN， FU C， et al. Family assessment of usual sugarcane parents and their cross combination[J]. Journal of Plant Genetic Resources， 2011， 12（2）： 234-240. （in Chinese）

[5] GAO SJ， CHEN RK， FU HY， et al. Genetic variation and selection of main economic traits in sugarcane （Saccharum spp.） seedling populations[J]. Chinese Journal of Tropical Crops， 2006， 27（1）： 49-53. （in Chinese）

[6] JING YF， BIAN X， LANG RB， et al. Fuzzy comprehensive evaluation of 32 sugercane cross combinations of Yunrui innovative parents[J]. Subtropical Agriculture Research， 2015， 11（4）： 217-221. （in Chinese）

[7] ZHAO PF. General combining ability of thirty-four sugarcane varieties/clones developed in Yunnan and economic traits in their progeny populations[D]. Beijing： Chinese Academy of Agricultural Sciences， 2011.

[8] LIU LX， ZHAO SL. Breeding for stability and adaptability of crop cultivars[J]. Shaanxi Journal of Agricultural Sciences， 1992， 38（5）： 45-47. （in Chinese）

[9] LIN YQ， ZHOU YF， CHEN ZJ. Comparison of nine stability parameters applied to the regional test of sugarcane varieties[J]. Journal of Fujian Agricultural College， 1990， 19（2）： 138-143. （in Chinese）

[10] XU LN， DENG ZH， CHEN RK， et al. Stability analysis on yield and quality characters of new sugarcane varieties[J]. Chinese Journal of Tropical Crops， 2006， 27（2）： 50-54. （in Chinese）

[11] LUO J， XU LP， QIU J， et al. Evaluation of sugarcane test environments and ecological zone division in China based on HA-GGE biplot[J]. Acta Agronomica Sinica， 2015， 41（2）： 214-227. （in Chinese）

[12] WANG ZT， SU WH， QUE YX， et al. Analysis of yield stability and test site representativeness of sugarcane trials using combined AMMI and HA-GGE biplot models[J]. Chinese Journal of Eco-Agriculture， 2016， 24（6）： 790-800. （in Chinese）

[13] DENG ZH， LIN YQ， CHEN RK， et al. Adaptability analysis of new promising sugarcane varieties[J]. Journal of Fujian Agricultural University （Natural Science Edition）， 2000， 29（2）： 145-150. （in Chinese）

[14] XU LN， DENG ZH， CHEN RK， et al. Analysis on heritability and combining ability of CL parents of sugarcane[J]. Journal of Plant Genetic Resources， 2006， 7（4）： 445-449. （in Chinese）

[15] LIU LF， MAO SX， HUANG YZ. Crops quantitative genetic[M]. Beijing： Agriculture Press， 1984. （in Chinese）

[16] TANG QY. DPS data processing system[M]. Beijing： Science Press， 2016. （in Chinese）

[17] WU YQ， PAN GT， RONG TZ. Study progress in crop stability[J]. Journal of Sichuan Agricultural University， 2005， 23（4）： 482-489. （in Chinese）

[18] GAUCH HG. Model selection and validation for yield trials with interaction[J]. Biometrics， 1988， 44（3）： 705-715.

[19] ZHANG XF， LI HG， YOU LX， et al. Variation and genetic stability of two-year-old Liriodendron seedling growth for 39 mating combinations[J]. Journal of Zhejiang Forestry College， 2011， 28（1）： 103-108. （in Chinese）

Editor： Yingzhi GUANG  Proofreader： Xinxiu ZHU