| Cotton (Gossypium spp.) is the most important fibre crop in the world. As a main by-product of cotton, cottonseed is an important source of edible oil and protein meals. Due to its important economic value, there was an increasing interest for cotton breeders in improving cottonseed quality. One of the most important aspects for cottonseed quality improvement is to understand its genetic mechanism and develop fast testing methods for seed quality. Thus, the aims of present study are (1)to investigate the feasibility of quality measuring method with near-infrared reflectance spectroscopy, using 49 upland cotton (Gossypium hirsutum L.) germplasms and 188 recombinant inbred lines (RILs), which were harvested from different cotton growing regions in different years; (2)to determine genetic effect, heritability, correlation, the exploitation of F2 heterosis of cottonseed quality traits through the genetic models of diploid plant seed quantitative traits and corresponding statistical approaches with two years data under two kinds of environment, using twelve transgenic upland cotton parents with a diallel mating design; (3)to construct an "immortalized F2" (IF2) population including 188 single crosses; and (4)to analyze the frequency of molecular marker loci in IF2 population, using a molecular map with 388 different markers. The main results were summarized as follows:(1) Protein and gossypol NIRS calibration equations of cottonseed were obtained with modified partial least squares (MPLS) through analyzing the scanning data with ISI software. The multiple correlation coefficients (RSQ) and statistic 1-variance ratio (1-VR) for the protein content NIRS calibration in cottonseed kernels were 0.93 and 0.93, respectively, and its standard error of calibration (SEC) and standard error of cross validation (SECV) were 0.62 and 0.64, respectively. The results indicated that NIRS could be used as a substitute for chemical methods in determination of protein content in cottonseed kernels. Also, the RSQ, SEC,1-VR and SECV for gossypol content determination of NIRS were 0.84,0.81,0.07 and 0.08, respectively. It showed a good fitting with chemical method although it was weaker than that of protein content NIRS.(2)The genetic effect, heritability and genetic correlation of cottonseed quality traits including seed index (SI), kernel index (KI), kernel percentage (KP), kernel/hull (K/H), protein content (PC), oil content (OC) and gossypol content (GC) in twelve parents and their diallel mating were analyzed through the genetic models of diploid seed quantitative traits. The results showed that①KP, PC and OC were mainly affected by genetic effect, and SI, KI, K/H and GC were mainly controlled by genetic×environmental interaction effect. In addition, all seed quality traits in this experiment were mainly controlled by maternal effects, followed by cytoplasmic effects. Four physical traits including SI, KI and K/H were dominantly controlled by environment effects. For cottonseed nutrient quality traits, PC and OC were mainly controlled by maternal additive effects. And the total narrow-sense heritabilities for PC and OC were high. GC was controlled by maternal dominance×environment interaction and cytoplasmic effects and having a high cytoplasm heritability. The results may supply some useful information for the improvement of seed quality by the genetic and breeding method.②The phenotypic correlation and genotypic correlation components for quality traits indicated that genetic main correlations and GE interaction correlations from different genetic systems including diploid embryo, cytoplasm and diploid maternal plant were mostly significant for the pairwise quality traits excepted for K/H. The relationship for SI and GC, KI and PC, KI and OC, KI and GC were influenced by the environment conditions. A negative correlation between PC and OC, PC and GC was observed. However, the correlation between OC and GC was positive. Therefore, some cottonseed quality traits could be indirectly improved by the selection based on other traits effectively.③Heterosis including the general heterosis and the GE interaction heterosis could be detected in the cottonseed quality traits, but the performance and main effect varied with the traits. The most important general heterosis of SI, KI, K/H, PC, OC and GC were maternal heterosis, while KP was mainly affected by cytoplasmic heterosis. In addition, SI, KI and KP were also controlled by interaction heterosis, but the interaction heterosis was not important for K/H, PC, OC and GC.④The predicted values of genetic effects for 12 parents were also analyzed in this study. It showed that testing parents could play different role in cottonseed quality improvement breeding. For example, SGK 3 and H 556 have the effect to increase SI and KI in these offsprings; SGK 3, ZMS 44, ZHE 905, LU 2015 and H 556 could increase KP significantly; 126XI 9, AY-4, ZHE 1793, LU 2015 and H 556 could increase K/H in the following progenies; AY-4, ZHE 905 and Z 1138 were better than other for increasing PC; 126XI 9, ZHE 1793, R 808, H 556 and J 220 were important in the improvement of OC in seeds; and ZMS 44, AY-4, ZHE 905,R 808 and LU 2015 were the good germplasm to decrease GC in cottonseeds.(3) An "immortalized F2" (IF2) population including 188 single crosses was constructed by using a set of recombinant inbred line (RIL) derived from the hybrid of HS46×MARCABUCAG8US-1-88. With this IF2 population, the frequency of molecular marker loci, QTL mapping and effect of single loci for seed quality traits were analyzed using a molecular map with 388 markers by software of QTL Network. Seventeen main effects and fifteen epistasis in QTLs were identified for cottonseed quality traits including SI, KI, KP, K/H, PC, OC and GC. Two of main QTL and epistasis were found closely correlated with SI, and the total variation explained was 22.27%; And two main QTL and four epistasis attribute to KI, and explained 23.04% variation; and four main QTL and 2 epistasis could explain 19.70% total variation of KP. Further, present study located the main QTL of K/H in the 15 chromosome. Also, the number of closely related main QTL and epistasis for protein, oil, gossypol content are 3-1,3-3, and 4-2, respectively. And the variation explained are 2.42%～23.04%. These QTLs detected for seed quality traits in cotton are expected to be useful in the breeding program for cottonseed nutrient quality.
|
PDF Full Text Request