Morphological Vaiation And QTL Mapping Of Morphological Quantitative Traits Duringadaptive Evolution In Nine-spined Sticklebacks

Explaining the genetic basis of phenotypic diversity in natural populations has always been one of the basic goals of evolutionary biology. Morphological divergence has been reported between marine and freshwater populations of nine-spined sticklebacks, and similar phenotypes have been independently and repeatedly evolved in several nine-spined stickleback freshwater habitats. Hence, nine-spined sticklebacks become anew model fish to gain insights into the genetic basis of phenotypic variation in species. One marine(HEL)and three freshwater populations(RYT, PYO and BYN) of nine-spined sticklebacks in Fennoscandia were selected in this study. Using RAD sequencing technology and the combination of coarse-mapping and fine-mapping, QTL analysis of body morphological variation was performed to study the genetic mechanisms of morphological divergence between freshwater and marine populations in nine-spined sticklebacks(Chap. 2); besides, in order to study the genetic pathways of similar morphological divergence in different freshwater populations, RAD seq and two steps QTL mapping methods were usedto address the extent to whether the same genetic process was responsible for the variance of shape(Chap. 3), linear measurement traits(Chap. 4), and pelvic structure(Chap. 5). Our study provided basic genetic searching and experimental supports for the understanding of morphological divergence under the same or different natural environmental selections in species.The main results of this study are listed as follows:(1) Morphological divergence and QTL mapping between marine and freshwater populations in nine-spined sticklebacksDifference were existed in the shape and ten candidate linear traits between HEL and RYT nine-spined stickleback wild populations. The lateral plate number of HEL population was higher than that of RYT population. Principal component analysis of shape variance in HEL x RYT F2 population showed that the first three PCs accounted for 61.3% of the total shape variation, capturing primarily variation in body length, depth and caudal peduncle length. 18,691 unique SNP markers distributed across 21 LGs were obtained from RAD sequencing. Twelve QTL were identified to be associated with the variation of shape coordinates, 4 QTL for the variation of caudal peduncle length, lower jaw length, body depth and snout length, and 4 QTL for the changes of lateral plate number. The percentage of variance explained(PVE) value by the QTL varied from 6.3% to 15.3%. Two QTL could be responsible for the divergence of several traits at the same time, which means these QTL were pleiotropic. No major effect QTL(percentage of variance explained >60%) was found in this cross, indicating that body shape change, linear traits variance and different lateral plate numbers between HEL and RYT populations might have polygenic basis.This study provided experimental evidence for the differentiation of caudal peduncle length, lower jaw length, body depth, snout length, and lateral plate number from genotypic level, and also succecessfully developed candate genes for most QTL.(2) Genetic basis of shape divergence among freshwater populations in nine-spined sticklebacks13,060 unique SNP markers distributed across 21 LGs were obtained from RAD sequencing in HEL and PYO cross. Using QTL mapping analysis, 50 significant QTL were mapped with shape coordinates variance, with the PVE value ranging from 5.1% to 14.1%. Five QTL were found to be linked with two or more coordinates, suggesting that these QTL were pleiotropic. No major effect QTL was found for the divergence of body shape between HEL and PYO population, indicating that body shape change between HEL and PYO populations might has polygenic basis. This body variancehad been proved to not be associated with the allometry of body size.Despite polygenic basis was both happened in RYT and PYO populations, but parallel coordinates were found to be associated with different QTL. Hence, we supposed that the genetic basis of body shape differentiation was different between these two freshwater nine-spined stickleback populations, and the evolution of similar adaptive shape change between these two populations is convergent evolution.(3) Genetic basis of linear measurement traits divergence among freshwater populations in nine-spined sticklebacksIn mapping analysis of HEL and PYO cross, 38 significant QTL were mapped with linear traits variance, with the PVE value ranging from 3.0% to 22.9%. Five QTL were found to be associated with two or more traits, suggesting that these QTL were pleiotropic.9,832 unique SNP markers distributed across 21 LGs were obtained from RAD sequencing in HEL and BYN cross. 30 significant QTL were mapped with linear traits variance, with the PVE value ranging from 1.7% to 7.5%. Two pleiotropic QTL were identified to be associated with two or more traits. No major effect QTL for linear traits variation were found in this cross.After comparing the QTL locations for linear traits variation in the three freshwater populations(RYT, PYO, and BYN), most QTL for the same linear traits differentiation were mapped on different linkage groups, indicating that the phenomenon of similar linear traits in these three freshwater populations was due to convergent evolution.(4) Pelvic reduction of freshwater populations in nine-spined sticklebacksOne major effect QTL was found to be associated with pelvic spine and girdle reduction in HEL and RYT cross, with the mean PVE value of 70.2%. This major QTL was located on the end of LG7, and Pitx1 gene was found to be the candidate gene for pelvic reduction in this cross. It seems that one major gene(Pitx1 gene), not multiple gene, conducted the pelvic reduction in RYT population.In HEL and BYN cross, reduction of pelvic spine, girdle, branch length, and branch width were mapped to 6, 6, 4, and 2 QTL markers. None of these QTL located on LG7. Furthermore, four of these markers were large effect QTL(percentage of variance explained >10%), and no major effect QTL was found, which indicating that Pitx1 gene was not associated with the pelvic reduction in BYN cross, and the pelvic reduction in BYN cross might have polygenic basis.Complementation cross of RYT and BYN population showed that several F1 individuals developed pelvic spine though their parents has no spines at all. This results referred that different allele were involved in pelvic reduction in RYT and BYN population, and also provided another evidence for convergent evolution of pelvic reduction among freshwater populations in nine-spined stickleback.