Aim  Eurycorymbus cavaleriei (Lévl.) Rehd. et Hand.‐Mazz. (Sapindaceae) is a Tertiary relict tree endemic to subtropical China. This area is a centre for speciation and evolution within the East Asia biome and one of its most important refugial locations. In this study we aim to elucidate the phylogeographical patterning in E. cavaleriei, in order to identify the locations of the species’ main refugia and the predominant patterns of migration that have led to the contemporary spatial genetic structure of chloroplast variation. Location  Subtropical China. Methods  We sampled 18 populations of E. cavaleriei throughout its geographical range. Chloroplast DNA (cpDNA) sequence data from two non‐coding regions ((trnS/trnG and pl20/5′‐rps12) were obtained from 170 individuals for phylogeographical analyses. Relationships among cpDNA haplotypes were determined using median‐joining networks. Genetic structure was examined by spatial analysis of molecular variance (SAMOVA). Population differentiation was estimated by GST and NST statistics. Results  Ten distinct haplotypes were identified. The level of differentiation among populations was relatively high (GST = 0.817), and NST was significantly higher than GST (P < 0.05), indicating that strong phylogeographical structure is exhibited by this species. The SAMOVA revealed five diverging groups of related haplotypes, which coincide with major landscape features in this region. Main conclusions  The high differentiation among populations of E. cavaleriei may be a combined effect of historical and contemporary processes, such as the low effective population size for the chloroplast genome of a dioecious species, long‐term range fragmentation and limited seed dispersal for the species. Clear‐cut geographical distributions of ancestral haplotypes of the species suggest multiple potential refugia across subtropical China. The identified refugial regions have long been recognized as centres of plant diversity and endemism for China and have also been suggested as glacial refugia for many other plant species. The combination of these factors means that these locations should be considered as the highest priority for inclusion in conservation policies and sustainable forest management strategies for subtropical China.

Lycopene content is an important factor for determining watermelon fruit quality. However, the low DNA polymorphism among cultivated watermelon (Citrullus lanatus) has hindered the ability to establish high quality genetic maps and study the quantitative trait loci (QTL) controlling the lycopene content trait. In this study, we successfully constructed a genetic map of watermelon to determine lycopene content and other horticultural fruit traits using a F 2 population developed from a cross between the two lines of watermelon LSW-177 and Cream of Saskatchewan. The genetic map contained 16 linkage groups covering a total length of 2,039.5 cM, which included 37 SSRs (Simple Sequence Repeat) and 107 CAPSs (Cleaved Amplified Polymorphic Sequences), with all of the CAPS markers developed from high-throughput re-sequencing of data from this study. Three CAPS markers (WII04E07-33,WII04E07-37,WII04E07-40) caused the F 2 population to perfectly co-segregate for each F 2 population plants. We also obtained 12 QTLs for all of the traits measured. Only one QTL (LCYB4.1) was detected with a high value of trait variation (83.50 %) that related to lycopene content and mapped on Chromosome 4 between CAPS markers WII04E07-33 and WII04E07-40, which could nearly account for all of the differences in lycopene content between the two parental strains. In this study, we highlighted 2,458 CAPS loci that were suitable for primer design with a polymorphism of 48.9 %, which is approximately a 12-fold increase from previous studies. The present map and QTLs will facilitate future studies on determining lycopene content related genes and cloning watermelon genes, while also providing for useful markers for breeding for lycopene content.

Domestication and improvement are two important stages in crop evolution. Melon (Cucumis melo L.) is an important vegetable crop with wide phenotypic diversity in many horticultural traits, especially fruit size, flesh thickness and aroma, which are likely the results of long-term extensive selection during its evolution. However, selective signals in domestication and improvement stages for these remarkable variations remain unclear. We resequenced 297 wild, landrace and improved melon accessions and obtained 2 045 412 high-quality SNPs. Population structure and genetic diversity analyses revealed independent and two-step selections in two subspecies of melon: ssp. melo and ssp. agrestis during melon breeding. We detected 233 (~18.35 Mbp) and 159 (~17.71 Mbp) novel potential selective signals during the improvement stage in ssp. agrestis and spp. melo, respectively. Two alcohol acyltransferase genes (CmAATs) unique to the melon genome compared with other cucurbit crops may have undergone stronger selection in ssp. agrestis for the characteristic aroma as compared with other cucurbits. Genomewide association analysis identified eight fruit size and seven flesh thickness signals overlapping with selective sweeps. Compared with thinskinned ssp. agrestis, thickskinned ssp. melo has undergone a stronger selection for thicker flesh. In most melon accessions, CmCLV3 has pleiotropic effects on carpel number and fruit shape. Findings from this study provide novel insights into melon crop evolution, and new tools to advance melon breeding. classified into two subspecies (C. melo ssp. melo and ssp. agrestis) based on the presence of ovary hairs (Jeffrey, 1980; Kirkbride, 1993). The ssp. agrestis is distributed mainly in Africa, Asia and Australia with momordica, acidulus, conomom, makuwa and chinensis groups, in Africa with the tibish group and in Central America with the chito group. The ssp. melo including the cantalupensis, inodorus, chandalak, ameri, flexuosus, chate or dudaim groups is distributed mainly in India, central and western

No abstract