The past 10 years has seen a steep increase in production of seedless watermelon (Citrullus lanatus) in the United States. Seedless fruit is produced on triploid plants that require pollination from diploid pollenizers for fruit set. Synchronization of the staminate flowers on the pollenizers with the appearance of pistillate flowers on the triploids is a fundamental requirement for this production system. Previous research suggested that pistillate flowers reach peak production early in the season, but data are only available for a small number of triploid cultivars. We compared the flowering patterns of 29 triploid cultivars and 20 pollenizers, including 10 harvested pollenizers, during the first 6 weeks after transplanting over 2 years. The average number of days from transplanting (DAT) to the first staminate flower was between 5.3 days and 19.1 days in 2012 and 9.7 days and 24.4 days in 2013 for the pollenizers and between 18.7 days and 27.6 days and 22.1 days and 32.7 days for the pistillate triploid flowers in the 2 years, respectively. K-means clustering of the weekly percentage of plants with staminate and pistillate flowers for the different cultivars shows that different triploid and pollenizer cultivars have different flowering patterns and that some combinations have better synchronized flowering than others. Growers should take particular care when choosing pollenizers for early-flowering triploid cultivars. Harvested pollenizers are better suited to late-flowering triploids and growers should choose triploid and pollenizer cultivar combinations with flowering patterns that best satisfy their specific production goals.
Cecilia E. McGregor and Vickie Waters
Cecilia E. McGregor and Vickie Waters
Crop wild relatives (CWRs) are important sources of variation for domesticated crops like watermelon (Citrullus lanatus) where cultivated varieties have a very narrow genetic base. The use of CWRs in plant breeding can be hampered by low fertility, chromosomal rearrangements, marker distortion, and linkage drag in the progeny. Pollen viability can be a quick and easy way to estimate male fertility, which can be a cause of marker distortion and an indicator of chromosomal rearrangements. Pollen viability was determined for F1 hybrids between cultivars and resistant citron and egusi types and the data were used to determine whether the parental cultivars/lines used or the directionality of the cross play a role in pollen viability. F1 hybrids between cultivars and the egusi type showed no reduction in pollen viability compared with parental lines, whereas pollen viability of hybrids with citron types varied between 61.8% and 91.7%. Significant main effects were observed for the cultivar and donor lines used, but the directionality of the cross did not affect pollen viability. F1 hybrids with ‘Crimson Sweet’ as the cultivar parent had significantly higher pollen viability than those with ‘Sugar Baby’ or ‘Charleston Gray’. Our results indicate that the directionality of the crosses between watermelon cultivars and infraspecific CWRs does not affect pollen viability but that the specific cultivars and donor lines used can have a significant effect. The high pollen viability of cultivar–egusi hybrids is supported by previous genetic data and strongly suggests that it should be easier to introgress traits from egusi types than citron types.
Geoffrey Meru and Cecilia E. McGregor
Fusarium wilt of watermelon (Citrullus lanatus), caused by Fusarium oxysporum f. sp. niveum (FON), is a devastating soil-borne disease limiting watermelon production across the world. Although many watermelon cultivars have been bred for resistance to FON races 0 and 1, the only released cultivars that are resistant to FON 2 are nonharvested pollenizers. The lack of FON 2–resistant edible cultivars is thought to be associated with linkage drag and/or preferential inheritance patterns observed when crossing the resistant, wild source (Citrullus amarus), with edible watermelon. A potential way to overcome these obstacles is to use a resistant C. lanatus as the source of resistance and to develop molecular markers to increase the efficiency of selection. Here we describe the identification of a quantitative trait locus (QTL) associated with FON 2 resistance in watermelon. The genotyping by sequencing (GBS) platform was used to generate single nucleotide polymorphisms (SNPs) in an F2 population (n = 178) developed from a cross between UGA147 (resistant) and ‘Charleston Gray’ (susceptible). Five hundred and one SNPs were placed on the watermelon physical map and used in the mapping of QTL. F3 lines were phenotyped for resistance to FON 2 in the greenhouse. An intermediate QTL associated with resistance to FON 2 was identified on chromosome 11 (Qfon11). This QTL is a potential target for marker-assisted selection (MAS) for FON 2 resistance in watermelon.
Cecilia E. McGregor and Don R. LaBonte
`White Jewel' is a yellow-and-orange fleshed spontaneous mutant of the orange-flesh sweetpotato [Ipomoea batatas (L.) Lam.] cultivar Jewel. Mutations in storage root flesh color, and other traits are common in sweetpotato. The orange flesh color of sweetpotato is due to β-carotene stored in chromoplasts of root cells. β-carotene is important because of its role in human health. In an effort to elucidate biosynthesis and storage of β-carotene in sweetpotato roots, microarray analysis was used to investigate genes differentially expressed between `White Jewel' and `Jewel' storage roots. β-carotene content calculated from a* color values of `Jewel' and `White Jewel' were 20.66 mg/100 g fresh weight (FW) and 1.68 mg/100 g FW, respectively. Isopentenyl diphosphate isomerase (IPI) was down-regulated in `White Jewel', but farnesyl-diphosphate synthase (FPPS), geranylgeranyl diphosphate synthase (GGPS), and lycopene β-cyclase (LCY-b) were not differentially expressed. Several genes associated with chloroplasts were differentially expressed, indicating probable differences in chromoplast development of `White Jewel' and `Jewel'. Sucrose Synthase was down-regulated in `White Jewel' and fructose and glucose levels in `White Jewel' were lower than in `Jewel' while sucrose levels were higher in `White Jewel'. No differences were observed between dry weight or alcohol insoluble solids of the two cultivars. This study represents the first effort to elucidate β-carotene synthesis and storage in sweetpotato through large-scale gene expression analysis.
Cecilia E. McGregor, Vickie Waters, Tripti Vashisth and Hussein Abdel-Haleem
U.S. watermelon (Citrullus lanatus) production is worth ≈$0.5 billion annually to growers and nearly all of them are dependent on reliable synchronized flowering time of triploid cultivars and diploid pollenizers in their production fields. One aspect of this synchronization is time to flowering, the change from the vegetative to reproductive phase of a plant. Flowering time has emerged as one of the key traits in horticultural and agronomic crops to breed for escape from biotic and abiotic stresses. However, very little is known about the control of flowering time in watermelon. The number of genes involved, mode of inheritance, heritability, and the possible candidate genes are all unknown. In this study, quantitative trait loci (QTL) associated with days to first male flower (DMF), days to first female flower (DFF), and the female-male flower interval (FMI) were identified in a ‘Klondike Black Seeded’ × ‘New Hampshire Midget’ recombinant inbred line population over 2 years. Heritability for DMF, DFF, and FMI were 0.43, 0.23, and 0.10, respectively. Control of flowering time was oligogenic with a major, stable, colocalized QTL on chromosome 3 responsible for ≈50% of the phenotypic variation observed for DMF and DFF. This region of the draft genome sequence contains 172 genes, including homologs of the flowering locus T (Cla009504) and tempranillo 1 (Cla000855) genes associated with flowering time in other species. Cla009504 and Cla000855 represent excellent candidate genes toward the development of a functional marker for marker-assisted selection of flowering time in watermelon. In addition to the major QTL on chromosome 3, two other QTL were identified for DMF (chromosomes 2 and 3) and DFF (chromosomes 3 and 11) and one for FMI on chromosome 2. Understanding the genes involved in this trait and the ability to select efficiently for flowering time phenotypes is expected to accelerate the development of new watermelon cultivars in changing environmental conditions.
Cecilia E. McGregor, Douglas W. Miano, Don R. LaBonte, Mary Hoy, Chris A. Clark and Guilherme J.M. Rosa
Sweet potato virus disease (SPVD) is one of the most devastating diseases affecting sweetpotato (Ipomoea batatas), an important food crop in developing countries. SPVD develops when sweetpotato plants are dually infected with sweet potato feathery mottle virus (SPFMV) and sweet potato chlorotic stunt virus (SPCSV). To better understand the synergistic interaction between these viruses, global gene expression was previously studied in the susceptible cultivar Beauregard. In the current study, global gene expression between SPVD-affected plants and virus-tested control plants (VT) were compared in ‘Beauregard’ (Bx) and resistant ‘NASPOT 1’ (Nas) sweetpotato cultivars at 5, 9, 13, and 17 days post inoculation (DPI). Titer levels of SPFMV and SPCSV were significantly lower in inoculated resistant plants (Nas_SPVD) than in susceptible plants (Bx_SPVD) at most of the time points. Chloroplast genes and cell expansion-related genes (including xyloglucan endotransglucosylase/hydrolases) were suppressed in Bx_SPVD, while stress-related genes were induced. This trend was not observed in resistant NAS_SPVD. Genes related to protein synthesis (e.g., ribosomal proteins and elongation factor genes) were induced in resistant NAS_SPVD at 5 DPI before returning to levels comparable with NAS_VT plants. At this time (5 DPI), individual viruses could not be detected in NAS_SPVD samples, and no symptoms were observed. Induction of protein synthesis-related genes is common in susceptible plants after virus infection and is generally in proportion to virus accumulation. Our results show that induction of protein synthesis genes also occurs early in the infection process in resistant plants, while virus titers were below the level of detection, suggesting that virus accumulation is not required for induction.