Click on author name to view affiliation information
Powdery mildew (Podosphaera xanthii) can cause severe damage to cucurbit crops grown in open fields and greenhouses. In recent years, there has been an increased interest in the United States in grafting watermelon plants onto various cucurbit rootstocks. Bottle gourd plants (Lagenaria siceraria) are being used throughout the world as rootstocks for grafting watermelon. Although gourd plants are beneficial, they may still be susceptible hosts to various soilborne and foliar diseases. Bottle gourd plant introductions (PI) resistant to diseases and pests can be a valuable source of germplasm in rootstock breeding programs. We evaluated 234 U.S. PIs of L. siceraria for tolerance to powdery mildew in two greenhouse tests. Young seedlings were inoculated by dusting powdery mildew spores of melon race 1 on the cotyledons. Plants were rated 2 weeks after inoculation using a 1 to 9 scale of increasing disease severity. Although none of the L. siceraria PIs were immune to powdery mildew, several PIs had significantly lower levels of powdery mildew severity compared with susceptible watermelon cultivar Mickey Lee. The experiment was repeated with 26 select PIs on whole seedlings and cotyledon disks. Significant variability in the level of resistance to powdery mildew on plants within PI was observed. Moderate resistance in several PIs to powdery mildew was confirmed. PI 271353 had consistently lower ratings in the various tests and can be considered the most resistant to P. xanthii race 1 among the L. siceraria accessions evaluated in this study. A few other PIs with moderate resistance to powdery mildew included PI 271357, PI 381840, and PI 273663. These results suggest that novel sources of resistance could be developed by careful selection and screening of several of the PIs with moderate resistance described in our study.
Bottle gourd [Lagenaria siceraria (Mol.) Standl.], which belongs to the Cucurbitaceae family, is considered to be one of the earliest species of plants to be domesticated by humans. Bottle gourds are regularly grown and consumed in various parts of the world, particularly Asia. Fruits of bottle gourds are harvested young and used as a vegetable. The mature, dried fruits have been used throughout the world for making storage jars and utensils and hence the name bottle gourd (Decker-Walters et al., 2001). Lagenaria siceraria is also known as ‘Calabash’ or the white-flowered gourd (Jeffrey, 1967). In the United States, Lagenaria gourds are currently being tested as rootstocks for commercial watermelon [Citrullus lanatus var. lanatus (Thunb.) Matsum. & Nakai] production (Hassell, 2007; King and Davis, 2006; Taylor et al., 2006). In many parts of the world, watermelon are grafted on diverse rootstocks (Lee and Oda, 2003; Oda, 2002). In Japan, most of the commercially cultivated watermelon is grafted on Lagenaria rootstocks (Oda, 1993, 2002). In Israel, 60% to 70% of the watermelon produced are grafted (Cohen et al., 2007; Koren and Edelstein, 2004). Grafting watermelon onto gourd rootstock for commercial production first began in Japan and Korea in the late 1920s (Lee, 1994). Theoretically, grafting of a susceptible scion onto a resistant rootstock can quickly provide a resistant plant particularly against soilborne diseases, thus eliminating the need for a prolonged breeding program (Cohen et al., 2007). One of the primary benefits of using Lagenaria as a rootstock is to manage soilborne diseases, particularly fusarium wilt of watermelon. Fusarium wilt, caused by Fusarium oxysporum Schlechtend.:Fr. f. sp. niveum (E.F. Sm.)W.C. Snyder & H.N. Hans., is the most commonly reported disease worldwide on watermelon and it can be successfully managed with grafted transplants (Cohen et al., 2007; Lopez-Galarza et al., 2004; Miguel et al., 2004; Oda, 2002). Interest in grafting has increased in the United States as an alternative pest prevention strategy to the use of methyl bromide fumigation. Grafting also has other advantages such as drought tolerance, tolerance to a high water table, ability to thrive in a wide range of soils, tolerance to low temperatures, and increased plant vigor and yield (Cohen et al., 2007; Oda, 2002; Yetisir et al., 2006). In addition, a recent study has shown that grafting can increase lycopene content in seedless watermelon (Perkins-Veazie et al., 2007). Although gourd plants are beneficial as rootstocks, they may still be susceptible hosts to various soilborne and foliar diseases. Bottle gourd PIs resistant to certain diseases and pests could be a valuable resource for rootstock breeding programs. Although the process of grafting for managing diseases appears to be straightforward, many factors will play a role in selecting suitable rootstocks that will ultimately be successful, including, but not limited to, the adaptation of grafted plants to the local environment, resistance to prevailing diseases, and the appropriate rootstock–scion combination (Cohen et al., 2007; Yetisir and Sari, 2003).
Powdery mildew caused by Podosphaera xanthii (Castagne) Braun & Shishkoff (syn. Sphaerotheca fuliginea auct. p.p.) can be a severely limiting factor in cucurbit production in open fields and greenhouses. Powdery mildew can result in reduced vigor of the seedlings and, in some instances, death of the seedlings. Powdery mildew on cucurbits also is caused by Golovinomyces cichoracearum (D.C.) V.P. Heluta (formerly Erysiphe cichoracearum D.C.); however, in the United States, P. xanthii is considered to be the more prevalent powdery mildew pathogen (McGrath and Thomas, 1996; Thomas et al., 1984). Seven races of P. xanthii have been identified using melon (Cucumis melo L.) differentials (Cohen et al., 2004; Pitrat et al., 1998; Thomas et al., 1984). Recently, variability within races 1 and 2 in P. xanthii populations was described using 32 melon cultigens with the potential for the existence of 28 races (McCreight, 2006). However, the actual significance of these races on commercial melon cultivars is not clear. Physiological races of this pathogen have not been classified for other cucurbits because of the lack of differentials and fully resistant germplasm (Cohen et al., 2004). However, it would not be unreasonable to expect new races of P. xanthii to be described based on other cucurbit differentials that may be developed in the future. Until true sets of differentials for other cucurbits, including watermelon, are developed, race designation using the melon differentials can be considered the most appropriate because it is the most studied to date. Powdery mildew has been reported on various cucurbits, including Lagenaria spp. (Hammett, 1977; McGrath and Thomas, 1996). It also has been shown that a powdery mildew isolate from one cucurbit can infect all other tested cucurbit spp. (Hammett, 1977). A detailed description of the cucurbit powdery mildew pathogen and race classification system has been provided by Cohen et al. (2004).
We have observed severe powdery mildew on young bottle gourd plants in the greenhouse that were being evaluated for resistance to soilborne pathogens. Similarly, several commercial Lagenaria and Cucurbita rootstocks being evaluated in other trials also were infected with powdery mildew to an extent necessitating the use of fungicides. For a rootstock material to be useful in a grafting program, the hypocotyls and cotyledons of the rootstock must be healthy and free of disease to provide the initial impetus for the rapid growth of the scion. There are no known reports of bottle gourd PIs being evaluated for resistance to powdery mildew or bottle gourd germplasm with resistance to powdery mildew. With a thought to these greenhouse observations and rootstock health requirements, the objective of our present study was to evaluate the entire L. siceraria U.S. PI collection for resistance to powdery mildew caused by P. xanthii race 1. The results of our study will be useful for commercial breeders involved in developing rootstocks for grafting cucurbits. It should also be useful to breeders in countries where bottle gourds are bred for human consumption.
Powdery mildew melon race 1 strain was maintained on susceptible melon cultivar Iran H and on susceptible squash cultivar ESPN in a growth room. The growth room was maintained at 21 °C and a 12-h photoperiod. The race of the powdery mildew pathogen was determined by inoculating the various differentials described by McGrath and Thomas (1996). The melon differentials used were: ‘Iran H’, Vedrantais, Ananas, PMR-45, PMR-5, WMR-29, Edisto 47, MR-1, and PI 414723. The race of the pathogen also was monitored throughout the studies in the greenhouse using these melon differentials. The pathogen was further confirmed as P. xanthii based on sequence similarity of the Internal Transcribed Spacer (ITS) region (Hirose et al., 2005; Takamatsu and Kano, 2001) to previously described sequences in the National Center for Biotechnology Information database and amplification of ITS regions using specific polymerase chain reaction primers (Takamatsu and Kano, 2001).
Seedlings of L. siceraria PIs were grown in 50-cell jiffy trays filled with Metro Mix (Sun Gro Horticulture, Seba Beach, Alberta, Canada). Five seeds of each PI were planted. This preliminary trial did not have replications; however, the entire test was repeated two times. Two days before inoculation, air was blown over leaves of ‘Iran H’ or squash plants to clear the old conidia and allow new conidia to form. Three weeks after seeding, seedlings were dusted with powdery mildew conidia as described by Thomas et al. (2005) and others (Davis et al., 2006b, 2007). Seedlings were maintained in the greenhouse where conditions were optimal for development of powdery mildew during the winter months. Normal day/night conditions and temperatures of ≈20 °C were prevalent in the greenhouses during the winter months in Charleston, SC. Two weeks after inoculation, seedlings were rated on a 1 to 9 severity scale described by Thomas et al. (2005). The 1 to 9 scale was based on increasing disease severity where 1 = no evidence of infection, 2 = trace infection of cotyledons only, 3 = low infection of cotyledons only, 4 = moderate infection of cotyledons only and very slight infection of first true leaf, 5 = severe infection of cotyledons and slight infection of first true leaf and hypocotyls, 6 = severe infection of cotyledons, moderate leaf infection and sporulation on first true leaf and hypocotyls, 7 = severe infection with abundant sporulation on cotyledons, first true leaf and hypocotyls, 8 = severe infection of all seedling with abundant sporulation on cotyledons and some necrosis of cotyledons and first true leaf, and 9 = plant dead as a result of powdery mildew. The 1 to 9 rating scale was used for evaluating whole seedlings of all the available L. siceraria PIs as this is the rating scale used by the Germplasm Resources Information Network (GRIN) online database to report disease reactions. Seedlings of watermelon cultivar Mickey Lee were used as the susceptible control. The experiment was conducted two times. Data from each experiment was considered as a replication for analysis. Weighted averages were calculated for each of the entries as done previously for watermelon by Thomas et al. (2005). Based on the results of the initial screening, 26 PIs were selected and evaluated in a replicated test. There were four replications for each selected PI with four plants per replication. The experiment was set up as a randomized complete block design. Seeds of each selected PI were planted in 2.5-inch plastic pots filled with Metro Mix. After 3 weeks, the plants were inoculated as described previously and maintained in a growth room. The growth room was maintained at 21 °C and a 12-h photoperiod. Powdery mildew disease ratings were based on the 1 to 9 severity scale described previously. The data were analyzed with the Kruskal-Wallis test using the PROCNPAR1WAY procedure of SAS (version 8.2; SAS Institute, Cary, NC) and the entries ranked using Wilcoxon rank scores.
Seedlings of the 26 selected L. siceraria PIs that were being grown for the previously mentioned study were used for the cotyledon disk experiment. There were four replications for each selected PI with four plants per replication. The experiment was set up as a randomized complete block design. Three weeks after seeding, one circular disk of 2.5-cm diameter was cut out of one of the cotyledons of each plant for each PI and placed in sterile petri dish lined with one moist blotter paper (Anchor Paper Company, St. Paul, MN). Thus, four cotyledon disks from each replication for each PI were placed in each petri dish. The cotyledon disks were inoculated by brushing dry powdery mildew conidia in the center of the disk on the adaxial surface. Approximately 1205 conidia (range, 580 to 1850) were brushed onto each cotyledon disk. The petri plates were then placed in an incubator maintained at 18 °C. Ten days after inoculation, the cotyledon disks were rated using a 1 to 5 scale developed by Ishii et al. (2001), where 1 = 0% to 5%, 2 = 6% to 25%, 3 = 26% to 50%, 4 = 51% to 75%, and 5 = greater than 75% disk surface covered with powdery mildew. A second cotyledon disk experiment was performed, but as a result of limited seed availability, and in some cases poor seed germination, we tested a mean of five single plant replications for 18 PIs. One circular disk of 2.5-cm diameter was cut out of one of the cotyledons of each plant for each PI and placed in sterile petri dish lined with one moist blotter paper. Data were recorded on the 1 to 5 scale as described previously. For statistical analysis, the rating data were first converted to percentages based on the midpercentage point for each rating, then arcsine-transformed and analyzed using PROC GLM procedures of SAS. Mean separation was performed using Fisher's protected least significant difference.
In these studies, we used the prevailing powdery mildew pathogen in Charleston, SC. Although it is well established that powdery mildew of cucurbits can be caused by two different pathogens (Cohen et al., 2004; McGrath and Thomas, 1996; Thomas et al., 1984), we have not observed the presence of G. cichoracearum (formerly E. cichoracearum) in recent years in South Carolina. We confirmed that the prevailing powdery mildew pathogen in Charleston, SC, was P. xanthii using specific polymerase chain reaction primers (PM5 forward and ITS4 or P3 as reverse primers) for amplifying the ITS region as described previously by Takamatsu and Kano (2001). We further sequenced the fragment for additional confirmation of the pathogen (data not presented). The race of P. xanthii was identified as melon race 1 based on the melon differentials (McGrath and Thomas, 1996; Thomas et al., 1984). This strain was found to cause disease on the following melon differentials: ‘Iran H’, Vedrantais, and Ananas; however, it did not cause disease on melon differentials PMR-45, PMR-5, WMR-29, Edisto 47, MR-1, or PI 414723 confirming that it was race 1. The reactions of these differentials have been considered to be consistent worldwide (McCreight, 2006). The melon race 1 in our study could also be designated as race 1M based on suggestions by Davis et al. (2007). Recently Davis et al. (2007) described race 1W from watermelon and suggested that this race did not infect melons. However, we have noticed in our greenhouse experiments that powdery mildew conidia obtained from watermelon plants can infect melon and vice versa. Similarly, in a study of powdery mildew isolates from seven states in the United States, it was observed that over 50% of the isolates from muskmelon grew on watermelon and some isolates from watermelon grew on melon (Shishkoff and McGrath, 2001). The description of race 1W is relatively new (Davis et al., 2007) and based on a strain from Oklahoma; however, it remains to be determined if such strains are present in other areas of the United States. It is possible that there may be some variants within the race 1 in South Carolina; however, the reactions of the race 1 in our greenhouse in South Carolina were exactly the same on the nine common sets of differentials as those observed by McCreight (2006). Using additional sets of differentials, McCreight (2006) reported eight variants of race one; however, the actual significance of these variants on other cucurbits still needs to be determined. Until true sets of differentials for other cucurbits, including watermelon, are developed (hopefully in the near future), race designation using the melon differentials can be considered the most appropriate because it is the most studied and consistent worldwide. Hence, we prefer to call the race prevailing in our greenhouse race 1.
None of the L. siceraria PIs were immune to powdery mildew infection. Our data indicate that most of the bottle gourd PIs that we evaluated tended to be susceptible (Fig. 1). The skewness for the data was –1.247 and Kurtosis was 6.03 indicating that more PIs were susceptible. The Shapiro-Wilk statistic was 0.913 (Pr < W = < 0.0001) confirming that the data were not following a normal distribution. Although no immune reactions to the pathogen were found, a few PIs were observed to have a moderate or higher level of resistance. In addition, there was enough variability among the PIs for their reaction to powdery mildew to allow identification and future development of resistant germplasm. In Table 1 and Figure 1, two PIs with a rating of 1 are listed; however, subsequent tests indicate them to be moderately resistant but not immune. Based on the first two tests, 21 PIs had powdery mildew ratings of 4.5 or less. Extensive variability in the levels of powdery mildew development on plants within each PI was observed. Similar variability has been observed within PIs during evaluation of watermelon germplasm for resistance to powdery mildew (Davis et al., 2006b, 2007; Thomas et al., 2005) and papaya ringspot virus (Strange et al., 2002) and bottle gourd germplasm for Zucchini yellow vein mosaic virus (Ling and Levi, 2007). Several bottle gourd PIs had significantly lower levels of powdery mildew severity compared with the susceptible ‘Mickey Lee’ plants that were rated greater than 7 with a mean of 7.5 (Table 1). Many of the ‘Mickey Lee’ plants died as a result of powdery mildew infection on the hypocotyls. Data for all PIs tested are presented in Table 1. Because of poor seed germination, several of the PIs (22 PIs) were tested in only one or the other primary test. If we observed some level of resistance (less than 4.5 actual mean rating of two replications), we included these entries in our subsequent replicated test as this test was used to confirm the resistance in these selected PIs.
Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1359
The mean Wilcoxon rank scores for each PI and the mean ratings on the 1 to 9 scale are presented in Table 2 for the rescreen with the select PIs. The χ2 for the Kruskal-Wallis test was 38.98 (Pr > χ2 = 0.0490) indicating a significant difference (P = 0.05) among PIs. A significant correlation (R = 0.716, P < 0.0001) in the ratings between the first test (Table 1) and the subsequent replicated study (Table 2) for the 26 select PIs was observed, although the overall ratings were higher in the subsequent study. This replicated study was conducted in an enclosed growth room with controlled conditions, which could have led to higher disease pressure compared with the studies conducted in the greenhouse on 234 PIs. PI 271357, PI 271353, and PI 271359 all collected from India had ratings between 4 and 5 and can be considered moderately resistant. Powdery mildew infection was not observed on hypocotyls of many (greater than 50%) plants of these three PIs. Powdery mildew infection of hypocotyls was low for PI 381840, PI 271351, and PI 280635 and these three PIs had mean rating of 5.3. Most (90%) of the seedlings of PI 271357 were rated 4.0 or less. Similarly, greater than 50% of the seedlings of PI 271359, PI 271353, and PI 280635 were rated 4.0 or less indicating the availability of resistance in whole seedlings to powdery mildew. Plants of Mickey Lee had a mean rating of 8.3 in this test and severe powdery mildew infection of the hypocotyls was observed. Many of the Mickey Lee plants in this test were either dying or dead at the time of rating, and all were rated >8 on the 1 to 9 scale.
Results from the first cotyledon disk experiment are presented in Table 3. Of the 26 PIs, four had overall ratings of <10% surface area covered with powdery mildew. However, we did observe disease reaction variability within each of these PIs. Overall, PI 271353, PI 381848, and PI 487482 appeared to be resistant to powdery mildew based on the cotyledon disk ratings compared with PI 432341, which was significantly more susceptible. Over 56% of the cotyledon disks were rated 1 for PI 271353, PI 381848, and PI 273662, indicating the availability of highly levels of resistance within each of these PIs. Similarly, several of the other PIs had cotyledon disks that were rated 1 (Table 3).
Results from the second cotyledon disk experiment are presented in Table 4. PI 271353, PI 271354, PI 271477, and PI 487482 again had significantly less powdery mildew compared with some of the other PIs or Mickey Lee in the second cotyledon disk test. PI 273662, which had the lowest rating in the first cotyledon disk experiment, was observed to be susceptible in the second cotyledon disk test and in the replicated whole plant evaluation. However, in all these tests, we identified a few moderately resistant plants within PI 273662 and selected them for seed increase and future testing. Leaf or cotyledon disk inoculation methods have been successfully used to determine resistance in various cucurbit species (Cohen, 1993; Kuzuya et al., 2006; Morishita et al., 2003). In the second cotyledon disk experiment, because of limited availability of seeds and poor germination in several accessions, we considered cotyledon disks from single plants as a replication for analysis. Similarly, single plants have been considered as replications for evaluation of cucurbits for resistance to various diseases before (Davis et al., 2006b; Strange et al., 2002).
Because the entire seedling will be used as a rootstock for grafting, we feel the whole seedling ratings should be given more weight and the cotyledon disk method used for additional confirmation. Based on the whole seedling rating of less than 5 on the 1 to 9 scale and consistent leaf disk ratings of less than 2 on the 1 to 5 scale, PI 271353 can be considered the most resistant to powdery mildew among the bottle gourd PIs evaluated in this study.
Thomas et al.'s (2005) study of resistance to powdery mildew in watermelon concluded that only moderate to intermediate resistance to powdery mildew was present in watermelon PIs. In addition, the same study suggested that careful evaluation of PIs with intermediate resistance followed by careful selection of segregating individuals may provide a more robust source of resistance to powdery mildew. Similarly, in a more recent study by Davis et al. (2007), 1573 watermelon PIs were evaluated and none were found to be immune to powdery mildew; however, several resistant and moderately resistant PIs were identified. None of the watermelon PIs or cultivars tested were rated below 5 in the Thomas et al. (2005) study, and ratings of 4 to 6 were considered intermediate. In our study, we only selected PIs with ratings below 5.0 for further evaluation. In our replicated whole seedling evaluations, only three PIs had ratings below 5.0 (Table 2), and 10 PIs had ratings below 6. We have selected individual, resistant plants from these PIs for seed increase and further screening and selection.
Shifts in the powdery mildew pathogen populations with respect to race are common (Cohen et al., 2004; Sowell, 1982). In 1996, race 1 was considered to be the more prevalent in the eastern United States (McGrath and Thomas, 1996). Recently, a powdery mildew melon race 1 resistant watermelon line was released based on its reaction to a naturally occurring race 1 in Oklahoma (Davis et al., 2006a). All this information suggests that powdery mildew race 1 is important to the industry and resistance to this race in cucurbits needs to be developed. Cohen et al. (2004) suggested that race identification is extremely important when conducting resistance studies for basic research and for the commercial industry when releasing powdery mildew-resistant lines. In the same vein, additional studies to determine the reaction of the select L. siceraria to melon race 2 and other potential races will be especially helpful because of the potential for race shifts in powdery mildew pathogen.
There are several different ways in which the watermelon scion is grafted on the bottle gourd rootstock (Oda, 1993, 2002). These include the tongue approach, slant cut, tube grafting, hole insertion, and so on (Oda, 1993, 2002). In some of the grafting procedures such as the tongue approach and the hole insertion grafting, the rootstock seedlings are allowed to grow until the formation of the first true leaf before grafting. Hence, in our whole seedling evaluations, we evaluated the seedlings including the first true leaf. In some of the grafts, one cotyledon of the rootstock is cut off to create the area for the graft, whereas in others, both cotyledons may be cut off and the graft made on the hypocotyls. In such cases, it is necessary that the hypocotyls is healthy and free from diseases. In our rating system, if trace amounts of powdery mildew were observed on the hypocotyls, the plants received a rating of 5. Such plants generally had moderate to severe infection of cotyledons. PI 271357, PI 271353, and PI 271359, which were collected from India, can be considered as potential sources of resistance in hypocotyls to powdery mildew. However, as mentioned before, some plants within these PIs were susceptible, thus indicating a need for further screening and selection. Recently, a melon race 1 powdery mildew-resistant watermelon line was released (Davis et al., 2006a). It will be interesting to determine if the combination of resistance in the rootstock and scion will outperform the resistance in the scion by itself.
Recently Clarke et al. (2006) suggested a dual origin for L. siceraria, one from America and one from Asia. However, GRIN indicates that the majority of the PIs in the collection originated in Africa (48%), followed by Asia (29%), America (12%), and Europe (10%). In our first study (Table 1), only two PIs from the United States had disease ratings of 4.0 on the 1 to 9 scale. However, both these PIs did not fare as well in subsequent evaluations. Because of the suggestions of dual origins for bottle gourd (Clarke et al., 2006), it may be necessary to collect more germplasms from Asia and America and identify sources of resistance to various diseases prevalent in America.
Recently, resistance to zucchini yellow mosaic virus (ZYMV) was described in bottle gourd PIs (Ling and Levi, 2007). Several of the PIs with moderate resistance to powdery mildew described in our article also were reported to have various levels of resistance to ZYMV. PI 271353 was reported to have partial or moderate resistance to ZYMV, whereas PI 271357 and PI 271359 were reported to be highly resistant to ZYMV (Ling and Levi, 2007). PI 271353 was also reported to have resistance to ZYMV in an earlier study (Provvidenti et al., 1984). Bottle gourd PI with multiple disease resistance will be useful in developing multiple disease-resistant rootstocks.
Contributor Notes
We acknowledge the technical assistance of William Cook, Jessie Dufault, and R. Carrington in conducting the greenhouse experiments. The help of Drs. Jennifer Ariss and William Turechek for statistical analysis is greatly appreciated.
Mention of a trademark name or proprietary product does not constitute a warranty or guarantee by the U.S. Department of Agriculture nor does it imply exclusion of other products that may also be suitable.
To whom reprint requests should be addressed; e-mail shaker.kousik@ars.usda.gov
