Squash Team members
- Florida: Geoffrey Meru, Yuqing Fu, Prerna Sabharwal, Swati Shrestha and Shailesh Acharya
- Michigan: Mary Hausbeck, Carmen Medina-Mora, David Perla, Matthew Uebbing
- New York: Michael Mazourek, Chris Smart, Colin Day, Libby Indermaur, Gregor Inzinna, Taylere Herrmann
- Puerto Rico: Angela Linares
Obj. 2. Map and develop markers for disease resistance
QTL mapping of resistances: Powdery Mildew – C. pepo
Geoffrey Meru, Prerna Sabharwal, and Yuqing Fu & Michael Mazourek and Gregor Inzinna
In 2023, screening of 207 USDA germplasm accessions of C. pepo for PM resistance was conducted at the University of Florida (greenhouse) and Cornell University (field) XP GWAS. Seeds of each accession (n=15) were grown in three replicates in a randomized complete block design. Success PM (carrying PM-0) and Early Prolific Straightneck cultivars were used as resistant and susceptible checks, respectively. Pathogen inoculum was provided through naturally infected plants and by spraying spore suspension on the foliage. At the 6th true-leaf stage, symptom severity data were recorded on a scale of 0-100%, based on visible pathogen sporulation on the surface of top 4th leaf, bottom 4th leaf, stem above 4th leaf, stem below 4th leaf and the whole plant level (Figure 1).
Figure 1 Greenhouse screening at the University of Florida for powdery mildew resistance in the USDA C. pepo collection (n= 207).
Across locations, Success PM/ 209 (R check) and Early Prolific/208 (S check) were consistently tolerant and susceptible, respectively. On the other hand, wide phenotypic variation was observed across the 207 C. pepo accessions (Figure 2).
Figure 2 Variation in powdery mildew symptoms in a few USDA accessions (189, 59, 56 & 167) and the resistant (209) and susceptible (EP) checks.
The greenhouse trial in Florida revealed the most resistant USDA accessions across each disease severity parameter (Table 1): top 4th leaf (68), bottom 4th leaf (188), stem above 4th leaf (188, 189), stem below 4th leaf (125, 188, 189) and whole plant (188, 189).
Table 1 The top 10 C. pepo accessions tolerant to powdery at the greenhouse trial in Florida. Genotypes denoted by an asterisk (*) were tolerant across all the recorded parameters.
Powdery mildew disease severity data (top 4th leaf, bottom 4th leaf, stem above 4th leaf, stem below 4th leaf and the whole plant) was used for GWAS analysis. The phenotype data was derived from PM screening trials in Florida (greenhouse- 2023), NY (greenhouse-2022), and Michigan (field 2022). Corresponding genotype data for approximately 4 million SNPs was obtained for the 207 C. pepo accessions from the Boyce Thompson Institute (Fei Lab). The SNP dataset was first converted from its original Variant Call Format (.vcf) to a numeric text (.txt) format, followed by the execution of the Genome Association and Prediction Integrated Tool (GAPIT) through the University of Florida HiPerGator computing infrastructure. Three statistical models (MLM, FarmCPU and Blink) were deployed to detect genomic loci significantly associated with powdery mildew resistance. GWAS analysis using greenhouse data (Florida) revealed significant genomic loci associated with PM resistance (Figure 3) for top 4th leaf (Chr 11 and Chr 20), stem above 4th
leaf (Chr 4, Chr 14 and Chr 16) and whole plant (Chr 13, Chr 15, Chr 18 and Chr 20). Resistance loci for ‘top 4th leaf’ and ‘whole plant’ co-located on Chr 20, suggesting linkage/ pleiotropy for the two traits.
Figure 3 GWAS for PM resistance using phenotypic data collected in Florida (greenhouse).
Significant loci were detected for top 4th leaf, stem above 4th leaf and the whole plant phenotype. On the other hand, GWAS analysis using phenotypic data collected in NY (greenhouse) revealed significant hits for PM resistance (Figure 4) for top 4th leaf (FarmPCU Chr 2, Chr 4, Chr 7, Chr 13 ad Chr 19//Blink: Chr 3, Chr 4, Chr 5 and Chr 19), bottom 4th leaf (Chr 6, Chr 14 and Chr 19) and stem above 4th leaf (Chr 12). The genomic loci on Chr 4 and Chr 19 for ‘top 4th leaf’ parameter were consistently detected using both the FarmCPU and Blink models.
Figure 4 GWAS for PM resistance using phenotypic data collected in Florida (greenhouse).
Significant loci were detected for top 4th leaf, stem above 4th leaf and the whole plant. The lack of overlapping resistance loci between Florida and NY is not surprising because of the low correlations (r = 0.1 to 0.25) observed between the phenotypic data recorded at the two locations. This disparity may be a result of differences in powdery mildew strains in NY and FL, as well as contrasting experimental conditions. GWAS analysis using phenotypic data collected in Michigan (Field) did not yield any significant hits.
PM was very slow to develop in NY resulting in plants being very large and overgrown in the field. Acting on helpful advice from our colleagues in Charleston, XP-GWAS was conducted where we were able to identify extreme phenotypes where plants were either completely covered with mildew or surprisingly clean. Using the resequencing data aligned to the C39 reference genome, XP-GWAS revealed highly significant hits on chromosome 2 and 6 (Figure 5).
Figure 5 Manhattan plots illustrating XP-GWAS hits on C. pepo chromosomes 2 and 6
QTL mapping of resistances: Phytophthora – C. pepo
Geoffrey Meru, Prerna Sabharwal, and Yuqing Fu
In 2023, we continued population advancement for the introgression of crown rot resistance from 181761-36P (C. pepo) and 394-1-27-12 (C. moschata) into elite germplasm of various C. pepo market groups through backcrossing. More than 1,600 seedlings from 90 segregating families were screened for crown rot resistance in the greenhouse (Figure 6). A majority (68%) of which were susceptible (DS >4 out of 5), some (18%) showed moderate resistance (DS of 2-3.9 out of 5) while 14% showed high resistance (DS <2).
Figure 6 Segregation in Phytophthora crown resistance was observed among the ninety C. pepo families screened in the greenhouse at the University of Florida.
Among the promising families, SS2503, SS2528, 69-52 and SS2722, SS2523 and SS2636 were the most resistant (Table 2). These lines will be advanced to the next generation for further selection. As previously reported by Hausbeck’s group, we noted that backcrosses or crosses involving subspecies pepo were more tolerant than those involving subspecies ovifera. Crosses involving the latter will necessitate an additional selfing step to eliminate a potential susceptibility gene in subspecies ovifera.
Table 2 Out of the 90 segregating families screened in 2023, eighteen C. pepo and bridge lines [C. moschata x C. pepo] families were most resistant or tolerant to Phytophthora crown rot.
2.2 Marker development and verification
Identifying Genomic Regions Associated with the Novel Powdery Mildew Resistance Native to Cucurbita moschata
Michael Mazourek and Gregor Inzinna
Currently, squash growers are relying on one gene for powdery mildew resistance (Pm-0) introgressed from Cucurbita okeechobeensis martinezii into C. moschata over 50 years ago by Henry Munger at Cornell. Towards building a more robust resistance package for squash growers, we have identified a native PMR in C. moschata. This novel resistance protects the leaf surface, stem and petiole, while the dominant Pm-0 gene primarily confers resistance to the stem and petiole. However, the physiological mechanisms and segregation patterns of this new resistance were unknown.
As previously reported, in the winter of 2021-2022 we genotyped resistant and susceptible bulks of an F2 population of a cross between Waltham and this novel resistance source and found regions associated with this trait by genotyping the 22 most resistant, 22 most susceptible individuals and the parents and identified genomic regions associated with the resistance using a BSA approach. Initial results indicated chromosome 13 was the most significantly associated region of the genome (Rifu_version1); segregation ratios for the resistance indicated a recessive resistance at the outset and we used this to guide our path forwards. Over the course of 2022 we created markers that spanned the 2MB region on chromosome 13 and differentiated the resistant and susceptible parents (Table 2).
Table 2 Fine mapping Chr 13 QTL
In the past year, we have worked to refine the resistance locus. In the spring of 2023, 190 F2, 90 F2 derived from the reciprocal cross (rF2), 120 BC1F2, and relevant controls were grown and genotyped for crossovers using a combination of HRM and PCR markers that were identified from the BSA. DNA was extracted using a CTAB method and SNP calls confirmed using Sanger sequencing. The resulting 60 individuals with crossovers were transplanted into the field for PM phenotyping and seed increase of F2:3 families. During the field season, we were able to self pollinate 48 of these individuals and had enough seed for further greenhouse PM screens of 39 resulting F3 families.
Using the same method in our 2021-2022 winter greenhouse screen, 30 plants each of the 39 F3 families were screened in the greenhouse during the winter of 2023 to 2024. We also evaluated 300 susceptible(S) by resistant(R) F2, 300 (RxS) revF2, 300 ([SxR]xS)BC1F1, and 450 ([RxS]xR)revBC1F1 plants to identify resistant individuals for further genotyping and calculating more significant segregation ratios for the phenotype. In the greenhouse screen of the F3 familes, only 2 families showed a strong resistance phenotype, and 4 segregated for both leaf and petiole resistance. Based on the genotyping data of the crossovers in the F3 familes, it
became obvious that 2-3 more loci may be required to reproduce the PM leaf resistance of the parent.
This winter, we were also able to rerun the BSA-S using the original 22 resistant and susceptible individuals and a new C. moschata reference genome provided by Zhangjun Fei. This new reference genome was much more complete. This BSA run increased the significance of other regions of the genome and indicated as many as 4 other regions of the genome may be associated with this resistance (Figure 5). We have developed a new marker set to clarify which of these regions is necessary for the resistance and have extracted DNA from resistant and susceptible individuals from these individuals from the 2023-24 winter greenhouse screen. We hope to run these markers of the next few weeks to clarify our findings and create marker sets for introgressing this trait into common cultivars.
Figure 5 BSA-Seq results for native PMR in C. moschata with updated reference genome
Obj. 3.4. Introgress, pyramid/stack resistances into advanced breeding lines
Evaluate the susceptibility of squash to Phytophthora crown rot under greenhouse conditions, 2023.
Carmen Medina-Mora and Mary Hausbeck with Michael Mazourek and Gregor Inzinna
A greenhouse trial at Michigan State University Plant Science Greenhouses in East Lansing, MI was established to evaluate the susceptibility of Cucurbita species to Phytophthora capsici. The trial consisted of 16 winter squash accessions (PIs) and 1 C. maxima Hubbard type, ‘Golden Delicious’ provided by Dr. M. Mazourek (Cornell Univ., NY). On 8 Sept, 18 seeds per line were directly seeded onto 3×3 in plastic containers containing SureMix soilless medium. On 10 Oct, when seedlings were at 3-4 leaf stages, all seedlings were inoculated with 5 mls of a zoospore suspension (1×105/ml) consisting of a 1:1 mixture of 2 strains of P. capsici (M. Hausbeck P. capsici collection, strains SP98 and 12889). Seedlings were distributed in a randomized complete block design with three blocks and six replications per block. Disease ratings based on a 0-5 categorical scale (0= healthy, 1=small lesion at crown, 2= water-soaked lesions beyond cotyledons, 3=wilt and plant partially collapsing, 4= severe wilt and plant completely collapsing, and 5=plant death) were conducted 7 days after inoculation on 16, 23, and 30 Oct. Data were analyzed using SAS PROC GLIMMIX procedure (ver.9.4; SAS Institute, Cary, NC) with maximum likelihood estimation method and Kenward-Rogers degrees of freedom as options. Differences among treatments were analyzed using Least-Square Means comparisons (P=0.05).
Table: Average disease ratings of Squash varieties based on a 0-5 categorical scale (0= healthy, 1=small lesion at crown, 2= water-soaked lesions beyond cotyledons, 3=wilt and plant partially collapsing, 4= severe wilt and plant completely collapsing, and 5=plant death)
The disease progressed rapidly; 7 days after inoculation (on 16 Oct) susceptible accessions were partially collapsing. On 30 Oct, 21 days after inoculation, at the end of this trial a high number of seedlings were severely diseased or dead. Overall, 10 out of the 12 accessions tested were as susceptible as the susceptible standard ‘Golden Delicious’. The remaining 2 accessions (‘Queensland Blue’ and ‘Buttercup Burgess Strain’) were less susceptible than ‘Golden Delicious’ but not as resistant as the C. maxima Kabocha type included in this trial.
Evaluate Age-Related Resistance (ARR) of squash breeding lines to fruit rot caused by Phytophthora capsici, 2023.
Carmen. Medina-Mora, Matthew Uebbing, and Mary Hausbeck with Michael Mazourek and Gregor Inzinna
A pollination plot was established at Michigan State University Plant Pathology Farm, East Lansing, MI; the Capac loam soil was plowed and disced on 9 May and 15 May, respectively, and amended with 130 lb Urea and 130 lb Potash on 16 May. On 13 Jun, 4-week-old seedlings were transplanted every 18-in onto raised beds (rows 100-ft X 16-ft center to center) covered with black polyethylene plastic. Until fruits were harvested, 28% fertilizer (1gal/A) was applied weekly and non-target diseases and insects were controlled. Starting on 5 Jul, Admire Pro (10.5 fl oz) was delivered through drip tape to control cucumber beetle and squash bug. In addition, on 28 Jul, squash plants were sprayed with Warrior (1.9 fl oz) to control beetles. Starting on 13 Jul, a fungicide program including chlorothalonil, quinoxyfen, and cyflufenamid [Bravo (3pt/A), a mixture of Quintec (6 fl oz) and Bravo (3 pt/A), a mixture of Torino (1.6 fl oz) and Bravo (32 fl oz), Torino (1.6 fl oz), and Quintec (6 fl oz)] was applied, as needed, to control the incidence of powdery mildew. To facilitate fruit set and reduce natural flower abortion, flowers at anthesis were hand-pollinated using an artist’s paint brush and marked at the petiole using colored flagging tape. Because flower development is asynchronized among breeding lines, the presence of flowers at anthesis was monitored daily for 32 days and flowers were hand-pollination every other day. A total of 500 flowers were hand-pollinated to harvest a maximum of 8 fruits corresponding to 21 days post-pollination (dpp) and 8 fruits corresponding to 28dpp per line. A 9mm mycelial plug of a 7-day-old culture of P. capsisci (strain SP98) was placed onto the skin of each disinfected fruit (10% bleach and rinsed with water) on same day each fruit was harvested. Fruit rot was evaluated 5 days post-inoculation (dpi) and disease assessment included: 1) lesion size, 2) incidence of hyphae beyond inoculation point, and 3) disease severity based on a 0-4 categorical scale (0= healthy, 1=water-soaked tissue, 2= light visible mycelial growth, 3=moderate mycelial growth, 4= dense mycelial growth). Incidence of fruit rot symptoms was recorded as binary data, where evidence of water-soaked tissue or mycelia was considered 100% incidence and the absence of these symptoms as 0% incidence. The size (width and length) of fruit lesions were recorded in cm and the area of an ellipse was calculated for each fruit lesion as follows: Area = 3.14*1/2 width *1/2 length. Data for fruit lesion size were analyzed SAS PROC GLIMMIX (ver.9.4; SAS institute, Cary, NC) with Least-Square Means (P=0.05) for pairwise comparisons. The incidence of water-soaking in the pulp was recorded as binary data, where evidence of water-soaked tissue was considered 100% incidence and the absence of water soaking as 0% incidence.
Fruits of ‘Buttercup Burgess Strain’ and ‘Thunder F1’ that developed at 28dpp had lower fruit rot incidence (12.5% and 0.0%), smaller fruit size lesions (0.39 and 0.00 cm2), and lower incidence of water-soaked tissue in the pulp (13.0% and 13.0%) than fruits that developed at 21 dpp, indicating age-related resistance. The incidence of fruit rot and water-soaked tissue in the pulp for fruits of ‘Dickinson’ that developed at 28dpp were unexpected; might be an effect of mechanical injury to the skin of the fruit. Fruits of ‘Queensland Blue’ and ‘Kestane’ did not demonstrate age-related resistance and were as susceptible as the susceptible control ‘Golden Delicious’.
Table: Incidence of fruit rot, area of fruit lesion, and percent incidence of water soaked lesions
Improved Dickinson (Cucurbita moschata Duchesne) breeding lines and their impact on processing quality: Chris Smart, Libby Indermaur, Colin Day, Taylere Herrmann with Michael Mazourek and Gregor Inzinna
To expand on preliminary work from 2022, a protocol was developed in consultation with Olga Padilla-Zakour (Cornell University, Department of Food Science) and the Cornell Food Venture Center Pilot Plant in Geneva, NY to process fruit at an industrial scale. Accessions, including two C. moschata breeding lines, plus both parents, ‘Dickinson’ and ‘Bugle’, were grown in a research field in Geneva, NY. Representative images of fruit from each accession are shown in Figure 6. For processing, fruit were washed, seeds were removed, and fruit flesh with skin from 20 kg of representative squash were steam cooked for 20 minutes (‘Dickinson’, TR2-03, and TR2-06) or 40 minutes (‘Bugle’) in 100% humidity at 210˚F in three replications. Cooked squash was processed, also in triplicate, in a Bertocchi CX extractor at 1950 rpm. Squash purée was cooked in a kettle to 180˚F and subsequently filled in ten 14oz, 300×407 2pc tinplate cans with lacquered bodies and ends per replication. Cans were sealed with an atmospheric closing machine and products were pasteurized in an industrial retort with a chamber temperature of 250˚F for 62 minutes. Cans were stored at room temperature (approximately 68˚F) for one month for later analysis.
For each accession, three cans from each processing run (nine cans total) were evaluated for ˚Brix, color, consistency, moisture content, and pH. Purée from a single brand of commercially available canned pumpkin (Libby’s) was used as a comparison. Color was assessed with an UltraScan VIS Spectrophotometer by measuring the CIE color space values of L*, a*, and b*.
Consistency was measured with a Bostwick consistometer at both 30 seconds and 1 minute, and is reported as distance flowed in cm. Moisture content (%) was measured with a MX-50 Moisture Analyzer. Representative images of opened cans are shown in Figure 7.
