Cucurbit Powdery Mildew Race Identification by Triplet-Septet and Disease Progress Estimation

ABSTRACT The Triplet-Septet (TS) set of melon cucurbit powdery mildew (CPM) race differentials (CPMRD) was established to provide an international means for objective and uniform identification and designation on CPM races. The Area Under Disease Progress Stairs (AUDPS) method for disease progress estimation was derived from the Area Under Disease Progress Curve (AUDPC) method, and both have been used to evaluate disease progress on other crops. We aimed to identify a melon CPM race on the TS melon CPMRD, and estimate disease progress thereon using AUDPC and AUDPS. Plants were inoculated at the 3 to 4 true leaf stage. Severity of CPM infection was evaluated on the 21 TS melon CPMRD at 15, 22, 32, and 41 days after inoculation (DAI) using a visual scale. The CPM population in the greenhouse was identified as race S based on reactions of a set of 11 commonly used melon CPMRD, and it may also be designated as 127.127.126 on the TS melon CPMRD. AUDPS identified higher levels of disease than AUDPC, and its results agreed with those obtained by the commonly used melon CPMRD conventional race identification methods (current and Triplet-Septet). AUDPS can be used to evaluate the disease progress on CPM.

The fungus predominantly colonizes leaf surfaces and may occasionally infect stems and fruits.Infection can decrease photosynthetic area by early loss of leaves, resulting in low production of photoassimilates.Consequently, infected plants produce fewer and smaller fruits, with low commercial quality, especially lower soluble solids levels, and sun burned rinds (Viana et al., 2001).
Control of CPM has been done by using chemical products and resistant cultivars.The pathogen has, however, developed resistance to most fungicides used for control (McGrath, 2001;2006).Cultivars with genetic resistance Cucurbit Powdery Mildew Race Identification by Triplet-Septet and Disease Progress Estimation are advantageous, because resistance is equally distributed throughout all plants in a field, and its use may decrease, or even exclude, the application of fungicides.Resistance to CPM was first available in western United States shipper type cantaloupe melon in the 1930s with the development and release of 'PMR 45' (Pryor et al., 1946).
The effectiveness of genetic resistance is decreased by the occurrence of pathogen races.Numerous (> 46) P. xanthii races on melon have been identified, and approximately 36 sources of resistance are known in melon (McCreight, 2006;McCreight et al., 2012).The development of new cultivars resistant to all races is impractical.Thus, breeding programs seek to provide resistance for regional needs, regarding the races with higher likelihood of occurrence, and that requires the constant monitoring of races.
There has been confusion about powdery mildew race identification due to different methods of evaluation and sets of melon race differentials, and nomenclature.A new objective and uniform system of pathogenic race identification and denomination has been proposed based on a uniform set of race differentials, a uniform code for scoring disease reactions, and a uniform screening methodology to classify CPM races based upon a three-part, numerical code, named Triplet-Septet (Lebeda et al., 2016).This proposal established a specific set of 21 race differentials, divided into three subgroups of seven that when challenged with any isolate will generate a three-part score or code that will range from 0.0.0 (all resistant) to 127.127.127(all susceptible), and the method is mathematically able to identify 2,097,152 races (Limpert & Müller, 1994).Stadnik et al. (2001) pointed out the lack of methods that suitably measure powdery mildew in quantitative ways.Van der Plank (1963) proposed a method that allows the progressive evaluation of disease development, Area Under Disease Progress Curve (AUDPC).Simko & Piepho (2012) proposed the Area Under Disease Progress Stairs (AUDPS) to better estimate the first and last evaluations of disease severity, as compared with AUDPC.
We confirmed the pathogenic race identification of the CPM population resident in a greenhouse at Salinas using whole plants and characterizing it on melon using the alphanumeric nomenclature that developed over time and the Triplet-Septet method.Disease reactions of the melon CPM race differentials were also characterized using two methods for estimating disease progress and their respective standardized estimators.

MATERIALS AND METHODS
The experiment was carried out at U. S. Department of Agriculture (USDA), Crop Improvement and Plant Protection Research Unit, Salinas, CA, from August to October 2016.We used the triple septet melon race differentials suggested by Lebeda et al. (2016) (Table 1).
Seeds of the 21 Cucurbit Powdery Mildew Race Differentials (CPMRD) were germinated on moistened paper towels in plastic boxes at 25 ºC and a 12-h photoperiod, in a CPM-free growth chamber (Conviron ® model 6050).Seedlings were transplanted at the cotyledon stage to 0.5 L plastic pots filled with potting mix (Sun Land™, Watsonville, CA) and grown under the same conditions as described above.Due to poor germination of 'Amarillo' seeds, only a single plant of this differential was used in the test.
Seedlings were transferred 16-days after transplanting to a greenhouse naturally infested by CPM and arranged in three randomized blocks with one plant of each CPMRD per rep.Plants were watered daily with dilute (1:100) 20N-20P-20K fertilizer solution applied via drip irrigation.Identity of P. xanthii was confirmed by observing conidia with fibrosin bodies, with the aid of a light microscope.
The resident CPM population was previously confirmed to be P. xanthii based on morphological and molecular characteristics (Bojorques Ramos et al., 2011).Inoculum was subsequently collected from infected melon plants in the same greenhouse.The 3 rd or 4 th true leaf of each test plant was inoculated with fragments of mycelia immediately after transplanting with the aid of an artist brush.Inoculated true leaves were evaluated at 15, 22, 32, and 41 days after inoculation (DAI).Levels of infection were assessed using a 1 to 9 visual scale as follows: 1 = no evidence of disease; 2 = trace of hyphae, no detectable sporulation; 3 = hyphae restricted, no detectable sporulation; 4 = few colonies present, sporulation; 5 = scattered colonies, sporulation; 6 = numerous colonies, sporulation; 7 = ≈ 50% of adaxial surface covered with hyphae and spores, few colonies on abaxial surface, abundant sporulation; 8 = 50-75% of adaxial surface covered with hyphae and spores, colonies on the abaxial surface, abundant sporulation; and 9 = > 75% of adaxial surface covered with hyphae and spores, numerous or coalesced colonies on abaxial surface.
Powdery mildew race identity was determined according to the reactions of the 21 differentials 41 DAI.
Susceptibility was attributed to those with means ≥ 4.0 and means < 4.0 indicated resistance, and race nomenclature was performed using both alphanumeric (McCreight et al., 2012) and Triplet-Septet methods, the latter as proposed by Lebeda et al. (2016) but, with modifications, given the fact that the authors suggested that race identification essays should be done using excised leaf discs under axenic conditions.Each member within a triplet is assigned a unique weight that becomes its score when susceptible; when resistant it is scored zero.The total score for a triplet reveals at a glance which differentials were susceptible and which were resistant.
Disease progress was estimated using AUDPC (Van der Plank, 1963) and AUDPS (Simko & Piepho, 2012) methods.Standardized estimates for both methods were calculated (sAUDPC and sAUDPS), as stipulated by Simko and Piepho (2012).Simulated disease progress estimates were generated for each level of the 1-9 disease severity scale at the same intervals the test data were recorded in order to compare AUPDC and AUPDS estimates, and identify susceptible and resistant differentials via these disease progress estimates.Differentials with mean AUDPC, AUDPS, sAUDPC, and sAUDPS estimates lower than those obtained for the disease rating scale score 4 simulations were, thus, considered resistant.
Data were subjected to Analysis of Variance and means comparisons by Tukey HSD test.AUDPC and AUDPS, sAUDPC and sAUDPS analyses were compared by Student's t-test.All analyses were performed with SAS 9.4 University Edition.Data from 'Amarillo' were not included in the statistical analyses, as there was only a single plant of this differential.However, its reaction was used to identify the powdery mildew race using the alphanumeric and Triplet-Septet methods.

Reactions of CPM race differentials
Infection was uniform, as represented by the low coefficient of variation for disease reaction (Table 1).P. xanthii colonies developed on 20 of the 21 differentials, and mean disease reaction scores ranged from 1.0 (PI 313970) to 9.0.This reaction pattern was consistent with race S. The first reports of race S were from commercial fields and experimental plots in Yuma, AZ, and Holtville, CA in 2003, when all commercial cultivars and all the commonly used melon CPMRD were susceptible; only PI 313970 was resistant (McCreight et al., 2005;McCreight & Coffey, 2011).
Disease severity differed significantly among the CPMRD.On the most infected differentials, leaves were fully covered by the fungus, with intense sporulation, and mycelia were observed on stems as well.P. xanthii mycelia were found after the test was terminated on fruits of Ames 31282, a kind of infection not frequently reported.This accession is resistant to race pxCh1 that was reported in China (Liu et al., 2010)2010; its susceptibility to race S was previously reported (McCreight et al., 2012).
Intermediate levels of infection were found on MR-1, PI 124111, and PI 124112; the latter two had the same mean disease reaction values and were statistically greater than MR-1 (Table 1).It is interesting to note that MR-1 was selected from PI 124111 for uniform reaction to downy mildew (Pseudoperonospora cubensis (Berk.And Curt.) Rostow) and P. xanthii races 1, 2, and 3 (Thomas, 1986).The lower means of these genotypes were due in part, perhaps, to the presence of blisters on their leaves, evidence of a foliar response that is a non-race-specific form of resistance (Sedlářová et al., 2009) first recognized in hops (Salmon, 1917;Salmon, 1919).Blisters have the appearance of water-soaked and raised lesion spots, sometimes chlorotic or necrotic at the center, that restricts or blocks the pathogen development on the foliar surface (McCreight, 2003;McCreight & Coffey, 2011).Nevertheless, mycelial growth was consistently observed on leaves from secondary stems of MR-1, and on both primary and secondary stems of PI 124111 and PI 124112.Republic (Lebeda & Sedláková, 2004;Sedlářová et al., 2009), as well as races SD and SDW that were isolated in Yuma, AZ and Imperial Valley (Coffey et al., 2006).Almost all (20/21) melon powdery mildew race differentials are susceptible to race 127.127.126.Thus, it is expected that cultivars developed using any of the sus-ceptible differentials as sources of CPM resistance will be susceptible to this race as well.That information may help the choice of cultivars to be eventually cultivated in areas affected by powdery mildew.Hudson de Oliveira Rabelo et al.

Application of the
The Triplet-Septet method was based on a similar approach adopted for downy mildew (Bremia lactucae) races on lettuce, proposed by Van Ettekoven & Van Arend (1999).This method is recommended by the International Bremia Evaluation Board (IBEB) (ISF, 2016), and internationally adopted by lettuce seed companies.
The Triplet-Septet is a recently proposed method and being slowly adopted.Lebeda et al. (2012) es (55.78.124, 23.0.124, 55.13.125, 51.12.116, 127.63.127, 55.14.125, 23.4.125, 55.5.125, 55.15.125, 55.0.126, and 55.47.125) were found.The aforementioned codes were obtained using a preliminary set of 21 differentials, that utilized 'Solatur' not Ames 31282 (Lebeda et al., 2016) as the last member of the third triplet (group 3.7).Thus, respective comparisons of the scores of the first two septets of the 2012 (preliminary) and 2016 (current) triple septets reveal the unique characteristic of race S, namely it infects all commonly used melon CPM race differentials, with the exception of PI 313970.The third septets differ by one member (Ames 31282 vs. 'Solatur') and may not, therefore, be directly compared, though the two alternates are highly susceptible to many isolates (Lebeda et al., 2016).Lebeda et al. (2012) noted that the use of the triplet code revealed a great potential to identify new sources of resistance to powdery mildew, which may favor gene pyramiding.The same authors also emphasized that some differentials, previously considered as equivalents for powdery mildew resistance, showed different reactions, possibly due to the complexity and variability of mechanisms for resistance in the hosts, as well, the wide number of powdery mildew races, especially P. xanthii.Another explanation may result from the fact that many differentials possess two or more genes (Pitrat et al., 1998) for resistance to P. xanthii, that though defeated may contribute some level of resistance (Simko et al., 2014).

Disease progress
Analysis of variance identified significant differences (P < 0.01) among the differentials for all disease progress measurements in the greenhouse study (Table 2).
Student's t-tests showed significant differences (P < 0.01) between AUDPC and AUDPS, and between sAUD-PC and sAUDPS (Table 2).Means separations by the four methods were similar, nearly identical, with a single difference between AUDPC and AUDPS, and no differences between sAUDPC and sAUDPS.
AUDPS estimated higher disease levels than AUDPC, with overall means of 220.05 and 168.12, respectively.In contrast, lower overall means were observed for sAUDPS (6.35) in comparison to sAUDPC (6.46).Nevertheless, Simko and Piepho (2012)   Simulations must be done for each experiment, because the resistance delimiters may vary according to the rating scale used, number of assessments, and intervals among assessments (Simko & Piepho, 2012).PI 313970 is the only CPM differential with values lower than those aforementioned, thus, its resistance was confirmed quantitatively as well.The AUDPS method of disease progress estimation proved better than AUDPC in this test for evaluation of melon genotypes for resistance to cucurbit powdery mildew.Resistance delimiters must, however, be calculated in all experiments, because they may vary as a function of the rating scale, number of assessments and intervals between assessments (Simko & Piepho, 2012).

CONCLUSION
The previous race S is equivalent to the race 127.127.126 by method Triplet-Septet.The Genotype PI 313970 is resistant to the race 127.127.126 of Podosphaera xanthii.
The differentials showed different levels of susceptibility.
The method AUDPS is indicated to evaluate genotypes for resistance to powdery mildew.However, resistance delimiters must be calculated in all experiments, since they may vary in function of the rating scale, number of assessments and interval between assessments.
Powdery mildew isolates identified by previous methods should be reclassified with the method Triplet-Septet, to be possible associations among previous researches and the new ones, aiming the standardization.
PI 313970 was the only resistant differential; it exhibited blister-like resistance, and it was statistically different from the other differentials.Resistance to race S is controlled by a single, recessive gene(McCreight & Coffey, 2011).Although rated resistant for foliar reaction, P. xanthii colonies were observed on stems of this differential; yet this kind of infection is not considered by the severity scales so far.This discrepancy between the foliar and stem reactions was previously observed on some plants of the same genotype byMcCreight & Coffey (2011), who attributed it to a possible heterogeneity of virulence factors in powdery mildew populations.PI 313970 was susceptible to race F in Czech Triplet-Septet method to this data set generated the Triplet-Septet code for race S: 127.127.126(Table 1).The totals for the first two triplets indicate all members of the two triplets were susceptible.The score of the third triplet indicates the first member (weight = 1) was Cucurbit Powdery Mildew Race Identification by Triplet-Septet and Disease Progress Estimation resistant and the others susceptible.
Powdery mildew isolates identified by previous methods should be characterized using the Triplet-Septet method in order to relate previous alpha numeric race designations with their respective Triplet-Septet codes.So doing would facilitate communication among established and new plant breeders and pathologists, and would help them to more clearly relate race designations (with respect to the pathogen identity and host resistance profiles of cultivars and breeding lines) with extensionists, professional crop advisors, and farmers.The susceptible differentials as determined by the alphanumeric and Triplet-Septet methods exhibited different levels of susceptibility.Disease progress estimates provide accurate disease assessments when done at equally distributed time intervals, and permit identification of multiple QTL for disease reaction.Such an approach may, therefore, be used to identify non-race specific resistance to CPM to stabilize reactions of melon genotypes to powdery mildew whenever new pathogenic races appear in endemic P. xanthii populations.

Table 1 :
(Lebeda et al., 2016) cucurbit powdery mildew race differentials in the melon Triplet-Septet(Lebeda et al., 2016)to an isolate of Podosphaera xanthii in a greenhouse 41 days after inoculation, their summary disease reactions, and Triplet-Septet groups, weights and scores Coffey, 2011.wed by the same letter do not differ significantly at P < 0.05 by Tukey HSD test.yR= resistant, S = susceptible; **Significant at P < 0.01 by F test.x Reaction based on a single plant, not included in analysis of variance.wMcCreight andCoffey, 2011.

Table 2 :
(Simko and Piepho, 2012)es for 20 melon powdery mildew race differentials using Area Under Disease Progress (AUDPC), Area Under Disease Progress Stairs (AUDPS), and standardized values for the two disease progress estimators(Simko and Piepho, 2012); 41 days after inoculation in a greenhouse, Salinas, CA Simko and Piepho, 2012.imko and Piepho, 2012.xMeanswithin columns followed by the same letter do not differ significantly at P < 0.05 by Tukey HSD test.**Significant at P < 0.01.Hudson de Oliveira Rabelo et al.

Table 3 :
Simulated data for disease progress using Area Under Disease Progress (AUDPC), Area Under Disease Progress Stairs (AUDPS), and standardized values for the two disease progress estimators, sAUDPC and sAUDPS and Piepho, 2012).A 1 to 9 disease rating scale was used to generate simulated raw data 15, 22, 32 and 41 days after inoculation (DAI) Simko and Piepho, 2012.imko and Piepho, 2012.