Significance of seedborne Stemphylium vesicarium in Stemphylium leaf spot of spinach

Objective 1: Quantify the relative abundance of Stemphylium species pathogenic on spinach that occur on spinach seed lots.

The total incidence of Stemphylium on the 12 seed lots seed ranged from 2.5 to 73.5% (Table 1). S. vesicarium was the predominant species, with the incidence on seed ranging from 1.1 to 70.6%. This species made up 42.7 to 100% of the Stemphylium isolates obtained from each seed lot. The next most predominant species was S. beticola, at an incidence ranging from 0.2 to 5.9% of seed. This species made up 2.3 to 56.3% of the Stemphylium isolates obtained from each seed lot. Other species that were isolated from fewer seed per lot included S. eturmiunum, S. amaranthi, S. astragali, S. simmonsii, S. chrysanthemicola, S. lucomagnoense, S. gracilariae, and S. drummondii. The only species documented to be pathogenic to spinach are S. vesicarium, S. beticola, and S. drummondii (Liu et al. 2020; Spawton et al. 2019; Zhou et al. 2011). A subset of isolates of each species from these seed lots, besides those of S. vesicarium tested, was tested for pathogenicity on the spinach cv. Mandolin, including S. beticola (n = 7), S. eturmiunum (4), S. amaranthi (4), S. astragali (1), S. simmonsii (5), S. chrysanthemicola (1), S. lucomagnoense (1), S. gracilariae (1), and S. drummondii (1). None of these isolates caused leaf spots on spinach, even though S. beticola and S. drummondii are known spinach pathogens. The lack of pathogenicity of isolates of the latter two species may reflect loss of virulence when isolates are stored for an extended period, as has been documented by other researchers working with Stemphylium species that are pathogenic on spinach (Stephen Koike, TriCal Diagnostics).

Table 1. Incidence (%) of Stemphylium species isolated from 12 spinach seed lots

^a   Seed of each lot was surface-sterilized in 1.2% NaOCl for 90 s, triple-rinsed in sterile, de-ionized water, dried, plated on NP-10 agar medium (Sorensen et al. 1991), and incubated as described by the International Seed Federation (2017). Single-spore isolations were carried out as described in the main text. Species determination was based on the best match of the cmdA sequence of the isolate to cmdA accessions in GenBank. All sequences were 95.29% identical to the best matching GenBank sequences of the species listed. The number in parentheses is the percentage of isolates identified as that species out of all Stemphylium isolates from that seed lot.

^b   - = the species was not isolated from the seed lot.

^c   Seed lot planted for a spinach crop in Texas in 2020 that developed symptoms of Stemphylium leaf spot.

^d   Seed lot was treated with a fungicide, so seed was rinsed under running deionized water for 15 min before sterilization and incubation on NP-10 agar medium for isolations.

The percentage of S. vesicarium isolates pathogenic to spinach from each of the 10 seed populations used in Objective 2 ranged from 0 to 100% (Table 2). Many vegetable seed companies select which spinach seed lots to treat with hot water for certified organic control of seedborne Stemphylium spp. based on the incidence of seed on which Stemphylium spp. are observed. The presence of significant proportions of non-pathogenic isolates on seed lots in this study suggests that companies may overestimate the amount of pathogenic S. vesicarium.

Table 2. Summary population statistics for non-clone corrected data for populations of Stemphylium vesicarium from 10 spinach seed lots used in a worldwide seed population study, and a subset of isolates tested for pathogenicity to spinach, with isolates genotyped using nine simple sequence (SSR) markers

^a   Number of isolates genotyped in each population.

^b   Number of multilocus genotypes (MLGs) in that population. This is also the number of individuals in each population after clone-correction.

^c   Expected number of MLGs after rarefaction at the smallest sample size.

^d   Number of MLGs found only in that population.

^e   Clonal fraction, estimated as 1 - (MLG/N).

^f    Simpson’s diversity index (Simpson 1949). Numbers in parentheses is the 95% confidence interval after 1,000 bootstraps.

^g   At least 50% of the isolates from each seed lot were tested for pathogenicity to the spinach cv. Mandolin.

^h   Summary statistics calculated for the population of S. vesicarium from each of the 10 seed lots

ⁱ    Summary statistics calculated for the isolates from each seed lot tested for pathogenicity, with each tested isolate categorized as pathogenic or non-pathogenic. The number of isolates from each seed lot that comprised the pathogenic and non-pathogenic groups is displayed.

Of the 110 isolates of S. vesicarium for Objective 3 that were tested for pathogenicity, 56 were pathogenic to spinach and present in three of the four populations: TXfoliar1, TXfoliar2, and TXseed1 (Table 3). In contrast, the 54 isolates that did not cause Stemphylium leaf spot were only present in the seed populations, i.e., all the foliar isolates were pathogenic to spinach.

Table 3. Summary statistics of non-clone corrected data for four spinach populations of Stemphylium vesicarium for a seed vs. foliar population study, and for the subset of isolates tested for pathogenicity, with isolates genotyped using nine simple sequence (SSR) markers

^a   Number of isolates genotyped in each population.

^b   Number of multilocus genotypes (MLGs) in that population. This is also the number of individuals in each population after clone-correction.

^c   Expected number of MLGs after rarefaction at the smallest sample size.

^d   Number of MLGs found only in that population.

^e   Clonal fraction, calculated as 1 - (MLG/N).

^f    Simpson’s diversity index (Simpson 1949). Numbers in parentheses is the 95% confidence interval after 1,000 bootstraps.

^g   At least 50% of the isolates from each of the four populations were tested for pathogenicity to the spinach cv. Mandolin.

^h   Summary statistics calculated for the populations of S. vesicarium from each of the four populations.

ⁱ    Summary statistics calculated for isolates from each population tested for pathogenicity, with each tested isolate categorized as pathogenic or non-pathogenic. The number of isolates from each population that comprised the pathogenic and non-pathogenic groups is displayed.

Objective 2: Assess the role of global spinach seed trade in moving isolates of S. vesicarium, causal agent of Stemphylium leaf spot of spinach, by characterizing populations of S. vesicarium genetically from seed lots grown in the five major spinach seed-producing countries of the world.

In total, 199 multilocus genotypes (MLGs) were detected in the 400 S. vesicarium isolates from spinach seed, ranging from 2 for seed lot USA1 to 39 for lot FR2 (Table 2). The number of MLGs unique to each population (i.e., only occurring in one of the 10 populations) ranged from 0 for lot USA1 to 36 for FR2, with 184 of the 199 total MLGs unique to specific seed lots. Diversity, based on Simpson’s diversity index (1 - λ), ranged from 0.219 for USA1 to 0.974 for FR2. The most frequent MLG was MLG.167, which was detected in 118 of the 400 isolates, in 7 of the 10 seed populations (1 isolate from DK1, 24 from DK2, 35 from FR1, 1 from FR2, 5 from NZ1, 35 from USA1, and 17 from USA2; Fig. 1A). The next most frequent MLG was MLG.115, detected in 35 of the isolates from five of the seed populations (1 isolate from DK2, 2 from FR1, 9 from NZ1, 5 from USA1, and 18 from USA2). Of the 199 MLGs detected, 175 (88%) were comprised of only one isolate.

Of the subset of 248 seed isolates of S. vesicarium that were tested for pathogenicity, the 84 that were pathogenic to the spinach cv. Mandolin originated from 7 of the 10 seed lots: DK1, DK2, FR1, FR2, NZ1, USA1, and USA2 (Table 2). The 164 isolates that did not cause Stemphylium leaf spot originated from 8 seed lots: DK1, DK2, FR2, NL1, NL2, NZ1, NZ2, and USA2. Therefore, 169 MLGs were detected in the 248 isolates tested for pathogenicity to spinach, of which 158 MLGs (93%) were each comprised of a single isolate. The 84 pathogenic isolates belonged to 11 MLGs, all of which were unique to the pathogenic isolates, while the 164 non-pathogenic isolates belonged to 158 MLGs (82.30 eMLGs), with all 158 MLGs unique to the non-pathogenic isolates. Diversity, measured as the Simpson’s diversity index (1 - λ), was greater for the non-pathogenic isolates (0.993) than the pathogenic isolates (0.628). Therefore, there appears to be an inverse correlation between pathogenicity to spinach and genetic diversity, i.e., populations of S. vesicarium on seed comprised mostly of pathogenic isolates tended to be less diverse genetically. MLG.167, the most frequent MLG detected in the original 400 S. vesicarium seed isolates, was also the most frequent MLG in the 248 isolates tested for pathogenicity, present in 47 of the isolates from six seed lots: 1 isolate from DK1, 9 from DK2, 15 from FR1, 1 from NZ1, 15 from USA1, and 6 from USA2, with all of these isolates pathogenic to spinach (Fig. 1B). Similarly, MLG.155, the second most frequent MLG among the 400 seedborne isolates, was also the second most frequent MLG detected in the 248 isolates tested for pathogenicity, and occurred in 19 of the isolates from five seed lots: 1 from DK2, 2 from FR1, 2 from NZ1, 5 from USA1, and 9 from USA2, all of which were pathogenic to spinach (Fig. 1B).

The DAPC of the 40 isolates of S. vesicarium isolates from each of the 10 spinach seed lots, did not differentiate the 10 seed populations by seed lot or country of origin (Fig. 2A). Similarly, the minimum spanning network of Bruvo’s distance for the non-clone corrected data of S. vesicarium from the 10 seed lots illustrated no clear clustering of MLGs based on seed lot, as the MLGs making up each seed lot were dispersed across the network (Fig. 1A). The AMOVA corroborated results of the DAPC and minimum spanning network. When the country in which the seed lot was grown was used as a grouping factor for the clone-corrected data of the S. vesicarium isolates from 10 seed lots, a large proportion of the genetic variation existed within seed lots (97.74%, P = 0.001; Table 4). The effect of the country in which the seed lot was grown was not significant (1.22%, P = 0.122). The effect of the seed lot within country was significant at P = 0.055, but only comprised 1.04% of the genetic variation. Therefore, the AMOVA indicated there was insignificant local or regional structure of this worldwide population of seedborne isolates of S. vesicarium. This could be due to unrestricted gene flow among the major countries in which spinach seed is produced, likely as a result of frequent movement of seed that limits regional and local genetic differentiation of populations of S. vesicarium. Given the nature of spinach seed production, whereby stock seed produced in one location typically is moved to various countries where seed crops are produced, and the harvested seed is then moved to even more countries where processing and fresh market spinach are grown, it is not surprising there was no evidence of genetic differentiation of seedborne populations of S. vesicarium based on the country of seed production. It is also possible that a lack of genetic differentiation is indicative of spinach being a new host for S. vesicarium, as this disease was only first described in 2001 (Koike et al. 2001).

Table 4. Analysis of molecular variance (AMOVA) of clone corrected data for populations of Stemphylium vesicarium from 10 spinach seed lots using a worldwide seed population study, in which isolates were genotyped with nine simple sequence repeat (SSR) markers, with geographic location used as a grouping factor

^a   An analog of F_ST that estimates genetic differentiation based on covariance components (Excoffier et al. 1992). A larger value suggests greater genetic differentiation.

^b   Significance of Φ was tested using a randomization test of 999 permutations in the R package arlequin (Excoffier and Lischer 2010).

When a second DAPC was conducted in which clone-corrected S. vesicarium isolates were organized by the 15 groups of pathogenic or non-pathogenic isolates within seed lots, the first DAPC axis separated most of the pathogenic isolates within seed lots (from DK1, DK2, FR1, NZ1, USA1, and USA2) from the non-pathogenic isolates, although there was some overlap between the pathogenic and non-pathogenic groups (Fig. 2B). The second DAPC axis (vertical axis) separated the one pathogenic isolate in FR2 from the other pathogenic and non-pathogenic isolates from all 10 seed lots. The minimum spanning network of Bruvo’s distance for the non-clone-corrected populations indicated that the three MLGs that were most frequent in the pathogenic isolates within each seed lot were more closely related to non-pathogenic MLGs than to one another, as these pathogenic MLGs were distributed across the network (Fig. 1B). The pathogenic isolates comprising these three major MLGs were similar genetically to several pathogenic isolates comprising one or two other MLGs, suggesting that some pathogenic isolates may be closely related, but not all pathogenic isolates are closely related. Additionally, there was no clear clustering of the MLGs detected in non-pathogenic isolates from the same seed lot (Fig. 1B), similar to what was demonstrated in the first DAPC with the data set for the original 400 seedborne isolates of S. vesicarium from the 10 seed lots (Fig. 2A).

When pathogen vs. non-pathogen was used as a grouping factor in the AMOVA, most of the variation occurred within seed lots (89.73%, P = 0.001; Table 5). The effect of pathogen vs. non-pathogen was also significant (9.16%, P = 0.001), as was the effect of seed lot within pathogenic and non-pathogenic groups (1.11%, P = 0.006), but to a lesser degree than the variation within seed lots. This suggests that pathogenic vs. non-pathogenic status of isolates does impact the genetic variation among worldwide populations of S. vesicarium on spinach.

Table 5. Analysis of molecular variance (AMOVA) of clone corrected data for populations of Stemphylium vesicarium from 10 spinach seed lots for a worldwide seed population study using isolates tested for pathogenicity to spinach, with isolates genotyped using nine simple sequence repeat (SSR) markers, and pathogenic vs. non-pathogenic as a grouping factor

^a   An analog of F_ST that estimates genetic differentiation based on covariance components (Excoffier et al. 1992). A larger value suggests greater genetic differentiation.

^b   Significance of Φ was tested using a randomization test of 999 permutations in the R package arlequin (Excoffier and Lischer 2010).

Objective 3: Determine the relative significance of seedborne S. vesicarium as an inoculum source for outbreaks of Stemphylium leaf spot by comparing isolates of the fungus from leaves of baby leaf spinach crops symptomatic with the disease to isolates from seed lots used to plant these spinach crops.

Sixty-eight MLGs were observed among the 180 isolates of S. vesicarium (Table 3). The foliar populations had 6 (TXfoliar1) and 5 (TXfoliar2) MLGs, while the two seed populations had 30 (TXseed1) and 35 (TXseed2) MLGs. The number of MLGs unique to each population was far less for the foliar populations (2 for TXfoliar1 and 1 for TXfoliar2) compared to the seed populations (27 for TXseed1 and 34 for TXseed2). Diversity, based on the Simpson’s diversity index (1 - λ) was 0.536 and 0.221 for TXfoliar1 and TXfoliar2 populations, respectively, and was much greater for the seed populations at 0.959 and 0.968 for TXseed1 and TXseed2, respectively. The most frequent MLG was MLG.68, found in 76 isolates, including 32 isolates from TXfoliar1 and 44 from TXfoliar2. The next most frequent MLG was MLG.167, which was also the most frequent MLG detected in the worldwide seed population study. MLG.167 was found in 13 isolates from three populations in this foliar vs. seed population study (11 isolates from TXfoliar1, 1 from TXfoliar2, and 1 from TXseed1). The prevalence of MLG.167 in both the worldwide seed populations and the foliar populations of S. vesicarium suggests that this MLG may be common in both seed and foliar populations beyond the populations evaluated in these studies. Of the 68 MLGs, 53 (78%) were each detected in only one isolate, with all but one of the 53 MLGs detected in one of the two S. vesicarium populations from seed, exemplifying the high diversity of the seed populations compared to the foliar populations of S. vesicarium.

The pathogenic isolates belonged to 11 MLGs, while the 54 non-pathogenic isolates belonged to 52 MLGs, with only two MLGs shared between the pathogenic and non-pathogenic isolates. MLG.68, the most frequent MLG detected in the original 180 S. vesicarium isolates in the study, was also the most frequent MLG detected in the 110 isolates tested for pathogenicity, occurring in 32 isolates (57% of pathogenic isolates) from the foliar populations, 13 from TXfoliar1 and 19 in TXfoliar2, but not in any of the seed isolates (Fig. 3). Similarly, MLG.167, the second most frequent MLG among the original 180 isolates, was the second most frequent MLG in the 110 isolates tested for pathogenicity, occurring in 7 pathogenic isolates (13% of pathogenic isolates): 5 in TXfoliar1, 1 in TXfoliar2, and 1 in TXseed1.

The DAPC of the subset of 110 isolates tested for pathogenicity from the four populations in the seed vs. foliar population study clustered the pathogenic isolates of S. vesicarium together (Fig. 4). The first DAPC axis separated the two pathogenic foliar populations and the single pathogenic seed population (from TXseed1) from both non-pathogenic seed populations, with limited overlap of the pathogenic isolates with non-pathogenic isolates in TXseed1, and limited overlap of the two non-pathogenic populations. The second DAPC axis separated the non-pathogenic TXseed1 population from the other three populations. This suggests that the structure of the two pathogenic foliar populations was most similar, and these foliar populations shared some genetic similarities with the pathogenic isolates from TXseed1, and pathogenic isolates from TXseed1 had some genetic similarities to the non-pathogenic isolates from TXseed1. The non-pathogenic TXseed1 isolates shared some genetic similarities with all four populations.

The minimum spanning network of Bruvo’s distance indicated that the two MLGs frequently detected in the pathogenic isolates (MLG.68 and MLG.167) were closely related, along with four MLGs that had one or two pathogenic isolates each, while the remaining five MLGs present in the pathogenic isolates were distantly related (Fig. 3). Only three MLGs (MLG.23, MLG.32, and MLG.37) were shared by pathogenic isolates from both foliage and seed, and these MLGs were distantly related. These shared MLGs suggests that S. vesicarium isolates that cause Stemphylium leaf spot of spinach can originate from infected seed, as demonstrated previously for seedborne S. beticola (Hernandez-Perez 2005). The fact that few pathogenic isolates from seed were found is not surprising given that very little infected seed is needed for the disease to occur in fresh market crops under optimal conditions of high temperatures and humidity (du Toit and Ocamb 2023; Koike et al. 2001), especially given that baby leaf spinach crops are planted at 7 to 9 million seed/ha, with sequential plantings under sprinkler irrigation. Therefore, Stemphylium leaf spot could occur even at a seed transmission rate of 10^-6 seed planted (which would result in >1 infected spinach seedling/ha). Given the high diversity of isolates on seed, there are likely more MLGs than those detected from just 40 isolates per seed lot, some of which may be present in the leaf populations.

Based on the AMOVA to determine if pathogenicity of isolates of S. vesicarium to spinach differentiated the two populations from seed and the two populations from foliage, most of the variation among isolates was detected within each of the four populations (94.22% P = 0.013, Table 6). The effect of pathogenic vs. non-pathogenic groups was also significant (5.80%, P = 0.001), but the effect of population within pathogenic and non-pathogenic groups was not significant (-0.02%, P = 0.350). The significant effect of pathogenicity corroborates results of the DAPC for this study (Fig. 4), as well as results from the worldwide seed population study.

Table 6. Analysis of molecular variance (AMOVA) of clone corrected data for four spinach populations of Stemphylium vesicarium used in a seed vs. foliar population study and tested for pathogenicity to spinach, with isolates genotyped using nine simple sequence repeat (SSR) markers, and pathogenic vs. non-pathogenic groups used as the grouping factor

^a   An analog of F_ST that estimates genetic differentiation based on covariance components (Excoffier et al. 1992). A larger value suggests greater genetic differentiation.

^b   Significance of Φ was tested using a randomization test of 999 permutations in the R package arlequin (Excoffier and Lischer 2010).

References

Bruvo, R., Michiels, N. K., D’Souza, T. G., and Schulenburg, H. 2004. A simple method for calculation of microsatellite genotype distances irrespective of ploidy level. Mol. Ecol. 13:2101-2106.

du Toit, L. J., and Ocamb, C. M. 2023. Spinach (Spinacia oleracea) - Stemphylium leaf spot. In: Pacific Northwest Plant Disease Management Handbook [online]. J. W. Pscheidt and C. M. Ocamb, eds. Oregon State University, Corvallis. https://pnwhandbooks.org/plantdisease/host-disease/spinach-spinacia-oleracea-stemphylium-leaf-spot. Accessed 23 May 2023.

Excoffier, L., and Lischer, H. E. L. 2010. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux or Windows. Mol. Ecol. Resourc. 10:564-567.

Excoffier, L., Smouse, P. E., and Quattro, J. M. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genet. 131:479-491.

Heck, D. W., Hay, F., and Pethybridge, S. J. 2023. Enabling population biology studies of Stemphylium vesicarium from onion with microsatellites. Plant Dis In press.

Hernandez-Perez, P. 2005. Seed transmission of Stemphylium botryosum and Cladosporium variabile in spinach. Pp. 79-101, in: Management of seedborne Stemphylium botryosum and Cladosporium variabile causing leaf spot of spinach seed crops in Western Washington. M.S. thesis, Washington State University, Pullman.

Hernandez-Perez, P., and du Toit, L. J. 2006. Seedborne Cladosporium variabile and Stemphylium botryosum in spinach. Plant Dis. 90:137-145.

International Seed Federation. 2017. Methods for the detection of Verticillium dahliae on spinach seed. http://worldseed.org/wp-content/uploads/2017/07/Spinach_Verticillium_July2017.pdf

Jombart, T. 2008. Adegenet: a R package for the multivariate analysis of genetic markers. Bioinform. 24:1403-1405.

Jombart, T., Devillard, S., and Balloux, F. 2010. Discriminant analysis of principal components: a new method for the analysis of genetically structured populations. BMC Genet. 11:94.

Kamvar, Z. N., Tabima, J. F., and Grünwald, N. J. 2014. Poppr: an R package for genetic analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ 2:e281.

Koike, S. T., Henderson, D. M., and Butler, E. E. 2001. Leaf spot disease of spinach in California caused by Stemphylium botryosum. Plant Dis. 85:126-130.

Liu, B., Stein, L., Cochran, K., du Toit, L. J., Feng, C., Dhillon, B., and Correll, J. C. 2020. Characterization of lead spot pathogens from several spinach production areas in the United States. Plant Dis. 104:1994-2004.

Sorensen, L. H., Schneider, A. T., and Davis, J. R. 1991. Influence of sodium polygalacturonate sources and improved recovery of Verticillium spp. from soil. Phytopathology 81:1347. (Abstr.)

Spawton, K. A., Derie, M. L., Correll, J. C., Stein, L., Olaya, G., Raid, R. N., Sandoya, G. V., Peever, T. L., and du Toit, L. J. 2019. Characterization of Stemphylium spp. from spinach based on molecular data, host response, and azoxystrobin sensitivity. Phytopathology 109:S2.167. Poster presented at Plant Health 2019, Annual Meeting of American Phytopathological Society, 3-7 Aug. 2019, Cleveland, OH.

Woudenberg, J. H. C., Hanse, B., van Leeuwen, G. C. M., Groenewald, J. Z., and Crous, P. W. 2017. Stemphylium revisited. Studies Mycol. 87:77-103.

Zhou, Y., Shi, Y., Xie, X., Guo, Y., and Li, B. 2011. Leaf spot of spinach caused by Stemphylium spinaciae sp. nov. Mycostema 30:379-383.