1. Introduction
Soybean (
Glycine max (L.) Merr.) is a crop species of enormous importance. It is a good source of both fat and protein and is used for human consumption either fresh or after processing as salad oil, margarine, mayonnaise, tofu, or, recently, also as milk, yogurt, and cheese. The meal fraction from oil production is used as high-protein animal feed for pigs, poultry, and cattle. Together, these make soybean almost indispensable for human nutrition directly or indirectly and, as a legume, soybean can also fix elemental nitrogen, making it very useful in crop rotations, improving soil fertility. Soybean cultivation is concentrated in the Americas and East Asia. The total soybean area in Europe is currently estimated at approximately 5.5 million hectares [
1], and within the European Union (EU) it has recently exceeded 1 million hectares, with continued growth expected, in part because of increasing demand for locally produced non-GMA legumes and government support aimed at making Europe independent from protein imports.
Species of genus
Diaporthe, also known as the
Diaporthe/Phomopsis Complex [
2], can cause several different diseases on soybeans. These are pod and stem blight, seed decay, stem canker, and seedling blight [
3]. Together, these diseases are responsible for substantial yield losses worldwide, surpassing the impact of any other pathogen affecting soybeans. The pathogens can be controlled by cultural practices like residue management, extended crop rotation, testing seed against
Diaporthe spp., and by fungicide application [
3]. However, the most effective control strategy is planting resistant soybean cultivars. Therefore, the identification of cultivars resistant to the different species or different strains of
Diaporthe is highly relevant. These cultivars can either be prioritized in cultivation where
Diaporthe is a problem or used as a source for resistance in breeding programs.
In Europe,
Diaporthe longicolla (Hobbs) J.M. Santos, Vrandečić & A.J.L. Phillips and
D. caulivora (Athow & Caldwell) J.M. Santos, Vrandečić & A.J.L. Phillips are considered the most important pathogens of soybeans within the genus
Diaporthe.
D. longicolla is primarily associated with seed decay but can also cause pod and stem blight, often together with
D. sojae [
4].
D. caulivora is the causal agent of northern stem canker and is also capable of infecting seeds [
5,
6,
7,
8].
D. aspalathi E. Jansen, Castl. & Crous, causes southern stem canker [
9] and, together with
D. caulivora, is the only species against which systematic resistance breeding has already been performed [
10].
D. aspalathi seems to be restricted regionally and has not yet been detected in Europe. Other
Diaporthe species associated with soybean and capable of infecting and causing disease include:
D. aseana Dissan., Tangthir. & K.D. Hyde (syn
D. tectonigena Doilom, Dissan. & K.D. Hyde),
D. bacilloides K. Petrović, Skaltsas & F.M. Mathew,
D. eres Nitschke,
D. flavescens K. Petrović, Skaltsas & F.M. Mathew,
D. foeniculina (Sacc.) Udayanga & Castl.,
D. goulteri R.G. Shivas, S.M. Thomps. & Y.P. Tan,
D. gulyae R.G. Shivas, S.M. Thompson & A.J. Young,
D. insulistroma K. Petrović, Skaltsas & F.M. Mathew,
D. kongii R.G. Shivas, S.M. Thompson & A.J. Young,
D. masirevicii R.G. Shivas, L. Morin, S.M. Thomps. & Y.P. Tan,
D. novem J.M. Santos, Vrandečić & A.J.L. Phillips (syn.
D. pseudolongicolla K. Petrović, L. Riccioni & M. Vidić) (The species is treated here as
Diaporthe novem, following current usage in Index Fungorum [
11] and Species Fungorum [
12]. However, an alternative interpretation recognizing the name
D. pseudolongicolla has been proposed by Petrović et al. [
13].),
D. rudis (Fr.) Nitschke,
D. ueckeri Udayanga & Castl. (syn.
D. ueckerae Udayanga & Castl.,
D. miriciae R.G. Shivas, S.M. Thomps. & Y.P. Tan), and
D. unshiuensis F. Huang, K.D. Hyde & Hong Y. Li [
3,
14,
15,
16,
17].
D. goulteri was only very recently isolated from soybeans for the first time [
18]. So far, not all species of
Diaporthe are present in all soybean-producing regions, but they are spreading further.
The taxonomy of the genus
Diaporthe is complex. In recent multilocus phylogenetic studies of the genus, many species names were found to be synonymous, and species boundaries were redefined [
19,
20]. Many species can infect more than one host and can either behave as pathogens or endophytes, depending on the host and weather conditions. The interactions of different
Diaporthe species with soybean also widely vary, as can be deduced from the different diseases that are caused. For example,
D. longicolla can cause pod and stem blight, black zone lines, and seed decay. This is accompanied by the observation that in different conditions, plants may either be killed by seedling blight, stem blight, or stem canker, or appear healthy throughout the growing season and only after the onset of senescence, pycnidia and perithecia appear on the plant surface and seeds may contain
Diaporthe mycelium [
8]. These late symptoms might be due to late infection, but under some conditions, the species behave like endophytes during plant growth and only switch their lifestyle at the onset of plant senescence. Neither possibility can be excluded at present. These things are still not entirely clear, even for the most important pathogens
D. caulivora, D. aspalathi, and
D. longicolla, for which genome sequencing has been performed already [
9,
21,
22,
23,
24,
25], and even less is known about the other species associated with soybeans. It is not entirely clear how they penetrate the plant, how they spread through the plant and how, when, and why sporulation is induced.
Given the taxonomic complexity and partially overlapping symptomology of these pathogens, diagnostic procedures were, until recently, limited to identifying the genus
Diaporthe. This has changed recently. Now, species-specific molecular diagnosis based on real-time PCR is available for several species [
10,
18,
26,
27]. Likewise, studies are performed to differentiate the species or isolates of the species based on their virulence against different soybean cultivars or accessions [
8,
14,
28,
29,
30]. For this, methods have been developed to inoculate soybean plants with
Diaporthe and time schedules and grading scales for assessing disease severity have been optimized. With a view to obtaining results fast and with a minimal use of resources, assays have been developed where germinating seeds are immersed in a spore suspension [
16]. Alternatively, in the greenhouse, seedlings are inoculated on the stem. Here, either the stem can be pierced with a toothpick overgrown with fungal mycelium, or an agar block with
Diaporthe mycelium is applied to the stem that can also be wounded in different ways [
28]. For the latter methods, humidity at the inoculation site and in the greenhouse is critical, since it has a strong influence on whether there is successful infection and can determine the nature of the symptoms that develop after infection.
Building upon these recent advances, the present study evaluates the susceptibility of different early maturing European soybean cultivars against different Diaporthe species. We inoculated the eight different soybean cultivars Magnolia, Salsa, Nessie, Tofina, Kofu, Orakel, Selena, and Lela with the four species D. longicolla, D. caulivora, D. novem (syn. D. pseudolongicolla), and D. goulteri. The susceptibility or resistance and biochemical responses of the cultivars and the virulence of the pathogens were assessed both visually and by measuring biochemical parameters of the infected plants.
2. Results
When the experiment was performed first, the eight soybean cultivars Magnolia (MG 00), Salsa (MG 00), Nessie (MG 0), Tofina (MG 00-0), Kofu (MG 000-00), Orakel (MG 0), Selena (MG 0), and Lela (MG 0) were inoculated with one isolate each of D. longicolla, D. caulivora, D. novem (syn. D. pseudolongicolla), and D. goulteri. The two inoculation methods Stem Cut and Stem Wound were used. In this initial experiment both inoculation methods produced reliable infections under conditions of high humidity. Clear differences in disease severity of the different species × cultivar combinations could be observed. To confirm the results, a repeat experiment was performed. Since the differences in the biochemical parameters appeared to be more pronounced for the Stem Wound method, this method was selected for the repeat experiment and the selection of cultivars was reduced to the five cultivars Kofu, Salsa, Tofina, Magnolia (German cv.), and Selena (Serbian cv.) that had shown the biggest differences in the first experiment. The Stem Cut method was not repeated because, while it yielded similar results, these data could not be integrated into the evaluation of the Stem Wound data.
2.1. Visual Evaluation of Disease Severity
Table 1 shows disease ranking and relative treatment effect for eight different cultivars and four different
Diaporthe species in the experiment using the Stem Cut method.
The RTE was highest for Salsa infected with D. caulivora, significantly higher than all other combinations except for Selena and Nessie, also infected with D. caulivora. From all combinations a tendency could be observed for severe infections by D. caulivora and D. longicolla, while D. goulteri and D. novem seem to be less aggressive. Apparently, the species generally considered the major pathogens from the genus Diaporthe are also the most aggressive. However, there are exceptions, for example, in cv. Nessie, D. goulteri caused an RTE similar to that of D. longicolla and even D. caulivora. Since the lowest RTE was shared by several combinations, a most resistant cultivar could not be determined. However, cv. Lela had the lowest RTE for three of the pathogen species and only showed medium RTE when infected with D. caulivora. Generally, cv. Magnolia also exhibited lower RTEs. Cultivar Kofu displays an interesting pattern, since it showed moderate but similar RTEs. Seemingly, there is a low level of resistance present against all species in this cultivar. That the rankings differ between Diaporthe species and cultivars indicates that there are specific interactions.
Table 2 shows disease ranking and relative treatment effect for five different cultivars and four different
Diaporthe species in two experiments using the Stem Wound method.
The highest disease severity based on RTE was found for the combinations Salsa × D. caulivora (0.79), Tofina × D. caulivora (0.77), and Selena × D. longicolla (0.72). The infections of Selena × D. novem (0.29), Magnolia × D. goulteri (0.26), and Kofu × D. goulteri (0.21) were least severe. The homogeneity of variances between the two experiments was satisfied using the Fligner–Killeen test (p = 0.30), so there was no objection against evaluating the repeats together. In addition, the trends seen with the Stem Cut method were the same as those seen with the Stem Wound method, indicating that the results are stable and reliable.
Figure 1 shows representative pictures with disease development for all cultivars inoculated with all species.
The data were also analyzed for differences in aggressiveness of the different
Diaporthe species across all cultivars (
Table 3).
Overall, D. caulivora was most aggressive, followed by D. longicolla. D. novem was much less aggressive, but still significantly more aggressive than D. goulteri. This nicely aligns with the findings for the specific pathogen on cultivar results, where D. caulivora was most aggressive on three cultivars (Magnolia, Salsa and Tofina), while D. caulivora and D. longicolla were equally most aggressive on two cultivars (Tofu and Selena). Only for one cultivar (Magnolia), D. novem and D. longicolla showed similar aggressiveness. D. goulteri was least aggressive on all five cultivars, whereas D. novem was least aggressive on all cultivars except Kofu.
To identify the most resistant cultivar, disease severity was analyzed across all species of
Diaporthe (
Table 4).
In this experiment, cv. Magnolia exhibited the lowest overall disease severity, indicating that this cultivar possesses enhanced resistance against or tolerance towards Diaporthe species. Differences between the other four cultivars were small and not statistically significant. When resistance against the individual species is considered, Magnolia showed intermediate values for D. caulivora and had the lowest disease severity for D. longicolla, D. goulteri, and D. novem.
Because of the different rating scales, the symptoms resulting from the two inoculation methods could not be evaluated together, but, nevertheless, they can be compared. Cultivars Lela, Nessie, and Orakel were only tested in the Stem Cut assay, so these were excluded from the comparison. In both assays, the combination Salsa × D. caulivora showed the most severe symptoms. Kofu × D. goulteri, which showed the mildest symptoms in the Stem Wound assay, displayed higher severity in the Stem Cut assay but was still among the lower-ranked combinations. Although differences between the two assays were observed, the overall patterns were comparable. These differences probably reflect normal biological variation rather than methodological inconsistency. Taken together, these results are clear and robust.
2.2. Biochemical Responses of Infected Soybean Plants
The visual assessment of symptoms was complemented by measuring biochemical responses in infected plants. The two inoculation methods were evaluated separately. Because the experiment was only repeated for the Stem Wound method, data for the Stem Cut assay are not presented. The two experimental repeats were evaluated independently and are presented separately in
Figure 2. Fewer cultivars were included in the second repetition than in the first. The results of the first repetition are shown on the left side of the figure, those of the second repetition on the right.
There was substantial variation between the two repetitions of the experiment. The relative differences between the values were inconsistent between experiments. An example is the values for SOD activity for cv. Selena. In the first repeat, SOD activity was relatively low in all plants, but in the second repeat, it was higher for all plants and considerably higher than for all others in plants infected with D. longicolla. Similar inconsistencies were observed for all the parameters studied.
Searching for correlations between disease severity and the biochemical parameters, it seems a good option to compare, in general, Magnolia, the most resistant cultivar, against the other cultivars, and especially Salsa, which appeared to be the most susceptible. Here, it appears that over both experimental repeats, Magnolia has slightly elevated levels of total phenolics and flavonoids. A significant increase in these compounds could not be observed in both repeats, however. This might indicate that phenolics and flavonoids inhibit Diaporthe but are not necessarily induced as a response to the infection.
Differentiating between the pathogens, where D. caulivora and D. longicolla proved most aggressive, the induction of SOD in Magnolia by these two pathogens in the first repeat is most striking. Though much less pronounced, this induction is also visible in the second repeat and can also be observed in the second repeat for Salsa, Selena and (× D. longicolla) in Kofu. Since SOD is activated as a response against oxidative stress, it might be assumed that Magnolia shows reduced disease symptoms because it is best able to deal with oxidative stress. Overall, high levels of SOD coincide with high RTSs, however. So, high SOD primarily is an indicator of plant stress.
MDA accumulation is an indicator of damage to cell membranes. At least partially, accumulation of MDA also coincides with disease severity. This shows that severe infections with Diaporthe on soybean also lead to increased cell damage.
3. Discussion
Our study could show clear and significant differences in virulence or aggressiveness between the different tested
Diaporthe species based on visually observed symptom severity. These findings are consistent with earlier reports documenting substantial variation in aggressiveness among
Diaporthe species infecting soybean [
5,
6,
7,
8,
10,
24,
25,
31]. In both inoculation assays,
D. caulivora consistently ranked as the most aggressive, followed by
D. longicolla, whereas
D. novem (syn.
D. pseudolongicolla) and
D. goulteri generally induced only mild to moderate symptoms. This confirms that the species previously associated with major stem and seed decay symptoms (
D. caulivora and
D. longicolla) indeed pose the highest pathogenic risk.
Overall, cv. Magnolia displayed lower symptom severity than the other cultivars tested. This was most pronounced for D. longicolla. This seems to indicate that there is general resistance against all four species, but stronger resistance, that may be species-specific, against D. longicolla. In practical terms, this means that cv. Magnolia can be recommended to farmers when there are problems with Diaporthe, especially when there is a high incidence of D. longicolla. Possibly, the specific resistance against D. longicolla can be used for breeding purposes. In the Stem Cut assay, also cv. Lela performed well. Here, also, the low RTE for D. longicolla is conspicuous. Since these results are only based on a single experiment, they are not as solid as those for cv. Magnolia; nevertheless, the same conclusions as for cv. Magnolia might be drawn for cv. Lela.
Our approach, however, has several inherent limitations. First, little is known about how the fungi infect and spread in the plant. Therefore, it is difficult to determine how well artificial inoculation methods represent natural disease progression. To be able to test several combinations of pathogen and cultivar, it is necessary to use a method that requires a minimum amount of work. This is true for the methods using inoculation of soybean plants at the stem. Based on earlier studies comparing these methods [
28,
32], the Stem Cut and Stem Wound methods were chosen as the most suitable.
A second limitation is the use of only one isolate per species. It is possible that different isolates would have shown different degrees of aggressiveness [
30]; therefore, it is not entirely clear whether the differences we observed are general differences between the species or just between the specific isolates. Indeed, scoring plants three months and later after planting and inoculation, Hosseini et al. [
7] found differences between
D. caulivora,
D. longicolla, and
D. novem isolates. Nevertheless, in this instance, the scale of this experiment required prioritizing inter-species comparisons. Given greenhouse space and the need to inoculate all plants within a short timeframe to maintain uniform humidity, including multiple isolates was not feasible. For
D. goulteri, only one isolate was available. The assumption that inter-species differences exceed intra-species variation appears supported by our results; still, future studies using multiple isolates per species would refine species-level aggressiveness profiles. Other studies have compared different isolates of the same
Diaporthe species [
30]. These studies were especially interested in finding differences in the virulence of isolates from different regions. To determine how much of the differences in virulence between different
Diaporthe sp. isolates are due to different traits of the species and how much are due to interspecies variability, studies would be necessary, encompassing all strains of all tested species while comparing at least two species. This, of course, is impossible; any study can only include a small selection of the available isolates, and these may or may not represent a large part of the variability of a species. As more datasets using comparable procedures become available, meta-analyses may offer deeper insights into pathogenic variability in the
Diaporthe complex.
The need for testing multiple isolates is also illustrated by a discrepancy of our results for the virulence of
D. novem (syn.
D. pseudolongicolla) with findings from Petrović et al. [
13]. In contrast to the relatively low RTE indicating little aggressiveness, the
D. pseudolongicolla isolate of [
13] caused extensive pathogenicity on soybean stems using the Stem Wound method. The four isolates tested by Hosseini et al. [
7] differed in their disease score. The more recent isolate DPC_HOH41 used in this study is also different. What also should be mentioned here is that, in contrast to
D. longicolla, where ITS,
TEF1 and
TUB sequences used for molecular phylogenetic analysis are 100% identical for all isolates, isolates of
D. novem differ in these marker sequences [
7]. Indeed, DPC_HOH41 is also slightly different in those sequences both to the isolates by Hosseini et al. [
7] and Petrović et al. [
13]. So, it can be speculated that
D. novem is a species with especially high intraspecific variation.
High MDA levels as a marker for lipid peroxidation and cell membrane damage indicate that the plants are suffering from stress [
31]. In their study, Petrović et al. [
31] chose an inoculation method they termed “mycelium contact” because, based on MDA levels, this method caused the lowest level of stress. The Stem Wound method used here is most similar to what is labeled as the “plug method” by Petrović et al. [
31]. Presumably because of the wounding of the stem that is performed in this procedure, the plants suffer elevated levels of stress from the inoculation procedure. This aligns with our results: in all cultivars except Salsa, MDA levels in control plants are comparable or higher than those in the plants infected with
D. goulteri or
D. novem.
Only in cultivars Selena, Kofu, and Salsa, MDA levels in those plants infected with D. longicolla rose significantly above the control values. It seems that, though Diaporthe infections do cause oxidative stress, cultivars like Salsa that seem to be able to efficiently deal with this kind of stress are not consequently also resistant against the pathogens.
In their study, Petrović et al. [
31] studied the activities of superoxide dismutase (SOD), catalase, and peroxidase and the glutathione content, in addition to MDA. The results for these other parameters could not be correlated with tolerance, as well as the MDA level. Therefore, in this study, we only studied SOD again and came to the same conclusion, that SOD is primarily a marker of oxidative state rather than a predictor for host resistance.
Instead of enzyme activities, we studied total phenolics and total flavonoids. Among the phenolics, and especially the flavonoids, are many substances with fungicidal or fungistatic properties. These are either permanently present in plants, contributing to pre-formed resistance as phytoanticipins, or are produced upon induction of plant defenses as phytoalexins. Therefore, elevated levels of phenolics and flavonoids can indicate resistance against fungal pathogens. Since we found Magnolia to be most resistant and Salsa to be most susceptible, it might be expected that Magnolia has comparably high levels of phenolics and flavonoids, and Salsa has low levels. This could be observed, but to a very small extent and not consistently. While in the first repeat, the levels were slightly higher for Magnolia than for Salsa, they were very similar for the two cultivars in the second repeat of the experiment. Moreover, Salsa × D. caulivora, with the highest RTE, had unexpectedly high flavonoids accumulation. Thus, it must be concluded that either phenolics and flavonoids do not play a major role in defense against Diaporthe, or that certain Diaporthe species can suppress or bypass those defense phenolics- and flavonoids-mediated defense pathways, preventing these metabolites from acting as reliable resistance indicators.
Taken together, the biochemical markers measured here did not provide reliable predictors of resistance or susceptibility. Visual disease severity remained the most consistent and biologically meaningful indicator of Diaporthe aggressiveness and soybean cultivar response.
These considerations highlight the complexity of Diaporthe–soybean interactions and emphasize the need for multi-isolate, multi-cultivar studies using standardized inoculation procedures to better resolve species- and genotype-specific pathogenic behaviors.