From poor information to a representative survey for the German territory
A recent survey in the states of Brandenburg and Berlin identified an unexpected high infection rate of U. laevis by Ca. P. ulmi [26]. This small survey reflected overall poor knowledge on the distribution of the pathogen in Germany and prompted a nationwide survey in which ca. 6,500 samples were collected at 339 sites. The results of this survey can be regarded as representative for the German territory, as most of the natural habitats of the three elm species were covered. Preliminary results of this survey have been published as a contribution for the IPWG conference in 2019 [35].
At the beginning of the survey, no information was available on the colonisation of Ca. P. ulmi in infected elm trees. To obtain meaningful real-time PCR results for a large number of samples, the most suited ‘plant-part-to-collect’ had to be determined. In the well-studied phytoplasma diseases of pome fruit trees, the phytoplasma titre displays strong seasonal fluctuations, reaching a low in late winter to early summer [36, 37]. The examination of an U. glabra tree, however, quickly revealed that Ca. P. ulmi colonises roots and shoots all year round in a constantly high titre. Therefore, branches or trunk material were sampled throughout the survey. Even though petioles or leaf midribs would have been easier to collect, the choice fell on branches or trunk material, as the sampling period could be extended with this material.
Spacer-specific real-time assay enables reliable detection of Ca. P. ulmi infection
The real-time PCR assays corroborated the results of the seasonal course study, as in most trees a fairly high number of Ca. P. ulmi was identified. More than 86% and 72% of the positive samples showed Ct values lower than 28, with the universal phytoplasma or specific Ca. P. ulmi spacer-based real-time PCR assay, respectively, representing an organism titre of 1006 per gram of phloem tissue and higher. Both real-time PCR assays worked reliably, although the species-specific assay showed a slightly lower performance in respect to the Ct values. This was most likely caused by the lower G+C content of the primers and probe and the reduced binding strength compared to the oligonucleotides of the universal phytoplasma assay. The substitution of adenine bases with 2,6-diaminopurin, to increase the melting temperature of the TaqMan spacer probe, did not change performance significantly compared to the assay with the non-modified probe. Nevertheless, the spacer assay proved to be a reliable diagnostic tool for the species-specific DNA amplification of Ca. P. ulmi strains in this study.
Occurrence of Ca. P. ulmi in Germany in comparison to the situation in other European countries
It is evident that Ca. P. ulmi was present in almost 28% of the elm samples, thus demonstrating the high number of infected trees in German elm populations. A higher overall Ca. P. ulmi infection rate of 46% was found in Croatian elm populations [38]. However, if the German survey would had focused only on phytoplasma hotspots, an even higher infection rate would have resulted, as 77 of the 339 sites examines showed infection rates over 66.7%. Therefore, the high overall incidence rate in Croatia might simply be a matter of sample size. In other studies, a preferential collection of material from symptomatic trees has been executed and infection rates of 85% and higher have been found, which also might not reflect the real infection rate [39]. The three elm species showed different disease rates. While almost one-third of the U. laevis samples tested Ca. P. ulmi-positive, U. glabra and U. minor were infected to a lesser degree. This difference is not due to sample size, as the number of tested accessions from each species was about similar. In Croatia, different elm species infection rates were also found, with almost 75% of infected U. laevis, followed by U. minor (10.8%) and U. glabra (4%) [38]. However, in this work, the latter two species were also infected by other phytoplasmas, and therefore the infection rates are not directly comparable.
Occasional detection of other phytoplasmas in elm
In all, except for two infected elm samples, Ca. P. ulmi was identified. These phytoplasmas were only identified by the different specificity of the two real-time PCR assays employed. Sequence analyses of ribosomal fragments, however, revealed that both phytoplasmas were related to Ca. P. ulmi and belong to the elm yellows or 16SrV group. Phytoplasmas with other taxonomic affiliations, such as Candidatus Phytoplasma solani and Candidatus Phytoplasma asteris, have been described in U. glabra and U. minor displaying rather unspecific symptoms of leaf yellowing and drying [38]. These symptoms could not be observed in the two phytoplasma-infected trees. The presence of other phytoplasmas in elms is most likely caused by an occasional feeding of infected phytoplasma insect vectors. However, the fact that only 16SrV-group phytoplasmas were found in elm trees might indicate a certain host-pathogen group specificity.
Rare symptom formation indicates tolerance in a long-living host
Despite the high number of phytoplasmas in the sieve tubes, German elms seem to react in a tolerant way upon infection, which stands in striking contrast to the reactions of American and Asian elm species [12, 16]. The few symptomatic U. glabra and U. minor trees that were found, however, displayed typical Ca. P. ulmi symptoms with witches’ brooms, stunting and leaf chlorosis. Mittempergher (2000) also concluded after extended observations in Italian elm breeding stations that European elm species tolerate a Ca. P. ulmi infection quite well, although the number of symptom sightings from southern Europe [38-40] for U. minor and U. laevis seem to be more frequent compared to reports from central and northern regions of the continent. However, this could also be caused by a stronger awareness of plant pathologists or breeders, regional differences of Ca. P. ulmi strain virulence, the genetic background of elm populations or other undetermined stress factors.
The different infection rates of the elm species and within the age categories are difficult to explain with our present knowledge. The progressive infection in aging U. glabra populations is easily comprehensible, but the plateau phase for U. laevis, and the constant infection rate of U. minor age classes, is not as easy to deduce. Of particular interest is the situation of old and monumental trees of more than 100 cm in trunk diameter. This category was represented by 124 European white elms, 29 of which were phytoplasma-infected, and five field elms, with one phytoplasma-infected individual. How some of these old trees escaped phytoplasma infection remains unclear, as they were well located in regions of high infection pressure. Therefore, it seems that a certain degree of resistance is present in the populations. However, other factors might be involved too, like the protective role of the plants’ microbiome [41], albeit insufficient information is available to assess its influence so far.
Unequal distribution of Ca. P. ulmi might be linked to vectors
Infection hotspots were located in the eastern, central and south-western parts of Germany, whereas infected sites became rare towards the north and north-western regions. The most plausible explanation is that an insect vector migrated from the southern-to-eastern side into the territory moving towards the north and west. Hence, the distribution map given in Fig. 3B might simply reflect a snapshot. Beside the verified Ca. P. ulmi vector Macropsis mendax for Italy [42], the phytoplasma has only been identified by PCR in Hyalesthes luteipes in Serbia, but no transmission experiment was performed [39]. For Germany, no data are available. The fact that infected trees were found from sea level up to an altitude of 750 m might indicate a vector different to M. mendax, as this insect is only known to occur up to an altitude of 400 m [43]. A spread of the bacterium through the exchange or trade of infected elms for planting purposes can be excluded, as elm timber has little forest use. In addition, root bridges can be discounted, as this would only explain spread in a small area.
Ca. P. ulmi infection is common in European elm stands
This survey demonstrated a nationwide distribution of Ca. P. ulmi in all three native elm species. The number of infected individuals is such that the eradication of this quarantine pathogen is impossible. In 2017, the EPPO realised the more general occurrence of Ca. P. ulmi in its member states and moved the pathogen from Annex I/A 1 to Annex I/A2, recognising its regional presence. However, even this categorisation does not seem to be justified in light of the general distribution demonstrated in this survey. The recent finding of Ca. P. ulmi in Belgium [23] demonstrates its presence also west of Germany, although the confirmed numbers of infected elms are still low. The new EU Council Directive 2016/2031, effective from mid-December 2019, will deprive Ca. P. ulmi of its quarantine status in member states of continental Europe and implement a legal change supported by the findings of this survey and also from other recent surveys.
Regardless of the legal status of the disease, this work has established a sound basis for future research with respect to transmission and phytoplasma-host interaction. Examples of tolerance are often overlooked by phytopathologists and are rarely reported. It is remarkable that the closely related elm yellows and alder yellows phytoplasmas share this feature [44], which might be the result of a long-term co-evolution of these phytoplasmas with their hosts . A deeper understanding of such a particular phytoplasma-host interaction may provide the key to developing new strategies to cope with phytoplasmosis in agriculture.