Diversity of Pectobacteriaceae Species in Potato Growing Regions in Northern Morocco

Dickeya and Pectobacterium pathogens are causative agents of several diseases that affect many crops worldwide. This work investigated the species diversity of these pathogens in Morocco, where Dickeya pathogens have only been isolated from potato fields recently. To this end, samplings were conducted in three major potato growing areas over a three-year period (2015–2017). Pathogens were characterized by sequence determination of both the gapA gene marker and genomes using Illumina and Oxford Nanopore technologies. We isolated 119 pathogens belonging to P. versatile (19%), P. carotovorum (3%), P. polaris (5%), P. brasiliense (56%) and D. dianthicola (17%). Their taxonomic assignation was confirmed by draft genome analyses of 10 representative strains of the collected species. D. dianthicola were isolated from a unique area where a wide species diversity of pectinolytic pathogens was observed. In tuber rotting assays, D. dianthicola isolates were more aggressive than Pectobacterium isolates. The complete genome sequence of D. dianthicola LAR.16.03.LID was obtained and compared with other D. dianthicola genomes from public databases. Overall, this study highlighted the ecological context from which some Dickeya and Pectobacterium species emerged in Morocco, and reported the first complete genome of a D. dianthicola strain isolated in Morocco that will be suitable for further epidemiological studies.


Introduction
Pectinolytic Pectobacterium and Dickeya spp. are causative agents of severe diseases in a wide range of plants of high economic value [1,2]. On potato tubers and stems, the diseases caused by pectinolytic pathogens are soft rot and blackleg, respectively. These pathogens produce a large set of extracellular enzymes that degrade the plant cell wall, resulting in plant tissue decay and maceration. This rotting process causes losses in the production of potato tubers sold both on the food market and as certified seed tubers [3]. The pathogens may be acquired by the host plants from soil and/or from contaminated seed tubers [4]. On plants, pathogen populations remain at a low level in asymptomatic

Sampling and Isolation of Pectinolytic Bacteria
In 2015, 2016 and 2017, blackleg symptoms were searched for in potato fields in four regions (Meknes, Guigo, Boumia and Larache) in Northern Morocco. Pectinolytic bacteria were isolated from symptomatic plant tissues using crystal violet pectate (CVP) medium as described previously [45]. The CVP plates were incubated at 28 • C for 3 days and colonies that had formed pits were re-streaked onto Tryptone (5 g/L) yeast extract (3 g/L) agar medium (TY). The purified isolates were spotted again on CVP to confirm the pectinolytic activity. The obtained cultures from single colonies were stored in 25% glycerol at −80 • C.

Molecular Characterization of Pectobacterium and Dickeya Isolates
The primer couples Y1/Y2 and ADE1/ADE2 (Table S1) were used for the identification of isolates belonging to Pectobacterium and Dickeya genera [46,47]. The reaction was carried out in a final volume of 25 µL, containing 1 µL of bacterial DNA (50 ng/µL), 2.5 µL of PCR buffer (10×), 2 µL of MgCl2 (25 mM), 2.5 µL of dNTPs (1 mM), 1U Taq polymerase and 1 µL of each primer (1 µM) and water. The temperature settings for PCR were the same as described before [46,47]. The analysis of PCR products was done by electrophoresis on 2% agarose gels, using PCR products of P. atrosepticum CFBP1526 T and D. solani IPO2222 T as control along with the 1 Kb DNA ladder.
Positive strains for either Y1/Y2 or ADE1/ADE2 PCR were further characterized using the gapA barcode procedure [37]. All the gapA PCR products obtained with gapA-F/gapA-R primers (Table S1) were sequenced using Sanger technology (GATC Biotech, Konstanz, Germany). The sequences were trimmed using the CLC genomic workbench (V10.1.1, Aarhus, Denmark) and aligned using ClustalW. The phylogenetic analysis of the gapA gene was performed as follows: the evolutionary distances were computed using the maximum composite likelihood method (Mega7 software, Pennsylvania State University Park, PA, USA) with 1000 bootstrap. The obtained sequences were deposited in GenBank (Table S2).

Genome Sequencing
A total of 10 strains representing 5 species identified with gapA sequencing were selected for genome sequencing ( Table 1). DNA of the 10 isolates (listed in Table 1) was extracted from overnight cultures in TY medium using the MasterPure™ Complete DNA and RNA Purification Kit (Epicenter, Madison, WI, USA) followed by an ethanol precipitation. The quantity and quality control of the DNA was completed using a NanoDrop (Wilmington, DE, USA) device and 1.0% agarose gel electrophoresis.
Paired-end libraries (500 bp in insert size) were constructed for each strain, and DNA sequencing was performed by Illumina NextSeq technology. Sequencing of the library was carried out using the 2 × 75 bp paired-end read module. Illumina sequencing was performed at the I2BC sequencing platform (Gif-sur-Yvette, France).
In the case of the D. dianthicola strain LAR. 16.03.LID, Nanopore sequencing was also performed. Library preparation and sequencing were performed at the GeT-PlaGe core facility, INRA Toulouse, using the "1D Native barcoding genomic DNA kit (EXP-NBD103 and SQK-LSK108)", according to the manufacturer's instructions. At each step, DNA was quantified using the Qubit dsDNA HS Assay Kit (Life Technologies). DNA purity was tested using a NanoDrop device (Thermofisher, Waltham, MA, USA) and size distribution and degradation was assessed using the fragment analyzer (AATI) High Sensitivity DNA Fragment Analysis Kit. Purification steps were performed using AMPure XP beads (Beckman Coulter). Quantities of 5 µg of each DNA (five samples) were purified then sheared at 20 kb using the Megaruptor1 system (Diagenode, Seraing, Belgium). A DNA damage repair step was performed on 3 µg of sample. Then END-repair and dA-tailing of double stranded DNA fragments were performed on 1 µg of each sample. Then, a specific index was ligated to each sample. The library was generated by an equimolar pooling of these barcoded samples. Then adapters were ligated to the library. The library was loaded on a R9.4.1 flowcell and sequenced on MinION instrument at 0.15 pmol within 48 h.

Genome Assembly
Assembly of the Illumina reads was performed using the CLC Genomics Workbench v10.1.1 software (CLCInc, Aarhus, Denmark). After quality (quality score threshold 0.05) and length (above 40 nucleotides) trimming of the reads, contigs were generated by de novo assembly (CLC parameters: automatic determination of the word and bubble sizes with no scaffolding). The draft genome sequences of each strain were deposited at NCBI and annotated using the NCBI Prokaryotic Genome Annotation Pipeline. Statistics of all the ten draft genomes are presented in Table 1.

Potato Tuber Rotting Assays
Bacterial strains from Morocco were cultivated in TY broth for 24 h at 28 • C in a rotary shaker set at 125 rpm. Bacterial cultures were washed twice, resuspended in 0.8% NaCl, and the optical density was adjusted to OD 600 = 1.0. Potato tubers (cv. Bintje) were surface-disinfected by submerging them into a 5% sodium hypochlorite solution for 5 min. They were subsequently rinsed twice in distilled water and air dried at room temperature one day before inoculation. A total of 10 potato tubers were infected with 10 µL of cell suspension of each strain, along with 10 tubers with NaCl 0.8% alone as a negative control. After 5 days of incubation at 24 • C, the tubers were cut vertically through the inoculation points. Disease symptoms were evaluated to define five aggressiveness classes [52]. Significance of the observed differences was assessed using a Kruskal-Wallis test (p < 0.05).

Diversity of the Pectinolytic Dickeya and Pectobacterium in Northern Morocco
From 2015 to 2017, our field inspections revealed the occurrence of blackleg symptoms in several potato growing areas, located in many townships distributed in three regions (Meknes, S1; Guigo, S2; Larache, S4) in northern Morocco (Figure 1). No symptoms were found in fields in the Boumia region (S3). Out of 200 strains isolated from plant symptoms, 140 provoked cavities on the pectate-containing medium. These were tested by PCR to evaluate whether they belonged to the Pectobacterium and Dickeya genera: 119 isolates generated amplification signals for either the Y1/Y2 or ADE1/ADE2 primer couples. Most of the isolates (83%) generated a signal with the Pectobacterium primers Y1/Y2, while the others (17%) did so with the Dickeya-specific ADE1/ADE2 primers.
All these PCR-positive Pectobacterium/Dickeya isolates were further characterized at species level based on their gapA gene sequence. Phylogenetic analyses using the Neighbor-Joining method (Mega7) of the gapA sequences are presented in the Figure S1. The regional diversity of the Dickeya and Pectobacterium isolates is summarized in Figure 1 (a detailed list is given in Table S2). The samples of the Larache region (S4) showed the highest diversity of taxons with the presence of D. dianthicola (20 isolates), P. polaris (6 isolates), P. brasiliense (5 isolates) and P. carotovorum (23 isolates). On the other hand, our investigations revealed the presence of only two species, P. brasiliense (53 isolates) and P. carotovorum (3 isolates), in the Meknes region (S1), and only a single one, P. brasiliense (9 isolates), in the Guigo region (S2). and Pectobacterium isolates is summarized in Figure 1 (a detailed list is given in Table S2). The samples of the Larache region (S4) showed the highest diversity of taxons with the presence of D. dianthicola (20 isolates), P. polaris (6 isolates), P. brasiliense (5 isolates) and P. carotovorum (23 isolates). On the other hand, our investigations revealed the presence of only two species, P. brasiliense (53 isolates) and P. carotovorum (3 isolates), in the Meknes region (S1), and only a single one, P. brasiliense (9 isolates), in the Guigo region (S2). Important information was also collected from the farmers regarding the potato variety, origin of the seed tubers (local production or importation), irrigation mode (surface water from a dam or underground water from wells) and geography (the sampled regions). We tested whether a correlation between these different parameters and the diversity of pathogens (the combination of Dickeya and Pectobacterium species) existed. Statistical analysis with the SAS (Statistical Analysis System, version 9.00, SAS Institute, 2002, Cary, NC, USA) software (Qui2 test with p < 0.05) revealed Important information was also collected from the farmers regarding the potato variety, origin of the seed tubers (local production or importation), irrigation mode (surface water from a dam or underground water from wells) and geography (the sampled regions). We tested whether a correlation between these different parameters and the diversity of pathogens (the combination of Dickeya and Pectobacterium species) existed. Statistical analysis with the SAS (Statistical Analysis System, version 9.00, SAS Institute, 2002, Cary, NC, USA) software (Qui2 test with p < 0.05) revealed that the higher diversity of pathogens was associated with three confounding factors: geography (the unique Larache region), surface water irrigation and imported seed tubers (Table S1).

Draft Genomes of 10 Pectinolytic Bacteria from Northern Morocco
A draft genome (Illumina technology) was used to consolidate the taxonomic position of 10 isolates belonging to the collected taxons. Between 705,755 and 17,016,482 trimmed reads were used for the contigs assembly of each of the 10 genomes. Characteristics of the draft genomes are presented in Table 1. Genome data were exploited to retrieve 13 housekeeping genes from each genome using BLAST. The concatenated genes were used for MLSA. The MLSA tree ( Figure 2) showed a similar topology to the one generated by the gapA analysis ( Figure S1).
Genomic data were also used to calculate ANI and isDDH values. Most of the Moroccan strains exhibited an ANI value higher than 95%, and an isDDH value higher than 70% with the closest type strains, confirming their taxonomic assignation. Strains belonging to the P. versatile clade showed an isDDH lower than 70%, but an ANI value higher than 95% with the strain SCC1. Recently, another study confirmed the classification of the strains S4. 16.03.3I (= CFBP8660), S4.16.03.3F (= CFBP8659) and SCC1 into the P. versatile species [7].
A draft genome (Illumina technology) was used to consolidate the taxonomic position of 10 isolates belonging to the collected taxons. Between 705,755 and 17,016,482 trimmed reads were used for the contigs assembly of each of the 10 genomes. Characteristics of the draft genomes are presented in Table 1. Genome data were exploited to retrieve 13 housekeeping genes from each genome using BLAST. The concatenated genes were used for MLSA. The MLSA tree ( Figure 2) showed a similar topology to the one generated by the gapA analysis ( Figure S1). Genomic data were also used to calculate ANI and isDDH values. Most of the Moroccan strains exhibited an ANI value higher than 95%, and an isDDH value higher than 70% with the closest type strains, confirming their taxonomic assignation. Strains belonging to the P. versatile clade showed an isDDH lower than 70%, but an ANI value higher than 95% with the strain SCC1. Recently, another study confirmed the classification of the strains S4.16.03.3I (= CFBP8660), S4.16.03.3F (= CFBP8659) and SCC1 into the P. versatile species [7]. for Dickeya strains. The alignment of the concatenated genes fusA, rpoD, rpoS, acnA, purA, recA, mdh, mtlD, groEL, secY, glyA, gapA and rplB was generated using ClustalW; the evolutionary history was inferred using the Neighbor-Joining method and the evolutionary distances were computed using the Maximum Composite Likelihood method. Phylogenetic analyses were conducted using MEGA7 software. The name of the Moroccan isolates is underlined.

Aggressiveness of the Pectinolytic Bacteria from Northern Morocco
Of the 10 Moroccan strains whose genome sequence is available (Table 1), all but 1 (P. brasiliense S4. 16.03.1C) were tested for aggressiveness on potato tubers. P. brasiliense strain S4.16.03.1C was isolated from the same field as P. brasiliense strain S1.16.01.3K, and they showed 100% identify by ANI (Table 2). Hence, we retained one of the two for tuber maceration assays. For each of the nine strains, 10 tubers were inoculated, and the resulting maceration symptoms were classified into five symptomatic classes ( Figure 3). In addition, 10 tubers were used as uninfected control. The aggressiveness was compared between all the strains using a Kruskal-Wallis test. All the pathogens provoked symptoms that are different to the control condition. No significant difference was observed between strains belonging to the same species. In contrast, D. dianthicola strains showed a higher aggressiveness when compared with Pectobacterium strains (p value < 0.05). strains, 10 tubers were inoculated, and the resulting maceration symptoms were classified into five symptomatic classes (Figure 3). In addition, 10 tubers were used as uninfected control. The aggressiveness was compared between all the strains using a Kruskal-Wallis test. All the pathogens provoked symptoms that are different to the control condition. No significant difference was observed between strains belonging to the same species. In contrast, D. dianthicola strains showed a higher aggressiveness when compared with Pectobacterium strains (p value < 0.05). . Virulence test of Pectobacterium and Dickeya on potato tubers. The symptoms provoked by each strain were compared by infecting 10 tubers per strain. The 10 non-inoculated tubers were used as control. Symptoms were classified into five classes (CLA0, CLA1, CLA2, CLA3, CLA4, according to increasing severity). The typology of these classes was illustrated by a picture of an example. Data were statistically analyzed by a Kruskal-Wallis test (α = 5%). Lower case letters on the right of the graph indicate statistical differences between the different inoculated pathogens.

Complete Genome of D. Dianthicola LAR.16.03.LID
The genome of D. dianthicola LAR.16.03.LID was the first complete genome of a D. dianthicola strain collected in Morocco, and the third D. dianthicola genome in the NCBI database that already hosted those of strains ME23 and RNS04.9 ( Figure 4). Phylogenetic relationships between all D. dianthicola genomes available in NCBI were determined using MLSA. In the phylogenetic tree, the two Moroccan strains LAR.16.03.LID and S4.16.03.P2.4 appeared to be highly related. This could be explained by the close isolate locations, that were two potato fields separated only by a road. The two Moroccan D. dianthicola strains clustered with D. dianthicola strain ME23, which has been recently collected in a potato field in USA (Figure 4). The genomic relationship between D. dianthicola ME23 and D. dianthicola LAR.16.03.LID was confirmed using SNP/InDels calling. The SNP and InDel number in the LAR. 16.03.LID genome reached 12,259 and 16,335, using the ME23 and D. dianthicola RNS04.9 genomes as a reference, respectively. The LAR. 16.03.LID genome was annotated using the NCBI Prokaryotic Genome Annotation Pipeline. A graphical genome map is provided in Figure 5. The D. dianthicola LAR. 16.03.LID genome exhibited a high synteny with that of D. dianthicola strains Figure 3. Virulence test of Pectobacterium and Dickeya on potato tubers. The symptoms provoked by each strain were compared by infecting 10 tubers per strain. The 10 non-inoculated tubers were used as control. Symptoms were classified into five classes (CLA0, CLA1, CLA2, CLA3, CLA4, according to increasing severity). The typology of these classes was illustrated by a picture of an example. Data were statistically analyzed by a Kruskal-Wallis test (α = 5%). Lower case letters on the right of the graph indicate statistical differences between the different inoculated pathogens.

Complete Genome of D. dianthicola LAR.16.03.LID
The genome of D. dianthicola LAR.16.03.LID was the first complete genome of a D. dianthicola strain collected in Morocco, and the third D. dianthicola genome in the NCBI database that already hosted those of strains ME23 and RNS04.9 (Figure 4). Phylogenetic relationships between all D. dianthicola genomes available in NCBI were determined using MLSA. In the phylogenetic tree, the two Moroccan strains LAR.16.03.LID and S4.16.03.P2.4 appeared to be highly related. This could be explained by the close isolate locations, that were two potato fields separated only by a road. The two Moroccan D. dianthicola strains clustered with D. dianthicola strain ME23, which has been recently collected in a potato field in USA (Figure 4). The genomic relationship between D. dianthicola ME23 and D. dianthicola LAR.16.03.LID was confirmed using SNP/InDels calling. The SNP and InDel number in the LAR. 16.03.LID genome reached 12,259 and 16,335, using the ME23 and D. dianthicola RNS04.9 genomes as a reference, respectively. The LAR. 16.03.LID genome was annotated using the NCBI Prokaryotic Genome Annotation Pipeline. A graphical genome map is provided in Figure 5. The D. dianthicola LAR. 16.03.LID genome exhibited a high synteny with that of D. dianthicola strains ME23 and RNS049, with the exception of some large insertion/deletions scattered in the genomes ( Figure 6). These regions contained strain-specific genes with no counterpart in the other D. dianthicola genomes. In strains LAR.16.03LID and ME23, the analyses evidenced one strain-specific region that contained some mobile elements, such as genes from transposons and prophages; strain-specific regions 1 and 2 are presented in Figure 6. Additional information about these strain-specific regions are available in Table S3-1 and Table S3-2. Strain RNS049 exhibited four strain-specific regions which also contain mobile elements ( Figure 6 and Table S3-1-6). Phaster analysis suggested the presence of intact prophages ( Figure 6 and Table S4). The CRISPR elements are very important features of bacterial genomes as they provide acquired immunity against viruses and plasmids [53]. The three D. dianthicola genomes hosted three or four CRISPR loci ( Figure 6 and Table S5). specific regions 1 and 2 are presented in Figure 6. Additional information about these strain-specific regions are available in Table S3-1 and Table S3-2. Strain RNS049 exhibited four strain-specific regions which also contain mobile elements ( Figure 6 and Table S3-1-6). Phaster analysis suggested the presence of intact prophages ( Figure 6 and Table S4). The CRISPR elements are very important features of bacterial genomes as they provide acquired immunity against viruses and plasmids [53]. The three D. dianthicola genomes hosted three or four CRISPR loci ( Figure 6 and Table S5).  The genes fusA, rpoD, leuS, rpoS, purA, infB, gyrB, recA, groEL, secY, glyA, gapA, rplB, dnaX and gyrA were concatenated. The alignment was generated using ClustalW; the evolutionary history was inferred using the Neighbor-Joining method and the evolutionary distances were computed using the Maximum Composite Likelihood method. Phylogenetic analyses were conducted using MEGA7 software. The Moroccan D. dianthicola isolates are indicated in bold face.     The D. dianthicola LAR. 16.03.LID genome exhibited an arsenal of virulence genes similar to that described for D. dianthicola RNS049 [39]. All the pectinase-encoding genes described in the model strain D. dadantii 3937 [54] were conserved in the three D. dianthicola genomes, with the noticeable exceptions of the lacking pehK gene (which encodes a predicted polygalacturonase) and the presence Figure 6. Synteny between the complete genomes of D. dianthicola LAR. 16.03.LID, RNS049 and ME23 strains. Synteny analysis was performed using MAUVE software. The numbers indicate the position of strain specific genomic regions. The secondary pathway gene clusters were searched using AntiSMASH, the prophages were identified using PHASTER, and CRISPER loci were localized using CRISPER finder.
The D. dianthicola LAR. 16.03.LID genome exhibited an arsenal of virulence genes similar to that described for D. dianthicola RNS049 [39]. All the pectinase-encoding genes described in the model strain D. dadantii 3937 [54] were conserved in the three D. dianthicola genomes, with the noticeable exceptions of the lacking pehK gene (which encodes a predicted polygalacturonase) and the presence of a truncated form of the pelA gene (which encodes Pectate lyase A). Aside macerating enzymes determinants, other genes implicated in different stages of the host infection were conserved in the three D. dianthicola strains, including those involved in the resistance to oxidative stress, acidic pH (cfa, asr) and antimicrobial peptides (arnB-T, sapABCDF), synthesis of cell envelope components (such as bscABCD and wza-wzb-wzc), and siderophore synthesis and uptake (acsF-A and cbrABCDE for achromobactin and fct-cbsCEBA for chrysobactin) [54].
We strengthened this analysis using the resistance gene identifier (RGI) and AntiSMASH softwares. We searched for genes involved in the resistance to different families of antimicrobial compounds. No differences were observed between the three D. dianthicola genomes (Table S6). The AntiSMASH analysis identified many secondary metabolite biogenesis clusters in D. dianthicola genomes that were already described in several Dickeya species, like those responsible for the synthesis of siderophores, cyanobactin with cytotoxic activity, bacteriocin, nonribosomal peptide-synthetase (NRPS) and arylpolyene ( Figure 6 and Table S7).

Discussion
The main objective of this study was to characterize the pectinolytic populations isolated from symptomatic potato plants in Morocco between 2015 and 2017. A set of 119 pectinolytic bacteria, belonging to the genus Pectobacterium or Dickeya, were isolated and characterized using the gapA gene marker in combination with MLSA and ANI. Most of the isolates (83%) belonged to the Pectobacterium genus: the P. brasiliense species dominated in the Meknes and Guigo regions, while D. dianthicola was identified in the Larache region only. P. brasiliense caused major economic losses to several crops (potato, cucumber, paprika, etc.) in many countries, including Canada, USA, South Africa, China, Korea and New Zealand [55][56][57][58][59][60]. The wide host range of this pathogen could facilitate its survival even in harsh environments, by parasitizing many alternative host plants. In several studies, P. brasiliense isolates have been shown to be more aggressive than other Pectobacterium spp., except in the case of three Canadian strains exhibiting low aggressiveness [61]. In our study, the P. brasiliense isolates were as virulent as the other Pectobacterium strains. P. brasiliense was described in Morocco in 2012, and by now is the dominant species in two regions (S1 and S2). In the S2 region, the farmers use seed tubers produced in the S1 region, confirming the effective adaptation of this pathogen to the northern parts of Morocco.
The northern region (S4) exhibited the highest diversity of the pathogens. P. versatile and P. Polaris, described for the first time in Morocco, along with P. brasiliense and D. dianthicola, were isolated in the region S4. In this region, the majority of seed tubers were imported, and the irrigation water was derived from a dam. Either one or both agronomic practices could contribute to the wider diversity of pathogens in the Larache (S4) region than in the other investigated regions. While previous studies identified the P. carotovorum species as the most prevalent soft rot pathogens in Morocco [41,42], our study extended the diversity to other Pectobacterium and Dickyea species, including the recently described species P. versatile [7]. This species encompasses isolates, including the P. versatile strain SCC1 isolated in 1980 in Finland [62], which had been collected from a wide diversity of environments (host plants, surface waters) and geographic areas [7]. The presence of members of this clade in the potato field could be linked to irrigation, as P. versatile is also able to survive in this environment. This hypothesis remains to be investigated by sampling water from the dam. In addition, our study extended the worldwide distribution of this species to Morocco.
The international distribution of the genus Pectobacterium increases concerns about the economic losses caused by this bacterium to the potato growers. A study in the neighboring country Algeria, carried out between 2014 and 2015, revealed the presence of pectinolytic bacteria causing soft rot in potatoes that belonged to P. brasiliense and P. carotovorum, as judged by MLSA [63]. Ozturk et al. reported the presence of P. atrosepticum, P. brasiliense, P. carotovorum and P. parmentieri species in Turkey [64]. In Europe, the prevalence of different species belonging to the genera Pectobacterium and Dickeya detected in diseased potato plants differs from year to year and between countries, five bacterial species being the main causative agents of blackleg, namely P. atrosepticum, P. parmentieri, P. brasilense, D. solani and D. dianthicola [65].
The D. dianthicola species has been detected in Morocco in 2017 [44], and described as the main species causing losses in potato in North America [66]. More studies are needed for monitoring the spread of this highly aggressive pathogen. To reach this objective, we used Illumina and ONT sequencing technologies to assemble a complete genome of one D. dianthicola isolate that could be used, in the future, as a reference for studying the clonal variability of D. dianthicola populations in Morocco and elsewhere. Comparison of the three complete genomes available indicated the presence of several clusters that encode the biosynthesis of a number of secondary metabolites implicated in stress defense, possibly playing an important role during plant-bacteria interactions. For instance, bacteriocins are small molecules with bactericidal activity usually restricted to closely related species, increasing the competition during infection, while the production of arylpolyene, implicated in the protection against reactive oxygen species [67], has been recently described in D. fangzhongdai genomes [68].

Conclusions
This study revealed a wide diversity of Pectobacterium and Dickeya pathogens in northern Morocco, including P. polaris and P. versatile, that are reported for the first time in this country. In tuber maceration assays, the tested isolates of the emerging pathogen D. dianthicola were more aggressive than the Pectobacterium isolates. This feature should alert stakeholders to the threat that this pathogen poses to potato tuber production in northern Morocco. The nucletotide sequence data of the Dickeya and Pectobacterium Moroccan isolates, including a complete genome of D. dianthicola, 10 draft genomes and 119 partial sequences of the gapA gene, were deposited in a public database (NCBI GenBank) to be used as genetic resources for monitoring the spread of these pathogens in Northern Africa and elsewhere.