Genetic Diversity And Population Structure of Zymoseptoria Tritici Populations of Southern Ethiopia Using SSR Markers

The fungal disease Zymoseptoria tritici causes Septoria tritici blotch, which is one of the most serious challenges to wheat production in Ethiopia and around the world. Understanding the pathogen's genetic structure is critical for developing and implementing effective management methods. Therefore, the present study targeted to explore the genetic structure of 51 Z. tritici isolates collected from four wheat producing zones of South and Southwestern parts of Ethiopia using nine microsatellite markers. In all of the examined isolates, a Z. tritici specic diagnostic marker that targets the ITS rDNA had amplied a predicted fragment size of 345bp. The number of alleles, gene diversity, and polymorphic information content per locus ranged from 9 to 14, 0.80 to 0.88, and 0.70 to 0.87, respectively, indicating a signicant degree of genetic variety within populations. The results of an analysis AMOVA revealed a moderate (0.14) genetic differentiation, with 86 percent of total genetic variability (3.93) occurring within populations. Due to the existence of considerable gene ow, the dendrogram produced by UPGMA and PCoA also revealed a moderate population clustering in which the populations were not clearly clustered according to their sample areas. Furthermore, population structure analysis using a Bayesian model loosely grouped the population into ve (K) sub-groups with substantial genetic mixing. The populations of the Kembata-Tembar and Hadiya zone have higher genetic variability than the other populations studied, and hence can be considered STB hot sites for future research on pathogen dynamics, germplasm screening, and host-pathogen interactions.


Introduction
Wheat is the third most widely produced cereal crop in the world next to maize (Zea masys L.) and rice (Oryza sativa L.). It is the fourth most important staple crop in Ethiopia after Tef (Eragrostis tef), Maize (Zea mays) and Sorghum (Sorghum bicolor) Next to South Africa; Ethiopia is the second largest wheat producer in Sub-Saharan Africa (Letta et al. 2013). Both bread wheat (Triticum aestivum L.) and durum wheat (Triticum turgidumssp. durum L.) are widely cultivated in Ethiopia for multiple purposes including food, feed and income generation (Dixon et al. 2006). In 2017, about 1.7 M ha of land was covered with wheat and the national annual production and productivity were 4.8million metric tonnes and 2.8 t /ha, respectively (FAOSTAT, 2018). The potential wheat growing regions of the country includes Oromiya, Amhara, Tigray and Southern Nations Nationalities and Peoples region (SNNPR). In spite of its larger production coverage and multiple uses, the average productivity of wheat in Ethiopia is 2. 8 t/ha; which is by below the global average of 3.27 t/ha (Alemar Said and Temam Hussien, 2016).
Septoria tritici blotch (STB) caused by the ascomycete Mycosphaerella graminicola (asexual stage: Zymoseptoria tritici) is the major wheat devastating fungal disease next to rust in Ethiopia ( Abera Takele et al. 2015; Tilahun Mekonnen et al. 2019;2020) and elsewhere in the world (Eyal et al. 1985). Under favorable growing conditions with high relative humidity (85%) and optimal temperature (22°C), STB could decrease yield by 30 to 70% (Eyal et al. 1987). The disease is mainly a foliage disease and the primary infection may arise from airborne or rain splashed asexual pycnidospores and sexual ascospores from infested crop debris (Shaner, 1981;Gilchrist and Dubin, 2002). Changes in farming practices (higher sowing densities and nitrogen fertilization), mono cropping system and limited number of cultivars contributed signi cantly to the increase of the disease (Berraies et al. 2013). In Ethiopia up to 82% wheat yield loss has been reported (Mengistu Huluka et al. 1991;Abreham Tadesse, 2008). STB is becoming serious problem in southern and southwestern part of Ethiopia including Haddiya, Kambata, Silte and Southwest Shewa (Alemar Said and Temam Hussien, 2016). Integrated disease management strategies including genetic resistance (resistant varieties), crop rotation, appropriate fertilizer and fungicide applications, proper seeding rates and dates would be alternatives to control STB disease (Berraies et al. 2013). Genetic resistance remains the rst line of defense against this foliar disease, especially in developing countries for resource poor farmers, and is the most environmentally friendly and pro table strategy for farmers (Teklay Abebe et al. 2015). However, in Ethiopia most of the high-yielding wheat cultivars grown today are susceptible to STB. All commercial and candidate wheat varieties are affected by the disease at varying intensity (Alemar Said and Temam Hussien, 2016). This calls for searching for new source of resistance to the diseases. In this regard, knowledge of the genetic diversity of the pathogen could have direct implications for development of effective and durable disease management strategies (Schnieder et al. 2001) So far, different molecular marker systems have been used to explore the genetic structure of Z. tritici populations including Sequence Characterized Ampli ed Region (SCAR), simple sequence repeats (SSRs), restriction fragment length polymorphisms (RFLPs), ampli ed fragment length polymorphisms (AFLPs) and Rapid ampli ed polymorphic DNAs (RAPDs) (Schnieder et al. 2001). Simple sequence repeats (SSRs) also called microsatellites are considered to be the marker of choice for assessing genetic diversity of population as they are co-dominant, easy to use, highly polymorphic, multiallelic, locus speci c, and highly reproducible (Czembor and Arseniuk, 1999).
In breeding for disease resistance cultivars, knowledge of pathogen's genetic diversity has a great importance (Sebei and Harrabi, 2008). However, information on the pathogen's geographic distribution and genetic diversity in Ethiopia is limited. By considering the greater e ciency and speed of SSR markers, this research aimed to use them to study genetic diversity of Z. tritici from southern part of Ethiopia.
In Ethiopia, Tilahun Mekonnen et al. (2020) used SSR markers to explore the genetic diversity of Z. tritici populations of central highlands and south eastern parts of Ethiopia and con rmed the existence of higher genetic diversity among the pathogen populations. However, information on the genetic structure of the pathogen populations of Southern Ethiopia is lacking. Therefore, the present study was designed to assess the molecular diversity of Zymoseptoria tritici populations collected from southern part of Ethiopia to generate useful baseline information for the development of durable Septoria resistant cultivars.

Description of the Study Area
The Z. tritici isolates used in this study were isolated from STB symptomatic wheat leaf samples collected from naturally infected wheat elds of three zones (Kembata-tembaro, Hadiya, Silite) of the SNNPR state, and one bordering zone i.e Southwest Shewa zones from Oromia in 2018 main cropping season (Fig. 1). The zones were selected based on their wheat production potentials in Southern part of Ethiopia. From each zone, Woreda /District and Kebels (the lowest local administration) were subsequently selected based on their wheat production potential and road accessibility. Sample Collection STB symptomatic wheat leaf samples were collected from naturally infected wheat elds (Fig. 5) from the four zones at the beginning of October in 2018 main cropping season following the main roads and accessible routes at 5-10 km intervals based on vehicle odometers. Hundred samples were collected at medium milk and early dough growth stages (GS) as described by Zadoks et al. (1974). We have visited about 120 naturally grown wheat elds and 0.83 average samples were collected from each eld.
During collection, green leaf samples in the wheat eld naturally infected with STB with black spots (pycnidia) on the necrosis area were collected. Scissors were swiped with 70% ethanol prior to the next sample collection to avoid cross contamination.
Collected samples were placed in paper envelopes, and followed by recording of collection date, sample code, latitude, longitude, altitude, and disease severity score. The samples were left to dry at room temperature for a week, and then were placed in zipped plastic bag and stored at 5°C until isolation begins.
Diagnostic and Disease Assessment in the Field Diagnosis of the disease on wheat is based on the observation of the typical symptom caused by Z. tritici. The STB was identi ed by chlorotic spots and light tan lesions with small blackings (Ponomarenko et al. 2011). Disease severity was scored based on double digit scale (00-99) where the rst digit (0-9) indicates the necrotic leaf area on the four uppermost infected leaves of 10 -20 plants and the second 0-9 digit represent the blotch development up the plant height (for instance 5 if the disease reached at the middle (50%) of the plant height, 8 for ag leaf and 9 for spike), and the second digit stands for disease severity as a percentage but in terms of 0-9 (1=10%, 2=20% … and 9=90%). Depending on the size of the wheat eld three to ve stops were made in an "X" pattern and average result was taken to describe the disease severity of the eld (Eyal et al. 1987) Isolation of Septoria tritici Pathogen Isolation of Zymoseptoria tritici fungal spore and following activities were conducted at National Agricultural Biotechnology Research Center (NABRC), Holeta, 29 Km west of the capital city-Addis Ababa. The isolation was carried out as described by Eyal et al. (1987) with some modi cations. During isolation a necrotic wheat leaf with a pycnidium was placed on wet lter paper in Petri dish (Fig. 2a) at room temperature (20-25°C) for four hours. Under high humidity ooze that contains pycnidiospores were emerged from the opening of the pycnidium (ostiole) and formed a drop (cirrhus) on top of the dark pycnidium. Observing under stereoscopic dissecting microscope (Fig. 2b), mono-pycnidial oozing drops were transferred onto potato dextrose agar plate (PDA) containing 250 mg/l chloramphenicol succinate using ame sterilized needle. Inoculated Petri plates were kept at 24°C for 10 days until fungal growth was observed. Developed pinkish-orange colony was streaked on a new PDA plates without antibiotics and kept at the same conditions.
Developed mono-spore derived colonies ( Fig. 2c) were transferred to yeast-sucrose broth (YSB) (1% sucrose, 1% yeast extract) and incubated at room temperature on orbital shaker (180 rpm) for two weeks for spore multiplication (Fig. 2e). Spore pellets were recovered (Fig. 2f) by centrifugation at 10, 000 rpm for 5 minutes and stored at -80°C until used for DNA extraction. For long term storage fungal isolates were preserved in yeast extract (4g/l), malt extract (4g/l), and sucrose (4 g/l), supplemented with 30% glycerol and stored at -80°C Assessing the Genetic Diversity of Zymoseptoria tritici isolates Genomic DNA isolation Fungal genomic DNA isolation was carried out using diversity array technology (DArT) with minor modi cation. The DNA was quantitatively and qualitatively checked using a Nano drop spectrophotometer (ND-1000) and gel-electrophoresis. DNA concentration was adjusted to50 ng/µl using sterile nuclease free water and stored at -20˚C for further use.

Molecular detection of Zymoseptoria tritici
All 51 isolates were con rmed by running polymerase chain reaction (PCR) using a race speci c diagnostic markers ITS1 (5'-TCCGTAGGTGAACCTGCGG-3') as forward and JB446 (5'CGAGGCTGGAGTGGTGT-3') as reverse primer (Beck and Ligon 1995). The PCR reaction was performed in a volume of 12.5 µl containing 6.25 µl (2x Taq plus master mixes), 2µl Nuclease free water, 1µl of each of the forward (ITS1) and reverse (JB446) primers and 2.5 µl of 50 ng template DNA. Ampli cation conditions were set as an initial denaturation at 94°C for 3 min followed by 35 cycles with 1 min denaturation at 94°C, 1 min annealing at 60°C, primer extension at 72 C for 2 min followed by nal extension step of 10 min at 72°C. The PCR ampli ed products were fractionated in 3% agarose gel electrophoresis by loading 5µl of each of the PCR product mixed with 2 µl of loading dye with 1x gel red using 1x TAE buffer at 100V for three hours. Amplicon fragment size was estimated using 100 bp DNA molecular ladders.
The gel was visualized and also photographed using gel-documentation system (BioDoc-It TM imaging system).

SSR marker based genotyping
For molecular diversity and population structure analysis, the con rmed Z. tritici isolates were genotyped using nine published single locus microsatellite primers as described by (Owun et al. 1998 and Berraies et al. 2013) ( Table 2). PCR was performed in a total volume of 12.5 µl containing 6.25 µl master mix, 2 µl DNA tamplate, 1µl of each of the forward and reverse primers, 0.25 µl DMSO and 2 µl nuclease free water. PCR ampli cation was conducted on a Biometra thermocycler program as follows: 3 min at 94°C for initial denaturation followed by 40 cycles of denaturation at 94°C for 1min, annealing at 52 -65°C for 1 min, and extension at 72°C for 2 min. The nal extension was adjusted to at 72°C for 10 min. PCR products were resolved on 3% agrose gels electrophoresis using 1× TAE buffer at 100 V for 3 h. The gel was stained with gel red and visualized under UV light and subsequently photographed. To estimate the amplicon size A 50 Bp and 100 Bp DNA ladder were used.

Data scoring and statistical analysis
The PCR ampli ed SSR regions fragment size on gel were estimated using PyElph 1. Genetic dissimilarity (GD) between isolates was calculated according to the formula of Nei (1973). The UPGMA (unweighted pair-group method with arithmetic averages) clustering method was used to obtain the dendrogram using DARwin var. 6.0.14 (Perrier and Jacquemoud-Collet, 2006

Molecular Based Identi cation of Zymoseptoria tritici
Conserved regions of the ribosomal DNA (ITS-derived primers) were used to amplify speci c fragments from the isolates for the detection of Z. tritici species. Diagnostic analysis using a pair of primers (ITS and JB446) revealed that all the 51 tested Z. tritci isolates resulted in positive unambiguous ampli cation of the expected fragment size of about 345 bp. This con rms that all the morphologically identi ed study materials are Zymoseptoria tritici and considering their molecular diversity is thus important.

Molecular Diversity Analysis
Microsatellite markers level of polymorphism The analysis revealed that all the nine loci were found to be polymorphic and produced a total of 439 bands with an average of 48.77 bands per locus ( Table 4). The highest number (51) of bands per locus was recorded for AG-0003 marker, out of which 48 (94.11%) were polymorphic. The highest (95.65%) percentage of polymorphic bands was produced by the locus ST1E7, followed by AG-003 and then AC-002 which resulted in 94.11% and 94% polymorphic bands, respectively. Five (55.56%) of the considered SSR marker resulted in above average percentage of polymorphic bands. The lowest (90%) percentages of polymorphic bands were exhibited by two SSR loci (ST2E4 and AC-0001) ( Table 3). The study revealed that the number of alleles per locus varied from 9 to 14 with an average of 12 alleles per locus. The highest number of alleles (14) was resulted from two loci (ST2E4 and AG-0009). The analysis showed that 57% of the alleles were scarce (frequency between 0.01 and 0.05). The frequency of 15 (14%) alleles was between 0.05 and 0.1, while 31 (29%) alleles had a frequency of 0.1-1.00 (Table 4). The analysis showed that the major allele frequency per locus ranged from 0.18 ( were recorded for the microsatellite locusAG-0009 (Table 5).On the other hand, the highest major allele frequency (0.43) and the lowest gene diversity (0.73), polymorphic information content (0.70) and Shannon's information index (0.96) were observed for primer ST1D7 (Table 5). The PIC values of all the SSR loci were found to be high (>0.5) con rming their high informativeness (Table 5). Genetic variability within and among the populations Summary of the different genetic diversity indices over the entire SSR loci for the four populations is presented in Table 5. The analysis showed the existence of high diversity among the four populations of Z. tritici with regard to number of alleles, effective number of alleles, genetic diversity, private allelic richness, Shannon's information index and percentage of polymorphic loci (Table 5) (Table 6). While isolates of Southwest Shewa showed lowest number of alleles (3.22), effective number of alleles (2.60) genetic diversity (0.57) and Shannon's information index (1.00) ( Table 6). The analysis con rmed that all the studied populations showed 100% percentage of polymorphism. Based on the number of locally common alleles (Freq. >= 5%) which is found in 50% or fewer populations, populations of Hadiya showed the highest value (2.11).

Analysis of molecular variance (AMOVA)
There was moderate degree of genetic differentiations among the tested Zymoseptoria tritici populations as revealed by Analysis of Molecular Variance (AMOVA) (PhiPT = 0.14; p = 0.001). The analysis showed that among populations genetic variation accounted for only 14% of the overall (3.93) genetic variations allocating the 86% for the within populations molecular diversity ( Table 6). The analysis also con rmed the presence of considerable (3.14) gene ow (Nm) or gene migration (3.14) among the studied populations (Table 7).

Measures of Genetic identity and Genetic distance between the populations
Genetic grouping of populations based on Nei's genetic diversity showed that the pair wise genetic distance between the populations ranged from 0.36 to 0.82 and the maximum genetic distance (0.82) was observed between population from Kembata-Tembaro zone and population from Southwest Shewa zone ( Table 7). The lowest genetic distance (0.36) was observed between populations of Silte and Southwest Shewa zone. On the other hand, Nei's genetic identity between the populations ranged from 0.43 to1.00 (Table 8).  . 7). None of the clusters consisted of individuals of single population exclusively. The rst cluster constituted isolates from Kembata-Temibaro (61.8%), Silte (17.6%), Hadiya (8.8%) and Southwest Shewa (11.8). The second cluster consisted of isolates from Hadya (71.4%), Southwest Shewa (21.4%) and Silte (7.1%). Whereas the third cluster comprised of isolates of Hadiya zone (Fig. 7).

Principal coordinate analysis
The principal coordinate analysis (PCoA) also con rmed the presence of poor population structure in the pathogen. It showed that the rst three coordinates accounted for about 37.75% of the genetic variation. The rst, second, and third principal coordinates explained about 15.46%, 12.27% and 10.02% of the gross variation, respectively. The PCoA analysis in the twodimensional plot displayed in Fig. 8 showed that isolates from different collection sites were clustered together. It is not cluster the isolates distinctly based on their geographical areas of sampling (Fig. 8) complementing the result of NJ cluster analysis.
Population structure and admixture pattern Population structure analysis was carried out using STRUCTURE v 2.3.4 software based on Bayesian phylogenic method. The Structure harvester detected two picks (both at K = 5 and K=7 (Fig. 9A). In such cases the most likely number of genetic clusters corresponds to the smallest value of K that captures the major structure in the data (Pritchard et al. 2000). Hence, the model detected the presence of ve (K = 5) subpopulations. Based on this value, Clumpak result (bar plot) showed wide admixtures and hence there was no clear geographic origin-based on structuring of the populations (Fig. 9B).

Discussion
In the present study a diagnostic molecular marker speci c to Zymoseptorai tritici was used for identi cation of the pathogen. Ethiopia. The study con rmed that Z. tritici is a widespread and critical problem of wheat production in SNNP regional state of Ethiopia as well.
Analysis of the molecular diversity showed that the Z. tritici populations sampled from naturally infected wheat elds had high level of genetic diversity. In total, 439 bands were detected of which 92.71% where polymorphic. This gure is signi cantly higher than Kabbage et al. (2008) who reported 75% -81% polymorphic bands at different population scales. This is because of primary source of inoculum which was due to airborne ascospores that would be dispersed uniformly within the eld and long year's exposure of the pathogen populations to diverse resistance genes in the host materials.
Per locus, wide range of number of alleles (9 -14), gene diversity ( 0.73 -0.88) and polymorphic information content (0.70 -0.87) were recorded indicating that all the microsatellite loci were highly informative, and useful genetic tools to disclose the genetic structure of the pathogen populations.
The PIC values provide discriminating power of a marker by taking into account not only the number of alleles at a locus but also relative frequencies of these alleles. Lower PIC values might be result of closely related genotypes and vice versa. Marker loci with an average number of alleles running at equal frequencies will have the highest PIC value. Markers polymorphism with > 0.5are considered as highly informative, between 0.5 and 0.25 moderate and those with values below 0.25 are low informative (Smith, et al. 1997). In the current studies PIC value close to 1 suggests the presence of more alleles indicating polymorphism in that population which showed more uniform distribution of polymorphism bands among the isolates.
The average gene diversity (0.82) and number of alleles (12) per locus observed in the present study were signi cantly higher alleles, respectively using 45 isolates of Z. tritici in Tunisia. The probable reason may be the number of SSR markers used in the present study is higher than they used and the genotype difference due to different in terms of locations of isolates collection may create the variation. Gene diversity is the probability that two randomly chosen alleles from the population are different. The difference between the highest (0.88) and lowest (0.73) gene diversity indicates the presence of variability among 51 Z. tritici isolates.
The observed gene diversity in the present study was also by far greater than the report of Owen et al. (1998) and Razavi and Hughes (2004b) who found an average gene diversity of 0.49 and 0.44 for 12 UK isolates and for 90 Canadian isolates, respectively. Moreover, the mean number of alleles (12) per locus we reported is signi cantly higher than the level reported by Medini

Conclusion
In the present study, a total of 51 Z. tritici isolates from ve zones of South and southwestern parts of Ethiopia were pro led using molecular tools. The diagnostic marker that target the ITS region of rDNA ampli ed an expected fragment size from all the tested isolates. All the used markers were found to be highly informative with PIC information contents ranging from 0.7 -0.87) and gene diversity of 0.73-0.88, con rming that they are useful genetic tools to depict the molecular diversity and population structure of the pathogen populations. Authors' contributions The study's inception and design were aided by all of the writers. Messele Molla was in charge of sample collection, data collecting, and analysis. All co-authors contributed to the interpretation of the data, the preparation and revision of the earlier version of the paper, and the nal approvals.
Funding Ethiopian Institute of Agricultural Research and Addis Ababa University collaborated on the project.

Compliance with ethical standards
Con ict of interest: The authors disclose that they do not have any competing interests.
Ethical approval: There are no human or animal participants in this study.
Ethical responsibility: The data in this manuscript is unique to us and has never been published before. Other people's work and words have been acknowledged appropriately.
Author agreement/ declarations: We check that all named authors have read and approved the paper, and that no other individuals who meet the criteria for authorship but are not listed have done so. We also a rm that the manuscript's authorship order has been approved by all of us. Z.ymoseptoria tritici isolation procedure. a)10 cm long STB symptomatic leaves on samples on wet lter paper in Petri dish b) transferring oozing drops to PDA observing using stereoscopic dissecting microscope under sterile cabinet, c) Pure Z. tritici culture on PDA ready for multiplication, d) Z. tritici spores under 10x magni cation objectives, e) Spore multiplication in YSB, and f) Spore pellets recovered through centrifugation. Neighbor-joining tree for 51 isolates of Z. tritici based on Jaccard coe cient of similarity with 1000 replication as revealed by 9 SSR molecular markers. Kem stands for isolates from Kembata-Tembaro zone, Had refers to Hadiya isolates, S.W.S stands for South West Shewa collections and Sil stands for Silte zone collections.

Figure 5
Two dimensional plot of PCoA analysis of 51 Z. tritici isolates and four Population (1= Kembata-Tembaro, 2 = Hadiya, 3 = Silte and 4 = Southwest Shewa populations) based on 9 SSR markers. Samples coded with the same symbol and colors belong to the same population.