Diversity of Ethiopian Durum Wheat Landraces for Resistance to Stem Rust Seedling Resistance Genes

Stem rust caused by Puccinia graminis f. sp. tritici ( Pgt ) is one of the most important diseases of wheat worldwide. Breeding for resistance to diseases is the most important approach for mitigation of yield losses. This study was conducted to estimate the diversity of all stage stem rust resistance (ASR) genes on the 142 durum wheat landrace accessions at seedling stage. The study was conducted in greenhouse at Ambo Plant Protection Research Center on the 142 durum wheat landrace accessions using 20 differential lines, one susceptible line (McNair), and eight Pgt races. The result depicted the presence of Sr7b, Sr8a, Sr9b, Sr10, Sr11, Sr13, Sr17, Sr30, Sr31, Sr36, and SrTmp in the Ethiopian durum wheat accessions. Among the 142 durum wheat accessions, 83 accessions were identified for possessing single ASR genes, and four accessions including the universal susceptible line (McNair) did not have effective resistance genes to the pathogen races tested in this study. The remaining 55 accessions had either a combination of two resistance genes, unknown number and kind of genes, or unidentified genes displaying resistance across all the pathogen races. This study demonstrated the prevalence of significant genetic diversity for stem rust ASR genes in the Ethiopian durum wheat landraces.


Introduction
In sub-Saharan Africa (SSA), Ethiopia is the largest producer of wheat with approximately 1.7 million ha of land under cultivation [1]. Wheat is fourth in area coverage and third in amount of grain production following maize and te in the country [2]. In Ethiopia, more than 90 bread and 36 durum wheat varieties have been released for production since 1950s. However, the national average yield is 2.42 t/ha, which is far less than the world average of 3.43 t/ha [1]. e low productivity is attributed to lack of varieties resistant to the prevalent wheat rusts, namely, the stem rust (Puccinia graminis f. sp. tritici Eriks. and E. Henn), leaf rust (P. triticina Eriks.), and stripe rust (P. striiformis Westend. f. sp. tritici Eriks.), which are the major important diseases.
Stem rusts are managed by cultural control, chemical applications, and use of resistant varieties, in which case the third option is the best strategy [3,4]. Wheat producers in Ethiopia require disease resistant varieties which were farmers-friendly, environmentally safe, and cost e ective. It is important to identify sources of resistance genes in order to develop disease resistant wheat cultivars. One of the rich sources of stress resistance germplasm is landraces or farmers' varieties, which are also known to be reservoirs of genetic resources like resistance genes for several plant diseases including wheat rust [5][6][7].
Use of crop diversity is a key approach to improve productivity and achieve food security [8]. Ethiopian durum wheat landraces are diverse and possess high variation for economically important agronomic traits including resistance/tolerance to both biotic and abiotic stresses but are not exploited enough [9][10][11]. e durum accessions contributed to the world wheat varietal improvement; for instance, the Ethiopian durum wheat landrace ST464 was one of the major sources of Sr13 [12].
Most commonly three of the stem rust resistance genes (Sr9d, Sr9e, and Sr13) are present in many T. durum genotypes alone or in various combinations [13]. However, the adequacy of the resistance genes believed to be present in those cultivars may not be effective in providing full protection against the pathogen [14]. erefore, it is important to expand our knowledge on the response of the Ethiopian durum wheat accessions to the current pathogen populations. Hence, we need to search new sources of stem rust resistance genes, particularly host plants possessing durable resistance genes/nonrace specific resistance genes [5,6]. e source of seeds used by majority of durum wheat growing farmers in Ethiopia is landraces consisting of large numbers of different genetic backgrounds [9]. For identification of resistant sources of genes, germplasms were assessed from known sources and screened for triple resistance to wheat rust diseases [15,16]. Colomba and Gregorini [17] reported that there are six types of Triticum species grown in Ethiopia, namely, T. dicoccum, T. turgidum, T. durum, T. polonocim, T. pyramidale, and T. carthilcum. Among those, T. durum (durum wheat) and Triticum aestivum are the most dominantly grown species. For future use in research and maintenance of the available germplasm, the Institute of Biodiversity Conservation (IBC) has collected more than 12,726 accessions of Triticum species from various agroecological zones of Ethiopia and, out of those, tetraploid wheat species accounted for 72% of the germplasm collection [18]. erefore, durum wheat accessions collected from different agroecologies and locations are considered to vary for resistance to diseases and pests, grain yield, and adaptation to specific environmental situations and are generally considered initial ground for durum wheat improvement program [18]. Hence, this study was conducted with the general objective of evaluating the genetic diversity of durum wheat accessions grown in Ethiopia for resistance to stem rust pathogen (Puccinia graminis f. sp. tritici). Pot experiment for evaluating seedlings in the greenhouse was conducted to infer resistance genes in the 142 wheat accessions including one control susceptible cultivar (McNair) and 20 differential lines following the method described by Roelfs and Marten [19]. 25 seeds from 142 durum wheat accessions and 15 seeds of each differential line were pregerminated on moist filter paper in 90 cm Petri dish. After two days, five sprouting seeds were transplanted into a 5 cm diameter plastic pot filled with sterilized soil, sand, and compost at the ratio of 2 : 1 : 1. Each pot was replicated twice and placed in seedling growth chamber room until two primary leaves were emerged for inoculation. In the same manner, the differential accessions were planted in pots and were arranged in four sets of five groups according to the following orders (

Inoculum Preparation and Inoculation.
For the eight predominant stem rust races (TTKSK (Ug99), TKTTF, JRCQC, TRTTF, TTRTF, TTKTT, TKKTF, and TTTTF), their virulence spectra on the stem rust differentials are described in Table 6. Techniques for inoculum production, collection, storage, and inoculation followed the standard guideline produced by Roelfs et al. [21]. Increasing of urediospore bulk sample was conducted on susceptible cultivar (McNair). e method used for deriving single isolate, characterization, and nomenclature was described by Fetch and Dunsmore [22]. One gelatin capsule of freshly harvested urediospores prepared by suspending 14 mg in 0.75 ml lightweight mineral oil, Soltrol 170 (Chevron Phillips Chemical Company, e Woodlands, Texas, United States) was used to inoculate on 48 accessions (240 seedlings per tray). Inoculation was done using atomized inoculator by spraying when the seedlings have fully expanded primary leaves and the second leaves begin to grow after seven days at seedling stage [23]. Inoculated seedlings were moistened with fine droplets of distilled water produced with an atomizer and placed in a dew chamber in darkness for 18 hours at 18 to 22°C temperature and 98 to 100% relative humidity. Upon removal from dark chamber, plants were exposed to 4 hours of fluorescent light to provide condition for infection and allowed to dry dew for about 2 hours. Inoculated plants were then transferred to greenhouse benches where conditions were regulated at 12 hours' photoperiod, with temperature range of 18 to 25°C and relative humidity (RH) of 60 to 70% [24].  Accessions  TTKSK  TKTTF  JRCQC  TRTTF  TTRTF  TTKTT  TKKTF  TTTTF  Postulated genes  222469 3 −

Disease Parameters.
In the greenhouse, data recording for seedling infection types began after 14 days of inoculation using the 0-4 scoring scale developed by Stakman et al. [25], where 0 indicates immune or fleck; 1 indicates small uredia with necrosis; 2 indicates small-to-medium uredia with chlorosis or necrosis; 3 indicates medium-size uredia with/without chlorosis; and 4 indicates large uredia without chlorosis or necrosis. e scales up to the rate of 2 were considered to be incompatible, while the rates above 3 were regarded as compatible reactions. e infection types were defined by modifying characters as follows: − , uredinia somewhat smaller than normal; +, uredinia somewhat larger than normal for the infection type.

Data Analysis and Interpretation.
Seedling resistance genes were postulated based on the gene-for-gene specificity hypothesis between the host resistance genes and the avirulence genes in the pathogen by correlating the response of the differential sets of each pathogen with the response of the host genotypes. To interpret the results from the multipathotype tests, a differential response key for a given gene based on responses of the target differential genotype against an array of pathotypes was generated (Table 5).
is key was used to postulate the resistance genes which occurred singly in various entries. Postulation of more than one resistance gene per entry was performed when deviation from the usual infection type was shown by a particular resistance gene, as evidenced by the specific infection types.

Results and Discussion
e gene postulation result depicted the presence of diverse types and numbers of stem rust resistance genes in the Ethiopian durum wheat accessions (Tables 1-4). Among the 142 durum wheat accessions evaluated for the presence of stem rust resistance genes, 83 were found to possess nine different kinds of (Sr7b, Sr8a, Sr9b, Sr10, Sr11, Sr30, Sr31, Sr36, and SrTmp) singly postulated stem rust resistance genes. Four accessions (5180, 204463, 222433, and 203968) including the universal susceptible line (McNair) did not have effective resistance genes to the pathogen races tested in this study. e remaining 55 accessions had either a combination of two resistance genes (like Sr13 and Sr17  Accessions  TTKSK  TKTTF  JRCQC  TRTTF  TTRTF  TTKTT  TKKTF  TTTTF  Postulated genes  206627 3 Advances in Agriculture postulated on 23 of the accessions), unknown number and kind of genes (where the ITs displayed could not be correlated/matched with the ITs of the tested 20 differential lines), or unidentified genes displaying resistance across all the pathogen races (Table 4).

Group 1: Durum Wheat Accessions Postulated for Possessing Single Resistance
Genes. e largest portions (58.45%) of the durum wheat accessions were found to possess only one (a single) stem rust resistance gene. Among the singly postulated stem rust resistance genes, the most frequent resistance gene was Sr11 (18.31%), followed by Sr10 (12.67%), Sr36 (9.15%), and Sr9b (5.6%). On the other hand, Sr31 (3.5%), Sr8a (2.8%), SrTmp (2.1%), Sr7b (1.4%), and Sr30 (0.7%) were the least postulated genes. e durum wheat accessions and their phenotypic expressions (disease reaction) against each of the eight Pgt races with the corresponding disease reaction of the differential cultivars are presented in Table 1. e above nine kinds of Sr resistance genes were postulated by comparing the IT patterns of the eight different Pgt races on the 83 durum wheat accessions with those of the differential lines possessing the known resistance genes as displayed.

Group 2: Durum Wheat Accessions with Two All Stage
Resistance (ASR) Genes. Accessions that showed low ITs similar to a combination of two resistance genes with compensating pathotypic specificities were postulated to carry the corresponding two Sr genes. In this case, 23 durum wheat accessions exhibited exactly identical ITs displayed by the differential cultivar "Combination V" that is known to carry Sr13 + Sr17 together (Table 2).

Group 3: Durum Wheat Accessions Postulated to Carry
More than Two Genes. e three durum wheat accessions 226867, 222422, and 208206 were postulated to carry three or more (Sr11, Sr36, and other unknown) resistance genes in combination (Table 3). ese accessions displayed low infection types to six of the Pgt races (TTKSK, TKTTF, JRCQC, TTKTT, TKKTF, and TTTTF) and high ITs for the remaining two Pgt races (TRTTF and TTRTF). e Sr36  Table 3: List of durum wheat accessions postulated to carry more than two stem rust resistance genes. Accessions  TTKSK  TKTTF  JRCQC  TRTTF  TTRTF  TTKTT  TKKTF  TTTTF  Postulated genes  226867 2

Discussion
Stem rust is economically one of the most important diseases of wheat that could lead to 100% yield loss if susceptible cultivar is used. Germplasm development and enhancement activities should be given more attention for mining novel stem rust resistance genes by screening diverse sources of germplasm collections in order to combat the devastating effects of stem rust disease. Ethiopia is highly endowed with durum wheat genetic diversity owing to the presence of large number of collections of tetraploid wheat germplasms maintained in the Institute of Biodiversity Conservation. Additionally, cultivation of landraces of durum wheat by millions of smallholder farmers in the central highlands ensures the continuation of genetic conservation of the crop for future use. Landraces are known to be reservoirs of genetic diversity for several economically important genes that can be easily deployed to modern cultivars through conventional and modern breeding methodologies. Gene postulation studies (multipathotype tests) conducted at seedling stages in the greenhouse facilitate identification of resistance genes controlled by major genes in the host genotypes (durum wheat plant in this case). In this study, 142 Ethiopian durum wheat accessions obtained from IBC, Ethiopia, were screened for resistance to stem rust pathogens in the greenhouse during the 2019 cropping season. e greenhouse study depicted 11 stem rust resistance genes (Sr7b, Sr8a, Sr9b, Sr10, Sr11, Sr17, Sr24, Sr30, Sr31, Sr36, and SrTmp). In agreement with the current finding, Belayneh et al. [26] postulated 11 stem rust resistance genes (Sr5, Sr7a, Sr7b, Sr8a, Sr9e, Sr11, Sr21, Sr27, Sr29, Sr30, and Sr37) from a set of 60 wheat genotypes constituted from 30 durum wheat and 30 bread wheat genotypes using 40 differential cultivars and 10 Pgt races isolated from Ethiopian Pgt population obtained from annual Pgt pathotype surveys. Similarly, Randhawa et al. [27] postulated seven kinds of stem rust resistance genes (Sr7b, Sr8a, Sr12, Sr15, Sr17, Sr23, and Sr30) from a set of 87 Nordic spring wheat cultivars using eight Australian Pgt races. In general, the multipathotype test indicated the presence of relatively appreciable diversity (11 Sr genes including Sr13+Sr17 postulated in combination) of stem rust resistance genes in the Ethiopian durum wheat landrace accessions; in particular, the set of germplasm accessions that displayed unknown and or uncharacterized Sr genes could add to additional Ug99 effective resistance genes in combinations or novel genes which are effective against all the pathotypes used in this study. Similar views were endorsed by Belayneh et al [26], Randhawa et al. [28], and Dakouri et al. [29] who identified uncharacterized/unknown genes effective against all the Pgt races used in their respective seedling tests in the greenhouse.
Regarding the number and type of stem rust resistance genes, Beteselassie et al. [16] postulated Sr7b, Sr8b, Sr9a, Sr9b, Sr10, Sr14, Sr24, Sr27, Sr28, Sr29, Sr30, Sr31, Sr32, and SrTt-3 + Sr10 on 16 emmer and 5 durum wheat landrace accessions (obtained from IBC, Ethiopia) using 10 Pgt races and 33 differential accessions. ese authors reported more diverse types of Sr genes compared to the current study that might be due to the use of 33 differential lines and 10 types of Pgt races and evaluation of two different tetraploid wheat species (emmer and durum wheat). In the same manner, Source: [20].

Advances in Agriculture
Belayneh et al. [26] used 40 differential cultivars and lines representing 40 different Sr genes and 10 Pgt races. ose authors were able to postulate Sr9e, Sr14, Sr21, Sr27, Sr28, Sr29, Sr32, and Sr37, whereas this study lacks the differential cultivars that help to detect these genes in the durum wheat germplasm. More findings similar to the results of this study have been reported by Dyck and Sykes [30] who carried out genetic analysis of stem rust resistance in Ethiopian durum wheat germplasm that showed the presence of Sr6, Sr8a, Sr9a, Sr9d, Sr9e, Sr11, Sr13, Sr30, and Sr36. e presence of Sr13 + Sr17 in the Ethiopian durum wheat germplasm has been reported by Banchigize-Getie [31] who evaluated 45 Ethiopian durum wheat landraces along with 35 differential cultivars and nine Pgt races for seedling resistance using multipathotype test conducted at the University of Sydney, Australia. Klindworth et al. [12] and Periyannan et al. [32] reported that durum wheat is the major source of Sr13, which is in agreement with this study.
e Pgt race JRCQC in Ethiopia and other races of stem rust in several countries were reported to show virulence to Sr13; however, this gene was found to be effective against the Pgt race Ug99 (TTKSK) and its derivatives that gave it major importance worldwide to be deployed in new cultivars in order to combat this aggressive Pgt race [32]. Hence, these 23 durum wheat accessions could be good sources of stem rust resistance to be incorporated in breeding lines to develop wheat cultivars.
Several researchers postulated combination of three or more rust resistance genes in a single genotype [16,26,27,31]. e durum accessions carrying a greater number of resistance genes could be excellent candidates (sources of germplasm) for gene pyramiding in transgressive breeding.
In order to solve the mysteries of the USR genes presented in Table 4, carrying out of additional such experiments with other additional differential lines and foreign Pgt races might help to identify the genes conferring resistance to stem rust in some or more of the genotypes [26]. e other option is to conduct molecular marker analysis using diagnostic DNA markers for the genes frequently found in durum wheat germplasm [33]. e last option could be to conduct genetic analysis for each of the landraces and study the stem rust inheritance to determine the type and number of gene(s) conferring resistance to these genotypes [34,35]. Discovery of USR genes in such kinds of (multipathotype tests) experiments conducted for Ethiopian durum wheat landraces had been reported by Naod-Beteselassie et al. [16], Belayneh et al. [26], and Getie [31] who have endorsed similar views indicated above.
Generally, this study depicted 11 stem rust resistance genes either singly or in combinations of two, three, or greater number of genes. Most of these ASR genes (60% or 83 accessions) are not effective when used singly to protect the host from Pgt in the field. Hence, they should be used in combination with other effective ASR and/or APR genes; particularly the ASR genes should be selected based on their compensating race specificities. For instance, 23 of the 142 durum wheat accessions possessed Sr13 + Sr17 in combination; also 3 of them possessed Sr11 + Sr36 + USR genes. ese genes showed compensating race specificities; due to this, the Sr11 + Sr36 + USR possessing accessions were resistant to six of the eight races, whereas the Sr13 + Sr17 possessing accessions were resistant to all races of the pathogens tested. e remaining 29 accessions possessed unidentified stem rust resistance genes. Such findings are not uncommon in such kinds of experiments. ese types of genetic materials could be potential sources of novel genes for rust resistance genes. Further studies such as genetic analysis (inheritance studies) and molecular marker analysis would lead to identification and characterization of the USR genes.

Data Availability
All the data related to this manuscript are included in the manuscript. If any additional information is needed and is available, the authors can provide it upon request to the corresponding author.

Conflicts of Interest
e authors declare that they have no conflicts of interest.