INTRODUCTION

Rice is perhaps the most widely cultivated food crop world over, but its production is constrained by diseases of fungal, bacterial and viral origin. Bacterial Leaf Blight (BLB) of rice, caused by Xanthomonas oryzae pv. oryzae (Xoo) is one of the oldest known diseases and was first noticed by the farmers of Japan in 1884. Subsequently, its incidence has been reported from different parts of Asia, northern Australia, Africa and USA (Mew et al., 1993; Awoderu et al., 1991; Sigee, 1993; Sere et al., 2005).

In Asia, BLB became a major disease after the introduction and widespread cultivation of high yielding but susceptible rice cultivars. Consequently, comprehensive studies on pathogen virulence diversity were undertaken, which provided useful information on the Xoo virulence population structure and resistance genes used in Asian breeding programs. However, little information is available on Xoo virulence population structure in Africa. This makes it very important to study the status of this bacterial disease in West African countries. Recent studies on BLB disease survey and samplings at different rice ecologies in Niger, Burkina Faso, Nigeria, Benin and Mali, revealed that BLB frequently occurred in farmers fields across these countries with incidence ranged from 70-85% and yield loss ranged from 50-90% (Sere et al., 2005). This indicates a wide spread of BLB in farmers fields across West African countries. Some selected Xoo isolates have shown high level of pathogenicity and virulence on the cultivated rice varieties (Sere et al., 2005). Research studies have also revealed that BLB is an important rice disease in irrigated rice ecosystems in West Africa.

Bacterial leaf blight is characterized by a high degree of race-cultivar specificity. There are over 30 reported races of isolates from several countries (Adhikari et al., 1999; Mew et al., 1993). A set of races identified in the Philippines using five differential rice cultivars (Mew et al., 1993) has been used widely for identifying and classifying resistance to BLB in other cultivars (Lee et al., 2003). It has been noted, however, that screening for resistance to pathogen populations specific to particular geographical locations and tailoring regional breeding programmes accordingly are important (Mew et al., 1993). Xoo also has a high degree of genetic diversity among different isolates, based on Restriction Fragment Length Polymorphism (RFLP) and pathotype analyses of more than 300 strains from different parts of Asia, using a repetitive Insertion Sequence (IS) element as the RFLP probe (Adhikari et al., 1999). In the study, isolates formed five clusters, each with more than one pathotype. Some correlation of clusters with geographical distribution and specific pathotypes was observed, indicating that tailoring breeding programmes for specific regions isindeed a tenable approach to control, although there was also evidence of movement of strains among regions. However, the present study aimed at conducting pathotyping analysis of 50 Xanthomonas oryzae pv. oryzae (Xoo) isolates from seven West African countries against 18 rice cultivars in order to identify and characterize Xoo virulence. The characterization of Xoo virulent population structure in West Africa will provide wide useful information for selection and deployment of cultivars with durable resistance.

MATERIALS AND METHODS

Bacterial Isolates
Xanthomonas oryzae pv. oryzae (Xoo) isolates (Table 1) used in this study were obtained from Plant Pathology Unit, Africa Rice Center (WARDA), Cotonou, Benin Republic, where their identity had been confirmed by oxidative biochemical test.

Near-Isogenic Lines (NILs)
Fourteen near isogenic lines (NILs) (Table 2) used for Xoo pathotyping study were obtained from International Rice Research Institute (IRRI). These are rice NILs with known resistance gene to Bacterial Leaf Blight (BLB). Other varieties such as IR64 and PNA647F4-56 were included as susceptible check (SCK) to BLB while Gigante and TOG5681 (highly resistant to Rice yellow mottle virus) were also included to study their current BLB resistant status.

Experimental Design
Split plot design with 3 replications was used. Fifty Xoo isolates (Table 1) were used to screen 18 varieties (Table 2) inside the screenhouse at Africa Rice Center (WARDA), Cotonou, Benin Republic. The experiment was carried out between February to May 2007. Rice grains were first pregerminated in steriled petri dishes under sterile condition. One plastic pot per variety per isolate in three replications was used.

Fertilizer Application
At transplanting 1.0 g of NPK per pot was applied and at 21 days after transplanting 0.2 g of Urea per pot was applied.

Table 1:	List of Xanthomonas oryzae pv. oryzae isolates used for the study

Isolates Inoculation
Fifty Xoo isolates of 20 μL each from stock culture were grown in Glucose Yeast Extract (GYE) liquid medium at 28°C to promote accelerated growth at incubation time of 36 h to obtain most active and effective bacterial cells (Sere et al., 2005). Inoculum was prepared by suspending the bacterial cells in sterile distilled water and adjusted to a concentration of 10⁹ cfu mL^-1 (OD₆₅₀ = 0.5) prior to inoculation. Inoculation was by clipping method (Sere et al., 2005). The whole leaves of each plant in each plastic pot were clip inoculated 21 days after sowing.

Measurement of Parameters
Temperature and relative humidity within the screen house were measured by thermohygrometer throughout the experiment period. At 14 days after inoculation, lesion length and total leaf length from the cut leaf tip were measured in centimeter (Sere et al., 2005). From the collected lesion length and total leaf length data, percentage lesion length was estimated. BLB disease reaction was categorized according to percentage lesion length.

Data Analysis
Using the percentage lesion length data, Analysis of Variance (ANOVA), genotype by environment (GxE) interaction and additive main effect and multiplicative interaction (AMMI) analyses were conducted using IRRISTAT software to identify different Xoo pathotypes and virulence groups (Ebdon and Gauch, 2002; Bruckner and Slanger, 1986; Aleong and Howard, 1985; Xiaoping and Ognjen, 2005).

RESULTS AND DISCUSSION

Considerable diversity was observed in the reactions of 50 Xoo isolates to 18 rice cultivars in terms of percentage lesion length due to Bacterial Leaf Blight (BLB) disease. Analysis of Variance (ANOVA) for percentage lesion length due to BLB disease caused by inoculated Xoo isolates from seven West African countries revealed significant strong interaction, between Xoo isolates and rice cultivars and between isolates country of origin and rice cultivars (Table 3). Means percentage lesion length were significantly different both for BLB disease caused by Xoo isolates of the same and different country of origin (Table 4, 5).

Table 2:	List of near isogenic lines (NILs) and other varieties used for Xoo pathotyping
SCK = Susceptible check

Table 3:	Analysis of variance for percentage lesion length due to BLB diseases caused by inoculated Xoo isolates from different West African countries
** = Significant at 1% level

Percentage lesion length relative to Xoo isolate country of origin was between 8.2-23.6% (Niger), 6.3-26.6% (Benin Republic), 9.3-28.3% (Nigeria), 12.2-53.6% (Burkina Faso),14.4-31.4% (Mali), 17.6-46.2% (Guinea) and 10.8-57.9% (The Gambia) (Table 4, 5). Xoo isolates from The Gambia produced the highest percentage lesion length of 57.9%, followed by those from Burkina Faso (53.6%), Guinea (46.2%), Mali (31.4%), Nigeria (28.3%), Benin Republic (26.6%) and Niger (23.6%) (Tables 4, 5). According to mean percentage lesion length by isolates country of origin, Burkina Faso produced the highest mean percentage lesion length of 34.4%, followed by Guinea (31.9%), The Gambia (28.3%), Nigeria (20.5%), Mali (19.5%), Niger (17.7%) and Benin Republic (17.1%) (Table 5; Fig. 1).

Table 4:	Analysis of means comparison for percentage lesion length due to BLB disease caused by inoculated Xoo isolates
In a column, means followed by a common letter are not significantly different at the 5% level by Duncan’s Multiple Range Test.Country: C1 = Niger; C2 = Benin Republic; C3 = Nigeria; C4 = Burkina Faso; C5 = Mali; C6 = Guinea; C7 = The Gambia. Variety: V1 = IRBB1; V2 = IRBB2; V3 = IRBB3; V4 = IRBB4; V5 = IRBB5: V6 = IRBB7; V7 = IRBB8; V8 = IRBB10; V9 = IRBB11

Table 5:	Analysis of means comparison for percentage lesion length due to BLB disease caused by inoculated Xoo isolates
In a column, means followed by a common letter are not significantly different at the 5 % level by Duncan’s Multiple Range Test. Country: C1 = Niger; C2 = Benin Republic; C3 = Nigeria; C4 = Burkina Faso; C5 = Mali; C6 = Guinea; C7 = The Gambia. Variety: V10 = IRBB13; V11 = IRBB14; V12 = IRBB21; V13 = IRBB53; V14 = IRBB59; V15 = IR24: V16 = PNA647F4-56; V17 = Gigante; V18 = TOG5681

Xoo isolates that produced percentage lesion length greater than 15% were considered virulence (Table 6). Xoo isolates from Burkina Faso have the highest percentage virulence of 45.4%, followed by Guinea (42.2%), The Gambia (38.3%), Nigeria (30.0%), Benin Republic (27.8%), Mali (26.4%) and Niger (23.1%) (Table 7). At the level of individual isolate, all the 50 Xoo isolates were virulence with the exception of 5 isolates (XN-6, XB-7, XNG-16, XBF-17 and XTG-49) which were less virulence (Fig. 2).

According to additive main effects and multiplicate interaction (AMMI) analysis, all the Xoo isolates were responsible mainly for unfavourable interactive conditions leading to significant increase in percentage lesion length in all the rice cultivars (Fig. 3, 4 and 5). Xoo virulence and interactive conditions were similar among isolates from Guinea and Burkina Faso, Mali and Nigeria, Benin and Niger with the exception of the Gambia which was distinct (Fig. 4, 5). Based on cluster dendrogram classification of Xoo isolates virulence, two major Xoo pathotypes (Pta and Ptb) were revealed (Fig. 6). Pta pathotype was made up of 29 virulence (Vr) Xoo isolates while Ptb pathotype constituted 21 mildly virulence (MVr) Xoo isolates (Fig. 6). Pta pathotype was further divided into three subgroup pathotypes (Pta1,

Fig. 1:

Virulence status of 50 Xoo isolates against 18 rice varieties. Isolate: I1 = XN-1; I2 = XN-2; I3 = XN-3; I4 = XN-4; I5 = XN-5; I6 = XN-6; I7 = XB-7; I8 = XB-8; I9 = XB-9; I10 = XB-10; I11 = XB-11; I12 = XNG-12; I13 = XNG-13; I14 = XNG-14; I15 = XNG-15; I16 = XNG-16; I17 = XBF-17; I18 = XBF-18; I19 = XBF-19; I20 = XBF-20; I21 = XBF-21; I22 = XBF-22; I23 = XM-23; I24 = XM-24; I25 = XM-25; I26 = XM-26; I27 = XM-27; I28 = XM-28; I29 = XM-29; I30 = XM-30; I31 = XG-31; I32 = XG-32; I33 = XG-33; I34 = XG-34; I35 = XG-35; I36 = XG-36; I37 = XG-37; I38 = XG-38; I39 = XG-39; I40 = XG-40; I41 = XTG-41; I42 = XTG-42; I43 = XTG-43; I44 = XTG-44; I45 = XTG-45; I46 = XTG-46; I47 = XTG-47; I48 = XTG-48; I49 = XTG-49; I50 = XTG-50

Table 6:	Analysis of means comparison for percentage lesion length due to BLB disease relative to inoculated Xoo isolates country of origin
In a column, means followed by a common letter are not significantly different at the 5 % level by Duncan’s Multiple Range Test. SCK = Susceptible check. N = Number of Xoo isolates. Country: C1 = Niger; C2 = Benin Republic; C3 = Nigeria; C4 = Burkina Faso; C5 = Mali; C6 = Guinea; C7 = The Gambia. Variety: V1 = IRBB1; V2 = IRBB2; V3 = IRBB3; V4 = IRBB4; V5 = IRBB5: V6 = IRBB7; V7 = IRBB8; V8 = IRBB10; V9 = IRBB11; V10 = IRBB13; V11 = IRBB14; V12 = IRBB21; V13 = IRBB53; V14 = IRBB59; V15 = IR24: V16 = PNA647F4-56; V17 = Gigante; V18 = TOG5681

Fig. 2:	Virulence status of Xoo isolates relative to country of origin. Country: C1 = Niger; C2 = Benin Republic; C3 = Nigeria; C4 = Burkina Faso; C5 = Mali; C6 = Guinea; C7 = The Gambia

Pta2 and Pta3) and Ptb pathotype into two subgroup pathotypes (Ptb1and Ptb2). Pta1 was made up of 10 isolates (XTG-47, XBF-20, XBF-21, XG-35, XG-33, XG-32, XBF-18, XM-23, XBF-22 and XG-31), Pta2 was 8 isolates (XG-36, XTG-46, XTG-41, XTG-42, XTG-48, XN-1, XN-2 and XN-3), Pta3 was 11 isolates (XB-10, XNG-13, XB-11, XNG-15, XNG-14, XBF-19, XM-25, XM-30, XG-40, XG-39 and XG-38), Ptb1 was 10 isolates (XG-37, XTG-43, XTG-50, XN-5, XN-4, XB-8, XB-9, XNG-12, XBF-17 and XM-24) and Ptb2 was 11 isolates (XM-26, XM-27, XM-28, XM-29, XG-34, XTG-44, XTG-45, XTG-49, XN-6, XB-7 and XNG-16) (Fig. 6).

Table 7:	Xoo virulence and variety resistance status
Country: C1 = Niger; C2 = Benin Republic; C3 = Nigeria; C4 = Burkina Faso; C5 = Mali; C6 = Guinea; C7 = The Gambia. Variety: V1 = IRBB1; V2 = IRBB2; V3 = IRBB3; V4 = IRBB4; V5 = IRBB5: V6 = IRBB7; V7 = IRBB8; V8 = IRBB10; V9 = IRBB11; V10 = IRBB13; V11 = IRBB14; V12 = IRBB21; V13 = IRBB53; V14 = IRBB59; V15 = IR24: V16 = PNA647F4-56; V17 = Gigante; V18 = TOG5681.R = Resistance (% lesion length less than 15 %); S = Susceptible (% lesion length greater than 15% ); Vr = Virulence

Fig. 3:

Genotype (cultivar) by environment (isolate) interaction effects on percentage lesion length using additive main effects and multiplicate interaction (AMMI) analysis. Genotype (G): G1 = IRBB1; G2 = IRBB2; G3 = IRBB3; G4 = IRBB4; G5 = IRBB5: G6 = IRBB7; G7 = IRBB8; G8 = IRBB10; G9 = IRBB11; G10 = IRBB13; G11 = IRBB14; G12 = IRBB21; G13 = IRBB53; G14 = IRBB59; G15 = IR24: G16 = PNA647F4-56; G17 = Gigante; G18 = TOG5681. Environment (E): E1 = XN-1; E2 = XN-2; E3 = XN-3; E4 = XN-4; E5 = XN-5; E6 = XN-6; E7 = XB-7; E8 = XB-8; E9 = XB-9; E10 = XB-10; E11 = XB-11; E12 = XNG-12; E13 = XNG-13; E14 = XNG-14; E15 = XNG-15; E16 = XNG-16; E17 = XBF-17; E18 = XBF-18; E19 = XBF-19; E20 = XBF-20; E21 = XBF-21; E22 = XBF-22; E23 = XM-23; E24 = XM-24; E25 = XM-25; E26 = XM-26; E27 = XM-27; E28 = XM-28; E29 = XM-29; E30 = XM-30; E31 = XG-31; E32 = XG-32; E33 = XG-33; E34 = XG-34; E35 = XG-35; E36 = XG-36; E37 = XG-37; E38 = XG-38; E39 = XG-39; E40 = XG-40; E41 = XTG-41; E42 = XTG-42; E43 = XTG-43; E44 = XTG-44; E45 = XTG-45; E46 = XTG-46; E47 = XTG-47; E48 = XTG-48; E49 = XTG-49; E50 = XTG-50

The occurrence and distribution of Xoo pathotypes varied among isolates country of origin (Table 8). Pta1, a virulence (Vr) pathotype, was known to exist in four countries (Burkina Faso, Mali, Guinea and The Gambia) with 20% occurrence, Pta2 (Vr) has 16% occurrence in three countries (Niger, Guinea and The Gambia) and Pta3 (Vr) has 22% occurrence in five countries (Benin Republic, Nigeria, Burkina Faso, Mali and Guinea) (Table 8). Ptb1, a mildly virulence (MVr) pathotype, was known to exist in all the seven countries (Niger, Benin Republic, Nigeria, Burkina Faso, Mali, Guinea and The Gambia) with 20% occurrence and Ptb2 (MVr) has 16% occurrence in six countries (Niger, Benin Republic, Nigeria, Mali, Guinea and the Gambia) (Table 8). Thus, in Niger the study revealed the presence of Pta1, Ptb1 and Ptb2 Xoo pathotypes, in Benin and Nigeria (Pta3, Ptb1 and Ptb2), in Burkina Faso (Pta1, Pta3 and Ptb1), in Mali (Pta1, Pta3, Ptb1 and Ptb2), in Guinea (Pta1, Pta2, Pta3, Ptb1 and Ptb2) and in The Gambia (Pta1, Pta2, Ptb1 and Ptb2) (Table 8).

Fig. 4:	Genotype (cultivar) by environment (isolate relative to country of origin) interaction effects on percentage lesion length using additive main effects and multiplicate interaction (AMMI) analysis

Fig. 5:	Relationship among rice variety and Xoo isolate country of origin as revealed by genotype (cultivar) by environment interaction effects on percentage lesion length using additive main effects and multiplicate interaction (AMMI) analysis

Fig. 6:	Analysis of Xoo isolates virulence as revealed by pathotyping using Additive Main effects and Multiplicate Interaction (AMMI) analysis

Table 8:	Xoo isolate group, virulence and distribution relative to country of origin
Pta = Pathotype a; Ptb = Pathotype b; Vr = Virulence; Mvr = Mildly virulence

Information on the existing population structure of the pathogen in a region can be useful in the identification and characterization of useful resistant germplasm (Choi et al., 1998). For example, Nelson et al. (1994) used knowledge of the Xoo population structure in the Philippines to select representative strains for screening rice germ plasm collections for resistance. The Additive Main effect and Multiplicative Interaction (AMMI) analysis was shown to be effective in understanding complex Genotype by Environment (GE) interactions typical of National Turfgrass Evaluation Program (NTEP) variety trials (Ebdon and Gauch, 2002). Interactions in such complex data sets are difficult to understand with ordinary Analysis of Variance (ANOVA). Genotype by environment interaction can be defined as the differential response of varying genotypes under changes in the environment. AMMI analysis used in this study has revealed diversity and extent of 50 Xoo isolates interaction among 18 rice cultivars that lead to the classification of Xoo isolates virulence into two major pathotypes (Pta and Ptb) that were responsible for unfavourable interactive conditions that lead to significant increase in percentage lesion length in all the rice cultivars. The existence of Pta1, Pta2 and Pta3 and Ptb1and Ptb2 subgroups pathotypes were likely due to mutations and interactions among isolates and strains that originally constituted Pta and Ptb pathotypes (Innes et al., 2001).

The movement of Xoo pathogens has important implications for the control of BLB disease. If migration is far reaching, the development and deployment of resistant germplasm would require knowledge about the structure and dynamics of distant, as well as local populations of the Xoo pathogen. Thus, it is important to understand pathogen migration and how it influences population genetic structure and potential for disease. Several examples of pathogen migration have been documented, including intercontinental movement of Phytophthora infestans (Fry et al., 1992) and Puccinia graminis (Burdon et al., 1982) and movement of Erysiphe graminis from continental Europe to England (Brown et al., 1991). In most cases, the arrival of immigrant genotypes has been associated with increased BLB disease problems. In the present study, Pta pathotype a virulence (Vr) type and Ptb pathotype a mildly virulence (MVr) type were known be to present in Mali, Nigeria, Benin, Burkina Faso, Niger, Guinea and the Gambia and Xoo virulence and interactive conditions were similar among isolates in these countries suggesting possible Xoo pathogen migration between these countries (Adhikari et al., 1995). However, such migration could be from germplasm exchange of contaminated seed among regional countries (Cheng et al., 1994). Since previous studies (Sere et al., 2005) were only conducted with Xoo isolates from Mali, the studies could not reveal Xoo pathogen migration and germplasm exchange of contaminated seeds across regional countries as revealed by the present study. The two major pathotypes (Pta and Ptb) of Xoo virulence obtained in this study has revealed information on Xoo virulence population structure in West Africa and this would provide wide useful information for selection and deployment of cultivars with durable resistance to BLB disease in West Africa.

ACKNOWLEDGEMENTS

We express our appreciation to International Rice Research Institute (IRRI) for providing the near-isogenic lines rice varieties used in this study.