Chapter 5: Resistance to Rice yellow mottle virus (RYMV) in a rice germplasm collection from
5.5 Discussion
The majority of the rice accessions in the present study were found to be susceptible to the different RYMV isolates. Although direct comparisons of RYMV resistance of the accessions used in the current study with previous studies conducted elsewhere are not reliable for different reasons (due to differences in sample sizes, species composition, genetic differences among accessions, difference in RYMV isolates used for disease evaluation, the experimental designand the method of inoculation), the use of the same reference varieties such as Gigante, IR64 and TOG5681 make comparison possible. Overall, the present study confirmed the high susceptibility of most of O. sativa accessions compared with most O. glaberrima accessions. Only seedling stage resistance was found in O. sativa accessions, becausenone of the O. sativa accessions were found to be highly resistant until 49 days after inoculation while a few accessions of O. glaberrima remained resistance throughout the disease scoring period. The susceptibility of most O. sativa varieties to RYMV has been reported by Rakotomalala et al. (2008) after screening a rice collection from Madagascar. In addition to this, none of the accessions from Niger and Mali were found to have comparable level of resistance as that of O. sativa varieties Gigante and Bekarosaka. Nevertheless, the indica accession DS86 expressed a high level of resistance to RYMV isolate B27, an intermediate phenotype to Ng177b, Ng122 and Ng144.
However, it was susceptible to RYMV isolate BF1 at 21DAI. In contrast to this, a popular modern irrigated variety, IR1529-680-3 (coded as TY49) from IRRI, and a farmer variety TY51 named "Kassimo" or "Waihidjo", both covering more than 80% of the rice irrigated areas of the country were found to be susceptible to RYMV during all the trials. Indeed, the susceptibility of the IR1529-680-3 had been reported earlier (Reckhaus and Adamou, 1989), more than 10 years after it was released. This variety is suspected to be the main cause of the first outbreak of the RYMV epidemic in Niger in 1984 (WARDA, 2001). The Kassimo variety occupied 60% of rice cropping area in Tillabéry in 2008 (data from ONAHA- Tillabéry, 2008) and was recently said as highly susceptible (Issaka, et al., 2012). Therefore the combination of these two varieties could explain the recent importance of RYMV in the region (Chap. 2). The relatively larger proportion of O. glaberrima accessions with high levels of resistance compared to O. sativa was in agreement with several other studies (Coulibaly et al., 1999; Ndjiondjop et al., 1999; Thiémélé et al., 2010). Indeed, the gene RYMV1 has shown a high allelic variability in O. glaberrima compare to O. sativa, whiles the resistance gene RYMV2, was found only in O. glaberrima (Albar et al.,
164
2006; Thiémélé et al., 2010). Thiemélé et al. (2010) predicted the occurrence of new resistance genes or alleles in the African rice. Although some of the resistant accessions found in the present study were similar to that of TOG5681 and TOG5672, they could be useful in breeding, in terms of other phenotypic traits. Indeed, TOG5681 has already been used as a donor parent to the lowland NERICAs developed by AfricaRice (Sié et al., 2008). Thus its genetic and phenotypic potential are being used, while any of the new resistant accessions have not yet been used in any breeding program. However, caution should be taken when using this kind of monogenic resistance because of the high selection pressure it applies on RYMV populations and
"fortuitous" molecular incident, such as a single mutation, can lead to the emergence of new virulent strains as it happened with the resistance gene Cf4 of tomato to the fungus Cladosporium fulvum (Joonsten et al., 1994) and rymv1-2 in rice (Hébrard et al., 2006). Moreover, molecular evidence revealed that RYMV's evolution is as faster as animal viruses (Fargette et al., 2008b).
Therefore priority should be given to combining monogenic resistance with partial or intermediate resistance. Plants with partial polygenic resistance cause a delay in the appearance of symptoms. The development of the disease will therefore be reduced and subsequently its re- transmission to neighbouring plants and the increase of the epidemic in the field (Van Der Plank, 1966). The interaction between the vertical resistance identified in the RYMV-rice pathosystem suggest caution when using this type of gene in breeding programmes, because of the risk of rising of virulent pathotypes (Robinson, 1995).
The use of several pathogenically different virus strains was helpful in identifying such accessions. Some of our O. sativa varieties and most of the O. glaberrima varieties could be used as parents to develop comprehensive resistance to RYMV. Another interesting point suggested by the present study was the non-significance of the virus isolate on both the disease score and the reduction of chlorophyll content of leaves over time. In addition, the likelihood of one virus strain inducing more or less symptoms than another, were also low over time. This indicates that short periods should be used for disease score assessment post-inoculation. The effect of the time of infection of three rice varieties, with different levels of resistance to RYMV were assessed in a screenhouse by inoculating rice plants at their seedling, tillering, booting and flowering stages (Onwughalu et al., 2011). The observation period varied between inoculation and harvest. The level of sterility of spikelets was not related to the period of infection, while the difference in the mean yield and the plant height were significant for plants inoculated at the seedling stage, even
165
in the susceptible cultivar. Ndjiondjop et al. (1999) observed symptoms from 14 to 62 days post infection while Thiémelé et al. (2010) recorded symptoms up to 12 weeks post inoculation.
However, in this series of experiments at 14 DAI the disease score, as well as the virus content assessed by ELISA, were able to separate the highly resistant Gigante and TOG5681 from the susceptible IR64 and the partially resistant Azucena.
The response to inoculation depends on the RYMV isolates used. Earlier studies reported that RYMV isolates from Niger belong mainly to the S1 serological group, but are related to a Central African lineage, although the country is located in West Africa (Traoré et al., 2005).
Ng117b and Ng122 appear to be more aggressive than Ng144, because of their ability to induce more severe symptoms in the susceptible IR64. However, the RYMV isolate BF1, belonging to the S2 group, was the most aggressive, not only based on its ability to damage IR64, but also the rapidity with which symptoms appeared. Presently a country-wide study of the genetic and pathological structure of RYMV is ongoing and the first results indicated that both S1-West Africa and S1-Central Africa strains of RYMV are present in the Niger, within four major pathotypes (Issaka, 2011; Issaka 2012). Additionally, "T" type strains, many of which are virulent to the resistance allele rymv1-3, were found predominantly in Niger, while a few "E" types that are virulent to rymv1-2 were also found (Traoré et al., 2010). Most of the few "T" and "E"
isolates virulent to rymv1-2, were also virulent to rymv1-3 in Niger (Issaka, 2011)
The secondary traits measured in this study, namely chlorophyll content of leaves, plant height, and AUSPC, have also been successfully used by several authors in the screening for resistance to RYMV (Albar et al., 1998; Coulibaly et al., 1999; Séré et al., 2008); as well as, in the management of nitrogen levels in rice (Esfahani et al., 2008). However, the measurement of the chlorophyll content of leaves was subject to variations due to the position of the SPAD-meter and the number of points taken on the plant leaf for the same plant, the age of the plant used for the notation, and the intensity of the sunshine during the period of the recording. This technology needs a standardization protocol to solve these problems.
The screening of a large collection of rice accessions from Niger and Mali revealed the overall vulnerability of most rice varieties to RYMV in these countries. In particular, most rice production is being undertaken in irrigated and lowland areas, cropped mainly with O. sativa
166
varieties, most of which are highly susceptible to RYMV. However, the survey has unlocked the hidden potential of the few remaining traditional landraces and wild relatives of the country for breeding for comprehensive resistance to RYMV. Those findings, coupled with the pending information on the structure of the viral population will make a great impact on rice breeding for resistance to RYMV in Niger. Furthermore, the sequencing of the rice genome and resulting applications, such as genome-wide selection (GWS) and marker-assisted recurrent selection (MARS), offer great opportunities to successfully combine both monogenic and polygenic modes of resistance in rice breeding programs.