Irrigated and Rain-Fed Lowland Rice Breeding in Uganda: A Review

Since introduction of rice into Uganda in 1904, improvement of the irrigated and rain-fed lowland types was undertaken to address a number of production and quality constraints in three consecutive and overlapping phases. The initial phase was achieved through evaluation of introduction, selection of promising lines and subsequent release of the selected lines for production by the farmers. In the second phase, genetic potential of traits and characteristics of interest were analyzed and used to guide selection of suitable parents for hybridization and the third phase employed genotyping approach in screening and selection of the parental lines and the segregating populations to enhance the breeding efficiency for the traits of importance. Simultaneously, the key production constraints addressed included resistance to rice yellow mottle virus (RYMV), rice blast, bacterial leaf blight and narrow leaf spot diseases as well as submergence tolerance and cold tolerance. The quality traits considered for the improvement alongside the grain yield parameters were the grain aroma, amylose content, shape and size. These interventions have resulted into release and wide adoption of seven rice varieties in Uganda besides several breeding lines which have informally diffused into different major rice production agro-ecology. Subsequently, it can be concluded that a substantially strong and functional breeding platform for rice in Uganda has been established.


Introduction
Uganda is a tropical country that lies astride equator, but with modified climatic conditions due to large water bodies and high peaked mountains. The altitude varies from 614 to 5,111 metres above sea level (masl) with much of the rice production areas falling within an altitude ranging from 1,000 to 1,400 masl. The least and the highest recorded temperature within the rice production agro-ecology is 8°C and 38°C, respectively, whereas on the basis of the average for the entire country, lowest temperatures range between 10°C and 17°C and highest from 23°C-25°C. The annual rainfall intensity in Uganda varies from 600 mm to 2,500 mm with much of the country receiving between 900 mm and 1,800 mm of rainfall. Owing to the diversity in the climatic conditions, rice production ecologies in Uganda are component is eliminated in double haploid technology, implying higher irritability of these economic traits in rice [10]. It was reported that in the case of F 3 and F 4 , both additive and dominance gene effects contribute to phenotypic differences between individuals, which tends to mask the expression of the desired traits [11] whereas variation in doubled haploid progeny is only due to some environmental effects.
The elite breeding lines were then evaluated in replicated yield trials and multilocation testing to identify promising lines for onwards evaluations on-farm and testing in National Trials following the National Variety Release guidelines and subsequently the promising lines basing on the key preferred production and quality attributes are nominated to the National Variety Release Committee of Uganda for approval and release as new rice varieties.
Key attributes of preferred rice for production are highlighted below: 1. Agronomic traits: This target breeding for high yielding and preferred plant type such as high yielding rice exhibiting 20% yield increase compared to the existing rain-fed lowland and irrigated rice varieties, but with plant height ranging from 90 to 110 cm and maturity period of between 90 and 135 days after planting.
2. Biotic stress: The focus in regards to the abiotic stress has been development of new climate resilient varieties tolerant to abiotic stresses such as drought, iron toxicity, low nitrogen (high nitrogen use efficiency), submergence and cold stresses.
4. Grain quality: Development of rice with quality attributes preferably with low amylose content for diabetic and elderly persons of less than 20%, noodles (>29%), high milling recovery (>65%) and preferred cooking quality of nonsticky and aroma characteristics. Preference for aromatic, non-sticky, whole grain and white rice grains traits are also commonly preferred by most rice consumers in Uganda [7,12]. 6. Produce quality breeder and foundation seed. This is one of the prescribed roles of a breeder to ensure availability of quality seeds in the right quantity either directly or through the agro-input dealers to the farming communities.

Introduction of rice varieties to Uganda
Irrigated and rain-fed lowland rice varieties were introduced into Uganda in two phases namely, first from 1921 to 1970 and second during 1971-2020. In the first phase, a total of eight irrigated and rain-fed lowland rice varieties, specifically Jaggery, Cakala, Matama, Kawemba, Kigaire, Seena, SUPA LOCAL and Bungala were introduced and grown ( Table 1). On the basis of aroma, all the eight rice cultivars except Bungala were aromatic types.
In the second phase covering the period from 1971 to 2020, six released varieties and up to 70 unreleased but informally released rice varieties were commonly being grown by the farming communities. The 70 informally released varieties are classified into two different groups according to their generations. The first are called the K-series of rice introduced from China under technical assistance program, while the second are introductions from major breeding centers detailed in Table 1. Both groups are modern varieties with high-yielding capacity and tolerance for various biotic and abiotic field stress conditions. The K-series is an acronym of Kibimba lines, named so probably due to the series being grown then in the Kibimba government rice irrigation scheme. Fortunately, most of the K-series had a desirable combination of intermediate amylose content and intermediate gelatinization temperature and a notable variety is IR64, which has been widely accepted as a high-quality rice. Later, however, there were several devastating stresses that undermined the importance of these varieties for which reasons more introductions from other centres were requested and received.

Rice diseases
Rice yellow mottle virus (RYMV) disease is apparently the most serious disease of rice under irrigated and rain-fed lowland conditions in Uganda. In order to identify lines with resistance to RYMV disease, 934 rice lines were screened in the years 2015, 2016, 17 and 2018. These list excluded IR 64, K 34, K 38, K 85 and KOMBOKA (susceptible control), Gigante from AfricaRice (resistant check), Namche 2, NERICA 4 (MET P71), NERICA 8 (MET P72) and WITA 9 (local resistant check) that were included in each set [20,21]. Each line was sown in the field at high plant population of 10 grams in a land area measuring 20 cm x 20 cm in size. Mechanical inoculations were carried out on seedlings (10 plants test line) at 3 weeks post germination as described by Thouvenel and Fauquet [22]. Symptom appearance was monitored on daily basis to assess the stage of disease initiation and thereafter, disease severity was scored using a scale of 1 (no symptoms) to 9 (severe symptoms) [23,24] at 45 days post-inoculation (DPI). Of the 934 lines evaluated, a total of 54 either highly resistant or just resistant were identified ( In 2018, another set of 112 germplasm was screened for rice yellow mottle virus disease (RYMVD) resistance. The germplasm comprised of O. barthi interspecific lines generated from crosses between Oryza sativa L. and O. barthi. These breeding lines were selected for their high yield potential, resistance to diseases and other desirable culinary qualities. Seventeen promising genotypes were identified, among which six comprising of [ARS126-3-B-1-2 (11), ARC36-2-P-2 (2), ARC39-145-P-2 (5), Gigante, IRL 2 (GP 54) and IRL 4 (69 GP 54)] were highly resistant to RYMV disease. List of the 11 resistant lines for RYM derived from O. barthi crosses are shown in Table 2.
In 2019, a total of 307 lines including the 247 KAFACI lines were introduced into East African Regional Rice Research and Training centre at NaCRRI, Namulonge-Uganda and each of the introduced seed samples were divided into three parts. The first part was planted using a single row plots in the screen house at NaCRRI. Each row had 10 cm spacing with 15 hills and the rows were. Spaced 15 cm apart. A prepared RYMV inoculate was used to inoculate 5 plants in each row following procedure described by [22]. Symptom appearance was monitored on daily basis to assess the stage of disease initiation. Also, symptom expression post inoculation (DPI), severity on weekly intervals from 21 DPI, through 28 DPI, 35 DPI and 42 DPI were collected. We also took record of plant height for inoculated and noninoculated plants and the percentage reduction for each of the lines were calculated The results revealed that 3 lines ofIR64/rymv1-2, IR64/rymv1-3 and IR64/RYMV3 showed severity score of 1 on score 1-9; 14 lines showed severity score of 3. These 7 lines are; SR34574-HB3565-284, ARS1958-1, SR34574-HB3565-285, SR34574-HB3565-290, HR32068F1-4-20-1-6-3-2, HR32068F1-4-20-1-6-4-2 and HR32068F1-4-20-1-6-4-3, which were also found to exhibit resistance to all other diseases assessed, namely rice blast, BB, sheath rot and narrow leaf brown spot. A total of 39 lines showed severity score of 5 while up to 57 lines showed reduction in plant height by at most 30%. Further analysis based on a rating scale for Susceptible (7)(8)(9), Medium resistance (4-6) and Resistance (0-3) indicated line, namely, HB4057, rymv1-2, rymv1-3, RYMV3, Hannam, NamChe-2 and Ungwang as the resistant donors.
In respect to bacterial leaf blight (BLB) disease, a total of 18 isogenic lines developed for BLB disease were screened in three hotspot areas of Namulonge-Wakiso, Olweny-Lira and Kibimba-Bugiri districts in Uganda. The results revealed IRBB21 (Xa 21) as the most resistant line in all the three locations followed by IRBB8 [27], implying that lines with these genes could be used to pyramid for multiple stress resistance in our breeding programme.
In another study, 32 lines were screened for resistance to BLB and six other lines, namely, CT 12, WITA 132 x NERICA 14, NERICA 10, NERICA 4 and NERICA 1 were reported resistant to BLB [27]. These four lines could be donor parents in breeding for BB resistance as currently, several rice lines have shown resistance to leaf bacterial blight ( Table 3).
Rice blast disease has been mentioned in several articles as a major constraint in rice production in Uganda [2,3,5,6]. Accordingly, 450 lines were progressively screened in 2014, 2015, 2016, 2017, 2018 and 2020 for sources of resistance to the rice blast disease. In 2014, a total of 50 lines introduced from South Korea though the KAFACI were screened alongside a resistant (IR-64) and a susceptible (NERICA 1) checks. These were the breeding population from a cross of an African cultivated rice, Oryza glaberrima of Niger Delta origin and Milyang 23, a Tongil-type Korean rice variety and a total of 29 lines were resistant to rice blast [26] (Table 4). 11 Irrigated and Rain-Fed Lowland Rice Breeding in Uganda: A Review DOI: http://dx.doi.org /10.5772/intechopen.97157 Another set of 46 rice genotypes introduced from South Korea though the KAFACI were screened for resistance to rice blast under screen house condition and in the field at NaCRRI Uganda in the year 2015 [28]. The screening exercise involved infecting and selecting the infected rice plants and by observing the symptoms on the leaves based on the rice blast identification guide. Data on leaf blast severity, lesion size, area under disease progress curve (AUDPC) for leaf blast severity and lesion size, panicle blast and yield were collected on five randomly selected plants in the field and on three plants in the screenhouse from each plot according to the standard evaluation system of rice [24]. Results revealed that genotypes SR33859-HB3324-133, SR33859-HB3324-93 and SR33701-HB3330-78 were highly resistance to rice blast and had good performance for yield [28] (Table 4).
In a related study aimed at understanding transmission of genes for resistance to rice blast, it was found that both additive and non-additive effects contributed to transmission of resistance genes for rice blast disease to the progenies. The inheritance of rice blast resistance has been indicated to be mainly controlled by additive gene effects, besides a small influence of a non-additive effects [28].
In an effort to combat brown spot disease caused by Biplaris oryzae pathogen, a total of 100 lines were screened for resistance to rice brown spot in 2017 and 2018 at Namulonge. The results showed that 18 lines were rated as highly resistant, 52 resistant, 27 moderately resistant and three lines including the checks were susceptible [29]. The list of the highly resistant lines recommended for further breeding work is presented in Table 5.
Further studies where F 2 progenies from the crosses involving parents with distinct phenotypic classes of brown spot revealed information of segregation ratios for the different crosses. In particular, crosses TXD 306 Â NERICA 4, NERICA 1 Â NERICA 4, NERICA 4 Â NERICA 1 and E 22 Â PAK conformed to the 3:1 ratio, suggesting the presence of at least one gene showing dominance [30].
Another study conducted to identify lines resistant to bacterial leaf streak (BLS) identified three lines of NERICA 1, NERICA 6 and IURON 2015-1 as highly resistant to the pathogen causing bacterial leaf streak [31].

Insect pests
A study was also conducted to identify rice lines resistant to African rice gall midge (AfRGM) and 20 rice genotypes with diverse breeding background were evaluated for the resistance to AfRGM under controlled conditions in a screen house and under the field conditions at NaCRRI, Namulonge -Uganda [32]. Infestation was done in accordance with the method use by Ogah [33] where 3 females and 2  [27]. SCA effects were observed in the crosses from NERICA 6 X E 22, NERICA 1 X K 85, NERICA 1 X KOMBOKA and NERICA 4 X KOMBOKA. Low (desirable) GCA values were witnessed in the case of genotypes of NERICA 6, NERICA 1, NERICA 4 and MET P-7, indicating the importance of the parents in contributing resistance towards AfRGM in rice.
In another study to identify sources of resistance to stalk-eyed fly pest (D. longicornis) in rice plants, four out of eight lines namely, NERICA 4, TXD 306, K 85 and NM7-22-11-B-P-1-1, showed high resistance to stalk eyed fly upon screening on-station at Namulonge in 2015 [34]. These four-high resistant (HR) lines were crossed with moderately susceptible lines NERICA 1, NERICA 6, Namche 2 and PAKISTAN in a North Carolina II mating design to determine their combing abilities for the insect pest resistance. The results showed that NERICA 4 and K 85 were good general combiners for resistance to the pest. The crosses Pakistan Â TXD 306 and NERICA 1Â NM7-22-11-B-P-1-1 were identified as promising lines for advancement. Further analysis revealed that the stalk-eyed fly in rice seems to be controlled both by additive and non-additive genes, thus selection at early Source: Marco et al. [29]. generations (F 1 and F 2 ) would not be effective. Therefore, selection can be appropriately delayed to a later generations, between F 4 and F 6 . Advancement of selected breeding lines (Pakistan Â TXD 306 and NERICA 1Â NM7-22-11-B-P-1-1) is, therefore, recommended for further evaluation for resistance to the stalk-eyed fly in later generations [34]. The parental lines NERICA 4 and K 85 are recommended as good general combiners and could be used as the donor parents in the breeding programme for the pest.

Abiotic stresses
Further, an investigation aimed at identifying rice lines with tolerance to cold was conducted and 41 lines at panicle initiation growth stage were subjected to controlled environmental condition of 17°C for 30 days prior to screening and the results revealed three lines, namely, Yunertian, Yunkeng and Zhongeng to exhibit tolerance to cold stress. Furthermore, a study was carried out to determine mode of transmission of genes resistance to cold stress and the crosses from Agoro x Zhongeng, TXD x Zhongeng, 1189 x Zhongeng, 1052 x Zhongeng, TXD x Yunertian and 1189 x Rumbuka presented high and positive specific combining ability (SCA) effect to cold stress at seedling stage indicating that the crosses could be used in selection of cold tolerant lines.
Submergence is a salient yield decimating factor in rice partly because water control and field level under irrigated and rain-fed conditions is weak. In order to address this challenge, a total of 29 rice genotypes were screened for submergence tolerance. Of these, six genotypes namely, O. barthi interspecific lines which were obtained from a cross of O. barthi and O. sativa, where O. glaberrima is a monocarpic annual derived from O. barthi. Two are released varieties (Namche 5, Namche 2) and six are candidate lines for release (ARS 126-3-B-1-2, ARU1189, ARU1190, ARU1191, E20 and E22) are potential candidate varieties for rain-fed lowland condition [35]. Evaluation of submergence tolerant rice genotypes following the IRRI standard protocol revealed a significant difference in seedling height assessed immediately after submergence stress. The genotypes were screened by submerging 14 days old seedling at 100 cm water depth for 14 days and another set at 45 cm water depth for 14 days following IRRI standard protocols. The study revealed four rice genotypes of Swarna, IRRI SUPA 3, KOMBOKA and SUPA 5 to be tolerant to submergence at 45 cm water depth for 10 days, with at least 85% survival rates. While varieties Swarna, SUPA 5, IRRI SUPA 3, KOMBOKA Mahsuri and IR 64 showed stable survival rate at both water depths with ≥75% survival rates.

Grain quality
Aroma in rice is a trait of high economic importance, thus are highly valued by consumers and thus commanding higher prices compared to the non-aromatic genotypes. However, the popular global varieties, namely, Basmati of India and Pakistan origins could not be adopted in Uganda because they succumbed to multiple biotic stresses rendering them not suitable for production in Uganda. In response to this challenge, the rice breeding program in Uganda, therefore, have employed two intervention strategies. The first was to screen all rice germplasm for aroma and initiate breeding program for improvement of aroma characteristics and through cooking and leaf sample testing, 39 aromatic rice varieties were identified ( Table 6). In the second study, screening was conducted based on 2-acetyl-1pyrroline (2-AP) concentration and the genotypes MET 3, SUPA 1052, Namche 1, ART-4 and BASMATI 370 exhibited not only high, but also stable 2-AP levels. parents identified with strong aroma characteristics could be used in the subsequent breeding program for aroma characteristics. Amylose level influences grain cooking quality and therefore rice with high amylose content are not preferred, for example in Uganda, rice with intermediate amylose level ranging from 15-22% are the commonly grown varieties. Therefore, in a bid to maintain preferred amylose content within the Uganda's rice collections, a study was under taken to screen 60 lines for amylose content for two seasons in 2018 [38]. Of the 60 lines screened, seven lines consistently were of intermediate amylose content (AC), namely Namche 1 (21.84%), P62H17 (20.86%), Namche 1 (21.84%), Namche 2 (16.74%), Namche 3 (14.64%), Namche 5 (22.77%) and ARU 1190 (27.86%) in both environments. A study to understand the mode of transmission of genes for amylose content revealed that six crosses, namely, 1052 x Suparica, 326104 x NERICA 4, 1052 x Namche 2, Namche 2 x Namche 3, Namche 1 x NERICA 4 and Namche 3 x NERICA 4 with significant (P ≤ 0.05) negative SCA effects indicating that there was reduced AC% in the crosses whereas the remaining seven crosses with positive significant SCA effects for amylose content indicated increase in the AC% of the progenies in these crosses.

Variety release
In light of the challenges of rice production and increasing demand for climate smart agriculture, varieties that are tolerant to known biotic and abiotic stresses were developed and released in Uganda (Table 7). Overall, 7 rainfed lowland rice varieties were released in addition to existing local rice varieties. Of the 7 varieties 2 were aromatic and 5 non-aromatic varieties [14,39].

Current focus
This information will guide selection of parents to use in rice improvement in the country. In the development of improved rice varieties, core sets of population are critical. Identified SNP markers are accelerating this process. Currently, rice germplasm available are being genotypes for presence of known genes of importance in rice breeding. Over 50 SNP markers covering major biotic and biotic, grains quality, yield and physiological traits are the current target are being used on the Uganda germplasm. With SNP markers being developed already aiding the process of selecting core populations for breeding and accelerating selection of promising lines, we believe that efforts to identify more SNPs in populations that show presence of genes through morphological but not positive under current SNP panel be given urgent attention. This will provide broad accumulations of preferred genes at genome level for O sativa. This is critical considering that core populations may

Conclusion
This paper provides information on the trends in the development of irrigated and lowland rice in Uganda. It reveals that there are more unreleased rice genotypes under cultivation than the released varieties. This observation points to the fact that rice improvement and variety release is a recent development in the country. Also, that more effort has been in screening introductions and conducting adaptation trials. These efforts contributed to selection of widely adapted genotypes that became accepted in the rice breeding program in Uganda. The panel of adapted lines have diverse parental background that includes Oryza barthii, Oryza glaberrima and the parent O. sativa. These lines will form basic germplasm for use in integrating the classical breeding to molecular biology led breeding.