Integrating genetic resistance with biocontrol against rice blast and drought.
Annual demand for rice in Liberia far outstrips local production, mainly as a result of rice blast and drought. The use of resistant cultivars has been widely considered as the most economical means of controlling the disease and mitigating the impact of drought stress (DS) on rice (Oryza sativa L.). However, despite the high variability of the rice blast pathogen (Pyricularia oryzae Cavara), efforts aimed at genetically managing the disease have focused mainly on the development of cultivars with vertical resistance, which has repeatedly failed soon after the release of new cultivars because new pathotypes evolved with matching virulence genes. The current research was consequently undertaken to devise breeding strategies aimed at durably controlling rice blast and mitigating the impact of DS on rice by integrating host plant resistance with biocontrol agents. Towards these objectives, separate studies were undertaken to appraise local perspectives of rice production in Liberia and to identify key rice varietal traits preferred by farmers, to characterize upland rice locally cultivated in Liberia for resistance to rice blast and tolerance to drought, to determine suitable environments for screening rice for drought tolerance and quantitative resistance to rice blast, to determine the gene action and combining ability for the inheritance of key rice blast resistance traits, to determine the effects of a combination of ethephon and gibberellic acid (GA3) on the hybridization of rice, and to investigate the effects of two strains of Trichoderma harzianum on blast resistance and drought tolerance of rice. Results of both factor and conjoint analyses of farmers’ selection criteria and trait preferences revealed that farmers considered a combination of specific agronomic and morphological characteristics, including grain yield, grain quality, stress resistance and post-harvest traits when selecting rice varieties for cultivation. Farmers expressed a strong preference for hardy, early maturing varieties with intermediate stature, which generate soft but discrete (loose) grains when boiled. Pathogenicity tests involving 38 upland genotypes, inoculated with three pathotypes of P. oryzae under controlled environment (CE) conditions, revealed that the blast resistance of selected varieties of upland rice from Liberia was characterized by the occurrence of both race-specific and race-non-specific patterns of resistance. Additive main effects and multiplicative interaction analysis revealed significant (p < 0.05) genetic variability among the genotypes for the number and size of sporulating lesions per plant. Rice plants subjected to intermittent DS both under CE and field conditions exhibited larger and more lesions with sporulating centres than those grown under well–watered conditions, thereby indicating the value of managed DS for evaluating rice for horizontal resistance to blast disease. Studies on drought tolerance of 22 of the upland rice genotypes showed that drought imposed for 21 days during the booting-to-grain-filling stage of development severely diminished grain yield (GY) and its components, and that the difference in grain yield between plants subjected to DS during the tillering stage and those grown under continuously well-watered conditions was not significant (p = 0.05). This study identified the booting-to-grain-filling stage as the primary phenological stage for improving drought tolerance of upland rice from Liberia. Principal component analysis showed that the 22 genotypes exhibited the greatest variation for GY, tiller productivity (TP) and chlorophyll content index (CCI) under well-watered conditions, and for leaf rolling, GY, CCI, spikelet fertility (SF), and TP, under drought imposed during the booting-to-grain-filling stage. Simple correlation and path analyses identified high levels of biomass at heading (BM), stomatal conductance (SC) and SF as important selection criteria for achieving high GY under well-watered conditions, and TP, BM and 1000 grain mass as the key criteria for selecting high yielding genotypes under DS occurring around the booting-to-grain-filling stage. Field screening for drought tolerance of rice over two seasons, using up to 24 genotypes, revealed that a novel technique using a plastic mulch soil cover system, designed at the University of KwaZulu-Natal, was effective for excluding rainfall from the plants, and is therefore a valuable tool for facilitating controlled drought screening of rice plants during a season with normal rainfall levels. Evaluation of eight blast susceptible genotypes and their F2 progenies derived from a full diallel mating system indicated that both additive and non-additive gene actions were involved in the inheritance of LN, LS and area under the disease progress curve (AUDPC), which are three key parameters measured to quantify the levels of horizontal resistance against rice blast. However, the genes with additive effects were more important, since the level of general combining ability (GCA) for each trait was higher than that of specific combining ability (SCA). These results indicated that future breeding of Liberian rice varieties aimed at increasing the level and durability of rice blast resistance should emphasize recurrent selection for reduced LS and lower LN and AUDPC, using genotypes that exhibit negative GCA effects for these traits as progenitors. Greenhouse studies with plant growth hormones showed that application of ethephon at 4000 to 6000 μL L-1 during microsporogenesis, combined with gibberellic acid applied at 90 or 150 μL L-1 during anthesis can provide for reliable male sterilization of flowers without diminishing female fertility. This can be used to facilitate the large scale cross-pollination of rice genotypes and their progeny, which is essential for recurrent selection breeding programmes. The results of both CE and field studies demonstrated that two biocontrol products, Eco-77® and Eco-T®, containing different strains of T. harzianum, were useful for managing rice blast and the effects of drought on rice, respectively. Notably, the impact of the biocontrol treatments tended to increase with the levels of host plant resistance. Overall, the results of this research show that new high yielding cultivars with durable resistance to rice blast and tolerance to drought can be bred in Liberia by adopting a recurrent selection scheme, using locally adapted genotypes as progenitors, and evaluating breeding lines under managed drought stress and well-watered conditions. Integrating the current biocontrol agents with improved host plant resistance will be important for increasing rice production in Liberia and other regions where rice suffers from rice blast and drought.