**3. Hybrid rice technology**

Hybrid rice is the commercial rice crop grown from F1 seeds of a cross between two genetically dissimilar parents. This could only be possible through the use of a male sterility system due to the strictly self-pollination nature of the rice plant [4, 8]. Hybrid varieties exploit the phenomenon of hybrid vigor (heterosis) to increase the yield potential of rice beyond the level of modern inbred rice varieties. A yield advantage of 15–30% over conventional inbred varieties grown under similar conditions has been reported [14]. The hybrid rice technology concept dates back to 1964 in China. However, it was only after 1976, when a wild abortive pollen plant was identified in Southern China, did the idea begin to materialize [4, 8]. Since the expression of heterosis is confined to the first generation only, farmers have to buy fresh seeds every season to raise commercial crop. Since the hybrids yield 15 to 30% more than pure line varieties, farmers prefer hybrid seeds if the price is economically beneficial and seeds are readily available. Hybrids can offer biological intellectual property protection which attracts and encourages private-sector involvement in seed production research and development [15].

#### **3.1 Genetic basis of heterosis**

For a long time, two theories; the dominant and overdominance theories were put forward to explain the phenomenon of heterosis [14]. According to the dominant theory, hybrid vigor is due to the action and interaction of favorable dominant alleles. The overdominance theory on the other hand suggests that heterozygous loci are superior to homozygous loci. Thus, two alleles complement each other and there is overexpression of genes in the heterozygous state [16]. Recently, epistasis or interactions among loci has been recognized as a major contributor to heterosis [17]. Estimates based on mating designs of the relative magnitude of additive, dominance and epistatic components of variance indicate that the magnitude of epistatic variance is small compared to additive and dominance components. This is because statistical designs cannot predict epistasis. Using molecular markers,

Yu *et al*. [18] provided evidence of the importance of epistasis as a possible genetic basis of heterosis in rice. Since none of these theories exclusively explains heterosis, suggestions are to ascribe the phenomenon to a combination of the three models.
