Pigeonpea Improvement at ICRISAT______________________________

The problem
Despite the proven potential of hybrid cereals, few, if any leguminous food crops are grown as hybrids. This is not for lack of trying. Cross-pollination systems are usually difficult to develop in leguminous species, which tend to exhibit very limited natural outcrossing, and seed multiplication rates are low, resulting in un-economic costs of hybrid seed production. For example, although heterosis levels of 25% had been reported for pigeonpea since the 1950s on an experimental basis (Solomon 1957), this knowledge was not exploitable in the absence of a practical hybrid seed technology.

The solution
A major and innovative contribution of ICRISAT was to challenge the assumption that hybrid systems were not achievable for food legumes. It took 15 years of research and development to prove this dogma wrong - but the world's first pigeonpea hybrid variety, ICPH 8, finally reached farmers' fields in 1991 (ICRISAT 1993, Saxena et al. 1996).

By challenging conventional wisdom, ICRISAT had opened up a whole new vista for pigeonpea improvement. The orphan crop had become the pacesetter. How was it done?

First, studies were carried out to understand the dynamics of pollination in pigeonpea (Saxena et al. 1990). Insects actively distribute pigeonpea pollen, but this only occasionally results in outcrossing, because the anthers dehisce

a day before the flower opens, in most cases self-pollinating the stigma before exsertion. ICRISAT gambled that with male sterility, this self-pollination could be prevented, making 100% outcrossing achievable.

In 1974, an extensive search for sterility sources was carried out on a trial of 7,216 accessions from the ICRISAT germplasm collection, and 124 lines derived from crosses with wild relative species. Seventy-two plants with aberrant floral characteristics were found, and grouped into five types.

The "translucent anther" type (found within accessions ICP 1555 and ICP 1596) appeared most promising, because it was devoid of functional pollen grains and it was easy to visually recognize in the field, which is critical when rogueing offtypes. A second genetic source was later identified in Australia, and dubbed "brown arrow-head" for the appearance of its anthers. Genetic analyses found that the two sources (ms1 and ms2) are non-allelic, and both are under single-gene recessive control.