| dc.description.abstract | Hybridization occurs in many crops that are sexually compatible with their wild relatives, leading to crop-to-wild pollen gene flow. Gene flow may result in evolution of super weeds, loss of genetic resources or even extinction of endangered species. In this study, gene flow between cultivated rice (Oryza sativa) and wild rice (0. langistaminatat was studied in Kenya. For hybridization to be possible, dormancy breaking methods for 0. longistamianta were investigated. Incubated and non-incubated hulled seeds were germinated in 1) Gibberellic acid (GA), 2) Hydrogen peroxide (H202), and 3). Water (H20) at two temperature regimes (3 (C and 2SoC). A similar set up for dehulled seeds was performed. Germination occurred in hulled but not in dehulled seeds. Highest germination percentage occurred in Gibberellic acid for all treatments. Significant (P<O.OS) differences were observed within treatments but not (P>O.OS) between treatments.
No interaction effects (P>O.OS) were observed between germination media and incubation and also with temperature. Highest germination was observed at higher temperatures (31 DC).
Dispersal distance of rice pollen was investigated in Tana River district in the Coast Province of Kenya. Seeds of 0. sativa were planted in a SOm diameter experimental design. Pollen traps placed at different heights (1.2 and 1.8 m) and distances (North, East South and West) from the pollen source were used. Pollen count decreased rapidly with increasing distance from the pollen source. Differences (P<0.05) in pollen count were observed in different directions but no difference (P>O.OS) between the two heights. The highest pollen count was observed between 11 :00 am and 12:00 noon.
F, hybrid seeds were generated from crosses between the two Oryza species under screenhouse conditions at Mwea Irrigation Scheme. Data on growth patterns, morphological traits and yield were scored in parents and hybrids. Only 6 % of the F 1 hybrid seeds set occurred with O. sativa as paternal plant but no seeds were formed in reciprocal crosses. Morphological differences (P<0.05) were observed among the plant types. In hybrids, yield correlated (r=+0.843) with flag leaf length but in O. /ongistaminata yield correlated with plant height (r=+0.767) and panicle length (r= +0.664). In 0. sativa, yield correlated with plant height (r= +0.741) and panicle exsertion. (r= +0.854).
In order to confirm that the hybrids identified by morphological markers were true hybrids, molecular characterization was performed. DNA extraction and Polymerase Chain reactions (peR) were conducted at the University of Nairobi. 22 rice simple sequence repeat (SSR) primers were screened to select markers that could identify the two parents and their F 1 hybrids. Six markers were found to be polymorphic across the samples. Polymorphic information content ranged from 0.223 for RM234 to 0.354 for RM280. A total of 12 alleles were detected and the number of alleles per primer was two. Through genotyping, F 1 hybrids were identified among the parents.
Generally, the main findings ofthe study were: 1) Dehulling ofO. /ongistaminata seeds was found to be necessary for effective germination., 2) Hybridization can occur between cultivated rice (0. sativa) and wild rice (0. longlstaminata), 3) To minimize such hybridization, a minimum isolation distance of more than 250 m between the two species is recommended, and 4) Use of morphology followed by molecular characterization is recommended for identification of' F, hybrids in the field. | en_US |
