Rice (Oryza sativa L.) is the most important crop planted in Cambodia. It is largely grown under unpredictable rainfall conditions that generally encourage the establishment and growth of many weed species at high density. Of these, six species are commonly found nationwide and have a considerable negative impact on national rice production. They are awnless barnyard grass [Echinochloa colona (L.) Link], barnyard grass (Echinochloa crus-galli (L.) P. Beauv.), small umbrella sedge (Cyperus difformis L.), two-leaf fimbristylis (Fimbristylis miliacea (L.) Vahl), water primrose (Ludwigia octovalvis (Jacq.) Raven), and goose weed (Sphenoclea zeylanica Gaertn.). Weeds of this kind, within the rainfed rice ecosystem, are difficult to manage, as field water levels cannot be maintained throughout the cropping season. This means weed management, through water control is not fully effective and farmers have to undertake hand weeding, which is time consuming and laborious. Selective herbicides could be used for this additional weed control, however most are too expensive to the Cambodian farmer and may pose a health and environmental hazard. An ideal solution to the Cambodian weed problem is, therefore, to use a rice line that can provide its own weed management through its physical and chemical interference of weed growth.
This project examines the possibility of using allelopathic rice lines to chemically suppress the growth of the six most important weed species in Cambodia. The germplasm that is used for this purpose is predominantly traditional Cambodian rice lines, but also includes some improved lines and two wild rice species. To detect the allelopathic activity in the rice lines, the project employed a range of bioassays with varying sensitivities. The first step in this procedure used a laboratory bioassay known as the "relay-seeding" technique. In this technique, rice and weed seedlings were grown together for 10 days in a Petri dish with an inert support medium. Controls, consisting of weed or rice seedlings alone, were also set up. The possible within-assay interference from factors such as temperature, competition for nutrient and water was minimized, and the differences in root and shoot lengths between weed seedlings grown in monoculture and those grown in the rice treatments were taken as an indication of allelopathy. The results obtained from the laboratory bioassay were then verified in glasshouse pot studies using soil and finally in field experiments.
The laboratory bioassay indicated that among 359 lines studied, there were two, the only two wild rices (Oryza nivara Sharma et Shastry and Oryza rufipogon Griff.) included, that stimulated the germination and growth of awnless barnyard grass, while there were another 15 Oryza sativa lines that could significantly reduce the root and shoot growth of the same weed.
These 17 lines, with a further 79 lines of Oryza sativa that showed a non-significant tendency to suppress the growth of awnless barnyard grass, were then studied in a second laboratory bioassay involving six weed species. The results showed that there were 28 lines that could suppress the seedling growth of awnless barnyard grass, 11 that could suppress barnyard grass, six that could suppress small umbrella sedge, four that could suppress two-leaf fimbristylis, four that could suppress water primrose, and three that could suppress goose weed. Overall, there were 22 Oryza sativa lines that could suppress the growth of one or more weed species, 13 lines that could suppress two or more, five that could suppress three or more, two that could suppress four or more and one that could suppress the growth of five species. All active lines came from the traditional Cambodian seed collection in the Cambodian Agricultural Research and Development Institute (CARDI).
The best 18 lines were then studied in a third glasshouse pot bioassay involving six weed species. Into this, and all subsequent studies, were introduced two Philippine rice lines that had a known allelopathic or non-allelopathic potential. These were added to act as controls to which the new selections could be compared. The results showed that all 18 lines were able to significantly suppress the growth of one or more weed species. Their growth suppressing activity was greater than that of the known allelopathic rice control.
The six best lines were then studied in a forth field bioassay involving four weed species. The results showed that all lines could suppress germination and/or growth of one or more weeds. On average, the suppressive effects of all six lines were greater than those seen for the known allelopathic control. Lines worthyof further study for their allelopathic potential, good yield and pest-resistance are the CAR lines 3, 4 and 8.
A final field bioassay revealed that straw residues, left over after the rice harvest, had a significant inhibitory effect on the germination and growth of both a follow-on rice crop as well as all common weeds studied. However, the observed growth suppression was greatest on the weed, with the crop better able to withstand the toxic effect of the residues. All lines tested, whether they were identified as potentially allelopathic or non-allelopathic, produced the inhibitory effect. This suggested that the toxin(s) involved in this reaction from residues was unlikely to be the same as those involved in the earlier observed allelopathic effect.
In summary, the application of rice crop protection strategies based on allelopathy deserves more consideration. It may be possible to achieve a significant degree of weed suppression through the planting of rice lines that have been shown to have an allelopathic potential. This would be especially true in areas where there is a high weed pressure and rice crop residue incorporation is practiced as a fallow weed control strategy. The use of weed-suppressive crops in combination with other weed management options has several advantages over the present systems used in Cambodia. For example, the use of herbicides can be avoided, or at least reduced, and the time spent on hand weeding can be considerably reduced. These changes would create a decrease in rice production costs and possibly less environmental contamination. A control method that is already in the rice seed provides poor farmers with an affordable weed management option.