Document(s) 30 de 38

Abstract: Eco-edaphic conditions under which rice grows range from deepwater swamp to rainfed upland, with this wide range of zones being a contributing factor to its great varietal diversity. The genetic wealth of cultivated and wild rice germplasm in eastern India is rich and diverse. There is little genetic erosion owing to the release of only a few varieties for flood-prone areas and the rate of replacement of landraces is slow. Breeders and genetic resource personnel have made efforts to collect landraces and their wild relatives, but flood-prone areas located in remote places (10-15% of flood-prone regions) are yet to be explored. Efforts have been by breeders to maintain farm diversity through the launching of a farmer participatory program whereby synthetic landraces and a 'genetic soup' (a method of testing bulk hybrids from F4 stage) have been employed. Testing of hybrid bulks on farmers field enables the breeder to create new landraces within a few seasons through continuous natural selection. It also provides feedback on the relative the merits of elongation ability and submergence tolerance as well as base breeding materials on which breeders can perform line selections in experimental sites. This kind of testing approach provides a considerable opportunity for broadening farmers' farm diversity through including landraces of sub-ecological groups, including wild type races. Farmers adopt only those varieties which are well-adapted to their existing set of environmental conditions. Criteria to be considered for using this diversity-enhancing approach should include: rate of replacement of landraces by improved types; range of environments; richness of genetic diversity; and assessment of past collection efforts.

Introduction

Table 1: Status of deep water varieties in IRRI germplasm bank
Country	No. of landraces
Bangladesh	443
Cambodia	25
China	21
India	42
Indonesia	13
Myanmar	8
Sri Lanka	82
Thailand	34
Vietnam	22
Others	43
Total	743

In eastern India, about 2.3 million ha of land annually suffer from excess water, ranging from 50 cm to 400 cm where rices of different cultural types are grown. Accordingly, unlike other ecosystems, in flood-prone areas, we find a full range of varieties under cultivation, from very primitive to improved types. Uncontrolled flooding is the dominant feature of this ecosystem, with depth and duration of floodwater in the field varying from place to place and from year to year. Therefore, each variety group is adapted only to a specific set of environmental conditions. Flood-prone rice which is mainly close to wild types varies considerably in traits such as morphological character, grain pigmentation and quality, tolerance to many abiotic stresses (e.g. excess water tolerance, drought and soil toxicities), and tolerance to biotic stresses (including resistance to blast, bacterial blight and sheath blight). Studies have revealed that O. nivara, a wild rice, is the only source of grassy stunt virus and grows in seasonal ditches of plateau regions. O. rufipogon is a perennial, photoperiod-sensitive, floating type, with thick and long culm, well-exerted and spreading panicles, and an ability to elongate under abrupt flooding. This indicates that genetic wealth is hidden in the flood-prone ecosystem and that such a resource requires the special attention of policymakers and scientists in order to be conserved.

Alternative approaches for identifying cultivars that are acceptable to the farmers have been suggested. Earlier in rice, Maurya et al (1988) tested advanced lines with villagers in Uttar Pradesh and successfully identified superior material that was preferred by farmers. 'Synthetic landraces', a bulk hybrid population & genetic soup approach, will be discussed as a method which may be particularly suited to adverse conditions. The relationship between the release of a variety, its adoption rate and their joint effect on genetic erosion in flood-prone races will also be emphasized.

A new approach for testing material on farmers' fields at an early stage

In a `synthetic landraces' and `genetic soup' approach, early hybrid populations (F3-F4) are subjected to naturally-flooded rice areas which are representative of most of farmers' fields. Planting should place in adjacent plots marked 'submergence-tolerant bulk' and 'elongating bulk'. As these early segregating materials have potential variability, populations will segregate into early, late, tall, short, good-, poor-survivals, etc. Two bulk populations may survive differently across water depths. Only limited panicles which mature at the appropriate time (as per requirement of the area) should be harvested and then replanted 2-3 generations in the field. Breeders can later select lines at the breeding station with reasonable confidence that the line selected will at least survive the deepwater conditions prevailing at the site.

Obervations on on-going work in the flood-prone ecosystem

Diversity related to flood-prone rice adaptation

Table 2: Diversity with respect to deep water adaptation
Factors	Choices
Deepwater survival strategy (3 choices)	- Submergence tolerance - Elongation ability - Fast emergence
Maturity time (7 choices)	150 d. to 220 d. Oct. to Jan/15 d. interval
Plant Stature (3 choices)	- Short (130 cm) - Medium (130-160 cm) - Tall (>160 cm)
Soil types (2 choices)	- Neutral - Acidic
Photoperiod sensitivity (2 choices)	- 12 to 12.5 day length hour (Latitude up to 14.00o N) - 13 to 14 day length hour (Latitude 22.5 to 27o N)

The role of synthetic landraces and genetic soup in the early adaptation of material

The relationship among the development of varieties, their adoption and farm diversity conservation

Table 3: List of varieties developed by different countries
Country	Varieties
Bangladesh	Indra sail, Tilakkachari, Biasbish, Gabura, Malibhanga, Habiganj Aman 1, Habiganj Aman 2, Habiganj Aman 3, Habiganj Aman 4, Habiganj 5, Habiganj 8. Habiganj Aman 6*, Habiganj Aman* 7, BR 118-3B-117**
Thailand	Leb Mue Nahng 111, Khao Nahng Nuey 11, Pin Gaew 56, Tapow Gaew 161. RD 19*, HTA 60*. BKNFR 76 106-16-0-1-0**
Vietnam	Nang Tay, Tao Binhc
* Cross bred varieties. ** Submergence tolerant advanced breeding lines.

As regards the adoption of newly developed DWR varieties, experience shows that farmers grow only the varieties with highest adaptability and stability. A new variety that requires different or improved management practices will not be easily accepted by the farmers unless the replacement variety shows yield superiority under similar environmental and management conditions. Therefore, the objectives for developing varieties should be based on an analysis of the weaknesses of the varieties already grown. This requires good knowledge of these varieties, especially of their local tolerance of abiotic stresses. At present, above 80% of the released varieties for flood-prone ecosystems in eastern India are reselections of landraces (Table 4). Keeping in mind the widely varying hydrological and flooding trends from year to year on the same location, scientists have developed a plant type concept where need based elongation is the main feature. It prevents unnecessary excess growth and directs the additional food resources to enhanced grain yield. However, landraces/floating rices lack this ability. Improved plant types include RD19, IR 11141-6-1-4 and Jalpriya for deepwater environments.

Table 4: List of rice varieties released in different states for flood-prone & deepwater ecologies.
State	Flood affected areas	Deepwater & floating rice
Assam	T 2205, T 2208	A R C 146
Bihar	BR 13, BR 49	BR 14*, BR 46*, Jai Suria*, BR 7, BR BR 8, Janki, Sudha, Varidhi
Kerala	ITB 15, PTB 16	AR 614 - 25 B
Orissa	FR 13 A, FR 43 B	Marsh 1, Marsh 2, Utkal Prabha
Tamil Nadu	CO 14, PTB 7, ADT 7	TNR 1, TNR 2
Uttar Pradesh	Madhukar, Barh-Avarodhi NDGR 24	Jalmagna*, Jalnidhi*, Chakia-59, Jalpriya
West Bengal	Jal plavan 1, Chin 31, Tilakachari	Jaladhi 1*, Jaladhi* 2, Chinsura 21, Suresh, Biraj, Savita, Mendira, Gogen, Nalini, Natangini, Dinesh, Amulya
*Floating rice (suitable for 1 m water depth)

Farmers are better judges than scientists in choosing cultivars for any particular area. There are cases where varieties have been widely used in areas but formally released later, e.g. Mahsuri, a rainfed lowland variety is very popular and has been cultivated for the last 10-12 years in different areas but was only notified recently. Pant Dhan 4 and Indrasan serve as similar examples in irrigated ecosystems. However, simultaneous testing of materials on farmers' fields under and International Fund for Agricultural Development (IFAD) supported Indian Council of Agricultural Research (ICAR) project for strengthening rainfed/deep water ecosystems led to the release of Jalpriya for deepwater up to 100 cm water depth and Jalnidhi floating rice for 2-3 m depth (Table 5).

Table 5: Relationship between varietal adoption & release of varieties (few examples) in Uttar Pradesh
Variety	Status	Ecosystem
Mahsuri	Fully adopted then notified & released	Rainfed lowland
Pant Dhan 4	Identified by State Govt. officials & farmers first, then released	Irrigated
Indrasan	Well saturated but yet to formally release/notify	Favourable
Jalnidhi	Simultaneous testing on farmers field & released/modified	Very deep water (floating group)
Jalpriya	Simultaneous testing on farmers field & released by SVRC	Deepwater
Barh Avarodhi	Simultaneous testing on farmer's field and released by SVRC	Flash flood
*SVRC - State Variety Release Committee

Experiments conducted at the Crop Research Station, Ghagharaghat, Uttar Pradesh, India with IRRI and Thai materials revealed that the transfer of fixed breeding materials and varieties is not very useful in the Indian context due to the problem of phenotypic acceptability (including correct flowering date and adaptation to new flooding patterns). Therefore, in order to get fruitful results, early segregating populations from F2-F3 onwards need to be tested under natural fields conditions over long periods. Such materials need to be exposed to natural selection for the desired flowering date, level of stability and sustainable yield. If progress is to be made, extensive work has to be done directly in the target environments, in close contact with farmers, and on farmers' farms.

Flowering date is affected due by slight variations in latitude within the country. Sabifa, a most popular and adapted variety of deepwater released for the West Bengal area, flowers 10-12 days in Uttar Pradesh due to a change in latitude. Similarly Jalpriya, Jalmagna, Jalmagna and Jalnidhi flower early in West Bengal. This suggests that a variety should be bred with the local context in mind as its adaptation is limited only to a set of specific environmental conditions.

Specificity in adaptation helps to maintain farm diversity and enables farmers: to fit their cropping system to their heterogenous conditions; to enhance the food security of their household; and to exploit a range of varieties.

Concluding comments

• A number of local varieties and wild types have been collected from different states of India where flood-prone rice are grown. However, it is believed that still 10-15% landraces have not yet been collected from DW and rainfed lowland areas which are located in remote place only accessible by boat.

• Due to poor germination ability of certain landraces, it becomes necessary to recollect cultivars from their original areas.

• Most of the rices from flood-prone races are late maturing, tall, and with varying flowering dates. Therefore, a germplasm collection program should be mapped chalked to ensure that the varying types are collected during limited trip to the areas.

• Since farmers have a better understanding of adaptation of these flood-prone rice types, their involvement in the breeding program should be encouraged.

• As regards the conduct of on-farm trials on farmers' fields within a participatory approach, some incentive should be given to the farmers as production is very risky.

• It was most surprising that only 733 accessions related to flood-prone rice are documented in the rice germplasm bank of IRRI. Timely evaluation of gene bank accessions is also problematic as only a few accessions are evaluated annually.

• As regards the extent of genetic erosion in flood-prone rice, it can be mentioned that the rate of genetic erosion is slow as most of the varieties released in flood-prone ecosystems are reselections from the landraces. Even the improved varieties developed for deepwater rice context include existing varieties as one of their parents in order to retain the local adaptation for a specific situation. Therefore, the genetic situation is not very alarming in the flood-prone ecosystem.

References

Maurya, D.M., A. Bottrall, and J, Farrington, 1988. Improved livelihoods, genetic diversity and farmers participation: a strategy for rice breeding in rainfed areas of India. Experimental Agriculture, 24:311-320.

Document(s) 30 de 38