This report summarizes the main achievements of the ACIAR project "Increased productivity of rice-based cropping systems in Lao PDR, Cambodia and Australia" conducted during 2000/2001 in Laos, Cambodia and Australia. The major goal of this project is to increase the productivity of the rice-based cropping systems in Lao PDR, Cambodia and Australia. This can be achieved through improved efficiency of the rice breeding programs to increase yield and improve stability of yield of lowland rice in these countries and by identifying appropriate strategies for agronomic improvements of rice yield and rice-based cropping systems. The delivery of new technology requires adequate definition of the extrapolation targets. The project therefore is developing agro-ecological characterization (maps) which will be used in Laos to assist the evaluation of the potential of cropping system strategies in the future.

Drought is a major problem in rice cropping systems in Laos and Cambodia. It can occur at any time during plant growth but it occurs most commonly early in the season delaying or preventing transplanting, and later during the reproductive and grain filling periods. Improved cultivars and better agronomic practices are needed to alleviate some of the detrimental effects of drought.

A series of experiments were conducted in Laos and Cambodia during 2000 and 2001. Genotype adaptation for different agro-ecological regions were studied in both countries. Large genotype-by-environment (GxE) interactions for grain yield was common in multi-environment trials (MET) conducted in Laos and Cambodia. These interactions were partly explained by genetic variation in phenological development in relation to soil moisture availability in Laos, and floods during early growth in Cambodia. One of the most important cultivar requirements for drought-prone rainfed lowland environments was an appropriate phenology that matched water availability. In Cambodia, most of the ACIAR lines and CARDI selection produced better yields than the traditional cultivars.

The availability of water in the plant was determined by the position of the field in the local topo-sequence. The large GxE interactions for different topo-sequences in Laos suggested that lines selected at the higher level of the topo-sequence might not necessarily perform well in the lower position of topo-sequence where moisture is available for most of the vegetative and reproductive phases. Therefore, the current practice of relying heavily on single location selection prior to the evaluation across environments in METs in the breeding program in NARC is unlikely to be effective in developing new cultivars that are widely adapted to the target domain or specifically adapted to important environmental constraints.

Preliminary surveys and community meetings were conducted to identify the main constraints in adapting double cropping systems in selected regions in Cambodia. The study showed that many farmers living in areas where water is not limiting are practicing rice based double cropping systems. However, studies of the system as a whole are needed to improve the economic efficiency of these rice based double cropping systems.

The effects of time of planting as it influence low temperature effects on seedling establishment and grain yield have been investigated for dry season irrigated rice in Laos. Variation was found across locations for the best time of planting for seedling survival and grain yield. In 1990/2000 where there was exceptionally low temperature period in late December in Northern Laos, seedlings from December sowing were often affected by low temperature. However, in Vientiane and Champassak, the highest yield was obtained in December sowing. Crops sown in mid January were often affected by rodents, birds and insects.

Temperature maps were developed first time in Laos using GIS technology. This is based on the relationship between temperature and altitude, which varies from month to month within a year. This was a significant achievement in the agro-ecological characterisation project in Laos.

An International Workshop for Increased Crop Production for Lowland Rice in SE Asia was conducted at Vientiane, Laos, where about 100 scientists from several countries attended and 32 papers read during the Workshop. These papers and recommendations for future research will be published as ACIAR Proceedings in 2001.

This report summarizes the main achievements of the ACIAR project "Increased productivity of rice-based cropping systems in Lao PDR, Cambodia and Australia" conducted during 2001/2002 in Laos, Cambodia and Australia. The major goal of this project is to increase the productivity of the rice-based cropping systems in Lao PDR, Cambodia and Australia. This can be achieved through improved efficiency of the rice breeding programs to increase yield and improve stability of yield of lowland rice in these countries and by identifying appropriate strategies for agronomic improvements of rice yield and rice-based cropping systems. The delivery of new technology requires adequate definition of the extrapolation targets. The project therefore is developing agro-ecological characterization (maps), which will be used in Laos to assist the evaluation of the potential of cropping system strategies in the future.

Drought is a major problem in rice cropping systems in Laos and Cambodia. It can occur at any time during plant growth but it occurs most commonly early in the season delaying or preventing transplanting, and later during the reproductive and grain filling periods. Improved cultivars and better agronomic practices are needed to alleviate some of the detrimental effects of drought.

A total of 33 experiments were conducted in Laos and Cambodia in 2001/2002. The adaptation of genotypes for different agro-ecological regions was studied in multi-environment trials (MET) in both countries. There was a large genotype-by-environment (GxE) interaction for grain yield in both countries, partly explained by genetipic variation in phenological development in relation to soil moisture.

A number of high yielding lines have been identified with superior performance over the environments. In Laos, yield of these lines exceeded the current leading cultivar TDK1 by about 10 % and another cultivar RD6 by about 20 %. In Cambodia, the high yielding lines produced about 20-25% higher yield than commonly grown cultivars. The studies show that genotypes that are adapted to rainfed lowland environments have high yield potential and can respond to well watered conditions, compared with less adapted cultivars. They are intermediate in height and are photoperiod insensitive or only mildly sensitive with early-medium maturity.

In Laos, low temperature during the seedling nursery stage in the dry season is a main constraint to intensifying cropping systems. Experiments on time of planting were conducted for 3 seasons in 4-6 locations to identify the optimum time of sowing. Low temperature is more severe in paddies located in the high altitude areas in Northern Laos. Rice crops failed in 2 years at the highest altitude areas examined (Xiean Khoung) when they were seeded in December. The main problem was poor germination and seedling growth in nurseries, with establishment failure at mean minimum air temperature below 120C for 30 days in the seedbed period. At mean minimum temperature above 16 0C there was no low temperature damage. The December-January seeding should be avoided in the high altitude areas in Northern Laos. However, in the low altitude areas, December planting produced higher yield than November planting. Hence the optimum time of seeding is different in different areas in Laos. The project developed GIS based maps to provide the minimum temperature probability that can be used as a guide for time of sowing.

In Cambodia, the focus of the cropping system research has been to diversify and intensify rise-based systems. Early wet season rice can be grown provided there is supplementary irrigation (usually from tube wells). The early wet season rice crop has consistently failed at Baneav where farmers rely only on rainfall. However, an early wet season and dry season legume (mungbean) crop produced some yields from rainfall alone. Rice-legume cropping systems in Slakou and Bat Rokar were successful to some extent, while farmers in other areas have preferred planting two crops of rice with supplementary irrigation.

Direct sseding of rice can shorten the planting time and reduce labour needs for transplanting. The direct seeding experiment at Savannakhet in Laos showed that higher seed rate (200kg ha-1) in broadcasting could produce higher grain yield than lower seed rates. However, higher seed rates reduced the seed size in both varieties; TDK 1 and SK12. The yield differences among seed rates were not significant both in broadcasting and row seeding at Vientiane. In direct seeding experiment at CARDI, Cambodia there is some evidence of genotypic variation in response to direct seeding and yield losses were observed in those lines with heavy lodging.

The serial field trials in NSW contained 50 cultivars from diverse origins and shallow water was maintained to expose plants to the cool night temperatures. Susceptible cultivars such as Sprint (from Russia) and Sasanishiki (Japan) had high levels of spikelet sterility while tolerant cultivars such as M103 (USA) and HSC55 (Hungary) were not damaged. The information has provided accurate crop phenology data for these 50 cultivars. Approximately 30 cultivars flower in response to temperature alone (eg HSC55, Jarrah) while the remainder responds to photoperiod as well as temperature (eg Sasanishiki, Amaroo). Increasing the photoperiod sensitivity of Australia's rice cultivars is one possible way to minimise the exposure of the crop to low temperatures during the young microspore stage.

Twenty-three cultivars were exposed to cool water (19C) from a shallow aquifer during the critical reproductive stage in the 2001/2002 season in Australia. Two cultivars, Jyoudeki and HSC55 had low levels of spikelet sterility (35%). The susceptibility of Sasanishiki, YRL 39 and Doongara was confirmed with approximately 90% sterility. The cold tolerance was observable at flowering with cold tolerant cultivars having large anthers, many engorged pollen grains and large stigma size. A large cool water system (10m x 2m) is now being developed for next season, which will have the capacity to test 400 genotypes.

ACIAR annual general meetings were conducted in Laos and Cambodia separately. The scientists involved in the ACIAR project presented their results and project proposal for the 2002/2003 seasons.

2.1 Plant breeding strategies for rainfed lowland rice

In Cambodia, the first extensive multi-location yield trials were conducted over 3 years at 8 locations using 32 lines from various sources. Despite a large genotype by environment interaction for yield, the magnitude was smaller than the level reported in Thailand from a similar experiment. The general conclusions from the experiments is that a) genotypes can be selected for broad adaptation to these main sites (ie there is one main target population of environments for the Cambodian breeding program); and b) multi-location trials need to be conducted for at least 2 years and at a minimum of 6 locations to allow selection of the broadly adapted cultivars.

The drought screening method developed increases the likelihood and severity of drought by draining standing water from the crop during the reproductive phase of growth. This method is being integrated into the breeding programs in Laos and Cambodia in the present ACIAR project, and is now widely used throughout Asia. Drought resistant lines identified in these projects are being used as donors for new populations that will assist in the development of agronomically acceptable and widely adapted cultivars. These populations are being advanced rapidly (3 generations per year) using techniques introduced by the project.

In addition to the drought resistance characteristics, the genotypes that are adapted to rainfed lowland environments have a high yield potential under irrigated conditions, compared with less adapted cultivars. The adapted cultivars are mostly intermediate height of early to medium maturity.

2.2 Intensification of rice-based cropping systems in rainfed lowlands
The success of double cropping of rice-based systems in the rainfed lowlands of Cambodia primarily depends on the effective management of water resources. The major limitations identified for double cropping systems include the lack of reliable rainfall for the early wet season (EWS) rice crop, risk of soil water saturation for early season mungbean crop, and lack of water for establishment and subsequent growth of mungbean after the wet season (WS) rice. The current practice is to minimise the risk and maximise the yield of the WS rice crop, and hence there is rather limited flexibility for other crops. The EWS mungbean crop should be planted as early as possible to reduce the risk of soil water saturation, whereas the dry season (DS) mungbean crop should be planted immediately after the harvesting of the WS rice. The availability of supplementary irrigation even for limited, strategic irrigation greatly reduces the risk of crop loss and increases the practice of double cropping in rainfed lowlands.

In Lao PDR, rice-rice double cropping is possible in irrigated rice growing areas. It is possible to achieve a combined yield of 9t ha-1 from two seasons. However, the DS rice crop should be established in November to avoid the low temperature damages in nurseries at Xieng Khoung and Luang Namtha.

2.3 Development of direct seeding technology
Direct seeding can be used in both the WS and DS when labour is in short supply. However, direct seeding in the DS is more effective as soil moisture can be controlled more effectively during seedling establishment. Good establishment of the rice crop is an important target for farmers to reduce competition from weeds. In the absence of weeds, genotype performance was similar under transplanting and direct seeding. However, lodging and poor establishment were problematic for some genotypes. It was concluded that there is no need to conduct a major, separate breeding program for direct seeding, provided that there is adequate selection against lodging.

2.4 Increasing productivity of dry season irrigated rice
One of the problems in increasing productivity of irrigated rice in Lao PDR is poor nursery establishment under low temperature conditions. An improved nursery management system using half-burnt paddy husk and plastic dome was developed for the low temperature areas in Northern Laos. The new technology overcomes the problem of raising enough seedlings during the critical low temperature period at Xieng Khoung and Luang Namtha.

Nine cultivars from various origins were included in a yield potential trial at Yanco Agricultural Institute, Australia. The cultivar, TDK 33-9 had the highest yield (10.3t ha-1) among the three Laos cultivars. However, the other two cultivars from Laos, TDK 2 and PNG 1 produced yields less than 5t ha-1.

2.5 Agro-ecological characterization
Agro-ecological maps that characterise the factors important for agricultural production have been developed. The temperature maps and associated documents that describe the development and application of the agro-ecological maps have been published. GIS-based maps of annual and weekly rainfall for the country were developed for Laos from 26 meteorological stations and 5 hydrological stations in Laos and 18 meteorological stations in Thailand. The daily data from 1994 to 1999 was used in the production of rainfall maps. Potential evapo-transpiration has been calculated for several stations where sunshine data was available enabling the completion of GIS based maps of weekly potential evapo-transpiration for Laos.

Year 4 (01/07/2003-30/06/2004)
Plant breeding strategies for rainfed lowland rice
Rainfed lowland rice is the main rice production system in Laos and Cambodia and the objective was to minimize the constraints for lowland rice production in the Mekong region through new breeding strategies for the development of varieties tolerant to drought and low soil fertility. Extensive multi-location experiments for grain yield were conducted over four years at eight locations using 16-32 lines from various sources. The combined analysis with 2003 wet season data revealed a strong relationship between yield reduction and water deficit at flowering for rainfed lowland rice in Cambodia. The grouping of environments in multi-location trials was not associated with soil type, but with water availability at flowering. A new method of quantifying water availability at flowering was developed, and this assisted in the environmental characterisation for multi-location trials. This technique also enables the testing of breeding populations for drought adaptation.
Intensification of rice-based cropping systems in rainfed lowlands
The development of double cropping systems to increase on-farm productivity is vital for farmers in Cambodia and Laos. Experiments have been conducted since 2001 in wet and dry seasons to examine the most successful and low risk cropping systems. In Cambodia, areas with high rainfall or supplementary irrigation had a low risk of double cropping for rice in the early wet season (EWS) and wet season (WS). There was a high risk of crop failure of EWS rice in low rainfall areas where supplementary irrigation was not available. Rice-rice double cropping was possible for most areas where good rainfall was received in early April in EWS. There were difficulties for double cropping mungbean with rice in the WS due to high rainfall causing water-logging in the EWS and drought conditions in the dry season (DS).
Experiments were also conducted at four locations in 2003 in northern Laos to explore the likelihood of rice-rice double-cropping under irrigated conditions. Crop establishment in the DS was successful when the mean minimum temperature was above 12C. The annual grain yield was increased to 8-9 t ha-1 when rice-rice double cropping was adopted under irrigated condition.
Development of direct seeding technology
The specific objective of experiments in 2003 was to investigate the characters and mechanisms associated with genotypes adapted to direct seeding. In the 2003 wet season, the effect of weed competition on grain yield and the relationship between stem diameter and lodging were investigated among 34 genotypes, which were direct seeded at three locations in Cambodia. The yield was similar under direct seeding when weeded 30 days after sowing to that of transplanted crops. Varieties that had a small first internode length had a low risk of lodging under direct seeding.
Increasing productivity of dry season irrigated rice
The increase in rice-rice double cropping in northern Laos has meant farmers are encountering new problems in the dry season (DS), primarily related to low temperature during seedling establishment. Experiments at different locations were conducted in the 2003-04 DS to investigate ways to improve seedling establishment for DS rice. The relationship between establishment temperature and seedling growth showed that when the mean minimum temperature was below 12C the rice crop was more likely to have poor establishment. Seedlings were protected from low temperature in the nursery with a plastic sheet applied as a flat cover and also as a dome. On-farm experiments showed the higher yields resulted when seedlings were established under the plastic dome compared to conventional farmer practice. The plastic dome could also increase sowing flexibility by reducing the duration of seedlings in the nursery by 20 days.
Agro-ecological characterisation
In the rainfed lowlands, where droughts and flood problems are related to water availability in the field, rainfall is an important climatic factor. Rainfall, temperature and potential evapo-transpiration GIS maps have now been completed for Laos. In 2003 these factors were used to produce GIS maps, which showed the length of growing period for Laos. GIS maps are now being combined with crop models in Savannakhet in southern Laos to examine the limits to high yield potential, particularly caused by water deficit. Climatic factors, particularly rainfall and temperature, and factors associated with water movement within and between rice paddies are identified as the key for agro-ecological characterisation of lowland ecosystems.

2.1 Plant breeding strategies for rainfed lowland rice
The information on the variety adaptation across varying rice growing environments in Lao PDR is limited. Previous ACIAR projects have investigated the magnitude of G-E interaction under different fertilizer conditions in three provinces and found relatively stable conditions of varieties across varying nutrient conditions. However, there is no evidence of the magnitude of G-E interaction in Laos with a large environmental and climatic variation.
The main objectives of this Sub-Project were a) to determine the magnitude of G-E interaction across the rice growing environments in Lao PDR and b) to examine whether there are any genotypes among the existing breeding population that can be used for wide adaptation to different growing conditions in the country. In total 48 genotypes were selected from the Lao breeding program, drought screening experiments and introductions from Thailand's breeding program for inclusion in 2004 wet season (WS) and 2004/05 dry season (DS) experiments. These selections were tested at Savannakhet, Champassak, Luang Namtha and Xieng Khoung under normal farmer practice using transplanting establishment methods. The results suggested that location, genotype, genotype-by-location variations for the yield, yield components and days to flower were significant. The genotype by environment variance component was four times larger than the genotype variance component for grain yield. This indicated that the yield performance of a selected set of genotypes is inconsistent across four provinces. Fifteen high yielding genotypes were identified. There were a number of better performing genotypes (e.g. TDK21-B-6-2-1-B and TDK47-2B-6-1-1) than the existing popular genotypes, TDK1 and TSN1 at four locations. These lines were selected from the breeding program and most of the improved lines had a higher yield than the traditional and popular varieties. The results of 2004/05 DS experiments will be analysed with the 2004 WS data in order to test the genotype by season interactions.
In the 2004WS, 23 genotypes were examined in multi-location trials (MLTs) in 8 locations throughout Cambodia. Most of the high yielding varieties selected from 2003 experiment and some new varieties selected from drought screening trial were planted in this experiment. These MLTs identified some genotypes with a higher grain yield than the traditional landrace material, which will be used as parents for the rice breeding program. In addition, a farmer participatory selection program was initiated with 25 farmers in Chrey Veal. The farmers selected three varieties based on visual assessment of plant type, duration, grain shape and grain yield.
2.2 Intensification of rice-based cropping systems in rainfed lowlands
The development of double cropping systems to increase on-farm productivity is vital for farmers in Cambodia and Laos. In Cambodia, areas with high rainfall or supplementary irrigation had a low risk of double cropping for rice in early wet season (EWS) and WS. There was a high risk of crop failure of EWS rice in low rainfall areas where supplementary irrigation was not available. Rice-rice double cropping was possible for most areas where good rainfall was received in early April during the EWS. There were difficulties for double cropping mungbean in EWS with rice in the WS due to high rainfall causing water-logging in the EWS and drought conditions in the DS.
Experiments were also conducted at Luang Namtha and Xieng Khoung in 2004 WS and 2004/05 DS in northern Laos to further test the likelihood of rice-rice double-cropping under irrigated conditions. Crop establishment in the DS was successful as there was no extremely low temperature during sowing in the DS. The annual grain yield was increased to 8-9 t/ha when rice-rice double cropping was adopted under irrigated condition in the DS.
2.3 Development of direct seeding technology
The specific objective of experiments in Laos in 2004 WS and 2004/05 DS was to investigate the nutrient requirement for different seed rates in Savannakhet province. In the WS, high nitrogen rates and increased plant density was appropriate for areas with weed problem.
A row seeder was used for direct seeding experiments and demonstration trials in irrigated conditions in DS at Vientiane Province. Several farmers demonstrated that they could significantly reduce the cost of establishment and weed control by adopting new technology. Direct seeding in the DS was recommended for Vientiane as it produced a consistently higher yield compared to transplanting. Direct seeding in the WS failed several times due to problems associated with water control.
2.4 Increasing productivity of dry season irrigated rice
The main constraint in growing rice in DS at higher altitudes is low temperature during seedling establishment. One option to provide farmers with a more flexible planting period in these high altitude areas is to use a plastic dome over the seedbed to protect the developing seedlings from low temperature in December-January. The project has successfully developed and demonstrated a simple nursery protection method using locally available plastics. In 2004/05 DS, several farmers from Xieng Khoung and Luang Namtha demonstrated that the technique was a practical solution to the low temperature problem. The farmers who have participated in on-farm experiments in 2003/04 DS have achieved a 20% higher grain yield compared to the yield of the crop that was established under the conventional unprotected nursery. The cost effectiveness and extension strategy of the technology needs to be evaluated in 2005/06 DS. Some extension material was developed to assist with the adoption of plastic protection during establishment in northern Laos.
2.5 Agro-ecological characterization
In 2004, six agro-climatic zones were developed for Laos based on temperature, rainfall and altitude data. The country has tropical, subtropical and mountainous zones. These zones are further divided into humid and sub-humid areas according to water availability. The maps also consider water availability at a greater level of specificity than what previously was available. Water availability in rice paddies is complicated because of downward and lateral water movement; the project quantified these components for better estimation of paddy water environments. The agro-climate zone map provides sufficient information to determine where different crops can be grown, based on climatic information alone. Crop suitability maps utilizing climate as well as soil and topography information, were developed for Savannakhet to identify areas suitable for different crops. These maps still require further testing for their accuracy in predicting crop suitability in different areas before they can be used by development planners.
The water balance study conducted in 2000-2002 across topo-sequence from field experiments in Southern Laos were analysed in 2004 and made the following findings. The water supply in paddy fields under rainfed conditions is determined primarily by rainfall, and frequent rainfall events can retain standing water in the field until a critical growth phase, i.e. anthesis. In the sloping lowlands, water losses by downward and lateral movement are greater in areas higher in a topo-sequence, and the lateral water loss possibly becomes a water input in the areas of lower topo-sequence. Therefore, the water availability in the fields that lies in lower topo-sequence is greater, resulting in higher productivity, while in the fields in higher topo-sequence, greater reduction in yield often occurs due to low plant water availability. However, this will not be the case where much water runs on to the fields in higher topo-sequence from catchments above the fields. In addition, the water balance study has shown that the rainfed lowlands in southern Laos potentially encounter the late season drought, implying low productivity in those years with low rainfall, and that the water productivity generally decreases with late flowering in the rainfed lowlands.