Variety improvement - Upland Rice

Transcription

1 Variety improvement - Upland Rice Khamdok Songyikhangs 1, Somphet Phengchan 1, Bruce Linquist 2 and Gary Atlin 3 1 Northern Region Agricultural and Forestry Research Center, Houay Khot, Luang Prabang Province, Lao PDR. 2 Lao-IRRI Project, P.O. Box 600, Luang Prabang 3 IRRI, Los Baños, Philippines Abstract Upland rice variety trials were initiated in Varieties are first evaluated in on-station observation nurseries, and then in replicated multi-location yield trials and finally in on-farm yield trials. The objective of these trials has been to evaluate and characterize traditional upland varieties and identify superior varieties suitable for a range of upland environments. In 2001, 178 black rice (black pericarp) varieties and 225 other traditional varieties were evaluated in the first observation nursery. Based on plant characteristics preferred by farmers and grain yield, 19 black rice and 22 ordinary rice varieties were selected for further evaluation in the 2002 second-year observation nursery. In the 2001 multi-location yield trial, 13 early and 12 medium maturing varieties were evaluated in three locations (two sites in Luang Prabang province and one site in Sayabouly province). Makfay, Pasiev and Do xao performed well in all sites. Based on good yield performance and plant characteristics, Makfay, Chao do (non-glutinous), Pasiev, Do xao, Phaesavanh and Med ngay were selected for further evaluation. In 2001 a participatory variety selection (PVS) program was introduced in the on-farm evaluation stage. The objective was to identify variety characteristics that farmers prefer and to identify varieties farmers liked and that also produced well. Eight early and eight medium duration varieties were evaluated by 32 farmers in five northern provinces. Farmers preferred varieties with large panicle, big seed, strong plant, a lot of seeds, high tillers and early maturity. In contrast, farmers did not like varieties that produced few tillers, had short and non-uniform panicles, small seeds, sterile spikelets, and short plant type. Based on farmer preference analysis and yield, six varieties were identified for wider evaluation in Mother- Baby trials in Key words: upland rice improvement, black pericarp rice, participatory variety selection 33

2 3 4 ñ Á ²ñ À í3 4Ä»È ö ñ ¹ Ó 3 4 õ 3 4Á ²ñ À í3 4Ä»ÈÄ É õ ªÒª3 4 ªº À ó õ 3 4²ñ À í3 4Ä»Èà ì1 2 ñ 1 2 «3 4 ó Áì1 2ªÒ ê1 4 ö 1 2ìò à ì1 2 ñ êéº «Áì1 2 ªº êé3 4 Á È ê1 4 ö 1 2ìò ÏøÈ 3 4 Ä»ÈÀ ó 1 2 ö À²º 1 2À ó ö Áì1 2 ñ Àìõº Á ²ñ À í3 4Ä»Èê ó ìñ ê ó Áì1 2Áê ÀÎ ñ Í ²3 4 À º Ä Ã À Ä»È. à ó 2001, 178 Á ²ñ À í3 4Ä»È Áì Á ²ñ ² À õº À í3 4Ä»ÈÄ Éê À ó à 3 4 õ 3 4²ñ óêî. ºó Ã È ìñ º ì ªí ê 3 4Ä»È ñ Áì1 2 ö 1 2ìò Ä É ñ Àìõº 19 Á ²ñ À í3 4 ÔÄ»È Áì Á ²ñ À í3 4ĻȲ À õº À²º õ ªÒÄ õ 3 4 ²ñ óê 2 ó 2002.à ó 2001, 3 4 êö ìº 1 4 ê1 4 Á ²ñ À í3 4Ä»Èà ì1 2 ñ êéº «ó À ó 13 Á ²ñ º3 4 ð Áì Á ²ñ º Ä É 1 2À ó µøè 3 (2 µøèá Í ² Áì1 2 1 µøè Á Ä 1 2 øìó). À í3 4 Ï3 4 ij, 3 4 ó Áì1 2 ð Àìó Àªó ÂªÄ É óã ê. ºó à ÈÁ ²ñ Ã¹É ö 1 2ìò óã ê êö ìº Áì1 2 ìñ êó ó º ªí À í3 4, À í3 4Ï3 4 ij, À í3 4 É3 4 ð (À í3 4 É3 4 ), À í ó, À í3 4 ð 3 4, À í3 4Á²È 1 2¹ñ Áì1 2À í3 4À ñ Ã¹È Ä É ñ Àìõº Ä 1 2À ó à ªÒÄ. à ó 2001, òêó 3 4 ñ Àìõº Á ²ñ Á ó È»È (PVS) Ä É Ã ÉÀ í3 4à À ó Á ²ñ À í3 4Ä»Èà ² ê º 3 4 Ä»ÈÀ ó 1 2 ö À²º õ 3 4 ìñ º Á ²ñ À í3 4 Ä»È Áì1 2 Á ²ñ à ê 3 4 Ä»È ñ Áì1 2Ã¹É ö ó. 8 Á ²ñ º3 4 ð Áì1 2 8 Á ²ñ º Ä É 1 2À ó»è ñ 3 4 Ä»È 32 ö à 5 Á ²3 4 ÀÎõº. 3 4 Ä»È È Ã¹È ñ Á ²ñ À í3 4êùɻ ùÈ1 4, À ñ ùÈ, ì ªí Á Á» ó, ùÉÀ ñ «, Áª ðí3 4 Áì1 2 ɺ É3 4 º3 4 ð. à À ó ö ñ É Ä»È Ò ñ Á ²ñ ê᪠ðîéº,» Áì1 2«º» Ҳɺ ñ, À ñ ɺ, óà ñ ìó Í3 4 Áì1 2ªí Àª. ºó Ã È ìñ º ì ªí ê 3 4 Ä»È ñ Áì1 2 ö 1 2ìò, Ä É ñ Àìõº 6 Á ²ñ À í3 4Ä»ÈÀ²º Ä 1 2À ó ûø Á 3 4 êö ìº Á Èìø ó Introduction Rice is the most important crop in Lao. More than 80% of the total cropped area is planted to rice. Upland rice area accounts for approximately 22% of total rice area and about 12% of total rice production. Most upland rice cultivation takes place in northern Laos where upland rice is the main component of the slash and burn systems. Germplasm collection and the identification of superior varieties have been important components of upland ecosystem studies in the northern part of Laos since their inception in Between 1991 and 1993 a total of 544 traditional upland rice cultivars were collected in seven Lao provinces. Between more than 7300 traditional upland rice varieties have been collected by the SDC-supported project (Collection and Preservation of Rice Biodiversity). These varieties are being evaluated to 34

3 identify superior lines. Varieties are initially evaluated in observational nurseries and multi-location yield trials. Following selections from these trials, on-farm testing and evaluation has been attempted, but with limited success. Despite ten years of research and 27 varieties reaching the on-farm evaluation stage, no varieties have been released or recommended. In 2001, a high emphasis was placed on the onfarm evaluation component. This involved the introduction of a participatory variety selection (PVS) program at the on-farm testing stage. The objectives were to identify (1) variety characteristics that farmers prefer and (2) varieties that farmer farmers like and that have good productivity. Materials and methods In the year 2001, upland rice variety trials involved two observational nurseries, a multi-location yield trial (three locations) and on-farm evaluations (32 farmers in 5 provinces). Observation Nursery 1 (OBN-1) There were two separate OBN-1 trials, one evaluating ordinary glutinous and non-glutinous rice varieties, and the other evaluating the specialty black rice (black pericarp) varieties. Trials were conducted at the Northern Regional Agriculture and Forestry Research Center (NAFRC) in Houay Khot. Both trials were located side by side in a flat field where rice was grown the previous year. The site was generally low in soil fertility and it was also highly heterogeneous. In the first-year observational nursery, 215 traditional upland rice varieties selected from the Lao Gene Bank at the National Agriculture Research Center (NARC) and 10 introduced non-glutinous rice varieties (from CIAT) were evaluated to select varieties with superior characteristics. Two common traditional check varieties were used (Khao Vieng and Hom). These checks were placed every 10 varieties within the trial. The trial was laid out in randomized complete block augmented design with 2 replications. The plot size was 0.75 m x 2m. Six to eight seeds in mixture with Sevin were dibbled using a hill spacing of 25 x 25 cm. Nitrogen in form of urea was applied at rate of 20 kg/ha to all plots after the first weeding to minimize field variation. Weeding was done manually as required. Observations included: days to flowering, number of missing hills, plant height, panicle number per hill, panicle length, weight of 10 panicles and whole-plot grain yield (adjusted to 14% moisture). Qualitative scorings were made for early vigor, panicle exertion, panicle threshability, spikelet fertility, lodging, pest and disease incidence, and a general agronomic score. At harvest, whole plots were harvested for grain yield. Yield estimates were adjusted for the number of missing hills. In the black rice OBN-1, there were 178 varieties. The traditional check variety was Khao kam. The experiment was managed as above except that plot size was 2m x 1m and there was only one replication. Multi-location yield trial Selected higher yielding varieties identified in the 2000 OBN-2 and multi-location yield trials were further evaluated in replicated multi-location yield trials to determine yield potential and adaptability of those varieties under a range of soil and climatic conditions. The trial was conducted at two sites in Xiengngeun district of Luang Prabang province (NAFRC, Houay Khot - elevation 360 m and Ban Houay Hia - elevation 800 m) and one site in Sayaboury province. At the Houay Khot site, rice was grown in the previous year and therefore this site was of low soil fertility. There were 13 early and 12 medium varieties evaluated (Table 1). There was one non-glutinous variety in the evaluation (Chao do). The experiment was laid out separately for early and medium duration varieties. 35

4 The experiment was laid out in randomized complete block design with 4 replications. The plot size was 4 x 2m. Six to eight seeds in mixture with Sevin were dibbled at a hill spacing of 25 x 25 cm. Nitrogen in form of urea was applied at rate of 20 kg/ha to all plots after the first weeding to minimize field variation. Weeding was done manually as required. Observations included days to flowering, number of missing hill, plant height, panicle number per hill, panicle length, weight of 10 panicles and whole-plot grain yield (adjusted to 14% moisture). Qualitative scorings were made for early vigor, panicle exertion, panicle threshability, spikelet fertility, lodging, pest and disease incidence, and a general agronomic score. Table 1. Varieties evaluated in the 2001 multi-location yield trials. Early maturity Medium maturity Accession # Variety Name Type Accession # Variety Name Type LG-6501 Deng a G LG Do xao a G LG Makfay a G LG Pa siev a G LG Chao do a N LG Med gnay a G LG Keo a G LG Phae savanh a G LG Khiekhang a G LG Khaia G LG Chaomeuang kay a N LG-7084 Paneur (peek) b G LG-6911 Dengsaid b G LG-6432 Mak hine soung b G LG-7771 Kou yong ke b G LG-7051 Do hom b G LG-6905 Sang b G LG-7023 Da cheung b G LG-1724 Dam b G LG-6593 Sangon b G LG-7347 Taaroon b G LG-6655 Lebmeu b G LG-6499 Nok b G Vieng b G Hom G Local check Local check a Varieties selected from 2000 second-year observation nursery b Varieties selected from 2000 multilocation yield trials. On-farm participatory selection (PVS) On-farm PVS was conducted with 32 farmers in five provinces: Luang Prabang (10 farmers), Oudomsay (6 farmers), Luang Namtha (4 farmers), Sayaboury (6 farmers) and Xieng Khouang (6 farmers). Seed availability limited the number of farmers able to evaluate the 16 glutinous upland rice varieties included in the trials. Twelve varieties were selected from the 2001 multi-location yield trials and 4 from previous trials (Vieng, Hom, Dam and Makthoua). These varieties were divided in two groups depending on their duration (Table 2). For the 2001 PVS trials, participating farmers were given one set (early or medium) of eight varieties (110 g per variety). This was enough to plant an area of 15 m 2 of each variety. Farmers were given assistance with plot layout and planting but managed the plots according to their normal practices. Between flowering and maturity researchers visited each farmer and conducted a Preference Analysis (PA). In the PA the farmers were asked to rank each variety using seeds, placing more seeds for varieties they liked. They were then asked to indicate the positive and negative characteristics of the varieties. Finally, they were asked which varieties they would continue to evaluate next year on their own. At harvest, the researchers and farmers harvested each plot for grain yield. 36

5 Of the 32 farmers evaluating varieties, 27 participated in the PA and we were able to collect yield data from 20.Yield data from four of the sites where the data was collected was not used in the analysis because of poor plot layout. Individual farmer variety rankings were standardized across farmers by expressing the ratings (seed placed by each farmer on the board) as a proportion of the total number of seed applied to the board by each farmer. The ratings were further standardized to a mean of 0 and SD of 1 for each farmer. Least-squares means were estimated using PROC MIXED METHOD=REML in SAS, with varieties considered fixed, and farms random effects. Correlations between measured grain yield and farmer preference ratings were estimated for those farms for which both types of data were collected. Table 2. Varieties and accession numbers evaluated during the 2001 PVS trials Early Maturing Medium Maturing Accession no. Variety name Accession no. Variety name LG-7347 Taa roon LG-7051 Do hom LG-6911 Deng said LG-6432 Mak hine soung LG-7771 Kou yongke LG-7023 Da cheung LG-6499 Nok LG-7084 Paneur LG- Vieng LG-6655 Lebmeu LG- Hom LG-1724 Dam LG-6905 Sang LG-6593 Sangon LG-6501 Deng LG-2387 Makthoua OBN-1 Results and discussion In the black rice OBN-1, there was a wide range of variety traits (Table 3, Annex 1). Yields averaged 104 g/m 2 (ranged from 10 to 255 g/m 2 ). The correlation between measured and adjusted yields is very poor (Annex 1) due to many missing hills. Further discussion will only relate to the adjusted grain yields that have been adjusted for missing hills. Days to flowering ranged from 85 to 149 days. The correlation matrix shows that grain yield was positively correlated with plant height, number of panicles and 10 panicle weight (Annex 2). Based on grain yield, other measures of crop performance and plant characteristics preferred by farmers, 19 black rice varieties were selected for further evaluation in OBN- 2 in 2002 (Annex 1). For the ordinary rice varieties, adjusted yields (adjusted for missing hills) averaged 56 g/m 2 (range from 5.5 to 356 g/m 2 : Table 4; Annex 3). Days to flowering ranged from 85 to 149 days. A correlation matrix indicates that grain yield was positively correlated with all selected plant characteristics except days to flowering where there was a significant negative relationship. In Annex 3, using rankings based on yield adjusted for missing hills, fourteen varieties yielded greater than the highest control variety (Khao Vieng) and twenty-two greater than the control average. Twenty-two varieties were selected for further evaluation in observation nursery-2 trials. 37

6 Table 3. Summary statistics for the 2001 Observation Nursery 1 at Houay Khot Station (black rice). Observations Mean n Range Std Error CV (%) Lao Traditional Grain yield (g/m 2 ) a Varieties Grain yield (g/m 2 ) b Days to flowering Height (cm) per hill Panicle length (cm) Weight of 10 panicles (g) Controls Days to flowering (means of 1 variety) Height (cm) a measured b adjusted for missing hills per hill Panicle length (cm) Table 4. Summary statistics for the 2001 Observation Nursery 1 at Houay Khot (Ordinary rice). Observations Mean n Range Std Error CV (%) Lao Traditional Grain yield (g/m 2 ) a Varieties Grain yield (g/m 2 ) b Days to flowering Height (cm) per hill Panicle length (cm) Weight of 10 panicles (g) Controls Grain yield (g/m 2 ) a (means of 2 varieties) Grain yield (g/m 2 ) b Days to flowering Height (cm) per hill Panicle length (cm) Weight of 10 panicles (g) a measured b adjusted for missing hills Table 5. Control varieties in 2001 Observation Nursery 1. Control Variety Grain yield Days to flowering Height Panicle length Weight of 10 panicles g m -2 cm hill -1 cm g Vieng a 99 a 112 a 9.8 a 22.6 a 18.2 a Hom b 99 a 105 b 5.5 b 18.0 b 16.8 a ANOVA (PR>F) ** ns ** ** ** ns CV (%) a measured b adjusted for missing hills 38

7 Multi-location yield trial Results of the 2001the multi-location yield trials are shown in Annex Average grain yields were highest in Sayabouly (Tables 6 and 7) for both the early (2.24 t/ha) and medium (2.30 t/ha) duration trials and lowest at Houay Khot (0.71 and 1.18 t/ha for the early and medium duration trials, respectively). Yields are low at Houay Khot probably because this year was the second consecutive year of rice production on these soils. In the early variety trials, Makfay had good yields at all sites. Khiekhang had good yields in both Sayabouly and Houay Hia, but low yields at Houay Khot, which suggests that it does not do well under low soil fertility conditions. In spite of lower yields at all sites, Chao do (non-glutinous) had the longest panicle, a characteristic desired by farmers. Two short duration varieties were selected for further testing: Makfay and Chao do. Table 6. of the early duration multi-location yield trails. An * beside a variety name indicates varieties that were selected for further evaluation. Houay Khot Houay Hia Sayabouly Accession # Variety Name Grain Yield (t ha -1 ) LG-6501 Deng 1.31 a 1.5 b 2.56 abc Check 1.30 a 2.3 ab 1.76 de LG Makfay * 1.12 a 2.8 a 2.77 ab LG-6911 Dengsaid 0.77 b 1.8 b 1.41 e LG-7771 Kou yong ke 0.67 b 1.5 b 2.32 e-d LG Chao do * 0.62 b 2.3 ab 2.16 b-e LG Keo 0.56 b 2.3 ab 1.51 a-e LG Khiekhang 0.54 b 2.3 ab 2.97 a LG Chaomeuang kay 0.54 b 1.8 b 2.88 ab LG-6905 Sang 0.53 b 2.3 ab 1.94 cde LG-1724 Dam 0.52 b 2.3 ab 2.74 ab Vieng 0.51 b 2.3 ab 2.13 b-e LG-7347 Taaroon 0.45 b 2.0 ab 1.91 cde LG-6499 Nok 0.44 b 1.8 b 2.25 a-d Site Mean In the medium duration trials, in Sayabouly, there was not a significant yield difference among varieties. Pa siev and Do Xao were the highest yielding varieties at both Houay Khot and Houay Hia (Table 7). Based on plant characteristic preferred by farmers and good yield performance in all tested sites, six varieties: Chao do, Do xao, Pasiev, Makfay, Phaesavanh and Med ngay were selected for further evaluation in on-farm trial in

8 Table 7. Yields of the medium duration multilocation yield trails. An * beside a variety name indicates varieties that were selected for further evaluation. Houay Khot Houay Hia Sayabouly Accession # Variety Name Grain Yield (t ha -1 ) LG * Pa siev 2.13 a 1.99 ab 2.31 LG-11297* Do xao 1.68 ab 2.65 a 2.18 LG-6432 Makhine soung 1.48 bc 1.55 cd 2.04 Check 1.29 bcd 1.83 c 2.75 LG-7084 Paneur (Peek) 1.22 b-e 1.66 cd 2.12 Check Vieng b-e 1.73 cd 2.29 LG * Med gnay 0.87 c-f 1.69 cd 2.31 LG-7023 Da cheung 0.82 def 2.14 ab 2.30 LG * Phae savanh 0.80 def 1.52 cd 2.72 LG Khai 0.77 def 1.43 cd - LG-6593 Sangon 0.69 def 1.41 cd - LG-7051 Do hom 0.60 ef 1.44 cd 2.38 LG-6655 Lebmeu 0.50 f 1.03 e 1.92 Site Mean On-farm participatory selection (PVS) Compared to previous years, the 2001 on-farm trials were very successful. PA data was collected from 84% of the farmers and yield data from 63%. Good success can be at least partially attributed to timely training (particularly in participatory research methodology) and continuous follow up during the course of the season. The PA could not be conducted in some cases due to early crop damage by livestock. Yield data could not be collected from some farmers due to damage by livestock, wild pigs or rats. These are factors that the researchers can do little about. However, at four sites, PA and yield data were collected, but due to poor plot layout, the data could not be used in the analysis. Also, related to this, some of the trials were in fields where the local variety (used as the control) had longer growth duration than those being evaluated. This resulted in worse than normal pest damage. In future, there will need to be training in these areas, and simpler designs involving fewer varieties will be used. During the PA, farmers were asked to identify reasons why they liked or did not like particular varieties. These reasons were summarized from all the PAs and are given in Figure 1. Large panicles, big seed and strong stems were the most commonly referred to positive characteristics. In contrast, farmers did not like varieties that produced few tillers, had short panicles or had panicles that matured non-uniformly. All the varieties evaluated were of early and medium duration, therefore, preference for early duration did not stand out prominently in this PA. However, early maturation was mentioned as a positive character and late maturation as a negative character. Earlier research had identified farmer preference for early- and medium-maturing varieties, and therefore, only such varieties were included in the onfarm evaluation. 40

9 Positive plant characteristic Few leaves Weed competitiveness Resistant to lodging Rapid emergence Many panicles Grows in different soils Uniform panicle Looks good Tall plant Early duration High tillering Many seeds Strong stem Large grain Large panicle Number of times mentioned by farmers Negative plant characterisitc Itchy seed Non-uniform plant Many leaves Late duration Few grain Suseptible to pests Non-uniform ripening Grows in some soil Poor emergence Sterile spikelets Small plant Small grain Non-uniform panicle Short panicle Low tillering Number of times mentioned by farmers Figure 1. Results from the preference analysis (PA) summarizing positive and negative plant characteristics most mentioned by farmers. Using these criteria, varieties can now be eliminated early in the evaluation process, saving time and limited resources. For example, passport data (including some of these characteristics) exists on most varieties in the collection. Using this database some varieties could easily be eliminated from the evaluation process. Also, in the OBN-1, variety scoring should include these characteristics, to ensure that varieties selected for further testing will not be rejected by farmers for reasons that are already obvious. In the early trial the highest grain yield (GY) was 2.14 t/ha (Nok; Table 8). This was significantly higher than the average of the local early-maturing varieties that yielded 1.5 t/ha. In the medium trial the highest yield was 1.94 t/ha (Makthoua) that was similar to the average yield of the local varieties (1.82 t/ha). An analysis of data for only those farmers from whom both yield and preference data were obtained showed significant differences in yield and preference in the medium maturity group, but not in the early group (yield, P=0.06; Table 8). Yields and farmer preference ratings (PR) (standardized) were associated. The correlation between variety means over farms for GY and PR was 0.82 in the medium maturity trial and 0.54 in the early trial. Occasionally, high-yielding cultivars were not preferred, but the lowest-yielding cultivars were never preferred. This indicates that agronomic selection for grain yield helps to select varieties farmers prefer. A test of cultivar by province interaction was conducted, using the within-province cultivar x farmer interaction as an error term. There was no detectable cultivar x province interaction for GY. A significant interaction was observed for PR in the early cultivar set; the cultivar Kouyonke was strongly preferred in Luang Prabang, but had a very low preference score in Sayabouly. The reason for this rank reversal is not clear. Certainly, one may expect interactions between cultivars and different farmers, ethnic groups, or environments. However a larger number of farmers must be surveyed for such an analysis to be reliable. 41

10 Table 8. Average grain yields and ranking in Varieties in bold indicate those selected for evaluation in Early varieties Medium varieties Variety Grain yield Standardized Variety Grain yield Standardized (t ha -1 ) farmer preference * (t ha -1 ) farmer preference * Nok Makthoua 1.94 a 0.46 Hom Makhine 1.75 ab 0.62 Taa roon Do hom 1.72 ab 0.49 Kouyongke Dam 1.55 bc 0.39 Vieng Peek 1.48 bc Deng said Da cheung 1.33 c Deng Sangon 1.31 c Sang Lebmeu 1.19 c LSD ns ns * Farmer rank is standardized so that the mean is 0.0 Based on farmers preferences and grain yield, Taa roon, Nok, Do hom, Makhine soung, Dam and Makthoua were selected for further evaluation in Seed of these varieties are being multiplied during the 2002 dry season. A Mother-Baby design will be employed in the 2002 analysis. One Mother trial will be placed in each village which will permit many farmers to evaluate varieties side by side. Baby trials will consist of one farmer receiving one variety to compare with their variety. In each village, we anticipate three or more farmers evaluating each variety (a total of 18 or more farmers per village). The number of villages depends on the success of the seed multiplication. Yield data will be collected from the Mother trial. In the Baby trials, farmers will be asked to compare the two varieties as well as providing information collected on growing conditions, soil type, etc. This information can be gathered during and after (provide some comparison of grain quality) the growing season. Farmers need to be involved earlier in the selection process. We have already indicated that results from the PA could be used to help screen out varieties early in the evaluation process. However, farmers could also be involved in the Multilocation Yield Trials. This was done in 2001 but results are not yet available. Farmer involvement at this stage requires a different methodology. Some lessons learned from this years PVS trails include: Compared to earlier on-farm work the success of on-farm evaluation using participatory methods was very good. In the future, more training emphasis needs to be placed on good plot layout if a number of varieties are to be evaluated in a single location. Better care needs to be taken in working with farmers, so that the duration of the varieties being tested matches the duration of the local check variety. Too many varieties were evaluated with each farmer. This resulted in several problems (1) farmers could not remember variety names, (2) the PA took too long and (3) it was easy for farmers and researchers to get varieties mixed up. Based on our experiences no more than five varieties should be evaluated with any one farmer. When evaluating many varieties a simple question could be, What varieties will you plan to further evaluate next year? 42

11 Farmers were always quick to identify such varieties. These would naturally be the one the farmer is most interested in. In future evaluations it would be good to gather additional information from the farmer related to the environment [such as, soil type, soil color, drought occurrence (if so, early middle or late season)]. This would allow us to test interactions between cultivar and environment. In future evaluations, we need to ask farmers to evaluate post-harvest characteristics of the rice such as threshability, milling quality and taste. This was not done this year due to the limited amount of seed each farmer received. More seed (at least 0.5 kg and preferably 1 kg) needs to be provided to each farmer for evaluation. This will require that a system of seed multiplication is in place. Seed multiplication during the dry season is currently being evaluated. Acknowledgements We would like to express my deepest and sincerest appreciation and gratitude to the students from Nabong Agriculture Collage in Vientiane and workers at NAFRC, Houay Khot for their help and cooperation in the conduct of study. A special thanks to the Swiss Agency for Development and Cooperation (SDC) and the Upland Rice Consortium (URRC) for financial support. 43

12 Annex Annex black rice Observation Nursery-1. Varieties selected for evaluation in the 2002 Observation Nursery 2 are denoted by an *. # Accession no. Days to flowering Plant height per hill Panicle length 10 pan weight (g) Yield measured Yield adjusted * * * * * * * * * * * * * * * * * * * Note: varieties, which yielded < 0.93 tha-1 were not shown in above table. Ranks 44

13 Annex 2. Correlation of selected characteristics of black rice in 2001 Observation Nursery 1. Observation hill -1 length Height 10 panicle weight Grain yield measured Days-to-flower 1.00 Traditional VARIETIES N=133 hill ns 1.00 length ns ns 1.00 Height ns ** ** panicle weight ** ** ns ** 1.00 Grain yield measured ** ** ns ** ** 1.00 Grain yield adjusted Grain yield adjusted ns ns ** ** ** ** 1.00 Observation Height Days-toflower Days-toflower hill -1 length Days-to-flower 1.00 Controlled N=15 hill ns 1.00 length ns ns 1.00 Height ns ns ** 1.00 r 2 ; probabilities of significance: > 0.05 = ns 0.01 to 0.05 = * < 0.01 = ** 45

14 Annex Observation nursery 1 at Houay Khot. Varieties selected for evaluation in 2002 Observation nursery 2 are denoted by an *. Accession No. Variety Name Type Height (cm) Panicle length (cm) Panicle per hill 10 Pan weight (g) Days to flower Yield (t/ha -1 ) measured adj. miss hills Rank 168 LG * G LG * G LG * G LG * G LG * G LG * G LG G LG G LG G LG * G LG * G LG * G LG G LG * G Check (Control) Hom G LG G LG * G LG * G LG * G LG * G LG G LG * G LG G LG G LG * G LG G LG * G LG * G LG G LG G LG * G LG * G LG * G LG * G LG G LG G Average Control LG G LG G LG G LG G LG G Note: varieties, which yielded < 0.93 tha-1 were not shown in above table. 46

15 Annex 4. Correlation of selected characteristics in 2001 Observation Nursery 1 (ordinary rice). Observation hill -1 Panicle length Height 10 panicle weight Grain yield measured Grain yield adjusted Days-to-flower 1.00 VARIETIES N=366 hill ns 1.00 Panicle length ns 0.20 ** 1.00 Height ns 0.41 ** 0.64 ** 1.00 Ten panicle weight ** 0.20 ** 0.42 ** 0.45 ** 1.00 Grain yield measured ** 0.49 ** 0.39 ** 0.52 ** 0.76 ** 1.00 Grain yield adjusted ** 0.46 ** 0.40 ** 0.50 ** 0.76 ** 0.98 ** 1.00 Observation Days-toflower Days-toflower hill -1 length Height 10 Panicle weightt Grain yield measured Grain yield adjusted Days-to-flower 1.00 hill ns 1.00 Controlled N=92 length ns 0.71 ** 1.00 Height ns 0.48 ** 0.61 ** 1.00 Ten panicle weight * 0.26 ns 0.33 * 0.57 ** 1.00 Grain yield measured ns 0.71 ** 0.67 ** 0.72 ** 0.55 ** 1.00 Grain yield adjusted ns 0.70 ** 0.66** 0.72 ** 0.55 ** 0.99 ** 1.00 r 2 ; probabilities of significance: > 0.05 = ns 0.01 to 0.05 = * < 0.01 = ** ) 47

16 Annex 5. Grain yield and agronomic characteristics of early maturing varieties in the 2001 Multi-Location Yield Trial (Houay Khot Station). Accession # Variety Name Type Grain yield (t ha -1 ) Days to flowering (hill 1 ) Height (cm) Panicle length (cm) 10 Panicle weight (g) LG-6501 Deng G 1.31 a 102 ab 6.8 bcd 131 e 21.8 a-e 51 a Check Hom 1 G 1.30 a 103 b 11.8 a 120 d 23.3 ab 34 de LG * Makfay G 1.12 a 101 ab 7.0 bc 118 cd 21.5 b-e 31 e LG-6911 Dengsaid G 0.77 b 101 ab 8.3 b 100 ab 20.5 cde 31 e LG-7771 Kou yong ke G 0.67 b 100 a 5.8 cd 107 ab 21.5 b-e 48 abc LG * Chao do N 0.62 b 100 a 5.8 cd 108 ab 24.0 a 34 de LG Keo G 0.56 b 101 ab 6.0 cd 108 bc 19.5 ef 31 e LG Khiekhang G 0.54 b 104 b 7.0 bc 97 a 16.8 g 35 de LG Chaomeuang kay N 0.54 b 102 ab 6.3 cd 107 ab 19.8 def 39 cde LG-6905 Sang G 0.53 b 101 ab 6.3 cd 104 ab 22.0 a-d 41 bcd LG-1724 Dam G 0.52 b 111 c 5.0 d 118 cd 22.3 abc 50 ab Vieng G 0.51 b 110 c 6.5 bcd 105 ab 18.0 fg 36 de LG-7347 Taaroon G 0.45 b 103 b 7.0 bc 97 a 20.0 c-f 33 de LG-6499 Nok G 0.44 b 102 ab 5.0 d 107 ab 21.3 b-e 38 de ANOVA (PR >F) ** ** ** ** ** ** CV (%) Controls * Varieties selected for evaluation in 2002 on-farm yield trials. 48

17 Annex 6. Grain yield and agronomic characteristics of medium maturing varieties in the 2001 Multi-Location Yield Trial (Houay Khot Station). Accession # Variety Name Type Grain yield (t ha -1 ) Days to flowering LG * Pa siev G 2.13 a 103 a LG-11297* Do xao G 1.68 ab 101 a LG-6432 Mak hine soung G 1.48 bc 105 a Hom G 1.29 bcd 103 a LG-7084 Paneur (Peek) G 1.22 b-e 116 c Check Vieng 1 G 1.18 b-e 104 a LG * Med gnay G 0.87 c-f 102 a LG-7023 Da cheung G 0.82 def 103 a LG * Phae savanh G 0.80 def 103 a LG Khai G 0.77 def 102 a LG-6593 Sangon G 0.69 def 115 bc LG-7051 Do hom G 0.60 ef 102 a LG-6655 Lebmeu G 0.50 f 112 b ANOVA (PR >F) ** ** CV (%) (hill 1 ) 9.5 ab 10.5 a 7.3 cde 10.3 a 6.0 de 7.8 bcd 6.0 de 6.0 de 7.5 cd 6.3 de 6.5 de 9.0 abc 5.5 e ** 16.1 Height (cm) Panicle length (cm) 10 panicle weight (g) 134 de 22.8 a-d 39.0 a 133 cde 24.0 abc 38.0 a 137 e 25.8 a 63.0 a 126 b-e 24.0 abc 34.0 a 115 ab 22.8 a-d 46.0 a 120 abc 21.3 cd 46.0 a 124 b-e 24.3 abc 51.0 a 125 b-e 24.8 ab 47.0 a 120 a-d 23.0 a-d 40.0 a 118 ab 23.3 a-d 44.0 a 114 ab 22.0 bcd 55.0 a 112 ab 22.8 a-d 36.0 a 109 a 20.8 d 31.0 a ** ** **? Controls * Varieties selected for evaluation in 2002 on-farm yield trials. 49

18 Annex 7. Grain yield and agronomic characteristics of early maturing varieties in the 2001 Multi-Location Yield Trial (Ban Houay Hia). Accession # Variety name Type Grain yield (t ha -1 ) Days to flowering (hill 1 ) Height (cm) Panicle length (cm) 10 panicle weight (g) LG-6501 Deng G 1.5 b 110 e 5.8 b 117 ab 22.5 cde 41.3 cde Check Hom 1 G 2.3 ab 102 a 7.5 ab 122 b 24.5 abc 35.0 e LG * Makfay G 2.8 a 106 b-e 7.5 ab 119 ab 22.8 b-e 48.8 a-d LG-6911 Dengsaid G 1.8 b 105 abc 6.5 ab 112 ab 24.0 a-d 35.0 e LG-7771 Kou yong ke G 1.5 b 107 cde 6.0 b 118 ab 23.0 b-e 56.3 a LG * Chao do N 2.3 ab 103 ab 7.5 ab 119 ab 25.3 a 35.0 e LG Keo G 2.3 ab 104 ab 6.8 ab 119 ab 23.0 b-e 41.3 cde LG Khiekhang G 2.3 ab 106 a-d 6.3 b 115 ab 21.3 e 40.0 de LG Chaomeuang kay N 1.8 b 105 a-d 8.3 a 114 ab 21.8 e 47.5 a-d LG-6905 Sang G 2.3 ab 102 a 7.5 ab 118 ab 22.0 de 42.5 b-e LG-1724 Dam G 2.3 ab 108 de 6.5 ab 118 ab 25.5 a 51.3 abc Vieng G 2.3 ab 109 de 6.8 ab 118 ab 21.5 e 50.0 a-d LG-7347 Taaroon G 2.0 ab 110 e 5.8 b 110 a 23.0 b-e 52.5 ab LG-6499 Nok G 1.8 b 109 de 6.8 ab 114 ab 24.8 ab 51.3 abc ANOVA (PR >F) ns ** ns ns ** ** CV (%) Controls * Varieties selected for evaluation in 2002 on-farm yield trials 50

19 Annex 8. Grain yield and agronomic characteristics of medium maturing varieties in the 2001 Multi-Location Yield Trial (Ban Houay Hia). Accession # Variety Name Type Grain Yield (t ha -1 ) Days to Flowering (hill 1 ) Height (cm) Panicle length (cm) 10 Panicle weight (g) LG-11297* Do xao G a 109 ab 10.9 a 123 a-e 23.0 e-g 71.3 LG-7023 Da cheung G ab 109 abc 5.9 g 129 ab 26.9 a 75.0 LG * Pa siev G ab 110 ab 10.4 ab 115 gh 22.1 i 74.8 Hom G c 110 ab 10.4 ab 120 a-e 23.6 c-f 67.5 Local Phae deng 1 G cd 110 ab 9.6 abc 112 i 23.1 c-f 70.3 LG * Med gnay G cd 106 efg 5.9 g 130 a 24.3 c-f 75.0 LG-7084 Paneur (peek) G cd 111 a 5.7 g 117 e-g 22.7 gh 65.0 LG-6432 Mak hine soung G cd 109 ab 5.6 g 125 a-d 25.0 abc 75.0 LG * Phae savanh G cd 105 gh 7.0 e-f 117 e-g 24.5 cde 72.5 LG-7051 Do hom G cd 107 cde 7.7 c-e 119 a-e 25.2 abc 77.5 LG Khai G cd 104 i 6.6 g 121 a-e 24.2 c-f 74.0 LG-6593 Sangon G cd 110 ab 8.3 a-d 128 a-c 25.3 ab 75.5 LG-6655 Lebmeu G e 108 abc 7.2 ef 118 c-f 25.1 abc 73.3 ANOVA (PR >F) * ** ** ns ** ns CV (%) Controls * Varieties selected for evaluation in 2002 on-farm yield trials. 51

20 Annex 9. Grain yield and agronomic characteristics of early maturing varieties in the 2001 Multi-Location Yield Trial (Sayabouly province). Accession # Variety Name Type Grain yield (t ha -1 ) Days to flowering (hill 1 ) Height (cm) Panicle length (cm) 10 Panicle Weight (g) LG-6501 Deng G 2.56 abc 97 ab Check Khao tia 1 G 1.76 de 90 f LG * Makfay G 2.77 ab 96 ab LG-6911 Dengsaid G 1.41 e 96 abc LG-7771 Kou yong ke G 2.32 e-d 96 ab LG * Chao do N 2.16 b-e 93 def LG Keo G 1.51 a-e 96 ab LG Khiekhang G 2.97 a 93 cde LG Chaomeuang kay N 2.88 ab 95 bcd LG-6905 Sang G 1.94 cde 91 ef LG-1724 Dam G 2.74 ab 97 ab Vieng G 2.13 b-e 97 a LG-7347 Taaroon G 1.91 cde 96 ab LG-6499 Nok G 2.25 a-d 95 abc ANOVA (PR >F) *** ** ns ns ns ns CV (%) control * Varieties selected for evaluation in 2002 on-farm yield trials. 52

21 Annex 10. Grain yield and agronomic characteristics of medium maturing varieties in the 2001 Multi-Location Yield Trial (Sayabouly province). Grain Yield Days to Height (cm) Panicle length 10 Panicle weight Accession # Variety Name Type (t ha -1 ) Flowering (hill 1 ) (cm) (g) 6.0 LG * Pa siev G bc e 4.7 LG-11297* Do xao G bc e 6.2 LG-6432 Mak hine soung G a b 5.7 Hom G bc d 6.2 LG-7084 Paneur (Peek) G a a 5.0 Check Deng du 1 G c d 7.0 LG * Med gnay G bc d 5.2 LG-7023 Da cheung G b d 6.0 LG * Phae savanh G bc bc - LG Khai G LG-6593 Sangon G LG-7051 Do hom G c c 3.5 LG-6655 Lebmeu G a c ANOVA (PR >F) ns *** ns ns ns *** CV (%) Controls * Varieties selected for evaluation in 2001 on-farm yield trials. 53