International Rice Research Notes

The International Rice Research Notes (IRRN) expedites communication among scientists concerned with the development of improved technology for rice and ricebased systems. The IRRN is a mechanism to help scientists keep each other informed of current rice research findings. The concise scientific note are meant to encourage rice scientists to communicate with one another to obtain details on the research reported. The IRRN is published three times a year in April, August, and December by the International Rice Research Institute; annual subject and variety indexes are also produced.

Contents
April 1996
Germplasm improvement

Focus on rice genetics

The specter of food shortages is looming once again, with the annual rate of increase of rice production slowing to where it is lower than the rate increase of rice consumers. Recent advances in cellular and molecular genetics of rice have come perhaps in the nick of time to provide us with new tools to develop rice varieties for the future. Only 10 years ago, the status of rice genetics was considered far behind that of other food crops, such as maize and wheat. The past decade, however, has seen an explosion of knowledge in this arena. Rice is now considered a model plant for such research on cereal crops. In October 1995, IRRI hosted the Third International Rice Genetics Symposium. More than 500 scientists from 31 countries attended. Along with a dramatic increase in the attendance over the years has come a major shift in the complexion of the program. During the first symposium in 1985, around 90% of the papers were on classical genetics; at this symposium, about 80% of the papers addressed topics on cellular and molecular genetics. The key papers presented are being published as an IRRI book. The posters displayed at the symposium appear as notes (in a modified format) throughout this issue of IRRN, and will also be featured in the next two issues. They are denoted by the symbol. We hope you find these notes to be valuable source of information.

Genetic resources Mobile genetic elements useful for classifying and determining relationships between rice strains 4 Cytological analysis of a natural hybrid between Oryza minuta and O. officinalis 4 Screening RFLP probes for differentiating indicas and japonicas 5 Overwintering ability of Dongxiang wild rice in Wuhan District, China 6 Genetic basis of variation in weedy rice collected from Republic of Korea 7 Allozyme variability and inter- and intrapopulational gene diversity of Oryza glumaepatula in the Amazon Basin 8 Germplasm use in Ecuador and Bolivia 8 Distribution of ribosomal DNA polymorphism in wild and cultivated rice from China 9 A research program for on-farm conservation of rice genetic resources 10 Genetics Genetics of the new plant type 11 Molecular basis of heterosis in hybrid rice and hybrid maize revealed by mRNA amplification 12 Linkage analysis of a photoperiod-Sensitive gene Se1 locus with neighboring loci in rice 13 Genetic analysis of plant regeneration ability from cell suspension cultures of rice 14 Classification of Korean rice germplasm based on isozyme polymorphism 15 Heritability and genetic correlation for component characters of yield sink capacity in rice 15 Heading-time genes control photoperiod insensitivity of rice cultivar Norin 20 15 QTL analysis of rice seedling vigor in japonica and indica genetic backgrounds 16 RFLP mapping of QTLs for yield and other related characters in rice 17 Use of DNA markers in constructing multilines 19 Breeding methods Development of cold-tolerant rice through anther culture A comparison of the efficiency of four breeding methods A promising dwarf rice mutant induced through gamma irradiation 21 Combining ability analysis on restorability for cytoplasmic sterility in hybrid rice 22 Male sterile line in rice (Oryza sativa) developed with O. glumaepatula cytoplasm 22 Pattern of segregation for grain weight involving bold- and grained rice varieties 23 20 20

IRRN production team

. . . . . . . . . . . . . . .

male

Editor: Carolyn Dedolph Assistant editor: Teresita Rola Layout and design: Erlie Putungan Production supervisor: Millet Magsino Editorial assistant: Luisa Gelisan Typesetting: Erlie Putungan Cecilia Gregorio Artwork: Jess Recuenco

slender-

IRRN 21:1 (April 1996)

Yield potential The yield potential of three crosses between diverse parents of indica type 24 Stability analysis of 13 early-duration upland rice genotypes in Bastar Plateau Zone, Madhya Pradesh, India 24 Variability, heritability, genetic advance, and genetic divergence in upland rice 25 Response of rice genotypes to N fertilization under temperate conditions of Kashmir Valley 27 Grain quality Enzyme isolation and regulation with lysine biosynthesis and degradation in developing seeds of rice 27 Pest resistancediseases Use of genetic male sterile-facilitated recurrent selection for blast resistance in rice 28 Mapping a new gene for resistance to bacterial blight using RFLP markers 30 Pest resistanceinsects Influence of rice plant morphology on leaffolder incidence A virulent rice gall midge biotype in Manipur 31 Pest resistanceother pests Selection for weed competitiveness in upland rice 31

Biochemical identification and genetic analysis of molybdenum cofactor mutants in rice 41 lntegrated germplasm improvement Altering associations between characters in rice through gamma radiation 42 lntegrated germplasm improvementirrigated Liangyou Peite, a new two-line hybrid rice released in China 42 Khushboo, a quality rice cultivar for Rajasthan 43 Bright-pearl 1, a fast grain-filling japonica rice 43
Crop and resource management

Physiology and plant nutrition Growth characteristics of IRRI-developed new rice plant type breeding lines in Japan 44 Fertilizer managementorganic sources Breaking seed dormancy in different leguminous forage species 45 Fertilizer managementinorganic sources Responses of promising rice genotypes to nitrogen levels in irrigated lowlands 46 Nitrogen use efficiency for three fertilizers in irrigated rice 47 Crop management Response of rice hybrid PMS2 A/IR31802 to seedling vigor and nitrogen levels in Haryana, India 47 lntegrated pest managementdiseases Fine and physical mapping toward the positional cloning of an indica-derived blast resistance gene Pi-b 48 Effect of botanicals on managing sheath rot of rice 49 Efficacy of lpomoea cornea in controlling rice sheath rot 50 lntegrated pest managementinsects Effect of steam distillate extracts of rice cultivars on rice thrips Adulterated pesticides in Cambodia 51
Research methodology

32

Stress tolerancedrought Genetic variability of osmotic adjustment under drought stress in rice 33 Stress toleranceexcess water Characterization of the pyruvate decarboxylase gene family of rice and its potential application to submergence tolerance 33 Molecular implication of submergence tolerance in rice using expressed sequence tags as probes 34 Stress toleranceadverse temperature Mapping of genes responsible for cold tolerance at the booting stage of rice 35 Variation in leaf color and thylakoid protein analyzed in chlorophyll rice mutant 36 Cold tolerance at booting stage of highly cold-tolerant rice lines derived from a javanica and a japonica 37 Stress toleranceadverse soils Variation in salt tolerance of rice plants regenerated from saltselected calli of a susceptible variety 38 Genetics of salinity tolerance and ionic uptake in rice 38 Stress tolerance Epigenetic control of tolerance for transient low light stress in rice 39 Stress toleranceother stresses Screening somaclonal variants with tolerance for both shade and photooxidation in rice 40
I

50

Leaf carbohydrate analysis-a simple method for integrating daily canopy photosynthesis 52 A weather-based empirical model to predict rice leaf blast in Thailand 52 Use of lesion number per leaf area to estimate lesion size for leaf blast quantitative studies 53 A simple method for estimating yield loss of rice due to severe panicle blast 54 Inoculation method for rice bunt 55

Note to readers To save on costs, the IRRN is now being published three times a year: April, August, and December. We have also discontinued the IRRN sections on news about research collaboration and announcements. The material previously covered in these sections is now included in the Rice Reporter.

IRRN 21:1 (April 1996)

Germplasm improvement
Genetic resources
Mobile genetic elements useful for classifying and determining relationships between rice strains
R. Motohashi, H. Ohtsubo, and E. Ohtsubo, Institute of Molecular and Cellular Biosciences, University of Tokyo, Yayoi 1-1-1, Bunkyo-ku, Tokyo 113, Japan
Distribution of p-SINE 1a Species O. sativa cv Nipponbare cv IR36 O. glaberrima W025 O. longistaminata W1451 P1 O. meridionalis W1625 O. glumaepatula W1192 p-SINE 1-r30 + + pSINE 1-r29 +b + + + + p-SINE 1-r25 + + p-SINE 1-r38 +c +c +c + +c

was examined by PCR using a relevant pair of primers that hybridize to the regions flanking the p-SINE 1 member. + and indicate that the fragments with or without p-SINE 1, respectively, are generated by PCR. bPCR fragments with pSINE1 have a tandem duplication. c PCR fragments with p-SINE 1 contain Tnr3. d indicates that both fragments with and without p-SINE 1 are generated by PCR.

a p-SINE1 members identified in O. sativa. The presence or absence of each p-SINE 1 member at the respective locus

Morphological and physiological studies have played an important role in classifying species in the genus Oryza. Although detailed information on the classification of rice species has been accumulated with the development of molecular biological technology, relationships between Oryza species with AA genome have not been clarified. We have identified 37 members of p-SINE1, a plant SINE (short interspersed elements) first found in rice (see figure), located at various loci. Most of these members are at the corresponding loci in Oryza species with the AA genome. However, several of them are present in a few Oryza species. Besides those described previously, a member of p-SINE1-r30 is only in O. sativa, p-SINE1-r29 is in all the AA genome Oryza species except for O. longistaminata, and p-SINE1-r25 is in all of the AA genome Oryza species except for O. meridionalis (see table). Another member, p-SINE1-r38, contained an insertion that was 1536 bp long, with imperfect terminal inverted repeats of about 13 bp beginning with 5'CACTA3'(see figure). The insertion is

Structures of p-SlNE 1 (a) and pSINE 1-r38 (b).

similar to that of the members of the En/ Spm transposable element family such as En/Spm in maize, Tam1 in snapdragon, Tgml in soybean, Pis1 in pea, and Tpn1 in Japanese morning glory. This insertion sequence is thus named Tnr3 (transposable element in rice #3). The Tnr3 also carries long subterminal regions containing direct and inverted repeats of short DNA sequences of 15 bp, another characteristic of Intergenetic or interspecific hybrids are important for studying the evolutionary relationships between plant species, particularly those occurring spontaneously in nature. One Oryza population collected in 1990 from San Roque, Hilongos, Leyte, Philippines (P90-18; by D. A. Vaughan, L. Engle, N. Altoveros, E. Quintana, and T. Borromeo) was completely sterile, with nondehiscent anthers, nonstainable pollen grains, and no seed set. Morphologically,

the En/Spm family. O. sativa, O. rufipogon, O. glaberrima, and O. glumaepatula contain p-SINE 1-r38 with Tnr3, but O. meridionalis contains only p-SINE1-r38 (see table). O. longistaminata, however, does not contain p-SINE1-r38. The p-SINE1 members listed in the table might become useful DNA markers for classifying and determining relationships among Oryza species with AA genome. this population resembled O. minuta J. S. Presl. ex C. B. Presl., a tetraploid Oryza species endemic to the Philippines that has 48 chromosomes containing two distantly related genomes (BBCC). This population was originally identified as sterile O. minuta and maintained in a nursery screenhouse for further observation. Cytological observation of both root tip cells and pollen mother cells (PMCs) confirmed this population is a triploid

Cytological analysis of a natural hybrid between Oryza minuta and O. officinalis

B. R. Lu and M. E. B. Naredo, IRRI; T. H. Borromeo, Agronomy Department, University of the Philippines Los Baos/Philippine Rice Research Institute, Muoz, Nueva Ecija, Philippines

IRRN 21:1 (April 1996)

Meiotic configurations in metaphase I of the sterile O. minuta. Cells observed (no.) 15

a

2n= 36 aa ra

Chromosome configuration I 8.60 (5-14) II 1 b 1.74 (1-4) II 2 b 10.13 (6-14) Ill 0.93 (0-2) IV 0.20 (0-1)

Chiasma/cell 24.80 (19-30)

a = average, r = range: b II1 = rod bivalents, II 2 = ring bivalents.

Anaphase I of the sterile O. minuta showing 36 chromosomes, including 11 laggards; two overlapping chromosomes are indicated by the arrow head.

(2n=3x) with 36 chromosomes. Meiotic analysis of its PMCs further indicated its irregular meiosis at different stages. An average of 8.60 univalents, 11.87 bivalents, and a certain amount of multivalents were scored at metaphase I (see table). Meiosis beyond metaphase I was complicated by

other irregularities. Numerous chromosomes were observed to lag at anaphase I and anaphase II (see figure). One chromatid bridge accompanied by a fragment was found at anaphase I. Micronuclei were scored in all quarters. We concluded that this sterile O. minuta was

a spontaneous hybrid between O. minuta and O. officinalis, which is classified as a diploid wild rice species (2n=2x=24) containing the CC genome. The sterile population was reported to grow sympatrically with O. minuta and O. officinalis at the original site. Numerical analysis of the meiotic configurations in this natural hybrid suggested the best fitting genomic model to be 2:1, with a c 2 value (relative affinity of the most closely related genomes) of 0.969 and WSSD (weighted sums of squares of differences) of 17.1. This indicates the presence of two closely related genomes (homologous) and one distantly related genome (homoeologous) in the natural hybrid. The two closely related genomes should therefore be the CC genomes derived, respectively, from O. minuta (BBCC) and O. officinalis (CC). The high frequency of bivalent formation reveals high homology between the CC genomes from the two species. The single distantly related genome should then be the B genome derived from O. minuta. The relatively high frequency of multivalents (>1 per cell) and low univalents indicates homoeologous pairing between either the CC genome or the B genome. Furthermore, reciprocal chromosomal translocations in the CC genome of one of the parental species probably causes multivalents to be formed. The chromatid bridge suggests an inversion of a chromosome segment in one of the genomes.

Screening RFLP probes for differentiating indicas and japonicas

Jie-Yun Zhuang, Hui-Rong Qian, Hong-Xuan Lin, Jun Lu, Kang-Le Zheng. Biotechnology Department, China National Rice Research Institute, Hangzhou 310006, China

Restriction fragment length polymorphism (RFLP) is an effective and reliable tool for classifying the subspecies of Oryza sativa L. However, a RFLP survey of the entire rice genome is time-consuming, laborintensive, and expensive. We report our progress on identifying a set of RFLP probes for differentiating indicas and japonicas.

In our previous RFLP survey of three indica testers and three japonica testers used for screening widely compatible varieties in China, 68 out of 160 probes produced identical hybridization patterns among rice varieties of the same subspecies and different patterns among rice varieties of different subspecies (Zheng et al 1994). Seven established varieties of various origins for both indicas and japonicas were used in RFLP analysis with 68 probes, combined with restriction enzymes DraI, EcoRI, EcoRV, and HindIII. Twenty-one probes regenerated different hybridization patterns between indica and japonica subspecies with at least one enzyme. The majority also regenerated identical patterns within a subspecies. A set of 13 probes for

differentiating subspecies was selected based on Southern blotting performance and chromosome distribution (Qian et al 1995). To determine the applicability of the indica-japonica differentiation probes, 52 cultivated Asian rice varieties with different geographic distributions were selected and assayed with 100 DNA probes in combination with a single enzyme. The proportions of shared DNA fragments and genetic distances between varieties were quantified according to Nei (1987) based on 184 polymorphic fragments of 65 polymorphic probes. Using the unweighted paired group method with arithmetic mean (Sokal and Michener 1958), the varieties were clearly clustered into two groups.

IRRN 21:1 (April 1996)

Indica-japonica discriminate parameters of specific fragments were then computed following the method of Morishima and Gadrinab (1987) with a slight modification (Zhuang et al 1995). Meanwhile, the 52 varieties were assayed with all the indica-japonica differentiation probes except RG869 and RG684. The discriminate parameters of specific fragments were calculated as (DiDj) = 1 for indica fragments and (Di-Dj) = 1 for japonica fragments. Average discriminate values of specific varieties were calculated and plotted (Fig. 1). Two sets of probes generated similar results, but the degree of indica-japonica differentiation was enlarged by the differentiation probes, as expected. A similar study was conducted for 20 accessions of O. rufipogon. Based on data from the 11 differentiation probes and 113 fragments of 40 DNA probes, average discriminate values of the specific accessions were calculated and plotted (Fig. 2). All followed a general trend. However, the differentiation probes revealed a distinguishable differentiation for indicas and japonicas within O. rufipogon while the randomly selected probes did not.

1. Average indica-japonica discriminate values of 52 rice varieties.

a

Based on data of 11 subspecies differentiation probes. b Based on data of 184 fragments of 65 polymorphic probes.

2. Average indica-japonica discriminate values of 20

Based on data of 113 fragments of 40 polymorphic probes b Based on data of 11 subspecies differentiation probes.

accessions of O. rufipogon.
a

Cited references
Morishima H, Gadrinad LU. 1987.Are theAsian common wild-rices differentiated into the indica and japonica types? In: Crop exploration and utilization of genetic resources. Proceedings of International Symposium, 6-12 Dec 1986; Chanhua, Taiwan. Taichung District Agricultural Improvement Station. p 11-19. Nei M. 1987. Molecular evolutionary genetics. New York: Columbia University Press. p 106- 107. Qian H-L, Zhuang J-Y, Lin H-X, Lu J. 1985. Identification of a set of RFLP probes for

subspecies differentiation in Oryza sativa L. Theor. Appl. Genet. 90:878-884. Sokal RR, Michener CD. 1958. A statistical method for evaluating systematic relationships. Univ. Kansas Sci. Bull. 28:1409-1438. Zheng K, Qian H, Shen B, Zhuang J, Lin H, Lu J. 1994. RLFP-based phylogenetic analysis of

wide compatibility varieties in Oryza sativa L. Theor. Appl. Genet. 88:65-69. Zhuang J-Y, Qian H-R, Lin H-X, Lu J, Cheng SH, Ying C-S, Luo L-J, Zhu X-D, Dong F-G, Min S-K, Sun Z-X, Zheng K-L 1995. RFLPbased analysis of the origin and differentiation of Oryza sativa L. [in Chinese]. Chin. J. Rice Sci. 9(3): 135-140.

Overwintering ability of Dongxiang wild rice in Wuhan District, China

Guangcun He and Lihui Shu, College of Life Sciences, Wuhan University, Wuhan 430072, China; Yichang Zhou, Huanggang Seed Farm, Huanggang, Hubei 431629, China; and Lanjie Liao, College of Life Sciences, Wuhan University

Dongxiang wild rice is found in Dongxiang County, Jiangxi Province, in the rice belt of central China. The climate is typically continental with extreme minimum temperatures below -10 C. Dongxiang wild rice belongs to Oryza rufipogon, which is widely distributed from 18 to 28 14' northern latitude in China.

Wild rice species have been studied to identify useful properties for rice breeding programs, but reports are few on their abilities to overwinter. Therefore, we studied the overwintering ability of Dongxiang wild rice. In 1993, 14 accessions of O. rufipogon were collected from Guangdong, Guangxi,

IRRN 21:1 (April 1996)

than 30 tillers per hill. Heading occurred in late September. Wuhan is 2.3 higher in latitude than is Dongxiang County. The extreme minimum winter temperatures were -8.0 C in 1993 and -3.5 C in 1994. The underground stems and roots of Dongxiang wild rice have strong cold tolerance, enabling the plants to survive the winter and to grow again in early spring. Dongxiang wild rice is an ideal source of cold tolerance. It could potentially be used to develop a new type of rice that can overwinter in rice-growing areas with climates similar to that of central China.

Sprouts from the overwintered stubble of Dongxiang wild rice.

Genetic basis of variation in weedy rice collected from Republic of Korea

H. S. Suh, College of Natural Resources, Yeungnam University, Kyongsan 712-749, Republic of Korea; Y. C. Cho, Crop Experiment Station, Rural Development Administration (RDA), Suweon, Republic of Korea; T. Y. Chung, National Agricultural Science and Technology Institute, RDA; and H. Morishima, National Institute of Genetics, Mishima 411, Japan

Hunan, and Jiangxi provinces and planted in ricefields in Wuhan District. Normal management practices were followed. Plants were left in the field to overwinter naturally. Many seedlings were growing in the stubble in the spring of 1994. Seedlings were pulled up to ascertain from where they came. Most had germinated from seeds shed during the previous year, but some had sprouted from underground stems. Dongxiang wild rice was retested in 1994. This time, the panicles were bagged after heading and the aboveground plant parts were cut off to prevent seed-germinated seedlings. Only stubble remained over winter. In the spring of 1995, when the seedlings were about 15 cm, stubble was removed to inspect growth (see figure). The underground stems were yellowish green and viable, and the roots were white and able to support new plant growth. The underground stem had three nodes, with 1-3 seedlings or buds regenerating from each (see figure). Sprouted seedlings were transplanted into pots and into the field. Those germinated from seeds served as a control. The plants in pots were subjected to a 10-h short-day treatment. Heading occurred in late July. The average seed-setting rate was 54%, which was the same as for the control. The spikelets turned black at maturity and were prone to shattering. The plants in the field grew vigorously and developed more

Variations in morphophysiological characters-such as grain shape (length/ width ratio), glume hair length, 100-grain weight, resistance to KClO3 phenol reaction, seed dormancy, and seed shatteringand allelic variation at 14 isozyme loci were tested for 141 strains of weedy rice selected randomly from more

than 1,000 strains collected from farmers' fields in the Republic of Korea. We investigated variations in 37 restriction fragment length polymorphism (RFLP) and 13 randomly amplified polymorphic DNA (RAPD) markers from 24 strains in addition to seed fertility of F1s between weedy and cultivated rices. Based on grain shape, Korean weedy rices were classified into long- and shortgrained types. The long-grained type came only from the southern regions of the peninsula, while the short-grained type was distributed throughout the peninsula. The F1s of long-grained weedy rice and indica testers had 48% seed fertility and those with japonica testers, 27%. Those of shortgrained weedy rice and indica testers had 37% seed fertility and those with japonica testers, 69%. By principal component analysis based on morphophysiological characters and isozyme variations, the weedy rice strains were divided into two distinct groups (see figure). All of the long-grained strains were grouped into the same cluster as indica cultivars IR36 and Peh-kuh; all of the shortgrained types were clustered with japonica cultivars Taichung 65 and Nakdongbyeo. In RFLP analysis, a high level of polymorphism was found between long-grained and short-grained type, while lower polymorphism was found within each grain type. Variations in the short-grained type were narrow but those in the long-grained type were broad. The long- and short-grained types were also distinctly classified by RAPD markers.

Scatter diagram of 141 strains of Korean weedy rices by the first vectors of principal component analysis based on morphophysiological (Im) and isozyme (li) variations.

IRRN 21:1 (April 1996)

Based on morphophysiological characters, F1 fertility, variations in isozymes, and RFLP and RAPD markers, the Korean short-grained weedy rices are a relatively uniform group similar to typical japonicas, while the long-grained weedy rices are diverse and more similar to indica cultivars than to japonica cultivars.

Summary of allozyme variation in O. glumaepatula for 30 loci of 16 enzymes within several regions: proportion of polymorphic loci (P), mean number of alleles per locus (A), observed heterozygosity (Hobs), and Nei's gene diversity parameters. a Populations (no.) Rio Negro 1 Rio Negro 2 Rio Solimes 1 Rio Solimes 2 Rio Solimes 3 Total lnterregion 3 6 11 8 6 34 5 P 0.20 0.27 0.67 0.53 0.23 0.73 A 1.27 1.27 1.77 1.63 1.23 1.93 Hobs 0.002 0.001 0.006 0.002 0.002 0.003 HT a 0.059 0.054 0.119 0.091 0.009 0.109 0.118 H sa 0.040 0.026 0.071 0.052 0.008 0.043 0.068 DSTa 0.018 0.028 0.048 0.039 0.001 0.067 0.050 G STa 0.310 0.521 0.403 0.430 0.109 0.610 0.421

Allozyme variability and inter- and intrapopulational gene diversity of Oryza glumaepatula in the Amazon Basin
M. Akimoto, Y. Shimamoto, Faculty of Agriculture, Hokkaido University, Sapporo, 060 Japan; H. Morishima, National Institute of Genetics, Mishima, 411 Japan

a H = expected heterozygosity, H = intrapopulational gene diversity, D T s ST = interpopulational gene diversity, GST = coefficient of genetic differentiation.

Oryza glumaepatula is a diploid wild rice species with the AA genome. It is found in tropical Central and South America. One of its unique ecotypes grows in the Amazon Basin, where water levels annually oscillate by as much as 10 m. The culms of these plants easily break at a specific growth stage. With no roots to anchor them to the ground, the plants float on the water and move downstream with the river current and wind. To analyze the gene flow system and genetic diversity of populations of Amazonian O. glumuepatula, we examined allozyme variability of 30 loci of 16 enzymes from 34 natural populations distributed in five regions. Several parameters, including fixation index and Nei's gene

diversity, were used to assess allozyme variability of each region. Allozymes were not so variable, although populations examined were collected from nearly 20,000 km 2 . Overall genetic diversity had low values compared with those of the Asian wild rice, O.rufipogon. But population genotypes tend to be differentiated among regions isolated geographically. In Rio Solimes, the main river of the Amazon, diversity, measured as the proportion of polymorphic loci (P), average number of alleles per locus (A), and total heterozygosity (HT = expected heterozygosity), was found to increase when going from the upper to the lower basin (see table). Gene flow most probably proceeds in a one-way direction from the upper to the lower Amazon basin. In Rio Negro, one of the Amazon's largest tributaries, diversity was not different between the upper and lower basins. The Rio Negro's water has poor

nutrition and low pH compared with that of other rivers, making it unfavorable for plant growth. O. glumaepatula colonization in the Rio Negro is limited. Observed heterozygosities (H obs ) were much lower than expected, and the fixation index averaged over P was nearly equal to 1 (FIS = 0.954). This indicates that Amazonian O. glumaepatula has developed selfpollination systems, enabling it to produce seed under unstable conditions. O. glumaepatula did not show clear population differentiation when compared with predominantly selfing annual populations of O. rufipogon that show higher interpopulational gene diversity (DST) rather than intrapopulational gene diversity (HS) (see table). The degree of gene flow among populations is probably much higher in O. glumaepatula than in O. rufipogon. The ability to float on water and selfpollination are important factors responsible for the genetic diversity observed in Amazonian O. glumaepatula.

Germplasm use in Ecuador and Bolivia

L. E. Berrio and E. P. Guimares, Centro lnternacional de Agricultura Tropical (CIAT), Apto. Aereo 6713, Cali, Colombia

IRRI and CIAT coordinate the International Network for Genetic Evaluation of Rice for Latin America and the Caribbean (INCERLAC). Since its inception in 1976, countries in the region have actively used the network to exchange germplasm and information. INGER-LAC has distributed 5,458 advanced lines to national partners through

2,272 nursery sets of 14 types from 1979 to 1993. To better understand and document the value of this service, we compared the use of introduced germplasm by national programs in Ecuador and Bolivia. We considered the number of nurseries (lines) introduced, number of varieties released from them, time taken to release a variety from an introduced line, and efforts needed to release a variety. Information was obtained from the INGER-LAC data base. From the nurseries requested and dispatched to the two countries, we determined the number of lines introduced yearly. Ecuador released

three varieties and Bolivia, six (Tables 1 and 2). The period covered was the time between the release of the first variety and the release of the most recent variety in each country. Having identified the first variety released, we then determined the year in which the line was introduced. The first variety was released in Ecuador in 1986, based on a line from a 1982 nursery, and in Bolivia, in 1987 from a line provided in a 1979 nursery. Ecuador introduced 3,201 lines from 64 nurseries and released three varieties from 1982 to 1994. The national program evaluated an average of 915 lines to release a

IRRN 21:1 (April 1996)

Table 1. Germplasm use by the national program of Bolivia. Year Nurseries (no.) Dispatched 7 6 9 2 1 2 2 1 1 2 3 5 6 5 3 55 3.7 Returned 3 3 7 2 1 1 0 1 1 2 3 3 1 1 2 31 2.1 Return (%) 42.9 50.0 77.8 100.0 100.0 50.0 0.0 100.0 100.0 100.0 100.0 60.0 16.7 20.0 66.7 56.4 Linesa (no.) 256 217 201 190 30 208 111 230 179 98 137 42 101 43 74 2117 141 Varietal release Number 0 0 0 0 0 0 0 0 2 0 0 0 0 0 2 2 6 0.4 Year of introduction 1979/1981 1987/1987 1988/1988 Time span (yr) 8/6 6/6 6/6 6.3

Distribution of ribosomal DNA polymorphism in wild and cultivated rice from China
K. D. Liu, National Key Laboratory of Crop Genetic Improvement (NKLCGI), Huazhong Agricultural University, Wuhan 430070, China; G. P. Yang, Crop and Soil Environmental Sciences Department (CSESD), Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA; Qifa Zhang, NKLCGI, Huazhong Agricultural University; S. H. Zhu, Genetics Department (GD), Wuhan University, Wuhan 430072, China; M. A. Saghai Maroof, CSESD, Virginia Polytechnic Institute and State University; and X. M. Wang, GD, Wuhan University

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Total Av

aBolivia evaluated 287 lines to release a variety. This average results from dividing the total number of lines intro-

duced from 1979 (the year the line was Introduced that gave rise to the first variety) to 1988 (the year the line was i ntroduced that resulted in the most recently released variety) by the number of released varieties.

Table 2. Germplasm use by the national program of Ecuador. Year Nurseries (no.) Dispatched 9 6 65 6 3 5 5 8 7 3 1 5 64 5.3 Returned 1 2 2 2 2 0 0 3 0 0 0 0 12 1.0 Return (%) 11.1 33.3 33.3 33.3 66.7 0.0 0.0 37.5 0.0 0.0 0.0 0.0 18.8 Linesa (no.) 409 411 373 336 381 290 217 235 94 214 157 84 3201 267 Varietal release Number 0 0 0 0 1 0 0 1 0 0 0 0 1 3 0.2 Year of introduction 1982 1986 1990 Time span (yr) 4 3 4 11 3.7

Chinese rice comprises a major portion of the worlds rice gene pool. The objective of this study was to gain understanding of the level and distribution of genetic diversity in Chinese rice germplasm using ribosomal DNA (rDNA) spacer length polymorphisms as markers. We surveyed three sets of rice: a sample of 83 accessions of Oryza rufipogon that represented most of its ecogeographical range; a sample of 348 entries of cultivated O. sativa, including both indica and japonica varieties from almost all ricegrowing areas in China; and 50 accessions of both indica and japonica cultivars from South and East Asia. Sampling emphasis was also given to a well-recognized center of diversity for cultivated rice (Yunnan in s outhwestern China) and to areas identified as being historically important for rice cultivation and evolution. Statistics pertinent to levels of rDNA polymorphisms were summarized (see table). In all, 42 spacer length variants (slvs) were resolved in samples of 481 accessions. These slvs comprised a stepladder, with step sizes varying from 21 to 311 bp. Thirty-six of 42 slvs were observed in O. rufipogon, but only 15 slvs were detected in O. sativa, despite it having more accessions than O. rufipogon. Nine slvs occurred in both O. rufipogon and O. sativa samples, and one of them was observed at a predominantly high frequency in both O. rufipogon (0.310) and O. sativa (0.428). This slv probably represents a widely adapted allele at one of the rDNA loci.

1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Total Av

aEcuador evaluated 915 lines to release a variety. This average results from dividing the total number of lines intro-

duced from 1982 (the year the line was introduced that gave rise to the first variety) to 1990 (the year the line was introduced that resulted in the most recently released variety) by the number of released varieties.

commercial variety, with an average of 3.7 yr elapsing between introduction and release. Bolivia tested 2,117 introductions from 55 nursery sets and released six varieties from 1979 to 1994. The national program evaluated an average of 287 lines per variety released, with an average of 6.3 yr between introduction and release. Bolivia thus tested only one-third of the total lines evaluated by Ecuador to release a variety but required twice as much time to achieve the same result. These results may

be attributed to differences in the national programs breeding strategies, target ecosystems, financial resources, and stability of breeder positions. In conclusion, the two national programs have efficiently used germplasm distributed through INGER-LAC. If efficiency is measured as the time from introduction to release, Ecuador is more efficient than Bolivia. But if the main criterion is the number of lines evaluated to release a commercial variety, then Bolivia is more efficient than Ecuador.

IRRN 21:1 (April 1996)

Summary statistics of rDNA polymorphisms in wild and cultivated rice. Cultivated Sample Spacer length variants (no.) Phenotypes (no.) Diversity Sample size (no.) Wild 36 40 2.99 83 Total 15 48 2.57 398 Indica 14 36 2.33 228 Japonica 13 28 2.38 170 Total 42 80 2.85 481

All six slvs occurring in O. sativa but not in O. rufipogon were rare (f < 0.01). In contrast, all of the slvs but one occurring in both O. rufipogon and O. sativa were observed at much higher frequencies (f>0.02) in O. sativa. A majority of the 27 slvs occurring only in O. rufipogon were infrequent or rare (f < 0.02), but a few of them were detected at sizeable frequencies (0.0250.095). Overall, no drastic difference existed in the composition of slvs between O. rufipogon and O. sativa. About equal numbers of slvs were observed in indica and japonica samples. However, distribution of slvs is clearly differentiated between indica and japonica rice groups: indica rice is preferentially associated with longer slvs and japonica rice with shorter ones. Furthermore, indica rice is dominated by one slv, whereas, in japonica rice, two common slvs were observed at about equal frequencies.

The combination of slvs observed in an accession is called a phenotype. Eighty phenotypes, made up of the 42 slvs, were observed in the total sample; the composition of the phenotypes varied from onebanded (containing only one slv) to sixbanded, with the one-banded type being the most common. In contrast to the slvs, it is surprising that more phenotypes were observed in O. sativa than in O. rufipogon. The phenotype occurring most frequently in the O. rufipogon group was also observed with highest frequency in O. sativa. However, only a few phenotypes were common between O. rufipogon and O. sativa. More phenotypes were detected in indicas than in japonicas. While all of the phenotypes occurring at frequencies more than 0.01 were observed in both indica and japonica groups, these groups differed substantially in phenotypic composition.

In summary, our analysis established that common wild rice is much more diverse in rDNA slvs than is cultivated rice, but cultivated rice has more slv combinations than does wild rice. This indicates that recombination, functioning to assemble various slvs into individual varieties, has played an important role in the evolution of cultivated rice. This observation might have some implications in resolving a recurring issue concerning the relative amount of genetic variation in wild and cultivated species. It is well recognized that the cultivated forms are usually phenotypically more numerous than their wild relatives, which leads many to believe that cultivated species contain more genetic variation than their wild relatives. The reverse had been argued with the availability of data from molecular marker analyses. More slvs (presumably with correspondence to alleles) with fewer phenotypes in the wild species compared with fewer slvs with more phenotypes in the cultivated species, as observed in this study, might explain the contrasting patterns in phenotype and DNA diversity of wild and cultivated species.

A research program for onfarm conservation of rice genetic resources

J.-L. Pham, M. R. Bellon and M. T. Jackson, IRRI

More than 75,000 accessions of cultivated Oryza sativa are stored in the International Rice Genebank at IRRI, thus ensuring their security and continued availability for rice improvement. However, from an evolutional point of view, this ex situ conservation is static: the adaptation of stored varieties to the biotic and abiotic environment is frozen. Thus, in addition to efforts toward genebank efficiency, research is needed on a complementary approach to ex situ conservation. On-farm conservation is dynamic and aims to promote adapting genetic resources by using the evolutionary processes that

created the diversity of rice varieties. Onfarm conservation of crop genetic resources is defined as the continued cultivation and management of a diverse set of crop populations by farmers in the agroecosystems where a crop has evolved. Despite increasing interest in on-farm conservation, knowledge on the meaning of this approach is still limited. Before implementing on-farm conservation, we need improved knowledge of its driving force: farmers management of diversity. Our research program for on-farm conservation started in April 1995 with funding from the Swiss Agency for Development and Cooperation. The objective is to identify opportunities to involve farmers managed systems in the conservation of rice genetic resources. To accomplish this, we will study the management of rice diversity by farmers and its genetic implications.

We have used two theoretical planning frameworks (Figs. 1 and 2). Our research will address the socioeconomic, sociocultural, environmental, and genetic aspects of farmers management of diversity. We will also study farmers management of diversity and its consequences using several combinations of factors that can influence diversity: market interaction (low/high), ethnic origin of farmers (minority/ majority), and environmental heterogeneity (low/high) (Fig. 1). We will pay particular attention to within-variety genetic polymorphism, which is the basis of evolutionary changes. The findings will lead us to assess other sources of genetic diversity (Fig. 2). The focus will be on the rainfed lowland ecosystem where the co-occurrence of modern and traditional varieties is expected. Study locations will be in India, Philippines, and Vietnam.

10

IRRN 21:1 (April 1996)

Bellon MR, J-L Pham, M T Jackson. 1995. Genetic conservation: a role for rice farmers. In: Maxted N. Ford-Lloyd BV, Hawkes JG, editors. Plant conservation: the in situ approach. London: Chapman & Hall. (in press) Pham J-L, Bellon MR, Jackson MT, 1995. What is on-farm conservation research on rice genetic resources? Paper presented at the Third Southeast Asia Symposium on Genetic Resources, 22-24 Aug 1995, SerpongJakarta, Indonesia. (in press).

Cited references

Genetics
Genetics of the new plant type
A. P. Bentota, D. Senadhira, IRRI; M. J. Lawrence, Wolfson Laboratory for Plant Molecular Biology, School of Biological Sciences, University of Birmingham, Birmingham B15 2TT, United Kingdom
1. A conceptual model of factors influencing farmers management of diversity (from Bellon et al 1995).

2. Factors influencing the genetic polymorphism of a land race (L 1) leading to a second land race (L 2 ). Farmers may also decide to create a distinct land race (L 3 ) (from Pham et al 1995).

One of the roles of on-farm conservation research is to test new strategies. These involve farmers managing a sample of genetic diversity in addition to their own varieties. Three ways to complement classical on-farm conservation are reintroducing varieties, introducing alien

varieties, and introducing composite populations. We will test the feasibility of these new strategies in developing the necessary links between ex situ and in situ conservation of rice genetic resources (Bellon et al 1995).

The genetical architecture of 12 quantitative characters of interest to rice breeders has been investigated in two crosses, Jinmibyeo/Gaok (cross 1) and Sangnambatbyeo/Kemandi Pance (cross 2), chosen at random from among those made in IRRIs new plant type program. Individuals of the basic generations (P1, P2, F1, F2, B1, and B2), F3 , and triple test cross families (F 2 P 1, F2 P 2 , and F2 F 1) were raised in a single, completely randomized block, one for each cross. Each block also contained plants of IR72 and IR74 as checks. The parents of these crosses differed significantly for most characters, except for the proportion of filled spikelets and grain yield in cross 1 and 100-grain weight in cross 2. While three characterspanicle number, grain yield, and dry matter displayed heterosis in the first cross, none did in the second. All of the characters in both crosses were heritable. Their heritabilities ranged from 0.65 for panicle length to 0.04 for harvest index in cross 1, and from 0.77 for dry matter to 0.14 for 100grain weight in cross 2. The genetical architecture of the characters in cross 1 was of the relatively simple type in which the genes involved displayed either additive effects only or additive and dominance

IRRN 21:1 (April 1996)

11

effects. The only exception was for those determining the related characters of tiller and panicle number, which also displayed some epistasis of the additive dominance type. On the other hand, the genetical architecture of the characters in cross 2 was more complex. With the exception of days to maturity and tiller number, all were controlled by genes that, in addition to displaying additive and dominance effects, also displayed epistatic effects, predominantly of the duplicate type. Predictions of the proportion of recombinant inbred lines, the means of which meet the desired new plant type targets, can be extracted by single seed

descent from these crosses. Based on information obtained from F3 families, this should be easy to achieve in both crosses for days to heading, days to maturity, tiller number, culm length, and panicle number. It will be less easy to achieve for spikelet number, doubtful to achieve for proportion of filled spikelets and harvest index. and very unlikely for grain yield in either cross. The pattern of genetic correlations between characters, in cross 1 estimated from F3 data, was very different from that in cross 2. Thus, while in cross 1 grain yield was moderately correlated (r = 0.56) only to days to heading, in cross 2 it was highly correlated (r = 0.75) to dry matter, tiller and

panicle number, panicle length, and spikelet number. This indicates that the chief cause of these correlations is the linkage disequilibrium of linked genes, rather than pleiotropy. This, in turn, suggests that it should be possible to extract inbred lines from appropriate crosses achieving the desired multiple new plant type. Results suggest, however, that the targets for proportion of filled spikelets, for harvest index and, in particular, for grain yield will be achieved only after several cycles of inbreeding, each of which is initiated by crossing the best inbred lines from the previous cycle.

Molecular basis of heterosis in hybrid rice and hybrid maize revealed by mRNA amplification
Jinshui Yang, Ninghui Chen, Yanping Gao, Minlian Xu, Koulin Ge, and C. C. Tan, Institute of Genetics, Fudan University, Shanghai 200433, China

We have detected the mRNA subpopulations derived from hybrid rice, hybrid maize, and their parental lines by differential display. Messenger RNA amplification reveals that profound alteration of gene expression occurs in the hybrid genetic background when compared with parental lines. Regulation of gene expression in the hybrid environment includes enhancement, selective transcription, silencing, cosuppression, and activity of specific genes that are silencing in parental lines. Comparing gene expression patterns in hybrids with those in parental lines provides the underlying information to analyze the biological processes controlling heterosis formation. To investigate the relationship between gene expression regulation and heterosis, we tried to use a differential display method to detect gene products from hybrid rice, hybrid maize, and their parental lines. Rice F 1 hybrids were from a cross between

two indica lines, Zhangsha 97 and Minfei 63, and maize F1 hybrids were from a cross between two inbred cultivars, YD4 and H24. The total mRNAs of the leaves of seedlings of hybrid rice, hybrid maize, and their parental lines were isolated using the method of Liang (1993). Two anchored oligo-dT primers, 3-T11CA and 3-T11GC. were used to reverse-transcribe the mRNA subpopulation. Two short arbitrary primers, 5-L1(5'-GATCGCATTG-3') and 5'-L2(5'TACAACGAGG-3'), were designed to make paired primer sets and to amplify cDNA subpopulations. Amplified labeling cDNAs were separated on sequencing gel. In side-by-side comparison, significant changes of the mRNA constitution and quantity appeared in hybrid rice and hybrid maize. Six kinds of gene expression alteration in the hybrids were detected based on amplified cDNA patterns: some new mRNAs that are absent in paternal lines appear in hybrid seedlings, meaning these genes are transcribed only in hybrids: maternal genes become inactive in hybrids; paternal genes are silencing in hybrids: allelic genes of two parental lines show cosuppression in hybrids; some genes are overexpressed in hybrids when compared with their expression in parental lines; and some genes are underexpressed in hybrids. We also found many genes expressed

normally in hybrids. We therefore conclude that in the hybrid genetic background, gene expression is altered in a variety of ways. It was reported recently that the major genetic basis of heterosis in rice is dominance as revealed by QTL analysis using molecularmarkers. In contrast, it seems that the prominent factors contributing heterosis in maize is overdominance (Stuber et al 1992, Xiao et al 1995). Further studies of gene expression alternation in hybrids will provide important information about the molecular basis of overdominance and dominance in heterosis.

Cited references

Allard RW. 1960. Inbreeding depression and heterosis. In: Principles of plant breeding. New York: John Wiley & Sons. p 213-223. Liag P, Averboukh L, Pardee AB. 1993. Distribution and cloning of eukaryotic mRNAs by means of differential display: refinements and optimization. Nucleic Acids Res. 21:3269-3275. Stuber et al. 1992. Identification of genetic factors contributing to heterosis in hybrid from two elite maize inbred lines using molecular markers. Genetics 132:823-839. Xiao J et al. 1995. Dominance is the major genetic basis of heterosis in rice as revealed by QTL analysis using molecular markers. Genetics 140:745-754.

12

IRRN 21:1 (April 1996)

Linkage analysis of a photoperiod-sensitive gene Se1 locus with neighboring loci in rice
I. Ohshima, Institute of Agriculture and Forestry (IAF), University of Tsukuba, lbaraki 305, Japan; F. Kikuchi, Agronomy Department, Tokyo University of Agriculture, Tokyo 156, Japan; K. Odaka, K. Nomura, and K. Tamura, IAF, University of Tsukuba; and H. Kato and R. Ikeda, National Agricultural Research Center, Tsukuba, lbaraki 305, Japan

The Se1 locus, which is located on chromosome 6, is a major locus controlling photoperiod sensitivity in rice. It has multiple alleles that adapt to different daylengths ranging from southern to northern latitudes. The Se1 locus is linked with the phosphoglucose isomerase isozyme (Pgi2) and blast resistance gene ( Pi-z t ) (Yokoo and Fujimaki 1971, Oosumi et al 1989). This study was undertaken to find restriction fragment length polymorphism (RFLP) markers linked closely to the Se1 gene and to construct the linkage map of Se1 with Pi-z t, Pgi2, and RFLP markers. Four near-isogenic lines (NlLs) were used as the crossing parents and developed as follows. A Japanese early variety, Fujisaka 5, was crossed with an indica variety Morak Sepilai from Malaysia. Fujisaka 5 was then backcrossed four times to the hybrids. The BC 4F1 was selfed up to the BC 4F13 generation by selecting the heterozygotes for heading time and blast resistance. Both an early-heading plant (Se-1 e /Se-1e) and late-heading plant (Se-1 u / Se-1 u) that segregate for blast resistance were selected from the BC 4F14 population. In the following generation, blast-resistant lines ( Pi-z t/Pi-z t ) and blast-susceptible lines (+/+) were obtained from each of early- and late-heading lines (Yokoo 1983) and designated as ER, ES, LR, and LS, respectively. In addition, ES, ER, and LS lines derived the Pgi2 allele 1 from Fujisaka 5 while LR camed allele 2 from Morak Sepilai. The F2 plants of LR/ES and ER/LS were sown in the field on 29 Apr 1992 at the University of Tsukuba (36 N). The heading time of each plant was recorded as

the date when the first panicle emerged. The genotypes of F 2 plants. in which the heading time was not clear. were estimated by investigating segregation of the heading time of F 3 progenies in 1993. The genotypes of blast resistance in two populations were determined from the segregation of F 3 progenies in the blast nursery field. About 30 plants of each line at the fifth or sixth leaf stage were inoculated with a spore suspension of blast fungus strain 007. Blast resistance was scored 3 wk after inoculation. The DNAs of parents and NILS were extracted from leaves by the CTAB method and were hybridized with rice genomic clones (RFLP markers) previously mapped to chromosome 6 (Saito et al 1991). Two of the 22 RFLP markers showed polymorphism among NILS. The DNA of 206 F 2 plants were extracted from leaves after heading and were digested with a restriction enzyme ( Eco RV). Southern blot analysis was carried out to determine the RFLP genotypes of F 2 plants by using these two clones (XNpb 294, XNpb 233). Recombination values were calculated by MAPL program (Ukai et al 1990) based on maximum likelihood. The segregation of heading time in two F2 populations, LR/ES and ER/LS showed trimodal distributions, respectively, and fit well to a 1:2:1 (ear1y:medium:late) ratio, indicating clear monogenic segregation of Se1 alleles. The phenotypic segregation of Pi-z t gene showed a good fit to the expected ratio (1:2:1) in two F 2 populations. The recombination values between Se1 and Pi-z t were 1.2 0.4 in LR/ES and 1.2 0.6 in ER/LS. We analyzed Pgi2 in the F2 population of LR/ES. The segregation for the genotype of Pgi2 fit well to the expected ratio (1:2:1) and showed the close linkage of Pgi2 and Se1. The recombination value was calculated as 2.8 0.6. Two RFLP markers (XNpb 294 and XNpb 233) revealed that only the LS line showed the identical band with the recurrent parent Fujisaka 5. The band of the other three lines was similar to that of the donor parent Morak Sepilai. The RFLP analysis was carried out for the F 2 population of ERLS by using XNpb 294 and XNpb 233 as probes. The segregation of

The linkage map of Pgi2, Se1, Pi-z t, and two RFLP markers on chromosome 6 in rice. All distances are given in centiMorgan units.

these two markers fit well to Mendelian monogenic inheritance. These markers were found to be tightly linked to Se1 and Pi-z t genes. The recombination value between XNpb 294 and Se1 was 3.7 1.0 and that between XNpb 294 and Pi-z t was 2.5 0.8. The recombination between XNpb 294 and XNpb 233 was not detected. The linkage map was arranged from recombination values (see figure). Based on these results, Pgi2, XNpb 294, and XNpb 233 may be useful as molecular markers of the Se1 gene for genetic and breeding studies. However, molecular markers more tightly linked must be identified to clone the Se1 gene using mapbased cloning procedures.

Cited references
Saito A, Yano N, Kishimoto N, Nakagahara M, Yoshimura A, Saito K, Kuhara S, Ukai Y, Kawase M, Nagamine T, Yoshimura S, Ideta O, Ohsawa R, Hayano Y, Iwata N, Suguira M. 1991. Linkage maps of restriction fragment length polymorphism loci in rice. Jpn. J. Breed. 41:665-670. Oosumi T, Miyazki A, Uchimiya H, Kikuch F, Yokoo M. 1989. Analysis of glucose phosphate isomerase in near-isogenic lines and cultivars of rice ( Oryza sativa L.) Bot. Mag. Tokyo 102:283-289. Ukai Y, Ohsawa R, Saito A, Hayashi T. 1995. MAPL: a package of computer programs for construction of DNA polymorphism linkage maps and analysis of QTL. Breed. Sci. 45:139-142. Yokoo M. 1983. Near-isogenic lines of rice with respect to a Pi-z t gene for resistance to blast disease. Jpn. J. Breed. 33:341-345. Yokoo M, Fujimaki H. I971. Tight linkage of blast resistance with late maturity observed in different indica varieties of rice. Jpn. J. Breed. 27:35-39.

IRRN 21:1 (April 1996)

13

Genetic analysis of plant regeneration ability from cell suspension cultures of rice
M. Tsukahara, T. Hirosawa, E. Nagai, Nursery Technology Inc., Kitsuregawa, Tochigi 329-14, Japan; H. Kato and R. Ikeda, National Agriculture Research Center, Tsukuba 305, Japan; and K. Maruyama, Agriculture, Forestry, and Fisheries Research Council, Secretariat, Tokyo 100, Japan

Genetic analysis of plant regeneration ability in rice was studied in 14 14 diallel crosses. The objectives of our study were to elucidate relationships between ecospecies and genetic properties of regeneration ability, to identify the varieties that improve the regeneration ability in F1, and to evaluate the possibility of obtaining F1s showing significant heterosis for both regeneration ability and biomass. We used 14 rice varieties with diverse geographical origin (4 japonica, 4 javanica, and 6 indica) and their reciprocal 164 F 1 hybrids (seeds from 18 F1s were not obtained). Callus induction and initiation of suspension culture were the same as that described by Tsukahara and Hirosawa (1992). For plant regeneration, after 7 d of suspension culture in a fresh subculture medium, 20 mg of cell clusters (200-1000 m in diameter) were transferred to 300-ml Erlenmeyer flasks containing 20 ml of liquid regeneration medium. Regeneration medium contained inorganic salts and vitamins of N6 medium

(Chu et al 1975) supplemented with 12 mM proline, 0.1 g casein acid hydrolysate L -1 , 0.1 mg NAAL-1, 0.1 mg kinetin L-l , and 5mM MES (pH 5.8). Three replications were prepared for each genotype. After 6 wk of culture, regenerated plantlets with green shoots of more than 1 mm were counted under a dissecting microscope. Data were transformed to stabilize variance using y= (x+1). A computer program for diallel analysis provided by Ukai (1989) was used for the statistical analysis. Among the 14 parent varieties, the number of regenerated plantlets varied from 108.3 (Kele) to 0 (Lemont, CP231, Nan Jing 11, and Qing Er Ai) with a mean of 14.1 (see table). The number of regenerated plantlets for the F 1 was widely distributed from 164 (Kele Banten) to 0. The mean was 36.9. Raffaelo, Kele, and Qing Er Ai were parents in the crosses of 18 F1s that regenerated more than 80 plantlets per flask. No regenerated plantlets were observed in 8 F1s, either of whose parents especially the femaledid not regenerate. In contrast, even though both parents did not regenerate, five of their seven F1s could regenerate. This suggests the participation of two or more genes in regeneration ability with complementary gene action. The graph of array variance (Vr) and array parent (offspring covariance [Wr] for regeneration in 10 10 diallel table in which Lemont, IR26, Nan Jing 11, and Qing Er Ai were excluded from 14 14 diallel table because some of the F 1 data were not available) suggests that no non-

allelic interaction exists and that the low regeneration ability was dominant. Analysis of variance of 10 10 diallel table showed that both additive and dominant effects were significant, as were reciprocal effects. No significant relationship exists between ecospecies and the above genetic properties. Among the 14 varieties, indicas Qing Er Ai and Kele would be useful for improving regeneration ability because of their significant high general combining ability and favorable maternal effects. Heterosis for regeneration ability in the crosses of japonica/indica, javanica/indica, and their respective reciprocal crosses was higher than that in the crosses of japonica/ japonica, japonica/javanica, and javanica/ japonica. Kato et al (1994) also had reported higher heterosis for shoot dry weight in crosses of japonica/indica than in japonica/ japonica and indica/indica using a half-diallel set from the same parents as in this study. There was no significant correlation (r = 0.199) between regeneration ability and shoot dry weight of each F1 hybrid, indicating that it was difficult to obtain F 1 hybrids with high heterosis in both characters.

Chu CC, Wang CC, Sun CS, Hsu C, Yin KC, Chu CY, Bi FY. 1975. Establishment of an efficient medium for anther culture of rice through comparative experiments on the nitrogen sources. Sci. Sin. 18:659-668. Kato H, Tanaka K, Nakazumi H, Araki H, Yoshida T, Ogi Y, Yanagihara S, Kishimoto N, Maruyama K. 1994. Heterosis of biomass among rice ecospecies and isozyme

Cited references

Regenerated plantlets per flask in a suspension culture of 14 rice varieties and their F1 hybrids. Male parent Female parent (origin) 1 Akihikari (Japan) Nipponbare (Japan) Raffaelo (Italy) Bomba 1 (Spain) Ketan Nangka (Indonesia) Banten (Indonesia) Lemont (USA) CP231 (Philippines) IR26 (IRRI) Jamuna (India) Nan Jing 11 (China) Qing Er Ai (India) Dular (India) Kele (India)
a = data not available.

Japonica 2 34.0 20.3 79.7 1.0 37.0 63.0 4.7 0.0 63.7 1.7 72.7 130.3 69.3 113.0 3 24.3 52.3 21.7 8.3 12.0 112.7 41.7 13.3 31.0 10.0 69.0 45.3 74.3 93.3 4 24.3 4.7 13.3 0.7 8.7 12.0 0.0 0.0 4.7 17.3 14.0 16.7 68.7 45.0 5 42.3 5.7 28.3 1.0 3.0 27.3 a 10.3 29.7 44.0 76.0 134.0

Javanlca 6 79.0 1.0 28.7 16.3 4.7 9.7 27.0 42.7 58.0 164.3 7 29.7 7.3 7.3 0.7 21.0 18.7 0.0 0.7 12.0 75.0 65.0 8 0.7 5.3 0.3 0.0 0.3 18.0 0.0 0.0 44.3 15.0 59.0 9 23.3 44.3 50.0 18.7 31.7 54.3 19.7 39.0 4.0 16.0 29.0 60.3 10 24.7 22.3 20.3 11.7 12.7 15.3 16.7 1.7 11.0 5.3 32.0 78.3 11

Indica 12 55.3 125.7 145.0 92.7 80.7 145.0 114.0 114.0 39.3 22.3 0.0 0.0 50.0 149.0 13 3.3 12.0 21.0 15.7 13.3 26.7 18.3 1.0 8.0 10.3 0.0 5.0 7.7 12.0 14 31.0 39.0 132.7 5.0 30.3 40.3 11.0 5.0 76.7 35.7 38.0 0.0 14.3 108.3

17.0 7.3 34.0 1.7 4.0 71.7 18.3 2.7 6.7 7.3 26.3 141.3 67.7 100.3

20.3 43.0 101.0 63.0 19.3 55.3 16.7 70.0 25.3 3.0 0.0 40.0 49.0

14

IRRN 21:1 (April 1996)

polymorphism and RFLP. Breed. Sci. 44:1271-277. Tsukahara M, Hirosawa T. 1992. Simple dehydration treatment promotes the plantlet regeneration from rice (Oryza sativa L.) callus. Plant Cell Rep. 11:550-553. Ukai Y. 1989. A microcomputer program DIALL for diallel analysis of quantitative characters. Jpn. J. Breed. 39: 107-109.

Genetic correlation and realized heritability for traits related to yield sink capacity of rice.a Trait b NP NG NPB NGPB NSB NGSB GL GW RGP
a

YSC 0.449 0.748 0.041 0.435 0.636 0.720 0.246 0.267 0.964

NP 6.690 0.467 0.347 0.286 0.294 0.030 0.559 0.046

NG 0.200 0.342 0.687 0.642 0.619 0.313 0.147 0.384

NPB 0.959 0.426 0.504 0.495 0.454 0.319 0.268 0.899

NGPB *0.281 *0.507 0.040 0.168 0.231 0.181 0.178 0.180

NSB *0.544 0.108 0.787 0.179 0.917 0.050 0.047 1.008

NGSB *0.130 0.122 0.187 *0.368 0.129 0.113 0.022 1.184

GL 0.256 0.155 0.136 0.103 0.124 0.270 0.052 0.393

GW 0.472 0.515 0.470 *0.319 0.074 *0.153 0.773 0.028

Classification of Korean rice germplasm based on isozyme polymorphism

K. S. Lee, D. Senadhira, and D. S. Brar, IRRI

Above the diagonal (underlined): genetic correlation coefficients estimated from selection responses in F 2 and F 3 populations of Nakateshinsenbon/Hieri. * = Sign of coefficient could not be determined. The data for YSC, NP, and RGP were not available. Below the diagonal: genetic correlation coefficients estimated from variance and covariance components in recombinant inbreds of Koshihikari/Nakateshinsenbon. Diagonal (underlined): realized heritabilities estimated from selection responses in F 2 and F 3 populations of Nakateshinsenbon/Hieri. b See text for the meanings of the abbreviations.

Isozymes are important biochemical markers in genetic and plant breeding research. Isozyme polymorphism has provided useful information for classifying rice cultivars, thus enabling breeders to transfer desirable traits through hybridization within and between groups. We analyzed traditional and modern rice cultivars from Korea using gel electrophoresis. Five isozyme loci (Pgi1, Pgi2, Amp1, Amp2, Amp3) were studied following the classification scheme of Glaszmann (1987). Zymograms were revealed following the staining procedure of Glaszmann et al (1988). Of the 286 cultivars, 42 (14.7% belonged to group 1 (indica types). This is the result of hybridization between indica and japonica types cultivated during 1976-80 as Tongil type in Korea. Most (242, or 84.6%) were classified as group VI (japonica types), which are currently heavily cultivated in Korea. Only two varieties were mixed.

Heritability and genetic correlation for component characters of yield sink capacity in rice
T. Kato, School of Bioresources, Hiroshima Prefectural University, Shobara, Hiroshima 727, Japan

Yield sink capacity (YSC), the maximum size of sink organs to be harvested per rice plant is a product of the number of grains plant -1 ( = number of panicles plant-1 [NP]

number of grains panicle -1 [NG]) multiplied by single grain weight. Grain ripening ability, measured as the ripened grain percentage (RGP), is generally superior for grains on primary branches compared with that for secondary branches. Therefore NG must be broken into components: number of primary branches panicle-1 (NPB), number of grains on primary branches primary branch -1 (NGPB), number of secondary branches primary branch-1 (NSB), and number of grains on secondary branches secondary branch-1 (NGSB). We estimated genetic correlation (rg) for these YSC components using two experiments. In experiment I, genetic correlation was estimated from variance and covariance components in 49 recombinant inbred lines of Koshihikari/ Nakateshinsenbon cultivated under four environments. In experiment II, rg was estimated from direct and correlated responses in selecting F2 and F 3 populations (Falconer 1989) of Nakateshinsenbon/ Hieri. Realized heritability was also estimated. In experiment I, YSC was positively correlated with NG (r = 0.748), but not with grain length (GL) or grain width (GW). A high negative correlation was detected between YSC and RGP (r =-0.964), perhaps due to the negative correlations between RGP and NSB (r = -1.008) and RGP and NGSB (r = -1.184). In contrast, NPB was positively and tightly correlated with RGP (r = 0.899). In experiment II, NPB showed a high positive correlation with NG (r = 0.959) and the highest realized heritability among the NG components.

In addition, NPB was not positively correlated with NSB and NGSB in either experiment. These results suggest a strategy for developing a promising panicle type that realizes both high NG and RGP that consists of increasing the NPB to provide numerous well-ripened grains and suppressing the NSB and/ or NGSB to not overproduce unfilled grains. Chau and Bhargava (1993) demonstrated that the large sink sizes of highyielding indica cultivars were mainly due to high NPB and high NPB NGPB. The genetic nature of NPB must be understood more precisely.

Chau NM, Bhargava SC. 1993. Physiological basis of higher productivity in rice. Indian J. Plant Physiol. 36:215-219. Falconer DS. 1989. Introduction to quantitative genetics. 3d ed. Harlow (UK): Longman Scientific & Technical. 438 p.

Cited references

Heading-time genes control photoperiod insensitivity of rice cultivar Norin 20

K. Ichitani, Y. Okumoto, and T. Tanisaka, Laboratory of Plant Breeding, Faculty of Agriculture, Kyoto University, Kyoto 606-01, Japan

Rice cultivars grown in Hokkaido, the highest latitude area (41-44N) in Japan, flower far earlier than any other Japanese cultivars. Previous work on geographical analysis for heading traits revealed that the

IRRN 21:1 (April 1996)

15

extreme earliness of Hokkaido cultivars depends on extremely weak photoperiod sensitivity. We analyzed the genetic factors controlling the weak photoperiod sensitivity of Hokkaido cultivars by comparing the gene constitution and degree of photoperiod sensitivity of photoperiod-insensitive Norin 20 with those of several other cultivars. A scatter diagram was constructed for 26 Hokkaido varieties according to basic vegetative growth (BVG) and photoperiod sensitivity (PS) (see figure). BVG was expressed as days to heading under short day length (10 h) condition, and PS was expressed as the difference between days to heading under whole daylength minus days to heading under short daylength. The cultivars could easily be classified into three groups based on PS: group I (PS < 3.0), group II (4.0 < PS < 10.0), and group III (PS >20.0). The PS of Norin 20 was 0.6, confirming that this variety has completely lost photoperiod sensitivity (Vergara and Chang 1985). We compared the gene constitution between Norin 20 (group I) and varieties randomly chosen from groups II and III. Seven varieties were crossed with 10 tester lines to study four photoperiod sensitivity loci: E1, E2, E3, and Se1. The dominant alleles of E1 and Se1 loci (E1 and Sel n (Se1u)) remarkably improved the photoperiod sensitivity of rice (Okumoto et al 1991), but the effects of those of E2 and E3 loci are small (Yamagata et al 1986). (See table for the results of the gene analysis.) For the E1 locus, all of the varieties examined carry e1, an allele for photoperiod
Estimated genotypes for heading-time loci E 1, E2, E 3, and Sel and degree of photoperiod sensitivitya of seven Hokkaido varieties. Hokkaido variety Norin 20 Kitahikari Yukara Kirara397 Shiokari Eiko Sasahonami Genotypeb E1 E2 E3 e3e3 e3e3 e3 e3 e3 e3 E3E3 e3e3 E3E3 Se1 Photoperiod sensitivity

International Rice Research Institute. p 778780. Yamagata H, Okumoto Y, Tanisaka T. 1986. Analysis of genes controlling heading time in Japanese rice. In: Rice genetics. Manila (Philippines): International Rice Research Institute. p351-359.

QTL analysis of rice seedling vigor in japonica and indica genetic backgrounds
E. D. Redoa, Philippine Rice Research Institute, Muoz, Nueva Ecija 3119, Philippines; and D. J. Mackill, United States Department of Agriculture-Agricultural Research Service, Agronomy and Range Science Department, University of California, Davis, CA 95616, USA

Scatter diagram of 26 Hokkaido varieties according to basic vegetative growth and photoperiod sensitivity.
a

(Days to heading under whole-day length) - (days to heading under short-day length [10 h]. bDays to heading under short-day length (10 h).

insensitivity. No alleles for photoperiod sensitivity were found among the varieties for the other loci. Norin 20 carries alleles for photoperiod insensitivity alleles at all the above loci, and its genotype was estimated tobe e1e1 e2e2 e3e3 Se1eSe1e. This suggests that the photoperiod insensitivity of Norin 20 can be attributed to the combining effect of these four alleles. However, the cultivars of the same genotype for four loci existed even in Groups II (Kitahikari, PS = 6.8) and III (Yukara, PS = 21.0). We conclude that e 1 is an essential gene for cultivating rice in Hokkaido. The combining effect of e1 with an allele(s) for photoperiod insensitivitybut not at E2, E3, and Se1 locicontrols the weak photoperiod sensitivity of Hokkaido varieties.

e1 e1 e2e2 e1e1 e2e2 e1e1 e2e2 e1e1 E2E2 e1e1 e2e2 e1e1 e2e2 e1e1 E2E2

Se1eSe1e 0.6 Se1eSe1e 6.8 Se1eSe1e 21.0 Se1nSe1n 8.0 Se1eSe1e 5.3 Se1eSe1e 23.3 Se1nSe1n 5.3

Cited references
Vergara BS, Chang TT. 1985. The flowering response of the rice plant to photoperiod. Manila (Philippines): International Rice Research Institute. Okumoto Y, Tanisaka T, Yamagata H. 1991. Genotypic difference in response to light interruption in Japanese rice varieties. In: Rice genetics II. Manila (Philippines):

alleles.

a See Fig. 1. b E , E , E , and Se1n= photoperiod-sensitive 1 2 3 aIleles. e 1, e 2, e 3, and Se1 e = photoperiod-Insensitive

Breeding for enhanced seedling vigora plant's ability to emerge rapidly from soil or water (Heydeker 1960)continues to be a major challenge for rice breeders in both tropical (Herdt 1991) and temperate (McKenzie et al 1994) rice-growing countries. Vigorous varieties are needed for optimum stand establishment under direct seeded rice cultures. The development of molecular marker maps in rice (McCouch et al 1988) makes it possible to identify and locate genetic factors or quantitative trait loci (QTLs) controlling polygenic characters. Rice seedling vigor is associated with shoot and root lengths under water-seeded rice culture (Jones and Peterson 1976) and coleoptile and mesocotyl lengths under drill seeding (Turner et al 1982). To identify and locate QTLs underlying these characters, two high-vigor cultivars, indica Black Gora (BG) and temperate japonica Italica Livorno (IL), were crossed to the same maternal parent, Labelle (LBL), a lowvigor tropical japonica. Two molecular maps were constructed based on 204 F 2 plants and 117 restriction fragment length polymorphisms (RFLPs) in LBL/BG and on 118 F2 plants and 129 random amplified polymorphic DNAs (RAPDs) and 18 RFLPS in LBL/IL. The LBL/BG map spanned 1496 Haldane cM, with an average marker spacing of 14 cM. Linkage analysis was performed using Mapmaker 3.0 (Lander et al 1987). Total length of the LL/lL ,map was

16

IRRN 21:1 (April 1996)

1168 cM with an average interval length of 9 cM. Lengths of shoot, root, coleoptile, and mesocotyl were scored on 172 F 3 families in LBL/BG using replicated slantboard tests at two temperatures (18 and 25 C) and on 113 F3 families in LBL/IL at 18 C. The slantboard test (Jones and Peterson 1976) is a standard laboratory procedure used for screening seedling vigor in California. Interval mapping (LOD = 2.5) using Mapmaker and single-point analyses (P < 0.05) using SAS programs (SAS Institute, 1989) identified 13 QTLs in LBL/BG, including four for shoot length, two each for root and coleoptile lengths, and five for mesocotyl length (see table). Three of these QTLs were expressed at both screening temperatures. In LBL/IL, 12 QTLs were identified including two for coleoptile length and five each for root and mesocotyl lengths (see table). Most QTLs, including those detected from the common maternal parent, mapped to different rice chromosomes in the two crosses. These results suggest that both genetic background and screening environment may influence QTL expression for seedling vigor. Different sets

of alleles conferring increased seedling vigor may be present in high-vigor indica and japonica genetic donors. The indica BG contributed QTLs with only minor effects for shoot length, the most important determinant of seedling vigor in Californias water-seeded rice culture. However, an allele from the temperate japonica cultivar IL accounting for 18.5% of the shoot length variation was identified by single-point analysis. Positive alleles at QTLs identified in the two populations may be useful for seedling vigor enhancement of both japonica and indica cultivars.

experimental and natural populations. Genomic 1:174-181. McCouch SR, Kochert G, Yu ZH, Wang ZY, Khush GS, Coffman WR, Tansley SD. 1988. Molecular mapping of rice chromosomes. Theor. Appl. Genet. 76:815-829. McKenzie KS, Johnson CW, Tseng ST, Oster JJ, Brandon DM. 1994. Breeding improved rice cultivars for temperate regionsa case study. Aust. J. Exp. Agric. 34:897-905. SAS Institute Inc. 1989. SAS/STAT users guide, Version 6, Fourth Edition, Volume 1. Cay (NC) SAS Institute Inc. Turner FT, Chen CC, Bollich CN. 1982. Coleoptile and mesocotyl length in semidwarf rice seedlings. Crop Sci. 22:42-46.

Cited references
Herdt RW. 1991. Research priorities for biotechnology. In: Khush GS and Toenniessen GH, editors. Rice biotechnology. Oxon (UK): CAB International. Heydecker W. 1960. Can we measure seedling vigor? Proc. Int. Seed Test. Assoc. 25:498512. Jones DB, Peterson ML. 1976. Rice seedling vigor at suboptimal temperatures. Crop Sci. 16:102-105. Lander ES, Green P, Abraham J, Barlow A, Daly M, Lincoln SE, Newburg L. 1987. Mapmaker: an interactive computer package for contructing primary genetic linkage maps of

RFLP mapping of QTLs for yield and other related characters in rice
H.-X. Lin, H.-R. Qian, J.-Y. Zhuang, J. Lu, S.-K. Min, and Z.-M Xiong, China National Rice Research Institute (CNRRI), Hangzhou 310006, China; N. Huang, IRRI; and K-L Zheng,CNRRI

Chromosomal locations of QTLs for length of shoot, root, coleoptile, and mesocotyl measured in slantboard tests in the Labelle/Black Gora and Labelle/ltalica Livorno populations as determined by interval and single-factor analysis. Chromosome no. Vigor trait Temp (C) Labelle/Black Gora % variance explained 7a 31 a 13 9 16 a 25 10 8a 13 15 a 28 10 a Nearest marker RG536 RG222 RG811 CDO718 RZ452 RZ448 RZ448 RG435 RZ67 RG716 RZ395 RZ422 Labelle/ltalica Livorno % variance explained 18 b 11 50 14 16 a 13 11 a 14 21 13 a 11 a 9a Nearest marker OPAD13 720 OPAJ201140 OPH191340 OPX171700 OPY6 1150 RG678 OPAA1640 OPX2700 OPZ122400 OPAN151050 RZ536 RG341

1 2 3 5

Shoot Root Mesocotyl Root Shoot Coleoptile Mesocotyl Shoot Root Mesocotyl Coleoptile Mesocotyl Root Mesocotyl Shoot Root Root Coleoptile Coleoptile Mesocotyl Root

6 7 9 10 11 12

18, 18, 25 18 18, 25 18, 18 18 25 18 25 18 18 18 25 18 18 18 18 18 18 18

25 25 25 25

aLabelle has the positive allele. bBy single-factor analysis.

Two F 2 populations were established from two indica/indica combinations: Tesanai 2/ CB (TSA/CB) and Waiyin 2/CB (WY/CB). CB, from California, USA, was the common male parent. Tesanai 2, from Guangdong, China, is a high-yielding variety noted for its large panicles. Waiyin 2, from IRRI, has large grains. Seventythree percent of 167 RFLP markers tested revealed polymorphisms between the parents of the two crosses. For each F 2 population, genotypes were determined for 171 plants with 93 RFLP markers in TSA/ CB and 101 in WY/CB, respectively. Two linkage maps, made up of 89 marker loci and 93 marker loci of 12 linkage groups, were constructed, respectively, using Mapmaker (Lander et al 1987). Each of eight characters were subjected to QTL mapping with both ANOVA (SAS Institute Inc. 1988) and interval mapping procedure (Lander and Botstein 1989). In TSA/CB, 22 QTLs for the eight characters were detected by one-way ANOVA (data not shown). The same QTLs were also detected by interval analysis (Table 1). In WY/CB, 19 QTLs for six traits were detected by one-way ANOVA, and, again,

IRRN 21:1 (April 1996)

17

Table 1. QTLs detected for yield components based on interval analysis (Mapmaker/QTL) in the Tesanai 2/CB F2 population. Traita GWT QTLb gwt1 twt2 gwt4 gwt5 gwt8 np2 np4 ng1 ng2 ng8 ng12 ns3 ns8 ns12 sf1 tgwt1 tgwt4 tgwt5 sd3 sd8 sd12 nfb8 Interval RG374-RG394 RG157-RG171 RG143-RG214 RG13-RG573 RZ66-RG598 RG157-RG171 RG143-RG214 RG374-RG394 RG25-RG157 RG978-RZ66 RG341-RG235 RG104-RG348 RZ562-RG978 RG457-RG341 RG374-RG394 RG173-RG532 RG143-RG214 RG182-RG13 RG104-RG348 RZ562-RG978 RG457-RG341 RG108-RZ562 LODc 2.93 4.11 3.10 2.39 2.31 3.62 9.68 4.41 2.79 2.28 2.48 2.32 4.84 2.34 3.86 3.01 2.74 2.73 2.13 3.73 2.21 6.49 % variationd explained 10.7 11.4 8.7 11.0 7.8 9.3 26.1 17.2 9.0 13.8 7.4 6.8 15.5 7.4 12.2 12.6 8.5 14.8 5.7 10.9 8.0 20.9 ae 1.96 9.14 7.91 1.27 8.04 1.39 2.92 1.39 21.46 30.55 1.77 16.56 35.70 37.89 5.10 1.63 1.39 1.55 5.43 11.79 14.06 1.91 df 12.36 7.10 0.28 12.72 0.51 2.27 0.18 48.18 27.65 9.49 31.53 45.87 53.64 9.61 25.00 1.50 0.55 1.33 15.04 14.69 4.04 0.49 d/[a] g 6.32 0.78 0.04 10.03 0.06 1.64 0.06 34.57 1.29 0.31 17.79 2.77 1.50 0.25 4.92 0.92 0.39 0.86 2.77 1.25 0.29 0.25

NP NG

NS

SF TGWT

SD

NFB

aGWT = grain weight plant -1, NP = number of panicles plant -1, NG = number of grams panicle -1, NS = number of spikelets panicle-1, SF = spikelet fertility, TGWT = 1000-grain weight, SD = spikelet density, NFB = number of first branches panicle-1. b QTLs are named by trait abbreviations plus chromosomal number. c Log10-likelihood. dPercent phenotyplc variance explained. eAdditive gene effect at the putative QTL loci. f Dominance effect at the putatlve QTL loci. g Degree of dominance.

Table 2. QTLs detected for yield components based on interval analysis (Mapmaker/QTL) in the Waiyin 2/CB F2 population. Traita GWT QTLb Interval RG374-RG394 RG788-RG190 CD082-RG360 RG324A-RG324B RG360-RG9 waxy-RG213 RG138-RG64 RG108-RZ562 RG1094-RG118 RG536-RG222 waxy-RG213 RZ562-RZ617 RZ649-RG381 RG171-RG437 RG241-RG561 RG520-RG256 RG138-RG64 RG108-RZ562 RG103-RG2 LODc 3.13 4.67 3.14 2.09 2.86 2.86 5.01 2.27 4.38 3.25 2.37 2.25 2.12 4.23 3.52 3.13 5.15 2.45 2.31 % variationd explained 11.5 22.3 10.1 5.6 9.9 12.1 13.1 7.3 12.5 9.4 12.5 6.7 5.8 14.8 22.8 12.3 13.3 7.9 6.3 ae 6.70 8.85 8.59 1.06 1.99 1.51 38.04 10.79 54.71 8.27 11.42 8.72 1.23 2.42 2.89 3.92 14.62 3.86 5.22 df 6.56 11.48 1.66 1.13 0.24 1.16 28.14 43.50 43.41 6.80 7.84 0.51 1.10 0.57 1.32 21.09 6.30 16.22 13.48 d/[a] g 0.98 1.30 0.19 1.06 0.12 0.76 0.74 4.03 0.79 0.82 0.69 0.06 0.90 0.24 0.46 5.38 0.43 4.20 2.58

gwt1 gwt4a gwt5a np2a np5 np6 tns6 tns8a tns11 sf1a sf6 sf8 tgwt1a tgwt2 tgwt10 sd2 sd6 sd8 sd11

NP

TNS

SF

TGWT

SD

a-gSee Table 1 for definitions.

all 19 QTLs were detected by interval analysis (Table 2). These findings indicate that both statistical procedures produce very similar results. The QTLs identified are named with trait abbreviation followed by chromosomal number (Tables 1 and 2). QTLs controlling grain weight plant-1 (GWT) were located within five intervals (gwt1, gwt2, gwt4, gwt5, and gwt8) in TSA/CB. The five QTLs for GWT are located on chromosomes 1, 2, 4, 5 and 8, with one in each chromosome (Table 1). In WY/CB, however, only 3 intervals containing QTLs (gwt1, gwt4a, and gwt5a) controlling GWT were detected. Three QTLs for GWT are located on each of chromosomes 1.4, and 5 (Table 2). Among the 41 QTLs for various traits detected in both populations. 23 QTLs can explain phenotypic variance larger than 10%. Some of the QTLs, such as np4, gwt4a, tgwt10, and nfb8, explained more than 20% of variance (Tables 1 and 2). If a QTL can explain larger variation and is significantly different from other QTLs (e.g., much larger LOD score), it can be assumed that this would be a major gene not a QTL. QTL np4 (Table 1) from TSA seems to have such characteristics. Number of panicles (NP) is generally considered to be under the control of both major and minor genes. QTL np4 might be a major gene that joins other QTLs with lesser effects, such as np2, to control panicle development. In general, 1 to 3 QTLs were detected for each trait in both populations (Tables 1 and 2). Comparison of QTLs for individual traits showed that a majority of QTLs are present in only one of the two populations. For example, three QTLs were detected for total number of spikelets panicle-1 (NS) but only one (tns8) was detected in both populations. Performance of female parents is very different, and a different set of operating genes is likely to control it. Altogether, we found three QTLs (gwt1, tns8, sd8) shared by both populations. Different QTLs detected in different populations provide an opportunity to combine them to produce new breeding lines with higher yield potential. As DNA markers are not subject to environmental effects, marker-aided selection of the QTLs

18

IRRN 21:1 (April 1996)

can be conducted in early generations of breeding programs. QTL mapping of rice yield and related characters is the first step to apply marker technology to breeding programs for improving yield potential.

Lander ES, Green P, Abbrahamson J, Barlow A, Daly MJ, Lincoln SE, Newburg L. 1987. Mapmaker: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations, Genomics 1:174-181. Lander EF, Botstein D. 1989. Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121:185-199. SAS Institute, Inc. 1988. SAS guide: statistics. Cary, North Carolina: SAS Institute.

Cited references

Use of DNA markers in constructing multilines

Autoradiograph showing hybridization pattern of RG213 in Daeseongbyeo (D), pooled three isolines (I), and Chucheongbyeo (R) DNA digested with the following enzymes: BamHI (1), EcoR I (2), EcoR V (3) Scal (41, and Hin dlll (5). The isoline resistant to race KJ-201 only showed the same restriction pattern as the donor with Sca l.

S. H. Ahn and Y. G. Kim, National Crop Experiment Station (NCES), Rural Development Association (RDA), Suweon, Republic of Korea; S. S. Han, National Agricultural Science and Technology Institute, RDA; H. C. Choi, NCES, RDA; S. R. McCouch, Plant Breeding Department, Cornell University, Ithaca, NY 14853, USA; and H. P. Moon, NCES, RDA

Reactiona of differential varieties and parents to four races of rice blast fungus. Variety Race KJ-101 KJ-201 KJ-301 KI-313 R R S S S S R R S S R S R R R S R S R S R S R S

Construction of multilines, each with a major resistance gene(s) for different Korean blast races, is under way to characterize resistance genes and to test whether deployment of multiple resistance genes in this fashion will provide durable resistance to blast disease. Resistance genes were introgressed from different donors into the commercial cultivars. Identifying molecular markers linked to these genes is essential to speed up the process of making the multilines. Three isolines (BC5F 3) constructed using Daeseongbyeo as a donor and Chucheongbyeo as the recurrent parent were used in the gene-tagging experiment. Each line shows resistance to blast isolates KJ-201, KJ-301, and KI-313, respectively (see table). These three isolates were used individually for inoculation during nearisogenic line (NIL) development. Genetic analysis of Daeseongbyeo indicated that

Differential variety Tetep Tongil Nongbaeg Jinheung Parents Daeseongbyeo Chucheongbyeo

a

R = resistant, S = susceptlble.

one dominant gene conditions resistance to isolates KJ-201 and KI-313 and two dominant genes condition resistance to KJ301 (Kim 1994). Fifteen RFLP markers previously reported to be linked to major blast resistance genes in different rice germplasm (McCouch et al 1993) and six microsatellite markers mapped to nearby regions were surveyed for polymorphism among the lines. One restriction fragment length polymorphism (RFLP) marker, RG213, on chromosome 6 showed polymorphism between the isogenic lines (see figure).

Using the segregating populations, this marker is being tested to determine whether it is linked to the resistance gene (KJ-201). We also have used random amplified polymorphic DNA (RAPD) technique to identify primers that amplify DNA that is polymorphic between the NlLs. Three hundred RAPD primers were used to analyze the genome. Each primer amplified 3.1 loci on average (corresponding to about 930 loci). This survey revealed one locus putatively linked to resistance and demonstrated that the lines were nearly isogenic to their recurrent parent, Chucheongbyeo. This marker also is being tested. Other sets of NILs were developed using two recurrent parents and four donor parents. Based on their reaction to races and phenotype, six NILs were selected and subjected to yield trials and hot-spot leaf blast nursery test. The reaction degree of six NILS appeared to be less severe than that of their respective recurrent parent.
Kim YG. 1994. Analytical studies on the inheritance of reaction to blast fungus in rice. Ph D dissertation. Korea: Cheongnam National University.

Cited references

IRRN 21:1 (April 1996)

19

McCouch SR, Nelson RJ, Tohme J, Zeigler RS. 1993. Mapping of blast resistance genes in rice. In: Zeigler RS, Leong SA, Teng PS, editors. Rice blast disease. United Kingdom and Manila (Philippines): CAB International and IRRI. p 167-186.

Table 1. Comparison of doubled haploid (DH) lines with F1 hybrids and parents. Barapani, India. 950 m above sea level.a Days to maturity P1 (IR70) F1 (IR70/Khonorullo) P1 (Khonorullo) DH1 DH7 DH10 DH21 CD at5% 145 136 129 133 139 130 140 Plant Panlcle- Splkelets panicle-1 height bearing (cm) tillers (no.) (no.) 57 85 129 87 81 78 101 6.82 5.0 7.0 4.0 3.8 3.2 3.2 4.1 1.01 123 147 145 203 132 77 159 19.4 Spikelet fertility (%) 8.1 62.2 60.7 77.3 91.4 94.8 81.5 2.25 1,000grain weight (g) 16.8 23.5 24.3 22.0 30.8 30.0 25.0 1.9 Yleld Grain plant-1 color (g) (p3-W) b 0.4c 8.1 5.2 7.9 5.9 6.2 7.1 0.51 W p2 p3 p1 p1 W W

Breeding methods
Development of cold-tolerant rice through anther culture
H. S. Gupta, R. N. Bhuyan, A. Pattanayak, and D. K. Pandey, Plant Breeding Division, Indian Council of Agricultural Research, Research Complex for NEH Region, Umroi Road, Barapani 793103, Meghalaya, India

a AV yield of 5 plants. b W = white, p3 = red, p2 = light red. p1 = very light red. c Low yield due to cold prevailing throughout the life cycle of the plant.

Table 2. Performance of DH lines in the field at Barapani, India. Variety IR70 Khonorullo DH1 DH7 DH10 DH21 Plant height (cm) 89.4 109.8 118.2 113.6 89.0 78.2 Paniclebearing tillers (no.) 10.4 4.4 16.0 15.6 5.4 10.0 Panicle length (cm) 24.4 18.0 22.0 26.4 20.8 21.4 Splkelets panicle-1 (no.) 105.8 64.8 104.8 172.3 54.4 147.6 Spikelet fertility (%) 72.6 76.2 66.0 63.1 81.6 74.1 1,000-grain weight (g) 23.0 29.3 15.0 22.2 26.7 22.3 Yield (t ha-1) 3.1 2.2 3.3 4.0 1.2 3.3

Rice grown in the high-altitude areas of the northeastern hills of India is exposed to cold at flowering and ripening stages. It often suffers from incomplete panicle exsertion, asynchronous flowering, and poor seed set, resulting in yield losses of 20-50%. In addition, due to a long crop season, only one crop of rice can be grown in these areas. Conventional breeding thus requires more time. Doubled haploid (DH) lines are increasingly being used to make breeding programs practical (Foroughi-Wehr and Wenzel 1990). We developed cold-tolerant lines by compressing the breeding cycle through anther culture. To incorporate cold tolerance into a high-yielding genetic background, IR70 was crossed with Khonorullo, a local redkernel variety with cold tolerance at the reproductive phase. F1-derived anther culture yielded 21 DH lines of which nine were selected. We tested these in the field for yield and yield-contributing characters. All nine yielded significantly more than did the better parent. Four lines-DH1, DH7, DH10, and DH21retained 97.5, 72.8, 87.6, and 87.5% of the yield of the F1, respectively. Two lines exhibited desirable white kernels, and one, DH21, had a high number of spikelets panicle -1 and good 1,000-grain weight. Among the light red kernel lines, DH1 was the most promising (Table 1).

In field tests during 1994-95 kharif (at 950 m above sea level), DH7 yielded the most (4.0 t ha-1), followed by DH21 and DHI, both at 3.3 t ha-1. The better parent. IR70, yielded 3.1 t ha -1 (Table 2). Yield and vigor of the DH lines are either better than those of both parents or as good as those of the better parent, possibly because of in vitro screening of recombinants. Anther culture hastened the fixation of vigor not only by fixing homozygosity but by discriminating against less fit recombinants. Increasing the sound grains panicle-1 of DH21 contributed to the higher yield over the better parent. We will conduct replicated yield trials during the next crop season.

A comparison of the efficiency of four breeding methods

M. Fahim and M. P. Dhanapala, Rice Research and Development Institute, Batalagoda, Sri Lanka; D. Senadhira, IRRI; M. J. Lawrence, Wolfson Laboratory for Plant Molecular Biology, School of Biological Sciences, University of Birmingham, Birmingham B15 2TT, United Kingdom

Cited reference

Foroughi-Wehr B, Wenzel G. 1990. Recurrent selection alternating with haploid steps-a rapid breeding procedure for combining agronomic traits in inbreeders. Theor. Appl. Genet. 80: 564-568.

Conventional breeding methods are successful in that they have been used to produce virtually all cultivars of small grain cereals, including rice. Their success rate per pedigree initiated is, nevertheless, low. Successful breeders must make hundreds of crosses to ensure that some promising lines emerge from each cohort of crosses. We compared the efficiency of three breeding methods with that of single seed descent (SSD) in each of two crosses, Bg850/IR50 and 88-5328/ob255. The performance of F6 lines extracted from each of these crosses by the pedigree (P), modified pedigree (MP), bulk (BL), and SSD methods were assessed for 13 quantitative characters in two completely

20

IRRN 21:1 (April 1996)

randomized trials, one for each cross. Analysis of the basic generations (P1, P2, F1, F2, B1, and B2) and F 3 families of these crosses showed that each of these characters was heritable. The narrow-sense heritabilities of 1,000-grain weight and the number of empty spikelets panicle-1 were high (0.77-0.97), and those of grain yield plant-1 were low (0.17-0.18) in both crosses. The proportion of recombinant inbred lines that could be extracted by SSD from each cross, the performance of which was better than that of the best parent, was predicted using information obtained from F3 families. While only 3% of these lines were expected to achieve this target for panicle weight and number of grains panicle-1 in the first cross, the expected proportion of superior lines was higher (8-82%) for other characters in both crosses. The actual proportion of the 100 F 6 lines produced by SSD from each cross, the performance of which was equal to or better

than the desired target, was close to that predicted for most characters. Furthermore, estimates of the genetic correlation between characters (obtained from the SSD lines) indicated that most of these correlations were less than 0.5. Where they were larger, this was generally favorable to the breeder. Exceptions were from the expected correlations between, for example, grain yield and its components. Both of these pedigrees, therefore, had considerable potential in their ability to produce superior recombinant inbred lines. This was not only for their single character performance, but also for the desired multiple phenotype. The average performance of the lines produced by the conventional methods was higher in both crosses than that of the SSD lines for grain yield plant-1 and number of grains panicle -1. Comparison of the best lines produced by each of the four methods for each character, however, showed that little difference existed between them for

grain yield or its components. The effect of the selection carried out during inbreeding with the P, MP, and BL methods was to cull low-yielding genotypes rather than to retain high-yielding material, despite attempts to do so. The average per line cost of producing the best SSD lines was only US$13.70 compared with US$89.50 for the P method, US$69.70 for the MP method, and US$22.30 for the BL method. The SSD method is also more rapid if three generations can be raised per annum at high densities in the nursery. These results are consistent with quantitative genetics theory, which indicates that attempts to carry out indirect selection visually of characters of low to intermediate heritability during inbreeding are unlikely to be successful. It is better to defer selection until homozygous material can be raised in a randomized trial in which characters of interest can be measured directly.

A promising dwarf rice mutant induced through gamma irradiation

N. R. Bai, R. Devika, and A. Regina, Rice Research Station (RRS), Moncompu, Kerala, India

M01.20-19-4 (IET13981) is a dwarf medium-duration mutant, developed

through gamma irradiation of local rice cultivar Chettivirippu. This cultivar is tall and susceptible to lodging. Its average yield is 2.5 t ha-1, and it is highly resistant to brown planthopper, stem borer, gall midge, blast, sheath blight, and sheath rot. The mutant is shorter, more resistant to lodging, and higher yielding than the parent variety (see table).

We selected plants up to the M 7 generation, after which the stable mutant, M01.2019-4, was included in yield trials (see table). M01.20-19-4 is suitable for cultivation in all of the rice-growing seasons of Kerala, and the people prefer its red grain. The mutant will be proposed for release shortly.

Characteristics, comparative grain yield and pest and disease scores of mutant M01.20-19-4 and parent variety. India, 1989-94. Variety Plant height (cm) Maturity Produc(d) tive tillers hill-1 (no.) Av yield (t ha-1) CYT a (3 seasons) 1989-90 RRS Moncompu (4 seasons) 1990-92 3.5 2.9 0.40 MLT b Research stations (5 locations) 1992 4.6 3.6 ns Farmers' fields (9 locations) 1992 5.2 4.2 0.42 AICRIP c trials Insect-disease score (0-9)

Brown Gall Stem Blast Sheath Sheath (17 (16 plant- midge borer blight rot locations) locations) hopper 1993 1994

MO1.20-19-4 98 (I ET13981) Chettivirippu 120 MO5 Ratna Ptb33 (resistant check) TN1 (susceptible check) CD (0.05)
a

120 110 5 -

4.2 2.7 0.42

4.8 4.0 -

4.4 4.1 -

2.2 2.0 2.1 9.0 -

2.8 2.1 2.5 9.0 -

2.2 2.0 3.5 8.5 -

2.8 2.7 2.1 9.0 -

1.5 2.9 2.0 9.0 -

1.9 2.6 2.0 8.0 -

CYT = comparative yield trial with 7 entries. b MLT = multilocation trials with 7 entries. c AICRIP = All lndia Coordinated Trials; 1993, 68 entries; 1994, 64 entries.

IRRN 21:1 (April 1996)

21

Combining ability analysis on restorability for cytoplasmic male sterility in hybrid rice
Wang Wenming, Wen Hongcan, Zheng Jiakui, Rice and Sorghum Institute of Sichuan Academy of Agricultural Sciences, Luzhou 646100, Sichuan, China

General combining ability (GCA) and specific combining ability (SCA) values a for restorability for CMS. 6323 72922 Ce-64 Hui-403 HR195 Restorers Hui-552 CDR22 Minghui63 (check) 1.62* 0.34 -1.96** -0.46 0.72 ** = significance at 0.05 and 0.01 level, respectively. GCA SE

CMS line nuclear background N-Qing N-17 N-Shan (check) GCA SE

a *,

Restorability, estimated as the mean spikelet fertility or seed set percentage in the F1 generation, refers to the potential of a certain cytoplasmic male sterile (CMS) line to restore fertility in the hybrid derived from it. It is caused by nuclear gene(s) of the restorer (R) lines, and nuclear (N) and cytoplasmic (C) background of the CMS (A) line, as well as the interaction between the two. To understand the relative importance of each factor, three series of isonuclear lines with three types of CMS backgrounds (nine CMS lines) were developed. Then we produced 72 combinations according to North Carolina Design II (9 CMS lines, 8 restorers). In the summer of 1994, each combination was grown in two rows with 9 plants each at 33.3- 20-cm spacing. The experiment was laid out in a split-plot design with three replications in which the main factor was the restorers. Data on seed set percentage were collected from the center five plants of either row. Statistical analysis was based on the sin -1 p transformation data.

1.49* -0.90 -1.02 -0.18 -0.56 0.11 -1.30* 1.45* 0.91 -4.25** 1.44 0.03

-1.74** -0.14 1.88** 2.00*

-0.84 -0.33 1.18 3.70**

0.61 1.38* -1.98** - 4.55**

0.78 -0.61 0.17 2.08*

1.80** -0.41 -1.39** 0.232 0.614

Variance analysis revealed that effects on restorability were significantly different among restorers, CMS line nuclear backgrounds, and N-R interactions, while not significantly different among the cytoplasmic background of CMS lines, C-R interactions, C-N interactions, and C-N-R interactions. Subsequent analysis of combining ability showed that HR195 and N-Qing had the highest general combining ability (GCA) among restorers and N-Qing among nuclear background of CMS lines. Among the restorers, the GCA for restorability of CDR22, Hui-403, 72922, and Ce-64 was higher than that of Minghui-63, a widely used restorer in hybrid rice production, and the GCA of 6323 and Hui-552 was lower. Both N-Qing and N-17 showed higher GCA

for restorability than did N-Shan, a commonly used CMS line nuclear background in China (see table). The relative importance analysis suggested that the restorer was ranked highest in restorability for CMS, followed by CMS line nuclear background, and N-R interaction. The GCA of the restorer and the nuclear background of the CMS line primarily determined the restorability for CMS (accounting for 86.4%), with specific combining ability (13.6%) having some influence. Thus, to improve CMS restorability, emphasis must be placed on the selection of restorers and maintainers (the nuclear background of CMS lines) with higher GCA. HR195 and N-Qing have therefore been exploited in hybrid rice breeding for their higher GCA for CMS restorability.

Male sterile line in rice (Oryza sativa) developed with O. glumaepatula cytoplasm
R. D. Dalmacio, D. S. Brar, S. S. Virmani, and G. S. Khush, IRRI

O. glumaepatula and the nuclear genome of IR64. Serving as female parents, 48 accessions of 3 wild species (AA genome) were crossed with IR64, one of the restorers of

WA cytoplasm. Data on pollen sterility were recorded in F 1 and backcross generations. Hybrids showing 70% or more pollen sterility were subsequently

Pollen fertility of F1 and backcross progenies derived from the cross of O. glumaepatula (Acc 100969) and IR64. Generation BC8 BC 7 BC 6 BC 5 BC 4 BC 3 BC 2 BC 1 F1 Pollen fertility (% of progeny) a CS 100 92 86 95 57 65 0 21 17 S 0 8 14 4 43 20 2 23 83 PS 0 0 0 1 0 10 16 30 0 PF 0 0 0 0 0 5 43 14 0 F 0 0 0 0 0 0 39 12 0 Plants analyzed (no.) 45 12 41 147 7 20 89 197 6

Most of the commercial hybrids of indica rice are based on the wild abortive (WA) source of cytoplasmic male sterility (CMS). More than 95% of all hybrid rice grown in China is based on the WA cytosterility system. Such cytoplasmic uniformity of hybrid rice could lead to genetic vulnerability to diseases and insects. To overcome this problem, diversification of the source of CMS is essential. We developed a CMS line with the cytoplasm of AA genome wild species

a Pollen fertility classes: CS = completely sterile, 0% pollen fertility; S = sterile, 1-10% pollen fertility: PS = partially

sterile, 11.30% pollen fertility; PF = partially fertile, 31.60% pollen fertility; F = fertile, 61.100% pollen fertility.

22

IRRN 21:1 (April 1996)

A new cytoplasmic male sterile line, lR69700 A, has the cytoplasm of O. glumaepatula and nuclear background of IR64.

backcrossed with the recurrent parent. Of all the backcross derivatives, one line with O. glumaepatula (Acc 100969) cytoplasm was found to be stable for complete pollen sterility (see table). This line was designated as IR69700 A. Earlier, we developed a CMS line (IR66707 A) with O. perennis (Acc 104823) cytoplasm and the nuclear genome of IR64. Crosses of IR69700A with nine restorers of WA cytoplasm show almost complete (88-100%) pollen sterility, indicating that the male sterility source of IR69700 A is different from WA cytosterility. The new CMS line is now in BC10 and resembles maintainer IR69700 B in morphological characters, except that it flowers 5-7 d later (see figure). This line is completely sterile and does not set any seed upon selfing. We are searching for restorers of this male sterile line for use in hybrid rice breeding.

Pattern of segregation for grain weight involving bold and slendergrained rice varieties
P. Vivekanandan and S. Giridharan, Tamil Nadu Rice Research Institute, Aduthurai, India

Fine-grained rice varieties are generally preferred in the market and sell for a premium price. Grain weight determines the fineness of the grain. We evaluated the pattern of segregation for grain weight in the F2 generation for four parents: TKM 9 (short bold), IR50 (long slender), White Ponni (medium slender), and Pusa Basmati 1 (extra long slender) with contrasting grain types. We grew 30 plants for each replication of P 1, P 2, and F 1 in two 3-m rows and 120 plants of F2 in eight rows during 1992 dry

Pattern of segregation for hundred-grain weight in F2 progenies.

season. The trial was replicated three times. One seedling hill-1 was planted using a spacing of 30 30 cm. Ten plants replication-1 in P1, P2, and F 1 and 70 plants replication-1 in F 2 were evaluated for hundredgrain weight.

Among the parents, White Ponni possessed the least grain weight (0.016 g). The F1 was intermediate between the parents in two crosses and exceeded the parental limits in IR50/Pusa Basmati 1. All crosses showed transgressive segregation for grain weight in both directions. White Ponni/Pusa Basmati 1 recorded the least mean grain weight in F 1 and F 2 generations with a high coefficient of variation (see table). It possessed high frequency (45 plants) of fine-grained segregants (0.014-0.017 g) and offers scope for selection. The cross involving bold- and slendergrained varieties (TKM9/Pusa Basmati 1) registered high frequency of progenies with grain weight of more than 0.022 g (see figure).

Mean performance, range, and coefficient of variation for hundred-grain weight. Mean Cross combination TKM9/Pusa Basmati 1 IR50/Pusa Basmati 1 White Ponni/Pusa Basrnati 1 Pistil parent 2.41 + 0.01 1.80 + 0.01 1.67 + 0.01 Pollen parent 1.88 + 0.01 1.88 + 0.01 1.88 + 0.01 F1 2.11 + 0.02 1.98 + 0.02 1.79 + 0.01 F2 2.13 + 0.01 1.94 + 0.01 1.76 + 0.01 Range (g) 1.6 - 2.8 1.6 - 2.5 1.4 - 2.3 CV (%) 9.44 8.15 10.40

IRRN 21:1 (April 1996)

23

Yield potential
~

The yield potential of three crosses between diverse parents of indica type
S. Sriyoheswaran, M. P. Dhanapala, Rice Research and Development Institute, Batalagoda, Sri Lanka; D. Senadhira, IRRI; and M. J. Lawrence, Wolfson Laboratory for Plant Molecular Biology, School of Biological Sciences, University of Birmingham, Birmingham B15 2TT, United Kingdom

A widespread belief exists that the yield of indica cultivars grown in the lowland tropics has reached a barrier of 10 t ha -1 because of the narrow genetic base of modern varieties. We tested this hypothesis by evaluating the yield potential of the crosses between high-yielding Sri Lankan varieties and high-yielding IRRI varieties: Bg34-8/IR58 (cross 1; 3-mo duration), Bg94-1/IR62 (cross 2; 3 1/2 mo duration) and Bg90-2/IR72 (cross 3; 4 1/2 mo duration). Individuals raised were of the basic generations (P 1, P2, F1, F2, B1, and B2) and F 3 and triple testcross families (F2P1, F2P2, and F 2F1) of each of these pedigrees in three completely randomized experiments, one for each cross. Each of the 14 characters scored was heritable. Estimates of the heritability of grain yield, obtained from F 3 families for

crosses 1, 2, and 3, were 0.29, 0.38 and 0.51, and for 1,000-grain weight over 0.57, 0.58, and 0.70, respectively. Though this was significant only in cross 1, the F 1 mean exceeded that of the best parent in all three crosses for grain yield. The dominance ratio, (H/D) , which was less than one in each cross, indicated that the increasing genes for this character were dispersed between the parents. Predictions of the proportion of recombinant inbred lines that could be extracted by single seed descent from each cross, the performance of which exceeds that of the best parent, showed that it should be relatively easy to obtain such lines from each cross for the majority of characters, including grain yield and 1,000grain weight. Estimates of genetic correlations between characters varied considerably over crosses, suggesting that their major cause is the linkage disequilibrium of linked genes rather than pleiotropy. Grain yield was strongly and positively correlated (r > 0.7) with tiller and panicle number, days to heading, height at flowering, and panicle weight in cross 3. However, in cross 2, grain yield was only strongly and positively correlated with number of grains panicle -1 and panicle weight, and in cross 1, with tiller number and panicle number. Results suggest that it should be possible to identify We evaluated 13 early-duration rice genotypes for grain yield and morphoagronomic attributes. The experiment was laid out in a randomized block design with three replications during the 1990-92 wet seasons. Each genotype was direct seeded in ten 5-m long rows using 25-cm spacing. We recorded plant height, panicle length, fertile grains panicle -1, and percentage spikelet fertility for five randomly selected plants from each plot of every replication. Days to flowering, days to maturity, and grain yield were computed for each plot. Number of panicle-bearing tillers m-2, from a randomly selected 1-m 2 plot was recorded. Means were used for stability analysis over three environments following the method of Ebherhart and Russel (1966).

crosses in a breeding program from which recombinant inbred lines can be extracted that are superior for any combination of traits of a desired multiple phenotype. Upon examining the means of the 38-40 F3 families raised in these experiments, one or more families achieved the desired target for each of the majority of characters, including grain yield in all three crosses. Evidence of transgressive variation for characters of interest as early as the F 3 generation leaves little doubt about their potential to produce superior inbred lines. The mean grain yield of the best F 3 family, cross 3, which is in the same age group as the new plant type, was 154.4 g. The predicted yield was 9.7 t ha-1, which is appreciably better than the predicted yield of 8.3 t ha-1 for the best parent of this cross, Bg90-2. Results from this investigation are inconsistent with the belief that the apparent yield barrier of indica cultivars is due to a narrow genetic base. This suggests that the inability of breeders to breach this barrier is because of other causes, including inefficient breeding procedures and employing indirect visual selection when attempting to obtain improvement for several quantitative characters simultaneously.

Stability analysis of 13 early duration upland rice genotypes in Bastar Plateau Zone, Madhya Pradesh, India
S. S. Rao, A. Bhatnagar, R. Verma, T. D. Pandey, and K. L. Nandeha, lndira Gandhi Agricultural University, Zonal Agricultural Research Station, Kumharawand, Jagdalpur 494005, Madhya Pradesh, India

Table 1. Mean grain yield (t ha-1) and estimates of stability parameters for 13 rice genotypes across three environments. Jagdalpur, Madhya Pradesh, India, 1990-92. Genotype R.281 PP 31-1 RWR 78-71-9 JR80-4-6 JR84-7-1-19 RNR1446 Culture 1 Poorva Annada Tulasi Rasi Aditya Tellahamsa Badoldhan (L) x 1.64 1.97 1.56 2.19 2.07 1.66 1.34 2.55 2.37 2.19 2.14 2.19 1.97 av b 0.88 1.33 1.34 0.78 1.17 0.03**a 0.11** 1.85** 2.0 1.95** 0.89 0.87 0.16** 1.99 s2d 2.84 4.44 1.16 8.75 5.88 6.59 6.55 4.72 4.75 4.22 3.68 0.06 6.28

Nearly 0.24 million ha of upland rice is grown in Bastar Plateau Zone in Madhya Pradesh during wet season (June-September). Traditional, low-yielding cultivars are dominant in the zone. Suitable upland early rice varieties need to be identified.

a** = significant at the 1% level.

24

IRRN 21:1 (April 1996)

Table 2. Morphoagronomic attributes (av of 3 yr) of upland rice genotypes at Jagdalpur, Madhya Pradesh, India. Genotype R281 PP 31-1 RWR78-71-9 JR80-4-6 JR84-7-1-19 RNR1446 Culture 1 Poorva Annada Tulasi Rasi Aditya Tellahamsa Badoldhan (L) Poorva/lR8608-298 T3/T141//T3 DR42/Ratna Tellahamsa/Rasi AC540/Ratna Saket 4/JR 2-331 MTU15/Waikaku Rasi/white gora TN1/CO 29 M63-83/Cauvery H12/TN1 Days to 50% flowering 82 77 84 72 80 67 70 80 81 83 66 81 78 Duration (d) 109 104 109 99 106 94 97 106 107 110 90 107 104 Plant height (cm) 71 76 66 70 81 95 65 71 79 76 75 80 120 Panicle-bearing tillers m -2 (no.) 302 351 313 261 311 324 320 300 305 316 288 279 217 Panicle length (cm) 19.0 18.2 18.5 19.6 18.8 20.6 17.4 20.4 18.2 17.8 17.8 18.5 22.1 Fertile grains panicle -1 (no.) 60 56 57 65 62 61 53 82 66 64 58 66 99 Spikelet fertility (%) 86 86 87 84 90 87 86 91 89 89 90 90 93

Regression coefficients (b) ranged from 0.16 to 2.01 (Table 1). The variance due to deviation from regression (s2d) was -6.598.75. The stability parameters of seven genotypes were not significant, indicating they are stable yielders across environments. Rice varieties Aditya and Tellahamsa had b values close to unity (0.89 and 0.87),

indicating their high stable performance across the environments tested. These varieties yielded above average and have been released for upland cultivation around India. Varieties Annada, Tulasi, and Rasi show slightly above average grain yield (Table 1). These b values were significantly deviated from unity, indicating they were more responsive under favorable environments.

Stability for grain yield in rice appears to be imparted by the stability of most of the yield attributes (Table 2). Genotypes Aditya, Tellahamsa, RNR1446, R281 PP 31-1, and JR84-7-1-19 were stable yielders over time. They are recommended for upland cultivation in Bastar Plateau Zone and for use in breeding cultivars with yield potential and adaptability.

Variability, heritability, genetic advance, and genetic divergence in upland rice

M. K. Sarma, A. K. Richharia, and R. K. Agarwal, Genetics and Plant Breeding Department, Institute of Agricultural Sciences, Banaras Hindu University (BHU), Varanasi 221005, India

Attempts to improve rice cultivars grown in the upland ecosystem have been limited. Systematic evaluation of the genetic variability and divergence of the existing upland rice germplasm is, therefore, important. We studied genotypic and phenotypic coefficients of variation, heritability, genetic advance, and genetic divergence of

39 upland rice genotypes. Most of them were early-maturing indigenous cultivars from Madhya Pradesh, India. The experiment was laid out in a completely randomized block design with three replications, during 1993 kharif (dry) season. The significant mean sum of squares indicated the strong variability among the

Table 1. Genetic parameters of variation in upland rice. BHU, Varanasi, India. 1993. Parameter Traits Days to 50% flowering Days to maturity Plant height (cm) 62.4 108.2 198.3 211.1 15.9 16.4 93.9 31.8 4.0 Primary branches panicle -1 (no.) 4.7 10.4 1.3 1.8 13.5 16.2 69.4 23.2 9.0 Secondary branches panicle -1 (no.) 6.6 32.2 27.3 33.9 28.8 32.2 80.4 53.2 14.3 100grain weight (g) 1.6 2.8 0.1 0.1 13.2 13.9 90.7 25.9 4.2 Panicle weight (g) 0.6 3.1 0.3 0.3 29.1 30.8 89.6 53.9 9.9 Panicle length (cm) 15.3 25.0 4.4 6.9 9.9 12.5 64.0 16.4 7.5 Spikelets panicle-1 (no.) 37.6 158.0 576.7 636.4 25.4 26.7 90.6 49.8 8.2 Grain Effective length tillers m -1 (mm) row length (no.) 6.2 9.9 0.8 0.8 10.6 10.6 99.9 21.9 0.4 36.7 251.0 1066.5 1518.1 44.8 53.4 70.2 77.3 29.1 Grain yield m-1 row length (g) 17.5 82.0 198.0 469.6 25.9 39.9 42.2 35.6 30.3

Range: Minimum 66.3 96.3 Maximum 47.0 Genotypic variance Phenotypic variance 50.7 8.6 Genotypic coefficient of variation Phenotypic coef8.9 ficient of variation 92.8 Heritability (%) Genetic advance 17.1 (% of mean) CV (%) 2.4

87.3 124.3 85.2 100.1 8.6 9.4 85.1 16.4 3.6

IRRN 21:1 (April 1996)

25

Table 2. Genetic divergence a in upland rice. BHU, Varanasi, India. 1993. Cluster Genotype b Days to 50% flowering 1 1 Ateya, Agnichar, Badaebutta, Badelwa Bademokado, Barangi, Barhi deshi, Bhadaili, Baril bhadainu, Lal bhadainu 1, Lal bhadainu 3, Bakki c Badsahbhog 1, Bastaria Batasi, Bhatta khuji, IR8 mutant d Akashie Amakoyeli, Alsengha, Bakhlasal, Barhasal, Barhi, Bayo Anokhi, Wishnoo bhog Tulasi f Badwani 5, Badwani-22, Bhatta dhan Baduri, Bala, Bhatta goyandi Badsah bhog 2 Bagri Aerramindo, Lal bhadainu 2 Badal phool, Annada f 77.7 (15.9) Days to maturity 2 102.7 (25.2) Plant height (cm) 3 93.4 (32.4) Primary branches panicle -1 (no.) 4 8.5 (19.1) Secondary 100branches grainpanicle -1 weight (no.) (g) 5 6 17.1 (50.1) 2.3 (61.9) Panicle weight (g) 7 2.0 (43.2) Panicle length (cm) 8 21.5 (23.8) Spikelets panicle -1 (no.) 9 93.6 (29.9) Grain Effective length tillers (mm) m-1 row length 10 (no.) 8.2 (44.1) 74.5 Grain yield m -1 row length (g) 59.7

2 3 4 5 6 7 8 9 10

83.0 82.5 83.3 73.0 83.5 77.0 66.3 71.0 88.2

108.1 (15.8) 112.4 104.3 95.4 113.8 105.0 96.7 111.0 123.0

83.3 (48.2) 99.8 (14.7) 94.7 89.9 86.1 65.6 63.7 78.6 63.7

8.1 (32.7) 8.6 (27.5) 7.7 (16.5) 8.9 8.4 7.1 4.7 8.8 8.4

17.5 (69.7) 19.8 (24.8) 17.3 (43.0) 17.6 (13.0) 25.3 22.1 6.6 18.3 16.2

2.2 (42.6) 2.6 (89.7) 2.1 (70.6) 2.4 (109.1) 1.8 (16.1) 1.9 2.5 1.8 2.1

1.8 (24.3) 2.4 (71.0) 1.5 (51.8) 1.9 (89.7) 1.8 (22.9) 1.7 (0.0) 0.6 1.7 1.6

20.0 (29.7) 21.6 (39.1) 20.7 (27.9) 22.9 (53.0) 21.5 (63.9) 17.7 (44.0) 15.3 (0.0) 21.7 20.1

95.5 (20.3) 98.9 (55.3) 80.7 (34.2) 96.3 (72.1) 114.4 (44.3) 112.0 (26.9) 37.6 (34.8) 100.1 (18.5) 78.5

7.6 (60.7) 9.1 (28.3) 8.4 (33.8) 9.7 (56.1) 6.4 (99.4) 7.0 (80.2) 8.1 (47.7) 7.5 (60.4) 9.2 (18.4)

72.6 17.9 64.8 60.7 57.7 73.0 251.0 52.3 57.3

43.9 61.9 52.6 61.8 32.6 50.4 51.5 61.3 36.8

a Figures without parentheses are mean values of the traits. Figures in parentheses are intercluster divergence values. Bold figures in parentheses are intracluster divergence values. bUnless otherwise indicated, other genotypes are indigenous cultivars of Madhya Pradesh, India. c Local variety of Varanasi. d Early mutant of IR8 developed at BHU. eRecently designated variety. fNational check varieties.

genotypes for the traits studied. The considerable range of variation expressed for the traits indicated good scope for genetic improvement (Table 1). Genotypic coefficient of variation (GCV) was the highest for effective tillers m -1 row length (44.8%) followed by panicle weight (29.1 %), secondary branches panicle -1 (28.8%), grain yield m-1 row length (25.9%), and spikelets panicle -1 (25.4%). Broad-sense heritability estimates ranged from 42.2% for grain yield m -1 row length to 99.9% for grain length. Effective tillers m-1 row length, panicle weight, secondary branches panicle-1, and spikelets panicle-1 showed high GCV along with high heritability, indicating effectiveness of selection based on these traits. Even though

days to 50% flowering (92.8%), days to maturity (85.1 %), plant height (93.9%), 100-grain weight (90.7%), and grain length (99.9%) had high heritability, they had low GCV. Genetic advance as a percentage of mean was highest for effective tillers m-1 row length (77.3%) followed by panicle weight (53.9%), secondary branches panicle -1 (53.2%), and spikelets panicle -1 (49.8%). Along with high levels of genetic advance, these traits also showed high GCV and should be considered when selecting to achieve high genetic gain. Genetic divergence was estimated using D2 analysis. Ten distinct clusters were obtained following Tochers method of clustering (Table 2). Mean values for the

traits in each cluster and inter- and intracluster distances were determined. Genotypes from the following clusters may be selected for use as parents in a hybridization program: cluster 8 for earliness: clusters 8 and 10 for dwarfness; cluster 6 for secondary branches panicle -1; cluster 3 for panicle weight; clusters 6 and 7 for spikelets panicle-1; and cluster 8 for effective tillers m-1 row length. The highest genetic divergence was observed between clusters 5 and 6, followed by clusters 5 and 8, and clusters 3 and 6. This suggests that the genotypes belonging to these cluster pairs would produce wide spectra of variation in segregating generations following hybridization.

26

IRRN 21:1 (April 1996)

Response of rice genotypes to N fertilization under temperate conditions of Kashmir Valley

A. S. Bali, K. N. Singh, and Amarjit S. Bali, Agronomy Division, Sher-e-Kashmir University of Agricultural Sciences and Technology, Shalimar, Srinagar-Kashmir, India

Yield and yield attributes of rice genotypes under different levels of N. Shalimar, Srinagar Kashmir, India. 1991 and 1992 kharif. Treatment Panicles m -2 (no.) 1991 N level (kg ha-1) 0 40 80 120 160 LSD (0.05) Genotype SKAU5 SKAU11 SKAU23 SKAU27 SKAU30 LSD (0.05)
a

Grains panicle -1 (no.) 1991 77 72 83 91 89 5 79 82 81 81 81 ns 1992 65 70 80 88 89 4 78 80 78 79 79 ns

1000-grain weight (g) 1991 23.4 24.0 24.6 24.2 24.2 0.5 23.9 24.1 24.2 24.1 24.1 ns 1992 23.1 23.6 24.0 24.1 24.1 0.2 23.8 23.7 23.8 23.9 23.9 ns

Grain yield (t ha -1) 1991 2.3 3.5 4.5 5.3 5.8 0.2 4.3 4.2 4.2 4.2 4.3 ns 1992 2.5 3.8 4.7 5.5 6.1 0.3 4.5 4.4 4.4 4.5 4.6 ns

Days to maturity 1991 129 130 132 132 133 1.9 133 133 126 132 132 1.21 1992 129 131 133 133 134 2.0 134 134 127 132 133 1.4

1992 249 327 381 421 435 9 361 361 360 365 366 ns

SKAU5 formerly K39 is a mediumduration (133 d) rice cultivar that has been widely cultivated during the past 20 yr in the main belt of the Kashmir Valley. Reliance on a single variety could prove disastrous over the long term. Diversifying the varieties grown in the area needs to be considered, particularly with the recent introduction of rice-based double cropping systems. Further, there are indications that the N needs of rice are likely to be greater than the current recommendation of 80 kg N ha-1. We studied the response of five rice varieties to five levels of N in a split-plot design during 1991 and 1992 kharif (dry) seasons (May-September). The experimental site was 1650 m above sea level, 34.1 N latitude and 74.9 E longitude. The soil was silty clay loam in texture, with pH 6.8, 1.21% organic C, 13.7 and 16.2 kg available P ha-1 (Olsens method), 112 and 128 kg available K ha -1 (ammonium acetate extraction method), 0.01 and 0.0 12% available N (alkaline permanganate

249 326 383 424 437 13 363 362 362 367 365 ns a

ns = not significant.

method) during 1991 and 1992, respectively. Thirty-five-day-old seedlings were fertilized before transplanting with 19.8 kg P ha-1, 16.6 kg K ha-1, and half of the designated amount of N. The remaining N was applied in two equal splits at tillering and panicle initiation. Management practices were identical during both years. Yield, yield components, and the nature of yield responses to N were similar over both years (see table). Rice grain yield increased with each increment of N up to 160 kg ha-1 , which registered the maximum yield of 5.8 t ha -1 during 1991.

The optimum dose of N was 143 kg ha-1, calculated from the pooled grain yield. Increase in grain yield occurred mainly through the contribution of panicles m-2. Grains panicle increased consistently up to 120 kg N ha-1 but 1,000-grain weight did not change with N rates. Yield and yield attributes did not differ among the genotypes tested. However, SKAU23 matured 1 wk earlier than the others. N rates need to be increased from the current 80 kg to 140 kg ha-1 for optimum yields. Cultivar SKAU23, which matures 1 wk earlier than SKAU5, can better fit in a double cropping system.

Grain quality
Enzyme isolation and regulation with lysine biosynthesis and degradation in developing seeds of rice
S. A. Gaziola and C. M. G. Teixeira, Genetics Department, State University of Campinas, Brazil; A. Ando, Genetics Department, University of Sao Paulo, Brazil; L. Sodek, Plant Physiology Department, State University of Campinas, Brazil: and R. A. Azevedo, Genetics Department, University of Sao Paulo

The seeds of cereals are nutritionally deficient in certain essential amino acids. Aspartic acid serves as a common precursor of several amino acids, including lysine

(Lys), threonine (Thr), methionine (Met), and isoleucine (Ile). As yet, the aspartic acid pathway has not been studied in rice. We are currently investigating the following enzymes in rice seeds: aspartate kinase (AK), the first enzyme of the pathway, and homoserine dehydrogenase (HSDH), which plays a major role in the biosynthesis of threonine, together with lysine ketoglutarate (LKR) and saccharopine dehydrogenase (SDH), both of which are involved in the catabolism of lysine. All of the enzymes were extracted from the developing seeds, roots, and leaves of rice cultivar IAC165, and LKR and SDH only from the cotyledons and developing seeds of soybean in Tris and phosphate

buffers containing DTT, EDTA, PMSF, PVP, KCl, and glycerol. The enzymes were partially purified with an ammonium sulfate (0-60% saturation) precipitation step and desalted on Sephadex G25 or G50 columns. Enzyme activities were measured using the methods of Azevedo et al (1992a, 1992b) and Brochetto-Braga et al (1992). All three stages presented AK and HSDH activities, with stage 2 showing higher levels of activity. Lys (5 mM) was the major inhibitor of AK activity, with inhibition varying from 30 to 90% among the stages tested (Table 1). Thr (5 mM) only slightly (4-12%) inhibited AK activity, and Lys or Thr did not inhibit HSDH (Table 1). Lys predominantly inhibited AK in rice as has been observed in all of the plant species

IRRN 21:1 (April 1996)

27

Table 1. Aspartate kinase (AK) (nkat mg prot -1 ) and Homoserine dehydrogenase (HSDH) (nmol min -1 mg prot -1 ) activities extracted from developing seeds of rice. Enzyme/treatment AK Control +Thr +Lys +Lys + Thr HSDH Control +Thr Developing seeds Stage 1 0.046 0.044 0.004 0.003 1.389 0.849 Stage 2 Stage 3 0.070 0.061 0.036 0.025 0.704 0.610 0.055 0.052 0.042 0.040 0 0

These preliminary tests showed that both enzymes are very sensitive to proteases and phenolic compounds during extraction. High concentration of salts in the buffer also appears to increase enzyme activity. The evidence suggests that both enzymes are seed-specific in rice but not in soybean.

Cited references
Azevedo RA, Blackwell RD, Smith RJ, Lea PJ. 1992a. Three aspartate kinase isoenzymes from maize. Phytochemistry 31:3725-3730. Azevedo RA, Smith RJ, Lea PJ. 1992b. Aspartate kinase regulation in maize:

evidence for co-purification of threoninesensitive aspartate kinase and homoserine dehydrogenase. Phytochemistry 31:37313734. Brochetto-Braga M, Leite A, Arruda P. 1992. Partial purification and characterization of lysine-ketoglutarate reductase in normal and opaque-2 maize endosperms. Plant Physiol. 98:1139-1147. Lea PJ, Blackwell RD, Azevedo RA. 1992. Analysis of barley metabolism using mutant genes. In: Shewry PR, editor. Barley: genetics, molecular biology and biotechnology. Oxford: CAB International. p 181-208.

Table 2. Lysine ketoglutarate (LKR) and saccharopine dehydrogenase (SDH) activities (nmol min -1 mg prot -1 ) extracted from rice and soybean tissues. Plant tissue Rice Developing seeds (stage 1) Developing seeds (stage 2) Developing seeds (stage 3) Root Leaf Soybean Cotyledons Developing seeds LKR activity 4.6 9.1 0 0 0 0.24 5.10 SDH activity 5.0 4.2 2.0 0 0 1.06 3.16

Pest resistancediseases
Use of genetic male sterilefacilitated recurrent selection for blast resistance in rice
B. N. Singh, West Africa Rice Development Association (WARDA) Rice Program, International Institute of Tropical Agriculture (IITA), lbadan Nigeria; T. M. Masajo, IRRI, Antananarivo, Madagascar; and O. A. Oladimeji, WARDA Rice Program, IITA, Ibadan, Nigeria

studied so far (Lea et al 1992), with the exception of Coix lacryma-jobi. It is currently difficult to assure that an isoenzyme sensitive to Thr is present in rice seeds. For HSDH, we did not observe a Thr-sensitive form of the enzyme, as isolated in other plant species studied. These results suggest that a bifunctional enzyme containing Tht-sensitive AKHSDH activities, such as the enzyme isolated from some higher plants (Azevedo et al 1992b), is not present in rice. LKR and SDH were extracted from rice and soybean tissues using phosphate and Tris buffers. The inclusion of PVP is essential to maintain both activities independent of the buffer system used. The presence of KC1 at 50 mM also helped to maintain the activity. LKR and SDH activities were observed in developing seeds at three different stages. However, LKR and SDH activities were not present in roots and leaves (Table 2). In soybean, both LKR and SDH activities were observed in developing seeds and cotyledons (Table 2).

Rice blast (Pyricularia grisea) is a serious disease in rainfed, upland, and lowland ecologies in Africa. Many blast races are rapidly changing (Awoderu 1970, Fomba and Taylor, 1994). Breeding resistant varieties by combining genes from various sources is one of the most efficient and costeffective approaches for blast management. We used genetic male sterile lines to incorporate genes from various improved donors.

In 1984 wet season, 18 crosses were initially made by hand pollination between 6 genetic male sterile lines from IRRI and 11 improved donors that are irrigated and rainfed lowland plant types with resistance to blast and tolerance for iron toxicity (see table). The male sterile lines have intermediate height and medium growth duration except for IR36, which is semidwarf and early-maturing. From the F 2 and subsequent generations, the seeds from male sterile plants were harvested after random mating. In each cycle, the population was subjected to natural blast infestation. In the third cycle, crosses were also made with eight new lines with improved plant type (see table). After five cycles of random mating, selections were made for male fertile recombinants and grown using the pedigree method. Lines were further screened for leaf blast resistance under spreader row inoculation method. Three to

Male sterile lines and fertile donors used in the study. Male sterile lines IR36 (ms) lR41512 lR41519 lR41523 lR41525 lR41530 Male fertile donors a ITA121 ITA212 ITA230 PELITA 1-1 TOX906-92-20-1-1-1 TOX725-1-8-20-1 ITA245 Suakoko 8 ITA247 Mahsuri ITA231 Male fertile donors
b

BKNFR76106-16-0-1-0 lR24637-38-2-21 ITA230 ITA315 Mat Candu TOX3058-28-1-1-1 TOX3118-2-E2-2-3 TOX3219-24-1-3

a Initial crosses made in 1984. b Crosses made in third cycle.

28

IRRN 21:1 (April 1996)

Frequency distribution of 568 lines for their reactions to leaf blast.

six plants were selected from each segregating line. The 27 fixed lines with blast resistance were evaluated for yield in 1994 wet season. Leaf blast resistance of 568 lines generated through genetic male sterilefacilitated recurrent selection (GMSFRS) were evaluated in a screenhouse under upland direct seeded condition during the 1994 wet season (August-October). The spreader row consisted of mixed seed of three highly susceptible lines: OB677, TOX3055-10-1-1-1-1 (TOX3055), and ITA306. The test entries (10 g each) were seeded in single, 50 cm-rows, 16 d later. Twenty days after seeding, the spreader rows were inoculated with mixed field isolates of leafblast at ITTA. Rainfall was favorable for disease development. Five checksIR46, ITA212, ITA305, TOX3055, and ITA306were randomly planted in each 10th row to monitor variation in leaf blast infection. The susceptible checks, such as OB677, TOX3055, and ITA306, had developed the susceptible lesions and more than 50% diseased leaf area (DLA) by 14 d. Leaf

blast infection was recorded from 14 to 42 d after inoculation at weekly intervals for both lesion type and % DLA using the Standard evaluation system for rice (IRRI 1988). We classified the 568 pedigree lines derived from the interbreeding composite population into four groups: Group I : Immune type: no lesion development, true or complete resistance. Group II: Moderately resistant type: initially resistant, and the disease developed to a score of 2 or 3 and up to 5% DLA (ITA305). Group III : Partial or field resistant type: initially resistant to moderately resistant, and the disease developed to a major lesion type score of 4 or 5 (moderately susceptible) and from 6 to 25% DLA. Group IV: Susceptible type: initial reaction was resistant, moderately resistant (IR46), or moderately susceptible (TOX3055, ITA306, or ITA212), and the disease score ranged from 7 (susceptible) to 9 (highly susceptible) and DLA was 26-100%.

Thirty percent of the lines were classified as highly resistant (group I), another 22% were moderately resistant (group II), 33% were partially resistant, and 25% susceptible to highly susceptible. Among partially resistant types, 12% were initially highly resistant (I, IIIii). Among susceptible types, 4% were initially resistant (IV-i), 16% had moderate resistance (IViii), and another 5% had a moderately susceptible reaction (IV-iii). The frequency distribution shows that 85% of the lines had blast resistance incorporated through GMSFRS (see figure). Variation was observed in the DLA of five checks. To evaluate the elite lines, the entries need to be replicated. Three highly susceptible lines and broad-based field isolates will be used to inoculate the spreader rows when selecting lines with horizontal resistance. In GMSFRS, the parent identity is lost to random mating and a composite population of lines is generated in each cycle. Male sterile-facilitated recurrent selection has been used in barley to develop composites (Jensen 1988). Genetic male sterile lines were effectively used to develop interbreeding composite populations for incorporating the genes for blast resistance from various improved donors. We are evaluating many resistant lines with rainfed and irrigated lowland plant type for their yield potential and resistance to biotic and abiotic stresses.

Awoderu VA. 1970. Identification of races of Pyricularia oryzae in Nigeria. Plant Dis. Rep. 54:520-523. Romba SN, Taylor DR. 1994. Rice blast in West Africa: its nature and control In: Zeigler RS, Leong SA, Teng PS, editors. Rice blast disease. UK: CAB International. p 343-355. [IRRI] International Rice Research Institute 1998. Standard evaluation system for rice. Manila (Philippines): IRRI. 54 p. Jensen NF. 1988. Male sterile facilitated recurrent selection method. In: Plant breeding methodology. New York: John Wiley & Sons. p 243-248.

Cited references

IRRN 21:1 (April 1996)

29

Mapping a new gene for resistance to bacterial blight using RFLP markers
Lin Xinghua, Zhang Duanpin, Xie Yuefeng, Zhang Qifa, and Gao Heping, State Key Laboratory in Genetic Improvement of Crops, Huazhong Agricultural University, Wuhan 430070, China

Performance of RFLP marker of plants in the susceptible population. Probe R543 G181 RZ536 C950 RG1109 G4001 Type of band 1 1 1 1 1 2 3 3 2 2 5 7 7 8 17 17 3 39 36 34 34 32 22 22
a

p (%)

Total 42 42 42 42 42 42 42 4.8 8.3 10.7 10.7 14.3 27.4 27.4 2.3 3.0 3.4 3.4 3.8 4.9 4.9

Genetic map distance (cM) 4.8 8.4 10.9 10.9 14.7 30.8 30.8 2.3 3.0 3.4 3.4 3.8 4.9 4.9

Bacterial blight (BB) of rice, caused by Xanthomonas oryzae pv. oryzae (Ishiyama) Dye, is prevalent throughout Asia. Some resistant cultivars were identified from 6,184 Yunnan local varieties by using 10 BB strains from China, Philippines, and Japan. Japonica Zha-Chang-Long (ZCL) was resistant to all BB strains tested (Chen et al 1990). Adominant gene controlled the resistance of ZCL to BB strain PXO61 and was nonallelic to Xa1, Xa2, Xa3, and Xa14 (Xie et al 1990). To identify this resistance gene, we mapped it based on restriction fragment length polymorphism (RFLP) markers by using bulked extremes and susceptible class analysis (Michelmore et al 1991, Zhang et a1 1994). The mapping population was the F2 of the cross between ZCL and ZhenZhui-Ai which is an indica susceptible to all BB strains. Equal amounts of DNA of 30 extremely resistant F 2 plants were mixed to make a resistant bulk. Equal amounts of DNA of 42 extremely susceptible plants (lesion length was longer than 9 cm) were mixed to make a susceptible bulk. The DNA samples of two parents and two bulks were digested with six enzymes (Bam HI, Dra I, Eco RI, Eco RV, Hind III, and Xba I) and were assayed for RFLP with 99 clone fragments provided by S. D. Tanksley and T. Sasaki. Of 99 probes, 35 (35.4%) were polymorphic between the combined parents tested and covered about 855 cM of the rice genome, which was about 58% of Cornell's RFLP linkage map length (Causse et al 1994). The ability to verify polymorphism using genomic clones was higher (44.1 %) than that using cDNA clones (15.6%). Twenty-eight polymorphic probes had intensity bands similar to those of their parents in resistant bulk and susceptible bulk.

a 1 = no. of plants with resistant parent band. 2 = no. of plants with two parent bands. 3 = no. of plants with susceptible parent band. b Recombination frequencies between markers and the gene for resistance to BB.

figure). The locus of this resistance gene was different from the loci of resistance genes previously mapped on this chromosome. A new dominant resistance gene to BB exists in Yunnan rice variety ZCL.

Cited references
Causse MA, Fulton TM, Cho YG, Ahn SN, Chunwongse J, Wu K, Xiao J, Yu Z, Ronald PC, Harrington SE, Second G, McCouch SR, Tanksley D. 1994. Saturated molecular map of the rice genome based on an interspecific backcross population. Genetics 138: 1251 1274. Chen Y, Liao XH, Dao SX, Xie YF, Zhang DP, Yu GX, Dai LY. 1990. Studies on resistance of Yunnan rice germplasms to bacterial blight. In: Zhu LH, editor. Advances in researches on resistance to diseases in major crops. Nanjing: Jiangsu Science-Technology Publishing House. p 21-30. Michelmore RW, Paran I, Kesseli RV. 1991. Identification of markers linked to diseaseresistance gene by bulked segregation analysis: a rapid method to detect markers in specific genomic regions by using segregation populations. Proc. Natl. Acad. Sci. USA 88:9828-9832. Xie YF, Zhang DP, Yu GX, Luo LJ, Zheng KP, Dai LY, Chen Y. 1990. Genetic studies of resistance to bacterial blight in fifteen Yunnan local varieties. In: Zhu LH, editor. Advances in researches on resistance to diseases in major crops. Nanjing: Jiangsu Science-Technology Publishing House. p 14-20. Zhang QF, Shen BZ, Dai XK, Mei MH, SaghaiMaroof MA, Li ZB. 1994. Using bulked extremes and recessive class to map genes for photoperiod-sensitive genic male sterility inrice. Proc. Natl. Acad. Sci. USA91:86758679.

The locus of the gene for resistance to bacterial blight in Yunnan rice variety Zha-Chang-Long on RFLP linkage 11 of rice.

Seven polymorphic probes were mainly susceptible parent bands in susceptible bulk. The resistance gene of ZCL was linked to the chromosome region of these seven probes on chromosome 11. Distances between the resistance gene of ZCL and the probes were determined (see table). The resistance gene of ZCL was located on the top of chromosome 11 (see

30

IRRN 21:1 (April 1996)

Pest resistanceinsects
Influence of rice plant morphology on leaffolder incidence
Z. Islam and A. N. M. R. Karim, Entomology Division, Bangladesh Rice Research Institute (BRRI), Gazipur 1701, Bangladesh
Rice plant morphological characters and incidence of folded leaves caused by leaffolders. BRRI, Gazipur, Bangladesh, 1993 aus.a Variety TKM6 lR32419-81-2-33 BR42419-44-2-32 S818B-10-2 BR4909-R1-R2 BR21 lR7156-J20-3-2-1-1 lR35366-40-3-32-2 lR31805-20-1-3-3 BR20 RP1442-2-2-3-5-1 BR1 lR42015-83-3-2-2 BR4490-B-69 B532B-PN-1-MS-1-KP-1 Purbachi Mean P < (ANOVA) cv %
a Data are av of 8 replications.

Plant height (cm) 147 90 97 91 107 103 104 93 96 112 87 82 78 108 90 98 99 0.01 3.9

Green leaves hill -1 (no.) 42 18 19 32 20 11 26 25 21 22 18 19 22 28 31 15 23 0.01 -

Leaf blade (cm) Length 56.1 34.4 32.4 32.2 36.1 30.2 33.6 32.6 32.7 37.2 30.5 33.1 31.3 36.9 29.0 28.6 34.2 0.01 9.7 Width 1.04 1.33 1.28 1.34 1.53 1.33 1.50 1.32 1.28 1.52 1.31 1.25 1.29 1.54 1.39 1.67 1.37 0.01 10.2

Folded leaves (no. hill-1) 0.38 0.46 0.58 0.63 0.71 0.75 0.92 0.96 1.00 1.00 1.21 1.25 1.29 1.46 1.54 1.88 1.01 0.01 (%) 0.9 2.9 3.0 2.2 3.9 7.4 4.2 3.9 5.7 4.8 6.9 7.1 6.4 5.1 6.0 14.1 5.3 0.01 -

Rice crops in Bangladesh are attacked by three rice leaffolder species: Cnaphalocrocis medinalis (Guene), Marasmia patnalis Bradley, and M. exigua (Butler) (Pyralidae: Lepidoptera). Larvae fold rice leaves lengthwise and feed on the green tissue, sometimes causing severe damage to the plant. We evaluated whether plant morphological characters influence leaffolder incidence. Sixteen rice varieties and breeding lines were tested in the 1993 aus (summer) season under irrigated condition at the BRRI experimental farm in Gazipur. The experiment used a randomized complete block design with eight replications. Fourweek-old seedlings of each entry were transplanted in 1-m -2 plot at 20-cm 2 hill spacing. Standard fertilizer and water management and other cultural operations were used. From booting to dough stages, three hills from each plot were selected at random for data collection. Border hills were not used. Plant height, number of total and folded leaves, and length and width of three flag leaves were determined.

Test entries differed in plant height, number of green leaves hill-1, and leaf blade length and width (see table). On average, leaffolders had folded 5.3% (0.9-14.1%) of the leaves. The entries also differed in rate of the incidence of folded leaves. Leaffolder-resistant variety TKM6, with narrow and long leaves, had the lowest level of folded leaves (0.9%), while Purbachi, with wide and short leaves, had the highest level (14.1%). There was no correlation between

folded leaves and plant height (r = -0.336) or green leaf number (r = -.0195). However, the incidence of folded leaves was negatively correlated with leaf blade length (r = -0.507, P<0.05) and positively correlated with leaf blade width (r= 0.551, P<0.05). Results indicated that narrow and long leaves may offer some resistance against leaffolders compared with wide and short leaves in rice plants.

A virulent rice gall midge biotype in Manipur

M. P. Singh, Entomology Department, College of Agriculture, Central Agricultural University, Imphal, Manipur, India

during the 1991-94 wet seasons. Thirtyday-old seedlings of each entry were planted in 3-m rows at 15- 15-cm spacing.

Local recommended agronomic practices were followed. Gall midge incidence was recorded at 50 d after transplanting.

Reaction of differential rice varieties to gall midge (GM). Iroisemba, Manipur, India. 1991-94. GM incidence (%) Group Differential 1991 Hills Tillers I II III W1263 ARC6605 Phalguna ARC5984 CR-MR1523 Velluthacheera Aganni Ptb10 T1477 TN1 0 0 70 60 65 25 45 30 45 80 0 0 14.5 11.9 13.5 7.5 8.2 8.1 10.8 22.7 1992 Hills 0 0 80 35 50 20 35 40 30 65 Tillers 0 0 19.0 11.7 15.2 8.2 14.8 10.2 10.9 17.5 1993 Hills Tillers 0 0 80 95 60 65 60 85 60 95 0 0 28.7 21.0 22.0 14.2 25.9 16.1 17.0 27.0 1994 Hills 0 0 50 70 35 10 15 35 55 60 Tillers 0 0 22.0 18.0 10.6 2.9 2.9 6.5 12.7 15.8 Mean Hills 0 0 70 65 52.5 30 38.8 47.5 47.5 75 Tillers 0 0 21.1 15.7 15.3 8.2 13.0 10.2 12.9 20.8

Five biotypes of rice gall midge (GM) Orseolia oryzae Wood-Mason have been identified in India. The GM population of Manipur was earlier described as biotype 3 of India, which conformed to the Ranchi GM biotype. However, a new pattern of reaction to differential rice varieties has been observed since 1991. We evaluated a standard set of 10 differentials in four groups against the GM population of Iroisemba, Manipur, India

IV

IRRN 21:1 (April 1996)

31

Only W1263 andARC6605 of group I showed resistant (R) reaction. Differentials of the other three groups showed a consistent susceptible (S) reaction (see table), thus depicting a new R-S-S-S type

response, which had not been previously reported in India. In supplementary studies, the Iroisemba population did not mate with any of the biotypes. We therefore propose

that this be considered a distinct biotype, 3M, that is different from the biotype 3 prevalent in Bihar (Ranchi) and Andhra Pradesh (Jagtiyal).

Pest resistanceother pests

Selection for weed competitiveness in upland rice
M. P. Jones, D. Johnson, B. Fofana, and T. Koupeur, West Africa Rice Development Association (WARDA), Bouak, Cte d'lvoire

Weed infestation is a major problem in upland and hydromorphic rice ecosystems in West Africa. Weeding is generally by hand, but in many cases effective weed control is impossible due to a shortage of labor. One way to reduce weed control inputs (hand weeding or herbicides) is to use weed-competitive cultivars to suppress and/or tolerate weeds. We conducted two experiments in 1994 to identify such cultivars. In the first experiment, 16 promising cultivars selected in 1993 for their ability to suppress or tolerate weeds were evaluated under high-, medium-, and low-input management at Mbe in Cte dIvoire. The experiment was laid out using a split-plot design with three replications and management levels as the main plot and varieties as the subplot. Two Oryza glaberrima varieties, CG14 and CG20, gave the most tillers and leaf area under all levels of management. They had rapid vegetative growth and their plots had low weed biomass at 18 and 40 d after sowing (DAS) (see figure). Tall O. sativa varieties with high tillering ability such as WAB96-1-1 and SP4, suppressed weed growth. A third group, WAB181-18, WAB56-50, and WAB56-125, with high tillering ability, large leaf area, and intermediate stature (90-100 cm), tolerated weeds during the ripening stage. These cultivars had superior yields and were selected for detailed studies on weed competitiveness in 1995. In a second experiment, 12 O. sativa and O. glaberrima lines, previously selected to form a range of distinctly different plant types for plant height, tillering capacity, and

Relationship between tiller number and weed dry weight (wdw) under low (a, b) and medium (c) Inputs.
a1 = CG20, 2 = CGN, 3 = WAB56-50, 4 = WAB181-18, 5 = WAB56-104, 6 = WAB56-125, 7 = WAB96-1-1, 8 =

IRAT144, 9 = WABC165, 10 = SP4, 11 = WAB99-1-1, 12 = IDSA10, 13 = ITAS257, 14 = IAC164, 15 = IRAT112, and 16 = CNA4136.

leaftype, were grown under two levels of weed control. The experiment was laid out in a factorial randomized complete block design with three replications and a plot size of 4 2.6 m.

Weed biomass at rice maturity differed among the varieties. There were interactions (P<0.003) between weeding regime and cultivar for grain yield. Regression analysis across varieties and replicates showed that

32

IRRN 21:1 (April 1996)

grain yield from weedy plots, expressed as a proportion of yield from the clean-weeded plots, was negatively cor-related with weed weight at rice maturity (r = 0.62, n=36) and positively correlated with rice root length at 49 DAS (0.79), and tiller number (0.75) and leaf area (0.73) at 36 DAS. These results are further evidence of the association between crop development during its vegetative phase and competitiveness with weeds. More studies will be conducted to confirm the factors that determine differences among eleven cultivars in weed competitiveness. Promising cultivars will be used in a breeding program and further evaluated for yield potential.

Stress tolerancedrought
Genetic variability of osmotic adjustment under drought stress in rice
R. Chandra Babu, and M. S. Pathan, Plant Molecular Genetics Laboratory (PMGL), Plant and Soil Science Department, Texas Technical University, Lubbock, Texas 79409, USA; J. C. O'Toole, The Rockefeller Foundation, Regional Office, Bangkok, Thailand; A. Blum, The Volcani Center, ARO, Bet Dagan, Israel; and H. T. Nguyen, PGML, PSSD, Texas Technical University

Drought is the major abiotic stress limiting rice yield in the rainfed lowland and upland ecosystems. Little progress has been made in incorporating tolerance for drought into rice. Understanding the physiological mechanism and their genetic basis will be helpful in developing selection strategies for improving drought tolerance in rice. Osmostic adjustment as a process of active solute accumulation under drought stress is receiving increasing attention as an important component of drought tolerance in crop plants (Blum 1989, Ludlow and Muchoe 1990). Osmotic adjustment results from the accumulation of solutes within cells, which lowers the osmotic potential and helps maintain turgor of both shoots and roots as plants experience water deficit stress. By using osmotic adjustment as a trait in dryland crop breeding programs, prospects are good for increasing potential yield and stabilizing yield under drought, as demonstrated in wheat (Morgan and

Condon 1986) and sorghum (Ludlow et al 1990). The objective of our study was to identify the extent of genetic variation for osmotic adjustment in cultivated rice grown under drought stress. Twelve cultivars representing a range of germplasm from traditional dryland types to improved rainfed lowland types were studied under greenhouse conditions. Plants were grown in large PVC containers using a potting soil mixture. Water was withheld from 50 d after sowing to induce drought stress. Changes in leaf osmotic potential and relative water content were measured every other day from initiation of stress until the plants wilted. The entire drought period was about 35 d. Osmotic adjustment was calculated following the method described by Morgan (1992). Results indicated large genetic variation for osmotic adjustment among the cultivars. The cultivar IR62266 expressed higher osmotic adjustment (1.76 MPa) than the rest. The cultivars Azucena, CT9993, Moroberekan, and IR52561 showed lower osmotic adjustment with values of 0.86, 0.81, 0.80, and 0.50 MPa, respectively. Despite the positive influence of osmotic adjustment as a drought tolerance mechanism, the adoption of this trait in breeding programs for crop improvement has been slow. Among other reasons, the inability to rapidly and precisely screen large breeding populations for osmotic adjustment limits progress toward using this trait as a selection criterion. Screening for osmotic adjustment currently requires time-consuming complex measurement procedures. Identifying molecular markers (such as restriction fragment length polymorphism) linked to osmotic adjustment may provide plant breeders with a new tool for selecting cultivars with higher osmotic adjustment capacity. The present study showed large genetic variation for osmotic adjustment that offers a good possibility for identifying suitable molecular markers for this trait in rice. Some of the genotypes tested in the study are parents of recombinant inbred and doubled-haploid line mapping populations that are being developed in collaboration with IRRI scientists for tagging genes associated with drought-adaptive traits.

Blum A. 1989. Osmotic adjustment and growth of barley genotypes under drought stress. Crop Sci. 29:230-233. Ludlow M, Muchow RC. 1990. A critical evaluation of the traits for improving crop yields in water-limited environments. Adv. Agron. 43:107-153. Ludlow MM, Santamaria FJ, Fukai S. 1990. Contribution of osmotic adjustment to grain yield of Sorghum bicolor (L.) Moench under water-limited conditions. II. Water stress after anthesis. Aust. J.Agric. Res. 41:67-78. Morgan JM. 1992. Osmotic components and properties associated with genotypic differences in osmoregulation in wheat. Aust. J. Plant Physiol. 19:67-76. Morgan JM, Condon AG. 1986. Water use, grain yield, and osmoregulation in wheat. Aust. J. Plant Physiol. 13:523-532.

Cited references

Stress tolerance excess water

Characterization of the pyruvate decarboxylase gene family of rice and its potential application to submergence tolerance
M. A. Hossain and E. Huq, Botany and Plant Pathology Department, Purdue University, West Lafayette, IN 47907, USA; A. Grover, Commonwealth Scientific and Industrial Research Organization (CSIRO) Division of Plant Industry, GPO Box 1600, Canberra, ACT 2601, Australia; Rue-Chih Su, Botany and Plant Pathology Department, Purdue University, USA; and E. S. Dennis, and W. J. Peacock, CSIRO Division of Plant Industry, Australia; T. K. Hodges, Botany and Plant Pathology Department, Purdue University, USA.

Flash flooding is one of the major causes of mass destruction of rice in many countries in Southeast Asia. Flooding produces anaerobic conditions because gases diffuse more slowly in water than in air. Under anoxia, alcoholic fermentation occurs at rapid rates during submergence. In maize, the two key enzymes, pyruvate decarboxylase (PDC) and alcohol dehydrogenase (ADH), show higher activities under anoxia (Baily-Serres et al 1988). Only minimal levels of ADH are required for ethanol production in maize

IRRN 21:1 (April 1996)

33

Transgenic rice plants with introduced pdc construct confirmed by PCR and Southern analysis. Plasmids used were pdc 29:35S promoter- adh intron-sense pdc cDNAl-nos; pdc 32:35S promoter- adh intron-antisense pdc cDNAl-nos; pdc 38:actin promoter-sense pdc cDNAl-nos; pdc 40:actin promoter-antisense pdc cDNAl-nos; pdc 42:6xARE promoter-adh intron-sense pdc cDNAl-nos; and pdc 44:6XARE promoter- adh intron-antisense pdc cDNAl-nos.
Genomic organization of three rice pdc genes, the cDNAs, and their expression in response to aerobic and anaerobic conditions. +++ = high induction, ++ = moderate anaerobic induction. UTR = untranslated region.

of transgenic plants, segregation of the transgenes in the progeny, and their response under submerged conditions are under investigation.

Genotype transformed IR54

Plasmids pdc pdc pdc pdc pdc pdc pdc pdc 38 40 29 42 42 44 32 38

Positive plants (no.) 6 2 20 4 8 15 2 4

Fertility

root tips (Roberts et al 1989). This permits survival under hypoxic (2-4% oxygen) conditions, suggesting a key regulatory role for PDC under anoxia. Furthermore, in maize roots under anoxia, the rate of adenosine triphosphate (ATP) synthesis and cytoplasmic pH control are more important in determining survival than are the actual levels of ATP or the energy charge (Xia et al 1995). It would therefore appear that a further enhancement in the activity of PDC should increase the production of ethanol and increase the rate of ATP synthesis under anoxia. We are cloning pdc and adh genes from rice. Three genomic clones, pdc1, pdc2, and pdc3 (Hossain et al 1994a), and two cDNA clones that correspond to pdc1 (Hossain et al 1994b) and pdc2 (Huq et al 1995) genomic clones, respectively, of the pdc gene family have been isolated, sequenced, and characterized from rice (IR54) (see figure). Comparison of the deduced amino acid sequences of the three rice pdc genes showed that pdc1 and pdc2 are 88% similar and 78% identica1, pdc1 and pdc3 are 89% similar and 79% identical, and pdc2 and pdc3 are 89% similar and 79% identical. The maize pdc (Kelley 1989) showed similarity and identity values of 95% and 90% to rice pdc1, 88% and 79% to rice pdc2, and 90% and 81 % to rice pdc3. At least four bands were observed from a Southem blot of IR54 genomic DNA when probed with a DNA fragment from the pdc coding region. Southern blots with probes specific to the 5' and 3' untranslated regions of differential bands of these cDNAs showed differential bands corresponding to these genes.

Radon

Fertile Fertile Fertile 3 fertile/1sterile 8 sterile 3 fertile/12 sterile Fertile 1 fertile/3 sterile

Northern blotting experiments showed that the pdc1 and pdc2 are inducible under anaerobic conditions; pdc3 did not give any detectable band on a Northern blot. Unlike pdcl and pdc2 genes, pdc3 does not have any introns and is shorter; it is probably a pseudogene. The pdc1 promoter has multiple copies of anaerobic responsivelike elements as in the maize adh gene. Deletion analysis of the pdcl promoter driving the gusA reporter gene was performed to evaluate the anaerobic responsive elements in transformed rice protoplasts. Following incubation under aerobic and anaerobic conditions, transient GUS activity assays indicated the presence of both positive and negative regulatory sequences. Six plasmid constructs have been made using the rice actin promoter, CaMV35S promoter, and 6XARE (anaerobic responsive element) promoter driving the pdc cDNA1 gene in both the sense and the antisense orientations. All of these plasmids have been transformed into rice cells by the biolistic or PEG uptake method (see table). For certain constructs, although many hygromycin-resistant calli were selected andregenerated small plantlets, they failed to grow to maturity. Some of the IR54 transgenic plants did have normal morphological appearance (height, tiller number, seed set) compared with seedgrown plants, but most of the Radon transgenic plants were sterile or produced only a few seeds. Further characterization

Baily-Serres J, Kloeckener-Gruissem B, Freeling M. 1988. Genetic and molecular approaches to the study of the anaerobic response and tissue-specific gene expression in maize. Plant Cell Environ. 11:351-355. Hossain MA, Huq E, Hodges TK. 1994b. Sequence of cDNA from Oryza sativa L. encoding pyruvate decarboxylase 1 gene, Plant Physiol. 106:799-800. Hossain MA, McGee JD, Grover A, Dennis ES, Peacock WJ, Hodges TK. 1994a. Nucleotide sequence of a rice genomic pyruvate decarboxylase gene that lacks introns: a pseudo-gene? Plant Physiol. 106:1697-1698. Huq E, Hossain MA, Hodges TK. 1995 Cloning and sequencing of cDNA encoding pyruvate decarboxylase 2, gene from rice. Plant Physiol. (in press) Kelley PM. 1989. Maize pyruvate decarboxylase mRNA is induced anaerobically. Plant Mol. Biol. 13:213222. Roberts JKM, Chang K, Webster C, Callis C, Walbot V. 1989. Dependence of ethanol fermentation, cytoplasmic pH regulation, and viability on the activity of alcohol dehydrogenase in hypoxic maize root tips. Plant Physiol. 89:1275-1278. Xia, J-H, Saglio P, Roberts JKM. 1995. Nucleotide levels do not critically determine survival of maize root tips acclimated to a low-oxygen environment.

Cited references

Plant Physiol. 108:589-595.

Molecular implication of submergence tolerance in rice using expressed sequence tags as probes
M. Umeda, C. Hara, and H. Uchimiya, Institute of Molecular and Cellular Biosciences, University of Tokyo, Yayoi 1-1-1, Bunkyo-ku, Tokyo 113, Japan

Large-scale sequencing of randomly selected cDNA clones was done to investigate feasibility of a method for isolating plant genes. Total RNA used for cDNA synthesis was prepared from suspension cultured cells of rice grown

34

IRRN 21:1 (April 1996)

Schematic representation of the relative amount of mRNA under submerged condition. Numerals represent the relative amounts of transcript as a percentage of the untreated samples: (1) phosphoglucoisomerase, (2) phosphofructokinase, (3) aldolase, (4) triose phosphate isomerase (5) triose phosphate dehydrogenase, (6) phosphoglyceryl kinase, (7) enolase, (8) pyruvic kinase, (9) dehydrogenase, (13) succinyl/CoA synthetase, (14) succinic dehydrogenase, (iron-sulfur protein subunit), (15) succinic dehydrogenase (flavoprotein subunit). Pyruvic decarboxylase, (10) alcohol dehydrogenase, (11) Dihydrolipoyl transacetylase (12) isocitric

under osmotic (6% or 20% sucrose), saline (2% NaCl), or N-starvation stresses. More than 2,000 cDNA clones were partially sequenced and compared with the GenBank and EMBL data bases. This allowed about 10% of the cDNA clones to be putatively identified as particular genes. These results indicate that stress treatment of suspension cultured cells makes it possible to effi-

ciently isolate various types of plant genes. To examine the usefulness of expressed sequence tags (ESTs) for the study of gene expression in a specific metabolic pathway, we analyzed transcript levels of genes engaged in ATP-generating pathways under submergence stress. We quantified each transcript level of genes associated with glycolysis and alcoholic fermentation

several times under submergence stress. We found two types of induction patterns: Type I and Type II. Maximum expression of Type I genes occurred after 24 h of submergence. Transfer to aerobic condition or partial exposure of shoot tips to air reduced expression. In the submergence-tolerant rice cultivar FR13A, several Type I genes were highly induced compared with intolerant cultivar IR42. This suggests that Type I genes may play an essential role in the activation of glycolysis and alcohol fermentation under submerged condition. Transcripts of Type II genes, such as aldolase and pyruvate kinase genes, reached the maximum after 10 h and did not show remarkable decrease by transfer to aerobic condition. These results suggest that expression of Type I and Type II genes is differentially regulated under low oxygen condition. It would be worthwhile to analyze mechanisms associated with coordinate or differential expression of the genes. This will lead to improved understanding in molecular breeding of means to manipulate a whole metabolic pathway in rice plants.

Stress toleranceadverse temperature

Mapping of genes responsible for cold tolerance at the booting stage of rice
A. Kato, K. Saito, K. Nagano, K. Miura, and H. Araki, Hokkaido National Agricultural Experiment Station, Hitsujigaoka 1, Toyohiraku, Sapporo 062, Japan

Kirara 397) was 0-95% in the field in 1993. The temperature at booting stage was 2-3 C lower than that in the average year. We analyzed the restriction fragment length polymorphism (RFLP) patterns and fertility of 47 selected B1F 4 plants. Fertility of the plants showing Silewah-type patterns for the loci of chromosome 3 or 4 was about 10% higher than for those plants with Hok-

kai 241-type patterns (see table). The results were almost the same as those obtained in a cold water experiment in which plants in a field were irrigated with cold water for about 1 mo, from the early stages of panicle formation to heading (Saito et al 1995), confirming that the loci on chromosomes 3 and 4 are involved in cold tolerance.

Norin PL 8 is a cold-tolerant parental rice line. It was developed by introducing the genes for cold tolerance from a javanica rice, Silewah, into a Japanese cultivar, Hokkai 241. The chromosome segments introgressed into Norin PL 8 from Silewah have been mapped on chromosomes 1, 3, 4, 7, and 8, and the loci on chromosomes 3 and 4 are probably involved in cold tolerance (Saito et al 1995). The fertility of B1F4 plants of Norin PL 8 and Kirara 397 (Kirara 397/Norin PL 8//

Genotype and fertility of B 1F4 plantsa of Kirara 397/Norin-PL8/ /Kirara 397 in 1993. RFLP marker XNpb100 XNpb345 XNpb102 XNpb235 XNpb177 XNpb379 XNpb278 Chromosome 3 3 4 4 4 7 8 Fertility (%) Kirara-type 50.4 50.5 50.6 52.5 52.5 51.2 58.7 Heterozygote 59.0 56.9 68.4 65.1 65.1 46.0 Silewah-type 63.8 63.8 66.1 66.8 66.8 66.8 57.9

a Plants (n = 47) were selected from 500 plants; different fertility classes are nearly equally represented. The temperature during July and August 1993 was 2-3C lower than that for the average year. The fertility of Kirara 397 was 40%.

IRRN 21:1 (April 1996)

35

Rice parental lines Hokkai PL 5. Hokkai PL 6, and Hokkai PL 7 are also cold-tolerant lines developed using the same procedures, except that the genes from javanica rice Lambayque 1, Mitak, and Thangone were introduced into Hokkai 244, the japonica parent. RFLP patterns of these parental lines indicated that the genes from javanica parents are located at least on chromosomes 1, 5, 10, and 11 in Hokkai PL 5, on chromosomes 8 and 10 in Hokkai PL 6, and on chromosomes 4, 5, and 7 in Hokkai PL 7. On the other hand, the RFLP markers for the loci on chromosomes 3 and 4 where the genes involved in cold tolerance in Norin PL 8 are located showed Hokkai 244- type patterns, suggesting that the genes involved in cold tolerance in Hokkai PL 5. Hokkai PL 6, and Hokkai PL 7 are different from those of Norin PL 8.

Cited reference

Saito K, Miura K, Nagano K, Hayano-Saito Y, Siato A, Araki H, Kato A.1995. Chromosome location of quantitative trait loci for cool tolerance at the booting stage in rice variety 'Norin-PL8'. Breed Sci. 45:337-340.

1. Fine-striped leaf (top) and normal leaf (bottom).

Variation in leaf color and thylakoid protein analyzed in chlorophyll rice mutant
C. Aoki, Y. Imai, A. Furuta, T. Nishimura, and K. Hattori, Laboratory of Plant Genetics and Breeding, School of Agricultural Sciences, Nagoya University, Chikusa, Nagoya 464-01, Japan

Some strains of chlorophyll-deficient mutants albina, viridis, chlorona, virescence, and zebra, obtained by -ray irradiation of rice, are preserved in our laboratory. We studied the modified leaf color and thylakoid protein of a low temperature-sensitive, chlorophylldeficient rice mutant. The mutant was derived from the Oryza sativa cultivar Nipponbare, which was irradiated by -ray at the total dose of 40 Gy (0.8 d -1). Seeds of the mutant and normal Nipponbare were sprouted at 30 C. At the 3d-leaf stage, seedlings were transferred to a phytotron in which they were exposed to controlled day-night temperatures of 30-24, 25-19, and 24-14 C

2. Polypeptides of thylakoid protein from the leaf blade of doubled haploid Nipponbare (C) and mutant (M) grown at 30-24 C (H), 25-19 C (M), and 20-14 C (L) in a phytotron. In the lanes of the mutant, closed arrows point to missing or faint bands in fine-striped and white leaves, arrowheads point to missing or faint bands in white leaves only, and open arrows point to added or intensified bands compared with patterns of green leaves.

36

IRRN 21:1 (April 1996)

under natural light. Thylakoid protein was analyzed according to Aoki et al (1995). The mutant plants produced expanded green leaves under 30-24 C, fine-striped leaves under 25-19 C, and white leaves under 20-14 C (Fig. 1). After a shift in temperature the mutants often had expanded nonhomogeneously colored leaves. In some plants transferred from 30 to 20-14 C, the upper parts of the leaf blades were green and the basal parts were white. However, in some plants transferred from 20-14 to 30-24 C, the upper leaf parts were white and the basal parts were green. Leaf color of the mutant became faint with low temperature. In the polypeptide patterns of the thylakoid protein in the finestriped and white leaves, some bands were missing or faint while others were added or intensified when compared with those in normal green leaves (Fig. 2). The plastids in the leaf tip were observed to mature earlier than those in the basal region (Robertson and Laetsch 1974). Therefore, it seems that low temperature only affects the plastids of the mutant in the developmental stage. It is possible that a light-harvesting protein was lost in finestriped and white leaves because about 23kDa bands were lacking. The mutant might be the result of a few nuclear gene mutations having been obtained by chronic d -ray irradiation. We need to analyze the genes of the mutant.

Cold tolerance at booting stage of highly cold-tolerant rice lines derived from a javanica and a japonica
K. Nagano and H. Araki, Faculty of Agriculture, Hokkaido National Agricultural Experiment Station, Hituzigaoka 1, Toyohira, Sapporo 062, Japan

Norin PL 8 is a cold-tolerant rice variety derived from a javanica (Silewah) and a japonica (Hokkai 241). Highly coldtolerant lines have been selected from crosses between different japonica rice cultivars at Tohoku (Fig. 1). We compared cold tolerance at the booting stage of these lines (see table) through cold water treatment. The experiment was conducted in both 4-liter pots and in the ricefield. In the pot experiment, 20 seeds/pot for Norin PL8 were seeded on
1. Pedigrees of rice varieties used in the experiment.

16 May and six other lines were seeded in a box on 25 Apr. Cold water (18.8 C) treatment began on 13 Jul and ended at heading date for the lines. Seedlings that were raised in a box were transplanted on 26 May, with cold water treatment (18.8 C) beginning on 27 Jun and ending 1 Sep. Water depth was 35 cm for both. In the field experiment, Norin PL8 was seeded on 4 May and 13 May and two other lines were seeded on 27 Mar and 8 Apr. Water depth was 15 cm, and water temperature was 19.0 C. Both 90C2013 and Chubo 59 showed seed fertilities higher than that of Norin PL 8 (Fig. 2). Results suggest that we can select highly cold-tolerant lines equal to that of javanica rice only from japonica rice. Accumulation of cold tolerance genes between javanica and japonica rices is possible.

Aoki C, Wada T, Nishimura T, Hattori K. 1995. Characterization and inheritance of variegated-leaf mutant in Petunia hybrida. Breed. Sci. 45:31-35. Robertson D, Laetsch WM. 1974. Structure and function of developing barley plastids. Plant Physiol. 54:148-159.

Cited references

2. Seed fertility of highly cold-tolerant lines treated with cold water in the pot experiment (18.8 C) (a) and in the field experiment (19.0 C) (b).

IRRN 21:1 (April 1996)

37

Stress toleranceadverse soils

Variation in salt tolerance of rice plants regenerated from salt-selected calli of a susceptible variety
B. B. Bong, Cuu Long Delta Rice Research Institute, Omon, Cantho, Vietnam; S. Tobita and T. Senboku, Japan International Center for Agricultural Sciences, Okinawa Subtropical Station, lshigaki 907, Japan
Efficiency in selection of salt-tolerant calli and its effect on plant regeneration. Description Calli cultured in salt medium (no.) Salt-tolerant calli selected (no.) Salt-tolerant calli transferred to regeneration medium (no.) Calli producing green shoots (no.) Calli producing green shoots out of total calli cultured in salt medium (%) Total R0, regenerated plants Number or Percentage 4050 214 (5.3%) 214 48 (22.4%) 1.2 120

The variety Mot Bui is popular in deepwater areas of the Cuu Long Delta of Vietnam. It is, however, salt-susceptible. We attempted to induce somaclonal variation for salt tolerance from this variety. The mature seeds were cultured on MS medium (supplemented with 2 mg 2,4-D L-1, 30 g sucrose L-1 and 8 g agar L-1) for callus induction. After three subcultures, the embryoderived calli were transferred to the same medium containing 1.5% NaCl. After 1 mo, the calli that survived and grew normally under salt stress were transferred

to the regeneration medium (MS medium supplemented with 2 mg kinetin L-1 0.5 mg , NAA L-1, 30 g sucrose L-1, and 8 g agar L-1). Once rooted, the regenerated plants (R0 generation) were grown in a nutrient solution; upon reaching 15 cm, they were transferred to the nutrient solution supplemented with NaCl to get a salt level expressed by EC of 12 dS m-1. The survival rate of the regenerated plants was recorded after 2 wk. The check varieties included Pokkali (tolerant), IR28 (susceptible), and the

parent variety Mot Bui. Another set of the regenerated plants was grown under normal conditions to obtain R1 somaclones, and their tolerance for salt was evaluated following the procedures described above. Out of 4,050 calli subjected to in vitro salt selection, most turned brown and died, but 214 showed normal growth. Of these calli, 48 produced green shoots when cultured on the regeneration medium (see table). The percentage of calli producing shoots to total number of calli cultured on the salt medium was 1.2. We screened regenerated plants (R0 generation) and found that 35% of them survived after 2 wk under salt stress, while all of the plants of Mot Bui and IR28 died. In the next generation, 34 R1 somaclones were evaluated for salt tolerance at the seedling stage (9 d old). Twenty-eight somaclones (23.5%) were salt-tolerant. This study suggests that variation for salt tolerance exists among lines derived from salttolerant calli of a salt-susceptible rice variety.

Genetics of salinity tolerance and ionic uptake in rice

B. Mishra, Crop Improvement Division, Central Soil Salinity Research Institute, Karnal 132001 (Haryana), India; M. Akbar, National Agricultural Research Center, Islamabad, Pakistan; D. V. Seshu and D. Senadhira, IRRI

Increasing salinity of soil and water is a serious threat to agriculture. Breeding for salt tolerance offers a more promising, energy-efficient, economical, and socially acceptable approach to solving these problems than do major engineering processes and soil amelioration. Salt-tolerant rice as a biological amendment for managing salt-affected soils has been demonstrated (Mishra 1994). Improving salt tolerance requires systematic genetic studies. Reliability of reproductive phase salinity tolerance score with grain yield was strongly correlated under both

controlled and field conditions (Mishra 1994). Grain yield was also strongly correlated with K+ uptake and Na-K ratio in the shoot. Based on these relationships, we investigated the genetics of salinity tolerance at the reproductive stage and K+ and Na-K ratio in the shoot in a 6 6 halfdiallel cross involving tolerant (CSR10, CSRl, and Nona Bokra) and susceptible (IR28, M1-48, and Basmati 370) parents in an artificially created saline soil environment (ECe = 10-14 dS m-1) at IRRI. Salinity stress was developed with NaCl + CaCl2 in a 1:1 ratio by weight and maintained by continuous recycling. Na and K concentrations in the shoot were analyzed on an oven dry weight basis. Data were analyzed by both the Hayman (1954) and Griffing (1956) methods. Analysis of variance showed highly significant genetic variations among the parents and F 1s for salinity tolerance score at the reproductive phase, K concentration, and Na-K ratio. The reduction in grain yield was 28.9% in tolerant variety CSR10

while susceptible variety IR28 showed 98.2% yield reduction. K+ absorption was maximum in CSR1 (2.31%) followed by CSR10 (2.27%); it was least in IR28 (2.06%). Low Na-K ratio resulted in better salinity tolerance. The lowest Na-K ratio was in CSR10 (tolerant) and highest in M1-48 (susceptible), thus revealing the significance of K+ uptake in salt tolerance. Estimates of genetic parameters following Haymans method showed significant additive and nonadditive gene action for salinity tolerance score, K+ and Na-K ratio (Table 1). Narrow-sense heritability estimates were 0.65 for salinity tolerance score, 0.55 for K+ absorption, and 0.41 for Na-K ratio. The presence of gene asymmetry for salinity tolerance score and Na-K ratio and gene symmetry for K + was detected. Average dominance was within the range of incomplete dominance. This suggested that one group of genes is involved in salinity tolerance score and NaK ratio and two groups in K+ uptake. Recessive alleles were more concentrated

38

IRRN 21:1 (April 1996)

Table 1. Estimates of genetic parameters for reproductive stage salinity tolerance score and K concentration and Na-K ratio in shoots of 6 6 diallel. ~~ ~ Genetic parameter D H1 H2 h2 F E Proportional values (H1/D) H2/4H1 [(4DH1) + F]/ [(4DH1) - F] r (Wr + Vr)/Yr h2/H2 h2 (narrow sense)
a

Estimate SE Salinity tolerance score 10.760* a 7.532* 5.536* 3.299 6.663* 0.290 0.837 0.184 2.175 0.96 0.596 0.65 1.080 2.916 2.645 1.786 2.698 0.441 0.056* 0.067* 0.056* 0.075* 0.004 0.012* 1.091 0.210 1.062 0.60 1.326 0.55 K 0.009 0.024 0.022 0.015 0.023 0.004 0.099* 0.129* 0.092* 0.030 0.096* 0.006 1.139 0.178 2.484 0.971 0.324 0.410 Na-K 0.017 0.044 0.039 0.027 0.043 0.007

Table 2. Combining ability analysis for salinity tolerance score at reproductive stage and K concentration and Na-K ratio in shoots of 6 6 diallel. Source df Mean square Salinity tolerance score at reproductive stage 15.36**a 3.71** 0.22 4.14 K Na-K ratio

gca 5 sca 15 Error 40 gca/sca

0.238** 0.065** 0.012 3.66

0.100** 0.033** 0.006 3.03

a** = significant at the 1% level.

Cited references
Griffing B. 1956. Concept of general and specific combining ability in relation to diallel crossing systems. Aust. J. Biol. Sci. 9:463-493. Hayman B. 1954. The theory and analysis of diallel crosses. Genetics 39:789-809. Mishra B. 1991. Combining ability and heterosis for yield and yield components related to reproductive-stage salinity and sodicity tolerance in rice Oryza sativa L. In: Rice genetics II. Manila, (Philippines): International Rice Research Institute. p 761. Mishra B. 1994. Breeding for salt tolerance in crops. Pages 226-259 In: Rao et al, editors. Salinity management for sustainable agriculture25 years of research at CSSRI. Karnal, (India): Central Soil Salinity Research Institute.

* = significant at the 5% level.

than dominant genes in the tolerant varieties CSR10 and CSR1 for salinity tolerance score at the reproductive phase and K + uptake while dominant alleles were more concentrated than recessive in Nona Bokra, CSR10, and CSRl in that order. Wr-Vr graphic analysis suggested the involvement of both major and minor genes for all the traits investigated. Combining ability analysis by Griffing's method confirmed the significance of both additive and nonadditive effects, with additive effects showing greater importance in the inheritance of the traits (Table 2). The

varieties CSR10 and CSRl were good general combiners for all three traits examined and could be useful in rice breeding programs aiming to improve salt tolerance. High heterotic effects found in some crosses suggest the potential of hybrid rice for salt-affected soils. CSR10 is now being used as a donor in an India-IRRI collaborative project on salinity tolerance and also as a low-cost biological amendment for managing salt-affected soils in India.

Stress tolerance
Epigenetic control of tolerance for transient low light stress in rice
Y. F. Chen, Institute of Agrobiological Genetics and Physiology, Jiangsu Academy of Agricultural Sciences, Nanjing 210014, China; C. G. Li, Institute of Food Crops, Jiangsu Academy of Agricultural Sciences, Nanjing 210014, China

Rice has smaller photosynthetic capacity and less accumulation of assimilates resulting in a low yieldunder low light stress than does rice grown under adequate solar radiation. For example, rice yield during the wet season is 65.2% of that

during the dry season (Murty and Sahu, 1987), and rice that receives half of the full sunshine has 57% less yield than rice receiving full sunshine (Murty 1992). Tolerance for low light stress is genetically controlled, but until recently little has been known about the genes involved. To explore whether cytoplasmic genes are involved in controlling this physiological character, we used several sets of reciprocal cross combinations in which double parents show typical tolerance for or sensitivity to transient low light stress. Detached flag leaves at the heading stage were inserted into tap water and equilibrated under near saturated light intensity (500 E m-2s-1). For a different set of leaves, light intensity was rapidly decreased to 110 E m-2s-1 and leaves were

incubated for 1 min under low irradiation. At the same time, net photosynthesis rates (Pn, mol CO2 m2s-1) in two light environments (expressed as Pn 500 and Pn 110) were determined with a LI-6200 photosynthesis system. To indicate the degree of Pn decrease immediately after irradiation, we used (Pn 500-Pn 110)/Pn 500. The threshold was determined to be half of the decrease of Pn(max) at 110 E m-2s-1. The leaves were tolerant when their Pn changes were below the threshold or sensitive when Pn changes were over the threshold under transient light stress. Data in the table were averages of 3-6 repetitions. Eight sets of combination among the 11 parents (see table) were classified into three kinds according to the degree of Pn reduction of the parents: 1) both parents have a

IRRN 21:1 (April 1996) 39

Changes in net photosynthetic rate (Pn) in flag leaves of reciprocal-cross F 1 and their parents under transient low light impact. Combination number (Pn 500-Pn 110)/Pn 500 Parent 1 Parent 2 Parent 1/Parent 2 1 LH422 (I) a 0.716 0.1446 YS8072 (J)
b

F1 Parent 2/Parent 1

Pal (I) 0.518 0.0339 Xiushui 04 (J) 0.435 0.0641 02428 (J) 0.482 0.0283 Minghui 75 (I) 0.356 0.0601 Zhongguo 91 (J) 0.361 0.1277 029 (J) 0.448 0.1916 Luweidao (I) 0.702 0.0693 IET9702 (I) 0.477 0.1209 0.446 0.1044 0.793 0.1392 1.09 0.153 0.693 0.0127 0.527 0.0636 0.621 0.0099 0.473 0.0233 0.702 0.1524 0.396 0.0820 0.891 0.0707 0.430Li 0.0848 0.776 0.1237 0.396 0.1266 0.747 0.1741 0.326 0.0874

0.625 0.0304 Sunong 30367 (I) 0.834 0.0767 LH422 (I) 0.716 0.1446

Luweidao (I) 0.702 0.0693

LH422 (I) 0.716 0.1446 YS8072 (J) 0.625 0.0304 02428 (J) 0.482 0.0283

small reduction in Pn, such as combination 8; 2) Pn reduction in both parents is large (such as combination 7); and 3) one parent has a large reduction in Pn and the other has a small reduction (combination 1). In the third kind of combination, performances of Pn reduction in the reciprocal cross F1 are basically identical to those in corresponding maternal lines under low irradiation impact. Results in the other combinations also support this. It can therefore be concluded that genes related to tolerance for transient low light stress in rice are mainly located in the cytoplasm. However, in some situations nuclear genes from both maternal and paternal lines interact with cytoplasmic genes from the maternal line to make tolerance stronger (in combination 1) or weaker (in combination 7). The large and positive correlations between Pn and accumulation of dry matter will make this genetic behavior of shared tolerance in rice useful in breeding for high yield.

0.439 0.0523

l = indica subspecies. bJ =japonica subspecies.

Stress toleranceother structures

Screening somaclonal variants with tolerance for both shade and photooxidation in rice
Y. F. Chen and L. H. Sun, Institute of Agrobiological Genetics and Physiology, Jiangsu Academy of Agricultural Sciences, Nanjing 210014, China

Rice plants must have increased adaptation to light stresses, including shade and photooxidation, before higher photosynthesis and yield can be achieved. Somaclonal variation is a new source of tolerance for shade and photooxidation. We have obtained some somaclonal variants with tolerance for both shade and photooxidation (double tolerance) in rice. Calli were induced on solid N6 medium with 2 mg 2,4-D L-1 liter from mature

caryopses of three varieties, 02428. Zhongguo 91, and 842 to both shade and photooxidation, and then subcultured on the same medium. Calli were transferred on to solid MS medium with 2 mg BA L-1 + 0.25 mg NAAL-1 + 0.25 mg IAA L-1 to regenerate plants (R1). Posterities of R 1 were sexually reproduced. Somaclonal variants from different parents were identified under shade and photooxidative conditions. Rice plants were grown from booting to grain tilling (3 wk) under two layers of light gray plastic nets, which allowed about 30% natural sunlight to penetrate. Under shade, specific leaf weight declined, with less than 20% of the varieties being classified as tolerant. The upper 2d leaves were gathered at booting stage and submerged in tap water containing low CO2 and low O2. They were

Differential responses of rice somaclonal variants under shading and photooxidative conditions. Somaclonal variant Grade of photooxidation a I II I I I V IV III II IV II IV II Decline of SLW (%) b 4.2 13.4 5.3 21.3 14.7 25.9 19.2 -c 31.7 24.9 18.6

02428 (parent) 02428hong (R 5) 02428h (R5) Zhongguo 91 (parent) CX-1 (R 4) 842 (parent) 842241-1 (R 2) 842241-2 (R 2) 842houshi (R 2) 84202-chou 1 (R 2) 84202-chou 7 (R2) 84202-chou 9 (R 2) 84202-chou 10 (R 2)
a

I = whole leaf area green, II = only leaf tip yellow, III = 1/3 of leaf area yellow, IV = 1/2 of leaf area yellow, V = 3/4 of leaf area or whole leaf area yellow. Upper 2nd leaves from the main stems were used, replicated three times. bSLW = specific leaf weight. The uppermost leaves from the main stems were used, except CX-1. c = not determined.

40

IRRN 21:1 (April 1996)

then incubated under 1400 Em-2 s -1 at 35 C for 6 d. Treated leaves were divided into five grades based on the degree of green lost, with grades I and II being considered tolerant (see table). Somaclonal variants have great differences in their responses to shading and photooxidative stresses (see table). Somaclonal variants with double tolerance

were obtained for both shade- and photooxidation-tolerant parent 02428 (such as 02428h and 02428hong), from photooxidation-tolerant but shade-sensitive parent Zhongguo 91 (such as CX-1), and from sensitive parent 842 to both shade and photooxidation-tolerant (such as 84202-chou 10). The variants with double sensitivity may be particularly useful for improving culti-

vars with good agronomic characters but sensitivity to light stresses. Frequency of variants with double sensitivity or with either tolerance for shade or photooxidation from double-sensitive parents is higher, however, than that for variants with double tolerance.

Biochemical identification and genetic analysis of molybdenum cofactor mutants in rice

H. Sato, Y. Imiya, Faculty of Agriculture, Kagawa University, Miki-cho, Kagawa, 76107, Japan; S. Ida, Research Institute for Food Science, Kyoto University, Uji, Kyoto 611, Japan; and M. Ichii, Faculty of Agriculture, Kagawa University

The molybdenum cofactor (MoCo) prosthetic group is essential for the activity of molybdo enzymes, such as nitrate reductase (NR), xanthine dehydrogenase (XDH), and

sulfite oxidase. However, very little information is available on the MoCo biosynthetic pathway in rice. We report the biochemical identification and genetic analysis of MoCo-deficient mutants in rice to provide genetic information on this pathway. Four MoCo-deficient mutants (cnx mutants), C25, C27, C32, and C33. were isolated as chlorate-resistant mutants from 100,000 M2 seedlings. Biochemical analysis showed that NADH-, NADPHNR, XDH activities. and MoCo biosynthesis abilities (MoCo activities) were deficient in these four mutants. indicating they were MoCo-deficient types (Table 1).

Table 1. Characteristics of the four cnx mutants and wild type IR30. Line IR30 C25 C27 C32 C33 NADH-NR activity Control 870.2 0.0 170.0 207.6 194.6
a

+0.5 mM Na 2 MoO4 432.4 0.0 465.4 467.7 517.8 ( 49) c ( 0) (273) (225) (266)

NADPH-NR activitya 11.3 0.0 0.3 1.1 0.0

MoCo activityb 3.8 0.0 1.4 1.6 0.4

XDH

Chlorate resistant Se R R R R

Tungstate sensitive Re R R R R
very

+d

a n mol NO - min -1 g fresh weight -1. b n mol NO - min -1 mg protein -1 . c Percentage of control, c + = XDH activity; = 2 2 e low XDH activity:

- = no XDH activity.

S = sensitive, R = resistant.

Table 2. Segregation of chlorate resistance in F 2 populations between cnx mutants and wild type IR30 or between cnx mutants. Cross combination Segregation in F2 plants a c 3:1 Chloratesensitive 83 82 83 90 102 0 0 Chlorateresistant 37 34 30 19 85 198 168 Total plants (no.) 120 116 113 109 187 198 168
2

value 9:7

Probability

C25/IR30 C27/IR30 C32/IR30 C33/IR30 C25/C27 C27/C32 C33/C27

2.117 1.149 0.144 3.330 0.220 No segregation No segregation

0.1 0.2 0.5 0.05 0.5

< < < <

P P P P

< < < <

0.2 0.3 0.7 0.1

< P < 0.7

a The F seeds were sown in 0.1 mM KC10 solution at 30 C. Fourteen days after sowing, chlorate resistance was 2 3 evaluated visually from the reduction in seedling height and the extent of brown spots on the leaves.

The cnx mutants could be divided into two groups: one that could recover NR activity by adding molybdate in growth medium and one that could not. To classify our cnx mutants, their 7-d-old seedlings were treated with Kimuras B solution containing 0.5 mM Na2 MoO4 for 2 d. Molybdate treatment did not recover NADH-NR activity in C25 but did so in C27, C32, and C33; the activity was double that of the control (Table I). We also investigated the tungstate sensitivity in the mutants because it has been reported that a tungstatesensitive group exists among Arabidopsis thaliana cnx mutants. We hydroponically cultured the mutants and their wild type IR30 with culture medium containing 0.1, 1, or 10 mM Na2 MoO4 for 7 d. In all treatments, no significant differences occurred between the mutants and IR30 in seedling height and root length. We therefore conclude that the four cnx mutants do not belong to the tungstate-sensitive group (Table I). Derived from the four cross combinations between cnx mutants and wild type IR30, the chlorate resistance in F 2 populations had a segregation of 3:1 wild type: mutant phenotype, indicating a single recessive gene controls each mutation (Table 2). The allelism test, using the three F 2 populations derived from the crosses between the four cnx mutants, indicated that the three mutant genes of C27, C32, and C33 are located at the same locus, and that the genes of C25 and these three mutants are located at different loci (Table 2). This result coincides with the phenotypic classification based on the restoration of NADH-NR activity by molybdate as mentioned above. Thus, we conclude that at least two loci are involved in the MoCo biosynthetic pathway in rice.

IRRN 21:1 (April 1996)

41

Integrated germplasm improvement

Altering associations between characters in rice through gamma radiation
L. M. Gonzlez, Nuclear Laboratory of the Agricultural Research Institute, Jorge Dimitrov, Gaveta Postal 2360, Bayamo 85100, Granma, Cuba
Total correlations between characters in control and irradiated progenies in the M 2 generation. a Characters Control Height Panicles plant -1 Panicle length Panicle weight Grains panicle -1 1,000-grain weight Height Panicles plant -1 Panicle length Panicle weight Grains panicle -1 1,000-grain weight Height Panicles plant -1 Panicle length Panicle weight Grains panicle -1 1,000-grain weight Height Panicles plant -1 Panicle length Panicle weight Grains panicle -1 1,000-grain weight
a

Yield plant-1

1,000-grain weight -0.228 -0.828** -0.145 -0.005 -0.005 0.269 0.042 0.100 0.167 0.314 0.017 0.053 0.354 0.572* 0.608* 0.208 0.487 0.675* 0.643* 0.528*

Grains panicle -1 0.142 0.209 0.789** 0.991*** 15 Krad 0.751** 0.289 0.767** 0.971*** 25 Krad 0.683* 0.455 0.605* 0.992*** 35 Krad 0.625* 0.228 0.497 0.966***

Panicle weight 0.119 0.155 0.754**

Panicle length -0.126 0.355

Panicle plant -1 0.291

We studied the usefulness of gamma radiation in altering character association in rice (Oryza sativa L.). Dry seeds of cultivar J112 were treated with gamma rays (60Co) with 15, 25, and 35 Krad doses. We conducted variability studies in the M2 and M 3 generations. The M3 progenies were grown along with parental varieties in an experiment that was laid out in a randomized block design during the 1993-94 dry season. Of 30 random plants, correlation coefficients were calculated for plant height, panicles plant-1, panicle length, panicle weight, grains panicle-1, 1000-grain weight, and yield plant-1. The treatments with gamma rays increased the correlation between panicle length and yield plant-1, panicle width, and grains panicle-1, and between panicle weight and grains panicle-1, and plant height and plant weight (see table). This increased correlation among traits can be used to increase the rate of selection response for a primary trait. Based on these results, the yield can be improved if selection is made for panicle length, panicle weight, and grains panicle-1.

0.311 0.699* 0.624* 0.685* 0.703** 0.532* 0.222 0.675* 0.778** 0.847** 0.850** 0.166 -0.045 0.609* 0.896** 0.812** 0.800** 0.360 0.434 0.626* 0.804** 0.809** 0.849** 0.518*

0.654* 0.378 0.836**

0.480 0.618*

0.367

0.622* 0.458 0.629

0.039 0.191

0.221

0.660* 0.218 0.683*

0.061 0.057

0.300

* and ** = significant at the 0.05 and 0.01% level, respectively.

control, but significant and negative in various treatments. Such character combinations included plant height and panicle weight and plant height and grains panicle -1 for all doses studied. The change in the direction of the correlation coeficients indicates the changes in relationships between characters.

Any change in the correlation (positive or negative) of one character with another indicates a radiation-induced effect. Changes in the relationships observed in this study coincide with other results and might be attributed to linkage effect, gene mutation, or change polygenic systems. For some of the character combinations, an association was nonsignificant in the

Integrated germplasm improvementirrigated

Liangyou Peite, a new two-line hybrid rice released in China
Bai Delang and Luo Xiaohe, Hunan Hybrid Rice Research Center (HHRRC), Mapoling, Changsha, Hunan 410125, China

Two-line hybrid rice has many advantages over three-line hybrid rice. Liangyou Peite (Pei Ai 64S/Teqing) was successfully bred over 10 yr at HHRRC. It was registered in I994 with the Hunan Varieties Evaluation Committee and suggested for national re-

lease as one of the first two-line hybrid rice combinations. It shows high yield potential, good grain quality, and resistance to multiple diseases and insects. It had wide adaptability trials in South China and Hunan Province in 1991-93 (Tables 1 and 2). In large-scale (3,330 ha) demonstrations in Hunan in 1991-94, Liangyou Peite yielded more than 7.5 t ha-1 in the late season with a maximum yield of 10.4 t ha -1. It yielded 9.0 t ha-1 as a middle season crop, with a maximum yield of 11.3 t ha -1. It

yielded a record of 17.1 t ha-1 in Yongsheng County, Yunan. Liangyou Peite had about 10% (0.811 t ha -1) yield advantage over its three-line hybrid rice counterparts. Moreover, it has strong ratooning ability with an average yield of 2.3-3 t ha-1 as a ratoon rice over the past 4 yr. In 1995, about 6,660 ha in China were planted to this hybrid. The female parent, Pei Ai 64 S, is the first practical thermosensitive genic male sterile line in China. It is an indica/javanica that has high combining ability and is compatible

42

IRRN 21:1 (April 1996)

Table 1. Performance of Liangyou Peite in South China and Hunan provincial and regional trials. South China Character 1991 Mid 23 sites, 10 provinces 143.0 1.5 6.2a 1.6b 99.0 478.5 288 157.8 118.6 75.2 22.8 101 1992 Late 12 sites, 8 provinces 125.9 3.4 6.7 8.9c 317 132.6 100.8 76.0 23.1 93 1992 Mid 7 sites 136.7 3.0 9.5 4.7bd 145.5 559.5 333 140.1 126.2 87.9 23.4 98 Hunan 1993 Mid 6 sites 135.8 5.5 7.7 5.8b 136.5 603.0 333 144.5 122.3 84.6 22.3 106

Table 2. Grain quality of Liangyou Peite. Grain length (mm) Grain width (mm) Length-width ratio Chalky grain (%) Chalky area (%) Brown rice (%) Milled rice (%) Head rice (%) Gelatinization temperature (C) Gel consistency (mm) Amylose content (%) Protein content (%) Eating quality Remark on quality 5.28 2.35 2.30 80.0 5.3 83.8 75.9 60.0 4.8 30 22.4 10.3 Good Fine

Growth duration (d) Growth duration (d) over check Grain yield (t ha -1) Grain yield (t ha -1) over check Basic tillers (no. m-2 ) Maximum tillers (no. m-2 ) Productive panicles (no. m-2 ) Spikelets (no. panicle -1) Filled grains (no. panicle -1) Seed set (%) 1,000-grain weight (g) Plant height (cm)

aBrown rice yield. b Check is Shanyou 63 (three-line rice hybrid). c Check is Shanyou Gui 99 (three-line rice hybrid). dSignificant at the 1% level.

for making indica and japonica hybrid rices. The male parent Teqing is a high-yielding inbred indica variety. The seed yield of Pei Ai 64 S averaged more than 3.0 t ha-1 with a maximum of . 5.5 t ha-1 F 1 hybrid seed yield was 2.3-3.0 t ha-1 in Hunan in 1995.

Khushboo, a quality rice cultivar for Rajasthan

K. B. Agrawal, Agriculture Research Station (ARS), Ummedganj Kota, Rajasthan, India

BK-805-36, a cross derived from Baran Basmati/Pusa 150, has been released as Khushboo in Rajasthan, India. It was identified in the F 6 generation at ARS, Kota. Khushboo is a semidwarf cultivar that matures in 118-123 d when transplanted. It is suitable for either normal or late sowing and possesses ideal plant type with upright leaves, compact tillering, late leafsenescence, and horizontal flag leaf position. Panicles are long (28 cm), compact, and fully exserted. Spikelets are partially awned and straw colored. Kernels are 7.5 mm with a length-breadth ratio of 4.7.
Table 1. Performance of Khushboo in varietal trials. Kota, Rajasthan, India. 1986-91. Year Yield (t ha -1) Khushboo 3.4 3.5 2.9 2.7 3.5 4.0 3.3 23.9 Basrnati 370 (check) 2.6 2.8 2.1 2.3 3.0 3.4 2.7 CD (0.05) 0.4 0.5 0.3 0.5 0.3 0.4

Kernel length after cooking is 13.4 mm with an elongation ratio of 1.8. The grains are long and slender, white, translucent, and strongly scented. Cooked rice is soft and well separated. In trials conducted at Kota from 1986 to 1991, Khushboo yielded consistently higher than Basmati 370 (Table 1). In coordinated trials conducted during 1989 at 11 locations and in 1990 at 7 locations, Kushboo yielded 14.7 and 9.17% more than Basmati 370, respectively. In on-farm demonstrations conducted. Khushboo yielded 41% in 1991 and 54% higher than Basmati 370 (Table 2). The cultivar is resistant to leaf blast and sheath rot, and moderately resistant to neck blast, brown spot, and white-backed plantopper.

Bright-pearl 1, a fast grain-filling japonica rice

Zhang Xiaoming and Zhen Rou, Crops Research Institute, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China; Jing Guoqiang, Zhejiang Yuhang Seed Company, Hangzhou 311100, China

Table 2. Khushboo grain yield in on-farm demonstrations. Kota, Rajasthan, India. 1991-92. Yield (t ha -1) Year Khushboo Basmati 370 (check) 3.5 2.5

1986 1987 1988 1989 1990 1991 Average Percent over Basmati 370

1991 (4 locations) Percent over Basrnati 370 1992 (7 locations) Percent over Basmati 370

4.9 41 3.9 54

All of the traditional late-japonica rice cultivars grown in the Yangtze River area of China require 45-50 d from heading to ripening. Sometimes the grain-filling period is longer because of unfavorable weather, resulting in low rice yield and delayed sowing of the next crop. Local farmers normally choose early-ripening rice cultivars. But the plants usually have insufficient growing time. With a longer vegetative stage and afast grain-filling rate, the late-japonica rice can be harvested in late October instead of early November, thus avoiding inclement weather and reducing moldy or sprouted seed. Using a fast grain-filling cultivar can increase rice quality. The ability to supply the market early means a better price for growers. Bright-pearl 1, derived from Bin 8103/ 713 in 1992, is a fast grain-filling japonica rice. Its grain filling is 10-15 d faster than that of check Xiushui 11. Bin 8103 is an early ripening/japonica cultivar resistant to brown planthopper (BPH) and blast. 713 is

IRRN 21:1 (April 1996)

43

Table 1. The major agronomic characters of Bright-pearl 1. Site Hangzhou, Zhejiang Sowing Planting (month/day) method a 7/8 7/15 6/16 7/10 7/25 7/10 6/20 7/10 6/24 6/30 5/27 5/20 6/30 TP BS DS DS DS BS TP TP TP TP TP TP DS 80% Growth Plant heading duration height (month/day) (d) (cm) 9/18 9/19 8/24 9/14 9/24 9/11 9/3 9/15 9/2 9/15 8/25 8/25 9/5 112 107 106 96 92 103 115 107 106 120 136 133 105 85.2 82.5 89.7 85.9 81.5 87.5 90.2 84.6 87.0 92.7 98.0 95.4 99.5 Grains Seed panicle -1 set (no.) (%) 69.9 72.8 62.6 65.1 64.2 77.2 76.5 78.9 71.6 74.6 78.0 71.4 80.5 88.6 83.2 92.3 86.7 84.7 85.6 89.5 84.3 85.1 91.2 89.7 93.4 94.3 Yield (t ha-1) 7.64 6.95 7.25 6.81 6.43 6.48 8.21 7.82 7.13 7.50 7.13 6.69 9.99

Jiaxing, Zhejiang Ningbo, Zhejiang Tanzhou, Zhejiang Dangtu, Aihui Nanlin, Aihui Jinshan, Shanghai Nantong, Jiangsu Xingfeng. Jiangxi

a TP = transplanting, BS = broadcast seeding, DS = direct sowing.

a late japonica cultivar that resists bacterial blight (BB) and blast. Bright-pearl 1 is a photoperiod-sensitive japonica rice with blast, BB, and BPH resistance. Its growth duration is about 135 d for a single crop and about 100 d for late season direct sowing in Zhejiang, Aihui, and Shanghai. The major agronomic characters were recorded (Table 1). Bright-pearl 1 has ideal plant type, with erect leaves, suitable stem angles, a strong root system, and tolerance for lodging. It also has good rice quality (Table 2). Bright-pearl 1 can be used in japonica rice breeding and can be planted by farmers.

Table 2. Milled rice properties of Bright-pearl 1. Variety Length (mm) Width (mm) Shape L/W Chalkiness (%) 1000grain weight (g) 26.2 27.5 Alkali spreading value 6.7 7.0 Amylose content Unbroken milled rice (%) 75.9 64.7

Bright-pearl 1 Xiushui 11 (check)

4.9 5.3

2.9 3.0

1.7 1.7

0.5 7.0

16.3 15.6

Crop and resource management

Physiology and plant nutrition
Growth characteristics of IRRIdeveloped new rice plant type breeding lines in Japan
S. Akita, Faculty of Agriculture (FA), University of Tokyo (UT), Yayoi, Bunkyo, Tokyo 113, Japan; T. Ishikawa, National Agriculture Research Center (NARC), Tsukuba, lbaraki 305, Japan; and H. Fujisawa, FA, UT

IRRI recently developed breeding lines of a new rice plant type (NPT) that promises increased yield potential in the tropics. These lines are tropical japonicas that have fewer but larger panicles than current modern varieties. They are known as panicle weight types. The physiological basis for the increased yield of the lines, however, has not been elucidated. We conducted field experiments at NARC to examine the factors related to the

yield potential of these NPT lines. Two NPT lines provided by IRRI and four cultivars (Table 1) were transplanted in 1994 during mid-May and mid-June. Planting density was 26.7 hills m-2 and 180 kg N ha-1 was applied (80 kg as basal, 60 kg at panicle initiation, and 40 kg during mid-reproductive phase). Standard management practices were followed. The dry weight of each organ and the leaf area were monitored at three growth stages for 15 plants. Yield and its components were observed in three replicated plots. The average solar radiation and average temperature during the growing season were higher than normal, contributing to good growth. Damage from disease and insect pests was minimal. Necrosis of the axis of the panicle and rachis in IR66740-ACI-3 plants was observed during the late ripening phase. The NPT lines grew slowly initially, but the rate during the reproductive phase was faster than that of the other cultivars. Dry weight at heading was almost the same for

the test materials, although it was sometimes higher for the NPT lines. However, the biological yield of the NPT lines was lower than that of most of the cultivars because the increase in their dry weight after heading was lower. Panicle number of the NPT lines was low, and IR65598-112-2 was the lowest. The spikelet number panicle-1 of IR65598-112-2 was the highest among the materials tested (Table 1). Leaves of both NPT lines were erect during the ripening phase, which ensured better canopy structure. No lodging was observed in NPT lines although most of the other cultivars from both plantings lodged during the mid-ripening phase (Table 1). Sink size was calculated by multiplying filled grain weight by spikelet number. No significant difference was found between the sink size of the NPT lines and most of the cultivars, although their economic yield was the lowest due to a lower grain-filling percentage (Table 2). The reduced percentage of filled grains was attributed to the grain sink being

44

IRRN 21:1 (April 1996)

Table 1. Growth characteristics of new plant type lines compared with those of high-yielding cultivars. Yawara, Japan. 1994. Cultivar/line May transplanting lR65598-112-2 IR66740-AC1-3 Kochihibiki Takanari IR36 IR72 June transplanting lR65598-112-2 IR66740-AC1-3 Kochihibiki Takanari IR36 IR72 Days to headinga (no.) 89 92 77 82 85 88 76 85 69 73 77 77 Biological yieldb (g m -2) 1737 1605 1539 2006 1800 1753 1494 1625 1575 1869 1654 1698 SD c Specific leaf area d (cm2 g -1) 185 180 217 206 232 212 159 171 208 208 228 182 Leaf area indexd 7.2 9.0 6.5 7.0 6.6 8.2 4.6 7.5 6.7 7.3 6.2 6.0 Panicle number (no. m -1 ) 144 317 499 281 580 501 118 217 399 239 445 395 Spikelet number (no. panicle -1) 367 131 70 182 88 109 326 171 83 200 101 122 LB/TDW e NSCf content (%) 39.1 31.8 10.5 10.8 6.8 4.7 13.8 34.9 4.4 8.9 5.1 8.7 ADWrg (g m -2) Lodging scoreh

146 344 104 65 60 137 2 139 54 25 35 122

0.33 0.37 0.30 0.28 0.27 0.32 0.29 0.36 0.32 0.30 0.26 0.29

392 150 474 786 645 458 444 305 505 619 504 498

0 0 3.2 0 4 3.5 0 0 3.3 0.5 4 4

aCounted from transplanting. b All data are shown on a dry weight basis. c SD= standard deviation. d Measured at heading stage. e Partitioning ratio of dry weight to leaf blade. fNonstructural carbohydrate content in culm at maturity. g Dry weight increase during ripening phase. h Scored on a 0-4 scale where 0 = no lodging and 4 = hardly lodged.

Table 2. Economic yield and yield components of new plant type lines. Tsukuba, Japan. 1994. Cultivar/line May transplanting lR65598-112-2 IR66740-AC1-3 Kochihibiki Takanari IR36 IR72 June transplanting lR65598-112-2 IR66740-AC1-3 Kochihibiki Takanari IR36 IR72 Spikelet number (no. m -2 ) 528 414 347 511 513 547 385 372 330 479 451 482 Grain filla (%) 44.9 37.2 71.1 88.5 80 62 49.7 36.6 81 83.4 84.5 72.4 Grain weightb (g m -2 ) 20.9 23.3 23.1 21.5 19.8 20.0 22.3 24.1 24.1 20.9 19.7 20.8 Sink sizeb,c (g m -2 ) 1104 965 802 1099 1016 952 859 897 795 1001 888 1003 Economic yield b (g m -2) 495 359 570 972 813 590 427 328 644 835 751 726 SD d Harvest indexe

48 78 16 19 15 52 7 28 61 28 16 41

0.25 0.20 0.32 0.42 0.39 0.34 0.25 0.18 0.36 0.39 0.39 0.37

yield/biological yield.

aPercentage of grains that sunk in water total spikelet -1. b Shown on a brown rice basis at 15% moisture content. The multiplication of grain or kernel welght and spikelet number m -2. d SD = standard devlation. e Harvest index = brown rice

inactivated along with the wider variation in the weight of fertilized kernels and increased nonstructural carbohydrate content in the culm at maturity (Table 1). Another factor was less dry weight increase during ripening, characterized by a higher dry weight at heading due to the larger partitioning of dry weight to leaves, especially in IR66740-ACI-3. These results indicate that panicle weight type rice may not always be highyielding unless accompanied by an effective grain sink. Less inactivation of the grain sink of NPT lines would ensure a yield level similar to that of IR72. For further increase in potential yield, a higher biological yieldsimilar to that of Takanariwill be essential.

Fertilizer management organic sources

Breaking seed dormancy in different leguminous forage species
W. S. Dhillon and U. Singh, International Fertilizer Development Center-IRRI Collaborative Project, IRRI

Clitoria (Clitoria ternatea L.), crotalaria (Crotalaria juncea L.), siratro (Macroptilium atripurpureum [DC] Urb.), and

desmanthus (Desmanthus virgatus (L.) Willd) have potential as leguminous green manures for rainfed lowland fields after the main rice crop in the Philippines. But their thick, hard seed coverings can cause problems for germination. To break seed dormancy, we subjected the seeds to a treatment of H2SO4 (commercial grade, 96%) for 15 min and to scrubbing for 15 min with sand, using mild force. The experiment was replicated three times. Seed germination was 98.0% for clitoria and 95.3% for siratro with the H2SO4 treatment, which was considerably higher than sand scrubbing and for the control

(Table 1). In crotalaria, seed germination was 99.3% without any treatment, 97.3% with sand, and 89.3% with H2SO4. Desmanthus did not respond well to any of the treatments; the highest germination was 17.3% with sand scrubbing. The seeds did not survive the H2SO4 treatment. To improve the germination of desmanthus, we conducted another experiment using various treatments (Table 2). Treating the seeds with H2SO4 for 8 min increased the germination to 87.3%. The seeds must be washed very quickly after treatment to prevent them from being subjected to a temperature of more than

IRRN 21:1 (April 1996)

45

Table 1. Seed germination (% after 192 h) of leguminous species under different treatments. a Treatment H2SO4 (concentrated) for 15 min Scrubbing with sand for 15 min Control Mean
a

Clitoria 98.0 55.3 49.3 67.6 b

Crotalaria 89.3 97.3 99.3 95.3 a

Siratro 95.3 18.0 16.7 43.3 c

Desmanthus 0.0 17.3 16.0 11.1 d

Mean 70.7 a 47.0 45.3 b b

Table 2. Germination of desmanthus seeds under different treatments. a Treatment Dry cold at 2 C for 24 h Wet cold at 2 C for 24 h Boiling water for 15 min Hot water at 65 C for 15 min H2SO4 (concentrated for 8 rnin) Control Germination rate (% after 192 h) 16.7 d 14.0 d 00.0 e 68.0 b 87.3 a 23.3 c

Means followed by the same letter are not significantly different at the 5% level by DMRT.

a Means followed by the same letter are not significantly different at the 5% level by DMRT.

45C. A more practical treatment of soaking the seeds in hot water for 15 min gave seed germination of 68.0%. Dry cold and wet cold treatments were not effective. The boiling water treatment damaged the seeds and resulted in no germination. The hot water (65 C) treatment for 15 min also improved the germination of

siratro, with a germination rate of 39% after 192 h. Clitoria, however, did not respond and germination was only 2%. In conclusion, crotalaria seeds do not require any breaking of dormancy to germinate. Applying concentrated H2SO4 for 15 min was most effective in breaking dormancy in clitoria and siratro. For

desmanthus, a shorter treatment time (8 min) with concentrated H2SO4 is recommended. Where high germination rate is not critical and H 2SO4 availability may be a problem, treating desmanthus and siratro with hot water (65 C) for 15 min is suggested.

Fertilizer management inorganic sources

Responses of promising rice genotypes to nitrogen levels in irrigated lowlands
A. K. Singh and A. Kumar, Agronomy Department, Rajendra Agricultural University, Pusa, Samastipur, Bihar 848125, India

Lowlands in Bihar constitute about 2.2 million ha, which is about 42% of the total rice area. Rice varieties selected for their nutrient requirementsparticularly for Nhave great relevance for boosting overall production of lowland rice in Bihar. The performances of genotypes IET 5760, IET 5914, and IET 8002 were evaluated against national check Pankaj and local checks Radha and Rajshree at five N levels (0, 40, 80, 120, and 160 kg ha-1).

The soil was silt loam in texture, with a pH of 8.6,0.46% organic C, and 213 kg available N ha-1, 8.4 kg available P ha-1, and 73 kg available K ha -1. The experiment was laid out in a split-plot design with the genotypes in the main plot and N in the subplots. Thirty-day-old seedlings ofdifferent genotypes were transplanted during the last week of July. The recommended doses of 17.6 kg P ha -1 and 16.6 kg K ha-1, along with one-fourth of the N according to the

Interaction effect of N application and genotype on grain yield (t ha -1), N uptake (kg ha-1), and net return (US$ ha -1).a Rice genotype IET5760 IET5914 IET8002 Rajshree Radha Pankaj Av 1.06 (88.8) 2.26 (60.6) 3.30 (188.4) 3.39 (195.7) 2.44 (85.5) 2.39 (79.2) 2.47 (86.8) N levels (kg ha-1) 0 [17.4] [38.7] [58.3] [60.6] [40.4] [41.8] [42.9] 40 2.04 (58.1) 3.29 (201.15) 4.42 (347.1) 4.44 (353.5) 3.28 (212.8) 3.39 (220.8) 3.48 (232.2) [41.8] [68.5] [91.3] [93.3] [73.6] [67.5] [72.7] 80 2.73 [60.3] (101.6) 4.06 [88.6] (261.5) 4.97 [108.6] (372.50) 5.04 [112.7] (384.6) 4.17 [93.1] (279.3) 4.44 [86.9] (293.2) 4.23 [91.7] (282.1) Grain yield 0.30 0.31 120 3.04 (130.4) 4.46 (301.7) 5.45 (420.7) 5.12 (384.8) 4.98 (367.90) 5.24 (394.1) 4.72 (333.3) [68.5] [101.1] [121.2] [116.1] [113.2] [114.2] [105.7] 3.08 (125.9) 4.56 (306.2) 5.74 (412.2) 5.09 (371.6) 5.46 (414.3) 5.36 (395) 4.88 (337.5) Net return 1009 1087 160 [70.3] [106.5] [114.8] [114.2] [124.4] [118.6] [108.1] Average 2.39 (65.4) 3.73 (226.2) 4.78 (348.1) 4.62 (338.1) 4.07 (272) 4.17 (276.4) [51.7] [80.7] [98.8] [99.4] [88.9] [85.8]

LSD (0.05)

For subplot treatment at the same level of main plot treatment For main plot treatment at the same level of subplot treatment

N uptake 8.2 8.3

a Values in [ ] indicate N uptake and in ( ), net return. b Main plot treatment is variety, and subplot treatment is N level.

46

IRRN 21:1 (April 1996)

treatment, were applied at transplanting. The remaining three-fourths of the N was applied in three equal splits at tillering, maximum tillering, and panicle initiation. No irrigation was required because rainfall was sufficient from active tillering to crop maturity. Significant differences were observed among genotypes and N levels for yield, N uptake, and net returns (see table). Genotype IET8002 gave significantly higher grain and straw yields than did the other

genotypes. In general, both yield and net return increased up to 120 kg N ha -1 . Interaction effect of N levels and genotypes was found to be significant for yield, N uptake, and net return. At 0, 40, and 80 kg N ha -1 , rice genotypes Rajshree and IET8002 were at par and significantly superior to other genotypes. IET8002 outperformed Rajshree at 120 and 160 kg N ha -1 . None of the genotypes but Radha responded significantly to 160 kg N ha -1 . The grain yield and N uptake of rice genotypes Rajshree and IET

8002 at 0 kg N ha -1 were superior to those of other genotypes at 40 kg N ha -1. A significant increase in net return occurred up to 80 kg N ha-1 for Rajshree and up to 120 kg N ha-1 for IET8002. Although genotype Radha responded significantly to N application up to 160 kg N ha-1 , the yield and net returns obtained at that level were lower than those obtained for IET8002even at 120 kg ha-1 . Rajshree and IET8002 are superior to other genotypes for yield, N use efficiency, and net return.

Nitrogen use efficiency for three fertilizers in irrigated rice

A. A. Jakhrom, Sindh Agriculture University, Tando Jam, Pakistan

We evaluated the N use efficiency for ammonium sulfate, urea, and ammonium nitrate in irrigated rice grown in a saline soil. We conducted a field experiment during the 1993 and 1994 summer seasons at Taluka Garhi Yaseen, Pakistan. The soil was clay loam in texture with a pH of 7.9, 0.09% total N, 6.5 ppm available P, 250.5 ppm exchangeable K, 90.0 ppm Ca, and 35.0 ppm Mg. The recommended fertilizer rate was 120 kg N ha -1 and 33 kg P ha -1 . The soil had adequate exchangeable K, so K was not applied. The experiment was laid out in a complete randomized block design with four replications. Three types of N fertilizers were used; the N level was divided into three equal splits. Before transplanting IR6 seedlings, one-third of the N was broadcast
Effect of source of N fertilizers on flooded rice.a Source of N Rate (kg N and P ha -1 ) 0- 0 120-33 120-33 120-33 -

and incorporated into the soil with a full dose of P as triple superphosphate. The remaining two-thirds was broadcast at tillering and panicle initiation. Whole plant samples were analyzed for N concentration at tillering and heading, as were the straw and grain after harvest. The N concentration at three physiological growth stages and in the grain significantly increased with the use of ammonium sulfate when compared with that from urea, ammonium nitrate, and the control. The low N concentration in the plant and grain from use of urea and ammonium nitrate suggests that with urea, more N is lost as ammonia because of the high soil pH, and with ammonium nitrate, anaerobic reduction causes losses of N2 O and N 2 gases (see table). Use of these two fertilizers therefore decreased grain yield. Nitrogen use efficiency was highest with ammonium sulfate compared with urea and ammonium nitrate because of lower N losses and higher N use for grain formation (see table).

Crop management
Response of rice hybrid PMS2 A/ IR31802 to seedling vigor and nitrogen levels in Haryana, India
Hari Om, Rice Research Station (RRS), Kaul 132021, India; S. K. Katyal, CCS Haryana Agricultural University, Hisar 125004, India; and S. D. Dhiman, RRS

N concentration (%) Tillering 0.80 c 1.90 a 1.60 1.30 9.80 b b Heading Straw Grain

Grain yield (kg ha -1 ) 2500 5100 4000 3500 12.3

N efficiency use GYF-GYC b NS

Control Ammonium sulfate Urea Ammonium nitrate CV %

1.10 c 0.30 c 2.71 a 0.82 a 2.20 2.00 8.76 b b 0.65 0.55 10.50 b b

c 1.50 2.00 a
1.75 1.60 11.51 b b

c
b b

21.67 12.50 8.33

(kg ha -1 ).

a Data are based on means of four replicatlons. Means followed by the same letter are not signiflcantly different at 0.05 level. b GYF = grain yield with fertilizer (kg ha -1 ), GYC = grain yield of control (kg ha -1 ), and NS = N applied

Heterosis in rice increases grain yield by 20% compared with the best pure lines. It is necessary, however, to assess the performance of promising hybrids at graded levels of N. Inadequate N application will not allow maximum grain yield. Excessive N may lead to excessive crop growth, creating favorable conditions for insect pests and diseases. The optimum nursery seeding rate for new hybrids also needs to be determined. We conducted a field experiment during the 1993 and 1994 wet seasons with rice hybrid PMS2 A/IR31802 using five N levels (0, 50, 100, 150, and 200 kg ha -l ) and three seeding rates in the nursery (20, 40, and 60 g m -2 ). The soil of the experimental plots was clay loam with a pH of 8.1, EC 0.26 mmho cm -1 , and 99.6 kg available N ha -l , 26.0 kg available P ha-1 and 390.0 kg available K ha -1. The experiment was laid out in a splitplot design with N levels as the mainplots and seed rates as the subplots with four replications. A uniform application 26.4 kg P ha -1 and 25 kg ZnSO 4 ha -1 was broadcast before transplanting, which was done in the last week of June at a spacing of 20 15 cm with 1 seedling hill -1 . N was topdressed in

IRRN 21:1 (April 1996)

47

three equal splits at transplanting, active tillering, and panicle initiation. Crop growth rate was calculated from dry weight recorded at regular intervals. Seedling vigor index was calculated from observations made 1 day before transplanting where
Vigor index = (shoot length + root length) germination (%)

Effects of N levels and nursery seeding rate on grain yield and harvest index of hybrid rice. Grain yield (t ha-1) Treatment N level (kg ha ) 0 50 100 150 200 LSD (0.05)
-1

Harvest index 1993 0.61 0.60 0.59 0.55 0.54 0.01 0.59 0.58 0.57 0.01 1994 0.59 0.58 0.57 0.54 0.52 0.01 0.57 0.56 0.56 0.01

Sterility (%) 1993 21.6 24.5 26.3 27.3 27.9 2.5 21.6 25.6 26.6 2.3 1994 27.0 29.7 31.0 33.6 33.6 2.7 30.0 31.1 31.6 2.6

1993 4.0 5.7 7.2 8.0 8.3 0.4

1994 3.4 5.4 6.8 7.5 7.7 0.3 6.5 6.2 5.9 0.2

Mean 3.7 5.6 7.0 7.8 8.0

Seedlings obtained from nursery beds with a seeding rate of 20 g m-2 were most vigorous. Their vigor index was 3585 in 1993 and 3269 in 1994, while for seedlings at 40 g m-2 it was 3129 and 2955 and at 60 g m-2, 2756 and 2636, respectively. Grain yield increased significantly as N application rose from 0 to 150 kg ha -1 (see table), although a further increase from 150 to 200 kg N ha-1 did not increase grain yield significantly. Harvest index decreased and spikelet sterility (%) increased progressively with increased N application. The most vigorous seedlings (from 20 g m-2 seeding rate) produced significantly higher grain yield (6.8 t ha-1) than did those from the higher seeding rates. The weakest seedlings produced thelowest yield (6.1 t ha-1). The harvest index was highest and sterility lowest in the most vigorous seedlings. The crop growth rate increased as the N rate was increased from 0 to 200 kg ha-1 (see figure). The highest rate was recorded with 200 kg N-1 ha between 61 and 80 d in 1993 and 41 and 60 d in 1994. After attaining maximum values, the crop growth rate decreased drastically during the final time interval during both years. Seedlings from the 20 g m -2 seeding rate generally had a higher crop growth rate during the times of maximum growth.

Nursery seeding rate (g m -2) 20 7.1 40 6.5 60 6.3 LSD (0.05) 0.4

6.8 6.4 6.1 -

Effects of N levels and nursery seeding rate on crop growth rate (g m -2 d -1).

Integrated pest managementdiseases

Fine and physical mapping toward the positional cloning of an indica-derived blast resistance gene Pi-b
M. Miyamoto, National Institute of Agrobiological Resources (NIAR), Tsukuba, Ibaraki, Japan, and Plant Biotechnology Institute (PBI), Iwama, Ibaraki, Japan; K. Rybka, NlAR and Plant Breeding and Acclimatization Institute, Warsaw, Poland; Y. Tanaka, NlAR and Faculty of Agriculture, Okayama University, Okayama, Japan; M. Kataoka, NIAR; I. Ando, National Agricultural Research Center, Tsukuba, Ibaraki, Japan; and S. Kawasaki, Presto, Research Development Corporation Japan, Saitama, Japan

Rice blast is one of the most devastating diseases of rice, especially in temperate areas. Therefore, abundant genetic data on rice blast resistance genes have been

accumulated in Japan. The indica-derived resistance genes introgressed into japonica backgrounds showed a wide spectrum of resistance and provided excellent materials for mapping the introgressed genes. We took Pi-b as a target for mapping and obtained two cosegregating markers and very close flanking markers on both sides. By using them, a contig of cosmid library clones was constructed around the gene.

48

IRRN 21:1 (April 1996)

Restriction fragment length polymorphism (RFLP) and random amplified polymorphic DNA (RAPD) analysis of near-isogenic lines (NILs) with Pi-b. The blast resistance gene Pi-b was introgressed from two Malaysian and two Indonesian cultivars into a japonica background by repeated backcrosses; several NILs were developed. It was suggested that the gene was on the end of chromosome 2 (Shinoda et al 1971). We studied patterns of RFLP markers around the region and summarized their results as a graphical genotype map (Fig. 1). To obtain markers closer to Pi-b, RAPD markers were sought using 800 polymerase chain reaction primers, and a marker cosegregating with Pi-b in all the NILs was found and named b-1 (Fig. 1). Although Pi-b was introgressed from four cultivars in Malaysia and Indonesia, the nonjaponica type RFLP patterns were identical among the four lines, suggesting a single origin of the gene in all of the cultivars. Fine mapping of Pi-b by F2 analysis. Through F2 analysis of 122 susceptible individuals (recessive homozygous) from the crosses of BL1/Nourin 22 and Aichi Asahi/BLl, the map distances between the Pi-b and nearby RFLP/RAPD markers were determined with an accuracy of 0.4 cM (Fig. 2a). RZ123 and b-1 cosegregated with Pi-b. The centromeric and telomeric flanking markers G1234 and RZ213 were 0.4 cM and 1.9 cM from the gene, respec-

1. Graphical genotype of the indica region introgressed into NILS with Pi-b. b1; RAPD marker, RZ, RG; Cornell University markers. G, cDK; RGP markers.

2. Fine mapping of Pi-b with F2 analyses with 170 susceptible individuals (a), and a contig constructed from the cosmid library (b).

tively. Large-fragment Southern blotting indicated RZ123 and b-1 are on the same fragment of 85 kB. Cosmid library of rice genome. A genomic cosmid library was constructed with an average insert size of 38 kb, with 5 genome equivalents using pWE 15 as a vector. RZ123 and b-1 were shown to be on the same clone. Four steps of walking were

done by the TAIL method (Liu et al 1994) to amplify the end portions of the inserts, and 100 kB was covered (Fig. 2b). Cited references
Liu, Y-G and R Wittier. 1995. Genomics. 25:674-681. Shinoda H et al. 1971. Bull. Chugoku Agric. Exp. Stn. Ser. A20: 1-25 [Jpn/Eng].

Effect of botanicals on managing sheath rot of rice

R. Jagannathan and K. Sivaprakasam, Plant Pathology Department, Agricultural College and Research Institute, Madurai, Tamil Nadu, India

Susceptible cultivar ADT36 was planted in the field during 1994-95. Leaf extracts, carbendazim, and monocrotophos were applied as foliar sprays at booting and 2 wk later. Each treatment was replicated three times.
Effect of botanicals on sheath rot incidence. Treatment Neem oil Neem seed kernel extract Dose (%) 3 5 10 10 10 500 g ha -1 500 ml ha -1

Sheath rot incidence was measured as the percentage of tillers affected in 25 randomly selected hills plot -1 at 20 d after the last spray. All treatments significantly reduced sheath rot compared with the control (see

We evaluated the efficacy of neem derivatives, such as neem oil (3%) and neem seed kernel extract (5%); leaf extracts (10%) from Nochi (Vitex negundo), white babul (Acacia leucocephala) and polyalthia (Polyalthia longifolia ); standard fungicide carbendazim (500 g ha-1); and insecticide monocrotopos (500 ml ha-1) to control sheath rot caused by Sarocladium oryzae (Sawada) W. Gams and D. Hawksw.

Mean sheat rot incidence (%) 8.18 7.79 11.25 8.14 10.83 4.71 14.54 18.58 2.88

Mean grain yield (t ha-1) 5.1 5.7 6.3 6.1 6.1 6.6 5.7 5.1 0.9

White babul Polyalthia Carbendazim Monocrotophos Untreated control CD (P = 0.05)

Nochi

IRRN 21:1 (April 1996)

49

table). Plants treated with neem seed kernel extract, Nochi, white babul, polyalthia, and carbendazim yielded more than the control (see table).

Efficacy of Ipomoea cornea in controlling rice sheath rot

S. Eswaramurthy, Plant Pathology Department (PPD), Agricultural College and Research Institute (ACRI), Tamil Nadu Agricultural University (TNAU), Killikulam, Vallanad 627252, India; V. Mariappan, PPD, ACRI, TNAU, Coimbatore 641003, India; and M. N. Alagianagalingam, K. S. Subramanian, A. Sankaralingam, and E. G. Ebenezer Ponsinghmal Raja, PPD, ACRI, TNAU, Killikulam, Vallanad 627252, India

simultaneously conducted in adjacent fields. In the artificial inoculation trial, the syringe inoculation method was used because of its greater efficacy when compared with other methods. In each plot, 200 unemerged panicles were syringeinoculated with 0.2 ml of spore suspensio n that contained 104 conidia ml -1. The first spraying was carried out 24 h after inoculation. The second spraying was done 15 d later. The plant products used were Ipomoea cornea leaf extract (10%), Prosophis julifera leaf extract (10%), neem seed kernel extract (5%), and neem oil (3%).

In the natural infection trial, no inoculation was done. Fifteen days before harvest, 100 inoculated panicles from the artificial inoculation trial and 100 panicles from the natural infection trial were randomly selected to assess sheath rot incidence. I. cornea extract, P. julifera extract, and Bavistin were highly effective, recording sheath rot incidence of 11.0, 16.3, and 11.7%, respectively, compared with 72.7% in the control under artificial inoculation conditions (see table). The same trend also occurred under natural infection conditions.

Efficacy of plant products and chemicals in controlling rice sheath rot.a Artificially inoculated plants Treatment Neem seed kernel extract, 5% Neem oil, 3% Bavistin, 0.1% + copper oxychloride, 0.25% Prosophis julifera leaf extract, 10% lpomoea cornea leaf extract, 10% Bavistin, 0.1% Dithane M45, 0.2% Ziram, 0.3% Copper oxychloride, 0.25% Control CD (P=0.05)
aMean of 3 replications. b ns = not significant.

Naturally infected plants Sheath rot incidence (%) 8.3 5.7 7.3 5.3 4.3 6.7 9.3 7.3 8.0 38.7 6.62 Yield (kg ha -1) 1208 1158 1225 1033 1067 1121 998 708 1158 700 ns

We studied the efficacy of plant products and chemicals to control rice sheath rot caused by Sarocladium oryzae. Field trials were conducted during 1993 dry (kharif) season. The experiment was laid out in a randomized block design with 10 treatments (see table) and three replications. The 25-d-old seedlings of cultivar ADT36 were planted at 15- 10-cm spacing in 5- 2-m plots. One trial was performed under artificial inoculation conditions and the other under natural infection conditions. The trials were

Sheath rot a incidence (%) 18.3 20.0 17.3 16.3 11.0 11.7 18.3 15.7 22.0 72.7 13.2

Yield (kg ha-1) 733 910 608 892 879 642 808 692 992 542 ns b

Integrated pest managementinsects

Effect of steam distillate extracts of rice cultivars on rice thrips
M. P. Parthiban and R. Veeravel, Entomology Department, Annamalai University, Annamalainagar, Tamil Nadu, India

We tested the reaction of 122 rice accessions against rice thrips Stenchaetothrips biformis (Bagnall) under field and screenhouse conditions at the Annamalai University experimental farm during July I993 and 1994. We used a 0-9 scale (Heinrichs et al 1985) to score the reactions. Steam distillates of six cultivarshighly resistant Ptb33 and IR62, resistant IET12008, susceptible ADT40, and highly susceptible IR64 and Rasishowing similar reactions under both field and

screenhouse conditions were prepared (Saxena and Okech, 1985) and bioassayed to assess the nymphal mortality of S. biformis and effect on fecundity of adults. Using a hand atomizer, the extracts (see table) were sprayed onindividual 21-d-old TN1 seedlings. One ml of suspension containing 2000 ppm of the active ingredient was delivered. Each treatment was replicated five times. Controls with acetone and water sprays were maintained. Each potted seedling was put into a mylar film cage (45 cm tall 10 cm diameter) and infested with a pair of freshly eclosed adult thrips. The eggs laid female -1 were counted 3 d after release. In another experiment with a similar set of treatments (see table), 21-d-old TN1 seedlings were sprayed with 1 ml of extract

Laboratory evaluation of steam distillate extracts of certain rice accessions against S. biformis.a Treatment (2000 ppm) Ptb33 IR62 IET12008 ADT40 IR64 Rasi Acetone Control Total eggs laidb (no.) 10.8 d 13.2 c 13.2 c 12.8 c 16.2 b 14.2 c 20.0 a 20.0 a 0.4 0.6 0.6 0.5 0.6 0.4 0.6 1.0 Larval mortality c (%) 70.0 a 4.7 60.0 a 4.7 64.0 a 4.7 43.3 b 2.7 40.0 b 4.7 39.6 b 2.7 6.6 c 2.7 3.3 c 2.7

a In a column, means ( SE) followed by the same letter do not differ significantly by DMRT (p = 0.05). bFive replications. c Three replications.

of each cultivar at the concentration of 2000 ppm. Plants treated with acetone and water alone served as controls. Each treatment was replicated three times. Treated seedlings were confined individually in test

50

IRRN 21:1 (April 1996)

tubes containing moist cotton to prevent wilting. The 10 first- instar nymphs were put into each test tube, and the top was fastened with muslin cloth. The percent of larval mortality was observed after 24 h. The extract of Ptb33 significantly inhibited oviposition significantly compared with the control and other cultivars tested. Mortality rates were significantly higher for treatments using extracts of Ptb33, IET12008, and IR62 than for the other cultivars and controls tested. The study revealed the differential reaction of rice accessions to S. biformis.

Adulterated pesticides in Cambodia

H. J. Nesbitt, Cambodia-IRRI-Australia Project, P. O. Box 01, Phnom Penh, Cambodia; A. Hickery and I. Phirum, International Development Research Center, P. O. Box 544, Phnom Penh, Cambodia

Cambodia has recently emerged from 25 yr of war and political isolation during which the countrys infrastructure and service networks were severely damaged or totally destroyed. During most of this period, agriculture existed in a state of low input/ low output technology. A village-based communal system controlled production, and the central government provided farmers with subsidized farm inputs. Toward the end of the 1980s, fertilizer and pesticides were imported almost exclusively from the Soviet Union and farm produce sales were tightly controlled. As a result, subsidized inputs were scarce and the country was self-sufficient neither in rice nor vegetables. At the beginning of the present decade, markets were opened and sales of farm inputs increased rapidly. Products were importedoften smuggled infrom a range of countries. No policy existed on the requisite quality or safety of the products and the government did not possess the necessary authority to implement any

regulatory programs. As a result, several dangerous chemicals banned in other countries were imported for sale in Cambodia. A range of products is presently available in Phnom Penh and provincial markets, including organochlorine insecticides, organophosphorus products, carbamates, and pyrethroids. The cheapest and most popular pesticides were generally the most dangerous, classified in the WHO class I category. By 1995, use of these products was concentrated in vegetable areas close to the capital, Phnom Penh, and in intensive rice-growing areas near the Vietnam border. Farmers observed that, despite the application of heavy doses of insecticides, pest kill was often low. Quality problems were suspected, so pesticide samples were collected from three markets within 15 km of Phnom Penh. Twenty-two products were analyzed by gas chromatography or liquid chromatography. Their active ingredients, their claimed strength, source, and measured concentration were determined (Table 1).
Analysis of 22 pesticide samples. Active ingredient Carbaryl Fenitrothion Methamidophos Parathion methyl Parathion methyl Carbofuran DDT Fenvalerate Methamidophos Methomyl Mevinphos Dichlorvos Mevinphos Mevinphos DDT Dichlorvos Dichlorvos Endrin Methamidophos Monocrotophos Monocrotophos "Mosquito poison" Source Vietnam Japan Vietnam Thailand Luxembourg Vietnam Vietnam Japan Vietnam Thailand Thailand Thailand Thailand Thailand Vietnam Vietnam Gov't store Gov't store Vietnam Vietnam Gov't store Khmer product Claimed strength 500 g L -1 500 g L -1 500 g L -1 500 g L -1 30 g kg -1 200 g L -1 500 g L-1 900 g kg -1 240 g L -1 500 g L-1 240 g L-1 240 g L -1 500 g L-1 725 g L -1 500 g L-1 -

Of the 22 samples analyzed, 5 were of the claimed strength or above, 6 were below strength, 3 were well below the claimed strength, and 8 possessed either none or extremely small amounts of active ingredient. Half of the analyzed samples were below 55% of the claimed strength. The active ingredient in the Khmer product could not be identified. Products were sourced from government stores, Vietnam, Japan, Thailand, and Luxembourg. All of the pesticides claimed to be from the government stores were inactive, and the remaining fakes were contained in bottles possessing Vietnamese labels. Some Vietnamese products, however, were analyzed and found to have high concentrations of active ingredient. The results of this study indicate that product regulation is necessary both for safety and quality control. This could be achieved by banning the importation of WHO category I pesticides and discouraging adulteration of products at the place of sale by imposing fines or other disincentives.

FAO a tolerance 475-525 g L -1 475-525 g L -1 475-525 g L -1 475-525 g L -1 27- 33 g kg -1 188-212 g L -1 475-525 g L -1 875-925 g kg -1 225-255 g L -1 475-525 g L -1 225-255 g L -1 225-255 g L -1 475-525 g L -1 475-525 g L -1 700-750 g L -1 475-525 g L -1 -

Measured strength 49 g kg -1 496 g L -1 528 g L -1 575 g L -1 539 g L -1 26 g kg -1 78 g kg -1 186 g L -1 470 g L -1 828 g kg -1 217 g L -1 266 g L -1 128 g L -1 131 g L -1 10.2 g L -1 <0.5 g L -1 73 g L -1 <0.5 g L -1 <5 g L -1 <5 g L -1 <5 g L -1 0

Concentration toleranceb 1 1 1 1 1 2 2 2 2 2 2 3 3 3 X X X X X X X X

a FAO, 1987. b 1 = within tolerable limits, 2 = below tolerable limits, 3 = well below tolerable limits. X = extremely low or nonexistent active ingredients.

IRRN 21:1 (April 1996)

51

Research methodology
Leaf carbohydrate analysisa simple method for integrating daily canopy photosynthesis
E. A. Conocono, T. L. Setter, and J. A. Egdane, IRRI

Current methods for assessing carbon assimilation require many photosynthesis measurements taken at different times during the day to get an accurate estimate on a daily basis. This is impractical and expensive in terms of labor and equipment, and it is sometimes impossible to do with more than one sample because of intermittent sunny and cloudy periods during the day. Carbon assimilation during the day will be integrated by the amount of carbohydrate assimilated by tissues. The nonstructural carbohydrate (soluble sugars and starch) that the plant produces during the day can be quantified by taking the difference between plant carbohydrate content at 6 a.m., when little or no photosynthesis has yet occurred, and at 6 p.m., when the plant has undergone photosynthesis for that day. This method may give a better estimate of carbon assimilated during the day compared with photosynthesis measurements taken at one point in time. We evaluated the relationship between plant carbohydrate content, daily canopy photosynthesis, and irradiance at the panicle initiation stage. The experiment used irrigated, direct seeded IR72. Light was varied using shade nets. Total soluble sugars was analyzed using anthrone reagent and starch by enzymatic hydrolysis with amyloglucosidase followed by glucose oxidase-peroxidase enzymes. Canopy photosynthesis was measured using a Licor 6200 portable photosynthesis system and a canopy frame (80 cm 80 cm and 130-cmhigh) with a removable top and four 120mm-diameter battery-operated fans to facilitate equilibration and rapid mixing of gas. Each measurement took 15 s and each replicate was measured 3-6 times every 2 h from 6 a.m. to 6 p.m. at four different radiation levels.

1. Total nonstructural carbohydrate profile of stems and leaves during the day. Plants were unshaded and samples taken at 0-2 days after panicle initiation. (error bars = standard deviation) 2. Correlation between daily canopy photosynthesis, net leaf carbohydrate content, and daily radiation. (error bars = standard deviation)

There is a linear increase in leaf carbohydrates with time (Fig. 1). This trend is not present in stem carbohydrates. meaning that daily canopy assimilation is more related to leaf carbohydrate content than stem carbohydrate content. The stem acts as storage tissue and may contribute up to 40% or more of grain carbohydrates such that diurnal changes in stem carbohydrates will be relative to high background concentrations. In contrast, leaf carbohydrates are at lower concentrations, and sugars are presumably readily mobilized from these tissues. Hence, leaf carbohydrates better reflect the results of current photosynthesis. A highly significant direct relationship exists between canopy photosynthesis per day and radiation (r2 = 1.00**), as well as between leaf carbohydrates and radiation (r2 = 0.99**) (Fig. 2). When leaf carbohydrate content and daily canopy photosynthesis are compared, we also find a highly significant correlation (r2 = 0.99**). Similar correlations were obtained in two other experiments. Based on these results, we suggest that leaf carbohydrate analysis is suitable for

quantifying daily canopy photosynthesis, especially when conditions do not allow efficient and accurate photosynthesis measurements.

A weather-based empirical model to predict rice leaf blast in Thailand

S. B. Calvero, Jr., IRRI; A. Surin, RPRGDepartment of Agriculture (DOA), Thailand; O. Pilansinchai, DOA, Thailand; and P. S. Teng, IRRI

Empirical models are important tools in rice disease management, especially in the context of sustainable agriculture. These models can guide farmers in choosing a disease management strategy to prevent an epidemic. For leaf blast, caused by Pyricularia grisea, models have been developed in some temperate countries where the disease naturally occurs. However, few of the models have been applied in actual production systems. In tropical countries such as Thailand, development of empirical models for rice blast is still in its early stages.

52

IRRN 21:1 (April 1996)

We used the multiple linear regression technique to generate an equation to serve as an empirical model to predict rice leaf blast on cultivar RD23. This model can be used as part of a blast management program in Thailand. The equation is based on leaf blast severity as the dependent variable assessed 32 d after sowing (DAS) from weekly sowings of the cultivar in 1993-94 under dry (upland) conditions at Uthong Field Crop Experiment Station, Suphanburi. Plants were grown and maintained in 1-m2 plots. Ammonium phosphate fertilizer (16-20-0) was applied to the plots along with 100 kg N ha-1. For each category of leaf blast seventy assessed, the likelihood of that severity correlating with a weather variable (e.g., rainfall frequency, number of days with relative humidity > 80%, mean maximum temperature) was estimated using WINDOW PANE, a weather variable searching computer program available at IRRI. This software served as an intercessory tool in developing the empirical model to better identify weather variables with direct effects on blast. The searching mechanism of WINDOW PANE is based on a series of time durations during which a particular weather variable occurs. For each time duration, correlation with leaf blast severity was estimated. Weather variables occurring at certain time durations that WINDOW PANE found to have high correlation with leaf blast severities were then chosen as independent variables in regression analysis. A best fitting equation was based on the following selection criteria: high adjusted coefficient of determination (r2), coefficient of variation (CV) < 25%, low predicted error sum of squares (PRESS), predictors with variance inflation factors (VIF) < 10, and regression coefficients ('s) with significance level (P) < 0.05. The best fitting equation used as an empirical model to predict percent leaf blast severity (Y) on RD23 is
Y = [56.184-8.261 MMAX0.710(MWS3)]2

Comparison of observed and predicted leaf blas severities observed 32 d after sowing on RD23. Uthong Field Crop Experiment Station, Suphanburi, Thailand. a Statistical parameter r 0 1 f tpooled Fitting sample (n = 20) 0.914** 2.146 ns 1.044** 0.096 ns 0.072ns Validation sample (n = 6) 0.955** 5.579ns 0.829** 1.587ns 0.375ns

a Values followed by ** and ns are significant (P < 0.05)

and not significant (P >0.05), respectively; n = number of observations.

satisfied these criteria reasonably well (see table) for both samples. A risk prediction chart for leaf blast in Suphanburi (see figure) was developed based on the predictions of the empirical model using different values of MMAX and MWS. From the chart, moderate to high blast risk exists at MMAX of 30-38 C and MWS of 0-2.0 ms -1 and a disease management intervention, such as fungicide application, may be needed. The chart also suggests that during days with 3.0 m s -1 windspeed, interventions would be unnecessary regardless of MMAX level, as blast would not substantially reduce rice yield.

Use of lesion number per leaf area to estimate lesion size for leaf blast quantitative studies
A. C. Calvero and R. S. Zeigler, IRRI

Risk prediction chart for leaf blast on RD23 sown at Uthong Field Crop Experiment Station, Suphanburi, Thailand, during 1993-1994. Infection: N = no infection; L = light (severity: 125%); M = moderate (severity: 27-66%); S = severe (severity: 67-100%).

where MMAX is the mean maximum temperature (C) at time durations from 40 d before sowing (DBS) to 14 DAS, and MWS is the mean windspeed from 40 DBS to 24 DAS. The equation has r2, CV, and PRESS values of 0.87, 11.62, and 23.86, respectively. All independent variables

have VIF values < 10 and their coefficients ('s) have P<0.05, which suggested uncorre-lated independent variables that are important in the model. The above equation was validated by comparing the actual and predicted severities for both fitting and validation samples (see table). The equation is useful in predicting leaf blast if the correlation coefficient (r) is > 0.85; the intercept (0) and slope parameters (1) have P > 0.05 and P 10.05, respectively; and the f [modified F statistic that tests simultaneously b 0 and 1 to be not different from zero (P>0.05) and unity (<0.05), respectively] and pooled tstatistic (tpooled) values have P > 0.05 in the validation sample. We found the equation

Leaf blast quantitative resistance studies to determine quantitative trait loci routinely use the parameters lesion number leaf area -1 (LnA) and lesion size leaf area -1 (LsA). Counting the number of sporulating lesions and measuring size using standard assessment keys make disease assessment prohibitively labor-intensive if samples are large. A high correlation between LnA and LsA has been reported in several studies. Lesion number is easily derived by simple counts, while lesion area is probably the most biologically significant. Therefore, we propose a mathematical model to describe their relationship, estimating LsA based on LnA alone. Data sets from two leaf blast polycyclic experiments done in the greenhouse using F7-recombinant inbred lines (RILs) of CO 39/IAC 165 and doubled haploid (DH) lines from IR64/Azucena, including their parent cultivars, were analyzed. Replications 3 and 4 were done on a staggered basis, with each replicate separated by 2 wk. LnA and LsA data were taken 2 wk after the plants were spray-inoculated with blast spore suspension. Isolates CA42 and PO6-6 were used to infect the RILs and DH lines, respectively. A total of 1318 disease observations were generated from the inoculation work. Out of this number, 988

IRRN 21:1 (April 1996)

53

observations were used to develop the model for LsA. The remaining 330 observations were used to validate the generated model. Multiple regression analysis indicated a significant exponential relationship between LnA and LsA (Fig. l). This relationship is described in the mathematical model as LsA = exp (1.545387+0.900306 (ln LnA)) where LsA is lesion size in mm2 cm -2 leaf area, ln is the natural logarithm with base e (e = 2.718), and LnA is lesion number cm -2 leaf area. The model fits the disease observations reasonably well with a coefficient of determination ( r 2 ) of 0.85. Both the 0 (y-intercept) and 1 (slope coefficient) coefficient terms have significant contributions to the model with t-values of 105.4 and 86.7, respectively. To test the validity of the model, the closeness in the values of LsA in the validation set and that of the model-predicted LsA was determined. Three criteria were used to determine such closeness: the variance (VARdiff ) of the difference between observed and predicted values (prediction error) is less than the mean square error (MSE) from the regression analysis of the full data set (n=l,318), average prediction error (APE)

1. Relationship between lesion number and lesion size in leaf blast.

2. Linear relationship of predicted and observed lesion size.

is close to zero, and the 0 and 1 obtained from regressing actual and predicted LsA is not significantly different from zero and unity, respectively. A 1:1 relationship between observed and predicted LsA for the validation data set (n = 330) was revealed (Fig. 2). The VARdiff (0.2704) was less than the MSE (0.2856) of the full data set and the APE was 0.0025, which is near zero value. Likewise, both 0 and 1 terms satisfy the third criterion,

suggesting that the model presented could be used to accurately estimate the LsA on the basis of LnA from new blast data sets. These data sets would likely come from various blast studies, specifically cultivar resistance screening, compatibility and virulence spectrum studies, and validation experiments for simulation models. These results apply to sporulating blast lesions.

A simple method for estimating yield loss of rice due to severe panicle blast
S. Koizumi, S. Fuji, and T. Tani, Mountainous Region Agricultural Research Institute, Aichiken Agricultural Research Center, Inabu, Kitashitara, Aichi 441-25, Japan

Seedlings were transplanted at 15- 30-cm spacing in 5.8-m 2 plots during 1992-94. Blast developed in the plots from natural infection. To obtain different levels of panicle blast severity, several fungicides were applied to Mineasahi, and mixed or

single plantings of NILs with different resistances to blast were made. Panicle blast severity was assessed by the Asaga scale (see table) on 20 representitive hills in each plot about 1 mo after heading. The mean of the assessed values was transformed to a

Asaga scale for assessing panicle blast severity in ricefields. Grade 0 1 3 4 5 6

and spikelet pedicels.

We investigated the relationship between panicle blast severity and yield loss of rice. We showed that yield loss can be estimated using a simple method under severe panicle blast situations. We used Mineasahi, a Japanese rice cultivar with a low level of partial resistance to blast, and near-isogenic lines (NILs) with different complete resistance genes to blast. These NILs were developed from Japanese rice cultivars Sasanishiki and Nipponbare.

Description No panicle blast observed. A few panicle branches diseased (slight). Diseased panicle branches are seen at a glance and neck blast is found in some cases (moderate). Many panicle branches are diseased and neck blast is observed (a lot). Many panicle branches are diseased and diseased neck nodes are moderate (severe). Panicle branches and neck nodes are severely diseased (very severe). All panicles are diseased.

Diseased spikelets a (%) 0 5 10 25 50 75 100

a Diseased spikelets include the spikelets prematurely killed by infection of P. grisea on neck nodes, panicle branches,

54

IRRN 21:1 (April 1996)

panicle blast for each of Mineasahi and the NILs. The estimated yield with no panicle blast was determined by extrapolating the regression line of yield and percentage of spikelets diseased with panicle blast to the percentage of spikelets diseased with panicle blast with a value of zero. In each experiment, yield loss was correlated positively with the percentage of spikelets diseased with panicle blast (significant at P < 0.01) and the yield loss almost agreed with the percentage of diseased spikelets. However, the agreement between panicle blast and the diseased spikelet percentage and yield loss was not severe compared with the other experiments (see figure). Yield loss caused by panicle blast varied with the time of P. grisea infection on rice panicles and was affected by leaf blast severity. Moreover, weather conditions and cultivar resistance may affect yield loss. However, our results show that yield loss due to panicle blast can be estimated by the percentage of diseased spikelets under severe disease conditions. The method is very simple and the percentage of diseased spikelets in rice hills can also be assessed directly without the Asaga scale. Farmers who cultivate rice under severe panicle blast conditions can use the method to estimate yield losses due to the disease.

Effect of inoculation method at different stages of growth on rice bunt infection. Bunted grains (%) Growth stage/method Boot - syringe spray Panicle exsertion - spray 1993 12.00 (20.26)a 3.00 (9.95) 6.67 (14.95) 4.00 (11.52) 0.00 (0.00) 0.00 (0.00) Uninoculated check CD (P = 0.05) CV (%) 0.00 (0.00) 1.10 7.78 1994 12.33 (20.56) 3.66 (11.02) 6.67 (14.95) 4.33 (12.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.97 6.63

Anthesis - spray Milk - spray

Dough - spray

a Figures in parentheses are angular transformed values.

Inoculation method for rice bunt

R. N. Singh and R. C. Rai, Plant Pathology Department, Rajendra Agricultural University, Pusa, Samastipur 848125, Bihar, India

Relationship between yield loss (%) and spikelets diseased with panicle blast (%).

percentage of diseased spikelets with the Asaga scale (see table). In each plot, 80 hills were harvested to determine grain yield for Mineasahi and rough rice yield for the NILs. Yield loss due to panicle blast was calculated each year from an estimated yield with no

Incidence of rice bunt (kernel smut) caused by Neovossia horrida (Tak.) Padwick and Khan is erratic, so screening germplasm for bunt or chemicals for use against bunt under natural incidence is not reliable. We studied the effect of inoculation method and crop stage on bunt development in a glasshouse in 1993 and 1994. Twenty-five tillers in three replications were inoculated at five growth stages (booting, panicle exsertion, anthesis, milk grain, and dough grain). Uninoculated panicles served as the check.

Two inoculation methods were used. In the syringe method, 2 ml of sporidial suspension (2500 sporidia ml-1) was injected into the boot. In the spray method, inoculum was sprayed on the panicles. The area enclosing the inoculated plants was covered with muslin cloth; water was sprinkled over the cloth to provide humidity. At booting, half of the plants were inoculated by syringe method and the other half by spray method. At the other stages, the spray method was used. The syringe inoculation method resulted in four times more bunted grains than did the spray method (see table). The best stage at which to spray using the spray method was panicle exsertion, at which time maximum disease occurrence was recorded. No disease developed when the crop was inoculated after anthesis. The syringe inoculation method should be adopted for screening germplasm, and the spray method should be used for evaluating antifungal compounds. The syringe inoculation method may not be feasible for inoculating crops on a large scale.

IRRN 21:1 (April 1996)

55

Instructions for contributors

NOTES General criteria. Scientific notes submitted to the IRRN for possible publication should be original work, have international or pannational relevance, be conducted during the immediate past three years or be work in progress, have rice environment relevance, advance rice knowledge, use appropriate research design and data collection methodology, report pertinent, adequate data, apply appropriate statistical analysis, and reach supportable conclusions.
Routine research. Reports of screening trials of varieties, fertilizer, cropping methods, and other routine observations using standard methodologies to establish local recommendations are not ordinarily accepted. Examples are singleseason, single-trial field experiments. Field trials should be repeated across more than one season, in multiple seasons, or in more than one location as appropriate. All experiments should include replications and an internationally known check or control treatment. Multiple submissions. Normally, only one report for a single experiment will be accepted. Two or more items about the same work submitted at the same time will be returned for merging. Submitting at different times multiple notes from the same experiment is highly inappropriate. Detection will result in the rejection of all submissions on that research.

category in which the note being submitted should appear. Write the category in the upper right-hand corner of the first page of the note.
GERMPLASM IMPROVEMENT

IRRN categories. Specify the

genetic resources genetics breeding methods yield potential grain quality pest resistance diseases insects other pests stress tolerance drought excess water adverse temperature adverse soils other stresses integrated germplasm improvement irrigated rainfed lowland upland flood-prone (deepwater and tidal wetlands) seed technology
CROP AND RESOURCE MANAGEMENT

soils soil microbiology physiology and plant nutrition fertilizer management inorganic sources organic sources crop management integrated pest management diseases insects weeds other pests water management farming systems farm machinery postharvest technology economic analysis
ENVIRONMENT SOCIOECONOMIC IMPACT EDUCATION AND COMMUNICATION RESEARCH METHODOLOGY

Manuscript preparation. Use international measureArrange the note as a brief ments. Do not use local units of measure. Express yield data statement of research objectives, a short description of in metric tons per hectare (t/ha) for field studies and in project design, and a succint grams per pot (g/pot) for smalldiscussion of results. Relate results to the objectives. Do not scale studies. Express all economic data in include abstracts. Do not cite references or include a terms of the US$. Do not use bibliography. Restrain acknowllocal monetary units. Economic information should be preedgments. Manuscripts must be in sented at the exchange rate English. Limit each note to no US$:local currency at the time more than two pages of doubledata were collected. spaced typewritten text. Submit When using acronyms or the original manuscript and a abbreviations, write the name duplicate, each with a clear copy in full on first mention, followed of all tables and figures. Authors by the acronym or abbreviation should retain a copy of the note in parentheses. Use the and of all tables and figures. abbreviation thereafter. Apply these rules, as Define any nonstandard appropriate, in the note: abbreviations or symbols used Specify the rice production in tables or figures in a footecosystems as irrigated, rainfed note, caption, or legend. lowland, upland, deepwater, and tidal wetlands. Each note can have no Indicate the type of rice more than two tables and/or culture (transplanted, wet figures (graphs, illustrations, or seeded, dry seeded). photos). All tables and figures If local terms for seasons are must be referred to in the text; used, define them by characterthey should be grouped at the istic weather (wet season, dry end of the note, each on a season, monsoon) and by separate page. Tables and months. figures must have clear titles Use standard, internationally that adequately explain the recognized terms to describe contents. rice plant parts, growth stages, and management practices. Do Review of notes. The IRRN not use local names. editor will send an acknowledg Provide genetic background for ment card when a note is new varieties or breeding lines. received. An IRRI scientist, For soil nutrient studies, selected by the editor, reviews include a standard soil profile each note. Reviewer names are description, classification, and not disclosed. Depending on relevant soil properties. the reviewer's report, a note Provide scientific names for will be accepted for publication, diseases, insects, weeds, and rejected, or returned to the crop plants. Do not use common author(s) for revision. names or local names alone. Quantify survey data, such as Comments. If you have cominfection percentage, degree of ments or suggestions about the severity, and sampling base. IRRN, please write to the editor. When evaluating susceptibility, We look forward to your resistance, and tolerance, continuing interest in IRRN. report the actual quantification of damage due to stress, which Mailing address. Send notes was used to assess level or and correspondence to the incidence. Specify the measureEditor, IRRI, P.O. Box 933, ments used. Manila 1099, Philippines. Use generic names, not trade Fax: (63-2) 891-1292 names, for all chemicals. E-mail: C.de-dolph@cgnet.com