ABSTRACTS OF POSTERS
P - 01 CHARACTERIZATION OF A AND C GENOME-SPECIFIC DISPERSED AND CLUSTERED REPETITIVE DNA SEQUENCES FROM AVENA SATIVA Evgueni V. Ananiev, M. Isabel Vales, Ronald L. Phillips, and Howard W. Rines Department of Agronomy and Plant Genetics, University of Minnesota and USDA-ARS, St. Paul, MN 55108
The genomic DNAs from eight species of oat with A, C, AB, AC, and ACD genomes were used as probes in Southern blot hybridization to a blot panel of cosmid clones from A. sativa (ACD) genomic DNA cut with different restriction enzymes. Several DNA subfragments were identified which specifically hybridized with labeled A or C genomic DNA. Those subfragments were recloned and characterized. Among them, eight A/D-genome and five C-genome specific DNA sequences have been identified as members of different families of dispersed repetitive elements. In situ hybridization of an A/D genome-specific repeat, termed As-A1, to metaphase chromosomes of A. sativa showed that it is dispersed along 14 out of 21 A. sativa chromosome pairs, representing the A and D genomes. One of the C genome-specific repeats, termed As-C1, is dispersed along 7 of the 21 pairs of A. sativa chromosomes. Six of the seven C genome chromosome pairs have chromosomal segments of different length at their termini which gave hybridization with the As-A1 but not with the As-C1 repeats while termini of six out of fourteen AD genome chromosome pairs gave hybridization with the As-C1 but not with the As-A1 repeats. Thus, these two probes, As-A1 and As-C1, enabled the detection of presumably ancient reciprocal terminal translocation segments of the C chromosomes and A or D chromosomes. A second C genome-specific repeat, termed As-C2, revealed clusters on six of the seven C chromosome pairs. Size, number and location of those clusters are specific for each chromosome pair of the C genome. These three genome-specific repetitive DNA sequences were used to screen the A. sativa genomic cosmid library to selectively isolate C or A/D genome-specific cosmid clones. Analysis of these new clones revealed that they contain additional genome- or chromosome-specific DNA probes.

P - 02 EARLY DETECTION OF CHROMOSOME RETENTION/ELIMINATION IN OAT X MAIZE HYBRIDS USING GENOMIC IN SITU HYBRIDIZATION Evgueni Ananiev, Gang Chen, Howard W. Rines, and Ronald L. Phillips Department of Agronomy and Plant Genetics, University of Minnesota and USDA-ARS, St. Paul, MN 55108
Interspecific hybridization between different species of cereals is widely used for the production of haploid plants or chromosome addition lines as the result of preferential elimination or occasional retention of chromosomes of one of the parents. The factors which affect chromosome elimination/retention are poorly understood and difficult to monitor. In most cases, cytological and molecular characterization requires well developed plants. We applied genomic fluorescence in situ hybridization for the early detection of chromosome elimination/retention in developing embryos, endosperm and early seedlings in oat x maize hybrids. Developing embryos and endosperm were isolated for analysis from caryopses at 3 and 6 days after pollination. In addition, root tips of young seedlings from rescued embryos were cut off, fixed in 45% acetic acid, squashed and made into permanent preparations. Maize genomic DNA or a multiprobe ( a collection of dispersed highly repetitive elements of the maize genome) were labeled with fluorescein and used for in situ hybridization. Maize chromosomal material was detected in metaphase plates as well as in interphase nuclei using this technique. All embryos from oat-maize crosses containing maize chromosomes were chimeric. Different cells in a developing embryo can have different numbers of retained maize chromosomes from zero to ten. In older embryos it is possible to detect sectors of cells without maize chromosomes. Most endosperm cells in contrast contain practically a full set of maize chromosomes. The cell division in endosperm usually ceases early in development. Different levels of chromosome retention/elimination were detected in developing young seedlings. The process of maize chromosome elimination appears to begin during the first cell divisions after fertilization, occurs with different rates in different tissues, and is highly asynchronous.

P - 03 THE TRANSFER OF CROWN RUST RESISTANCE GENE PC94 FROM A WILD DIPLOID TO CULTIVATED HEXAPLOID OAT Taing Aung¹*, James Chong¹ and Mike Leggett² ¹ Cereal Research Centre, Agriculture and Agri-Food Canada, 195 Dafoe Road, Winnipeg, MB, Canada, R3T 2M9; ²Institute of Grassland & Environmental Research, Aberystwyth, Wales

Crown rust Puccinia coronata Cda. f.sp. avenae Eriks. is the major foliar disease of cultivated oats in Canada and the United States. Estimates of the reduction of the grain yield, based on yearly reports of Cereal Rust Laboratories U.S. Department of Agriculture, caused by the pathogen vary from 10% to 35% depending on the severity of the infection and the cultivar used. The ability of the crown rust pathogen to produce new biotypes capable of overcoming previously resistant cultivars usually within a few years has led to a continuous search for further sources of resistance. Several wild diploid species of oat particularly, Avena strigosa, confer high degree of resistance to crown rust. Highly effective crown rust resistance gene(s) was identified in a genotype of diploid oat A. strigosa Schreb and the genotype was desiganated as (RL1697) at CRC, Winnipeg. The resistance diploid genome of RL1697 was combined with the hexaploid genome of oat cultivar SunII in the form of a synthetic octaploid (2n = 56). Resistant disomic addition lines were generated in the subsequent backcross progenies from the crosses between the octaploid and SunII recurrent parent. The resistance from the RL1697 chromosomes in the addition lines was introgressed into the SunII genome through induced chromosome pairing procedure. We report on the successful transfer of crown rust resistance from A. strigosa (RL1697) to cultivar SunII.

The diploid oat A.strigosa (RL1697) resistant to crown rust race CR152 was obtained from the Institute of Grassland and Environmental Research, Wales, UK. in 1991. The F1 plants resistant to CR152 were produced from the reciprocal crosses between (RL1697) and A. strigosa {Saia(Wpg)} which is susceptible to CR152. The selfed progeny of the diploid F1 plants segregated in a ratio of 3 R : 1 S, indicating that the RL1697 resistance is inherited as a single dominant gene. Oat cultivar SunII (pistillate parent) was crossed with RL1697 (pollen parent) and the F1 hybrids were treated with 0.2% colchicine to produce three synthetic octaploids (2n = 56). Resistance reaction observed in the synthetic octaploids (2n = 56) indicated that the resistance confer by RL1697 genome when combined with SunII genome is expressed. Resistant plants with 2n = 44, 45 chromosome (disomic addition) were selected from the selfed progenies of the BC1 plants produced from a cross between SunII and the octaploid. The selected resistant chromosome addition lines were cytologically unstable due to addition decay (loss of added RL1697 chromosomes resulted in the loss of the resistance) which indicate that breed-true resistance line cannot be establish from the disomic RL1697 chromosome addition line. . The disomic addition lines were then crossed to another synthetic octaploid (2n=56) derived from a cross between SunII (hexaploid) and A. longiglumis,CW57 (diploid) and the resistant F1 plants so produced were backcrossed to SunII as pistillate parent for five times. The homozygous resistant plants which breed true for resistance in the selfed progenies of BCV populations from heterozygous resistant lines which segregated in a monofactorial ratio of 3 resistant : 1 susceptible were produced. The homozygous and heterozygous plants with 42 chromosomes were selected from the F3 generation of the BCV progenies. The selected resistance lines exhibited no cytological abnormality (no loss or gain of chromosome, no unusual chromosome pairing at meiosis) all the heterozygous resistant lines segregated in 3R:1S ratio and all the homozygous lines breed true for resistance. This confirms the transfer of crown rust resistance gene from A. strigosa (RL1697) to cultivar SunII. One homozygous line which is cytologically stable (2n=42 and meiosis with regular 21 bivalents) and with agronomically desirable characteristics was selected, the line was designated as S42, the resistance gene transferred from RL1697 to SunII is designated as Pc94. Subsequent evaluations showed that Pc94 is highly resistant (0 ; reaction type) to crown rust tested in the green house and under field conditions in Manitoba and in New Zealand.

Introgression of genes from species of lower ploidy (diploids and tetraploids) into the genome of hexaploid oat cultivars is very difficult to achieve by means of regular backcrossing procedures. This is mainly due to the low frequency with which the chromosomes of the wild species pair with those of cultivated oat, consequently reducing the chance of recombination and gene transfer. It is well documented that a genotype of A. longiglumis (CW57) suppress the activity of the gene or genes controlling regular bivalent pairing in A. sativa and induced pairing between non homologous chromosomes that resulted in translocation between two homoeologous chromosomes as well as between unrelated chromosomes. The ability of CW57 gene(s) in inducing pairing between A.strigosa (RL1697) chromosome and A.sativa chromosome probably have affected the transfer of the RL1697 resistance into SunII genetic background.

S42 was used as pollen parent and crossed to 11 North American oat cultivars (AC Belmont, AC Marie, AC Preakness, Caliber, Dal, Derby, Dumont, Jasper, Riel, Robert, Steele). The resistance gene Pc94 is inherited as dominant gene in five oat cultivar (SunII, Calibre, Dal, Derby and Jasper) genetic backgrounds. However, Pc94 is inherited as a recessive gene in seven cultivar (AC Belmont, AC Marie, AC Preakness, Dumont, Riel, Robert and Steele) genetic backgrounds. Since crown rust resistance gene Pc38 is present in all the above seven cultivars and the expression of Pc94 is modified in all these cultivars it is logical to conclude that Pc94 expression is suppressed by Pc38 (J. Chong & T. Aung , see the following article). However, the modification of Pc94 gene expression could also be due to reversal of dominance to recessive nature of the gene in some genetic backgrounds similar to the phenomena reported in wheat with reference to leaf rust resistance genes. While Pc94 is proven to be a highly effective gene, there is still a need to deploy this gene in combination with other Pc genes with the aim to provide adequate protection to the oat crop from the ravages of the disease and to prolong the life of the resistance gene. The S42 genotype with the Pc94 gene has provided a valuable new source of resistance to crown rust for oat improvement programs in Canada.

*Corresponding author (Email: TAUNG@EM.AGR.CA)

*Corresponding author (Email: TAUNG@EM.AGR.CA

P - 04 OAT GERMPLASM RESOURCES IN THE USDA-ARS NATIONAL SMALL GRAINS COLLECTION Harold E. Bockelman USDA-ARS, National Small Grains Collection, Aberdeen
The USDA-ARS National Small Grains Collection (NSGC) is a large working collection of wheat, barley, oat, rye, triticale, and rice. The origins of these small grain collections date back to the beginning of the 20th century with the first USDA plant explorers. For the first half of the century the collections were maintained individually by USDA Cereal Investigators. The NSGC was officially organized in 1948, bringing together the various small grain collections. The oat (Avena) collection presently consists of 20,781 accessions representing 17 species originating in 95 countries and regions of the world. Important components of the oat collection include 9,904 accessions of the hexaploid cultivated species, A. sativa, and 8,264 accessions of the wild hexaploid species, A. sterilis. More than 225,000 evaluation data records are available, including important disease, insect, quality, and agronomic characters. Passport and evaluation data are maintained on the Germplasm Resources Information Network (GRIN), available online at http://www.ars-grin.gov/npgs.

P - 05 AGRONOMIC AND GRAIN QUALITY CHARACTERISTICS OF HULLESS OATS D. E. Burrup¹, D.M. Wesenberg¹, D.M. Peterson², and J.C. Whitmore^3. ¹National Small Grains Germplasm Research Facility, USDA, ARS, 1691 S. 2700 W., P.O. Box 307, Aberdeen, Idaho 83210, USA; ²Cereal Crops Research Unit, USDA, ARS, 501 Walnut St., Madison, Wisconsin 53705, USA; ³University of Idaho, Tetonia Research & Extension Center, 888 West Highway 33, Newdale, Idaho 83436, USA
Grower interest in hulless oats in Idaho and adjacent states is being maintained, but most planted acreages remain small and largely constitute on-farm trials. Inquiries received by the authors frequently involve growers interested in specialty markets, dairy producers, or horse trainers. Hulless oat improvement currently accounts for about 25% of the oat program effort at Aberdeen. One of the highest yielding selections in the program continues to be 86Ab1616, which averaged 5622 lbs/A in groat yield in irrigated trials conducted at Aberdeen, Idaho in 1994-97. The well adapted high yielding cultivar 'Monida' averaged 5275 lbs/A in groat yield in the same trials. The best Aberdeen hulless oat selections relative to expression of the hulless trait (e.g., 90Ab1500) averaged about 98% hulless kernels and some of the hulless oat selections developed by Agriculture and Agri-Food Canada, Ottawa, Ontario (e.g., NO51-1) averaged over 98% hulless kernels in these trials. These selections represent marked improvement over some of the earlier hulless oat cultivars relative to the expression of the hulless trait. Major problems frequently associated with hulless oats that continue to impede grower and user acceptance include lower yield, shattering, poor expression of the hulless trait, a severe itch factor, germ damage, and grain handling problems in elevators. Data that may impact on the latter issue are presented here for review. During the conduct of test weight determinations it was observed that hulless oat selections vary significantly in their rate of flow in test weight equipment. The time required for a standard amount of grain (one quart in this instance) to flow from a standard test weight device was measured. Trichome characteristics, including trichome dehiscence, were also studied. Among 11 oat cultivars grown in an elite trial for four years, three hulled cultivars ('Ajay', Monida, and 'Ogle') were very similar in flow time, averaging 2.4 seconds. The eight hulless cultivars averaged 2.8 seconds and ranged from 2.6 to 3.6 seconds. At least in some cases the presence of prominent, long trichomes appeared to be associated with the reduced rate of flow. In a larger series of 37 hulless oat entries grown in the same trial at Aberdeen in 1997, the flow time ranged from 2.6 to 3.6 seconds, with the hulled entries again averaging 2.4 seconds. In a second trial grown under irrigation at Aberdeen in 1997, the flow time for 23 hulless cultivars ranged from 2.7 to 4.1 seconds. This newly studied characteristic might be an additional useful criterion to consider in the development of hulless oat cultivars.

P - 06 RECURRENT SELECTION FOR INCREASED PARTIAL RESISTANCE TO OAT CROWN RUST> C. Castell¹, L. Savin², L. Federizzi³, G. Ochocki⁴, and D. Stuthman¹ Dep. of Agronomy and Plant Genetics, University of Minnesota; ²ENSAT, Toulouse France; ³ UFRGS, Porto Legre, Brazil; and ⁴USDA-ARS Cereal Disease Laboratory, St. Paul, MN
Major gene crown rust resistance has not ever been long lasting. Because it is usually race-specific, it can be overcome as new virulences emerge in the pathogen population. Partial resistance, especially if it is race non-specific, is likely longer lasting, or durable. This alternative type of resistance, also called incomplete, rate-reducing, slow rusting, or late rusting among other names, is often under polygenic control. Therefore using recurrent selection (RS) to accumulate genes with smaller effects should theoretically be effective. Another important difference between major gene and partial resistance is that partial resistance can not be measured in absolute terms. Rather, it is a relative measure comparing a possibly partially resistant test candidate to a fully susceptible genotype. Our objectives were to 1) practice RS to increase the level of partial resistance, 2) evaluate the parents from each cycle of selection for their relative level of partial resistance, and 3) determine which mechanisms of partial resistance (infection efficiency, latent period, or spore production), if any, were modified during the RS. F5 lines from cycle 7 of our original Closed Recurrent Selection System (CRSS) were used as the base population for this study. Twenty to 25 lines from 63 families were planted at Rosemount, MN, in May, 1996, and evaluated for partial crown rust resistance after exposure to natural inoculation. Based on family average, the 21 best families (least rust at maturity) were chosen as parents for the next cycle. These 21 parents were crossed in a circulant partial diallel in fall 1996 and the resulting 63 families of F1 plants were grown in the spring 1997 greenhouse. In May 1997, these resulting F2's were planted at Rosemount to initiate another cycle of RS for partial resistance after exposure to natural infection. For this cycle, selection was first for the 21 best families and then for the best plant in each selected family. This procedure was repeated for the next cycle and on May 5, 1998, F2 families were planted at Rosemount to initiate the third cycle of selection for partial crown rust resistance. Simultaneously, the 12 C0 parents from the CRSS, the 21 parents from cycle one RS for resistance, the 21 parents from cycle 2 and 'Starter', a susceptible check, were planted in a RCB design to determine the putative genetic progress for partial crown rust resistance. Using adult plant inoculation in the greenhouse on the same three sets of parents and Starter, possible resistance mechanisms of slow-rusting are being studied.

P - 07 CHARACTERIZATION OF MAIZE CHROMOSOME RETENTION IN DEVELOPING EMBRYOS AND PLANTS FROM OAT X MAIZE CROSSES Gang Chen¹, Evgueni Ananiev¹, M. Isabel Vales¹, Oscar Riera-Lizarazu¹, Howard Rines², and Ronald Phillips¹ ¹University of Minnesota; ²USDA-ARS, St. Paul, MN 55108
When the oat cultivar Starter was pollinated with maize hybrid Seneca 60, about 20% of the florets developed embryos. However, only 12% of the rescued embryos germinated and only about half of those survived to maturity. Maize chromosome identification with maize microsatellite (SSR) probes showed that all plants that ceased growth at an early stage retained more than one maize chromosome, 35% of the surviving plants retained one to two maize chromosomes and the remaining 65% surviving plants had no maize chromosomes. The retention of multiple maize chromosomes apparently retards embryo germination and seedling growth. In attempts to obtain faster and more complete maize chromosome elimination during embryo development, the effects of different temperatures were tested on the chromosome constitution of young embryos in oat x maize hybridization. Oat x maize crosses were made with the traditional emasculation and pollination method followed by a treatment of 10 mg/L 2,4-D solution to florets two days after pollination. After 2,4-D application, sets of plants were put in growth chambers under temperatures of 16°C/14° C (day/ night), 20°C/18°C, or 25°C/23°C. Six days after pollination embryos were harvested and analyzed for chromosome constitution by an in situ hybridization technique using fluorescein-labeled maize total genomic DNA as the probe. All of the embryos from the different temperature treatments at this stage were chimeric in terms of maize chromosome retention. In this initial test of 5 embryos per treatment, maize chromatin was detected in 93.1% of the embryo cells in the high temperature treatment, 47.9% in the middle temperature treatment, and 27% in the low temperature treatment. These results indicate that low temperature leads to less maize chromosome retention during embryo development.

P - 08 EVALUATION OF PARTIAL RESISTANCE TO OAT CROWN RUST FOR QTL IDENTIFICATION Gang Chen¹, Howard Rines², Kurt Leonard², Jerry Ochocki², Deon Stuthman¹, and Ronald Phillips¹ ¹University of Minnesota and ²USDA-ARS, St. Paul, MN 55108
The overall objective of this research is to identify molecular (RFLP, AFLP) markers associated with genes for crown rust partial resistance in oat to be able to transfer these genes to elite germplasm using marker assisted selection. A population of 158 F6:8 recombinant inbred lines (RILs) derived from a cross between MN841801, a partially resistant line, and Noble, a susceptible cultivar, were evaluated for their resistance to oat crown rust in 1997 and 1998 in both greenhouse and field tests. In the greenhouse experiments, a single rust isolate MNB236 was chosen based on its virulence to the parents in the seedling stage and used to inoculate the flag leaf at the floret emergence stage. The pustule numbers per leaf area were hand-counted every 2 days for 5 dates starting from 12 days after inoculation. Digital pictures of the inoculated leaves were taken for image analysis of pustule numbers per leaf area, percentage of diseased area vs. total leaf area, and pustule size. For the field experiments, a natural crown rust population was used to inoculate adult plants at floret emergence and a computer simulated scale was used as a reference to read the percentage of diseased area vs. total leaf area. The number of pustules per leaf area obtained by digital image analysis and by hand counting matched well , the correlation coefficient was 0.91. The values for the different progeny lines for the area under the disease progress curve (AUDPC) from the greenhouse test had the same patterns as those from the field. Tests of selected RILs with four other isolates gave no indication of race specificity in the resistance responses. Restriction enzymes BamHI,DraI,EcoRI,EcoRV and HindIII are being used to survey parents and progeny for RFLP polymorphisms for marker analysis.

P - 09 MOLECULAR CLONING, CHARACTERISATION, MAPPING AND FUNCTIONAL ANALYSIS OF DISEASE RESISTANCE GENES FROM OAT Davis W Cheng and K C Armstrong Eastern Cereal and Oilseed Research Centre, Agriculture & Agri-Food Canada, Ottawa, ON
Molecular cloning, characterisation, mapping and functional dissection of oat disease resistance genes will elucidate the mechanisms of gene mutation and reorganisation involved in the evolution of new forms of disease resistance. Cloning of the disease resistance genes would offer the possibility to use differences between the resistant and susceptible alleles as markers for selection in breeding and to engineer crop varieties with a higher level of resistance.

Resistance genes cloned from flax (L6, M), Arabidopsis (RPS2, RPM1, RPP5), tomato (prf, I2, Cf-9, Cf-2), sugar beet (HS1), tobacco (N), rice (Xa21), and others barley (b1-b9),rice (rz1-15 ), share conserved domains from which primers can be designed to amplify disease resistance gene analogues. We have used degenerate primers to these conserved domains and primers designed to other isolated genes eg wheat (Lr10), barley (mlo), peroxidase from wheat etc. We have cloned two peroxidase genes which may be induced by powdery mildew; two NBS-LRR type genes from primers for b1 of barley; three resistance gene analogues which contain a typical LRR motif from degenerant primers designed from L6. We have also cloned related sequences with primers from Lr10 of wheat, three related sequences to mlo from barley, two sequences related to rice pi-2(t), and thirteen related sequences to flax L6 and tobacco N. Blast search results indicate that several of these genes contain domains with high homology to domains from different genes from other species.

Crown and stem rust are probably the two most important diseases of oat. Our efforts will focus on cloning and identifying rust resistance gene analogues. In particular we will focus on the Pg3, Pg9, Pc68 etc gene cluster by amplifying sequences from the telo-VIII cytogenetic stock using the microdissection protocol. We have confirmed that the genes are probably located on this chromosome arm. Mapping the amplified products to the known map positions of rust genes will imply function and northern analysis after pathogen challenge will imply induction by the pathogen.

P - 10 MOLECULAR CLONING AND CHARACTERIZATION OF PEROXIDASE GENES FROM OAT ( AVENA SATIVA L. VAR. OGLE ) WHICH MAY BE INVOLVED IN INDUCED RESISTANCE TO ERYSIPHE GRAMINIS F. SP. TRITICI Davis W. Cheng and Ken C. Armstrong Eastern Cereal and Oilseeds Research Center, Agriculture & Agri-Food Canada, Bldg 50, Central Experimental Farm, Ottawa, Ontario K1A 0C6, Canada
Two peroxidase genes, temporarily designated pxc2 and pxc6, were cloned and characterized from oat based on a peroxidase gene sequence from common wheat. They are 1191 bp and 1197 bp in length and contain 3 exons and 2 introns in almost the same position . The mRNA splicing site is typical of eukaryotic genes. Their exon sequences are more conserved ( 79.4% - 88.7% ) than their intron sequences ( 42.1% - 45% ). Blast search results indicates that the gene sequences reported here are very similar to the barley mRNA for peroxidase induced by powdery mildew ( 85% - 91% ), the wheat pox2 gene induced by powdery mildew with haem-binding activity ( 85% - 89% ), and the wheat pathogen peroxidase gene induced to powdery mildew ( 86% - 100% ). The putative proteins encoded by oat pxc2 and pxc6 genes consist of 313 aa ( pI 7.98 ) and 314 aa ( pI 8.18 ) respectively compared to the wheat peroxidase gene which consists of 310 aa ( pI 7.98 ). The alignment of amino acid sequences indicates that the structural homologies between oat peroxidase PXC2 and PXC6 and wheat and barley peroxidases are still very high (64.3% - 75.4% ). Experiments is to study the developmental expression and chromosomal localization of these peroxidase genes from oat which will also be outlined.

P - 11 IDENTIFICATION OF OAT SPECIES AND VARIETIES USING MINISATELLITE AND MICROSATELLITE DNAs ChengDao Li, Brian G. Rossnagel, and Graham J. Scoles Plant Sciences Department & Crop Development Centre, University of Saskatchewan, 51 Campus Drive, Saskatoon, SK S7N 5A8

The detection and analysis of DNA polymorphisms in crops is an essential component of marker-assisted selection and cultivar identification in plant breeding. We have explored microsatellite DNA primers and the direct amplification of minisatellite DNA to identify oat species and varieties.

Five primers were designed based on minisatellite DNA core sequences. The PCR products of primers from A longiglumis, A wiestii, A canarien, A strigosa, A clauda, A abyssinica, A barbata, A maroccana, A murphyi, A sterillis, A fatua, A byzantina and A sativa demonstrated a high degree of polymorphism and the various genomes could be identified. However, the level of polymorphism among cultivated varieties was low.

Microsatellite DNA enrichment libraries have been constructed. Genomic DNA of oat (Avena sativa cv. Ogle) was digested with a combination of restriction enzymes DraI, AluI and RsaI, and 200 to 900 bp fragments were excised and purified from the gel. The SSR fragments were enriched with 3'-biotinilated oligos and cloned into a pCR 2.1 vector. More than 90% of the white clones showed positive hybridization with the synthesized SSR oligos. Sequence analysis demonstrated that the most enriched SSR library is a (GAA/CTT)n repeat, with 92.3% of the white clones containing that sequence. The (TC/AG)n and (TG/AC)n libraries have 69% and 51% of the SSR clones respectively. However, no (AT/TA)n repeat clone was identified in the present study, although it has been reported to be the most abundant SSR in plants. These libraries could provide enough SSR clones for constructing a saturated SSR linkage map for genetic analysis of oat. Twenty pairs of primers were designed and synthesized based on the above sequences. These primers were used to characterize 11 Avena species and 20 cultivated oat varieties.

P - 12 QUANTITATIVE TRAIT LOCI IDENTIFIED IN A HULLESS BY COVERED OAT (AVENA SATIVA L.) POPULATION David De Koeyer¹, Nick Tinker¹, Jitka Deyl¹, Vern Burrows¹, Corinne Chenier¹, Steve Molnar¹, Ken Armstrong¹, George Fedak¹, Charlene Wight¹, Darrell Wesenberg², Brian Rossnagel³, Deon Stuthman⁴, Charles Brown⁵, Fran Webster⁵, and Art McElroy¹ ¹ Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Food Canada, Ottawa, ON K1A 0C6; ² USDA-ARS, National Small Grains Germplasm Research Facility, Aberdeen, ID; ³ Dept. of Crop Science and Plant Ecology, U. of Saskatchewan, Saskatoon, SK; ⁴ Dept. of Agronomy and Plant Genetics, U. of Minnesota, St. Paul, MN; ⁵ The Quaker Oats Company, Barrington IL
Quantitative trait loci (QTLs) detected in adapted populations may ultimately be useful for improving germplasm via marker-assisted selection. An F₆ recombinant inbred oat (Avena sativa L.) population, "T/M", was developed from a cross between a hulless line and an elite covered seeded line. One-hundred and one T/M lines (59 covered and 42 hulless) were grown in several environments since 1992. Fourteen agronomic and quality traits were evaluated in trials at Aberdeen, ID, Rosemount and St. Paul, MN, Ottawa, ON, and Saskatoon, SK. The linkage map used for QTL analysis includes 193 RFLP, RAPD, and AFLP loci. These are organized into 35 linkage groups and 40 unlinked loci. For some traits, QTL analysis was performed separately on the covered and hulless portion of the population or only on the covered lines. Simple interval mapping was utilized for the QTL analyses (LOD =2.9). In total, 27 QTLs were identified affecting heading date, plant height, lodging, tertiary kernels, grain yield, kernel weight, test weight, thin and plump kernels, and groat beta-glucan, oil, and protein content. Two QTL regions affecting plant height and grain yield appear to correspond to QTLs or major genes identified in other oat populations. Many of the QTLs affecting kernel weight, % thin and % plump kernels were concentrated in the same genomic regions within T/M. There were several lines in this population with improved groat percentage compared to the covered parent. This would indicate that groat percentage QTLs exist in this population and further mapping efforts may identify these loci. Marker-trait relationships identified in this QTL study are being applied to marker-assisted selection projects to evaluate this strategy for developing high quality oat lines with superior agronomic performance.

This research was made possible by generous financial support from the Quaker Oats Company, and by collaborative dialogue and technical assistance from a consortium of Quaker funded institutions.

P - 13 DIFFERENCES AMONG PRIMARY, SECONDARY, AND TERTIARY OAT (AVENA SATIVA L.) KERNELS FOR MORPHOLOGICAL PARAMETERS AND GROAT PERCENTAGE David De Koeyer, Jitka Deyl, and Art McElroy Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Food Canada, Ottawa, ON, Canada, K1A 0C6
Improving groat percentage is a high priority for developing superior oat cultivars for both feed and milling markets. A sample of oat grain is comprised of sub-populations of primary, secondary, and sometimes tertiary kernels. The objectives of this study were to determine the variation that exists for the "kernel population" characteristics in elite breeding material and to assess the relationship between these parameters and groat percentage. Twenty cultivars and breeding lines were grown at Ottawa, ON in 1997. Panicles were randomly selected and the primary, secondary, and tertiary kernels were manually separated, counted, and weighed. Groat percentage was obtained for samples of the three kernel types. The proportion of each kernel type was significantly different among the lines tested, based on number of kernels and total weight. Germplasm with higher groat percentage was characterized as having high primary groat percentage and a smaller difference in groat percentage between the primary and secondary kernels. Only two lines had greater than 80% groat values for secondary kernels, a plateau that appears to exist for the other lines in this study. Improved selection efficiency may be obtained by preferentially selecting for groat percentage of primary kernels, the largest (61.0%) component on a total weight basis. Only 3 lines in this study exceed 75% for primary groat percentage, whereas, only 3 lines did not exceed 75% for secondary groat percentage. Increasing the proportion (weight) of secondary kernels relative to the primary kernels may also be an effective way of increasing groat percentage. This study has helped to identify some important factors that may be used to help develop varieties with higher groat percentage.

P - 14 ANTIOXIDANT ACTIVITY AND PHENOLIC ANTIOXIDANT CONTENTS OF OAT FRACTIONS Cheryld L. Emmons¹, David M. Peterson^1,2, Greg Paul³, Jeffrey B. Blumberg^4,5, Garry J. Handelman^4,5, Ronald L. Prior⁵, and Guohua Cao⁵ ¹Cereal Crops Research Unit, Agricultural Research Service, USDA, Madison, Wisconsin; ²Department of Agronomy, University of Wisconsin, Madison, Wisconsin; ³The Quaker Oats Company, Barrington, Illinois; ⁴Tufts University, Boston, Massachusetts; and ⁵Jean Mayer USDA Human Nutrition Research Center on Aging, Boston, Massachusetts
The potential of oat fractions to contribute to dietary antioxidants was evaluated. Antioxidant activity was measured in methanol extracts of flour, pearlings, bran fines, and trichomes from groats by 1) inhibition of coupled autoxidation of linoleic acid and -carotene (AOA), 2) inhibition of copper and peroxyl radical-induced low-density-lipoprotein oxidative damage (LDL-C, LDL-P, respectively) (Trolox equivalents), and 3) oxygen radical (peroxyl and hydroxyl) absorbance capacity (ORAC-P and ORAC-H, respectively) (Trolox equivalents). Pearlings had greater AOA (62-78% inhibition) than flour (53%), bran fines (57%), or trichomes (47%). Pearlings had greater ORAC-P (5.81) and ORAC-H (6.34) than flour (3.40, 3.70), bran fines (2.71, 1.80), and trichomes (2.74, 1.11). LDL-P and LDL-C were greatest for pearlings (3.82 and 3.42, respectively), followed by trichomes (1.54, 1.87), flour (1.26, 0.55), and bran fines (1.28, 0.46). AOA was correlated to LDL-C (r=0.75).

Total phenolic content (TPC), determined using the Folin-Ciocalteau phenol reagent, was expressed as gallic acid equivalents. Pearlings contained the greatest TPC (318 mg/kg), followed by flour (156 mg/kg), trichomes (139 mg/kg), and bran fines (107 mg/kg). TPC was correlated to AOA (r=0.85), ORAC-P (r=0.80), and LDL-C (r=0.93). Phenolic compounds were separated by HPLC on a C₁₈ column using a linear 1-40% acetonitrile gradient (pH 2.8). Peaks were identified by comparing retention times and UV spectra with known standards. Tannic acid, p-hydroxy-benzoic acid, vanillic acid (VA), caffeic acid (CA), syringic acid, vanillin, p-coumaric acid, and ferulic acid were identified and quantified in each sample. AOA was correlated to VA and CA contents (r=0.84 and r=0.79, respectively). LDL-C was correlated to CA (r=0.60). Three peaks correlated to AOA and LDL-C have spectra similar to hydroxy cinnamic acids and retention characteristics similar to avenanthramides (AV). They have been tentatively identified as AV1, AV3, and AV4 (r>0.83 and r>0.54 for AOA and LDL-C, respectively). A fourth peak, with a spectrum similar to catechin (a flavan-3-ol), has been denoted F3OL (r=0.80 and r=0.56 for AOA and LDL-C, respectively). Removing trichomes from the correlation analysis increased coefficients for some of the identified compounds, suggesting 1) oat fractions contain many phenolic compounds that contribute to antioxidant activity, and 2) trichomes have potential for good activity, but may contain interfering pro-oxidative compounds.

These data indicate a potential for oat products, especially those enriched in the outer layers of the kernel, to significantly contribute to dietary intake of antioxidant phytonutrients. Analysis of a wider selection of genotypes, fractions, and products is planned to determine the full potential of these materials as sources of dietary antioxidants.

P - 15 PARACENTRIC INVERSIONS IN F₁ HYBRIDS OF OAT CULTIVARS George Fedak, Nick Tinker, L. O'Donoughue, Vern Burrows, Ken Armstrong, Steve Molnar Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Food Canada, Ottawa, Ontario, Canada K1A 0C6
The presence of duplicated regions in the oat genome as detected by RFLP markers and other anomalies in segregation patterns of parental genotypes in segregating populations has prompted a study of meiotic chromosome behaviour in F₁ hybrids that gave rise to the mapping populations. A population obtained from a Hinoat × Newman F₁ hybrid showed an excess of Newman type parental gametes (L. O'Donoughue, p.c. N. Tinker p.c.)

Chromosome behaviour was examined at meiosis in F₁ hybrids derived from cultivars Tibor × Marion, Hinoat × Newman, Marion × Tibor and Hinoat × Marion. A very low frequency of quadrivalent configurations was observed in most plants of most hybrid combinations. A-I bridges were observed at frequencies of 30 - 49% in Hinoat × Newman and 7% in some but not all Tibor × Marion F₁ hybrids. In meiocytes where bridges occurred, single bridges were most common but up to five bridges were seen in some cells. The single A-I bridge configurations were the result of single crossovers or 3 strand double crossovers in the inverted segments in inversion heterozygotes. Double bridges were the result of four strand double crossovers in the inversion loops. The expected free fragments that accompany a bridge configuration were rarely seen at A-I but were frequently present in the cytoplasm of telophase I. There was some decline in proportion of stainable pollen in plants containing inversions.

The above observations indicate that the cultivars Hinoat and/or Newman contain paracentric inversions. Some of the consequences of the presence of inversions include asynapsis and nonhomologous pairing in regions adjacent to the inversion, thus causing reduced recombination in those regions. Chromatids with duplication and deficiencies would result from the recombination process leading to the production of like gametes. The viability of certain gametes could thus be affected. This concept requires additional study in other cultivar combinations.
P - 16 AN UPDATE OF THE MOLECULAR LINKAGE MAP OF CULTIVATED OAT Susanne Groh¹, Nick Tinker², Louise O'Donoughue³, Shahryar Kianian⁴, Stephen Fox⁵, Suzanne Livingston¹, Charlene Wight², Howard Rines⁶, Mark Sorrells⁷, Steve Molnar², George Fedak², Ken Armstrong², Ronald Phillips¹. ¹Dept. of Agronomy and Plant Genetics, University of Minnesota, St. Paul, MN 55108, USA; ²Agriculture and Agri-Food Canada, Research Branch, Eastern Cereal and Oilseed Research Centre, Ottawa, ON K1A 0C6, Canada; ³DNA Landmarks Inc., St. Jean Sur Richelieu, Quebec J3B 6Z1, Canada; ⁴Dept. of Plant Sciences, North Dakota State University, Fargo, ND 58105, USA; ⁵Crop Development Centre, University of Saskatchewan, Saskatoon, SK S6N 5A8, Canada; ⁶Plant Science Research Unit, USDA-ARS, and Dept. of Agronomy and Plant Genetics, University of Minnesota, St. Paul, MN 55108, USA; ⁷Dept. of Plant Breeding, Cornell University, Ithaca NY, 14853, USA
A molecular genetic linkage map of cultivated oat based on the cross Kanota (Avena byzantina) × Ogle (A. sativa) has been developed by a consortium effort. This map has provided the foundation for additional genetic studies and is a pivotal component of ongoing comparative-mapping efforts. We have continued to update this map since it was first reported by O'Donoughue et al. (Genome 38:368-380, 1995). Here, we report an updated version of the Kanota × Ogle linkage map. The former map was based on 561 loci mapped in 71 recombinant inbred lines (RILs). We have now mapped additional RFLP markers detected by cDNA clones from oat, barley, and wheat as well as probes from oat genomic DNA. After checking for duplicates and severe segregation distortion, data were available for 703 marker loci. The population has been extended to 137 RILs, in which 108 of the loci have been mapped to date. Markers were assigned to linkage groups at LOD 7. Some unlinked groups have been joined based on chromosome assignments determined by aneuploid studies (reported elsewhere), resulting in a total of 32 linkage groups. Marker ordering within linkage groups was performed for a subset of 259 loci providing a robust framework with a uniform marker density of approximately 5 to 10 cM. Most ordering within this framework was performed at LOD 3. Markers not included in the framework have been placed relative to this framework. The current size of this base map is 1770 cM excluding unlinked markers. Future work will include mapping of additional markers on the extended population, with emphasis on markers that will allow for comparisons with maps of other crop species. Updated versions of the map will be made available on the GrainGenes database (http://wheat.pw.usda.gov/).

P - 17 TERTIARY KERNEL EVALUATION IN OAT Cristine Handel¹, Melissa Huhn¹, Deon Stuthman¹ and Gary Fulcher² ¹Department of Agronomy and Plant Genetics, University of Minnesota; ²Dept. of Food Science and Nutrition, University of Minnesota
Seed physical traits, such as seed size and shape, are very important for oat processing industry. It is known that the number of kernels per spikelet is variable in oat (Avena sativa L.). Commonly there are only two kernels, but three and even four can be present. In most cases the primary kernel is bigger than the secondary one, and it is bigger than the tertiary kernel (TK). In studying TKs, a precise phenotypic rating of this trait is needed, especially if the data are used for molecular marker development. One way of rating TK is by looking at the top of the panicle of 5 plants in a row, and scoring zero (absence) or one (presence). This method is very fast but genotypes that are intermediate or even low for TK's might be misscored in some environments. Another way of rating TKs is by counting the number of florets and TKs in a panicle. This method is precise, but very time consuming. In our study these two methods were compared. In addition, the panicles were divided into three parts and TK numbers determined for each third of the panicle. The genetic material used for the study was composed of seven F6 pairs of iso-lines (high and low TK) planted in St. Paul MN (regular TK expression) and Aberdeen ID (high TK expression). Genotypes, locations, types of rating and the genotypeXlocation interaction were significant sources of variation. In both locations, the rating of the second third (middle of the panicle) gave the same results as rating the whole panicle for TKs, making the rating process faster but still adequately precise.

P - 18 AMPLIFIED DNA FRAGMENT LENGTH POLYMORPHISM BETWEEN OAT AND BARLEY MAPPING PARENTS D. L. Hoffman and J. E. McNeil USDA-ARS Small Grains Germplasm Research Facility, Aberdeen, Idaho 83210 USA
Genetic maps have become an important tool for understanding the basic genetics of crop plants. Detailed molecular maps for oat and barley have been constructed with the use of DNA restriction fragment length polymorphism (RFLP) markers. The use of amplified DNA fragment length polymorphism (AFLP) markers is an efficient method for mapping additional populations and adding more markers to existing maps. Before mapping with AFLP markers, it is essential to know which primer pairs yield numerous DNA amplification differences between parental cultivars.

The objectives of this study were to 1) identify which AFLP primer pairs will yield the most DNA amplification differences among oat mapping parents and 2) compare this information with that found for barley. Total genomic DNA was extracted from plants of oat mapping cultivars Kanota, Ogle, and TAM O-301, and from plants of barley mapping cultivars Steptoe and Morex. AFLP was conducted using protocols established by Gibco BRL Life Technologies, and the five cultivars were surveyed for potential AFLP markers with 62 EcoRI-MseI primer pairs.

The results indicated that some primer pairs are better than than others in producing AFLPs in both oat and barley, and that the conduct of a preliminary screen is helpful for efficient mapping. AFLP was more readily found in barley. Some adjustments to the procedure may need to be made to find more AFLP for oat. The next step will be to determine the actual number of AFLP markers with segregating populations. This information will be used to help construct an oat molecular genetic map in Ogle/TAM O-301 and increase the number of non-RFLP markers for barley.

P - 19 NINE CYCLES OF RECURRENT SELECTION FOR INCREASED GROAT OIL CONTENT IN OAT J.B. Holland and K.J. Frey Dep. of Agronomy, Iowa State University, Ames, IA 50010
Nine cycles of recurrent phenotypic selection for high groat (caryopsis) oil content were conducted in a genetically broad-based oat (Avena sativa L.) population including germplasm from the wild relative A. sterilis. The objectives of the experiment were to determine if selection for high groat oil content was effective throughout all nine cycles and if selection reduced genetic variation for oil content; to identify agronomically acceptable lines with high groat oil content; and to document the effects of selection on grain yield, groat fraction, or oil yield. We evaluated 100 random lines from the base (C0) population and each of the nine selection cycle populations in three environments. Mean groat oil content increased linearly over cycles of selection at a rate of 6.6 g kg^-1 per cycle (equivalent to 6.7% of C0 population mean per cycle) from a mean of 98.2 g kg^-1 in C0 to 158.5 g kg^-1 in C9. The rate of gain from selection did not decrease in later cycles. Oil yield increased at a rate equivalent to 1.9% of C0 population mean per cycle, while mean groat fraction did not change and mean grain yield decreased at a rate equivalent to 3.2% of C0 population mean per cycle. For all traits, genetic variation did not decrease over cycles. Selected S₀-derived lines from the population with groat oil contents greater than 150 g kg^-1 exhibited agronomic deficiencies such as lower yield and greater lodging and disease susceptibility compared to the best check cultivar.

P - 20 COMBINED AFLP AND RFLP MAPPING IN TWO HEXAPLOID OAT RECOMBINANT INBRED POPULATIONS H. Jin¹, L. L. Domier^1,2, F. L. Kolb¹, and C. M. Brown¹ ¹Department of Crop Sciences, University of Illinois, 1102 South Goodwin Ave., Urbana, IL 61801 USA; ²United States Department of Agriculture, Agricultural Research Service, Crop Protection Research Unit, 1102 South Goodwin Ave., Urbana, IL 61801 USA
The segregation of AFLP markers was analyzed in two different hexaploid oat recombinant inbred populations. The first population was derived from a cross of Kanota x Ogle and consisted of 71 recombinant lines, which were previously used to produce an RFLP map of the hexaploid oat genome. The second population was derived from a cross of Clintland64 and IL86-5698, barley yellow dwarf virus sensitive and tolerant oat lines, respectively. One hundred twenty-six lines of this population were used to identify loci contributing to tolerance to barley yellow dwarf virus infection. More than 300 AFLP markers were scored in each of the populations, of which 85 could be scored in both populations. These markers were linked to RFLP markers in 20 of 38 Kanota x Ogle RFLP linkage groups. The addition of the 306 AFLP markers to the Kanota x Ogle RFLP data set reduced the number of linkage groups to 33 at a LOD of 8.0 and a recombination fraction of 0.3. Most of the markers that showed distorted segregation in one population were more normally distributed in the other. All but eight of the AFLP markers scored in the Kanota x Ogle population were linked to previously mapped RFLP markers. Thirty six linkage groups were observed in the Clintland64 x IL86-5698 population. Two of the linkage groups from this population were consolidated by comparing the segregation of markers in both populations. Markers were linked in similar orders in the two populations and showed a very similar distribution to that of the RFLP markers.

P - 21 GENETIC ENGINEERING OF MODEL AND ELITE OAT VARIETIES WITH ANTIFUNGAL GENES USING A NOVEL SELECTION SYSTEM Heidi F. Kaeppler, Geeta K. Menon, Ronald W. Skadsen, and Anna M. Nuutila. Department of Agronomy, University of Wisconsin, Madison, WI, 53706, and USDA-ARS Cereal Crops Quality Unit, Madison WI, 53706
Genetic engineering can be utilized in oat as a complimentary tool to breeding to incorporate genes encoding novel or altered traits for improved grain quality or agronomic performance. Limitations in oat transformation include low transformation frequencies, difficulty in application of current transformation systems to elite oat varieties, and few choices of marker genes for efficient selection of transformants. The objectives of our research were to: 1. develop a new selection system based on visual selection for expression of the green fluorescent protein gene, and 2. transform the highly culturable oat genotype, GP-1, and elite oat varieties Belle, Dane and Gem, with transgenes encoding antifungal proteins using the new gfp-based selection system. The gfp expression construct and the antifungal transgene constructs were codelivered into friable, embryogenic oat tissue cultures via microprojectile bombardment. Transgenic cell lines of GP-1, Belle, Dane and Gem were recovered following visual selection for gfp-expressing cells. Fertile, transgenic plants were regenerated from all selected cell lines of GP-1. Transgenic cell lines of Belle, Dane and Gem are currently on regeneration medium. Strong expression of gfp was observed in callus, regenerated plant tissues and transgenic seedlings. Southern blot and PCR analysis confirm integration of the transgenes in primary regenerants and seedling progeny. The period of time between initiation of selection and placement of cells on regeneration medium was 4-5 weeks, compared to 8-12 weeks when selection was based on antibiotic resistance. Analysis of transgenic plants and progeny for expression of the antifungal proteins is underway.

P - 22 NUCLEAR MAGNETIC RESONANCE IMAGING FOR DETERMINING THE DISTRIBUTION OF OIL IN OAT (AVENA SATIVA L.) GROATS S. Kibite, H. Grondey and A. Blazek Agriculture & Agri-Food Canada, Lacombe Research Centre Lacombe, Alberta, Canada.
Nuclear magnetic resonance (NMR) imaging is an extension of NMR spectroscopy where the NMR signal is resolved by the application of 3-dimensional magnetic field gradient. NMR imaging was used in the present study to determine the feasibility of using this technology to examine the distribution of oil in the oat groat, and also to elucidate any differences that may exist among high- and low-oil genotypes. This information would be useful to plant breeders who may be interested in breeding new cultivars for various end-use applications.

Three oat genotypes, Waldern, LAO-505-001, and N326-7 were selected to give a range of oil contents from less that 6% to more than 16% on a dry groat-weight basis. The three genotypes were grown at the same location and in the same growing season to eliminate environmental differences, which have been shown to affect oil content in oat. A Brunker mico-imaging probe equipped with a 10-mm horizontal coil was used for imaging the groats. When a short TE (time to echo) was used, both the oil and water components of the groats were visible, and the resulting images were blurred due to the fact that the oil and water signals were offset slightly from each other. A longer TE dephased the signals from the water component and produced sharper images. The length of TE required to dephase the water signals was 23 ms for Waldern, 29 ms for N326-7 and 50 ms for LAO-505-001. These TE values appeared to be inversely related to the moisture content of the samples since the Waldern, N326-7 and LAO-505-001 groats contained 13.1, 10.1 and 9.5% moisture, respectively. In all three genotypes, the oil component appeared to be localized in the germ as well as in the periphery (presumably the bran fraction) of the groat, with very little oil present in the endosperm. Although the amounts of oil in the germ and bran fractions of each genotype were not quantified, it was evident from visual observation of the images that the major differences among high, medium- and low-oil genotypes were due to the amounts of oil that were present in the germ. These observations indicate that NMR imaging can be a useful indicator of the distribution of oil in the groat, and can aid plant breeders in cultivar selection for specific end-uses.

P - 23 GENETIC DIVERSITY FOR ALUMINUM TOLERANCE IN AVENA SPP S. Kibite. Agriculture & Agri-Food Canada, Lacombe Research Centre, Lacombe, Alberta, Canada.
Aluminum (Al) toxicity is a major factor limiting oat (Avena sativa L.) production in the Grey-Wooded soils of western Canada. Progress in breeding for Al tolerant oat cultivars has been limited by a lack of suitable sources of genes for Al tolerance. In an effort to find new genes for Al tolerance, ~3,500 germplasm accessions for the USDA World oat Collection were evaluated for Al response in the greenhouse using a modified hydroponics system. Accession lines were grown in Al-toxic (300 M of Al, pH 4.5) and Al-free (0 M of Al, pH 6.5) nutrient culture solutions. Ten seedlings from each treatment combination were harvested 14 days after seeding, and the roots measured to assess the level of Al tolerance in each accession. Using this method, 11 Al tolerant lines (5 A. strigosa, 4 A. brevis, 1 A. fatua and 1 A. abyssinica) were identified. Subsequent greenhouse experiments revealed that these 11 accession lines were more tolerant than previously reported Al tolerant genotypes, and suggested that they may carry new genes for Al tolerance.
P - 24 THE INHERITANCE OF ALUMINUM TOLERANCE IN THREE OAT CROSSES S. Kibite Agriculture & Agri0Food Canada, Lacombe Research Centre, Lacombe, Alberta, Canada
Aluminum (Al) toxicity is a major problem in oat (Avena sativa L.) grown in acid soils. Breeding for Al tolerance (AT) is hampered by a lack of information on the inheritance of this trait. Three experiments were conducted in a greenhouse to determine the mode of inheritance of AT in oat. The results showed that AT is a heritable trait and that it is governed by nuclear genes. Data from two experiments exhibited inheritance patterns consistent with those expected for a single gene with complete dominance for the tolerant reaction. Data from the third experiment showed a more complex inheritance suggesting that several genes may be involved in the expression of AT. Broad-sense heritability estimates in the three experiments ranged from 0.23 to 0.81.
P - 25 QTL MAPPING OF GROAT QUALITY TRAITS IN OAT: ACETYL-COA CARBOXYLASE, A CANDIDATE GENE FOR OIL CONTENT Shahryar F. Kianian¹, Susanne Groh², Margaret A. Egli², Ronald L. Phillips², Howard W. Rines³, David A. Somers², Burle G. Gengenbach², Francis H. Webster⁴, Suzanne M. Livingston², Louise S. O'Donoughue⁵, Mark E. Sorrells⁶, Darrell M. Wesenberg⁷, Deon D. Stuthman², and R. Gary Fulcher⁸

¹Dept. of Plant Sciences, North Dakota State University, Fargo ND 58105, USA
²Dept. of Agronomy and Plant Genetics, University of Minnesota, St. Paul MN 55108, USA
³Plant Science Research Unit, USDA-ARS, and Dept. of Agronomy and Plant Genetics, University of Minnesota, St. Paul MN 55108, USA
⁴The Quaker Oats Co., John Stuart Research Laboratories, Barrington IL 60010, USA
⁵DNA Landmarks Inc., St. Jean Sur Richelieu, Quebec, J3B 6Z1, Canada
⁶Dept. of Plant Breeding, Cornell University, Ithaca NY 14853, USA
⁷USDA-ARS, National Small Grains Germplasm Research Facility, Aberdeen ID 83210, USA
⁸Dept. of Food Science and Nutrition, University of Minnesota, St. Paul MN 55108, USA

Genetic manipulation of groat quality traits such as oil and beta-glucan content is desired for optimizing the value of oats in human and livestock diets. The localization and characterization of quantitative trait loci (QTLs) for these traits is necessary for identifying genetic markers to be used for marker-assisted selection (MAS) to improve the quality of oat groats. Two populations of recombinant inbred lines derived from the crosses Kanota × Ogle (K×O) and Kanota × Marion (K×M) were genotyped using RFLP markers and tested for oil and beta-glucan content across several locations and years. QTL analysis was conducted using composite interval mapping with 350 RFLP loci in K×O and single factor ANOVA with 60 loci in K×M. A major QTL affecting oil content was found consistently across environments in both populations on linkage group 11, explaining up to 43% of the phenotypic variance. A second locus was found in each population, explaining an additional 10 to 20% of the phenotypic variance. A cDNA clone encoding oat acetyl-CoA carboxylase (ACCase) mapped to the same region as the major QTL on linkage group 11 in both populations. ACCase, which catalyzes the first step in de novo fatty acid synthesis, is therefore likely to play a major role in determining groat oil content and provides an excellent marker for MAS. Furthermore, the genomic region around ACCase also showed an effect on beta-glucan, indicating that selection for this locus could simultaneously influence oil and beta-glucan content. However, the allele from Kanota increased the numerical value of both traits in K×O, while in K×M the allele from Kanota increased oil and the allele from Marion increased beta-glucan content. Additional research will be needed to gain a better understanding of the role of ACCase in determining groat oil content and possible effects on correlated traits.

P - 26 THE EFFECT OF RIPENESS AND HARVESTING TECHNIQUE ON THE GERMINATION OF NAKED OAT IN NORTHERN GROWING CONDITIONS Anna-Maija Kirkkari¹, Ari Rajala², and Pirjo Peltonen-Sainio³ ¹Department of Agriculture, TTS-Institute; ²Department of Plant Production, University of Helsinki; ³Department of Plant Production, University of Helsinki
In Finland, oat is the most cultivated cereal crop after barley (370 000 hectares, 34 % of the grain crop cultivation area in 1996). Presently, all of the cultivated oat in Finland is conventional hulled oat. The interest in naked oat is however increasing, due to its high quality grain material for various industrial purposes and also due to possibilities to utilise naked oat in novel healthy food production. In Scandinavian growing conditions, grain is most often harvested at moisture content between 15 - 30 % . Combine harvesting at high moisture content, exposes the naked grain to mechanical damages. After harvesting grain is dried in grain circulating, batch type dryers. Heavy mechanical handling of naked grain during harvesting and drying especially at high moisture levels often result in reduction in germination. Poor germinating grain material is unacceptable for seeding or for industrial processing that requires high germination.

The aim of this study was to optimise the harvesting procedure of naked oat in order to minimise grain damage and loss of germination. A practical guide, i.e. 'harvesting recipe' will be produced for the benefit of farmers interested in this new crop. In farm scale trials, supported with laboratory tests, naked oat will be harvested and dried with conventional machinery to establish the effect of ripeness and various grain moisture on the germination. Hand threshed grain material is used as reference. The trials are carried out in 1998-2000.

In preliminary trials, highest germination has been achieved when harvested at 19% moisture content. In field trials, the optimal moisture range will be established in relation to ripening and weather conditions.

In conclusion, naked oat is an attractive new crop in Finland for food and feed processing industry. Prior to introduction to the farmers, the especially problems related to harvesting and drying need to be solved.

P - 27 DEVELOPMENT OF BARLEY YELLOW DWARF VIRUS TOLERANT OAT GERMPLASM LINES F.L. Kolb¹, C.M. Brown¹, L.L. Domier ^2,1, and N.J. Smith¹ ¹Department of Crop Sciences, University of Illinois at Urbana-Champaign; ² USDA-ARS
A population from a four-way cross was used to develop barley yellow dwarf virus (BYDV) tolerant oat germplasm lines. Our objective was to combine genes for tolerance to BYDV into germplasm lines with a higher level of tolerance to BYDV than the parents. The four-way cross involved four BYDV tolerant parents: IL86-1156, IL86-5698, IL86-6404 and Ogle. The F₃ population was space-planted in the field in 1993 and inoculated with BYDV-PAV-IL when the plants were at Feekes GS 1. Plants that exhibited BYDV symptoms were destroyed. About 780 of the most tolerant plants (based on lack of symptoms) were harvested individually. In 1994, a single hill of each line was evaluated for BYDV tolerance in a BYDV-PAV-IL inoculated nursery, and 139 lines were selected. The 139 lines (plus the parents and checks) were evaluated in 1995 and 1996 using three replications of paired control and BYDV-PAV inoculated hills. Because these lines exhibit little or no BYDV symptoms, we evaluated BYDV tolerance based on virus titer using ELISA, percent stunting (height difference between control and inoculated hills), and percent yield loss (grain yield difference between control and inoculated hills). In 1996, 62 lines were evaluated for agronomic performance in a replicated experiment using 6 row plots at one location. In addition to BYDV tolerance, lines were selected for further evaluation based on grain yield per se, kernel morphology and absence of awns. In 1997, 42 lines were evaluated further in replicated trials at two locations. Based on all of the criteria used for selection, 7 lines are being increased for release as BYDV tolerant germplasm lines. These lines differ somewhat in height, maturity and ELISA virus titer. All of the lines have excellent tolerance to BYDV-PAV.

P - 28 NIRS PREDICTION OF GROAT PROTEIN AND OIL CONTENT USING WHOLE OAT (AVENA SATIVA L.) SAMPLES A.R. McElroy and D.T. Grégoire Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Food Canada, Ottawa, Ontario. K1A 0C6
Near infrared reflectance spectroscopy (NIRS) is routinely used for quality analysis of grains and oilseeds. Quality analyses are made on whole oat samples for feed purposes and on groats for milling purposes. Therefore, milling oat samples must be dehulled before quality analysis. NIRS prediction equations were developed to determine the protein and oil content of the groat using whole oat samples.

Groat samples, prepared by Codem dehulling, were evaluated for protein (Kjeldahl N x 6.25) and acid hydrolysis fat. Initial equations were developed from a 100 sample population from Ottawa and Elora, Ontario, and then enhanced with a 154 sample set from Saskatchewan (11 cultivars x 14 sites). The per cent groat ranged from 60 - 79 % (Ontario) and from 63 - 80% (Saskatchewan). The mean groat protein was approximately 15% for both locations (11 - 23% overall), while the mean groat oil content was 8.5 (Ontario) and 9.9 % (Saskatchewan) with an overall range from 5 - 13.5%.

Comparisons of the standard error of prediction (SEP), slope and ratio of standard deviation of predicted values to SEP (RPD) were made between two techniques: 1) whole oat prediction equations developed here, and 2) bias-corrected predictions using Quaker Oat universal groat equations.

The SEP using whole oat and groat equations were similar for protein (0.51 vs 0.68) and oil (0.27 vs 0.31) for the Ontario samples. Slope and RPD values were similar for both techniques. The SEP for the Saskatchewan samples, using the enhanced quation, was 0.72 for protein and 0.49 for oil. It is concluded that the protein and oil content can be predicted as accurately using whole oat equations as with traditional groat equations, despite a wide variation in the groat content of the samples.

P - 29 CURRENT STATUS AND PERSPECTIVES OF OAT BIOTECHNOLOGY IN SOUTHERN BRAZIL S.C.K. Milach UFRGS, Depto. Plantas de Lavoura, C.P. 776, Porto Alegre, RS, Brazil, 915011-970, milach@vortex.ufrgs.br
There are two main groups in southern Brazil that have been conducting research in oat (Avena sativa L.) biotechnology. One is at Federal University of Rio Grande do Sul (UFRGS) and the other at University of Passo Fundo. In the later, the main effort has been in using tissue culture to generate somaclonal variants. Several lines regenerated in vitro have shown variation for agronomic traits and one of these has been selected and tested in the official yield nursery; but has not been promoted to a variety. At UFRGS until recently the main effort was also in using tissue culture to select for new variation. In the last few years, however, this program has been expanded. In this paper we will present the current efforts and perspectives of oat biotechnology at UFRGS. Thirty-five oat genotypes grown in the Southern Brazilian environments are being characterized at the molecular level using RAPD markers. These genotypes were chosen because they represent a broad range of variability and have traits of interest to be mapped with markers in the future. Partial resistance (PR) to oat crown rust has being identified in the germplasm of UFRGS. Through a collaborative research project with Dr. Greg Penner from the Cereal Research Centre of the Agriculture and Agri-Food Canada Institute, current work is being conducted on mapping PR to crown rust using AFLP markers in one population from a cross between two UFRGS lines. The perspective on this project is to convert some of the AFLP markers identified as associated to PR to Amplicon Specific Alleles to screen oat lines for PR in the UFRGS breeding program. Finally, the development of a system to transform Brazilian oat lines is underway with the objective to obtain resistance to spot blotch, which is not available at desirable levels in the oat germplasm.

P - 30 VARIABILITY FOR OAT KERNEL MORPHOLOGY IN BRAZILIAN GENOTYPES AND ITS RELATION TO MILLING YIELD S.C.K. Milach, C.R.A. Bothona, L.M. Tisian, M.C.C. Teixeira UFRGS, Depto. Plantas de Lavoura, C.P. 776, Porto Alegre, RS, 915011-970, Brazil
Kernel morphology of oat (Avena sativa L.) Brazilian genotypes has been studied in the last few years at the Federal University of Rio Grande do Sul, in attempt to identify germplasm with superior quality. The objective of this paper is to present the main results of the studies carried out to characterize the genetic variability, to access the effect of years in the expression of kernel morphology, and to identify kernel traits possibly associated to milling yield. Nine genotypes were grown in 1995 and 1996 in a complete randomized experiment with four replications. Five panicles per replication were harvested and primary, secondary and tertiary kernels per panicle were separated and counted. A sample of 50 kernels of each type per genotype and replication was prepared and evaluated using digital image analysis (DIA). Kernel traits evaluated were area, length, width, perimeter and shape factor. In 1996, the 5 meters long and wide plots were harvested after the five panicles were taken; then test weight, milling yield, weight of a thousand kernels and grain yield were evaluated. Variability among oat genotypes and the interaction genotype x year were significant for all kernel morphology traits. However, the top (UFRGS911747) and bottom (UFRGS7) genotypes have been the same in both years for all traits. Significant correlation was not found between kernel morphology and milling yield traits. However, the genotype with the best milling yield, CTC3, did not have tertiary kernels and had high test weight. The results obtained so far indicate that morphology per se is not enough to select genotypes with high milling yield in the Brazilian oat germplasm.

P - 31 STRATEGIES TO CHARACTERIZE PARTIAL RESISTANCE TO OAT CROWN RUST OCCURRING IN GENOTYPES GROWN IN THE SOUTHERN BRAZILIAN ENVIRONMENTS S.C.K. Milach¹, G.C.H Thome¹, G.O. Mellos¹, L.C. Federizzi¹, J.A. Martinelli¹, G.A. Penner² ¹ UFRGS, Depto. Plantas de Lavoura, Porto Alegre, RS, C.P. 776, 91501-970, Brazil; ² Cereal Research Center, Agriculture and Agri-Food Canada, Winnipeg, Canada
Oat (Avena sativa L) crown rust (Puccinia coronata f. sp. avenae) encounters favorable environments and has large variability in Southern Brazil. For this reason, a major challenge for the Oat Breeding Program at the Federal University of Rio Grande do Sul has been to breed genotypes with more durable resistance to crown rust. Thus, we have been searching for alternatives to identify genotypes with partial resistance (PR). The objective of this paper is to outline our approach in identifying and characterizing PR to crown rust in oat Brazilian genotypes and to present the main results obtained so far. The strategies used include: identification of traits and methods to evaluate PR for the Southern Brazilian environments; identification and characterization of oat lines for PR; verification that PR will hold through different years; determination of the PR genetic basis in selected lines; and molecular mapping of quantitative trait loci (QTL) associated to PR. A digital image analysis procedure was developed to evaluate leaf area infected with rust. Other traits evaluated were pustule size and area under the disease progress curve. The results indicated that a more precise analysis was obtained by using the data of all the PR components. Twenty-seven oat lines were evaluated in 1995 using this criteria. Six of them were selected for the genetic studies and 15 populations developed from crosses among them. The diallel analysis for pustule size and leaf area affected by crown rust revealed that additive gene action was most important for both traits. The genotype UFRGS91906 was identified as one with the best General Combining Ability for components of PR and its phenotypic data was consistent for three years. The segregating population from the cross UFRGS7 x UFRGS91906 is being used to construct a molecular map for PR QTL identification.

P - 32 SCAR MARKERS FOR RUST RESISTANCE GENES PC68, PG3 AND PG9 DESIGNED FOR MARKER ASSISTED SELECTION IN OATS W. Orr, D. De Koeyer, C. Chenier, N. Tinker and S. J. Molnar Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Food Canada, Ottawa, Ontario K1A 0C6
Crown rust, caused by Puccinia coronata and stem rust, caused by P. graminis, are important diseases of oat (Avena sativa L.). The resistance genes Pc68, Pg3 and Pg9 form a single tightly linked cluster. To facilitate introgression of these genes into other germplasm, we converted strategic RAPD markers to PCR based Sequence Characterized Amplified Region (SCAR) markers. One RAPD marker (UBC195) has been linked to Pg3 in Rodney O-Pg3 and to Pg9 in OT328 x Dumont. The diagnostic 0.4kb fragments were sequenced. A primer pair designed from the Pg3 sequence consistently amplifies a band on the susceptible line (Rodney O) but not on the resistant line (Rodney O-Pg3). The resultant 0.4 kb fragment is digested by Rsa I to 0.24kb in some lines (example OT966-1 which carries Pc68) and to 0.2kb in others (example OT974-1 which carries Pc48) such that homozygotes and heterozygotes of such crosses can be distinguished. A second primer pair was designed from the sequence of the fragment produced by UBC195 from Dumont (Pg9). A single 0.4kb band is produced on Dumont (resistant) but not on OT328 (susceptible). This fragment is also amplified in lines known to carry Pc68 (OA966-1, AC Francis and OT274) and in OA974-1 (Pg9?) but not in genotypes with neither gene (OT328, OA971-2 and OA971-8). These "user-friendly" PCR based SCAR markers are currently being employed in breeding programs to expedite selection for rust resistance.

This research was made possible by generous financial assistance from Quaker Oats and by collaborative dialogue and technical assistance from a consortium of Quaker funded institutions.

P - 33 DEVELOPMENT OF A GENOMIC MAP OF CULTIVATED OAT IN TAM-0-301 X OGLE RIL POPULATION V. Portyanko¹, D. Hoffman², M. Lee¹, and J. Holland¹ ¹Department of Agronomy, Iowa State University, Ames, IA 50011; ²USDA-ARS Small Grains Research Facility, Aberdeen, ID 83210
Complete genetic maps have become an essential component of progress in crop improvement and basic genetics. The first RFLP map of the cultivated oat genome (n=21) involved 38 linkage groups and was based on Kanota x Ogle RIL population (O'Donoughue et al., 1995).

The major objectives of this study are 1) to develop a cultivated oat RFLP map with the number of linkage groups corresponding to the basic number of chromosomes; 2) to connect this map to genome maps of related grass genera; 3) to use this map for experiments to map QTLs affecting vernalization and photoperiod responses, disease resistance, and grain yield under diverse conditions.

To date 292 cDNA clones (including a core set of 175 anchor probes being mapped in other cereals) were screened in combinations with 6 restriction enzymes to identify polymorphisms between mapping parents, Avena byzantina, cv. TAM-0-301 and Avena sativa, cv. Ogle. About two thirds (or 215) of a total set of probes screened were found to be informative in combination with at least one enzyme. Mapping of these clones inTAM-0-301 x Ogle population of 136 F_6:7RILs is underway.

A number of important developmental and agronomic traits were scored in the mapping population to map underlying QTLs/genome regions. These include heading date and plant height under different vernalization and photoperiod regimes; heading date, grain yield, and test weight in the field under different planting dates and environments.

P - 34 OAT PLANTS WITH INTROGRESSED MAIZE CHROMOSOME SEGMENTS FROM GAMMA IRRADIATION OF PROGENY OF OAT X MAIZE CROSSES O. Riera-Lizarazu¹, M. I. Vales¹, E. V. Ananiev¹, H. W. Rines², and R. L. Phillips¹ ¹University. of Minnesota; ²USDA-ARS, St. Paul, MN.
Crosses of oat (Avena sativa L., 2n = 6x = 42) by maize (Zea mays L. 2n = 2x = 20) have resulted in the recovery of both haploid oat plants (2n = 21) and plants with 21 oat chromosomes plus one or more retained maize chromosomes. From these, self-fertile disomic addition lines (2n = 42 + 2) with a full complement of 21 oat chromosome pairs plus an added maize chromosome pair were recovered for 6 of the 10 maize chromosomes. The phenotypes of these oat-maize chromosome addition lines ranged from that of near normal oat plants (e.g., maize chromosome 4 additions) to quite marked effects on vigor and plant morphology (e.g., maize chromosome 3 addition). The added maize chromosome represents a potential source of new traits for oat such as disease resistance or seed composition modification. In addition, for maize genomics analyses, the addition lines represent a valuable source for mapping or isolating maize DNA sequences specific to a particular maize chromosome. Irradiation of these whole chromosome addition lines to induce the transfer of maize chromosome segments into the oat genome was seen as a way to further reduce the amount and complexity of maize chromatin present. For this purpose, we treated oat maize chromosome 9 monosomic seed with gamma rays (30, 40, and 50 k rad). Progeny plants from self-fertilization of plants from irradiated seed were screened for the presence of maize DNA by Southern analysis using a multiprobe of a mixture of maize-specific repeated sequences. The maize chromatin present was characterized by fluorescence genomic in situ hybridization of root tip cells and by analyzing DNA extractions with a series of 34 maize RFLP and SSR markers specific for chromosome 9. Various maize chromosome 9 rearrangements in oat were identified including translocation introgressions of maize chromosome segments onto oat chromosomes.
P - 35 INDEPENDENT TESTS AND VALIDATION OF PUTATIVE -GLUCAN QTLS IN OAT V. JM. Robert¹, R. L. Phillips^1,2, and D. D. Stuthman¹ ¹ Department of Agronomy and Plant Genetics; ²Plant Molecular Genetics Institute, University of Minnesota, 411 Borlaug Hall, 1991 Buford Circle, St. Paul, MN 55108, USA
Several studies have demonstrated the hypocholesterolemic effect associated with the consumption of oat soluble fiber [(1-3), (1-4)--D-Glucans]. This effect can reduce the risk of heart disease in humans. The development of a genetic map in cultivated oat (Avena byzantina X Avena sativa) has permitted the study and mapping of quantitative trait loci (QTLs) associated with inheritance of soluble fiber. This study revealed a complex inheritance as many QTLs were detected and it appeared that further validation would be useful prior to initiating a marker assisted selection (MAS) breeding program for soluble fiber. The objective of this research was to test the effect of each QTL previously identified, independently of other QTLs. This research should allow validation and selection of QTLs with large effects on soluble fiber inheritance.

Recombinant inbred lines (RILs) of the mapping population with residual heterozygosity at specific soluble fiber QTLs were identified. Sets of near isogenic lines (NILs) were generated by isolating a series of selfed progenies from each RIL. This material was grown for two years and at several locations. Fiber contents were obtained through near infrared technology and chemical analysis. These NIL sets were segregating for specific molecular markers in the vicinity of specific QTLs in an otherwise homogeneous genetic background. Therefore, allelic substitutions through regression analysis were tested at each independent QTL. Moreover, pleiotropy and/or genetic linkage with seed or plant characteristics were sometimes detected as correlated responses were observed in some sets of segregating NILs. This method allowed validation of the soluble fiber QTLs and provided furhter understanding of soluble fiber inheritance.

P - 36 GROAT BREAKAGE Brian G. Rossnagel Crop Development Centre, University of Saskatchewan, 51 Campus Drive, Saskatoon, SK, S7N 5A8 Canada
Groat breakage during dehulling, which leads to increased mill loss and lower mill yield, has been a problem for millers from time to time for many years. Traditionally it has been associated with certain environmental conditions around crop harvest, especially frost. However, when reported as a major problem with Canadian shipments to the US in 1993 and 1994, since this also coincided with an increase in the portion of those shipments which would be the variety Derby, we became suspicious that a genetic component might be involved. As a result a project was set up to evaluate genetic differences for groat breakage.

Several studies have been undertaken to evaluate methodology to evaluate groat breakage and to determine if genetic differences do exist. One study comparing samples from 11 varieties over 14 locations in 1996 demonstrated that significant differences between varieties exist, with Derby demonstrating greater and Elvy less breakage than others. Environment was also a factor. Unfortunately, it has been determined that hand separation of broken from whole groats is the only accurate method to evaluate this trait, since sieving to separate broken from whole groats is influenced by original groat size. One year of data from a 90 genotype SSD population segregating for groat breakage indicates that the trait is normally distributed and thus probably under multi-genic control. While there is some association of breakage with lower fat, greater plumpness and thin hulls, these relationships are not absolute and with careful monitoring of breeding materials breeders should be able to select good milling quality oat with minimal groat breakage.

Special project funding from the Saskatchewan Agriculture Development Fund for this groat breakage evaluation project is gratefully acknowledged.
P - 37 THE AMMI MODEL FOR GROUPING ENVIRONMENTS AND OAT GENOTYPES Jose J. Salmerón INIFAP-Campo Experimental Sierra de Chihuahua, Cd.Cuauhtémoc, Chihuahua, Mexico 31500

A study was performed in ten environments under rainfall conditions during the oat growing seasons of 1992, 1993 and 1994. Ten oat genotypes (six lines and four varieties) were seeded in the above environments. The AMMI model was used to obtain homogenous groups of genotypes and locations and identify stable and high yielding genotypes. Genotypes were arranged in three categories: One of them with three oat varieties clasiffied as instable due to their high and negative PCA1 scores Papigochi and Chihuahua are late, and Páramo is an early cultivar, also this group was susceptible to stem and crown rust. A second group formed by five lines had a high and positive PCA1 scores, these were the high yield potential genotypes, intermediate maturity and resistant to oat rust and a third group constituted by two genotypes with PCA1 scores of nearly 0 has small interaction effects and hence classified as stable genotypes, these were a variety Cusi an early resitantat to one rust and other advanced intermediate maturity and resistant line. According to grain yield, two groups of environment were identified: High interaction genotype-environment Tepatitlán, Jal., and Pedernales 92, Chih. with high yield potential and Sto. Tomas, Chih with intermediate yield potential. Another group were locations with low interaction effects: Campo 1A, Cuauhtémoc, Chih. in 1993 with high yield potential, Calera and Sombrerete, Zac. with intermediate yield potential. The low yield potential locations were: Campo 1A(1994) , Campo 38(1993 and 1994), and Pedernales(1994) all within Chihuahua State. The AMMI model represents another technique to visualize the genotype by environment interaction.

P - 38 DEVELOPMENT OF A "SUN II" OAT ANEUPLOID SERIES FOR RFLP MARKER ASSOCIATIONS TO CHROMOSOME. M.A. Start¹, H.W. Rines², and R.L. Phillips¹ ¹University of Minnesota; ²USDA-ARS, St. Paul, MN.
The molecular genetic linkage map for cultivated hexaploid oat (2n=6x=42) constructed from an Avena byzantina cv. Kanota by A. sativa cv. Ogle cross using principally RFLP markers consists of at least 532 loci assigned to 38 linkage groups. Additional mapping efforts and the use of aneuploids have lowered the number of linkage groups to 33. The objective of this research is to reduce the number of linkage groups to 21, the expected number for hexaploid oat, by assigning linkage groups and unlinked markers to physical chromosomes using a nearly complete Sun II monosomic series. Twenty oat monosomics each missing a different chromosome have been identified based on C-banding analysis. Fourteen of the monosomics were newly derived from Sun II oat haploids generated in oat by maize crosses. Three of the monosomics were derived by Hacker and Riley in a different selection of Sun II and have a 3C-14 interchange. Backcrosses have been made in an attempt to eliminate the interchange and to standardize the background of the Sun II monosomics. Three monosomics have only been recovered in the cultivar Kanota, as described by Morikawa. Backcrosses with Sun II have been made with these three Kanota monosomics. Each monosomic line available was crossed with Kanota and Ogle. Monosomic F1s from these crosses were identified by mitotic chromosome counts of root tips, and leaf tissue was collected for DNA extraction. RFLP autoradiograms will be analyzed and missing bands will allow assignment of sequences to the monosomic chromosomes.

P - 39 CONSTRUCTION OF A GENETIC LINKAGE MAP IN HEXAPLOID OAT USING AFLP: SEARCHING FOR GENETIC MARKERS FOR PARTIAL RESISTANCE TO CROWN RUST G.C.H. Thome¹, G.A. Penner², A. Zacharias², G.O. Mellos¹, S.C.K. Milach¹, L.C. Federizzi¹ ¹ Federal University of Rio Grande do Sul - Brazil; ² Cereal Research Centre, Agriculture and Agri-Food Canada, Winnipeg
Crown rust (Puccinia coronata f. sp. avenae) is the most important fungal disease of oat (Avena sativa L) in southern Brazil. In susceptible cultivars, severe infections prior to the seed development can result in reduced quality and yield losses as high as 50%. Since the pathogen has been able to easily overcome the major resistance genes, more attention is currently being paid to other forms of resistance, such as partial resistance (PR). Screening for PR to oat crown rust and the analysis of its components in Brazilian oat genotypes started in the Oat Breeding Program at the Federal University of Rio Grande do Sul (UFRGS) in 1995. The oat line UFRGS 910906 was observed to have components of PR in the field and it has maintained its PR for over 3 years under high rust inoculum pressure. The number of genes controlling PR in UFRGS 910906 is not known. This is currently under study at UFRGS. Due to the difficulty of screening for PR in the field and the effects of environment on its expression, selection for PR by conventional methods is not always reliable. The ability to identify individuals with PR at the seedling stage using molecular markers is desirable since it would increase the speed and reliability of selection in breeding programs. This poster presents preliminary results of a study under way to develop molecular markers for PR to crown rust using Amplified Fragment Length Polymorphism (AFLP) analysis in a cross between UFRGS 910906 and the susceptible cultivar UFRGS 7. In a field trial of the F2 generation (200 individuals), continuous variation of PR to crown rust was observed. Different primer combinations are being tested on this F2 population in order to develop a genetic map providing markers linked to QTLs or genes that control PR. To date, 20 different primer combinations have been tested yielding a total of 150 polymorphisms segregating across the population.

P - 40 HOMEOLOGOUS OR DUPLICATED GENOMIC REGIONS AFFECTING GROAT OIL CONCENTRATION Nick Tinker, David De Koeyer, Fran Webster, George Fedak, Steve Molnar Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Food Canada, Ottawa ON, Canada K1A 0C6
Oil concentration of oat groats has important feed and nutritional implications. Low oil is considered desirable in most food products, while high oil contributes to greater feed energy. We used a genetic linkage map to identify quantitative trait loci (QTLs) in a naked-by-covered population of cultivated oat (see companion poster). Here, we report and discuss a more detailed analysis of QTLs affecting groat oil concentration. Based on simple interval mapping (LOD=3), three genomic regions were associated with oil concentration. A fourth region, which showed marker homology to the other three QTL regions, was significant based on single factor analysis (P<0.01). The allelic effects at all four regions were approximately equal, with each pair of alleles showing substitution effects of up to 0.5% oil. In two regions, alleles from the covered parent gave increased oil, while in the other two regions, alleles from the naked parent gave increased oil. These four QTL effects were primarily additive (no significant epistasis), accounting for more than 50% the variation in oil concentration, and explaining much of the transgressive segregation seen in this population. Based on commonality of cDNA sequences mapped in these regions, it is possible that all four QTLs result from genetic variation at a homeologous or duplicated locus (i.e. a common point in a biochemical pathway). The above hypothesis is supported by the additivity of allelic effects shown by these QTL: variation at different points in a biochemical pathway would be more likely to have produced epistatic effects.

This research was made possible by generous financial support from the Quaker Oats Company, and by collaborative dialogue and technical assistance from a consortium of Quaker funded institutions.

P - 41 AN RFLP / RAPD / AFLP LINKAGE MAP OF CULTIVATED OAT FROM A HULLESS-BY-COVERED CROSS Nick Tinker, Charlene Wight, Louise O'Donoughue, George Fedak, Ken Armstrong, Vern Burrows, Art McElroy, David De Koeyer, Corinne Chenier, Anissa Lazreg-Lybaert, Shan He, Steve Molnar Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Food Canada, Ottawa, ON, Canada K1A 0C6
Molecular linkage maps provide important tools for gene discovery and genomic research. In oat, previously reported RFLP linkage maps have been developed in two diploid populations and in the interspecific hexaploid population Kanota (Avena byzantina) x Ogle (A. sativa) ("K/O"). Here, we report a linkage map of RFLP, RAPD, and AFLP markers developed in an intraspecific (A. sativa) hexaploid oat population. The mapping population consisted of 102 recombinant-inbred lines derived by single-seed descent (to the F6) from a hulless-by-covered cross ("T/M"). Approximately 170 RFLP markers, 24 RAPD markers, and 250 AFLP markers were mapped to 35 linkage groups (LOD 6). Genome coverage in the 35 linkage groups is approximately 750 cM. An additional 45 unlinked markers probably account for a large portion of the remaining genome. The AFLP markers appear to be randomly distributed on most linkage groups, with five linkage groups containing exclusively AFLPs. Most of the RFLP markers were based on cDNA clones mapped in K/O. Since most clones map to multiple homeologous loci, definitive locus identity between two maps requires an allele that is common to both populations. Thus, it was possible to confirm the identity of only 46 RFLP loci common to both the T/M and the K/O map. These loci were used as anchors to align linkage groups in T/M with the K/O map. Other loci were presumed to be allelic when the same clone was mapped within a linkage group containing a confirmed allelic relationship. Homology to the K/O map was determined for 19 of the T/M linkage groups and several of the unlinked T/M markers. As with the K/O map, it has not been possible to reconcile all linkage groups to 21 oat chromosomes. Future comparisons among these and other oat maps may eventually make this possible. Presently, the T/M map provides a useful tool for the discovery of QTL, and for comparative genome mapping.

P - 42 TRANSFORMATION OF ELITE OAT CULTIVARS K.A. Torbert¹, H.W. Rines², H.F. Kaeppler³, G.K. Menon³, and D.A. Somers¹ ¹Department of Agronomy and Plant Genetics, University of Minnesota, 411 Borlaug Hall, 1991Upper Buford Circle, St. Paul, MN 55108; ²Plant Science Research Unit, USDA-Agricultural Research Service, St. Paul, MN 55108; ³Department of Agronomy, University of Wisconsin, Madison, WI 53705
Mature embryos of a specific oat (Avena sativa L.) have shown to be a reliable explant for initiating tissue cultures that can be genetically engineered to produce transgenic, fertile plants (Torbert et al. 1998). The purpose of this research was to investigate mature embryos from 16 elite North American spring oat cultivars for tissue culture initiation and the capability to produce transgenic plants from elite cultivars. Fifteen of the genotypes tested yielded at least some embryogenic callus that was similar in appearance to transformable callus based on our previous work. One elite cultivar, 'Belle' was utilized in determining transformation efficiency. From 30 microprojectile bombardments, 17 independent genetically engineered tissue cultures produced transgenic, fertile plants. DNA sequences designed to influence a variety of traits have been introduced into Belle. These traits include aluminum tolerance, antisense for starch composition modification, and wheat high molecular weight glutenin in oat kernels to investigate the dough making properties of oats by these proteins. These results demonstrate that mature embryo-derived tissue cultures will be useful for transforming elite oat cultivars.
P - 43 BREEDING FOR LEAF RUST RESISTANCE IN OATS L. Song and Weining Song Leslie Research Centre, Farming Systems Institute, Department of Primary Industries, Toowoomba, Qld 4350, Australia
The constant emergence of new and virulent races of leaf rust has long been recognised as a threat to the grazing oat industry in Australia, where over 400,000 hectares are grown annually between southern Queensland and northern New South Wales. The breeding program addresses this problem by pyramiding rust resistance genes with the aid of molecular markers. Almost all the rust resistance genes were introduced from programs overseas. These genes have been identified and backcrossed to adapted cultivars. In this process, recombinants have been selected for other agronomic characters which include foliage yield, growth habit and recovery after cutting. This breeding program also looks into other sources of rust resistances, including slow rusting. Good collaboration has been established with other oat improvement programs in Australia as well as programs overseas. Some of the more advanced materials have reached the F₇ or F₈ stage and are now ready for commercialisation.
P - 44 BETA-GLUCAN CONTENT IN ELITE OAT GERMPLASM AND BREEDING POPULATIONS D.M. Wesenberg¹, D.M. Peterson², L. Herrin², and D.E. Burrup¹ ¹National Small Grains Germplasm Research Facility, USDA, ARS, 1691 S. 2700 W., P.O. Box 307, Aberdeen, Idaho 83210, USA; ²Cereal Crops Research Unit, USDA, ARS, 501 Walnut St., Madison, Wisconsin 53705, USA.
Beta-glucan content determinations are routinely completed along with evaluations of other biochemical and agronomic traits for elite oat germplasm grown in irrigated and dryland trials in southern Idaho. In irrigated trials grown at Aberdeen, ID in 1994-97, beta-glucan content for 26 oat cultivars ranged from 3.74% ('Derby') to 6.22% (92Ab300). The beta-glucan content for the same entries grown under irrigation at Tetonia, ID in 1994-97 ranged from 3.91% ('Newdak') to 5.91% (92Ab300). In dryland trials involving the same entries at Tetonia during the same time period beta-glucan content ranged from 4.38 % (Derby) to 5.83% (92Ab300). Data from these trials indicate that environment influences beta-glucan content to a modest degree, with beta-glucan content averaging 4.73% at Aberdeen (Aberdeen), 4.88% at Tetonia (irrigated), and 5.11% at Tetonia (dryland). The relative performance of genotypes was similar over locations and apparently influenced little by environment. No association of beta-glucan content with other traits evaluated was apparent in these trials. In a study involving F₃ and F₄ bulk selections from a cross of NPB 871754/'Monida' grown in hills at Aberdeen in 1996 and 1997 and Tetonia in 1997, beta-glucan content of 164 F3 selections ranged from 4.74% to 7.32% at Aberdeen in 1996 with the parents averaging 6.91% and 5.49%, respectively. Monida averaged 189% of NPB 871754 in yield at Aberdeen in 1996, but despite the yield contrast protein content was similar for the parental lines, averaging 19.1% for NPB 871754 and 18.3% for Monida. The relationship of beta-glucan to protein content and yield in these early generation studies indicated that selection for high or low beta-glucan content should not be significantly impeded by protein or yield considerations. Correlations between beta-glucan and protein content were positive. Correlations between beta-glucan content and yield were negative, but relatively low ranging from -0.15 to -0.40 in the three environments. Standard unit heritability estimates for beta-glucan content based on F₃ populations grown at Aberdeen and F₄ populations grown at Aberdeen and Tetonia were less than 0.40.

Return to the Table of Contents