Publicado en el No.140, Pgina 23 a 31 - (32015) caracteres
Characterization and analysis of a collection of Avena sativa L. from UruguayMariana Vilaró Mónica Rebuffo Carolina Miranda Clara Pritsch Tabaré Abadie
The oat crop is placed fifth in importance among the cultivated cereals of the world, following wheat, rice, maize and barley (FAO 2001). Oats are mainly used for cattle feed in Uruguay. A substantial portion of the crop is grazed directly in the fields, and afterwards harvested for hay or grain when livestock are removed (double purpose) (Rebuffo 1997). Oat ranks first among winter forage species in intensive production systems, particularly when forage availability is sought in fall or early in the winter.
The introduction of the crop into the country took place in the 18th century through the Spanish colonization, with animal feed being its main destination. From that moment, landraces evolved, developing local adaptations (Böerger 1943). In spite of the fact that breeding programmes for oats have been in existence since the beginning of the 20th century, and that modern cultivars are available (INASE 2002), oat production in Uruguay is currently based mainly on landraces descended from Estanzuela 1095a. These consist of a series of populations of multilines. This was originally a pure line, released in 1930 by the ‘Instituto Fitotécnico y Semillero Nacional La Estanzuela’ (‘La Estanzuela’) (Millot et al. 1981). This cultivar originated from an individual selection of an heterogeneous sample of seed from the north of the country, in the year 1925 by Dr Alberto Böerger (Böerger 1943).
Classification and origin of cultivated oats
The genus Avena shows great complexity, which is demonstrated by the fact that some classifications consider 7 species (Ladizinsky and Zohary 1971), 14 species (Ladizinsky 1989) and even 34 species (Baum 1977). Within the hexaploid oats, the four taxa (Avena byzantina, A. fatua, A. sativa and A. sterilis) represent a complex of wild types, weeds and cultivated derivatives that have been treated as different species in the taxonomic literature (Mordvinkina 1936). At present, these four taxa are classified as a single biological species (A. sativa L.) as they all share the same chromosome number and are mutually interfertile (Stanton 1955; Ladizinsky 1989; USDA–NRCS 2002). Under the biological species concept, they can be treated as subspecies (USDA–NRCS 2002) or even as botanic varieties (Loskutov 2002). Cultivated hexaploid oats show certain morphological characteristics that clearly differentiate them from the wild taxa (non-shattering florets, essentially glabrous lemmas and awns that are rudimentary or absent). The group of cultivated oats includes white oats (A. sativa) and red oats (A. byzantina), which are differentiated primarily by their rachilla fracturing pattern ( Stanton 1955). These two taxa are cited as synonyms by the United States Department of Agriculture (USDA) Plants Database (USDA–NRCS 2002).
There is agreement that A. sterilis represents the direct hexaploid progenitor of cultivated hexaploid types (Holden 1976; Zhou et al. 1999; Cheng-Dao et al. 2000; Jellen and Beard 2000). Seed from A. sterilis (that originated in the Turkey–Iran–Iraq region) was probably transported along with the primary crops of wheat and barley as weed contaminants through Europe, becoming domesticated in central or northern Europe 2000–3000 years ago (Holden 1976; Zhou et al. 1999). Apparently, two independent paths of domestication from the progenitor species occurred, one path led to the development of the A. sativa type and the other to the A. byzantina type (Zhou et al. 1999; Jellen and Beard 2000). Differences between these two types are based on intergenomic translocations in chromosomes 7C-17 (Zhou et al. 1999; Jellen and Beard 2000).
Both A. sativa and A. byzantina (white and red oats, respectively) are sown in Uruguay. Landraces are A. byzantina type, whereas modern cultivars belong mainly to the A. sativa type. In the past, oat production was based mainly on red oats (Fisher et al. 1937), because of their favorable agronomic characteristics such as resistance to intensive grazing and double purpose ability (Millot et al. 1981).
Landraces (also defined here as local varieties) are populations that have been formed by natural selection processes under the effects of weather, soil type and agronomic conditions of a particular region, and have suffered artificial selection pressures from the deliberate manipulation of farmers. They contain important genetic variability, which determines their ability to adapt to changes in their environment (Frankel and Brown 1995). In addition, landraces provide an important source of useful variability for breeding activities (Frankel and Brown 1995; Allard 1997) provided that they are accompanied by information on characterization and agronomic evaluation. This information is essential for the correct conservation of genetic variability and for the accessions to be of use in breeding programs.
The INIA ‘La Estanzuela’ oat collection
A diverse collection of more than 600 oat genotypes is kept at Instituto Nacional de Investigaciones Agropecuarias (INIA) ‘La Estanzuela’, maintained by the second author. The four main groups of accessions that represent the diversity of the collection are: (a) old varieties from the region (Argentina, Brazil and Uruguay) that had been kept in National Plant Germplasm System/United States Department of Agriculture (NPGS/USDA), and were donated in 1996; (b) modern lines from the International Quaker Oats Evaluation Network Collection; (c) selections for differential resistance to leaf rust (Puccinia coronata f. sp. avenae) from foundation seed of two traditional cultivars (Estanzuela 1095a and RLE 115); (d) farmer varieties, designated as landraces, which were collected from fields in different regions of the country where farmers that had kept and multiplied seed in their fields for more than three years.
The general objectives of this research were to characterize and analyze the oat collection of Uruguay, in order to contribute to the conservation and utilization of locally adapted genotypes. The specific objectives were to characterize the 120-accession subset, and compare the groups by means of univariate and multivariate methods.
Materials and methods
A subset of 120 entries was selected to represent the diversity of the collection, classified into four different groups:
Field methods and experimental design
The 120-accession subset was sown at the INIA ‘La Estanzuela’ Experimental Station, located at Colonia, Uruguay (34°20'S latitude, 57°41'W longitude) on 28 July 2000. Accessions were sown in a two-row 1-m plot in direct sowing (rows spaced 0.17 m apart). Sowing density was the equivalent to 100 kg ha -1 of the cultivar Estanzuela 1095a, corrected by germination and seed size. Harvest was undertaken manually on 23 December 2000. The experimental design was incomplete blocks with two replicates. For all traits, five plants per plot were assessed. A mixture of genotypes susceptible to leaf rust was sown between the plots and around the edges of the experiment, to standardize the amount of inoculum reaching the plots.
The collection was characterized for 20 traits during the vegetative (traits 1–4), reproductive (5–12) and maturity stages (13–20). Traits 1–18 were quantitative, whereas 19 and 20 were qualitative. A description of the traits is presented below, including three elongation rates (traits 5–7) calculated to describe the pattern of reproductive development (from differentiation of reproductive node until flowering):
Univariate methods (S tatistica: StatSoft Inc. 1998; G enstat: GENSTAT 5 1998) used for the study of diversity between groups consisted of the following: adjustment of means to environmental gradients (such as soil fertility), estimation of univariate statistics and comparison of means for all the traits studied. For the adjustment of means to environmental conditions, an analysis of variance with the REML (Restricted Maximum Likelihood) estimation method was undertaken. This analysis considers an incomplete block design and a mixed model of fixed variables (repetitions and accessions) and random variables (repetition×block). The univariate statistics calculated were: mean, standard deviation, maximum and minimum. The Brown–Forsythe Test of Homogeneity of Variances was performed. A univariate analysis of variance was performed on the variables with homogenous variances; those that were significant (p<0.05) were compared using the Minimum Significant Difference. The variables with heterogeneous variances were tested with a t-test with separate variance estimates. The two discrete traits analyzed (19 and 20) were considered for the estimations as continuous traits because of the high number of classes that were obtained after the adjustment of the means.
The multivariate analysis performed was the Discriminant Analysis (S tatistica: StatSoft Inc. 1998). This method was used to study the differences between groups, identifying the most discriminant traits and performing a classification of cases. Canonic functions (roots) were obtained in order to establish the maximum differences between groups and to display them graphically (Statistica: StatSoft Inc. 1998).
A good separation among groups can be observed, mainly between the farmer and modern lines, based on the assessed traits (Table 1). These two groups show the most extreme values for the majority of the traits, while the groups of old varieties and selections show intermediate values. Plant height/width relationship shows the lowest mean value in the farmer group (0.42), which corresponds with a prostrate plant growth habit. The group of modern lines, in comparison, has a mean value for this trait of 1.7, showing that these plants have an erect growth habit. In addition, farmer varieties show a significantly higher value of tillers (17 tillers per plant) than the modern group (10 tillers).
The traits analyzed in this study show important differences in their discriminatory ability (F—Fisher), with values that range from 37 (number of tillers) to 1 (plant height/width relationship and awn length) (Table 2). The model includes the first 14 traits, excluded from the analysis were those traits that have a small effect on group separation, such as hairiness at basal part of the lemmas, grain length and glume length.
High correct classification values were obtained from the Discriminant Analysis, with a mean value for the four groups of 85% (Table 3). The groups that displayed the highest values of correct classifications were the modern lines and farmers’ varieties (92 and 90%, respectively). The lowest value was displayed by the group of selections (71%).
Root 1 explains 65.3% of the total variation, the most determinant traits being: erectness of panicle, 1000 grain weight, number of tillers and rate of elongation to third node (Table 4) (high values of standardized coefficients and correlations have a greater weight on the roots). Roots 1 and 2 together account for 83.9% of the among–within group variation. Figure 1 shows the canonic value of each accession for roots 1 and 3. In spite of the fact that root 3 explains a smaller proportion of variation than root 2, Figure 1 shows a clearer separation of the four groups, particularly as it separates accessions belonging to the group of selections from the rest of the groups.
Description of the groups
The group of farmer varieties shows the highest value for root 1. This root is highly determined by the traits: number of tillers, erectness of panicle and leaf width (Table 4). Farmer varieties show plants with many tillers, drooping panicles and narrow leaves. This group is also characterized by a prostrate growth habit (low values of height/width of plant relationship) and the smallest flowering heights, as well as being the latest to initiate their reproductive phase. Another feature of this group is that the accessions displayed the largest rust infection indices (AUDPC) and the lowest grain yields (Table 1).
Modern lines present the lowest values for root 1. This group contrasts with the farmer group, showing plants with few tillers, wide leaves and erect panicles. The group is also characterized by an erect growth habit, high grain yields, the highest rate of internode elongation, the shortest life cycles and the lowest rust infection indices (AUDPC) (Table 1). Figure 1 graphically displays the clear separation between the modern lines and farmer varieties, being the two most differentiable groups.
Old varieties show intermediate values for the first root, being situated towards the center of the graph. This group presents intermediate values in the traits more highly correlated with root 1. Some of the traits under study display a broad range of values (Table 1), which results in the great dispersion observed in the graphic representation.
The selections show intermediate values for the first root, as observed in the old varieties. This group is characterized by semi-erect panicles and leaves of intermediate width, similar to old varieties. However, for the trait number of tillers (highly correlated with root 1), this group shows values that are as low as in the modern group.
Traits that best discriminate between the four groups of accessions are number of tillers, leaf width, grain yield and erectness of panicle (Table 2). This result is also verified in the univariate analysis, as these traits show the most contrasting values (Table 1). This is in agreement with Stanton (1955), who affirms that the most effectively used characters in classification of oat cultivars are number of tillers, type of panicle and growth habit, among others. This author also states that grain and glume characters are of little value for classification purposes, which is confirmed in this research. Traits used in this study are very useful for discrimination purposes between the four groups of accessions, as shown by the high correct classification values. This is especially evident for the groups of modern lines and farmer varieties.
The high proportion of variation accounted for by the first two roots (83.9%) can be represented in graphs of two or three dimensions without an important loss of information (Franco et al. 1997a). Correlation coefficients between the original traits and the roots provide a biological interpretation (Franco et al. 1997b). In fact, morphological and physiological differences between the four groups under study are established, with two groups being clearly differentiable: farmer varieties and modern lines (A. byzantina and A. sativa types of oats, respectively) (Table 1 and Figure 1). This is in agreement with the fact that accessions belonging to these two types of oats show important morphological differences, especially in traits such as growth habit and tillering capacity. These differences in plant architecture have important consequences and have determined the different uses of the two oat types.
The group of old varieties shows an important dispersal in values of root 1. This is because the group includes accessions from both A. byzantina and A. sativa oat types. In addition, these accessions come from different countries ( Argentina, Brazil and Uruguay) and show different degrees of evolution in terms of breeding (cultivars, cultivated, breeding lines and landraces). The criteria to include accessions in this group was that they should represent a sample of the old oats of the region, collected by USDA, but these accessions do not necessarily have the same origin.
The selections display intermediate values for the first root because the group shows intermediate values for the traits highly correlated with this root. However, this group shows the highest values for root 3. A reason for this may be because root 3 has a greater standardized coefficient and a greater correlation with the trait erectness of panicle than root 2, which could be causing a better separation between groups.
During the last 70 years, germplasm introgression between the two types of oats (A. byzantina and A. sativa) has been performed to a great extent (Zhou et al. 1999), as both taxa have been extensively crossed in the development of modern cultivars (Jellen and Beard 2000). However, Millot (personal communication, 2002) states that recombinations between these two types are not always possible to achieve, each type keeping its distinctive features. A possible explanation for this would be the presence of reciprocal chromosome translocations, which are found in cultivated (Ladizinsky 1970; Singh and Kolb 1991; Wilson and McMullen 1997) hexaploid oat genotypes. Translocations may have profound effects on recombination through physical suppression of pairing around the breakpoint (Burnham, 1934), as well as inviability of duplicate-deficient (Dp-Df) spores arising from interstitial recombination in translocation heterozygotes (Sansome 1932).
The morphological and physiological differences observed between accessions from the farmer varieties and modern lines may also have another cause, such as differential evolutionary processes. Accessions from the farmer group show clear adaptations to grazing conditions (that may have been very intensive in some cases) to which they have been exposed for many years (Rebuffo, personal communication, 2002) see below. The characteristics observed in the modern lines are probably the product of breeding activities. This is surely owing to the fact that these accessions belong to the Quaker collection, which has a main goal of increasing grain yield and quality in the region (Floss 1997). Since the domestication of oat, its adaptation to crop conditions has led to alterations in plant morphology (reduction in awn length, more compact panicles) and physiology as a result of changes in plant architecture (Peltonen-Sainio 1993).
The utilization of farmer varieties, as has already been mentioned, is mainly for forage (and not grain production). Intensive grazing selects in favor of the most prostrate and late individuals, as these are the ones that manage to survive these extreme conditions. Similar adaptations to grazing conditions have been described in different grass species by Jaindl et al. (1994), Carámbula (1977) and Hodkinson et al. (1989). This type of management is typical of the most intensive production systems in Uruguay and particularly of the places where these accessions were collected. Oat is an autogamous species, and its populations are mixtures of individuals with a high frequency of homozygotes. In these populations, changes in the frequencies of individuals can take place very quickly in response to extreme grazing systems, such as continuous grazing or late livestock removal in the spring. Even though the study of adaptation processes in autogamous species in response to changes in their environment is not frequent, Nichols and Cocks (1997) demonstrated that a population of Trifolium subterraneum could show changes in a very short period of time (3 years) in two contrasting environments.
Owing to the years of selection for adaptation to intensive grazing conditions that accessions of the farmer groups have experienced, they hold important genetic value. Having characterizations of these genotypes is of great importance, as it is a way of preserving valuable individuals from being lost. Conservation of these individuals is thus of great importance as they could provide sources of resistance to diseases or to particular environmental conditions (Frankel and Brown 1995). These landraces could provide sources of resistance to drought, for example, because farmers sow them early in the autumn, or even late in the summer. Thus, it is to be expected that they would show tolerance to high temperatures and low hydric availability.
This work has been a contribution to increase the knowledge about the oat germplasm conserved in Uruguay. This better understanding should allow a better conservation and use of the collection in breeding programmes. The research will also assist in the conservation of valuable germplasm, as is the case of local varieties, which hold important local adaptation and are of widespread use by farmers throughout the entire country.
We gratefully acknowledge the financial support provided by the INIA ‘La Estanzuela’, Uruguay. This study was a component of a research for the degree of Ingeniera Agronoma from the College of Agriculture of Uruguay for the first and third authors. The authors also thank Lucía Gutiérrez and J.C. Millot for their valuable suggestions and advice. This work would not have been possible had it not been for the landraces that were kindly donated by farmers throughout the country in 1999.
Allard RW. 1997. Genetic basis of the evolution of adaptedness in plants. In: Tigerstedt PMA, editor. Adaptation in Plant Breeding. Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 1–11.
Baum BR. 1977. Oats: wild and cultivated. A monograph of the genus Avena L. (Poaceae). Biosystematics Research Institute, Canada Department Of Agriculture, Research Branch, Ottawa, Ontario, Canada. Monograph no. 14. http://pas.byu.edu/Faculty/enj/oatsite/aventaxa.htm.
Böerger A. 1943. Investigaciones Agronómicas; Genética Fitotecnia Rioplatense (Vol. 2). Montevideo, Barreiro y Ramos.
Burnham CR. 1934. Chromosomal interchanges in maize: reduction of crossing over and the association of non-homologous parts. American Nature 68:81–82.
Carámbula, M. 1977. Producción y manejo de pasturas sembradas. 2nd edition. Montevideo, Hemisferio Sur, pp. 1–464.
Cheng-Dao L, Rossnagel BG, Scoles GJ. 2000. Tracing the phylogeny of the hexaploid oat Avena sativa with satellite DNAs. Crop Science 40:1755–1763.
FAO. 2001. FAOSTAT agricultural data. http://apps.fao.org/cgi-bin/nph-db.pl?subset=agriculture.
Fisher GJ, Brotos C, Bonjour A, Gheorghianov V. 1937. Los ensayos de avena para forraje verde realizados en el año 1934. Archivo Fitotécnico Uruguayo 2:483–529.
Floss EL. 1997. Breeding oat cultivars suitable for production in developing countries—progress and importance for South America. In: Rebuffo M, Abadie T, editors. Third South American Oats Congress. INIA La Estanzuela, Colonia, Uruguay, November 11 12, 1997, pp. 3–5.
Franco J, Crossa J, Diaz J, Taba S, Villaseñor J, Eberhart SA. 1997a. A sequential clustering strategy for classifying gene bank accessions. Crop Science 37(5):1656–1662.
Franco J, Crossa J, Villaseñor J, Taba S, Eberhart SA. 1997b. Classifying Mexican maize accessions using hierarchical and density search methods. Crop Science 37(3):972–980.
Frankel OH, Brown AHD. 1995. The conservation of plant biodiversity. Cambridge University Press, UK.
G enstat 5. 1998. Release. 4.1 for edition. Lowes Agricultural Trust. IACR, Rothampstead, UK.
Hodkinson KC, Ludlow MM, Mott JJ, Baruch Z. 1989. Comparative responses of the savanna grasses Cenchrus ciliaris and Themeda triandra to defoliation. Oecologia 79(1):45–52.
Holden JHW. 1976. Oats. In: Simmonds NW, editor. Evolution of Crop Plants. Longman, Edinburgh School of Agriculture, Edinburgh, UK, pp. 86–90.
INASE. 2002. Cultivos Forrajeros. Boletín no. 73. http://www.inase.org.uy/Cultivosforr.htm. Instituto Nacional de Semillas-Uruguay, Canelones, Uruguay.
Jaindl RG, Doescher P, Miller RF, Eddleman LE. 1994. Persistence of Idaho fescue on degraded rangelands: adaptation to defoliation or tolerance. Journal of Range Management 47(1):54–59.
Jellen EN, Beard J. 2000. Geographical distribution of a chromosome 7C and 17 intergenomic translocation in cultivated oat. Crop Science 40:256–263.
Ladizinsky G. 1970. Chromosome rearrangements in the hexaploid oats. Heredity 25:457–461.
Ladizinsky G. 1989. Biological species and wild genetic resources in Avena. In: Report of a Working Group on Avena (third meeting) held at the Plant Breeding and Acclimatization Institute, Radzikow, Poland, 7 9 March 1989, ECP/GR/IBPGRI. International Plant Genetic Resources Institute, Rome, Italy, pp. 19–32.
Ladizinsky G, Zohary D. 1971. Notes on species delimitation, species relationships and polyploidy in Avena. Euphytica 20:380 395.
Loskutov I. 2002. Systematical approaches to the genus Avena L. http://www.vir.nw.ru/avena/sys_app.htm.
Millot JC, Rebuffo MI, Acosta YM. 1981. RLE 115: Nueva variedad de avena. In: Avena. CIAAB, La Estanzuela. Miscelánea 36, pp. 1–12.
Mordvinkina AI. 1936. Oves—Avena. In: Kulturnaya flora SSSR. Khlebnye zlaki. Rozh, Yachmen, Oves. (Cultivated flora of the USSR. Grain cereals. Rye, barley, oats). Vol. 2., M.-L., pp. 333–438.
Nichols PGH, Cocks PS. 1997. Evolution in segregating genotype mixtures of subterranean clover. In: Buchanan-Smith JG, Bailey LD, MacCaughey P., editors. Eighteenth International Grassland Congress, Winnipeg, 10 June, Manitoba, Canada. Proceedings. Canadian Forage Council, Canadian Society of Agronomy, Canadian Society of Animal Science. Vol. 1, session 4, pp. 93–94, id no. 732, 1 CD.
Peltonen-Sainio P. 1993. Productivity of oats: genetic gains and associated physiological changes. In: Slafer GA, editor. Genetic Improvement of Field Crops. University of Buenos Aires, Argentina and The University of Melbourne, Parkville, Victoria, Australia, pp. 69–94.
Rebuffo M. 1997. Oat use and production in Uruguay. In: Rebuffo M, Abadie T, editors. Third South American Oats Congress. INIA La Estanzuela, Colonia, Uruguay, November 11-12, 1997, pp. 25–28.
Sansome ER. 1932. Segmental interchange in Pisum sativum. Cytologia 3:200–219.
Simpson GM. 1968. Association between grain yield per plant and photosynthetic area above the flag-leaf node in wheat. Canadian Journal of Plant Science 48:253–260.
Singh RJ, Kolb FL. 1991. Chromosomal interchanges in six hexaploid oat genotypes. Crop Science 31:726–729.
Stanton TR. 1955. Oat identification and classification. USDA Technical Bulletin no. 1100.
StatSoft, Inc. 1998. Statistica for Windows (computer program manual). Tulsa, OK. StatSoft Inc.
USDA–NRCS. 2002. The Plants Database, Version 3.5. National Plant Data Center, Baton Rouge, LA 70874-4490 USA. http://plants.usda.gov/.
Wilson WA, McMullen MS. 1997. Recombination between a crown rust resistance locus and an interchange breakpoint in hexaploid oat. Crop Science 37:1694–1698.
Zhoux, Jellen EN, Murphy JP. 1999. Progenitor germplasm of domesticated hexaploid oat. Crop Science 39:1208–1214.
Contact PGRNfirstname.lastname@example.org about this page
Key words / Descriptors
Copyright © Bioversity International - FAO. All rights reserved.
: mysql_close(): no MySQL-Link resource supplied in /Volumes/Danika1891 HD/WWW/PGR/article.php on line 204