Genetika, oplemenjivanje bilja i sjemenarstvo ORIGINAL SCIENTIFIC PAPER Morphological and AFLP variation in some genotypes of Poa angustifolia L. and Poa humilis Ehrh. ex Hoffm. Zsuzsa Lisztes-Szabó1, Ákos Zubor2, Béla Tóthmérész3, Mária Papp4, József Prokisch2 1 Institute of Plant Sciences, University of Debrecen Department of Food Science and Quality Assurance, University of Debrecen 3 Department of Ecology, University of Debrecen 4 Department of Botany, University of Debrecen Address: Institute of Plant Sciences, University of Debrecen, H-4015, Debrecen, P. O. Box 36, Hungary, (e-mail: szabozs@agr.unideb.hu) 2 Abstract Morphological and AFLP characters of P. humilis and P. angustifolia individuals collected from different habitats were compared after growing at the same site for three years. The study showed that the morphological plasticity of these species is considerable and the morphological characteristics were of only limited usefulness for taxonomic separation. AFLP analysis showed a high degree of polymorphism within the species. These populations of P. humilis and P. angustifolia could not be distinguished on the basis of molecular characters. The species group demonstrates that in cases of polymorphic taxa the genetic variability allows effective adaptation to slight environmental changes. Keywords: morphological characters, AFLP, Poa, multivariate analysis Introduction Poa angustifolia and Poa humilis belong to Poa pratensis L. aggregation. Many individuals with intermediate morphological characters can be observed. The morphological variability of P. pratensis genotypes is demonstrated by several authors (Bonos et al., 2000; Frenot et al., 1999; Rua, 1996; Reader and Bonser, 1993). Recently, a PCR-based assay for plant DNA fingerprinting, AFLP, has been developed which reveals significant levels of DNA polymorphism (Vos et al., 1995). The aim of this study is to contrast the morphological and genetic variability of P. angustifolia and P. humilis populations collected from East Hungary to provide additional information on the extent of genetic divergence. Material and methods Plant material Population for transplantation were collected in May 2003. The shoot clusters of 3 P. angustifolia populations and 3 P. humilis populations were collected from the sandy, 3 P. angustifolia populations and 3 P. humilis populations were collected from sandy loam and 3 P. angustifolia populations and 3 P. humilis populations were collected from alkaline soil pastures in Hajdú-Bihar County, Eastern Hungary. The shoots of the 18 populations were separated to small tiller units and planted in the Botanical Garden of University of Debrecen in the same plot. There were 18 rows in the plot, one row for each population. A row contained 15 tillers. The distance between tillers was about 10 cm, the width of a row is 50 cm. 44. hrvatski i 4. međunarodni simpozij agronoma 345 Genetics, Plant Breeding and Seed Production Morphometric analysis In May 2006, shoots were collected random from the signed shoot clusters of the original habitats (controls) and others were also collected random from the shoots growing in the experimental plot. We measured 20 individuals at sandy, 20 individuals sandy loam, and 20 individuals alkaline soil of both species from the original habitats and the experimental plot, too (altogether 240 individuals). The average of 10 data of the lateral tiller leaves and that of the spikelets were added to the given individuals. The following characteristics of the vegetative organs were measured (numbering follows the label of Fig. 1.): v1: height, v4: flag leaf sheath length, v5: flag leaf blade length, v6: ligule length of the flag leaves, v7: the widest width of leaves, v8: flag leaf blade width, v14: longest leaf blade lengths of lateral tillers. The following characteristics of the generative organs were measured (numbering follows the label of Fig. 1.): v1: height, v4: flag leaf sheath length, v5: flag leaf blade length, v6: ligule length of the flag leaves, v7: the widest width of leaves, v8: flag leaf blade width, v14: longest leaf blade lengths of lateral tillers. Morphological data were analysed by principal component analysis using the R statistical program (R Development Core Team, 2005). AFLP analysis For AFLP, each sample consisted of about one leaf per plant for ten plants per population. ZenoGene kit (ZenonBio Ltd, Szeged, Hungary) was used for DNA extraction following manufacturer‘s instructions. After the AFLP procedure (K. Szabó et al., 2006) the final amplified and selected PCR fragments were appeared by capillary electrophoresis with ABI Prism 3100 Genetic Analyzer in fragment analysis mode at the Agricultural Biotechnology Centre (Gödöllő, Hungary) using 360 mm length capillaries and POP-4 gel. GeneScan HD-400 Rox Size was applied as standard marker. AFLP fragments were identified by GeneScan 3.7 computer program. For cluster analysis of AFLP data each detected band above 80 threshold limit was scored as present (1) or absent (0) in all samples. The unweighted pair group method using arithmetic means (UPGMA) was used and a dendrogram was constructed based on Jaccard‘s similarity coefficient by the statistical software package SPSS 11.0 for Windows (SPSS Inc., USA). Genetic diversity and significant differences in the populations were estimated by the proportion of polymorphic loci. The allele frequencies and heterozygosity were estimated based on square root of recessive genotypes. These estimates and Neigenetical distances were analysed by TFPGA (Miller and Mark, 1997). Results and discussion Morphometric analysis We did not observe a sharp separation on the PCA biplot diagram based on all features of the two species (Fig. 1). Clusters of the individuals of the two species show complete overlap. However, the variables show unambiguously two different tendencies, depending on whether they are vegetative (v1-v8, v14), or reproductive (v9-v13) characteristics (except v7). Generative characteristics and the widest width of leaves (v7) are under influence of environmental factors less than vegetative characteristics. There was no separation based on the features of reproductive organs (Fig. 1). However, the individuals were concentrated into two, partly overlapping groups based on the characteristics of their vegetative organs. The first is the group of transplanted individuals which have larger vegetative organs. The second is the group of individuals with smaller organs which were collected from the original turf clones. An explanation for this is that in the experimental environment the interspecific competition was suppressed as seedlings of 346 44th Croatian & 4th International Symposium on Agriculture Genetika, oplemenjivanje bilja i sjemenarstvo other species were eliminated and the transplanted plants were able to spread and occupy all of the available area. The PCA results separated the transplanted and control populations of P. angustifolia into two groups but the transplanted and control populations of P. humilis did not show such separation (Fig. 1). The quantitative morphological characteristics of P. angustifolia responded more sensitively to the transplantation. Neither species nor transplanted populations were separated based on the reproductive characters. The two species did not show differences considering the vegetative characteristics, but the transplanted and control individuals of P. angustifolia separated slightly. 4 V14 PCA 2 V1 V5 V2 V4 2 V3 V8 0 V6 V11 V7 -2 V10 V9 V12 V13 -4 -6 -4 -2 0 2 4 6 PCA 1 Figure 1: PCA biplot of the two species based on all variables. (1-8, 14 vegetative, 9-13 reproductive variables.) □: transplanted P. angustifolia, ○: transplanted P. humilis, : control P. angustifolia, : control P. humilis. v1: height, v2: panicle length, v3: panicle width, v4: flag leaf sheath length, v5: flag leaf blade length, v6: ligule length of the flag leaves, v7: the widest width of leaves, v8: flag leaf blade width, v9: spikelet length, v10: upper glume length, v11: lower glume length, v12: palea length, v13: lemma length, v14: longest leaf blade lengths of lateral tillers. AFLP analysis In the AFLP analysis three primer combinations were used and generated clear banding patterns. A total of 338 DNA fragments were produced across 12 genotypes of four Poa species (plus P. pratensis and P. compressa) with 333 fragments being polymorphic (98.52% polymorphism) (Table 2). A representative example of the amplification products obtained at P. angustifolia genotypes using the primer combination EcoRI-ACC+Tru1ICAG. The number of bands from each primer combinations ranged from 84 (EcoRIACC+Tru1I-CAA) to 159 (EcoRI-ACC+Tru1I-CAG). In the present study, a wide variation in the number of polymorphic bands (83-157) and the percentages of polymorphic bands within a primer combination (98.80-98.74 %) were observed. The mean polymorphic rate was 98.47 %. A presence (1) or absence (0) binary data matrix containing 333 polymorphic AFLP fragments was used to generate the genetic similarity estimates. The Nei-genetical distance, between 0.87 and 0.95, shows remarkable genetic variation among the studied Poa populations. At the same time, the species of the P. pratensis group are close to each 44. hrvatski i 4. međunarodni simpozij agronoma 347 Genetics, Plant Breeding and Seed Production other, only P. compressa shows some separation. Inside the group, the similarity is the biggest between P. pratensis and P. humilis, and the smallest between P. pratenis and P. angustifolia. Table 2: The number of bands and degree of polymorphism revealed by AFLP primer combinations. Primer combinations EcoRI-ACC+Tru1I-CAG EcoRI-ACC+Tru1I-CAC EcoRI-ACC+Tru1I-CAA Total Total bands 159 95 84 338 Polymorphic bands 157 93 83 333 Polymorphism rate (%) 98.74 97.89 98.80 98.47 UPGMA clustering analysis was carried out in order to demonstrate graphically the genetic similarities among the Poa genotypes to present a dendrogram of 338 AFLP fragments obtained from three primer combinations at 12 Poa populations. But the populations of different species are not separated from each other as would be expected. The heterozygosity is bigger (25.6 %) in P. humilis than in P. angustifolia (23.4 %). The heterozygosity and polymorphism in P. humilis populations are bigger than P. angustifolia populations (Table 3). Table 3: Mean heterozygosity and polymorphism rate in populations of P. angustifolia and P. humilis. P. humilis P. angustifolia Heterozygosity (%) 25.60 23.40 Polymorphism (%) 94.40 83.80 The mean theta value of 0.2581 (0.2307 lower limit and 0.2827 upper limit) does not mean significant differences among the populations of studied species with 95% confidence interval and 1000 bootstrap repeat. Populations of different species were found to be genetically similar and there was a wide range of genetic variability among populations of a species. The different leaf blade width of the tillers of the two species is apparent, but sharp separation cannot be seen between the two species in the clusters on the basis of AFLP markers. This questions the value of the leaf blade width as a taxonomic character. The species level difference is ambiguous between the four species. The species level segregation of the two observed grass species is not supported by the morphological features measured. Rua (1996) reported similar results, as the quantitative morphological features did not reveal sharp statistical segregation between some Poa species. In our study the basis of the identification was the width of the leaf blades whilst other features, such as the length values, also used in identification keys, did not prove to be useful. Our previous observations suggested that the habitat adaptation, due to the habitat variations, destroyed the sharp border in morphological features between the two species. The results of the present study also supported that there was no phenotypic gap between these species, based on the measured characteristics. After three years growing in the same plot, tillers showed differences in morphological characteristics compared to individuals of original habitats. In some cases it seems there are not correspondences between 348 44th Croatian & 4th International Symposium on Agriculture Genetika, oplemenjivanje bilja i sjemenarstvo morphological characters and genotypes (McElroy, et al. 2002; Gillespie and Boles, 2001). These observations are consistent with the results of our study. The present study confirms the previous results that behind the habitat-sensitive features of grass shoots there is high within-species genetic variability. AFLP variation of population in Poa pratensis L. aggregation does not show species level separation between the studied species based on the generated AFLP markers. Our study showed that the morphological plasticity of our species was considerable, which was related to the small scale variability of the habitats. Thus, the measured quantitative morphological characters were not useful for taxonomic separation. There was no difference in the usefulness of the vegetative and generative features considering for the separation of the species. The study of AFLP variation of populations of the two species suggests complex population structures with high levels of introgression. The transplantation experiment showed that, despite considerable phenotypic plasticity, the morphology is determined by genotype (Howland et al., 1995). Poa taxa have high genetic potential for microevolutionary development. References Bonos, S. A., Meyer, W. A., Murphy, J. A. (2000). Classification of Kentucky bluegrass genotypes grown as spaces-plants. Hortscience 35 (5): 910-913 Frenot, Y., Aubry, M., Misset, M. T., Gloaguen, J. C., Gourret, J. P., Lebouvier, M. (1999). Phenotypic plasticity and genetic diversity in Poa annua L. (Poaceae) at Crozet and Kerguelen Islands (subantarctic). Polar Biology 22 (5): 302-310 Gillespie, L. J., Boles, R. (2001). Phylogenetic relationships and infraspecific variation in Canadian Arctic Poa based on chloroplast DNA restriction site data. Canadian Journal of Botany 79: 679-701 Howland, D. E:, Oliver R. P., Davy, A. J. (1995). Morphological and molecular variation in natural populations of Betula. New Phytol. 130: 117-124 K. Szabó, Zs.; B. Tóthmérész, Á. A. Zubor, M. Papp, Z. Győri, J. Prokisch (2006). AFLP variation in Poa pratensis L. agg. in Eastern Hungary. In: Proceedings of 41st Croatian and 1st International Symposium on Agriculture, ISBN 953-6331-39-X, p. 225-226 McElroy, J. S., Walker, R. H., van Santen, E. (2002). Patterns of variation in Poa annua populations as revealed by canonical discriminant analysis of life history traits. Crop Science 42 (2): 513-517 Miller, Mark P. (1997). Tools for population genetic analysis (TFPGA) 1.3: A Windows program for the analysis of allozyme and molecular population genetic data. Computer software distributed by author. R Development Core Team (2005). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.Rproject.org.http://her.bio.nau.edu/-miller/tfpga.htm. Reader, R. J., Bonser, S. P. (1993). Control of plant frequency on an environmental gradient – effects of abiotic variables, neighbors, and predators on Poa pratensis and Poa compressa (Gramineae). Canadian Journal of Botany - Revue Canadienne de Botanique 71 (4): 592-597 Rua, G. H. (1996). The genus Poa in Patagonia: a phenetic analysis of species of Poa subgenus Poa. Botanical Journal of the Linnean Society 121 (3) 229-241 Vos, P., Hogers, R., Bleeker, M., Reijans, M., van de Lee, T., Hornes, M., Frijters, A., Pot, L., Peleman, J., Kuiper, M., Zabeau, M. (1995). AFLP: A new technique for DNA fingerprinting. Nucleic Acids Res. 23: 4407-4414 44. hrvatski i 4. međunarodni simpozij agronoma 349