Yield traits of six clones of Miscanthus in the first 3 years following planting in Poland*

Stanisław Jeżowski

Instutute of Plant genetics, Polish Academy of Sciences, Strzeszyńska 34, 60-479 Poznań

Abstract

Miscanthus species are highly productive with low input and are excellent candidates for bioenergy feedstock production. A field experiment was conducted to characterize phenotypic differences in selected clones generated from interspecific hybrids of Miscanthus sinensis x Miscanthus sacchariflorus and intraspecific hybrids within M. sinensis. The field experiment was planted in plots of 20 m² at a density of 1 plant m^-2 in three randomized blocks. Te trial was monitored for 3 years for traits important to biomass production including plant height, tiller density, tuft diameter and shoot diameter. ANOVA showed significant genotypic variation in these traits once the stand was 2 years old. This study shows that tillering and tuft diameter in years 1 and 2 are the most important traits influencing biomass yield, but over 3 years when the highest yielding potential is reached, tillering and tuft diameter have the highest correlation with biomass yield. These results identifying high-yielding Miscanthus clones will be utilized in our plant improvement program.

Introduction

The natural habitat from which giant grasses from the genus Miscanthus originate is primarily Asia (Jeżowski, 1994; Deuter and Jeżowski, 1998; Xi and Jeżowski, 2004). These plants are frequently referred to as elephant grass or Chinese grass, but the name Miscanthus is also used. In Europe these grasses started to be cultivated, initially as ornamental plants, about 70 years ago. However, only in the 1980s Miscanthus plants started to be appreciated for their functional value (Deuter and Jeżowski, 1998, 2002).

At present giant grasses from the genus Miscanthus are considered to be key renewable raw materials for industry and energy production (El Bassam, 1997; Pude, 2000; Lewandowski, 2006). The main factor affecting this increased interest in these grasses is the fact that they exhibit (as plants of the C4 metabolic pathway) a very effective production of biomass (dry matter). Miscanthus is a perennial plant which may be cultivated at one site for 15-20 years. It has small nutritional requirements and yields well on a wide range of soils. Although this plant originate from Asia, it performs well in the European climate.

The best-known European Miscanthus is a triploid (3x) hybrid Miscanthus x giganteus, generated by crossing Miscanthus sinensis (2x) x Miscanthus sacchaliflorus (4x) (Greef and Deuter, 1993). This hybrid reaches full maturity in the third year after planting with peak autumn yields ranging from 10 to 30 t dry matter ha ^-1 (Lewandowski, 2006). An obstacle to the large-scale and prompt introduction of this triploid Miscanthus form for cultivation is the fact that it is sterile (unable to produce seeds) and therefore must be propagated vegetatively either as rhizomes or by micropropagation (Pude, 2000, Műnzer, 2000). Another disadvantage of Miscanthus x giganteus is its poor frost tolerance in the first year after planting (Pude, 2000).

Strategies for breeding work can either concentrate on selection of more frost-tolerant triploid forms of a triploid interspecific hybrid (Greef, 1996) or on obtaining fertile Miscanthus forms (2x or 4x) with the highest possible yields and best frost-tolerance. M. sinensis as well as diploid and tetraploid species of M. sacchariflorus (Deuter and Jeżowski, 1998, 2002) can both be propagated by seeds reducing establishment costs Christian et al., 2005). This study presents results of investigations on the variation and interdependence of yield structure characters in selected genotypes (clones) of grasses from the genus Miscanthus in the yield building phase over the first 3 years after planting.

Materials and methods

Experimental material consisted of six clones generated from hybrids of giant grasses from the genus Miscanthus. In this set of plant material there were two clones of a triploid (3x) hybrid Miscanthus x giganteus (MG/1, MG/2), which were generated by crossing M. sinensis (2x) x M. sacchariflorus (4x), two clones of hybrids generated by crossing M. sinensis x M. sacchariflorus (2x) (MS/3, MS/4) and two clones of hybrids generated within species M. sinensis (2x) (MS/5, MS/6). Initial plant material for their selection was provided by TINPLANT GmbH. All these forms are found in the collection of the Institute of Plant Genetics (PAS) in Poznań, Poland.

In autumn 2002 selected genotypes were planted at a density of 1 plant m ^-2 (10,000 plants ha ^-1 ) in three plots of 20 m² (4m x 5 m) each in the randomized block. Before planting subsoil ploughing was performed and mineral NPK fertilization was applied at 60, 80 and 120 kg ha ^-1 respectively. Soil used in the experiments belonged to the class of light loamy soils of the 4^th classification vale. During the growing season field observations (monthly) of plant height, tillering (the number of productive shoots per plant), plant diameter and shoot diameter (measured at half the canopy height) were made to characterize the growth and development dynamics of each genotype. The yields of standing dry biomass were measured by harvesting (at the end of March) ten plants from the central area of each plot. It is noteworthy that during the entire investigation, 2002-2005, the winter time was not very frosty in the Poznan area, and no loss in plants was noted. The autumn and winter of the season 2002/2003 were mild and these climatic conditions had a great impact on the survival of young Miscanthus plants. Te average temperature between November and March (about -5^oC), together with a thick snow layer that acted as natural insulator, contributed to the survival of the plants.

Analysis of variance was performed to determine significant differences between genotypes, traits and years. Analysis of canonical variables was performed for the purpose of a graphic presentation of the distribution of investigated forms characterized by all analyzed characters (Rao, 1965; Mardial et al., 1979) as it makes it possible to transform with a certain loss of information a multidimensional space to a two-dimensional space (plane). The interdependence between analyzed characters and the yield of biomass (dry matter) was measured using correlation coefficients calculated on plot means. Regression analysis was conducted to establish the significance of the effect of analyzed characters on the biomass yield in analyzed Miscanthus clones.

Results and discussion

Mean values of analyzed characters of biomass (dry matter) yield structure in three subsequent years of the study are presented in Table 1. Genotypes differed in plant height, tillering, tuft diameter and yield (tuft weight). The highest values for these characters, although only starting from the second year of growth and development, for the analyzed Miscanthus forms were found for two clones, MG/1 and MG/2, which are triploid forms. The lowest values for these characters were observed, already in the first year of growth and development, in clones MS/5 and MS/6, which are diploid forms.

Analysis of variance (data not presented in this work) showed that the variation in analyzed Miscanthus genotypes in terms of yields as well as yield structure characters were statistically significant. Such a trend was observed during the three successive years of plant growth and development, when plants reached full yielding potential.

The interdependence between yield (tuft weight) of analyzed Miscanthus clones are characters such as plant height, tillering, tuft diameter, shoot diameter, as well as the interrelationships between these characters, are presented in Table 2. This table shows that the yield of 1-year- old plants was significantly correlated with tillering and tuft diameter. Correlations between yield and these characters was r=0.78 and r=0.79, respectively. In contrast, the yield of older plants, 2- and 3-year-old, also depended significantly on these characters (the level of significance was similar to the level in the first year), but it was also correlated with plant height. In this case this correlation in subsequent years was r=0.67 and r=0.91, respectively.

Table 2. Correlations coefficients between Miscanthus biomass yield

and characters affecting yield over 3 years

Results of regression analysis for the dependence between biomass yield and analyzed characters are presented in Table 3. As a result of this analysis characters having no significant effect on yield in analyzed Miscanthus clones in individual years of the study were disregarded. On the basis of these regression equations it may be observed that tillering had a significant effect on biomass yield in the first year of growth and development. In the second year the yield of biomass was significantly determined by tuft diameter, while in the third year when Miscanthus reaches its full yielding potential, the volume of biomass was again determined, as it was in the first year, by tillering. A very similar dependence of biomass yield on the above mentioned morphological characteristics and the period of development in giant grasses from genus Miscanthus was previously observed by Dambroth (1991), El Bassa and Dambroth (1991), Dressler and Herzog (2004) and Pude and Jeżowski (2003).

Table 3. Regression analysis for the dependence of Miscanthus

biomass yield (y) on characters affecting yield over 3 years

In this study the differences and similarities of Miscanthus forms in terms of yield and all the other related characters jointly were evaluated using canonical variables (V1 and V2). This was performed by plotting the forms in a two-dimensional space (on a plane) described by all these characters. A graphic representation of the distribution of analyzed clones in individual years of the study (from planting to full yielding potential reached in the third year) is given in Figs. 1-3. This representation shows that in the first two years of growth and development only two clones, M/5 and M/6, were very similar and at the same time distant from other clones, MS/3, Ms/4, Mg/1 and MG/2 (Figs. 1 and 2). In the thirs year of the study, when Miscanthus plants reach full yielding potential, the graphic representation of the distribution of all analyzed Miscanthus clones changes so that three groups (pairs) of clones (located close to each other) were formed (Fig. 3), which in turn were distant from one another. The first pair was formed by clones MG/1 and MG/2, the second clones MS/3 and MS/4, whereas the third clones MS/5 and MS/6. Such a distribution of the analyzed forms may be explained by genetic background, as clones MG/1 and MG/2 came from a triploid hybrid M. sinensis (2x) x M. sacchariflorus (4x), clones MS/5 and MS/6 originated from hybrid M. sinensis (2x) x M. sacchariflorus (2x), while clones MS/3 and MS/4 were derived from hybrids generated within M. sinensis. A similar large genetic variation in different hybrids of grasses from the genus Miscanthus was described earlier by Greef and Deuter (1993), Deuter and Abraham (2000) and Xi and Jeżowski (2004).

In conlusion, this study shows that tillering and tuft diameter in the early years are the two most important traits influencing biomass yield. The most important results of this study, apart from the high amount of genetic variation observed among Miscanthus clones, is that the two most important components of yield are tillering and tuft diameter. This study also shows that effective selection in the breeding of hybrids can start in the first year by selecting for tillering and tuft diameter, and it appears that the best plant material for improvement through breeding are triploid hybrids.

References

Cristain DG, Yates NE, Rice AB, 2005. Establishing Miscanthus sinensis from seed using conventional sowing methods. Ind. Crops Prod. 21, 109-111.

Dambroth, M. 1991. Miscantus sinensis-Einfűrhung in die Thematik. KTLB-Arbeitspapier 158, 7-14.

Deuter M, Abraham J, 2000. Wiessenstand in der Miscanthus. Iniversität in Bonn. Beitrage zu agrarwisseschaften. Band 18, 8-14.

Deuter M, Jeżowski S, 1998. Szanse i problemy hodowli traw z rodzaju Miscanthus jako roślin alternatywnych. Hodowla i Nasiennictwo 4, 45-48.

Deuter M, Jeżowski S, 2002. Breeding conditions of the giant grasses Miscanthus genus-state art. Post. Nauk Rol. 2, 59-67.

Dressler, U.B., Herzog H. 2004. Biomasseproduktion mit Miscanthus-erste Ergebnisse zu Ertrag. Wassernutzung und Nitratauswaschung. Mitt. Ges. Pflanzenbauwiss. 6, 337-340.

El Bassam N, 1997. Renowable Energy. REU Tech. Ser. 46, 41-196.

El Bassam N, Dambroth M. 1991. A content of energy plant farm. In: Proceedings of the International Conference on Biomass for Energy, Industry and Environment, Greece, pp. 34-40.

Greef MJ, 1996. Etablirung und Biomassebildung von Miscanthus x giganteus. Cuvillier Verlag Gőttingen, 1-162.

Greef MJ, Deuter M, 1993. Syntaxonomy of Miscanthus x giganteus. Angewande Botanik 67, 87-90.

Jeżowski S, 1994. Miscanthus sinensis “Giganteus” – a grass for industrial and energetic purpose. Gen. Pol. 35A, 372-279.

Lewandowski I. 2006. Miscanthus – a multifunctional biomass crop for the future. In: Jeżowski S, Wojciechowicz KM, Zenkteler E (Eds.), Alternative Plants for Sustainable Agriculture. Institute of Plant Genetics PAS, Poznań, pp. 83-90

Mardial KV, Kent JT, Bibby JM, 1979. Multivariate Analysis. Academic Press Inc., Ltd. (London), pp. 1-509.

Műnzer W, 2000. Rhizompflanzen, eine Alternative? Beitrage zu Agrarwissenschaften. Universität in Bonn 19, 15-20.

Pude R. 2000. Anbau und ertrage von Miscanthus in Europa. Materiały konferencyjne. Polsko-niemiecka konferencja naukowa na temat wykorzystania trzciny chińskiej, 91-95.

Pude R. Jezowski S. 2003. Effect of selected morphogenetic traits on growth and development of Miscanthus spp. Biuletyn IHAR 227, 573-583.

Rao, CR. 1965. Linear Statistical Inference and its Applications. John Wiley, New York.

Xi Q, Jeżowski S. 2004. Plant resources of Triarrhena and Miscanthus species in China and its meaning for Europe. Plant Breed. Seed Sci. 49, 63-77.

*Full text of the article (including Figures) was published in Industrial Crops and Products 27 (2008), 65-68.