Bioenerg. Res. (2015) 8:1480–1491 DOI 10.1007/s12155-015-9682-2
New Insights into the Propagation Methods of Switchgrass, Miscanthus and Giant Reed Danilo Scordia 1 & Federica Zanetti 2 & Szilard Sandor Varga 3 & Efthymia Alexopoulou 4 & Valeria Cavallaro 5 & Andrea Monti 2 & Venera Copani 1 & Salvatore L. Cosentino 1
Published online: 21 October 2015 # Springer Science+Business Media New York 2015
Abstract A general obstacle to the development of perennial grasses is the relatively high cost of propagation and planting. The objective of the present study was to investigate new propagation and planting methods of giant reed (Arundo donax L.), miscanthus (Miscanthus x giganteus Greef et Deuter) and switchgrass (Panicum virgatum L.). Field and open-air pot trials were carried out in four different locations across Europe: hydro-seeding of switchgrass was tested in field trials at the experimental farm of the University of Bologna, Italy; stem propagation and bud activation methods of miscanthus were evaluated in field experiments in Péteri, Hungary; giant reed rhizome and stem propagations were compared in a field trial in Aliartos, Greece; finally, an open-air pot trial was carried out in Catania, Italy, using single-node stem cuttings of giant reed. Hydro-seeding emerged as a feasible and promising technique for switchgrass to ensure prompt seed emergence and weed control during plant establishment. Direct stem plantings of miscanthus were successful, and activated stem-buds were able to sprout in field conditions; however, timely stem transplant was determinant for shoot density and biomass yield. In giant reed, * Salvatore L. Cosentino
[email protected] 1
Dipartimento di Agricoltura, Alimentazione e Ambiente – Di3A, University of Catania, Via Valdisavoia 5, 95123 Catania, Italy
2
Department of Agricultural Sciences, University of Bologna, Viale Giuseppe Fanin 44, 40127 Bologna, Italy
3
Primus Ltd., Hatvani utca 25, H-1201 Budapest, Hungary
4
CRES – Center for Renewable Energy Sources and Saving, 19th Km Marathonos Avenue, 19009 Pikermi Attikis, Greece
5
CNR- IVALSA, UOS di Catania, Via P. Gaifami 18, 95126 Catania, Italy
rhizome propagation showed a higher stem density and biomass yield than stem propagation; however, the yield gap was not significant from the second year onwards. Single-node rooting was mainly driven by air temperature. Nodes from basal stems showed higher rooting rates than median and apical ones. Growth regulator pretreatments enhanced rooting rate only at transplanting times under suboptimal air temperatures. In general, these experiments provided insights into propagation strategies aimed at enhancing the establishment phase of perennial grasses. Keywords Crop establishment . Biomass production . Perennial grasses . Miscanthus x giganteus . Arundo donax . Panicum virgatum
Introduction The development of perennial grasses such as switchgrass (Panicum virgatum L.), giant reed (Arundo donax L.) and miscanthus (Miscanthus x giganteus Greef. et Deuter) depends on economic and environmental benefits as well as on qualitative characteristics required for bioconversion processes [1–5]. For example, a dramatic drawback of some perennial rhizomatous grasses unable to produce fertile seeds, such as giant reed or the triploid hybrid Miscanthus x giganteus, is the high cost of vegetative propagation (rhizomes and micropropagated plants); on the other hand, switchgrass, a fertile species which can be reproduced by seeds, needs timely and effective weed control strategies during its establishment, as well as careful soil tillage and a firm seedbed to achieve a successful stand. As known, giant reed does not produce viable seeds in most environments. The agamic propagation seems to be the only method for establishing this species [6]. In addition to
Bioenerg. Res. (2015) 8:1480–1491
agamic propagation strategies, rhizomes, stems cuttings and in vitro culture have been widely studied, with variable results [7–12]. Likewise, rhizome division is the most common method to agamically propagate Miscanthus x giganteus [13], although micropropagation [14] and stem propagation methods have also been evaluated [15]. While rhizome propagation ensures successful shoot emergence and, thus, establishment, albeit at a significant cost and with intensive labor, the number of new plants that can be propagated agamically from a single ‘mother’ plant is relatively small [16], leading to low plant density at establishment and low yield at first harvest. In addition, harvesting rhizomes makes it necessary to dig up the parent stand, leading to CO2 losses and leaving the soil vulnerable to erosion [15]. The propagation via stem-cuttings represents a more economic and environmentally friendly method than rhizome propagation. This method does not require the considerable work involved in digging up, breaking apart and replanting rhizomes, as the propagation material is the aboveground biomass; it also makes the multiplication rate several orders of magnitude greater than that achievable through rhizome propagation [15]. In summary, rhizomes and in vitro culture are the most successful but, at the same time, the most costly propagation methods for sterile species [17]. Stem cutting is a promising and relatively easy and cheap propagation method that still needs to be fine-tuned to these species, especially in regards to stem cutting portions, time of transplant in relation to the environmental conditions, and stem cutting pretreatments. Although switchgrass produces viable seeds, its establishment often fails because of the small seed size and morphology, seed dormancy, low early vigor, weed competition, sowing depth and methods [2, 18, 19]. It has been demonstrated that the low germination energy and the consequent quick weed colonization are often significant constraints during switchgrass establishment [19, 20]; thus, mechanical and/or chemical controls have to be scheduled to obtain the desired plant density [21]. Even if switchgrass could be grown under variable soil conditions [22], a firm seed bed is recommended in order to be able to sow it at shallow depth [19, 20]. The optimal soil condition for sowing could be achieved by both conventional and also by sod seeding, the latter being particularly successful in establishing switchgrass in degraded lands or in already established grasslands [22]. Furthermore, sod seeding allows quicker crop establishment, while it increases competitiveness toward weed emergence [23]. Among innovative techniques to colonize semi-arid lands, hydro-seeding emerged as a reliable way to establish permanent grassland in degraded surfaces or on hillsides [24]. Hydro-seeding also provides a more uniform coverage and advantages in soil erosion containment; it is generally faster than conventional broadcast seeding methods, and it offers the possibility to alleviate soil compactions, especially in finely textured soils when soil moisture
1481
conditions are inadequate for supporting equipment loads. This technique is widely used for this purpose in temperate climates [25] and, nowadays, has become very popular in semi-arid Mediterranean areas to restore grassland [26, 27]. Furthermore, hydro-seeding is also quite commonly used to establish lawns with satisfactory results and lower maintenance costs (i.e., reduced water requirements) [28]. The hydro-seeding technique consists in mixing selected seeds with colloidal substances in an aqueous solution usually along with mulch of various origin [29], and fertilizers [30]. The role of mulch is to cover the soil, to have the seeds partially covered and to interfere with weed emergence at early growth stages. In the frame of the European research OPTIMA project (www.optimafp7.eu), new propagation techniques to establish perennial grasses were studied. The objective of the present work was, therefore, to report the experiences of four partners of the OPTIMA project, in Hungary, Greece, and Italy (North and South), on new/alternative propagation methods of giant reed, miscanthus and switchgrass.
Materials and Methods Switchgrass Trials To examine the effectiveness and advantages of hydro-seeding of switchgrass compared to conventional seeding, two consecutive trials (in 2013 and 2014) were performed at the Experimental Farm of the University of Bologna, Italy (44° 54’ N; 11° 40’ E; 32 m a.s.l.) in an Ustochrepts soil (USDA 1999). The cumulate rainfalls between May and September were 206 and 370 mm in 2013 and 2014, respectively. The mean temperatures in the same period were 21.9 and 20.9 °C in 2013 and 2014, respectively. In 2013, hydro- (HS; Table 1) and conventional mechanical (MS, pneumatic disk seeder, Damax, Italy) seedlings were factorially combined with three soil tillage methods: conventional (CT, ploughing 40 cm depth then harrowing), minimum (MT, combined sub-soiler and discharrowing 20 cm depth) and no-tillage (NT, sod seeding). In 2014, only two tillage methods (MT and NT) were evaluated in combination with three hydro-seeding methodologies (Table 1). Appropriate mixing equipment for hydro-seeding was used for spreading 1 L m−2 of mulch containing 1.8 g of seeds (i.e., the adopted seed rate: 18 kg ha−1 pure live seeds). When mechanically seeded, switchgrass row spacing was 50 cm. In both years (2013 and 2014), the previous crop was maize (Zea mays L.). A strip-plot experimental design with three replications was adopted. Soil tillage was the main factor, while hydro-seeding was the sub-factor. All plots were irrigated by means of a sprinkler system until complete switchgrass emergence. Eight (80 mm in total) and nine (40 mm in total) irrigations were needed in 2013 and 2014, respectively. According to our previous experiences,
1482 Table 1
Bioenerg. Res. (2015) 8:1480–1491 Hydro-seeding (HS) mixtures used for switchgrass trials in 2013 and 2014 at the University of Bologna, Italy
Year
HS type
Commercial name and composition
Rate (g m−2)
2013a
HS
44
2014b
HS1
Provide Verde millennium [complete mixture of soil microorganisms (Penicillium sp.), algae, polysaccharides, mulch (1L ha−1), vegetal glue, organic fertilizer (4.4 % N, 2.2 % P2O5, 1.1 % K2O, 2.1 % Mg, 2.7 % Ca)] Envitotal (vegetal mulch mixture of cellulose 33 %, vegetal glue, soil N-fixing microorganisms 2.6 %, zeolites 38 %, organic carbon 11 %, total nitrogen 1.7 %) Envitotal (vegetal mulch mixture of cellulose 33 %, vegetal glue, soil N-fixing microorganisms 2.6 %, zeolites 38 %, organic carbon 11 %, total nitrogen 1.7 %) Cellugrȕn (cellulose fiber mulch: cellulose content 80 %, av. fiber length 1400 μm, av. fiber thickness 45 μm; bulk density 20–35 g L-1, pH 7.5)+soil control (adhesive hydro-colloid compound)
HS2 HS3 a
45 75 60+1
Provided by Geocentro S.R.L. (Italy); b Provided by Biasion S.P.A. (Italy).
av. = average
fertilization was not applied to reduce weed competition during switchgrass establishment. Switchgrass and weed presence were measured every 10 days until 19 days after sowing (DAS) in trial 1 and 54 DAS in trial 2 on two randomized sampling areas of 0.25 m2 per plot by counting seedlings of switchgrass and weeds. The aim of both trials was to compare sustainable hydroseeding technique with commonly used field practices for switchgrass establishment. Therefore, chemical weeding [Lontrel 72 SG (a.i.: Clopiralid 72 g L -1 ) at a rate of 120 g ha−1 +Trimmer SX (a.i.: Tribenuron-methyl 500 g kg−1) at a rate of 14 g ha−1] was performed only on MS plots, whilst in the HS plots, no weed control was applied. Considering switchgrass establishment unsatisfactory in 2013, trial 1 was destroyed at the end of September 2013, while trial 2 is still ongoing. For this reason, fresh and dry (105 °C until constant weight) aboveground biomass (AGB) were determined at the end of the growing season (24 November 2014) only in 2014, on sampling areas of 1.6 m2 per plot. Prior to biomass determinations, switchgrass was separated from weeds. In 2015, the number of switchgrass plants and weeds were measured on two sampling areas (0.25 m2 each) in all plots, 15 and 30 days after the beginning of spring re-growth (30 March 2015). All data were subjected to an analysis of variance (ANOVA). When F-ratios were significant, means were separated by the Fisher’s least significant difference (LSD) multiple range test (P≤0.05). Data from trials 1 and 2 were analyzed separately due to differences in the experimental set up. Giant Reed Experiments Rhizome vs. Stem Cutting Propagation A field experiment was established on 19 March 2012 in Aliartos, Greece (38°23’N; 23°06’E; 83 m a.s.l.) by using a local clone of giant reed. The following factors were compared in a split-plot design with three replications: (i) two propagation materials, rhizomes (RP) and stem
propagation (SP), and (ii) two plant densities, betweenrow 70 cm and within-row 75 (P1) or 150 (P2) cm, corresponding to 19,200 and 9500 plants ha−1, respectively. Propagation material was the main factor, while plant density was the sub-factor. Each plot measured 84 m2 (7× 12 m). The soil, which was managed as fallow the previous season, was ploughed and basic fertilization was applied (NPK, 11-15-15, 250 kg ha−1) during soil bed preparation by disk harrowing. Both rhizomes and stem cuttings were collected from a thirteen-year-old giant reed plantation located 5 km from the field trial. Stems were harvested in February 2012 and cut at 40–50 cm segments, subsequently stored in wet conditions for a few weeks prior to being directly transplanted in the field. One rhizome cutting with at least one visible bud was placed according to the desired density (P1 and P2, respectively). Stem cuttings were placed by laying two parallel stem segments inside the open furrow spaced at P1 or P2, respectively. No irrigation was applied after transplanting; however, supplemental irrigation was provided at the end of May 2012 for a total amount of 30 mm. Weed control was necessary only in some areas inside the plot and was done by hand. A top dress fertilization was applied at re-growth, both in 2013 and 2014, by using ammonium nitrate at a rate of 50 kg ha−1 yr−1. In the year following establishment, no weed control was needed due to strong competition of giant reed. At the end of each growing season (February 2013, 2014 and 2015), aboveground biomass (AGB) yield was measured by harvesting and weighting the whole fresh biomass in each plot. Biometric traits, namely plant height and stem diameter (only in 2014 and 2015), were measured on 10 random tillers per plot, while the number of stems per square meter was measured in 3 random sub-plots. Sub-samples of stems and leaves were oven-dried at 85 °C until constant weight to determine the dry matter yield. All data were subjected to a two-way ANOVA. When Fratios were significant, means were separated by the Fisher’s LSD multiple range test (P≤0.05).
Bioenerg. Res. (2015) 8:1480–1491
1483
Single-node Stem Cutting Propagation
Miscanthus Trial
Single-node stem cuttings of one-year-old giant reed stems (local clone ‘Fondachello’), collected at the Experimental Farm of the University of Catania (Italy), were used as propagation material at the Department of Agriculture, Food and Environment, University of Catania (37°31’N; 15°04’E; 73 m a.s.l.). In a completely randomized design replicated three times, three factors were compared: (i) transplanting time (25 February, 18 March, 14 May, 3 June, 2 July, 15 October and 25 November 2013), (ii) node position (basal, median and apical), and (iii) pretreatments [control-untreated nodes (C), soaking in distilled water (S), 1-naphthaleneacetic acid (NAA) and indole-3-butyric acid (IBA)]. Each replicate was represented by 10 stem cuttings. At each excision date, leaves were removed from stems and single-node segments 4 cm long were obtained. These node segments were excised from the basal, median and apical region of the stem. Two nodes at the distal part of each region were excluded in order to achieve better differentiation between regions. As regards pretreatment, node segments were soaked for 48 hours in distilled water (S) or in a solution containing 100 mg L−1 of NAA or IBA. The control (untreated nodes) was excised and planted at the same time as the pretreated nodes. One node segment was planted vertically in each pot (15-cm diameter×13-cm height and a capacity of 2 L), with the node just below the soil surface. Each pot was filled with a commercial horticultural substrate (Corg 34 %, N 0.2 %, organic matter 68 %, pH 5.0–6.5). Pots were placed in the open air under shade nets and irrigated up to soil field capacity (i.e., determining water loss from each pot gravimetrically on a daily basis). Eight weeks after each transplanting time, plants were removed from pots, thoroughly washed to remove the soil, and the number of full developed plantlets (i.e., with shoots and roots) was recorded. A two-way ANOVA was used to analyze the effects of transplanting time and node position in untreated node rooting, while a three-way ANOVA was performed to analyze the effects of transplanting time, pretreatment and node position on rooting rate. Data of the worst transplanting time (i.e., no rooting) was excluded from the ANOVA analysis. When Fratios were significant, means were separated by the Fisher’s LSD multiple range test (P≤0.05). Percentage data were arcsine pffiffiffiffi % transformed before statistical analysis. Linear regression analyses were performed by plotting average minimum and maximum temperature (during the eight weeks after each transplanting date) against rooting rate of basal, median and apical node cuttings. On the basis of the linear regression, when R2 was significant, an estimate of minimum base temperature was calculated at the abscissa intercept (null rooting, i.e., y=0).
Basal parts 1 m long of fresh collected stems from a 4-year-old miscanthus plantation were used as propagation material. Stems were cut at 3–5 cm aboveground and leaves were not removed. The stem-activating solution contained 10 mg L−1 6-Benzylaminopurine and 20 g L−1 sucrose. Sets of 500–1000 stems were bound and treated in a pressure chamber with approximately 1 atm difference in pressure and for a reaction time of 1 h to facilitate the nutrient uptake. The treated stem cuttings were directly transplanted in field plots at the experimental farm of Primus Ltd., Péteri, Hungary (47°23'N; 19°24'E; 147–150 m a.s.l.). In a randomized block design replicated three times, transplanting dates were compared. Plots were established in 2012 from early July to late September (6 and 23 July, 18 August, 7 and 25 September). Maximum and minimum air temperatures during each transplanting time are shown in Table 2. Each plot consisted of 4 parallel 10-m-long furrows, where 1-m-long stem cuttings were continuously laid down at a 70-cm row distance and about a 10-cm depth. Each plot measured 33 m2 (3×11 m). At establishment and the year after (2013), two irrigations were applied weekly by a drip irrigation system, for a total amount of 20 mm at each transplanting time. Irrigation was not necessary in 2014. Fertilization was not applied to reduce weed competition during establishment. Glyphosate (N-phosphonomethyl glycine) was used for weed control before soil preparation, two weeks after transplanting and in the subsequent early spring before stem sprouting. After emergence of new shoots, only selective herbicide (2,4dichlorophenoxyacetic acid) was used in order to control dicots. In the second year and only in plots that were established in midAugust and September, the same herbicide used to control dicots was necessary in the middle of summertime. At each transplanting time, the number of new developed shoots was counted in each plot in October 2012, 2013, and 2014. Fresh matter yield was measured in the second and third growing seasons by harvesting and weighting the biomass cut at 4 cm aboveground in each plot. Aboveground biomass Table 2 Maximum and minimum air temperatures (Tmax, Tmin, respectively; °C) during transplanting dates (2012) of miscanthus (Miscanthus x giganteus Greef et Deuter) stem cuttings at the experimental farm of Primus Ltd., Hungary Transplanting time July 6 July 23 August 18 September 7 September 23
Tmax
Tmin
25 34 29 28 19
15 21 14 5 4
1484
(AGB) was oven dried at 105 °C to determine dry matter yield. Harvest took place in March 2013 and 2014, respectively. All data were subjected to one-way ANOVA. When Fratios were significant, means were separated by Fisher’s LSD multiple range test (P≤0.05).
Results Switchgrass Hydro-Seeding Trials In 2013 (trial 1), a significant interaction sowing type x tillage (P≤0.05) was found on switchgrass emergence: CS/NT provided the highest emergence rate, and HS/NT the lowest (Fig. 1). HS/MT enhanced switchgrass emergence, but the amount of weeds was also considerable. The number of weeds was only affected by tillage (P≤0.05): CT showed a considerably higher amount of weeds (+400 %) than NT and MT. The majority of weeds were broadleaf, mostly purslane (Portulaca oleracea L.) and redroot pigweed (Amaranthus retroflexus L.). Weed competition was dramatic in CT, leading to a significant decrease of switchgrass seedlings; thus, stand establishment failed. In 2014 (trial 2), CS showed a faster emergence rate (data not shown), as well as a higher number of seedlings (Fig. 2). No significant effect of HS mixture was found on seedling emergence. Likewise, no significant differences emerged between MT and NT on seedling number and emergence rate. In contrast, weed emergence significantly differed between NT and MT, the latter showing fewer weeds (P≤0.05). As in 2013, broadleaf weeds clearly prevailed; common groundsel (Senecio vulgaris L.) was the most common weed in 2014. The aboveground dry biomass (AGB) was similar in all
Fig. 1 Switchgrass trial 1 (2013): switchgrass seedling (left axis) and weed (right axis) emergences at 19 days after sowing. Different letters represent statistically different means (P≤0.05) within switchgrass seedlings or weeds. CT, MT and NT: conventional, minimum and no tillage, respectively. HS and CS, hydro-seeding and conventional seeding, respectively
Bioenerg. Res. (2015) 8:1480–1491
treatments, averaging 4.7 Mg ha−1; HS3/MT showed the highest AGB (5.5 Mg ha−1). Weed biomass was similar in all treatments. Switchgrass re-growth (trial 2 only) was measured as the number of tillers in spring 2015. Significant differences (P≤ 0.01) in tiller number were found between seeding techniques (HS and CS) and between seedbed preparation methods (MT and NT): CS showed a higher number of tillers than HS (900 vs. 588 tillers m−2, respectively), while MT showed a higher number of tillers than NT (734 vs. 598 tillers m−2, respectively). Soil tillage effects on weed population were still detectable in the second year: NT showed a higher presence of weeds than MT (P≤0.05). Giant Reed Trial Rhizome and Stem Cutting Propagation At the first harvest (2013), stem height was significantly different between propagation methods (P≤0.01), while neither plant density nor interaction were. Averaged across the two densities, RP was taller than SP (4.14 and 2.44 m, respectively). A propagation method effect was also detected for stem density, 8.0 stems m−2 in RP and 4.6 stems m−2 in SP, averaged across the two densities. Neither plant density nor twofactor interaction were significant. The AGB was significantly higher in RP than SP (P≤0.01); again, neither plant density nor interaction were (Table 3). Both stem height and stem density increased at the second harvest (2014); however, no significant differences were found between RP and SP, P1 and P2 and their interaction. Stem diameter, on the other hand, was significantly higher in SP than RP (P≤0.05). Propagation material, plant density and
Bioenerg. Res. (2015) 8:1480–1491
1485
Fig. 2 Switchgrass trial 2 (2014): switchgrass seedling (left axis) and weed (right axis) densities at 54 days after sowing. Different letters represent statistically different means (P≤0.05) within switchgrass seedlings or weeds. CS, HS1-3, conventional sowing and three different hydro-seeding types (see Materials and Methods and Table 1). MT and NT, minimum and no-tillage, respectively
the interaction were not significant in terms of AGB. Nevertheless, these increased from the first to the second harvest, 4.6-fold in RPP1, 6.6-fold in RPP2, 4.2-fold in SPP1 and 7.3fold in SPP2.
Table 3 Biometric traits and aboveground biomass dry matter yield of giant reed (Arundo donax L.) propagated by means of rhizomes (RP) or stem cuttings (SP) at two densities, 70×70 cm (P1) and 70×150 cm (P2) at first, second and third harvest (2013, 2014, 2015, respectively) in Aliartos, Greece
At the third harvest (2015), neither stem height nor stem density were significant amongst factors, whilst significance at P≤0.06 was detected in AGB, with RP higher than SP (54.7 and 30.8 Mg ha−1, respectively, averaged across the two
Stem height (m)
Stem density (no. m−2)
Stem diameter (mm)
Biomass yield (Mg DM ha−1)
2013 RP P1 RP P2 SP P1
4.26 4.02 2.47
8.1 7.9 4.7
-
8.7 7.7 4.9
SP P2 Propagation
2.41 0.000
4.5 0.000
-
5.0 0.001
Density Propagation x Density 2014 RP P1 RP P2 SP P1 SP P2 Propagation Density
0.123 0.284
0.438 0.873
-
0.513 0.411
6.27 5.80 5.71 6.31 0.963 0.914
11.7 16.9 10.2 11.0 0.335 0.423
25.7 22.4 24.7 26.1 0.019 0.111
40.2 51.1 20.4 36.6 0.228 0.330
Propagationx Density 2015 RP P1 RP P2 SP P1 SP P2 Propagation Density PropagationxDensity
0.343
0.555
0.257
0.843
6.23 6.27 6.22 5.98 0.227 0.392 0.267
15.3 21.1 13.6 14.3 0.266 0.384 0.484
24.8 24.9 26.1 23.2 0.744 0.037 0.029
54.1 55.3 33.7 27.9 0.063 0.833 0.751
P-values of propagation, density and their interaction (i.e., "Propagation*Density") using the F-test at P
