Biomass yield and energy balance of giant reed (Arundo donax L.) cropped in central Italy as related to different management practices

Europ. J. Agronomy 22 (2005) 375–389

Biomass yield and energy balance of giant reed (ArundodonaxL.) cropped in central Italy as related to different

management practices

L.G. Angelinia,∗, L. Ceccarinia, E. Bonarib

a Dipartimento di Agronomia e Gestione dell’Agroecosistema, Via S. Michele degli Scalzi 2, 56100 Pisa, Italyb Scuola Superiore di Studi Universitari e di Perfezionamento S. Anna, Piazza Martiri della Libert`a, 33, 56100 Pisa, Italy

Received 12 May 2003; received in revised form 24 March 2004; accepted 12 May 2004

Abstract

In order to evaluate the possibility of reducing energy input in giant reed (Arundo donaxL.) as a perennial biomass crop, a fieldexperiment was carried out from 1996 to 2001 in central Italy. Crop yield response to fertilisation (200–80–200 kg ha−1 N–P–K),harvest time (autumn and winter) and plant density (20,000 and 40,000 plants per ha) was evaluated. The energy balance wasassessed considering the energy costs of production inputs and the energy output obtained by the transformation of the finalproduct. The crop yield increased by +50% from the establishment period to the 2nd year of growth when it achieved the highestdry matter yield. The mature crop displayed on average annual production rates of 3 kg dry matter m−2, with maximum values

hed to plant.utlue

le in anhar-

ue). Thisield out-th yearet energyyield and

obtained in fertilised plot and during winter harvest time.Fertilisation mainly enhanced dry matter yield in the initial period (+0.7 kg dry matter m−2 as years 1–6 mean value). T

biomass water content was affected by harvest time, decreasing by about 10% from autumn to winter. With regardensity, higher dry matter yields were achieved with 20,000 plants per ha (+0.3 kg dry matter m−2 as years 1–6 mean value)

The total energy input decreased from fertilised (18 GJ ha−1) to not fertilised crops (4 GJ ha−1). The higher energetic inpwas represented by fertilisation which involved 14 GJ ha−1 (fertilisers plus their distribution) of total energy costs. This varepresents 78% of total energy inputs for fertilised crops.

Giant reed biomass calorific mean value (i.e., the calorific value obtained from combustion of biomass sampadiabatic system) was about 17 MJ kg−1 dry matter and it was not affected by fertilisation, or by plant density orvest time. Fertilisation enhanced crop biomass yield from 23 to 27 dry tonnes per ha (years 1–6 mean val15% increase was possible with an energy consumption of 70% of the overall energy cost. Maximum energy yput was 496 GJ ha−1, obtained with 20,000 plants per ha and fertilisation. From the establishment period to 2nd–6of growth the energy production efficiency (as ratio between energy output and energy input per ha) and the nyield (as difference between energy output and energy input per ha) increased due to the low crop dry biomass

.

∗ Corresponding authorE-mail address:[email protected] (L.G. Angelini).

1161-0301/$ – see front matter © 2004 Elsevier B.V. All rights reserved.

376 L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389

the high energy costs for crop planting. The energy production efficiency and net energy yield were also affected by fertilisationand plant density. In the mature crop the energy efficiency was highest without fertilisation both with 20,000 (131 GJ ha−1) and40,000 plants per ha (119 GJ ha−1).© 2004 Elsevier B.V. All rights reserved.

Keywords: Arundo donaxL.; Giant reed; Biomass; Crop yield; Energy production; Fertilisation; Harvest time; Plant density; Energy efficiency

1. Introduction

Since the 1970’s, biomass crops have attracted in-creasing interest in future European energy suppliesbecause they may satisfy a relevant part of the energydemand in the EU, and at the same time reduce car-bon dioxide emission (CO2) (Venendaal et al., 1997;Hanegraaf et al., 1998). In the Mediterranean environ-ment several perennial grasses are the leading candi-dates to become energy crops because they produce lig-nocellulosic biomass that is ideal for fuel and becausethey also display a good adaptability to such environ-ments. Ideally, energy crops should give high yields ofuseable energy and be produced with minimum energyinput. Unlike annual crops, perennial biomass cropshave the advantage of being able to distribute theirplanting costs (the term “cost” is not meant economi-cally but physically) throughout the cultivation period,thus giving a more favourable energy output/input ratio.In the case of perennial grasses, it has been shown thatthese species have high production levels in high re-

with rapid growth and high yield capacity. Previous re-search carried out on giant reed has highlighted the high

iron-96;algotf theterialsts.

ghen-

,rate

contamination of ground water requires a careful es-timate of crop yield response to nitrogen fertilization.As in other biomass species, the giant reed yield levelcan be influenced by the harvest time and a delayedharvest time in winter may increase biomass yield andincrease dry matter concentration, thus making harvestand storing easy. Another agronomic aspect to be con-sidered is plant density because of its effects on yield,crop establishment and duration, and production costs.Agamic propagation by rhizomes is the chief way toplant this species and one of the most important prob-lems to be solved is the high planting costs, due to thedifficult mechanisation of the propagation practices byrhizomes (Pari and Venturi, 1999).

In order to determine the possibility of reducing cropinput and of improving cultivation methods in giantreed as a biomass crop, we carried out a study to eval-uate the effects of fertilisation, harvest time and plantdensity on its yield and energy balance.

reand rainfall data during the experimental period werecollected from a weather station that was 15 m fromthe experimental site and presented inTable 1. Physi-

evere

as 1,crop

ant

ingken

productive potential in several Mediterranean envments (Faix et al., 1989; Vecchiet et al., 1994, 19Merlo et al., 1998; Shatalov and Pereira, 2001; Hidand Fernandez, 2001; Lewandowski et al., 2003), buless attention has been paid to a better definition obest management system for producing plant mawith the highest yield and the lowest production coApplication of high fertilisers is required for a hibiomass production in energy crops, but the highergy input that this requires is well known (Lockeretz1980). Furthermore, the recent concern with nit

source environments (high inputs) (Petrini et al., 1996;Maryan, 1997; Ercoli et al., 1999; Vecchiet et al., 1994,1996), but there is little information on their ability toadapt to low input cultivation methods. Among peren-nial grasses, giant reed,ArundodonaxL. (Poaceae fam-ily) grows spontaneously and abundantly all over Italy,

2. Materials and methods

2.1. Crop culture

A field study was conducted in Pisa (43◦40′N,10◦19′E), Italy from 1996 to 2001. Air temperatu

cal and chemical soil properties are reported inTable 2.The soil was characterised by a water table that nfell below 1.2 m of top soil during the trial period. Thcrop was planted in 1996 and crop age at harvest w2, 3, 4, 5 and 6 years. The treatments consisted ofmineral fertilisation (200–80–200 and 0–0–0 kg ha−1

N–P–K), harvest time (autumn and winter) and pldensity (20,000 and 40,000 plants per ha).

Planting was carried out on 4 March 1996 usrhizomes with a couple of buds weighing 500 g ta

L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389 377

Table 1Climatic conditions along the experimental period (1996–2001) in comparison with long-term average (1918–1982) at the Rottaia Field Exper-imental Station (Pisa, 43◦N; 10◦E)

Months 1996 1997 1998 1999 2000 2001 Long-term

Rainfall (mm)January 2.8 118.0 73.6 77.7 97.3 92.5 89February 77.5 21.3 56.9 29.4 16.4 24.1 80March 16.1 21.1 49.0 41.9 85.2 123.7 79April 83.1 55.0 97.6 61.1 63.6 61.0 78May 49.7 25.0 24.2 16.6 7.1 34.7 67June 44.5 58.4 44.8 63.4 18.9 15.8 46July 12.2 7.2 1.7 2.8 9.0 18.1 25August 98.1 0 8.8 41.4 12.9 0 42September 91.7 45.9 98.2 160.2 7.7 88.7 84October 114.5 37.5 95.3 124.4 97.7 99.3 134November 224.6 82.4 3.3 131.5 323.3 114.9 118December 122.6 95.6 56.6 75.7 97.3 43.5 99

Total or mean 937.4 567.4 610.0 826.1 836.4 716.3 941

Maximum air temperature (◦C)January 12.6 13.7 12.9 12.5 13.0 12.3 11.2February 11.3 14.8 15.0 12.3 13.9 14.5 12.7March 14.3 17.9 15.8 16.1 15.0 17.6 15.1April 17.9 17.4 17.2 17.7 17.1 18.8 18.3May 21.6 23.2 23.3 23.1 23.4 25.4 22.4June 27.0 25.7 26.4 26.1 26.6 27.7 26.1July 28.3 28.0 28.8 29.6 26.7 30.2 29.1August 28.8 30.8 29.4 29.4 28.2 31.3 29.2September 23.7 27.7 24.9 27.1 25.3 25.9 26.2October 21.1 21.6 21.0 21.5 14.1 26.1 21.3November 16.5 16.1 14.1 15.4 15.3 16.5 15.8December 12.4 13.1 10.5 12.1 13.0 11.9 12.0

Total or mean 19.6 20.8 19.9 20.2 19.3 21.5 20.0

Minimum air temperature (◦C)January 5.3 4.5 3.6 2.4 6.2 6.1 2.8February 2.0 5.1 2.6 0.4 2.5 4.1 2.3March 4.4 4.6 3.0 4.4 5.0 10.0 5.3April 8.7 5.7 8.3 7.2 8.2 7.8 8.1May 11.9 12.0 10.5 12.2 12.5 14.7 11.6June 15.5 16.6 14.4 14.1 15.5 16.4 15.0July 17.1 16.5 16.3 16.8 15.5 19.8 17.2August 17.2 19.1 17.8 18.0 17.0 19.6 17.0September 13.1 16.0 14.1 15.4 14.1 13.4 14.8October 10.9 11.1 11.2 12.2 8.2 14.5 11.0November 9.2 8.1 4.9 5.4 8.5 7.3 6.9December 4.9 4.1 1.5 3.3 6.2 2.8 3.7

Total or mean 10.0 10.3 9.0 9.3 9.9 11.4 9.6

378 L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389

Table 2Soil characteristics (soil horizon 0–20 cm sampled on February 1996before planting)

Sand (2–0.05 mm; %) 25Silt (0.05–0.002 mm; %) 57Clay (<0.002 mm; %) 18pH 7.6Organic matter (%) 2.3Total nitrogen (Kjeldahl Method; g kg−1) 1.1Assimilable phosphorus (Olsen Method; mg kg−1) 16.55Exchangeable potassium (Dirks and Scheffer method;

mg kg−1)222.6

Limestone (%) 2.19Cation exchange capacity (method BaCl2, pH 8.1;

meq/100 g)23.09

CE (conducibility;�S) 292.5

from plants of a local selected clone. The rhizomeswere planted at 10–20 cm of soil depth. Plants werewatered immediately after planting to ensure good rootcontact with the soil.

The previous crop was barley. Tillage was conductedin the autumn of 1995 and consisted of medium-depthploughing (30–40 cm). Seedbed preparation was con-ducted in the spring, immediately before planting, bya pass with a double-disking harrow and a pass witha field cultivator. Pre-plant fertiliser was distributedat a rate of 80 kg ha−1 P (triple superphosphate) and200 kg ha−1 K (potassium sulphate). Nitrogen fertiliserwas applied in the establishment year (1996) as 50%pre-plant and 50% side dressing when giant reed plantswere 0.30–0.40 m tall. In the following years, P andK fertilisers were applied during the winter (approxi-mately at the end of January, after the 2nd harvest time)while N was applied entirely at the start of growth inthe spring (approximately during March). Nitrogen wasapplied as urea. Plots were kept weed-free by hoeing.

The experimental design was a split plot with fertil-isation rates as main plots (20 m× 20 m each), harvesttime treatments as sub-plots (10 m× 10 m each) andplant density as sub-sub-plots (5 m× 5 m each). Eachsub-plot was replicated four times. The main plots were5 m apart.

Giant reed (local ecotype) was grown in 1.00 and0.50 m wide rows at a population of 20,000 and 40,000plants per ha, respectively.

Following each growing season, harvests were car-r tw mi row

Table 3Summary of harvest dates in the different field trials

Year HarvestAutumn Winter

1996 28 October 27 February1997 23 October 17 February1998 28 October 28 January1999 9 November 19 January2000 30 November 8 March2001 11 October 12 February

were not included in the harvested area. Harvest datesin the different years are reported inTable 3. No cropdiseases were detected during the experimental periodand irrigation treatment was never necessary during thedifferent field experiments.

Plants in a 10 m2 area were harvested by cutting 5 cmabove-ground level and weighed to determine freshweight. Height, stem diameter and shoot number werealso determined on a small sub-sample (plants on 2 m2

area).The sub-samples were placed in a forced-draft oven

at 75◦C for 72 h and ground after determination of thedry weight. A homogeneous dried sample of the entireaerial plant part was milled in a Retsch SM1 rotor millto <297�m. The calorific value was determined usinga Leco AC 300 calorimeter according to the ASTMD2015 standard method. Data are reported on a drybasis. The energy yield of giant reed biomass was cal-culated by multiplying the calorific value per unit ofbiomass by the above-ground dry matter yield.

2.2. Energy analysis

Energy analysis of biomass production was carriedout by determining energy costs for machinery fabrica-tion and repairs, for fertiliser and planting material andfor fuel consumption for the various operations. Energycosts for delivering the production outside the field, forstorage and drying were not calculated. In our research,giant reed was propagated by rhizome cuttings. There-f tiono stedb ate-r uesu rialu cropy rgy

ied out on two occasions. One part (10 m2) of each ploas harvested in the autumn and the other part (102)

n the winter. The border plants in the outermost

ore, this method is considered in the determinaf crop energy cost following the approach suggey Heichel (1980). Because the harvested plant mial was anatomically similar to the organs or tisssed for establishing the crop, the quantity of matesed for propagation was subtracted from the totalield. In our analysis, the propagation material ene

L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389 379

cost was considered negligible, since the biomass in-volved was <3% of the total crop yield.

Energy inputs for machinery were determined, fol-lowing Doering (1980), by estimating energy con-sumption for the fabrication and the repair of the ma-chinery utilised for giant reed cultivation, and by cal-culating the annual per ha machinery cost. The energyinput for fabrication and repair parts and materials ofthe machinery was calculated taking into account theirweight and using an energy cost of 386 kJ kg−1 year−1

(Berry and Fels, 1973). We assumed that machineryand tools were used on 200 ha, and machine life was 10years. The energy cost for fertiliser manufacturing was59.9, 5.5, and 5.6 MJ kg−1 for N, P, and K, respectively(Lockeretz, 1980). The fuel costs of various manage-ment operations were calculated by determining dieselconsumption and by multiplying those values by theheat of diesel fuel combustion (44.4 MJ kg−1). Lubri-cant consumption was calculated for each operationand multiplied by the energy coefficient of 80 MJ kg−1.

The energy balance was assessed considering theenergy costs of production inputs and the energy outputobtained by the transformation of the final product. Theefficiency of crop energy production was evaluated asnet energy yield (calculated as the difference betweenenergy output and energy input per ha) and as energyproduction efficiency (as ratio between energy outputand energy input per ha).

2.3. Statistical analysis

byA tald ed;U in-t sp entm levelbG

2

dataf ith1 , arep asm ere

increasing from March to August, with minimum val-ues above 3◦C and maximum values above 15◦C fromthe beginning of March. Maximum values above 28◦Cwere observed from half of July to half of Augustand thereafter a decreasing trend was noticed. FromSeptember to November the minimum air temperaturesnever decreased below 5◦C and the maximum val-ues remained above 15◦C. Air temperatures decreasedrapidly from December to February but never droppedbelow 0◦C. Air temperatures during the growing sea-sons were close to the long-term mean and is suited togiant reed growth that can continue until November.

The summer air maximum temperatures from 1996to 2001 averaged 28.3◦C slightly higher than thosetypical for the long period (28.1◦C). Air temperatureshigher than usual occurred during autumn and winter1997 with favourable conditions for giant reed tillering.Furthermore, summer 2001 was a particular hot seasonwith mean summer air temperatures of 29.7◦C.

There was considerable variability in rainfallamounts and distribution from year to year. The meanannual amount of rainfall observed from 1996 to 2001(738 mm) was below the long-term average (941 mm).Lower annual rainfall values than long-term averagewere observed in 1997, 1998 and 2001. The rainfalldistribution typically shows two peaks, during autumnand spring, with a long dry period from July to August.During the experimentation the amount of rainfall fromJuly to August was 35 mm, below the long-term aver-age (67 mm). Summer 2001 was a particular dry seasonw

3

3

wasa en-s -a( )( ntsa ars.A pro-g tion,h pear-i

For each year, all variables were analysedNOVA using a standard spilt–split–plot experimenesign to test the effects of fertilisation (F: fertilisF: unfertilised), harvesting time (A: autumn; W: w

er), plant density (D2 = 20,000;D4 = 40,000 planter ha), and their interactions. Significantly differeans were separated at 0.05 and 0.01 probabilityy the last significant difference (LSD) test (Gomez andomez, 1984).

.4. Weather conditions

The monthly rainfall and average temperaturesrom 1996 to 2001 experimentation in comparison w918–1982 long-term data for the same locationresented inTable 1. The specific environment wainly characterised by air temperatures, which w

ith only 18.1 mm from July to August.

. Results

.1. Fresh biomass yield

Above-ground fresh biomass yield of giant reedffected by fertilisation (years 1, 3 and 4), by plant dity (years 1–3), by fertilisation× harvest time interction (years 2, 3 and 6), fertilisation× plant densityyear 2), and by harvest time× plant density (year 6Table 4). The differences imposed by the treatmere more obvious during the early crop growth yes the years go by and as the radical apparatusressively deepens, the differences due to fertilisaarvest time and plant density decrease until disap

ng (Table 5).

380 L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389

Table 4Summary of analysis of variance of the above-ground yield of giant reed from the crop establishment (1996) to the 6th year of growth

Above-ground biomass yield

1996 1997 1998

FW (kg m−2) DW (kg m−2) DM (%) FW (kg m−2) DW (kg m−2) DM (%) FW (kg m−2) DW (kg m−2) DM (%)

A ∗ ∗ ∗∗ NS ∗ NS ∗∗ ∗ ∗B NS ∗∗ ∗∗ NS NS ∗∗ NS ∗ ∗∗C ∗ ∗ NS ∗∗ ∗∗ ∗∗ ∗ NS NSA × B NS NS NS ∗∗ ∗∗ ∗ ∗ NS NSA × C NS NS NS ∗ NS NS NS NS NSB × C NS NS NS NS NS NS NS NS NSA × B × C NS NS NS NS NS NS NS NS NS

1999 2000 2001

FW (kg m−2) DW (kg m−2) DM (%) FW (kg m−2) DW (kg m−2) DM (%) FW (kg m−2) DW (kg m−2) DM (%)

A ∗ ∗ NS NS NS NS NS NS NSB NS NS ∗ NS NS ∗ NS NS ∗∗C NS NS NS NS NS NS NS NS NSA × B NS NS NS NS NS NS ∗ NS NSA × C NS NS NS NS NS NS NS NS NSB × C NS NS NS NS NS NS ∗ NS NSA × B × C NS NS NS NS NS NS NS NS NS

A = with fertilisation/without fertilisation; B = autumn harvest/winter harvest; C = 20,000/40,000 plants per ha.∗, ∗∗ and NS significance atP= 0.05 and 0.01 level of probability and non-significant. FW: fresh biomass yield, DW: dry yield, DM: dry matter concentration.

Mean fresh production rapidly increased from theyoung to the mature stand in the 2nd-year-old cropwhen the crop biomass fresh yield reached a peak value(from 4.7 1st year to 8.7 kg m−2 2nd year). The cropbiomass yield became more stable over the following

Table 5Above-ground fresh (FW) and dry yield (DW) and dry matter concentration (DM) of giant reed from the crop establishment (1996) to the 6thyear of growth

1996 1997 1998

FW (kg m−2) DW (kg m−2) DM (%) FW (kg m−2) DW (kg m−2) DM (%) FW (kg m−2) DW (kg m−2) DM (%)

Fertilised 5.2 1.9 36.8 9.5 4.6 48.3 6.7 3.1 46.3Unfertilised 4.2 1.7 41.2 7.9 3.7 47.4 4.5 2.2 48.8Autumn 4.4 1.5 33.7 9.3 4.1 44.5 5.4 2.4 44.4Winter 4.9 2.2 44.3 8.2 4.1 51.2 5.5 2.8 50.920,000 plants per ha 4.0 1.6 38.9 10.8 5.0 46.7 6.5 2.9 44.640,000 plants per ha 5.3 2.1 39.1 6.7 3.3 49.4 4.7 2.3 48.9

1999 2000 2001

FW (kg m−2) DW (kg m−2) DM (%) FW (kg m−2) DW (kg m−2) DM (%) FW (kg m−2) DW (kg m−2) DM (%)

Fertilised 6.1 2.8 45.5 5.5 2.6 46.7 4.7 2.2 45.1Unfertilised 3.8 1.8 46.3 4.2 2.0 46.2 4.0 1.8 46.0Autumn 5.3 2.3 44.1 4.8 2.2 45.4 4.3 1.9 42.4Winter 4.3 2.2 47.8 4.9 2.4 47.6 4.4 2.1 48.720,000 plants per ha 5.4 2.4 44.4 4.9 2.3 46.5 4.2 1.9 45.740,000 plants per ha 4.5 2.1 47.4 4.8 2.2 46.4 4.5 2.1 45.4

years and averaged 5.0 kg m−2 year−1. Concerning fer-tilisation in the 1st trial (years l, 3 and 4), the biomassobtained in the fertilised plots was significantly higher(mean value 6.0 kg m−2) with respect to that obtainedwithout fertilisation (4.2 kg m−2). Even in later years

L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389 381

fertilised plants produce more (5.1 kg m−2) with re-spect to those not fertilised (4.1 kg m−2), but the dif-ference is not significant. The highest fresh yield wasachieved in year 2 with fertilisation.

The harvest time did not affect the biomassyield during the different growing seasons (5.6 and5.4 kg m−2 average value in autumn and winter, respec-tively) (Table 5).

Plant density influenced this characteristic mainlyduring the 1st, 2nd and 3rd growing seasons but in adifferent way. Higher yield was obtained in the year ofplanting with higher plant density, but when the cropwas established, plants with lower density gave higherproduction (Table 5). This behaviour was explainedby the higher plant productivity and greater ability totiller and develop new stalks. The number of stalks persquare meter was greater in 40,000 plants ha−1 than20,000 plants ha−1 in year 1 (23.4 versus 38.1), whilethe opposite situation was observed in year 2 (33.8 ver-sus 26.6) and year 3 (29.2 versus 28.2) (Fig. 1).

In the 2nd, 3rd and in the final year, fresh biomassyield was also affected by the interaction of fertili-sation and harvest time (Table 4). In the plots thatwere given nutrients in the winter, the quantity offresh harvested biomass is significantly higher than inthe plants grown without fertilisation (Table 6). Eventhe interaction between the fertilisation and the plantdensity significantly affected the fresh biomass yield:the best results were obtained in the plants fertilisedw ntshu -tt timea theq rserp eo2 tshw

3

af-f rs 1a tion×

Fig. 1. Number of stems as affected by fertilisation (F: fertilised, UF:unfertilised), harvest time (A: autumn, W: winter) and plant density(D2 = 20,000 plants ha−1, D4 = 40,000 plants ha−1) from the 1st to6th year of growth. Mean values followed by different letters aresignificant different atP = 0.05 (small letter) andP = 0.01 (capitalletter) probability level according to LSD test.

ith sparser plant density (fertilised with 20,000 plaa−1 = 12.6a; fertilised with 40,000 plants ha−1 = 6.5b;nfertilised with 20,000 plants ha−1 = 9.0b; unfer

ilised with 40,000 plants ha−1 = 6.8b kg m−2). Fur-hermore, in year 6 the interaction between harvestnd plant density significantly affected fresh yield:uantity of biomass harvested in winter in the spalanting (3.9 kg m−2) was significantly lower than thne in the denser planting (5.0 kg m−2) (autumn with0,000 plants ha−1 = 4.5ab; autumn with 40,000 plana−1 = 4.1b; winter with 20,000 plants ha−1 = 3.9b;inter with 40,000 plants ha−1 = 5.0a kg m−2).

.2. Dry biomass yield

Above-ground dry matter yield of giant reed wasected by fertilisation (years 1–4), harvest time (yeand 3), plant density (years 1 and 2), and by fertilisaharvest time interaction (year 2) (Table 4).

382 L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389

Table 6Fresh and dry yield (FW, DW), plant height, number of stalks and basal stalk diameter as affected by fertilisation and harvest time

FW (kg m−2) DW (kg m−2) Stalks (n·m−2) Height (cm) Stalk diameter (mm)

Year of establishmentFertilised Autumn 5.0 1.6 26.2 299 10.9

Winter 5.3 2.2 35.6 316 12.1Unfertilised Autumn 3.8 1.3 22.5 254 9.9

Winter 4.6 2.2 38.6 275 12.3

Significance NS NS ∗ NS NS

2nd YearFertilised Autumn 8.2 3.7 28.4 382 13

Winter 10.9 4.7 39.3 406 14Unfertilised Autumn 10.4 4.5 27.8 441 16

Winter 5.5 2.8 25.5 375 13

Significance ∗∗ ∗∗ ∗ ∗∗ ∗∗

3rd YearFertilised Autumn 6.0 2.7 27.7 396 15.2

Winter 7.4 3.5 33.3 446 19.1Unfertilised Autumn 4.9 2.1 27.7 406 15.8

Winter 4.1 2.2 26.4 386 16.8

Significance ∗ NS NS NS ∗

6th YearFertilised Autumn 4.2 1.9 38 273 12.5

Winter 5.1 2.5 44.5 288 16.8Unfertilised Autumn 4.4 2.1 34.4 297 14.3

Winter 3.6 2.1 40.3 326 15.4

Significance ∗ NS NS NS NS

Comparison among the year of establishment and 2nd, 3rd and 6th growing cycle in which significant differences among treatments wereobserved.∗, ∗∗ and NS significance atP = 0.05 and 0.01 level of probability and non-significant.

Because of this characteristic, the differences causedby the treatments are more evident and significant inthe early years, and they tend to disappear in lateryears. The progress of dry biomass production is sim-ilar to the progress of fresh biomass production dis-cussed earlier. From the 1st planting year to the 2ndyear, we can observe a very rapid yield increase andthe highest values during the entire plant density pe-riod are attained (Table 5). Later, the production levelstabilises and we obtain mean annual yield values equalto 2.3 kg m−2 year−1 (mean value years 3–6).

During all the trial years the quantity of dry biomasson the surface unit was always higher in the fertilisedplots (years 1–6 mean value 2.9 kg m−2) than in theunfertilised plots (years 1–6 mean value 2.2 kg m−2).Chemical fertilisers improve dry matter yield aboveall in the 1st year of growth (we observed the biggest

differences between fertilised and unfertilised in the2nd year), while as the roots expanded, the differencesbetween fertilised and unfertilised plots were no longersignificant (Table 5).

Harvesting in winter generally means an increase inyield because in autumn (usually characterised by mildtemperatures) plant growth continues with new stems.Even if the differences between the two harvest timesare significant only in the 1st year of planting (years 1and 3) we generally observe that the dry matter yieldis on average higher if we harvest in winter (years 1–6mean value 2.6 kg m−2) instead of in autumn (years1–6 mean value 2.4 kg m−2) (Table 5).

With a higher investment we obtained, only inthe 1st year, a higher dry yield (1.6 kg m−2 versus2.1 kg m−2 in 20,000 and 40,000 plants ha−1, respec-tively) (Table 5). In the following year the opposite sit-

L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389 383

uation was observed and the higher dry matter yieldswere achieved with the lower density (2.9 kg m−2 ver-sus 2.4 kg m−2 years 2–6 overall mean values). In fact,as observed by other authors (Vecchiet et al., 1996),in the 1st year of cultivation, the production level fordifferent plantation density is generally the result ofthe multiplication of biomass production obtained forsingle rhizomes. In the following years, the yield is notproportional to the density, but a plant with more spacehas a stronger and higher development than one withless growth space.

In the 2nd year of growth, dry matter yield wassignificantly affected by fertilisation× harvest time:plants fertilised and harvested in winter gave a yieldtwo-fold greater than those unfertilised and harvestedin the same period (4.7 kg m−2 versus 2.8 kg m−2)(Table 6).

3.3. Dry matter concentration

The contents of the dry biomass substance is alwayshigher in the plants harvested in winter than in autumn(48 and 42% average winter and autumn value, respec-tively) (Table 5) and this fact is very important becauseit affects the biomass’s transformation and storage pro-cesses. The other factors of variation analysed did notaffect significantly this parameter.

3.4. Plant height

s 1a lantd fer-td -s

theu croph roph m inw res

F UF:u sity( tt ares ll

Plant height was affected by fertilisation (yearnd 5), harvest time (years 1, 2, 5 and 6), and by pensity (year 5). Furthermore, it was influenced by

ilisation× harvest time (year 2); harvest time× plantensity (years 2 and 6) and by fertilisation× plant denity (years 4 and 5) (Table 6).

In year 1, the fertilised plants were higher thannfertilised plants and the average height of thearvested in winter was higher than that of the carvested in autumn (276 cm in autumn and 296 cinter) (Fig. 2). Instead in year 2, the low temperatu

ig. 2. Plant height as affected by fertilisation (F: fertilised,nfertilised), harvest time (A: autumn, W: winter) and plant denD2 = 20,000 plants ha−1, D4 = 40,000 plants ha−1) from the 1so 6th year of growth. Mean values followed by different lettersignificant different atP = 0.05 (small letter) andP = 0.01 (capitaetter) probability level according to LSD test.

384 L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389

recorded after the autumn harvest stopped the growth ofthe plants so that the height of the plants in autumn wasstatistically higher than the height measured in winter(412 and 390 cm in autumn and winter, respectively)(Fig. 2). Fertilisation also interacts with harvest time:the height of the fertilised plants is higher than that ofthe unfertilised plants and in autumn both fertilised andunfertilised plants are always higher than in winter. Thehighest significant plant height was reached by plantsin 20,000 plants ha−1 harvested in autumn (427a versus396b, 388b and 393b cm in autumn with 40,000 plantsha−1, winter with 20,000 plants ha−1 and winter with40,000 plants ha−1). The height was not affected by thetreatments in the 3rd trial year.

In years 4 and 5, plants grown inD2 with fertilisa-tion were significantly higher than those without (389versus 316 cm in year 4; 359 versus 276 cm in year 5).

In year 6, when 40,000 plants ha−1 plants were har-vested in the winter, they were higher than the otherplants (winter with 40,000 plants ha−1 320a versus294b, 295b and 275b cm winter with 20,000 plantsha−1, autumn with 20,000 and 40,000 plants ha−1, re-spectively).

3.5. Number of stalks

The number of stalks per area was significantly af-fected by fertilisation (year 2), harvest time (year 1),and by plant density (years 1 and 2). Furthermore, itw s1 )(

itys lks.F um-b ,000t

emss itha ith2 0paa e itsg fer-t uarem

3.6. Stalk diameter

As a general feature, stalk diameter did not varywith the different treatments, being rather stable andnot significantly correlated with stalk number per unitarea. The stalk diameter was significantly affected byharvest time (years 1 and 2), and by plant density (years2 and 3). Furthermore, it was influenced by fertilisation× harvest time (years 2 and 3), by fertilisation× plantdensity (year 5), fertilisation× harvest time× plantdensity (year 1) (Table 6). In the 1st year, the diame-ter is strongly affected by harvest time: the plants har-vested in the winter, because of the favourable weatherconditions, continue to grow to 12 mm diameter. Alsosignificant is the interaction of all the treatments fromwhich we can conclude that the plants harvested in win-ter, fertilised and with a sparser density have the biggestdiameter. In the 2nd and 3rd year only the plant den-sity significantly influenced the diameter: the diametermeasured in the sparser plantation was bigger than theone measured in the thicker plantation. The harvest inautumn had stems with diameters bigger than in thewinter (Fig. 3).

3.7. Energy analysis

The total energy input of giant reed productionchanged from the establishment to the 2nd–6th yearof growth. It was higher in fertilised crops than unfer-tilised and ranged from 8 to 22 GJ ha−1 in the establish-iy sedb earo sev ts int vely( fer-t hi-z ostsr lcu-l ingy har-v edc nergyc

asswf rtil-

as influenced by fertilisation× harvest time (yearand 2); by fertilisation× plant density (year 2

Table 6).In the 2nd year of growth, plants with lower dens

howed a greater ability to tiller and develop new staertilisation enhanced tillering and therefore the ner of stalks per square meter was greater in 20

han 40,000 plants ha−1 and in fertilised crop (Fig. 1).These two factors interact and the number of st

ignificantly increases in the fertilised plants wlower density than the others (41 fertilised w

0,000 plants ha−1 versus 27 fertilised with 40,00lants ha−1, unfertilised with 20,000 plants ha−1

nd unfertilised with 40,000 plants ha−1). With mildutumn–winter temperatures the crop did not ceasrowth and therefore plants harvested in winter and

ilised showed the highest number of stalks per sqeter (Table 6).

ng year and from 3.5 to 17.5 GJ ha−1 in the followingears (Table 8). The higher energetic input was cauy fertilisation, which involved energy costs each yf 14 GJ ha−1 (fertilisers plus their distribution). Thealues represent 64 and 78% of total energy inpuhe establishing and in the following years, respectiTable 7). Mechanisation required less energy thanilisation and it mainly concerned the planting of romes and the crop harvest. It involved energy canging from 35 to 37% of the total energy costs caated in the 1st year for fertilised crops. In the followears the energy costs for the plant density and theest are only 22% of the total. With the unfertilisrop the mechanisation costs represent the total eosts.

The average calorific value of dry giant reed biomas 16 MJ kg−1in the establishing year and 18 MJ kg−1

rom the 2nd to 6th year. It was not affected by fe

L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389 385

Table 7Energy input (MJ ha−1) for the production of giant reed biomass considering the fertilised and non-fertilised crop with 20,000 (D2) and 40,000(D4) plants per ha

Fertilised Unfertilised

D2 (20,000 plants ha−1) D4 (40,000 plants ha−1) D2 (20,000 plants ha−1) D4 (40,000 plants ha−1)

Crop establishmentTillage (plowing + harrowing) 3102 3102 3102 3102Planting rhizomes 419 837 419 837Fertiliser distribution 491 491N fertilisera 11980 11980P fertilizer 440 440K fertilizer 1120 1120Hoeing (1 time) 703 703 703 703Harvest 3473 3473 3473 3473

Total 21728 22146 7697 8115

Cropping operations from 2nd yearb

Fertiliser distribution 491 491N fertilizer 11980 11980P fertilizer 440 440K fertilizer 1120 1120Harvest 3473 3473 3473 3473Total from 2nd year 17504 17504 3473 3473

Total 179264 179682 38954 39372

Mean value per yearc 17926 17968 3895 3937

a The energy costs for fertiliser manufacturing was 59.9, 5.5, and 5.6 MJ kg−1 for N, P and K, respectively (Lockeretz, 1980).b Mean value of years 2, 3, 4 and 5.c 10 Years estimated cultivation cycle.

isation, or by plant density or harvest time, and there-fore the effects of treatments on crop energy yield wereequal to those on biomass yield (Table 8). Fertilisa-tion enhanced crop biomass yield from 23 to 27 drytonnes per ha as overall mean from the 1st to 6th yearof growth. This 15% increase was possible with an en-ergy consumption of 70% of the overall energy cost(Tables 7 and 8). The maximum energy yield outputof giant reed was 496 GJ ha−1, obtained with 20,000plants per ha and fertilisation. On the contrary, withoutfertilisation and with 40,000 plants per ha the energyyield output was 387 GJ ha−1.

If we consider the mature crop (from the 2nd to 6thyear), the energy output considerably increased due toa higher dry crop yield and to its higher calorific value(18 MJ kg−1). The energy input decreased because theplanting operation costs were not included, thereforethe efficiency of crop energy production increased. Theefficiency of crop energy production was evaluated as

net energy yield and as energy production efficiency.From the establishment year to the following growingseasons, the energy efficiency increased (from 25 to77) as well as the net energy yield (280 to 463 GJ ha−1)(Table 8). This was explained by the low crop dry matteryield and high energy costs for crop planting in year1. One advantage of a perennial energy crop such asgiant reed is linked to the possibility of distributing thecrop establishment costs over a longer period (a cyclelength of 10 years was considered here) thus decreasingtheir incidence in the 1st year. Likewise, the lower yieldin the establishment yield is distributed over a longerperiod and its negative influence on energy efficiencydecreases.

If we consider the overall crop cycle (1st–6th year),we can see that net energy and the energy produc-tion efficiency were affected by fertilisation and plantdensity. With fertilisation, the crop with 20,000 plantsper ha showed the highest significant value of net en-

386 L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389

Table 8Global energy balance for giant reed. Mean values of fertilised and non-fertilised crop with 20,000 and 40,000 plant densities

Arundo donaxL. Dry yield(t ha−1)

Energy input(GJ ha−1)

Energy output(GJ ha−1)

Efficiency of energyoutput/input

Net energy yieldoutput−input(GJ ha−1)

Crop establishmentFertilisedD2 18.2 21.73 291.2a 13 269.5FertilisedD4 20.5 22.15 328.0 15 305.9UnfertilisedD2 13.3 7.70 212.8 28 205.1UnfertilisedD4 21.7 8.12 347.2 43 339.1

Mean 18.4 14.9 294.6 25 279.9

Years 1–6 cropFertilisedD2 29.15a 17.93 496a 28 478.1FertilisedD4 24.71b 17.97 421b 23 403.0UnfertilisedD2 23.24b 3.89 395b 102 391.1UnfertilisedD4 22.77b 3.94 387b 98 383.1

Mean 25.0 10.9 425.0 63 413.8

Years 2–6 cropFertilisedD2 31.3a 17.50 563a 32 545.5FertilisedD4 25.6b 17.50 461b 26 443.5UnfertilisedD2 25.20b 3.47 454b 131 450.5UnfertilisedD4 23.00b 3.47 414b 119 410.5

Mean 26.3 10.5 473.0 77 462.5

Data were submitted by ANOVA analyses considering fertilisation as main factor and plant density as 2nd factor of variation in a split-plotscheme. Mean dry yield and energy output values followed by the same letter are not significantly different forP < 0.05 according to LSD test.

a Calorific mean value per kg of oven dry weight biomass: in the crop establishment 16.0 MJ kg−1; from the 1st to 6th growing cycle17.0 MJ kg−1; from 2nd to 6th growing cycle 18.0 MJ kg−1.

ergy yield (478 GJ ha−1), while the highest energy ef-ficiency (102) was obtained in the same crop withoutfertilisation. Non-significant differences were observedbetween the two unfertilised crops with different densi-ties (391 and 383 GJ ha−1 for unfertilised with 20,000plants ha−1 and unfertilised with 40,000 plants ha−1,respectively). The energy efficiency was 102 and 98,respectively (Table 8).

When considering the mature fertilised crop (fromthe 2nd to 6th year), a significant higher energy effi-ciency and higher net energy yield with 20,000 than40,000 plant density was observed (32 and 26; 546and 444 GJ ha−1, respectively). Without fertilisation,no significant differences between the two plant den-sities were observed, even though with 20,000 plantsper ha, the highest value of energy efficiency (131)was obtained. Thus, in order to improve energy effi-ciency it was better to produce biomass from an unfer-tilised crop with 20,000 plants per ha. If fertilisationwas applied, it was better to use a lower plant density

(20,000 plants per ha) to improve net energy and effi-ciency (Table 8).

4. Discussion

The results of this study give indication of the gi-ant reed biomass production along the first 6 years ofgrowth as affected by different management practicesand contribute to a better knowledge of this crop in theMediterranean countries. Even if this species growsspontaneously and abundantly all over Italy, there arenot many research data on fertilisation, harvest time andplant density response of giant reed in the country. Ourresults have demonstrated that biomass yield rapidlyincreased from the young to the mature stand in the2nd-year-old crop, when the highest crop yields wereachieved with an increase of +50% from the year of es-tablishment. This agrees with other experiments donein southern regions of Italy (Foti and Cosentino, 2001)

L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389 387

Fig. 3. Stem diameter as affected by fertilisation (F: fertilised, UF:unfertilised), harvest time (A: autumn, W: winter) and plant density(D2 = 20,000 plants ha−1, D4 = 40,000 plants ha−1) from the 1stto 6th year of growth. Mean values followed by different letters aresignificant different atP = 0.05 (small letter) andP = 0.01 (capitalletter) probability level according to LSD test.

where the highest dry yield of 1.5–3.4 kg DM m−2 wasobtained in year 2.

From the 2nd year after planting, giant reed dis-played annual production rates of 3 kg DM m−2 withmaximum values achieved in fertilised plot and in win-ter harvest time.

The biomass yield potential of giant reed in thetested environment is similar to that observed by otherauthors in the north-east of Italy (Vecchiet et al., 1996)but higher than those reported byFoti and Cosentino(2001) in the arid environment of Sicily and byFaixet al. (1989)in the south of France (2–2.5 kg DM m−2).

Fertilisation enhanced dry matter yield mainly inthe initial years (+0.7 kg DM m−2 as year 1–6 meanvalue). Furthermore, N amount is an important fac-tor, which must be considered in order to maximiseplant utilisation and reduce losses. Water content ofgiant reed feedstocks depends on the harvesting time.The biomass water content ranges from 52 to 58% andit is considerably high to ensure good biomass stora-bility, thus the harvested material must be dried. Dataobtained from this experiment showed that feedstockwater content decreased by about 10% from autumn towinter. Winter harvest has other benefits for this en-vironment characterised by frequent and intensive au-tumn rainfall that prevents harvest or causes damage tosoil structure and physical properties.

As regards plant density, higher dry matter yieldswere achieved with the lower density of 20,000 plantsper ha, which produced a greater number of tillers perp unita rkedo den-s , infi ro-d isc freshsp sesso atedt dif-f gatedb lowp se ofr g, aso s-p ea ist nual

lant and therefore enhanced yield production perrea. With deep and fresh soils like the ones we won, it is possible to noticeably decrease the plantity without changing productivity. On the contraryelds characterised by low water retention, high puction yields are tied to high plant density. Thisonfirmed by studies carried out on the deep andoil in the Friuli Region byVecchiet et al. (1996)com-aring from 20,000 to 40,000 plants per ha to asptimal giant reed plant density. They demonstr

hat in the 1st two growing seasons, non-significanterences were observed. Giant reed crop is propay plant-rhizomes and the possibility of adoptinglant density has positive economic effects becaueduced energy costs of mechanical transplantinutlined byPari (1998). Another particular positive aect that must be considered in Mediterranean ar

hat giant reed can be grown successfully as polian

388 L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389

species because crop survival over the winter is high.On the contrary in central Europe giant reed behavesas an annual energy cop for the low soil temperaturesand lack of winter hardiness of the rhizomes.

Among the different herbaceous species the polian-nual rhizomatous crops, such as giant reed, proved tohave more significant interest due to their favourableenergy balance in comparison with annual herbaceouscrops that require about 50% of the total energy fortillage and seeding operations. The ratio of energy out-put to energy input in the production of giant reed inthe 1st 6 years of growth is 26–100 for fertilised tounfertilised crop. This value is thus much better thanthat of grain crops at 8.5 (Lewandowski et al., 1995)or that ofMiscanthus sinensisat 15–20, as reported byLewandowski and Kicherer (1997)and byErcoli et al.(1999)(22–47 depending on irrigation and/or nitrogensupply).

5. Conclusion

The main objectives of the present study were toevaluate the effects of fertilisation, harvest time andplant density on giant reed yield and energy bal-ance. After 6 years of observation, giant reed appearsparticularly suited for the cultivation environment ofcentral Italy because of its high biomass yield and for itsfavourable energy balance. The possibility of reducingcrop inputs, such as fertilisation and plant density, leadt on.T on-s op int s ofu re-v ofg s ofge chiete ino,2 t al.,2 m-m ge-m

deso ont estt nce.

It towards developing better management practices forthis crop in order to improve the efficiency of biomassproduction in Italy.

In conclusion, among the biomass herbaceous crops,the perennial giant reed is of particular interest forMediterranean environments for its high biomass yieldand favourable energy balance.

References

Berry, R., Fels, M.F., 1973. The production and consumption of au-tomobiles. An energy analysis of the manufacturing, discard, andreuse of the automobile and its component materials. Universityof Chicago Press, Chicago, IL.

Cristou, M., Mardikis, M., Alexopoulou, E., 2001. Propagation ma-terial and plant density effects on theArundo donaxyields. In:Kyritsis, S., Beenackers, A.A.C.M., Helm, P., Grassi, A., Chiara-monti, D. (Eds.), Biomass for Energy and Industry. Proceedingof the First World Conference, Sevilla, Spain, 5–9 June. Jamesand James, Science Publishers Ltd, London, pp. 1622–1628.

Doering III, O.C., 1980. Accounting for energy in farm machineryand buildings. In: Pimentel, D. (Ed.), Handbook of Energy Util-isation in Agriculture. CRC Press, Boca Raton, FL, pp. 9–24.

Ercoli, L., Mariotti, M., Masoni, A., Bonari, E., 1999. Effect of irri-gation and nitrogen fertilisation on biomass yield and efficiencyof energy use in crop production ofMiscanthus. Field Crops Res.63, 3–11.

Faix, O., Meier, D., Beinhof, O., 1989. Analysis of lignocellulose andlignins fromArundo donaxL. andMiscanthus sinensisAndress.,and hydroliquefaction ofMiscanthus. Biomass 18, 109–126.

Foti, S., Cosentino, S.L., 2001. Colture erbacee annuali e poliennalida energia. Riv. Agron. 35, 200–215.

G gri-

H ess-ops.

H ropa-k ofFL,

H opu--

ers,assfer-e Pub-

L ey

L ass:qual-

o improve the efficiency of crop energy productihe results of the conducted field experiments demtrated that this crop can be grown as an energy crhe Mediterranean area of Europe giving high yieldseable energy with minimum energy input for its. Pious results outlined the high potential productivityiant reed, but data were limited to the initial yearrowth or at least to the fourth growing period (Faixt al., 1989; Venendaal and Stassen, 1996; Vect al., 1996; Cristou et al., 2001; Foti and Cosent001; Shatalov and Pereira, 2002; Lewandowski e003) and further investigation were strongly recoended to improve the knowledge of crop manaent.The present long-term field experiment provi

riginal information along 6 years of investigationshe effects of fertilisation, plant density and harvime on giant reed biomass yield and energy bala

omez, K.A., Gomez, A.A., 1984. Statistical procedures for acultural research. Wiley, New York.

anegraaf, M., Biewinga, C., Gert van der Bijl, E., 1998. Assing the ecological and economic sustainability of energy crBiomass Bioenergy 15, 345–355.

eichel, G.H., 1980. Assessing the fossil energy costs of pgating agricultural crops. In: Pimentel, D. (Ed.), HandbooEnergy Utilisation in Agriculture. CRC Press, Boca Raton,pp. 27–33.

idalgo, M., Fernandez, J., 2001. Biomass production of ten plations of giant reed (Arundo donaxL.) under the environmental conditions of Madrid (Spain). In: Kyritsis, S., BeenackA.A.C.M., Helm, P., Grassi, A., Chiaramonti, D. (Eds.), Biomfor Energy and Industry. Proceeding of the First World Conence, Sevilla, Spain, 5–9 June. James and James, Scienclisher Ltd, London, pp. 1181–1184.

ewandowski, I., Kicherer, A., Vonier, P., 1995. CO2-balance for thcultivation and combustion ofMiscanthus. Biomass Bioenerg8, 81–90.

ewandowski, I., Kicherer, A., 1997. Combustion quality of biompractical relevance and experiments to modify the biomassity of Miscanthus×Giganteus. Eur. J. Agron. 6, 163–177.

L.G. Angelini et al. / Europ. J. Agronomy 22 (2005) 375–389 389

Lewandowski, I., Scurlock, J.M.O., Lindvall, E., Christou, M., 2003.The development and current status of perennial rhizomatousgrasses as energy crops in the US and Europe. Biomass Bioenergy25, 335–361.

Lockeretz, W., 1980. Energy inputs for nitrogen, phosphorus, andpotash fertilisers. In: Pimentel, D. (Ed.), Handbook of EnergyUtilisation in Agriculture. CRC Press, Boca Raton, Florida, pp.23–26.

Maryan, P.S., 1997. Future prospects for energy crops in the UK. In:Bullard, M.J., Ellis, R.G., Heath, M.C., Knight, J.D., Lainsbury,M.A., Parker, S.R. (Eds.), Biomass and Energy Crops AAB, vol.49, pp. 449–454.

Merlo, L., Morgana, B., Sardo, V., La Loggia, F., 1998. Experienceswith giant reed and perennial C4 grasses in Sicily. In: Kopetz, H.,Weber, T., Pals, W., Chartier, P., Ferrero, G.L. (Eds.), Proceed-ings of the International Conference on Biomass for Energy andIndustry. Wurzburg, Germany, 8–11 June. CARMEN, RimparGermany, pp. 1006–1008.

Pari, L., 1998. First tests of anArundo donax(giant reed) rhizomesharvester. In: Kopetz, H., Weber, T., Pals, W., Chartier, P., Fer-rero, G.L. (Eds.), Proceedings of the International Conferenceon Biomass for Energy and Industry. Wurzburg, Germany, 8–11June. CARMEN, Rimpar Germany, pp. 824–826.

Pari, L., Venturi, P., Propagazione delle colture da biomassa polien-nali rizomatose. Atti del XXXIII Convegno annuale della SocietaItaliana di Agronomia, Le colture “non alimentari” Legnaro (PD)20–23 Settembre, pp. 75–76.

Petrini, C., Bazzocchi, R., Bonari, E., Ercoli, L., Masoni, A., 1996.Effect of irrigation and nitrogen supply on biomass productionfromMiscanthusin Northern-Central Italy. Agricoltura Mediter-ranea 126, 275–284.

Shatalov, A.A., Pereira, H., 2002. Influence of stem morphology onpulp and paper properties ofArundo donaxL. reed. Ind. CropsProd. 15, 77–83.

Shatalov, A.A., Pereira, H., 2001.Arundo donaxL. (giant reed)as a source of fibres for paper industry: perspectives for mod-ern ecologically friendly pulping technologies. In: Kyritsis, S.,Beenackers, A.A.C.M., Helm, P., Grassi, A., Chiaramonti, D.(Eds.), Biomass for Energy and Industry. Proceeding of the FirstWorld Conference, Sevilla, Spain, 5–9 June. James and James,Science Publisher Ltd, London, pp. 1183–1186.

Vecchiet, M., Jodice, R., Smedile, E., Parrini, F., 1994. La cannacomune: riscoperta di una possibilita produttiva. Quaderni Agri-colture e Innovazione 30/31, 78–85.

Vecchiet, M., Jodice, R., Schenone, G., 1996. Agronomic researchon giant reed (Arundo donaxL.) management system and culti-vation of two different provenances. In: Proceedings of the NinthEuropean Bioenergy Conference, Copenhagen, 24–27 June, pp.644–648.

Venendaal, R., Stassen, H., 1996. European Energy Crops Overview.Country Report for Italy. EU FAIR Contract n. FAIR1-CT95-0512 (http://www.nf-2000.org/secure/Fair/S331.htm).

Venendaal, R., Jorgensen, U., Foster, C.A., 1997. European energycrops: a synthesis. Biomass Bioenergy 13, 147–185.