This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Author's personal copy b i o m a s s a n d b i o e n e r g y 4 6 ( 2 0 1 2 ) 6 4 5 e6 5 2

Available online at www.sciencedirect.com

http://www.elsevier.com/locate/biombioe

The effect of density on short rotation Populus sp. plantations in the Mediterranean area I. Can˜ellas a,*, P. Huelin a, M.J. Herna´ndez a, P. Ciria b, R. Calvo c, G. Gea-Izquierdo a, H. Sixto a a

INIA-CIFOR, Departamento de Selvicultura y Gestio´n de los Sistemas Forestales, Ca La Corun˜a km 7.5, 28040 Madrid, Spain CEDER-CIEMAT, Departamento de Energı´a, Autovı´a de Navarra, salida 56, 42290 Lubia, Soria, Spain c INIA, Servicio de Biometrı´a, Ca La Corun˜a km 7.5, 28040 Madrid, Spain b

article info

abstract

Article history:

In this study, the effect of density on poplar short-rotation forestry plantations was

Received 3 November 2011

analyzed in the context of maximizing biomass production under Mediterranean condi-

Received in revised form

tions. Data from 12 experimental sites in Spain with densities ranging from 6666 to 33,333

7 February 2012

cuttings$ha
1 were used. At the end of the first year, biomass production from the different

Accepted 27 June 2012

poplar plantations differed significantly among sites and densities. Furthermore, after one

Available online 25 July 2012

year of growth, biomass production was found to increase exponentially with plot density. However, after three years of growth, at the end of the first rotation, significant differences

Keywords:

were not found. When we analyzed the variables that affect the production, we found that

Poplar

weed control and therefore a reduction in competition explained much of the variability.

Density

The greatest biomass production was achieved in high density plots in more southern

Short rotation forestry

latitudes where high spring temperatures benefit trees growing in well-watered soils.

‘I-214’clone

Considering on the one hand the scope of current weed control methods and on the other,

Biomass production

the cost of plantation and/or the difficulties associated with management and mechanisation, plantation densities of more than 15,000 plants ha
1 would not be recommendable. ª 2012 Elsevier Ltd. All rights reserved.

1.

Introduction

In the current general context of climatic change, bio-energy is one of the key options in the short and medium term to mitigate greenhouse gas emissions [1]. Additionally, rural development could benefit from bio-energy guidelines which, in accordance with the rural development policies 2007e2013 [2], provide a viable alternative use for large marginal areas of agricultural and forestry land. In Short Rotation Forestry (SRF), fast growing species are grown under intensive management regimes using rotation periods of 2e10 years [3,4]. Although many studies have been undertaken with regard to SRF [5,6],

there is still a need for further research into many aspects related to the plantations. This research should focus on three key aspects: (i) improvement of genetic material and its adaptation to different sites, (ii) plantation design and rotation length, and (iii) tending operations such as weed control, fertilization and irrigation [7]. Lignocellulosic energy crops of poplar and willlow are now considered one of the most promising options in Europe for biomass production in the short term [8]. In the Mediterranean area, poplar (Populus spp.) is more suited to the edaphoclimatic conditions than Salix [7]. In terms of cultivation, poplar has a number of advantageous characteristics such as

* Corresponding author. Tel.: þ34 913476867; fax: þ34 913476767. E-mail address: [email protected] (I. Can˜ellas). 0961-9534/$ e see front matter ª 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biombioe.2012.06.032

Author's personal copy 646

b i o m a s s a n d b i o e n e r g y 4 6 ( 2 0 1 2 ) 6 4 5 e6 5 2

fast growth and ability for vegetative propagation and resprouting [9]. Additionally, the genus has a high potential for genetic improvement because of its broad genetic diversity and small genome (ca. 520 Mbp), which has recently been sequenced [10,11]. In the medium to long term, these features allow improvements to be made according to the purpose of production. The relationship between individual tree growth and stand growth depends on the balance of the trade-off between individual growth and competition. In turn, this relationship depends on external factors and will vary according to the species, ecological conditions and/or management regimes. Plantation density is one of the most important factors to be considered in SRF plantations for biomass production. Stools growing at low density are able to maximize individual plant growth. As density is increased, if individual plants still receive enough resources for their individual growth, then total yield rises initially in proportion to the number of plants per area. However, there is no specific density at which optimal production is attained, but rather a range within which production remains constant [12]. As regards the economic profitability of the crop, other factors apart from biomass yield must be considered such as plant material/plantation cost, weed control strategies or the operational capacity of the machinery employed. A number of studies concerning planting density in SRF have been undertaken over the last 20 years [13e15], although a comparison of the results obtained is not straightforward due to differences in plant material used, site conditions or cultural treatments such as irrigation. Several studies concerning the use of poplar as an energy crop have been published. The densities employed in these trials range from 5000 plants ha
1 up to 40,000 plants ha
1 [16,17], although few of them have compared production at two or more densities at the same time under similar ecological conditions [13,18]. The main objective of this study is to determine the effect of density on biomass production in SRF plantations for energy purposes under Mediterranean conditions. Our initial hypothesis is that total biomass production is similar over a wide range of densities and therefore the optimum density

in terms of economic viability will be the lower limit at which maximum biomass production is attained. The specific objectives are: i) to evaluate the effect of density on biomass production after the first (establishment) year and at the end of the first rotation (three-year old plants); ii) to evaluate the interaction between site characteristics (climate, soil and management) and density and determine the most relevant variables that could have an influence on biomass production in the study scenario.

2.

Material and methods

2.1.

Site description and data sampling

The study was carried out using data from different SRF trials located at 12 sites in Spain. To evaluate clone behaviour and to study the effect of density on growth and biomass production, the ‘I-214’ clone (Populus x canadensis (Dode) Guinier) was planted at these sites. The different sites covered a broad range of climatic and edaphic conditions within the Mediterranean climate, as shown in Table 1 and Fig. 1. The biomass production experimental sites include trials with areas ranging from 0.25 ha to 4 ha. Plots were established in early spring between 2005 and 2009, with 9e25 stools in each of the 2e11 replications per site and plant densities ranging from 6666 cuttings$ha
1 to 33,333 cuttings$ha
1 (3 m spacing between rows) (Table 2). Plantations were established using hardwood cuttings of 20e30 cm in length, planted in rows either by hand or with adapted planting machines. As is common practice in SRF, the whole trial area at each site was fertilized during soil tillage according to the specific soil characteristics and Oxifluorfen (4 l ha
1) was applied immediately after the cuttings were planted to control weed establishment and competition. Weed control, when necessary, was done mechanically. Irrigation was applied in all plots either by drip or flooding to site field capacity. Coppicing was undertaken at the end of the establishment year in order to promote sprouting at the S01, S02 and S11 sites.

Table 1 e Climate and soil (top 30 cm) properties at the 12 study sites. Sites

Coordinates Latitude

S01 S02 S03 S04 S05 S06 S07 S08 S09 S10 S11 S12

41 41 42 42 42 37 42 41 42 40 42 41

360 280 370 360 370 120 040 420 070 310 100 500

N N N N N N N N N N N N

Altitude m asl

P mm

T C

pH

OM %

Sand %

Silt %

Clay %

1090 827 560 540 560 602 20 52 730 595 263 705

625 560 707 845 693 429 703 729 483 464 450 492

9.7 12.0 11.8 10.9 11.8 16.5 15.1 13.3 11.7 13.8 14.0 12.0

6.8 8.7 7.1 6.0 5.9 8.6 8.5 6.1 8.2 8.4 8.6 5.4

0.6 2.5 0.5 0.9 0.8 0.9 0.9 0.58 1.0 0.5 1.4 1.5

86.2 11.0 48.3 60.3 56.8 40.0 67.8 77.9 72.9 68.9 28.4 57.0

5.6 56.5 29.6 23.7 23.6 35.3 16.3 10.3 10.3 15.3 36.4 20.0

8.1 32.5 22.2 16.0 19.5 24.7 15.9 11.9 16.8 15.8 35.2 21.0

Longitude 2 2 6 6 6 3 3 2 5 3 1 5

300 330 390 370 410 420 040 400 350 180 400 530

W W W W W W E E W W W W

P: annual rainfall, T: average annual temperature, OM: organic matter.

Author's personal copy b i o m a s s a n d b i o e n e r g y 4 6 ( 2 0 1 2 ) 6 4 5 e6 5 2

647

Fig. 1 e Location of the twelve experimental sites included in this study.

To study clone performance in relation to site and plant density, diameter (mm) over bark at 10 cm (d10) and at 130 cm (dbh) were measured once leaves had fallen using a digital calliper and total height (cm) of the highest stool (TH) was measured using a pole. These measurements were also evaluated at two different stages: (i) at all twelve sites after one vegetative period; (ii) at three sites at the end of the first rotation (3 yr old plants). In addition, two other variables were analyzed in this study: total plot basal area and aboveground biomass production. Basal area (BA) was calculated from the dbh of each independent stool, measured after the first and third vegetative periods. Biomass production was evaluated during the winter months when there were no leaves on the trees at the three locations (S01, S02, and S11) where the first rotation period (3 years) had ended. For these sites we calculated total aboveground dry biomass (TB) from the estimated dry weight of randomly selected whole plants from each of the plots after oven-drying to constant weight at 100  C. Plot basal area, including from 9 to 25 stools, (see Table 2) was used as an indicator of production due to the high correlation observed between stool biomass and diameter of dominant stools [7,19]. To evaluate the success of the weed control treatment applied we used a visual estimate of the percentage of weed cover in the plots according to the following scale: i) good weed control when weed cover was less than 10%, ii) medium level effectiveness when weed cover was between 10% and 50%, and iii) poor weed control when cover was more than 50%. Monitoring was carried out twice in each growing season

and involved three observers. The temperature data were provided by the Spanish National Meteorological Agency (AEMET) from the closest meteorological station to the trial sites.

2.2.

Statistical analysis

Prior to testing the effect of plant density on production, we explored the variables for normality. A multifactor analysis of variance (ANOVA) was performed for each of the two response variables (BA, in m2 ha
1, and TB, in t DM ha
1). Two fixed factors were taken into account: plantation density (DENSITY, 7 levels for the first vegetative period, and 4 levels for the third), and site (SITE, 12 levels for the first vegetative period and 3 levels for the third vegetative period). The standard model for the experimental design is expressed as follows: yijk ¼ m þ ai þ bj þ ðabÞij þ ekðijÞ where yijk is the response variable (plot BA and TB) with density i at site j in plot k, m is the global mean, ai is the effect of ith density, bj is the effect of jth site, (ab)ij is the interaction of the i level density factor with the j level site factor, and ek(ij ) is the independent random error following a normal distribution N(0,s2). Due to the fact that the data were unbalanced (meaning that not every DENSITY was present at every SITE) we used the Type IV Sum of Squares [20] and have only used the ANOVA analysis to determine the significance of the

Author's personal copy 648

b i o m a s s a n d b i o e n e r g y 4 6 ( 2 0 1 2 ) 6 4 5 e6 5 2

Table 2 e Crop density, vegetative period evaluated and mean stem size. dbh: diameter at 1.30 m. Sites

S01

S02 S03

S04 S05 S06

S07

S08 S09 S10

S11

S12

Density stump/ha

Replications

Number of evaluated stumps/plot

Age (vegetative periods)

Mean dbh 1st year (mm)

Mean dbh 3rd year (mm)

15,000 25,000 33,333 20,000 10,000 13,333 15,000 20,000 13,333 13,333 10,000 13,333 15,000 20,000 6666 10,000 13,333 15,000 20,000 6666 13,333 10,000 33,333 10,000 20,000 33,333 15,000 20,000 25,000 33,333 6666 10,000 13,333 15,000 20,000

3 3 3 8 2 4 2 3 4 4 3 8 3 3 5 3 6 3 3 3 3 3 4 3 3 4 3 7 3 3 4 3 11 3 3

16 16 16 16 16 16 16 16 25 25 16 16 16 16 16 16 16 16 16 25 25 9 9 9 9 9 16 16 16 16 25 16 16 16 16

1 and 3 1 and 3 1 and 3 1 and 3 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 and 3 1 and 3 1 and 3 1 and 3 1 1 1 1 1

10.81 8.83 9.05 4.29 6.93 9.95 7.85 6.03 7.52 7.76 15.69 13.45 15.88 15.17 9.64 8.40 9.05 7.48 6.24 3.94 5.79 5.60 10.75 7.87 11.39 15.70 9.45 16.91 12.24 10.77 10.37 10.00 10.00 10.82 7.17

41.17 31.62 26.13 46.82

interaction between our two factors. As the interaction was significant we analyzed the density for each site and the site for each density through “slice” analysis [21]. Least-squares means multiple comparison testing was carried out between two levels of each effect when these were significant in the ANOVA. To study the relationship between BA and the independent covariates, a multivariate analysis was performed. For the purposes of the multivariate analyses, BA was transformed to an ordinal response of 0, 1 and 2. The value of production was assigned to group “0” if the sample had less than 0.65 m2 ha
1, “1” if the sample had between 0.65 and 1.8 m2 ha
1, and “2” if it was taken from plots over 1.8 m2 ha
1 (Table 3). Covariates included cultural planting factors such as planting density and weed control; environmental factors such as latitude, altitude and mean temperature in spring; as well as other factors relating to the influence of plot characteristics, such as size of plantation, which might have a significant influence on estimated production (Table 3). Since the analyses included both categorical and continuous covariates, we first explored the relationship between dependent and independent covariates using multinomial logistic regression [22]. In a second

42.42 41.17 41.59 41.42

step, we carried out a canonical discriminant analysis (CDA) using the continuous covariates selected in the logistic regression to explain the biomass production according to different independent variables (Table 3). The level of significance in the analyses was set to p ¼ 0.05 and the SAS statistical software version SAS 9.1 [21] was used.

3.

Results

3.1.

Average effect of density and site on yield

Overall, our results suggest that total basal area increases with increasing density after the first year of establishment (Fig. 2). After the first year of establishment, the interaction of the two factors DENSITY and SITE was significant (F ¼ 7.22, p < 0.001). Thus the differences within each of these factors had to be analyzed separately at the different levels of the other factor. Significant differences as regards productivity were only found between different densities at four of the nine sites considered in the analysis: S06 (F ¼ 6.86, p  0.001), S09 (F ¼ 15.55, p  0.001), S10 (F ¼ 53.10, p  0.0001) and S11 (F ¼ 15.73, p  0.001). When

Author's personal copy 649

b i o m a s s a n d b i o e n e r g y 4 6 ( 2 0 1 2 ) 6 4 5 e6 5 2

Fig. 2 e Growth expressed as basal area in different density plots (number of cuttings per hectare) after the first vegetative period of growth. the all study sites were analyzed according to density levels, significant differences were only identified between sites for densities higher than 10,000 cuttings ha
1, although as can be seen from Fig. 3, not all levels of density within those four sites were significantly different. However, at the end of the first rotation (three-year old plants), densities above 20,000 cuttings ha
1 did not yield higher productions, as can be seen in Fig. 4. At the end of the first rotation, significant differences were found in overall mean BA (F ¼ 8.76; p  0.001) and biomass (F ¼ 12.97; p  0.001) for the different sites analyzed, although not between densities (Fig. 5). These differences were robust within factors, since the interaction was non-significant.

3.2.

Other factors explaining poplar yield

In the exploratory multinomial logistic regression and after a stepwise selection, the significant covariates selected were: weed control, plant spacing, latitude, altitude and mean

spring temperature (different climatic covariates were analyzed and the most significant proved to be spring temperature, data not shown). As weed control was the covariate which explained the greatest variability in the logistic model (c2 ¼ 41.75; p  0.001), the canonical discriminant analysis (CDA) was carried out separately in plots with ‘good’, ‘medium’ and ‘poor’ weed control. In plots with poor or medium weed control, no significant covariate was identified which explained biomass production when running the CDA (data not shown). However, in the analysis of those plots where good weed control was achieved; density, latitude, altitude and mean temperature in spring were selected. CDA only revealed substantial differences among yield classes in those plots where weed control was ‘good’, giving a squared canonical correlation of 0.67 for can1 and 0.17 for can2, indicating higher discriminatory power in the first canonical variable. The total-sample correlations between can1 and the original variables were 0.67 for density,
0.44 for altitude, 0.51 for mean temperature in spring and
0.46 for latitude.

Table 3 e Categorization of the dependent variable (BA) and covariates used in logistic regression. The independent variables are classified as categorical or numerical depending on the data. Variables Dependent

Production

Independent

Categorical

Numerical

Weed control Size of plantations Number of stumps per plot Planting density Altitude Latitude MTemp spring

MTemp spring: Mean temperature in spring; N: number of data in each category.

Values

N

0 if BA < 0.65 1 if 0.65 < BA < 1.8 2 if BA > 1.8 “0” if bad control “1” if normal control “2” if good control “0” if trial size 1” if demonstration size From 9 to 25 stumps

45 49 42 19 30 87 96 40

From 6666 to e33,333 cuttings ha
1 From 20 to 1100 m asl From 37.2 to 42.6 North From 14.7  C to 22.0  C

Author's personal copy 650

b i o m a s s a n d b i o e n e r g y 4 6 ( 2 0 1 2 ) 6 4 5 e6 5 2

Fig. 3 e Interactions in relation to growth in the first vegetative period at sites with significant differences. Density as the main independent variable and site as the conditional independent variable. Different letters stand for significant (p < 0.05) differences in basal area for each site.

Where good weed control was present, the plots showing the highest levels of production (basal area > 1.8 m2 ha
1) were those with the highest density (up to 33,333 cuttings ha
1) and to a lesser extent, those with a high mean spring temperature (up to 22  C). On the other hand, areas of medium production (basal area < 1.8 m2 ha
1) or low production (basal area > 0.65 m2 ha
1) (Fig. 6) were associated with higher altitudes (up to 1100 m) and latitudes (up to 42.6 North).

4.

Discussion

The profitability of SRF poplar plantations will be dependent on determining the optimum plant density since this will affect the cost of plantation (purchase of cuttings and machinery costs), which can be as much as 40% of the total cost of production [23]. Reducing the spacing between plants will negatively affect the size of individual trees but can increase biomass production per unit area, at least over short times scales, due to more rapid canopy closure and full site occupancy.

The existing literature regarding the effect of density on production is inconclusive. Larocque [24] and Proe et al. [18] reported 15% and 35% reduction in biomass for hybrid poplars when spacing was increased from 1.0 to 1.5 m over a four-year rotation respectively. De Bell et al. [25] reported that the increased production per unit area associated with closer spacing at planting was no longer observed in short rotation poplar growing in the USA after a period of 7 years. Other studies also report complex relationships between planting density and rotation length with the early advantage of closer spacing being lost as a result of increasing competition as rotation lengths increase [3]. Other studies have pointed to increases in production which are directly dependent on plantation density [12,18,26]. In our study, after the first vegetative period, the production measured via basal area was found to increase (m2 ha
1) as the number of plants per hectare increased. At the end of the rotation (3 yr old plants), the differences in both biomass and basal area associated with the DENSITY factor disappeared, while differences related to SITE factors were still present. No interaction was detected between the two factors. This would imply that the location of the plantation is more important in terms of productivity than the number of plants per hectare. Weed control in the establishment of plantations of the genus Populus from cuttings is an indispensable practice for appropriate cultivation [27,28]. In this case, of all the production variables, the categorical variable weed control (poor, medium and good) defined for all the study sites is that which explains the most variance. This is because the weeds compete with the poplars for resources, negatively affecting production, particularly at first ages [29,30]. Weed control must be particularly intensive during the first growing period in SRF plantations and it is essential to begin treatment at the soil preparation stage through the use of appropriate herbicides or tillage techniques [31], especially where perennial species are present since these may be difficult to control later on. Inadequate control of existing vegetation during the soil preparation work will result in a deficient plantation and cause high levels of tree mortality along with a reduction in productivity of up to 50% [32]. The necessity to apply treatments during the second or third vegetative periods is more questionable since although improvements in growth have been detected in certain combinations of site, planting densities and cultivation

Fig. 4 e Basal area (left) and total biomass (right) variability according to different densities in plots after three vegetative periods.

Author's personal copy b i o m a s s a n d b i o e n e r g y 4 6 ( 2 0 1 2 ) 6 4 5 e6 5 2

Fig. 5 e Interactions in relation to growth after three vegetative periods at sites with significant differences. Different letters stand for significant (p < 0.05) differences in mean basal area between sites.

practices [32], in others cases, weed control has been found unnecessary [27]. The early closure of tree crowns efficiently controls the proliferation of weeds. A lower dose of herbicide than that required for annual agricultural crops is normally sufficient [33]. A weed control protocol was clearly defined for all the trial sites. However, the efficiency of the herbicide treatment varied considerably from one site to another due to factors such as the size of the seed bank, the meteorological conditions at the time of application, the skill of the workers or specific site conditions which affect the efficiency of the herbicide. This must be taken into account when estimating real scale production since cultural treatment may be more heterogeneous over large areas. Other variables such as mean spring temperature indicate that at locations with a longer vegetative period, greater production is obtained. In this regard, altitude and latitude have been included with negative coefficients; the plantations at higher altitudes or latitudes will have lower productivity, since SRF plantations were watered and the precipitations were not a limiting factor. Therefore, the production potential

651

of Populus through SRF is high in Southern Europe, and the level of production can be improved through using genetic material which is better adapted to the specific growing conditions [34e36]. Although various different experimental projects were included in this study, these were either set up to investigate the effect of density (experimental plots) or other issues affecting production. Hence, the size and design of plantations, number of experimental units and stools measured in each, differed from one site to another. Trnka et al. [37] found that the highest poplar biomass production was achieved in intensively managed, small experimental plots, while the lowest yield corresponded to marginal areas with low management intensity and no weed control. In our study, no significant response in terms of production was detected in relation to the effect of design and size of plantation (demonstrative or experimental) or the number of stools measured in each experimental unit. More trials are necessary to establish the effect of density on production, following the development of plants over successive rotations. Furthermore, different clone material must be considered in order to give greater consistency to our results.

5.

Conclusions

The effect of density on the growth and production of Populus genus clones in SRF plantations was found to be significant in the first establishment year, when higher density results in greater growth. This suggests that in a hypothetical annual rotation, without taking into consideration the effects of management intensity on the vigour of regrowth, higher density would be the most advantageous during the first year. After three years, competition between plants increased and as a consequence the effect of density on growth and production was not significant, although there was a high level of variability from one site to another. Among the variables that explain production variability, weed control was the most influential factor. While density had a positive influence on production, latitude and longitude had a negative effect. Sites at more southerly locations were the most productive since the plants benefited from warmer climates and longer growth periods as long as they were kept well-watered. Hence, to maximize biomass production, taking into account planting costs as well as the difficulties associated with management and mechanisation of plantations, our overall results suggest that a density of 15,000 cuttings ha
1 should not be exceeded.

Acknowledgements

Fig. 6 e Representation of basal area production sorted into three levels of productivity (0 if BA < 0.65; 1 if 0.65 < BA < 1.8 and 2 if BA > 1.8) according to the two first canonical variable axes. Low production ‘0’ is represented by diamonds, medium production ‘1’ by squares, high production ‘2’ by triangles.

This work has been supported by RTA-2008-00025, On-cultivos and Lignocrop projects. The authors wish to thank Junta Castilla y Leo´n, IRTA Mas Badı´a Foundation, Soria Activa Foundation, CIUDEN, Acciona Energı´a enterprise, Asaja Granada and Prynconval SL. for maintaining the trials. We also thank the research support staff, P. de la Iglesia, A. Bachiller and E. Viscasillas for their dedication and

Author's personal copy 652

b i o m a s s a n d b i o e n e r g y 4 6 ( 2 0 1 2 ) 6 4 5 e6 5 2

efficiency in the data collection process. We also wish to thank Adam Collins for checking the English version of this article. [20]

references

[21] [22]

[1] Faaij APC. Bio-energy in Europe: changing technology choices. Energ Policy 2006;34:322e42. [2] European Commission. The EU rural development Policy: facing the Challenges. European Communities; 2008. [3] Makeschin F. Short rotation forestry in central and northern Europe e introduction and conclusions. For Ecol Manag 1999; 121:1e7. [4] Weih M. Intensive short rotation forestry in boreal climates: present and future perspectives. Can J Forest Res 2004;34(7): 1369e78. [5] Paris P, Pisanelli A, Sabatti M, Mareschi L, Cannata F, Scarascia-Mugnozza G. Short rotation forestry in Italy. II Sino-Italian Workshop on Agroforestry, Beijing; 2005. p. 9. [6] Di Nasso NNO, Guidi W, Ragaglini G, Tozzini C, Bonari E. Biomass production and energy balance of a 12-year-old short-rotation coppice poplar stand under different cutting cycles. GCB Bioenerg 2010;2:89e97. [7] Sixto H, Herna´ndez MJ, Barrio M, Carrasco JE, Can˜ellas I. Plantaciones del ge´nero Populus para la produccio´n de biomasa con fines energe´ticos: revisio´n. Investigacio´n Agraria: Sistemas y Recursos Forestales 2007;16(3):277e94. [8] Implementation Plan 2015. Spanish Biomass Technology Platform, p.69. [9] Dickmann D. Silviculture and biology of short-rotation woody crops in temperate regions: then and now. Biomass Bioenerg 2006;30:696e705. [10] Bradshaw HD, Ceulemans R, Davis J, Steller RF. Emerging model system in plant biology: poplar (Populus) as a model forest tree. J Plant Growth Regul 2000;19:306e13. [11] Tuskan GA, DiFazio SP, Teichmann T. Poplar Genomics is Getting Popular: the Impact of the poplar genome project on tree research. Plant Biol 2004;6(1):2e4. [12] Willebrand E, Verwijst T. Population dynamics of willow coppice system and their implications for management of short rotation forests. For Chron 1993;69:699e704. [13] Armstrong A, Johns C, Tubby I. Effects of spacing and cutting cycle on the yield of poplar grown as an energy crop. Biomass Bioenerg 1999;17(4):305e14. [14] Klasnja B, Orlovic S, Galic Z, Drekic M, Vasic M, Pilipovic A. Poplar biomass of high density short rotation plantations as raw material for energy production. Wood Res 2008;53(2): 27e38. [15] Laureysens I, Pellis A, Willems J, Ceulemans R. Growth and production of a short rotation coppice culture of poplar. III. Second rotation results. Biomass Bioenerg 2005;29(1):10e21. [16] DeBell DS, Clendenen GW, Zasada JC. Growing Populus biomass comparison of wood grass wider-spaced short rotation systems. Biomass Bioenerg 1993;4(5):305e13. [17] Facciotto G, Schenone G. Il pioppo fonte di energia rinnovabile. Sherwood 1998;35:19e26. [18] Proe M, Griffiths JH, Craig J. Effects of spacing, species and coppicing on leaf area, light interception and photosynthesis in short rotation forestry. Biomass Bioenerg 2002;23(5):315e26. [19] Laureysens I, Deraedt W, Indeherberge T, Ceulemans R. Population dynamics in a 6-year old coppice culture of poplar. I. Clonal differences in stool mortality, shoot

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31] [32]

[33]

[34]

[35]

[36]

[37]

dynamics and shoot diameter distribution in relation to biomass production. Biomass Bioenerg 2003;24(2):81e95. Pendleton OJ, Von Tress M, Bremer R. Interpretation of the four types of analyses of variance tables in SAS. Comm Statist Theor Meth 1986;15(9):2785e808. SAS Institute Inc. Getting Started with the SAS 9.1 Open Metadata Architecture. Cary, NC: SAS Institute Inc.; 2003. Hosmer DW, Lemeshow S. Applied logistic regression. John Wiley & Son; 2000. Manzone M, Airoldi G, Balsari P. Energetic and economic evaluation of a poplar cultivation for the biomass production in Italy. Biomass Bioenerg 2009;33:1258e64. Larocque GR. Performance and morphological response of the hybrid poplar DN-74 (Populus deltoides x nigra) under different spacings on a 4-year rotation. Ann For Sci 1999;56: 275e87. De Bell DS, Clendenen GW, Harrington CA, Zasada JC. Tree growth and stand development in short-rotation Populus plantings: 7-year results for two clones at three spacings. Biomass Bioenerg 1996;11:253e69. Bullard MJ, Mustill SJ, McMillan SD, Nixon PMI, Carver P, Britt CP. Yield improvements through modification of planting density and harvest frequency in short rotation coppice Salix spp. - 1. Yield response in two morphologically diverse varieties. Biomass Bioenerg 2002;22(1):15e25. Sixto H, Grau JM, Garcı´a-Baudı´n JM. Assessment of the effect of broad-spectrum pre-emergence herbicides in poplar nurseries. Crop Prot 2001;20(2):121e6. Otto S, Loddo D, Zanin G. Weed-poplar competition dynamics and yield loss in Italian short-rotation forestry. Weed Res 2009;50:153e62. Willoughby I, Clay D. Herbicides for farm woodlands and short rotation coppice. Forestry Commission Field Book. London: H.M.S.O.; 1996. Kauter D, Lewandowski I, Claupein W. Quantity and quality of harvestable biomass from Populus short rotation coppice for solid fuel useda review of the physiological basis and management influences. Biomass Bioenerg 2003;24(6): 411e27. Tubby I, Armstrong A. The establishment and management of short rotation coppice. Forestry Comm 2002:1e12. Buhler D, Netzer DA, Riemenschneider E, Hartzler RG. Weed management in short rotation poplar and herbaceous perennial crops grown for biofuel production. Biomass Bioenergy 1998;14(4):385e94. Karacic A. Production and ecological aspects of short rotation poplars in Sweden PhD. Uppsala: Swedish University of Agricultural Sciences; 2005. p. 42. Cruset JE, Voltas J. Adaptacio´n de clones de chopo en el valle del Ebro: ana´lisis retrospectivo de la red experimental establecida en la de´cada de los 80 mediante modelos mixtos, vol I; 2009. Actas del 5 Congreso Forestal Espan˜ol. Sixto H, Salvı´a J, Barrio M, Ciria MP, Can˜ellas I. Genetic variation and genotype-environment interactions in short rotation Populus plantations in southern Europe. New For 2011;42:163e77. Paris P, Mareschi L, Sabatti M, Pisanelli A, Ecosse A, Nardin F, et al. Comparing hybrid Populus clones for SRF across northern Italy after two biennial rotations: survival, growth and yield. Biomass Bioenerg 2011;35:1524e32. Trnka M, Trnka M, Fialova´ J, Koutecky V, Fajman M, Zalud Z, et al. Biomass production and survival rates of selected poplar clones grown under a short-rotation system on arable land. Plant Soil Environ 2008;54(2):78e88.