Description of the study area

The study was conducted in the eighth region of Chile, in an experiment established in La Aguada (37° 11’23” S and 72° 26’ 04” W), Yumbel commune. The predominating soil is known as “arenal”; it has coarse texture, it is deep, undeveloped, alluvial and derived from volcanic sands of andesitic and basaltic origin. The prevailing climate of the region is sub-humid with temperatures between 28.6 °C in January and 4.4 °C in July. Mean annual rainfall is 1093 mm (^{Novoa and Villaseca, 1989}).

Experimental design

The experiment was set up in August 2012 using a complete randomized block design with three replications. Each block was a 110 m square (12 100 m²) consisting of four quadrants, each containing nine square plots, 18 m on the side (324 m²). Each plot, had a border buffer zone and one experimental unit with 30 useful trees. The species planted were E. globulus, E. denticulata and A. dealbata, at three densities (5000, 10 000 and 15 000 trees ha^-1). The three species were managed with three coppicing frequencies. The quadrant were harvested the first, second and third years and one quadrant was left without cutting (Figure 1). The variables were measured in the harvested quadrants in December 2012 and corresponded to the two-year coppicing frequency (two-year-old stumps).

Figure 1 Distribution and design of experimental plots. Each quadrant represents one treatment (coppicing frequency), each color represents a species and the letters (A, B and C) are planting densities. 

Evaluated variables

In December 2012, the dynamics of number of proventitious shoots was evaluated in the obtained stumps. The number of shoots was counted in the experimental unit in September of 2014 and 2015, and height and diameter at collar height (DCH) of each shoot were measured. For combined analysis of shoots, shoots of each type (proventitious and adventitious) were counted after 10, 21 and 31 months of growth on each stump. Shoot type was defined by their physiological characteristics and position. Shoots on the stump crown in a horizontal position formed from scar buds were identified as adventitious, whereas those better joined to the stump (mother plant) and in a vertical position originating from latent buds within live stump tissue were identified as proventitious (^{De Simón and Bocio, 1999}). In two measuring periods (2014 and 2015) basic density of each shoot or stem was evaluated in slices made at different heights along the trunk. To this end, three stumps were selected in the buffer area of each plot, and all of the shoots were cut from each, packaged and labeled to identify the block, species, density, stump and shoot type (adventitious and proventitious), then taken to the laboratory for analysis.

Basic density of the wood was determined in all of the shoots from each of the stumps. In each shoot, total height was measured with a 12 m telescometer (Messfixs®) and each section’s diameter was determined with a caliper (Mitutoyo®) at 10, 50, 130 cm (diameter at breast height (dbh) and then at each meter until reaching diameters of less than 10 mm. At each of these points along the trunk, a slice (2.5 cm thick) was obtained. These were ordered and labeled in plastic netting, and completely submerged in recipients with water so that the slices would exceed the point of saturation of its fibers.

When the slices were completely saturated (constant weight), the volume of each sample was determined by the water displacement method (1 cm³ water=1 g), following the Norm NCh176/2 (^{INN, 1986}). The displaced water was weighed on an electronic balance (Snowrex® precision 0.01 g) on which a recipient with water was placed. In this recipient the samples were completely submerged and did not touch the sides or the bottom. Weight of the displaced water corresponding to the volume of the wood sample was obtained considering that the density of water as one (Valencia and Vargas, 1997). In paper bags, the samples were then dehydrated at 105 °C in an oven (Riossa®) until constant weight. The dry material (anhydrous) was weighed on an electronic balance (precision 0.01 g) immediately after removing them from the oven before environmental moisture could be absorbed. Basic density of the wood corresponded to the quotient of dry weight (anhydrous) over saturated volume:

where Db: basic density (g cm^-3); Vh: saturated (cm³).

Data analysis

Data were analyzed with descriptive and inferential statistics based on the species, density and evaluation date. For basic density, an ANOVA was performed and means were compared with Tukey test (p≤0.05). The statistical model was:

i= 1, ...3; j = 1, ...2; k = 1, ...3; l = 1, ...2

where Y_ijkl: value of the dependent variable observed in i, in year j, in plantation density k and shoot type l; µ: general mean; SP_i: effect of the i ^th species; A_j: effect of the j^th year; D_k: effect of the k^th plantation density; T_rl: effect of the l^th shoot type; SP(A)_ij: effect of the interaction between the i^th species and j ^th year; SP(D)_ik: effect of the interaction between the i ^th species and the k^th density; SP(Tr)_il: effect of the interaction between the i^th species and the l ^th type of shoot; E_ijkl: residual random error.

Data were analyzed with SAS Version 9.2 (^{SAS Institute, 2008}), Infostat® (Universidad Nacional de Córdoba, Argentina) and Sigma Plot® 8.1 (Systat Software, Inc. San José California, USA).

Proventitious shoots

Acacia dealbata had the highest number of proventitious shoots at the density of 5000 trees ha^-1, compared with the other treatments (Figure 2). In 2014, nearly 85 % of the stumps had four to five shoots in the densities of 10 000 and 15 000 trees ha^-1. The highest number was seven shoots, on average measuring 2.46 m high and 21 mm in diameter at collar height (DCH) in the three densities. Over time, the shoots reached 3.2 m and a DAC of 27 mm. The number of shoots in the densities of 5000 and 10 000 trees ha^-1 was adjusted among plantation densities in 2015 when 91 % of the stumps had five shoots and reached a maximum of seven, which the other 9 % produced (Figure 2). With 15 000 trees ha^-1, 87 % of the stumps had only three proventitious shoots, while the rest had up to five (Figure 2). Over time, in this density there was a decrease in the number of proventitious shoots, possibly due to the effect of inter- and intra- stump competition for space and nutrients. ^{Valero y Picos (2009}) obtained similar results in plots invaded by A. dealbata in Galicia. They recorded average sprouting of 7 to 12 shoots per stump in plots evaluated on the first dates of growth.

Figure 2 Proportion of Acacia dealbata stumps that produced proventitious shoots in two periods of evaluation (2014 and 2015). 

Eucalyptus denticulata showed considerable capacity for sprouting in all of the plantation densities, in number of shoots, height and diameter. In 2014, almost 80 % of the stumps produced three proventitious shoots with up to seven shoots, whereas the others had eight (Figure 3) with an average height of 2.33 m and a DCH of 21.57 mm in the three densities. The number of proventitious shoots decreased in 2015; in the plantation densities 10 000 and 15 000 trees ha^-1, 94 % of the stumps had two shoots (Figure 3). In the case of the plantation density 5000 trees ha^-1, 91 % of the stumps had three shoots and the rest had up to four. Average height of the shoots reached 4.77 m, and average DCH was 42.91 mm.

Figure 3 Proportion of Eucalyptus denticulata stumps that grew proventitious shoots in two periods of evaluation (2014 and 2015). 

In 2014, all of the E. globulus stumps in the three plantation densities had four to five proventitious shoots (Figure 4). Although this species had a smaller number of shoots, it exhibited better characteristics (on that year it reached a height of 3.35 m and a DCH of 29 mm). In 2015, shoots on almost all of the stumps decreased to between two and three with average height of 7.2 m and DCH of 55 mm. Moreover, a high proportion of stumps produced one to two shoots in the density of 15 000 trees ha^-1 (Figure 4). This affects shoot production because of the competition for space and light. Villegas (2003)⁴ evaluated the effect of stump spacing on the number of shoots on E. globulus after 60 d of growth. He attributed the non-significant effect to early evaluation (two months) at the time when initial growth of the shoot depends on the mother plant (stump), which provides support and nutrients. This did not occur after 24 months of growth in our study in which shoots engaged in inter- and intra-stump competition, in which plantation density had a significant influence.

Figure 4 Proportion of Eucalyptus globulus stumps that produced proventitious shoots in two periods of evaluation (2014 and 2015). 

The three species varied in height and diameter as adjustment in number of shoots in function of the number of trees ha^-1 and time. This information can be used to adequate management plans, reduce costs or avoid replantation and use the natural adjustment of number of shoots in short cycles to multiply the amount of biomass per unit of area for dendroenergetic purposes, or to improve wood quality through pruning to produce lumber for milling or to obtain pulp.

Combined analysis of adventitious and proventitious shoots

Evaluation on different dates allowed visualization of the dynamics of the number of adventitious and proventitious shoots in each species, plantation density and growth period (Table 1). In A. dealbata with 10 months of growth, with 15 000 trees ha^-1, a high number of adventitious shoots was observed, eight on average and up to 37 shoots exceptionally. The high number of adventitious shoots in the density of 15 000 trees ha^-1 decreased continuously, and at 21 months there were five and at 31 months there were three. In the density of 10 000 trees ha^-1, the response was irregular and reached an average of two adventitious shoots with a maximum of seven per stump at the end of the evaluation. In the density of 5000 trees ha^-1, the number of adventitious shoots was lower with an average fluctuation of two and three on the three dates.

Table 1 Number of proventitious and adventitious shoots on stumps of dendroenergetic species evaluated on three dates and plantation densities. 

Mean value, ±=standard deviation and observed range (minimum-maximum).

The first evaluation (10 months) of this species, at the density of 5000 trees ha^-1, showed an average of four proventitious-type shoots and an interval of two to eight shoots. At 21 months, there was a reduction of this type of shoot, between one and four and averaging three; four remained in the final evaluation (31 months). At higher densities, the number of shoots and the variability decreased over time. After coppicing, at the lower density the stumps were exposed to direct solar radiation and to wind, accelerating the scarring process in the cut area, which in turn reduced the production of callous tissue where adventitious type buds originate. Thus, the formation of proventitious shoots was favored, and an average of four were produced in all of the plantation densities 31 months after coppicing.

Eucalyptus denticulata had a large number of adventitious shoots on the first evaluation date, with five in the density of 10 000 trees ha^-1 and seven in that of 15 000 trees ha^-1 (Table 1). Moreover, there was broad variation among stumps (up to 25 adventitious shoots per stump). This response decreased in the evaluations at 21 and 31 months in all of the densities. The pattern was also seen with proventitious shoots (two to five), the highest quantity in the density of 5000 trees ha^-1.

In the first evaluation of E. globulus, after 10 months of growth, the highest average number of adventitious shoots (eight) was observed in the density of 5000 trees ha^-1, with a maximum of 36 (Table 1). In the following evaluations (21 and 31 months after coppicing), average values decreased to three and two, respectively. Also on the first date (10 months) the number of proventitious shoots in this species varied from five to six in the three densities and decreased over time to two shoots per stump in the three densities (Table 1). Over time, the number of shoots, both adventitious and proventitious, decreased in all of the species and plantation densities mainly due to natural elimination of the weak shoots, whose growth is affected by competition for space and solar radiation (^{Sarmiento and Ávila, 2007}; ^{Goel and Singh, 2008}). A reduction in light intensity affects initiation of regrowth (^{Meier et al., 2012}).

The importance of stump diameter in the probability of regrowth of some tree species has been demonstrated, and trees with thicker trunks have less capacity for regrowth (^{Sands and Abrams, 2009}). In our study, the young 2-year-old E. denticulata stumps with diameters between 5 and 8 cm and 7-cm-diameter E. globulus and A. dealbata stumps exhibited good regrowth. ^{Mc Carthy et al. (2014}) report differences in the number of shoots between clones, coppicing age, and stump height.

Basic density

Basic wood density was significantly different among species and evaluation periods; but shoot types (adventitious and proventitious) were not different in the three studied species. In 2014, E. denticulata had the highest values and A. dealbata the lowest (Table 2). The wood density values of E. globulus coincided with those documented by ^{Igartúa and Monteoliva (2010}) in single stem adults (480 to 503 kg m^-3).

Table 2 Mean basic density values of shoots in three species with dendroenergetic potential established in three plantation densities. 

Different upper case letters signal significant differences among species in the same year; lower case letters indicate differences between years (p≤0.05).

In 2015, the basic density values decreased in the three species, although in E. denticulata it was higher. The values of A. dealbata (380 and 460 kg m^-3) (Table 2) were similar to those obtained (318 up to 544 kg m^-3) by ^{Pinilla and Hernández (2010}) in sites of the Biobío region. They also observed that the lower values (414 to 466 kg m^-3) corresponded to young trees and varied among experimental sites and trunk section sampled; the highest value of basic density was found at the base of the trunk (558 to 590 kg m^-3).

Variation in wood basic density within the tree has also been observed in eucalyptus. This is related with the formative ages of the stem and fiber morphology (^{Igartúa and Monteoliva, 2010}). In E. regnans, basic density decreased (404 kg m^-3) between the base and 4.5 % of the total tree height but later increased (462 kg m^-3) up to 70 % of the tree height (^{Omonte and Valenzuela, 2011}). The reduction in basic density between years in our study was mainly due to the effect of primary growth, which modified the distance from the base to the sampling sites at different trunk heights (^{Igartúa et al., 2002}).