- Email: [email protected]
Hybrid and clonal variability of nutrient content and nutrient use efficiency in Eucalyptus stands in Congo
Forest Ecology and Management 210 (2005) 193–204 www.elsevier.com/locate/foreco
Hybrid and clonal variability of nutrient content and nutrient use efficiency in Eucalyptus stands in Congo R. Safou-Matondo a,*, P. Deleporte b, J.P. Laclau c, J.P. Bouillet d a UR2PI, BP 1291, Pointe-Noire, Congo CIRAD-Foreˆt/UR2PI, Pointe-Noire, Congo c CIRAD-Foreˆt/ESALQ-USP, Cx Postal 9, 13418-900 Piracicaba, SP, Brazil d CIRAD-Foreˆt, Campus International de Baillarguet, 34398 Montpellier Cedex 5, France b
Received 11 February 2004; received in revised form 21 January 2005; accepted 13 February 2005
Abstract For 20 years, there has been 42,000 ha estate of clonal Eucalyptus plantations around Pointe-Noire in Congo on sandy soils that have very low reserves of available nutrients. These plantations have been based on a natural hybrid (E. PF1). This hybrid is being replaced by E. urophylla E. grandis (UG), a more productive hybrid developed by the breeding program of UR2PI. A study of biogeochemical cycles showed that nutrient removal by harvesting is the main nutrient output in the E. PF1 ecosystem. It is therefore important to quantify the nutrient content (NC) in both hybrids to compare corresponding nutrient removal values. The work dealt with four UG clones and the most planted clone of E. PF1. Twelve trees per clone were sampled at the logging age (8 years) in a clonal test for UG clones and in a nearby stand for E. PF1. Tables were established to predict, from girth at breast height (C1.30 m), the biomass and nutrient content of stemwood, bark, dead and living branches, leaves, and were applied to the inventory of the different stands to evaluate corresponding biomass, NC and nutrient use efficiency (NUE) on a per-hectare basis. Total biomass differed between the two hybrids and among UG clones: 109 t ha1 for E. PF1 and 108–155 t ha1 for UG clones. In E. PF1 trees, total NC was globally lower for N, K, and Mg, but greater for P and Ca. In stemwood, nitrogen content was similar for both hybrids. By contrast, in UG clones, NC was much lower for P (72%) and Ca (40% to 55%). The same trends were observed for NUE: equivalent for both hybrids for N, but higher in UG clones for P (+72%) and Ca (+43% to +59%). A marked variability among clones was observed for K and Mg. UG clones allocated proportionally more nutrients in leaves than E. PF1. These results show that clones should not be selected only on growth traits but also on NUE and on the concentration of nutrients in tree components removed by harvesting. It will be then possible to limit the cost of fertilising needed to maintain stand growth and soil fertility. # 2005 Elsevier B.V. All rights reserved. Keywords: Biomass; Nutrient content; Nutrient use efficiency; Hybrid variability; Clonal variability; Eucalyptus
* Corresponding author. Tel.: +242 94 31 84; fax: +242 94 47 95. E-mail addresses: [email protected] (R. Safou-Matondo), [email protected] (P. Deleporte). 0378-1127/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.foreco.2005.02.049
194
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
1. Introduction More than 40,000 ha of Eucalyptus have been planted on savannas around Pointe-Noire in the Republic of Congo. They are mainly based on two natural hybrids (E. PF1 and E. tereticornis E. grandis (UG)), which appeared naturally among the first afforestation. With fertilising and weeding, these hybrids perform well (15–20 m3 ha1 year1), but they are replaced increasingly by the artificial hybrid E. urophylla E. grandis, that it is more productive (up to 40 m3 ha1 year1 of over-bark volume in clonal tests) (Nouguier, 1997; Saya, 1999). The duration of the stand rotation is 7–8 years. The current method of harvesting de-barks stems up to 7 cm in diameter for pulpwood, all residues of over 2 cm diameter being commercialised as firewood (Laclau et al., 1996, 2000). The dynamics of biomass and nutrient accumulation have been studied in an E. PF1 clone 1-41 chronosequence (Laclau et al., 2000). This clone has been the most extensively planted (7000 ha) because of its high growth and is used as a reference material in breeding and silvicultural trials. Biomass and nutrient content models have been established for each tree component (stemwood, bark, living and dead branches, leaves) for the major nutrients (N, P, K, Ca, Mg). These models can quantify the nutrient requirement of trees within the stand rotation and their nutrient content at harvesting. Nutrient content at harvesting is particularly important as a biogeochemical study showed that nutrient removal at that time is the main nutrient output in the E. PF1 ecosystem (Laclau et al., 2002). The nutrient content of the slash remaining after harvesting can also be estimated. This information is essential to identify methods for sustainable management of replanted site as soil fertility is closely linked to organic matter content (Bouillet et al., 2000; Nzila et al., 2002). In particular, the N nutrient budget— differences between N inputs and N outputs—is very unbalanced, and the need of N fertiliser increases markedly over successive rotations. Bouillet et al. (2001) estimated this deficit to 165 kg N ha1 at the end of planted crop rotation, 375 kg N ha1 after the first coppice rotation and 550 kg N ha1 at the end of the second coppice rotation. A study on the effects of site management on nutrient cycling and tree growth in Congo showed that
the nutrient content in the aboveground biomass of replanted Eucalypt stands was strongly dependent on slash and litter management practices (Nzila et al., 2002). A marked increase in nutrient concentration in foliage was observed as the nutrient content of organic residues on the soil surface increased. As UG hybrid is more productive than E. PF1, its nutrient content is likely to be higher than in E. PF1. However, the nutrient use efficiency (NUE) of mineral elements in biomass production could differ sharply between the two hybrids and/or among UG clones. The NC removed by harvesting could be then not proportional to the biomass produced. Different indexes have been proposed to represent NUE (Vitousek, 1982). One of the most common indexes represents the mean quantity of aboveground biomass (or of stemwood biomass) produced per unit of nutrient incorporated into the aerial components of the stands since planting (Morais et al., 1990; Herbert, 1996). This index may be then used to select highly productive genotypes with low nutrient accumulation that are better for adapted sustainable production. The current study compared the nutrient contents in stemwood, bark, dead and living branches, and leaves of four UG clones and the clone 1-41 of E. PF1. Its objectives were to establish, at the harvesting age (8 years), biomass and nutrient content for the different tree components, and to calculate the corresponding NUE. The variability between hybrids and clones was investigated.
2. Materials and methods 2.1. Site parameters The plantations are established on coastal plains around Pointe-Noire, Congo at latitude 48S and longitude 128E. The climate is sub-equatorial with a high atmospheric humidity (85% on average), with a rainy season from October to May and a dry season from June to September. Mean annual rainfall is around 1200 mm, and mean annual temperature is 25 8C with seasonal variations of about 5 8C. The soils are Ferralic Arenosols (FAO classification) characterised by a homogeneous sandy texture, acidic reaction, limited available nutrients, and very low
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
195
2000) and the same silviculture was applied: savannah ploughed before planting; spacing of 5 m 3 m; starter fertilisation of 150 g per plant of NPK (13% N, 6% P, 17% K); manual weeding in the planting row and mechanical weeding in the inter-row.
levels of exchangeable cations, organic matter and cation exchange capacity (Table 1) (Laclau, 2001). 2.2. Stand characteristics This study was carried out in two stands:
2.3. Methodology in evaluating biomass and nutrient content
A clonal test using a randomised complete block design with five replicates. Each unit plot consisted of 5 5 trees, at a stand density of 666 trees ha1 (Nouguier, 1997). Twenty-three UG clones were compared, among which four UG clones were sampled. Two clones (18-50 and 18-52) were full sibs, and the two other clones (18-64 and 18-65) were not related. A trial planted with the E. PF1 clone 1-41 located 1 km far from the clonal test, comparing various types of cuttings, with five replicates. No significant differences among treatments were observed.
The main stages adopted classically for this type of study have been followed (Laclau et al., 2000; Grace and Madgwick, 1987; Ranger et al., 1995): Inventory of the stands (125 trees per UG clone, and 600 trees for E. PF1 clone 1-41). Twelve trees per clone (four trees in three replicates distributed throughout six basal area classes defined from the inventory) were cut down, and the major components were isolated: stemwood (up to 7 cm and between 7 and 2 cm), stembark (up to 7 cm and between 7 and 2 cm), leaves, living branches (up to 2 cm and over 2 cm) and dead branches. The diameters of the 7 and 2 cm corresponded to the commercial limits for pulpwood and firewood,
The UG and the E. PF1 clones were not then planted in the same trial. However, it was decided to compare the results as the stands were established at the same date and harvested at the same age (8 years old), the soils properties were similar (Laclau et al.,
Table 1 Main soil characteristics in the Pointe-Noire region (from Laclau et al., 2000) Soil layer
Total elementsa
Particle size distribution
Clay (%) Silt (%) Sand (%) Orgamic material (%) N total (%) C/N Ca (%) Mg (%) K (%) P (%) Al (%) A11 (0–5 cm) A12 (5–70 cm) B1 (70–120 cm) B21 (120–200 cm) B22 (200–400 cm) B23 (400–600 cm)
4.5 4.0 1.7 4.8 6.9 5.8
2.4 3.4 7.3 6.7 5.9 8.3
Soil layer
pH (H2O)
A11 (0–5 cm) A12 (5–70 cm) B1 (70–120 cm) B21 (120–200 cm) B22 (200–400 cm) B23 (400–600 cm)
5.2 4.7 5.3 5.5 5.4 5.4
a b c d e
93.1 92.6 91.1 88.5 87.2 85.9
0.91 0.57 0.26 0.31 0.16 0.12
0.37 0.23 0.11 0.14 0.09 0.09
14.3 14.3 13.6 12.9 10.0 7.8
Exchangeable cationsb (cmolc kg1) K
Ca
Mg
Mn
Na
Al
0.03 0.05 0.01 0.01 0.02 0.02
0.19 0.10 0.09 0.08 0.08 0.08
0.07 0.04 0.02 0.02 0.02 0.02
0.01 0.01 0.01 0.01 0.01 0.01
0.02 0.04 0.04 0.01 0.02 0.02
0.10 0.17 0.08 0.13 0.10 0.07
Acid digestion and ICP determination of elements. Cobalt-hexamine extraction, ICP determination of cations. BC: exchangeable cations’ sum. CEC: cation exchange capacity. Available P (Duchauffour and Bonneau (1959) methodology).
1.41 1.91 0.85 1.09 1.11 1.28
0.23 0.26 0.22 0.23 0.26 0.25
0.15 0.16 0.16 0.19 0.22 0.20
0.05 0.04 0.05 0.05 0.04 0.04
0.63 0.64 0.92 0.97 1.14 1.08
BCc
CECd
BC/CEC (%)
Pavailablee (%)
0.31 0.23 0.16 0.12 0.14 0.14
0.43 0.43 0.26 0.32 0.28 0.30
72 53 61 37 50 47
0.03 0.02 0.02 0.02 0.03 0.02
196
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
respectively. The following variables were measured: Total length of the trees and stem circumference at the base, 0.30 m, 1.30 m (C1.30 m), and every 3 m up to a diameter of 2 cm. Fresh weight of the stem up to diameters of 7 cm and between 7 and 2 cm. Disks of wood and bark of constant thickness were taken from the centre of the stem pieces, every 3 m. These disks were dried until constant weight and the dry biomass of the components in each tree was calculated proportionally. All the branches and leaves were collected and samples of these components were dried. The branches with a diameter exceeding 2 cm were distinguished from the smaller ones, in order to quantify the biomass and nutrient content of firewood. 2.4. Chemical sample analyses Before the chemical analyses, every sample was dried at 65 8C, ground and homogenised. Total N was analysed by acid–base volumetry after Kjeldahl mineralisation. P was determined by cold-colorimetry from the Murphy and Riley reagents. Ca, K and Mg were analysed by atomic absorption spectrophotometry. 2.5. Statistical data processing To evaluate biomass and nutrient contents of each component of each clone, tables were established following the general model: Yi j ¼ a þ a j þ ðb Ci j Þ þ ðb j Ci j Þ þ ðc Ci2j Þ þ ðc j Ci2j Þ þ ðd Ci3j Þ þ ðd j Ci3j Þ þ ei j where Yij was the biomass or the nutrient content of a given component of tree i of the clone j; Cij was the circumference at breast height (C1.30 m) of tree i of the clone j; a, b, c and d were the global parameters of the model; a j, b j, c j and d j were the local parameters of the model, depending on clone j; eij was the residual error. The statistics were operated using GLM procedure of SYSTAT 9.0 software (SPSS Inc#). The backward stepping option was used to introduce only significant variables (at 0.05 threshold)
into the model. A tolerance threshold of 0.001 was chosen. Criteria to compare individual and global models were a maximum adjusted R2 value, a minimum root mean square error (RMSE) and no bias in the distribution of the residues. At first, non-weighted regressions were calculated for biomass and nutrient content for each component and each clone (8 compounds per tree 12 trees 5 clones) corresponding to 480 samples and 288 equations: 8 components 5 clones (biomass + 5 nutrients) and 48 equations for the global models. Clonal and global models were compared using F-test: Fobs ¼
ðSSE2 SSE1 Þ=ð p1 p2 Þ SSE1 =ðn p1 Þ
where p1 is the number of parameters with the clonal models, p2 is the number of parameters with the global model, SSE1 = sum of square error of the clonal model, SSE2 = sum of square error of the global model, N = number of observations, F obs was compared the value given in the Fisher’s table F( p1 p2, n p1). If F obs > F( p1 p2, n p1), then clonal model gave more precisely the biomass or the nutrient content better than the global model. These equations were then applied to the inventory of each plot of the stands to estimate, for each component, the biomass and nutrient content on a perhectare basis. The average of the five replicates gave an estimation of biomass and nutrient content per clone. ANOVA were performed to test differences between hybrids and clones, according to Bonferonni test ( p < 0.05). UG clones were compared through a twoway ANOVA, with five replicates and four clones, using Bonferonni test ( p < 0.05). E. PF1 clone 1-41 was compared with the most productive UG clone (18-52) through a one-way ANOVA with five replicates, at 0.05 threshold.
3. Results 3.1. Growth and stand production Differences in tree growth and stand production were observed among clones (Table 2 and Fig. 1). Even though clones 18-52 and 18-50 being full sibs, 18-52 exhibited a MAI 18% higher than the clone 18-50.
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
197
Table 2 Mean height, mean circumference, total volume of the stand and MAI of the Eucalyptus clones (results at 99 months) SEUGa
Clones
H (m) C1.30 m (cm) V (m3 ha1) MAI (m3 ha1 year1)
18-50
18-52
18-64
18-65
1-41
27.8 a 62.6 a 235.2 b 28.5
27.9 aa 66.4 aa 275.0 aa 33.3
25.9 b 53.1 b 176.0 d 21.3
26.6 ab 63.0 a 215.4 c 26.1
27.0 a 57.5 b 207.0 a 25.1
0.3 1.0 4.4
Latin letters indicate significant differences (threshold: 0.05) between clones UG using Bonferonni test. Greek letters indicate significant differences (threshold: 0.05) between the most productive clone UG 18-52 and the clone 1-41 using Bonferonni test. a SEUG: standard error for the comparison UG clones.
UG clones presented better MAI than E. PF1, except for the UG clone 18-64. This clone had therefore not been used in commercial plantations. 3.2. Biomass and nutrient content tables A comparison of the global model for all clones and the individual model of each clone showed that the global model was as precise as the individual model in 45 out of the 48 cases. The local parameters of the general model were different from zero in 46 out of the 48 cases: 37 equations (respectively, 9) exhibited one local parameter (respectively, 2) different from zero. The Figs. 2–4 give some examples of the differences observed among clones for some tree components, even between the full sib clones 18-50 and 18-52. The shape of the curves may vary markedly according to components, nutrients, and clones. R2 coefficients were generally higher than 0.85 (data not shown) and higher than 0.90 for stemwood,
except for Mg content (R2 = 0.79). The weakest R2 coefficients (0.53–0.75) were observed for wood and bark between diameters 7 and 2 cm, dead branches and living branches. The prediction of biomass and nutrient contents was then accurate for stemwood and bark over 7 cm and leaves but weaker for living branches and dead branches. 3.3. Biomass and nutrient content per hectare Large differences in total biomass were observed between the two hybrids and among UG clones, including the full sibs 18-50 and 18-52 (Table 3). The best clone (18-52) produced 20%, 27%, 42% and 44% more than clones 18-50, 18-65, 1-41 and 18-64, respectively. The biomass of stemwood ranged from 91 to 132 Mg ha1 in the UG clones, and 95 Mg ha1 for the clone 1-41. Stemwood represented the main part of the aerial biomass (Table 3). The proportion of the total biomass
Fig. 1. Changes in the individual basal area of the Eucalyptus clones according to stand age.
198
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
Fig. 2. Curves and parameters related to the model fitted for Ca content of leaves (the other parameters of the model are null).
Fig. 3. Curves and parameters related to the model fitted for P content of bark up to a diameter of 7 cm (the other parameters of the model are null).
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
199
Fig. 4. Curves and parameters related to the model fitted for N content of wood up to diameter 7 cm (other parameters of the model are null).
in this component was similar between hybrids: 86.5% for clone 1-41 and a mean of 84.6% for UG clones. In contrast, marked differences between hybrids were observed for bark and leaves: The percentage of bark was 28% higher in the clone 1-41 than in UG clones.
The percentage of leaves was 27% lower in the clone 1-41 than in UG clones. The clone 1-41 exhibited the lowest percentage of living branches. This component represented between 6.9% and 7.9% of the aerial biomass in clones 18-52 and 18-50, respectively.
Table 3 Total aerial dry biomass, corresponding MAI and proportion of the total biomass in the tree components of the Eucalyptus clones (results at 99 months) SEUGa
Clones Total biomass (t ha1) MAI (t ha1 year1) Proportion of total biomass (%) Stemwood (D > 2 cm) Bark (D > 2 cm) Leaves Living branches Dead branches
18-50
18-52
18-64
18-65
1-41
129.5 b 15.7
155.2 aa 18.8
107.6 c 13.0
121.9 b 14.8
109.4 b 13.3
83.8 3.8 2.6 7.9 1.7
83.6 4.5 2.7 7.8 1.4
86.5 5.0 1.9 4.9 1.7
86.1 3.5 2.4 6.9 1.2
85.0 3.8 2.7 7.9 0.6
2.5
Latin letters indicate significant differences (threshold: 0.05) between clones UG using Bonferonni test. Greek letters indicate significant differences (threshold: 0.05) between the most productive clone UG 18-52 and the clone 1-41 using Bonferonni test. a SEUG: standard error for the comparison UG clones.
200
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
Table 4 Biomass (t ha1) and nutrient content (kg ha1) of the tree components of the Eucalyptus clones (results at 99 months) Clone
Components Stemwood (D > 2 cm)
Bark (D > 2 cm)
Leaves
Living branches
Dead branches
Total
18-50
Biomass N P K Ca Mg
111.2 217.1 8.4 18.7 17.2 19.4
b ab ab c b b
4.5 19.9 14.6 12.1 30.7 16.8
b a a bc a a
3.1 60.6 5.4 17.3 12.7 12.3
bc b c c b b
9.0 35.6 8.8 15.1 13.7 9.8
b a a b b a
1.6 5.3 0.6 0.2 1.7 0.5
a a a c b c
129.5 338.5 37.7 63.4 76.2 58.8
b a b c b b
18-52
Biomass N P K Ca Mg
131.9 228.4 9.1 28.5 22.5 23.4
aa aa ab bb ab aa
5.9 17.8 15.8 12.7 31.3 16.5
aa bb aa ba aa ab
4.2 77.1 9.4 26.7 17.4 15.6
aa aa aa aa aa aa
12.2 29.2 8.4 17.7 15.3 9.7
aa ba aa aa aa aa
0.9 2.1 0.3 0.1 0.7 0.3
cb db ca db cb db
155.2 354.4 42.9 85.8 87.2 65.5
aa aa aa ba aa aa
18-64
Biomass N P K Ca Mg
90.8 151.1 7.9 27.0 20.9 9.4
c c ab b a d
4.1 14.6 7.6 10.9 19.5 10.8
b c b c b b
2.8 47.8 4.2 17.9 11.3 8.9
c c d c c d
8.2 21.3 7.6 15.2 12.1 7.3
b c a b c b
1.8 3.0 0.5 0.4 1.6 0.8
a c b a b a
107.6 237.8 27.9 71.2 65.4 37.2
c b c c c d
18-65
Biomass N P K Ca Mg
101.9 202.6 7.0 44.4 16.7 16.0
b b b a b c
5.5 19.1 6.9 17.7 33.4 11.9
a ab b a a b
3.3 68.0 7.7 23.3 11.5 10.9
b b b b bc c
9.5 32.4 7.6 19.7 12.4 9.8
b ab a a c a
1.7 4.2 0.6 0.4 2.0 0.7
a b a b a b
121.9 326.2 29.8 105.4 76.1 49.3
b a c a b c
1-41
Biomass N P K Ca Mg
94.6 190.5 28.5 39.4 37.3 16.7
b a a a a b
5.5 25.2 14.9 16.2 44.9 24.1
a a a a a a
2.1 42.7 3.3 9.2 7.1 6.5
b b b b b b
5.4 11.8 3.1 5.4 5.7 5.0
b b b b b b
1.9 6.3 0.5 0.2 2.8 0.8
a a a a a a
109.4 276.6 50.3 70.4 97.8 53.1
b a a a a a
SEUGa
Biomass N P K Ca Mg
2.1 4.4 0.4 0.9 0.4 0.4
0.1 0.4 0.3 0.4 0.8 0.3
0.1 1.8 0.2 0.5 0.3 0.3
0.4 0.8 0.3 0.5 0.3 0.2
0.1 0.1 0.02 0.01 0.05 0.02
2.5 7.2 0.7 2.3 1.7 1.1
Latin letters indicate significant differences (threshold: 0.05) between clones UG using Bonferonni test. Greek letters indicate significant differences (threshold: 0.05) between the most productive clone UG 18-52 and the clone 1-41 using Bonferonni test. a SEUG: standard error for the comparison UG clones.
The total accumulation of P and Ca was generally lower in UG clones than in the clone 1-41: on average 31% and 22%, respectively (Table 4). The opposite pattern was found for N (+14%) and K (+16%). No trend was found for Mg. The nutrient contents were always lower in the stemwood of UG clones than in clone 1-41 for P
(71%), Ca (48%) and K (25%). UG also exhibited systematically lower content of all nutrients in the bark, but a much higher accumulation of all nutrients in the living branches and the leaves. N accumulation increased with clone production, but not proportionally: (i) the total biomass of clone 18-65 was 13% greater than of the clone 18-64 but the
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204 Table 5 Biomass (t ha1) and nutrient content (kg ha1) removed by harvesting Clone
Biomass
N
P
K
Ca
18-50 18-52 18-64 18-65 1-41
116.2 136.1 93.9 105.7 98.1
231.7 234.5 156.2 213.6 195.8
13.1 11.6 12.2 11.2 31.5
22.6 32.5 30.4 48.7 41.6
24.3 28.3 25.3 22.9 41.1
24.3 28.2 12.1 19.8 20.1
UGa UG/1-41 (%)
113.0 115
209.0 107
12.0 38
33.5 81
25.2 61
21.1 105
a
Mg
UG: mean of the four E. urophylla E. grandis clones.
Table 6 Biomass (t ha1) and nutrient content (kg ha1) of slash after harvest Clone 18-50 18-52 18-64 18-65 1-41 UGa UG/1-41 (%) a
Biomass
Ca
Mg
13.3 19.1 13.7 16.1 11.3
N 106.8 119.9 81.6 112.6 80.8
P 24.6 31.3 15.7 18.6 18.8
K 40.8 53.3 40.8 56.7 28.8
51.9 58.9 40.1 53.2 56.7
34.5 37.3 25.1 29.5 33.0
15.6 138
105.2 130
22.5 120
47.9 1646
51.0 90
31.6 96
UG: mean of the four E. urophylla E. grandis clones.
corresponding increase in N content was 37%; and (ii) clone 18-52 was 20% more productive than the clone 18-50 but accumulated only 5% more nitrogen.
201
to the soil (leaves, branches and bark up to stem diameter of 7 cm) (Table 6). Nitrogen removal increased with biomass production, and corresponded on average to 66% of the total aerial N content. Nutrient removal was much lower on average in UG clones than clone 1-41 for P (62%), Ca (29%) and K (19%). Differences were observed among UG clones. For example, the nutrient content removed with clone 18-65 was the lowest for P but the greatest for K (Table 5). The contents of Ca and Mg in slash were quite similar for both hybrids (Table 6). In contrast, they were higher for K (+66%), N (+30%) and P (+20%) in UG clones than clone 1-41. Nevertheless, a noticeable variability was observed among UG clones. In particular, the clone 18-52 was characterised by the highest amounts of nutrients in slash. UG clones exhibited higher biomass production, except for the clone 18-64, and higher N removal by harvesting than the clone 1-41, but also greater N returns to the soil. The difference between nitrogen returning to the soil and nitrogen removed was then better for UG clones: 104 kg N ha1 on average for UG clones compared to 115 kg N ha1 for the clone 1-41 (Tables 5 and 6). The same conclusions could be drawn up for P (+10 kg ha1 versus 13 kg ha1), K (+14 kg ha1 versus 13 kg ha1) and Ca (+26 kg ha1 versus +10 kg ha1). No marked differences were pointed out for Mg (+11 kg ha1 versus +13 kg ha1).
3.4. Nutrient removals and cycling 3.5. Nutrient use efficiency We calculated the biomass and nutrient content removed by harvesting (stemwood and bark between stem diameters of 7 and 2 cm) (Table 5), and returning
The results found were consistent for the two formula of NUE: aerial biomass/aerial nutrient
Table 7 Values of NUE (in kg of dry biomass per kg of nutrient accumulated in the aerial biomass) according to the clone and the nutrient taken into account Clone
Total aerial biomass (t ha1)
NUE
18-50 18-52 18-64 18-65 1-41
129.5 155.2 107.6 121.9 109.4
382 438 452 374 395
SEUGa
2.5
N b aa c b b
P
3
c ba a c b
K
3430 3610 3860 4090 2130 60
c bca ab a b
Ca
2040 1810 1510 1160 1540 10
a ba c d b
1700 1780 1650 1600 1120 10
Mg b aa c c b
2200 2370 2890 2470 2030
d ca a b b
20
Latin letters indicate significant differences (threshold: 0.05) between clones UG using Bonferonni test. Greek letters indicate significant differences (threshold: 0.05) between the most productive clone UG 18-52 and the clone 1-41 using Bonferonni test. a SEUG: standard error for the comparison UG clones.
202
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
content; or stemwood biomass/aerial nutrient content. The UG clones exhibited much higher NUE than clone 1-41 for P and Ca (Table 7). Their NUE was also higher for Mg. In contrast, no real trends were observed for N and K. A strong variability in NUE was observed among UG clones, but the most efficient clone depended on the nutrient taken into account.
4. Discussion A general model, as precise as clonal models, was adjusted to predict stand biomass and nutrient contents. This general model is easier to use than individual models (in this study, the number of equations decrease from 240 to 48), but allows fairly good prediction by taking into account the clonal variability in within tree biomass allocation and in nutrient accumulation. The best prediction was observed for the components representing the main parts of the biomass and nutrient contents of the stand. A marked inter-clonal variability was observed in nutrient content. P and Ca contents aboveground in UG clones were lower than in the clone 1-41, even for clones 18-50 and 18-52, whose wood production was much higher. Discrepancies in nutrient content between hybrids were large in the bark but rather low in the wood. The amounts of nutrients in the canopy were higher in UG clones than in the clone 141. Considering the optimal foliar concentrations indicated by Herbert (1996) for E. grandis, the concentrations in N, K and especially Ca were weak. Foliar ratios N/P, N/K, P/K and Ca/Mg were far from the optimum for the UG clones (Table 8). This finding suggests a low soil availability in N and Ca and is
consistent with studies carried out in Congo showing: (i) a clear positive response of Eucalypt stand to N fertilisation (Bouillet et al., 2001); and (ii) that tree growth may be limited by a shortage of Ca (Nzila et al., 2002). By contrast, the very weak concentrations of K may have no consequential effects on the growth, as Na is likely to substitute partially to K (Marschner, 1995; Laclau, 2001) due to the proximity of the sea. The reciprocal recurrent selection scheme carried out in Congo (Vigneron et al., 2000) led to the development of UG clones with an architecture different from that of the natural hybrid E. PF1 (larger crowns, etc.). The UG clones exhibited higher use efficiency of P and Ca than the 1-41 clone. The use efficiency of K, Ca and Mg measured for UG clones were moreover higher than values reported for various Eucalyptus species in Brazil, India and South Africa (Laclau et al., 2000) (Table 9). By contrast, the use efficiency of P remained poor for UG clones compared to values reported in the literature for other species (Morais et al., 1990; Herbert, 1996; Negi and Sharma, 1996). The NUE value is very dependent on the availability of the element in the soil (Morais et al., 1990). The low use efficiency of P observed might be a result of the relatively high availability of P in the soil of the Pointe-Noire region, shown by the lack of response of trees to the input of P (Safou-Matondo and Bouillet, 1999). The high NUE of the trees sampled show that clones selected by the breeding program are particularly efficient to produce high amounts of wood despite the very low nutrient availability in the soil of the Pointe-Noire region. Thus, NUE might be a relevant index to integrate in the selection strategies
Table 8 Mean foliar concentrations of major nutrients and corresponding ratios for the Eucalyptus clones (results at 99 months) Clone
N (%)
P (%)
K (%)
18-50 18-52 18-64 18-65 1-41
1.96 1.82 1.71 2.05 2.05
0.17 0.22 0.15 0.23 0.16
0.56 0.63 0.64 0.70 0.44
E. grandis (Herbert, Minimum Maximum Optimum
1996) 1.25 3.35 2.8
0.10 0.35 0.15
0.36 1.19 0.75
Ca (%)
Mg (%)
N/P
N/K
P/K
Ca/Mg
0.41 0.41 0.40 0.35 0.34
0.40 0.37 0.32 0.33 0.31
11.3 8.2 11.3 8.9 12.9
3.5 2.9 2.7 2.9 4.7
0.31 0.35 0.24 0.33 0.36
1.04 1.12 1.27 1.06 1.10
0.56 1.82 >1.0
0.21 0.62 0.35
3.0 28.6 18
1.0 5.4 3.5
0.09 0.62 0.2
1.24 7.28 >3.3
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
203
Table 9 Nutrient use efficiency in the 7-year-old stand for adult Eucalyptus plantations from the literature (data expressed in kg of dry biomass per kg of nutrient accumulated in the aerial biomass) N
P
Congo E. PF1 hybrid, clone 1-41, 7 years (Laclau et al., 2000) E. PF1 hybrid, clone 1-38, 7 years (Loumeto, 1986)
592 611
2000 1638
Brazil E. grandis Hill, 8 years (Morais et al., 1990) E. saligna Smith, 8 years (Morais et al., 1990) E. cloeziana F. Muell., 8 years (Morais et al., 1990) E. citriodora Hook., 8 years (Morais et al., 1990) E. brassiana S.T. Blake, 8 years (Morais et al., 1990)
282–443 246–315 307–352 241–318 227–340
4269–8783 3997–6327 5997–11351 4064–7357 4419–8727
India E. PF1 hybrid, 7–10 years (Negi and Sharma, 1996) E. globulus Labill., 9 years (Negi and Sharma, 1996) E. grandis Hill, 9 years (Negi and Sharma, 1996)
318–352 155 204
1370–4333 2841 2000
South Africa E. fastigata Deane & Maiden, 7 years (Herbert, 1996) E. grandis Hill, 7 years (Herbert, 1996) E. macarthurii Deane & Maiden, 7 years (Herbert, 1996) E. nitens Deane & Maiden, 7 years (Herbert, 1996) E. smithii R. Baker, 7 years (Herbert, 1996)
487 538 438 510 822
since this parameter provides an indication of the amounts of nutrients exported at the harvest. The inputs of fertiliser needed to ensure a sustainable yield of commercial plantation will be therefore dependent on the NUE of the clones planted. However, correlation between stemwood production and values of NUE was low for the five clones sampled, and the most productive clone 18-52 did not have the best NUE for any component, except for Ca. This pattern indicates that maximizing yields and economic returns needs to take into account the efficiency of other factors involved in growth of the trees, as light or water (Dye, 2000, 2001; Gonzales et al., 2004).
5. Conclusion Clones of the natural hybrid E. PF1 are progressively replaced by UG clones in the industrial plantations around Pointe-Noire. This study emphasizes that, although the production of wood with the best UG clones is much higher than with the E. PF1 clones, the immobilization in the biomass is lower for P, K, Ca, and of the same extent for Mg. Moreover, the
5651 7604 3519 6338 11697
K
Ca
Mg
1475 1026
1022 780
1956 2071
550–621 456–609 957–1106 427–464 508–691
409–885 372–586 1014–1205 310–647 404–546
1313–1711 929–1340 1345–1858 974–1339 1828–2058
436–914 475 278
121–214 194 143
592–1557 1453 667
527 471 401 449 694
560 395 394 406 530
1673 1376 1361 1550 2074
amounts of P and K returning to the soil at the harvest were much higher for UG clones than for the clone 141. These features are therefore positive for the maintenance of soil fertility, especially in Congo where Eucalyptus production is largely dependent on conservative management of organic matter and nutrients (Nzila et al., 2002). The long-term availability of N in soils is more worrying since the input–output budget during the first rotation after afforestation on savanna is clearly unbalanced for the clone 1-41 (Laclau, 2001). A use efficiency of N of the same order of magnitude for the best UG clones (18-52, for example) and for the clone 1-41 shows that the losses of N at the harvest (mainly through stemwood removal and leaching) will remain large for UG clones. The deficit of N will therefore increase over successive rotations (Bouillet et al., 2001) and will impose to enhance the nitrogen inputs in these plantations.
Acknowledgements We thank the scientists involved in the breeding program in Congo for providing all facilities for the
204
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
harvest of the clonal test and for their valuable comments: Jean-Marc Bouvet, Raphael Gouma, Aubin Saya and Philippe Vigneron. We want also to acknowledge the Founders of UR2PI (Republic of Congo, ECO SA and CIRAD-Foreˆ t) for their financial support. We thank L. Veysseyre (IRD Laboratory) and G. Heral-Llimous (CIRAD Laboratory) for chemical analysis.
References Bouillet, J.P., Nzila, J.D., Laclau, J.P., Ranger, J., 2000. Eucalyptus commercial plantations in the equatorial zone, on the coastal plains of the Congo. In: Workshop Proceedings ‘‘Site Management and Productivity in Tropical Forest Plantations’’, Kerala, India, 7–12 December 1999, pp. 11–20. Bouillet, J.P., Safou-Matondo, R., Laclau, J.P., Nzila, J.D., Ranger, J., 2001. Changes over successive rotations in NPK fertiliser needs of Congolese clonal Eucalyptus plantations: Relationships with changes in soil fertility. In: Proceedings of the IUFRO Symposium ‘‘Developing the Eucalypt of the Future’’, Chile, 10–15 September 2001. Duchauffour, P., Bonneau, M., 1959. Une me´ thode nouvelle du dosage du phosphore assimilable dans les sols foretiers. Bull. Assoc. Franc¸aise pour l’Etude du sol. 4, 193–198. Dye, P.J., 2000. Water use efficiency in South African Eucalyptus plantations: a review. South. Afr. J. 189 . Dye, P.J., 2001. Modelling growth and water use in four Pinus patula stands with the 3-PG model. South. Afr. J. 191 . Gonzales, J.L., Stape, L., Laclau, J.P., Gava, J.L., 2004. Silvicultural effects on Eucalypt plantations. In: EucProd 2002, Hobart, Tasmania, 10–15 November 2002. Grace, J.C., Madgwick, H.A.I., 1987. Sampling procedures for estimating forest biomass in the Puruki watershed. N. Z. J. For. Sci. 17 (2–3), 272–282. Herbert, M.A., 1996. Fertilizers and Eucalypt plantations in South Africa. In: Attiwill, P.M., Adams, M.A. (Eds.), Nutrition of Eucalypts. CSIRO, Australia, pp. 303–325. Laclau, J.P., 2001. Dynamique du fonctionnement mine´ ral d’une plantation d’Eucalyptus. Effets du reboisement sur un sol de savane du littoral congolais; conse´ quences pour la gestion des plantations industrielles. Ph.D. thesis, Institut National Agronomique Paris-Grignon, Paris, 146 pp.
Laclau, J.P., Bouillet, J.P., Ranger, J., 2000. Dynamics of biomass and nutrient accumulation in a clonal plantation of Eucalyptus in Congo. For. Ecol. Manage. 128, 181–196. Laclau, J.P., Ranger, J., Bouillet, J.P., Deleporte, P., Nzila, J.D., 2002. Fonctionnement d’une plantation d’Eucalyptus et influence sur la fertilite´ des sols de savane du littoral congolais. Bois et Foreˆ ts des Tropiques. Laclau, J.P., Quillet, G., Wamba, P., 1996. Exploitation de plantations industrielles d’eucalyptus au Congo. Le Flamboyant 40, 8– 13. Loumeto, J.J., 1986. Contribution a` l’e´ tude de la distruibution mine´ rale dans les eucalyptus du Congo. The`se de l’universite´ de Rennes I. Sciences de la vie et de l’Environnement, 134 pp. Marschner, H., 1995. Mineral Nutrition of Higher Plants, second ed. Academic Press, 889 pp. Morais, E.J., Barros, N.F., Novais, R.F., Brandi, R.M., 1990. Biomassa e eficieˆ ncia nutricional de espe´ cies de eucalipto em duas regioes bioclimaticas de Minas Gerais. Revista Brasileira de Cieˆ ncia do Solo 14, 353–362. Negi, J.D.S., Sharma, S.C., 1996. Mineral nutrition and resource conservation in Eucalyptus plantations and other forest covers in India. In: Attiwill, P.M., Adams, M.A. (Eds.), Nutrition of Eucalypts. CSIRO, Australia, pp. 399–416. Nouguier, S., 1997. Se´ lection des clones de E. urophylla E. grandis. Analyse des tests clonaux: R86-10, R88-04, R89-05, R90-05, R90-06, R90-07, R91-03, R91-04 et R92-03. Note interne UR2PI, 47 pp. Nzila, J.D., Bouillet, J.P., Ranger, J., 2002. The effects of slash management on nutrient cycling and tree growth in Eucalyptus plantations in the Congo. For. Ecol. Manage. 171, 209–221. Ranger, J., Marques, R., Colin-Belgrand, M., Flammang, N., Gelhaye, D., 1995. The dynamics of biomass and nutrient accumulation in a Douglas-fir (Pseudotsuga menziesii) stand using a chronosequence approach. For. Ecol. Manage. 72, 167–183. Safou-Matondo, R., Bouillet, J.P., 1999. Essai factoriel NPK 33 sur replantation. Parcelle R97-8. Mensurations a` 9, 13 et 18 mois. Note interne UR2PI, 12 pp. Saya, R.A., 1999. Analyse finales du test clonal R91-03. Se´ lection des clones. Note interne UR2PI. Vigneron, P., Bouvet, J.M., Gouma, R., Saya, A., Gion, J.M., Verhaegen, D., 2000. Eucalypt hybrids breeding in Congo. In: Dungey, H.S., Dieters, M.J., Nickles, D.G. (Eds.), Hybrid Breeding and Genetics Forest Trees. Proceedings of QFRI/ CRC-SPF Symposium, 9–14 April, Noosa, Queensland, Australia. Vitousek, P.M., 1982. Nutrient cycling and nutrient efficiency. Am. Nat. 119, 553–572.
Hybrid and clonal variability of nutrient content and nutrient use efficiency in Eucalyptus stands in Congo R. Safou-Matondo a,*, P. Deleporte b, J.P. Laclau c, J.P. Bouillet d a UR2PI, BP 1291, Pointe-Noire, Congo CIRAD-Foreˆt/UR2PI, Pointe-Noire, Congo c CIRAD-Foreˆt/ESALQ-USP, Cx Postal 9, 13418-900 Piracicaba, SP, Brazil d CIRAD-Foreˆt, Campus International de Baillarguet, 34398 Montpellier Cedex 5, France b
Received 11 February 2004; received in revised form 21 January 2005; accepted 13 February 2005
Abstract For 20 years, there has been 42,000 ha estate of clonal Eucalyptus plantations around Pointe-Noire in Congo on sandy soils that have very low reserves of available nutrients. These plantations have been based on a natural hybrid (E. PF1). This hybrid is being replaced by E. urophylla E. grandis (UG), a more productive hybrid developed by the breeding program of UR2PI. A study of biogeochemical cycles showed that nutrient removal by harvesting is the main nutrient output in the E. PF1 ecosystem. It is therefore important to quantify the nutrient content (NC) in both hybrids to compare corresponding nutrient removal values. The work dealt with four UG clones and the most planted clone of E. PF1. Twelve trees per clone were sampled at the logging age (8 years) in a clonal test for UG clones and in a nearby stand for E. PF1. Tables were established to predict, from girth at breast height (C1.30 m), the biomass and nutrient content of stemwood, bark, dead and living branches, leaves, and were applied to the inventory of the different stands to evaluate corresponding biomass, NC and nutrient use efficiency (NUE) on a per-hectare basis. Total biomass differed between the two hybrids and among UG clones: 109 t ha1 for E. PF1 and 108–155 t ha1 for UG clones. In E. PF1 trees, total NC was globally lower for N, K, and Mg, but greater for P and Ca. In stemwood, nitrogen content was similar for both hybrids. By contrast, in UG clones, NC was much lower for P (72%) and Ca (40% to 55%). The same trends were observed for NUE: equivalent for both hybrids for N, but higher in UG clones for P (+72%) and Ca (+43% to +59%). A marked variability among clones was observed for K and Mg. UG clones allocated proportionally more nutrients in leaves than E. PF1. These results show that clones should not be selected only on growth traits but also on NUE and on the concentration of nutrients in tree components removed by harvesting. It will be then possible to limit the cost of fertilising needed to maintain stand growth and soil fertility. # 2005 Elsevier B.V. All rights reserved. Keywords: Biomass; Nutrient content; Nutrient use efficiency; Hybrid variability; Clonal variability; Eucalyptus
* Corresponding author. Tel.: +242 94 31 84; fax: +242 94 47 95. E-mail addresses: [email protected] (R. Safou-Matondo), [email protected] (P. Deleporte). 0378-1127/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.foreco.2005.02.049
194
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
1. Introduction More than 40,000 ha of Eucalyptus have been planted on savannas around Pointe-Noire in the Republic of Congo. They are mainly based on two natural hybrids (E. PF1 and E. tereticornis E. grandis (UG)), which appeared naturally among the first afforestation. With fertilising and weeding, these hybrids perform well (15–20 m3 ha1 year1), but they are replaced increasingly by the artificial hybrid E. urophylla E. grandis, that it is more productive (up to 40 m3 ha1 year1 of over-bark volume in clonal tests) (Nouguier, 1997; Saya, 1999). The duration of the stand rotation is 7–8 years. The current method of harvesting de-barks stems up to 7 cm in diameter for pulpwood, all residues of over 2 cm diameter being commercialised as firewood (Laclau et al., 1996, 2000). The dynamics of biomass and nutrient accumulation have been studied in an E. PF1 clone 1-41 chronosequence (Laclau et al., 2000). This clone has been the most extensively planted (7000 ha) because of its high growth and is used as a reference material in breeding and silvicultural trials. Biomass and nutrient content models have been established for each tree component (stemwood, bark, living and dead branches, leaves) for the major nutrients (N, P, K, Ca, Mg). These models can quantify the nutrient requirement of trees within the stand rotation and their nutrient content at harvesting. Nutrient content at harvesting is particularly important as a biogeochemical study showed that nutrient removal at that time is the main nutrient output in the E. PF1 ecosystem (Laclau et al., 2002). The nutrient content of the slash remaining after harvesting can also be estimated. This information is essential to identify methods for sustainable management of replanted site as soil fertility is closely linked to organic matter content (Bouillet et al., 2000; Nzila et al., 2002). In particular, the N nutrient budget— differences between N inputs and N outputs—is very unbalanced, and the need of N fertiliser increases markedly over successive rotations. Bouillet et al. (2001) estimated this deficit to 165 kg N ha1 at the end of planted crop rotation, 375 kg N ha1 after the first coppice rotation and 550 kg N ha1 at the end of the second coppice rotation. A study on the effects of site management on nutrient cycling and tree growth in Congo showed that
the nutrient content in the aboveground biomass of replanted Eucalypt stands was strongly dependent on slash and litter management practices (Nzila et al., 2002). A marked increase in nutrient concentration in foliage was observed as the nutrient content of organic residues on the soil surface increased. As UG hybrid is more productive than E. PF1, its nutrient content is likely to be higher than in E. PF1. However, the nutrient use efficiency (NUE) of mineral elements in biomass production could differ sharply between the two hybrids and/or among UG clones. The NC removed by harvesting could be then not proportional to the biomass produced. Different indexes have been proposed to represent NUE (Vitousek, 1982). One of the most common indexes represents the mean quantity of aboveground biomass (or of stemwood biomass) produced per unit of nutrient incorporated into the aerial components of the stands since planting (Morais et al., 1990; Herbert, 1996). This index may be then used to select highly productive genotypes with low nutrient accumulation that are better for adapted sustainable production. The current study compared the nutrient contents in stemwood, bark, dead and living branches, and leaves of four UG clones and the clone 1-41 of E. PF1. Its objectives were to establish, at the harvesting age (8 years), biomass and nutrient content for the different tree components, and to calculate the corresponding NUE. The variability between hybrids and clones was investigated.
2. Materials and methods 2.1. Site parameters The plantations are established on coastal plains around Pointe-Noire, Congo at latitude 48S and longitude 128E. The climate is sub-equatorial with a high atmospheric humidity (85% on average), with a rainy season from October to May and a dry season from June to September. Mean annual rainfall is around 1200 mm, and mean annual temperature is 25 8C with seasonal variations of about 5 8C. The soils are Ferralic Arenosols (FAO classification) characterised by a homogeneous sandy texture, acidic reaction, limited available nutrients, and very low
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
195
2000) and the same silviculture was applied: savannah ploughed before planting; spacing of 5 m 3 m; starter fertilisation of 150 g per plant of NPK (13% N, 6% P, 17% K); manual weeding in the planting row and mechanical weeding in the inter-row.
levels of exchangeable cations, organic matter and cation exchange capacity (Table 1) (Laclau, 2001). 2.2. Stand characteristics This study was carried out in two stands:
2.3. Methodology in evaluating biomass and nutrient content
A clonal test using a randomised complete block design with five replicates. Each unit plot consisted of 5 5 trees, at a stand density of 666 trees ha1 (Nouguier, 1997). Twenty-three UG clones were compared, among which four UG clones were sampled. Two clones (18-50 and 18-52) were full sibs, and the two other clones (18-64 and 18-65) were not related. A trial planted with the E. PF1 clone 1-41 located 1 km far from the clonal test, comparing various types of cuttings, with five replicates. No significant differences among treatments were observed.
The main stages adopted classically for this type of study have been followed (Laclau et al., 2000; Grace and Madgwick, 1987; Ranger et al., 1995): Inventory of the stands (125 trees per UG clone, and 600 trees for E. PF1 clone 1-41). Twelve trees per clone (four trees in three replicates distributed throughout six basal area classes defined from the inventory) were cut down, and the major components were isolated: stemwood (up to 7 cm and between 7 and 2 cm), stembark (up to 7 cm and between 7 and 2 cm), leaves, living branches (up to 2 cm and over 2 cm) and dead branches. The diameters of the 7 and 2 cm corresponded to the commercial limits for pulpwood and firewood,
The UG and the E. PF1 clones were not then planted in the same trial. However, it was decided to compare the results as the stands were established at the same date and harvested at the same age (8 years old), the soils properties were similar (Laclau et al.,
Table 1 Main soil characteristics in the Pointe-Noire region (from Laclau et al., 2000) Soil layer
Total elementsa
Particle size distribution
Clay (%) Silt (%) Sand (%) Orgamic material (%) N total (%) C/N Ca (%) Mg (%) K (%) P (%) Al (%) A11 (0–5 cm) A12 (5–70 cm) B1 (70–120 cm) B21 (120–200 cm) B22 (200–400 cm) B23 (400–600 cm)
4.5 4.0 1.7 4.8 6.9 5.8
2.4 3.4 7.3 6.7 5.9 8.3
Soil layer
pH (H2O)
A11 (0–5 cm) A12 (5–70 cm) B1 (70–120 cm) B21 (120–200 cm) B22 (200–400 cm) B23 (400–600 cm)
5.2 4.7 5.3 5.5 5.4 5.4
a b c d e
93.1 92.6 91.1 88.5 87.2 85.9
0.91 0.57 0.26 0.31 0.16 0.12
0.37 0.23 0.11 0.14 0.09 0.09
14.3 14.3 13.6 12.9 10.0 7.8
Exchangeable cationsb (cmolc kg1) K
Ca
Mg
Mn
Na
Al
0.03 0.05 0.01 0.01 0.02 0.02
0.19 0.10 0.09 0.08 0.08 0.08
0.07 0.04 0.02 0.02 0.02 0.02
0.01 0.01 0.01 0.01 0.01 0.01
0.02 0.04 0.04 0.01 0.02 0.02
0.10 0.17 0.08 0.13 0.10 0.07
Acid digestion and ICP determination of elements. Cobalt-hexamine extraction, ICP determination of cations. BC: exchangeable cations’ sum. CEC: cation exchange capacity. Available P (Duchauffour and Bonneau (1959) methodology).
1.41 1.91 0.85 1.09 1.11 1.28
0.23 0.26 0.22 0.23 0.26 0.25
0.15 0.16 0.16 0.19 0.22 0.20
0.05 0.04 0.05 0.05 0.04 0.04
0.63 0.64 0.92 0.97 1.14 1.08
BCc
CECd
BC/CEC (%)
Pavailablee (%)
0.31 0.23 0.16 0.12 0.14 0.14
0.43 0.43 0.26 0.32 0.28 0.30
72 53 61 37 50 47
0.03 0.02 0.02 0.02 0.03 0.02
196
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
respectively. The following variables were measured: Total length of the trees and stem circumference at the base, 0.30 m, 1.30 m (C1.30 m), and every 3 m up to a diameter of 2 cm. Fresh weight of the stem up to diameters of 7 cm and between 7 and 2 cm. Disks of wood and bark of constant thickness were taken from the centre of the stem pieces, every 3 m. These disks were dried until constant weight and the dry biomass of the components in each tree was calculated proportionally. All the branches and leaves were collected and samples of these components were dried. The branches with a diameter exceeding 2 cm were distinguished from the smaller ones, in order to quantify the biomass and nutrient content of firewood. 2.4. Chemical sample analyses Before the chemical analyses, every sample was dried at 65 8C, ground and homogenised. Total N was analysed by acid–base volumetry after Kjeldahl mineralisation. P was determined by cold-colorimetry from the Murphy and Riley reagents. Ca, K and Mg were analysed by atomic absorption spectrophotometry. 2.5. Statistical data processing To evaluate biomass and nutrient contents of each component of each clone, tables were established following the general model: Yi j ¼ a þ a j þ ðb Ci j Þ þ ðb j Ci j Þ þ ðc Ci2j Þ þ ðc j Ci2j Þ þ ðd Ci3j Þ þ ðd j Ci3j Þ þ ei j where Yij was the biomass or the nutrient content of a given component of tree i of the clone j; Cij was the circumference at breast height (C1.30 m) of tree i of the clone j; a, b, c and d were the global parameters of the model; a j, b j, c j and d j were the local parameters of the model, depending on clone j; eij was the residual error. The statistics were operated using GLM procedure of SYSTAT 9.0 software (SPSS Inc#). The backward stepping option was used to introduce only significant variables (at 0.05 threshold)
into the model. A tolerance threshold of 0.001 was chosen. Criteria to compare individual and global models were a maximum adjusted R2 value, a minimum root mean square error (RMSE) and no bias in the distribution of the residues. At first, non-weighted regressions were calculated for biomass and nutrient content for each component and each clone (8 compounds per tree 12 trees 5 clones) corresponding to 480 samples and 288 equations: 8 components 5 clones (biomass + 5 nutrients) and 48 equations for the global models. Clonal and global models were compared using F-test: Fobs ¼
ðSSE2 SSE1 Þ=ð p1 p2 Þ SSE1 =ðn p1 Þ
where p1 is the number of parameters with the clonal models, p2 is the number of parameters with the global model, SSE1 = sum of square error of the clonal model, SSE2 = sum of square error of the global model, N = number of observations, F obs was compared the value given in the Fisher’s table F( p1 p2, n p1). If F obs > F( p1 p2, n p1), then clonal model gave more precisely the biomass or the nutrient content better than the global model. These equations were then applied to the inventory of each plot of the stands to estimate, for each component, the biomass and nutrient content on a perhectare basis. The average of the five replicates gave an estimation of biomass and nutrient content per clone. ANOVA were performed to test differences between hybrids and clones, according to Bonferonni test ( p < 0.05). UG clones were compared through a twoway ANOVA, with five replicates and four clones, using Bonferonni test ( p < 0.05). E. PF1 clone 1-41 was compared with the most productive UG clone (18-52) through a one-way ANOVA with five replicates, at 0.05 threshold.
3. Results 3.1. Growth and stand production Differences in tree growth and stand production were observed among clones (Table 2 and Fig. 1). Even though clones 18-52 and 18-50 being full sibs, 18-52 exhibited a MAI 18% higher than the clone 18-50.
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
197
Table 2 Mean height, mean circumference, total volume of the stand and MAI of the Eucalyptus clones (results at 99 months) SEUGa
Clones
H (m) C1.30 m (cm) V (m3 ha1) MAI (m3 ha1 year1)
18-50
18-52
18-64
18-65
1-41
27.8 a 62.6 a 235.2 b 28.5
27.9 aa 66.4 aa 275.0 aa 33.3
25.9 b 53.1 b 176.0 d 21.3
26.6 ab 63.0 a 215.4 c 26.1
27.0 a 57.5 b 207.0 a 25.1
0.3 1.0 4.4
Latin letters indicate significant differences (threshold: 0.05) between clones UG using Bonferonni test. Greek letters indicate significant differences (threshold: 0.05) between the most productive clone UG 18-52 and the clone 1-41 using Bonferonni test. a SEUG: standard error for the comparison UG clones.
UG clones presented better MAI than E. PF1, except for the UG clone 18-64. This clone had therefore not been used in commercial plantations. 3.2. Biomass and nutrient content tables A comparison of the global model for all clones and the individual model of each clone showed that the global model was as precise as the individual model in 45 out of the 48 cases. The local parameters of the general model were different from zero in 46 out of the 48 cases: 37 equations (respectively, 9) exhibited one local parameter (respectively, 2) different from zero. The Figs. 2–4 give some examples of the differences observed among clones for some tree components, even between the full sib clones 18-50 and 18-52. The shape of the curves may vary markedly according to components, nutrients, and clones. R2 coefficients were generally higher than 0.85 (data not shown) and higher than 0.90 for stemwood,
except for Mg content (R2 = 0.79). The weakest R2 coefficients (0.53–0.75) were observed for wood and bark between diameters 7 and 2 cm, dead branches and living branches. The prediction of biomass and nutrient contents was then accurate for stemwood and bark over 7 cm and leaves but weaker for living branches and dead branches. 3.3. Biomass and nutrient content per hectare Large differences in total biomass were observed between the two hybrids and among UG clones, including the full sibs 18-50 and 18-52 (Table 3). The best clone (18-52) produced 20%, 27%, 42% and 44% more than clones 18-50, 18-65, 1-41 and 18-64, respectively. The biomass of stemwood ranged from 91 to 132 Mg ha1 in the UG clones, and 95 Mg ha1 for the clone 1-41. Stemwood represented the main part of the aerial biomass (Table 3). The proportion of the total biomass
Fig. 1. Changes in the individual basal area of the Eucalyptus clones according to stand age.
198
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
Fig. 2. Curves and parameters related to the model fitted for Ca content of leaves (the other parameters of the model are null).
Fig. 3. Curves and parameters related to the model fitted for P content of bark up to a diameter of 7 cm (the other parameters of the model are null).
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
199
Fig. 4. Curves and parameters related to the model fitted for N content of wood up to diameter 7 cm (other parameters of the model are null).
in this component was similar between hybrids: 86.5% for clone 1-41 and a mean of 84.6% for UG clones. In contrast, marked differences between hybrids were observed for bark and leaves: The percentage of bark was 28% higher in the clone 1-41 than in UG clones.
The percentage of leaves was 27% lower in the clone 1-41 than in UG clones. The clone 1-41 exhibited the lowest percentage of living branches. This component represented between 6.9% and 7.9% of the aerial biomass in clones 18-52 and 18-50, respectively.
Table 3 Total aerial dry biomass, corresponding MAI and proportion of the total biomass in the tree components of the Eucalyptus clones (results at 99 months) SEUGa
Clones Total biomass (t ha1) MAI (t ha1 year1) Proportion of total biomass (%) Stemwood (D > 2 cm) Bark (D > 2 cm) Leaves Living branches Dead branches
18-50
18-52
18-64
18-65
1-41
129.5 b 15.7
155.2 aa 18.8
107.6 c 13.0
121.9 b 14.8
109.4 b 13.3
83.8 3.8 2.6 7.9 1.7
83.6 4.5 2.7 7.8 1.4
86.5 5.0 1.9 4.9 1.7
86.1 3.5 2.4 6.9 1.2
85.0 3.8 2.7 7.9 0.6
2.5
Latin letters indicate significant differences (threshold: 0.05) between clones UG using Bonferonni test. Greek letters indicate significant differences (threshold: 0.05) between the most productive clone UG 18-52 and the clone 1-41 using Bonferonni test. a SEUG: standard error for the comparison UG clones.
200
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
Table 4 Biomass (t ha1) and nutrient content (kg ha1) of the tree components of the Eucalyptus clones (results at 99 months) Clone
Components Stemwood (D > 2 cm)
Bark (D > 2 cm)
Leaves
Living branches
Dead branches
Total
18-50
Biomass N P K Ca Mg
111.2 217.1 8.4 18.7 17.2 19.4
b ab ab c b b
4.5 19.9 14.6 12.1 30.7 16.8
b a a bc a a
3.1 60.6 5.4 17.3 12.7 12.3
bc b c c b b
9.0 35.6 8.8 15.1 13.7 9.8
b a a b b a
1.6 5.3 0.6 0.2 1.7 0.5
a a a c b c
129.5 338.5 37.7 63.4 76.2 58.8
b a b c b b
18-52
Biomass N P K Ca Mg
131.9 228.4 9.1 28.5 22.5 23.4
aa aa ab bb ab aa
5.9 17.8 15.8 12.7 31.3 16.5
aa bb aa ba aa ab
4.2 77.1 9.4 26.7 17.4 15.6
aa aa aa aa aa aa
12.2 29.2 8.4 17.7 15.3 9.7
aa ba aa aa aa aa
0.9 2.1 0.3 0.1 0.7 0.3
cb db ca db cb db
155.2 354.4 42.9 85.8 87.2 65.5
aa aa aa ba aa aa
18-64
Biomass N P K Ca Mg
90.8 151.1 7.9 27.0 20.9 9.4
c c ab b a d
4.1 14.6 7.6 10.9 19.5 10.8
b c b c b b
2.8 47.8 4.2 17.9 11.3 8.9
c c d c c d
8.2 21.3 7.6 15.2 12.1 7.3
b c a b c b
1.8 3.0 0.5 0.4 1.6 0.8
a c b a b a
107.6 237.8 27.9 71.2 65.4 37.2
c b c c c d
18-65
Biomass N P K Ca Mg
101.9 202.6 7.0 44.4 16.7 16.0
b b b a b c
5.5 19.1 6.9 17.7 33.4 11.9
a ab b a a b
3.3 68.0 7.7 23.3 11.5 10.9
b b b b bc c
9.5 32.4 7.6 19.7 12.4 9.8
b ab a a c a
1.7 4.2 0.6 0.4 2.0 0.7
a b a b a b
121.9 326.2 29.8 105.4 76.1 49.3
b a c a b c
1-41
Biomass N P K Ca Mg
94.6 190.5 28.5 39.4 37.3 16.7
b a a a a b
5.5 25.2 14.9 16.2 44.9 24.1
a a a a a a
2.1 42.7 3.3 9.2 7.1 6.5
b b b b b b
5.4 11.8 3.1 5.4 5.7 5.0
b b b b b b
1.9 6.3 0.5 0.2 2.8 0.8
a a a a a a
109.4 276.6 50.3 70.4 97.8 53.1
b a a a a a
SEUGa
Biomass N P K Ca Mg
2.1 4.4 0.4 0.9 0.4 0.4
0.1 0.4 0.3 0.4 0.8 0.3
0.1 1.8 0.2 0.5 0.3 0.3
0.4 0.8 0.3 0.5 0.3 0.2
0.1 0.1 0.02 0.01 0.05 0.02
2.5 7.2 0.7 2.3 1.7 1.1
Latin letters indicate significant differences (threshold: 0.05) between clones UG using Bonferonni test. Greek letters indicate significant differences (threshold: 0.05) between the most productive clone UG 18-52 and the clone 1-41 using Bonferonni test. a SEUG: standard error for the comparison UG clones.
The total accumulation of P and Ca was generally lower in UG clones than in the clone 1-41: on average 31% and 22%, respectively (Table 4). The opposite pattern was found for N (+14%) and K (+16%). No trend was found for Mg. The nutrient contents were always lower in the stemwood of UG clones than in clone 1-41 for P
(71%), Ca (48%) and K (25%). UG also exhibited systematically lower content of all nutrients in the bark, but a much higher accumulation of all nutrients in the living branches and the leaves. N accumulation increased with clone production, but not proportionally: (i) the total biomass of clone 18-65 was 13% greater than of the clone 18-64 but the
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204 Table 5 Biomass (t ha1) and nutrient content (kg ha1) removed by harvesting Clone
Biomass
N
P
K
Ca
18-50 18-52 18-64 18-65 1-41
116.2 136.1 93.9 105.7 98.1
231.7 234.5 156.2 213.6 195.8
13.1 11.6 12.2 11.2 31.5
22.6 32.5 30.4 48.7 41.6
24.3 28.3 25.3 22.9 41.1
24.3 28.2 12.1 19.8 20.1
UGa UG/1-41 (%)
113.0 115
209.0 107
12.0 38
33.5 81
25.2 61
21.1 105
a
Mg
UG: mean of the four E. urophylla E. grandis clones.
Table 6 Biomass (t ha1) and nutrient content (kg ha1) of slash after harvest Clone 18-50 18-52 18-64 18-65 1-41 UGa UG/1-41 (%) a
Biomass
Ca
Mg
13.3 19.1 13.7 16.1 11.3
N 106.8 119.9 81.6 112.6 80.8
P 24.6 31.3 15.7 18.6 18.8
K 40.8 53.3 40.8 56.7 28.8
51.9 58.9 40.1 53.2 56.7
34.5 37.3 25.1 29.5 33.0
15.6 138
105.2 130
22.5 120
47.9 1646
51.0 90
31.6 96
UG: mean of the four E. urophylla E. grandis clones.
corresponding increase in N content was 37%; and (ii) clone 18-52 was 20% more productive than the clone 18-50 but accumulated only 5% more nitrogen.
201
to the soil (leaves, branches and bark up to stem diameter of 7 cm) (Table 6). Nitrogen removal increased with biomass production, and corresponded on average to 66% of the total aerial N content. Nutrient removal was much lower on average in UG clones than clone 1-41 for P (62%), Ca (29%) and K (19%). Differences were observed among UG clones. For example, the nutrient content removed with clone 18-65 was the lowest for P but the greatest for K (Table 5). The contents of Ca and Mg in slash were quite similar for both hybrids (Table 6). In contrast, they were higher for K (+66%), N (+30%) and P (+20%) in UG clones than clone 1-41. Nevertheless, a noticeable variability was observed among UG clones. In particular, the clone 18-52 was characterised by the highest amounts of nutrients in slash. UG clones exhibited higher biomass production, except for the clone 18-64, and higher N removal by harvesting than the clone 1-41, but also greater N returns to the soil. The difference between nitrogen returning to the soil and nitrogen removed was then better for UG clones: 104 kg N ha1 on average for UG clones compared to 115 kg N ha1 for the clone 1-41 (Tables 5 and 6). The same conclusions could be drawn up for P (+10 kg ha1 versus 13 kg ha1), K (+14 kg ha1 versus 13 kg ha1) and Ca (+26 kg ha1 versus +10 kg ha1). No marked differences were pointed out for Mg (+11 kg ha1 versus +13 kg ha1).
3.4. Nutrient removals and cycling 3.5. Nutrient use efficiency We calculated the biomass and nutrient content removed by harvesting (stemwood and bark between stem diameters of 7 and 2 cm) (Table 5), and returning
The results found were consistent for the two formula of NUE: aerial biomass/aerial nutrient
Table 7 Values of NUE (in kg of dry biomass per kg of nutrient accumulated in the aerial biomass) according to the clone and the nutrient taken into account Clone
Total aerial biomass (t ha1)
NUE
18-50 18-52 18-64 18-65 1-41
129.5 155.2 107.6 121.9 109.4
382 438 452 374 395
SEUGa
2.5
N b aa c b b
P
3
c ba a c b
K
3430 3610 3860 4090 2130 60
c bca ab a b
Ca
2040 1810 1510 1160 1540 10
a ba c d b
1700 1780 1650 1600 1120 10
Mg b aa c c b
2200 2370 2890 2470 2030
d ca a b b
20
Latin letters indicate significant differences (threshold: 0.05) between clones UG using Bonferonni test. Greek letters indicate significant differences (threshold: 0.05) between the most productive clone UG 18-52 and the clone 1-41 using Bonferonni test. a SEUG: standard error for the comparison UG clones.
202
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
content; or stemwood biomass/aerial nutrient content. The UG clones exhibited much higher NUE than clone 1-41 for P and Ca (Table 7). Their NUE was also higher for Mg. In contrast, no real trends were observed for N and K. A strong variability in NUE was observed among UG clones, but the most efficient clone depended on the nutrient taken into account.
4. Discussion A general model, as precise as clonal models, was adjusted to predict stand biomass and nutrient contents. This general model is easier to use than individual models (in this study, the number of equations decrease from 240 to 48), but allows fairly good prediction by taking into account the clonal variability in within tree biomass allocation and in nutrient accumulation. The best prediction was observed for the components representing the main parts of the biomass and nutrient contents of the stand. A marked inter-clonal variability was observed in nutrient content. P and Ca contents aboveground in UG clones were lower than in the clone 1-41, even for clones 18-50 and 18-52, whose wood production was much higher. Discrepancies in nutrient content between hybrids were large in the bark but rather low in the wood. The amounts of nutrients in the canopy were higher in UG clones than in the clone 141. Considering the optimal foliar concentrations indicated by Herbert (1996) for E. grandis, the concentrations in N, K and especially Ca were weak. Foliar ratios N/P, N/K, P/K and Ca/Mg were far from the optimum for the UG clones (Table 8). This finding suggests a low soil availability in N and Ca and is
consistent with studies carried out in Congo showing: (i) a clear positive response of Eucalypt stand to N fertilisation (Bouillet et al., 2001); and (ii) that tree growth may be limited by a shortage of Ca (Nzila et al., 2002). By contrast, the very weak concentrations of K may have no consequential effects on the growth, as Na is likely to substitute partially to K (Marschner, 1995; Laclau, 2001) due to the proximity of the sea. The reciprocal recurrent selection scheme carried out in Congo (Vigneron et al., 2000) led to the development of UG clones with an architecture different from that of the natural hybrid E. PF1 (larger crowns, etc.). The UG clones exhibited higher use efficiency of P and Ca than the 1-41 clone. The use efficiency of K, Ca and Mg measured for UG clones were moreover higher than values reported for various Eucalyptus species in Brazil, India and South Africa (Laclau et al., 2000) (Table 9). By contrast, the use efficiency of P remained poor for UG clones compared to values reported in the literature for other species (Morais et al., 1990; Herbert, 1996; Negi and Sharma, 1996). The NUE value is very dependent on the availability of the element in the soil (Morais et al., 1990). The low use efficiency of P observed might be a result of the relatively high availability of P in the soil of the Pointe-Noire region, shown by the lack of response of trees to the input of P (Safou-Matondo and Bouillet, 1999). The high NUE of the trees sampled show that clones selected by the breeding program are particularly efficient to produce high amounts of wood despite the very low nutrient availability in the soil of the Pointe-Noire region. Thus, NUE might be a relevant index to integrate in the selection strategies
Table 8 Mean foliar concentrations of major nutrients and corresponding ratios for the Eucalyptus clones (results at 99 months) Clone
N (%)
P (%)
K (%)
18-50 18-52 18-64 18-65 1-41
1.96 1.82 1.71 2.05 2.05
0.17 0.22 0.15 0.23 0.16
0.56 0.63 0.64 0.70 0.44
E. grandis (Herbert, Minimum Maximum Optimum
1996) 1.25 3.35 2.8
0.10 0.35 0.15
0.36 1.19 0.75
Ca (%)
Mg (%)
N/P
N/K
P/K
Ca/Mg
0.41 0.41 0.40 0.35 0.34
0.40 0.37 0.32 0.33 0.31
11.3 8.2 11.3 8.9 12.9
3.5 2.9 2.7 2.9 4.7
0.31 0.35 0.24 0.33 0.36
1.04 1.12 1.27 1.06 1.10
0.56 1.82 >1.0
0.21 0.62 0.35
3.0 28.6 18
1.0 5.4 3.5
0.09 0.62 0.2
1.24 7.28 >3.3
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
203
Table 9 Nutrient use efficiency in the 7-year-old stand for adult Eucalyptus plantations from the literature (data expressed in kg of dry biomass per kg of nutrient accumulated in the aerial biomass) N
P
Congo E. PF1 hybrid, clone 1-41, 7 years (Laclau et al., 2000) E. PF1 hybrid, clone 1-38, 7 years (Loumeto, 1986)
592 611
2000 1638
Brazil E. grandis Hill, 8 years (Morais et al., 1990) E. saligna Smith, 8 years (Morais et al., 1990) E. cloeziana F. Muell., 8 years (Morais et al., 1990) E. citriodora Hook., 8 years (Morais et al., 1990) E. brassiana S.T. Blake, 8 years (Morais et al., 1990)
282–443 246–315 307–352 241–318 227–340
4269–8783 3997–6327 5997–11351 4064–7357 4419–8727
India E. PF1 hybrid, 7–10 years (Negi and Sharma, 1996) E. globulus Labill., 9 years (Negi and Sharma, 1996) E. grandis Hill, 9 years (Negi and Sharma, 1996)
318–352 155 204
1370–4333 2841 2000
South Africa E. fastigata Deane & Maiden, 7 years (Herbert, 1996) E. grandis Hill, 7 years (Herbert, 1996) E. macarthurii Deane & Maiden, 7 years (Herbert, 1996) E. nitens Deane & Maiden, 7 years (Herbert, 1996) E. smithii R. Baker, 7 years (Herbert, 1996)
487 538 438 510 822
since this parameter provides an indication of the amounts of nutrients exported at the harvest. The inputs of fertiliser needed to ensure a sustainable yield of commercial plantation will be therefore dependent on the NUE of the clones planted. However, correlation between stemwood production and values of NUE was low for the five clones sampled, and the most productive clone 18-52 did not have the best NUE for any component, except for Ca. This pattern indicates that maximizing yields and economic returns needs to take into account the efficiency of other factors involved in growth of the trees, as light or water (Dye, 2000, 2001; Gonzales et al., 2004).
5. Conclusion Clones of the natural hybrid E. PF1 are progressively replaced by UG clones in the industrial plantations around Pointe-Noire. This study emphasizes that, although the production of wood with the best UG clones is much higher than with the E. PF1 clones, the immobilization in the biomass is lower for P, K, Ca, and of the same extent for Mg. Moreover, the
5651 7604 3519 6338 11697
K
Ca
Mg
1475 1026
1022 780
1956 2071
550–621 456–609 957–1106 427–464 508–691
409–885 372–586 1014–1205 310–647 404–546
1313–1711 929–1340 1345–1858 974–1339 1828–2058
436–914 475 278
121–214 194 143
592–1557 1453 667
527 471 401 449 694
560 395 394 406 530
1673 1376 1361 1550 2074
amounts of P and K returning to the soil at the harvest were much higher for UG clones than for the clone 141. These features are therefore positive for the maintenance of soil fertility, especially in Congo where Eucalyptus production is largely dependent on conservative management of organic matter and nutrients (Nzila et al., 2002). The long-term availability of N in soils is more worrying since the input–output budget during the first rotation after afforestation on savanna is clearly unbalanced for the clone 1-41 (Laclau, 2001). A use efficiency of N of the same order of magnitude for the best UG clones (18-52, for example) and for the clone 1-41 shows that the losses of N at the harvest (mainly through stemwood removal and leaching) will remain large for UG clones. The deficit of N will therefore increase over successive rotations (Bouillet et al., 2001) and will impose to enhance the nitrogen inputs in these plantations.
Acknowledgements We thank the scientists involved in the breeding program in Congo for providing all facilities for the
204
R. Safou-Matondo et al. / Forest Ecology and Management 210 (2005) 193–204
harvest of the clonal test and for their valuable comments: Jean-Marc Bouvet, Raphael Gouma, Aubin Saya and Philippe Vigneron. We want also to acknowledge the Founders of UR2PI (Republic of Congo, ECO SA and CIRAD-Foreˆ t) for their financial support. We thank L. Veysseyre (IRD Laboratory) and G. Heral-Llimous (CIRAD Laboratory) for chemical analysis.
References Bouillet, J.P., Nzila, J.D., Laclau, J.P., Ranger, J., 2000. Eucalyptus commercial plantations in the equatorial zone, on the coastal plains of the Congo. In: Workshop Proceedings ‘‘Site Management and Productivity in Tropical Forest Plantations’’, Kerala, India, 7–12 December 1999, pp. 11–20. Bouillet, J.P., Safou-Matondo, R., Laclau, J.P., Nzila, J.D., Ranger, J., 2001. Changes over successive rotations in NPK fertiliser needs of Congolese clonal Eucalyptus plantations: Relationships with changes in soil fertility. In: Proceedings of the IUFRO Symposium ‘‘Developing the Eucalypt of the Future’’, Chile, 10–15 September 2001. Duchauffour, P., Bonneau, M., 1959. Une me´ thode nouvelle du dosage du phosphore assimilable dans les sols foretiers. Bull. Assoc. Franc¸aise pour l’Etude du sol. 4, 193–198. Dye, P.J., 2000. Water use efficiency in South African Eucalyptus plantations: a review. South. Afr. J. 189 . Dye, P.J., 2001. Modelling growth and water use in four Pinus patula stands with the 3-PG model. South. Afr. J. 191 . Gonzales, J.L., Stape, L., Laclau, J.P., Gava, J.L., 2004. Silvicultural effects on Eucalypt plantations. In: EucProd 2002, Hobart, Tasmania, 10–15 November 2002. Grace, J.C., Madgwick, H.A.I., 1987. Sampling procedures for estimating forest biomass in the Puruki watershed. N. Z. J. For. Sci. 17 (2–3), 272–282. Herbert, M.A., 1996. Fertilizers and Eucalypt plantations in South Africa. In: Attiwill, P.M., Adams, M.A. (Eds.), Nutrition of Eucalypts. CSIRO, Australia, pp. 303–325. Laclau, J.P., 2001. Dynamique du fonctionnement mine´ ral d’une plantation d’Eucalyptus. Effets du reboisement sur un sol de savane du littoral congolais; conse´ quences pour la gestion des plantations industrielles. Ph.D. thesis, Institut National Agronomique Paris-Grignon, Paris, 146 pp.
Laclau, J.P., Bouillet, J.P., Ranger, J., 2000. Dynamics of biomass and nutrient accumulation in a clonal plantation of Eucalyptus in Congo. For. Ecol. Manage. 128, 181–196. Laclau, J.P., Ranger, J., Bouillet, J.P., Deleporte, P., Nzila, J.D., 2002. Fonctionnement d’une plantation d’Eucalyptus et influence sur la fertilite´ des sols de savane du littoral congolais. Bois et Foreˆ ts des Tropiques. Laclau, J.P., Quillet, G., Wamba, P., 1996. Exploitation de plantations industrielles d’eucalyptus au Congo. Le Flamboyant 40, 8– 13. Loumeto, J.J., 1986. Contribution a` l’e´ tude de la distruibution mine´ rale dans les eucalyptus du Congo. The`se de l’universite´ de Rennes I. Sciences de la vie et de l’Environnement, 134 pp. Marschner, H., 1995. Mineral Nutrition of Higher Plants, second ed. Academic Press, 889 pp. Morais, E.J., Barros, N.F., Novais, R.F., Brandi, R.M., 1990. Biomassa e eficieˆ ncia nutricional de espe´ cies de eucalipto em duas regioes bioclimaticas de Minas Gerais. Revista Brasileira de Cieˆ ncia do Solo 14, 353–362. Negi, J.D.S., Sharma, S.C., 1996. Mineral nutrition and resource conservation in Eucalyptus plantations and other forest covers in India. In: Attiwill, P.M., Adams, M.A. (Eds.), Nutrition of Eucalypts. CSIRO, Australia, pp. 399–416. Nouguier, S., 1997. Se´ lection des clones de E. urophylla E. grandis. Analyse des tests clonaux: R86-10, R88-04, R89-05, R90-05, R90-06, R90-07, R91-03, R91-04 et R92-03. Note interne UR2PI, 47 pp. Nzila, J.D., Bouillet, J.P., Ranger, J., 2002. The effects of slash management on nutrient cycling and tree growth in Eucalyptus plantations in the Congo. For. Ecol. Manage. 171, 209–221. Ranger, J., Marques, R., Colin-Belgrand, M., Flammang, N., Gelhaye, D., 1995. The dynamics of biomass and nutrient accumulation in a Douglas-fir (Pseudotsuga menziesii) stand using a chronosequence approach. For. Ecol. Manage. 72, 167–183. Safou-Matondo, R., Bouillet, J.P., 1999. Essai factoriel NPK 33 sur replantation. Parcelle R97-8. Mensurations a` 9, 13 et 18 mois. Note interne UR2PI, 12 pp. Saya, R.A., 1999. Analyse finales du test clonal R91-03. Se´ lection des clones. Note interne UR2PI. Vigneron, P., Bouvet, J.M., Gouma, R., Saya, A., Gion, J.M., Verhaegen, D., 2000. Eucalypt hybrids breeding in Congo. In: Dungey, H.S., Dieters, M.J., Nickles, D.G. (Eds.), Hybrid Breeding and Genetics Forest Trees. Proceedings of QFRI/ CRC-SPF Symposium, 9–14 April, Noosa, Queensland, Australia. Vitousek, P.M., 1982. Nutrient cycling and nutrient efficiency. Am. Nat. 119, 553–572.