Annals of Botany 91: 879-892, 2003
© 2003 Annals of Botany Company
Nutrient Dynamics throughout the Rotation of Eucalyptus Clonal Stands in Congo
1 CIRAD-Forêt/UR2PI, TA 10/C, 34398 Montpellier Cedex 5, France, 2 INRA, Biogéochimie des écosystèmes forestiers, 54280 Seichamps, France and 3 UR2PI, BP 1264, Pointe-Noire, République du Congo
* For correspondence at: CIRAD-Forêt/USP-ESALQ, IPEF, Av. Pádua Dias 11, Caixa Postal 530, CEP 13·400-970, Piracicaba-SP, Brazil. E-mail: laclau{at}cirad fr
Received: 15 November 2002; Returned for revision: 2 January 2003 ; Accepted: 21 February 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONLITERATURE CITED |
|---|
The dynamics of the main nutrient fluxes of the biological cycle were quantified in a clonal Eucalyptus plantation throughout the whole planted crop rotation: current annual requirements of nutrients, uptake from the soil, internal translocations within trees, return to soil (litterfall and crown leaching) and decomposition in the forest floor. As reported for other species, two growth periods were identified in these short-rotation plantations: (1) a juvenile phase up to canopy closure, during which the uptake of nutrients from the soil reserves supplied most of the current requirements; and (2) a second phase up to harvest, characterized by intense nutrient recycling processes. Internal translocation within trees supplied about 30 % of the annual requirements of N and P from 2 years of age onwards, and about 50 % of the K requirement. The mineralization of large amounts of organic matter returned to the soil with litterfall during stand development represented a key process providing nutrients to the stand at the end of the rotation. The importance of the recycling processes was clearly shown by the small amounts of nutrients permanently immobilized in the ligneous components of trees, compared with the total requirements accumulated over the stand rotation which were two to four times higher. Small pools of nutrients circulating quickly in the ecosystem made it possible to produce high amounts of biomass in poor soils. The sustainability of these plantations will require fertilizer inputs that match the changes in soil fertility over successive rotations, mainly linked to the dynamics of organic matter in this tropical soil.
Key words: Biomass, nutrient content, dynamics, Eucalyptus, eucalypt, plantation, Congo, Africa.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONLITERATURE CITED |
|---|
A better understanding of the nutrient recycling processes that enable Eucalyptus plantations to produce large amounts of biomass in soils of low fertility is essential to optimize silvicultural practices so as to maintain soil fertility over successive rotations. Although some fluxes of nutrients have been studied intensively (mainly litterfall and nutrient content of the trees), studies quantifying the dynamics of the main fluxes of the biological cycle of nutrients during stand development are scarce for Eucalyptus plantations (e.g. Bargali et al., 1992; Gonçalves et al., 1997; Parrotta, 1999; Nambiar et al., 2000). Case studies are necessary to improve our knowledge of the mineral functioning of these stands and to validate models simulating nutrient cycling processes in forest ecosystems. Moreover, the rapid development of a fast-growing species makes this type of management a good model for dynamic studies (Ranger and Colin-Belgrand, 1996).
About 40 000 hectares of Eucalyptus plantations have been planted in littoral savannas in Congo. The short-rotation forestry carried out in these plantations can be considered representative of most of the Eucalyptus industrial plantations worldwide. Sustainable management of these plantations, i.e. long-term production and maintenance of the environment, has been identified as a priority for research in Congo. This question is particularly relevant as large amounts of nutrients are exported every 7 years with biomass removal from soils that are sandy, acidic, and have low reserves of available nutrients (Laclau et al., 2000).
The fluxes of nutrients studied in this ecosystem corresponded to the biological cycle defined by Switzer and Nelson (1972). The objectives were: (1) to quantify the dynamics of the main nutrient fluxes of the biological cycle throughout the development of a Eucalyptus plantation; (2) to assess the actual amounts of nutrients involved in stand growth; and (3) to identify and quantify the different sources of these nutrients. The aim of the study was to gain insight both into the nutrition of Eucalyptus plantations and into the influence of these stands on soil functioning.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONLITERATURE CITED |
|---|
Study sites
The ecological situation and the main characteristics of the plant material studied have been described previously (Laclau et al., 2000). In brief, the climate is characterized by an annual rainfall of approx. 1200 mm and a marked dry season from June to September. The atmospheric humidity is approx. 85 % on average with little seasonal variation (2 %). The average temperature is 25 °C with seasonal variations of approx. 5 °C. The geological bedrock is composed of thick detritic formations of continental origin, dating from the plio-pleistocene. The soils are Ferralic Arenosols (FAO classification) characterized by their homogeneity in the landscape, their colour (greyish on the surface and ochre at depth), their texture (sand content >85 %), their structure (always particulate), their depth (>10 m) and their chemical paucity. The ability of the soil to retain water is very low (about 12 % in volume).
The study was carried out at two sites in Congo located 20 km apart: (1) a chronosequence of Eucalyptus stands of the same clone representing the whole rotation and situated in a 500-m radius was sampled at Kissoko; and (2) a diachronic study was carried out on trees aged between 6 and 9 years old in a planted crop of the same clone at Kondi. Both sites are located on a plateau at an elevation of about 100 m, and are approx. 10–20 km from the sea. The main characteristics of soil and climate are similar at both sites. Reserves of nutrients in the soil were quantified in the 6-year-old stand at Kondi (Table 1).
|
Plant material
The stands sampled at Kissoko and Kondi are Eucalyptus crops consisting of the most productive clone of hybrid E. PF1. They were planted between 1989 and 1997 on savanna (Table 2). The hybrid E. PF1 derives from natural crosses between two or three individuals of Eucalyptus alba Reinw. ex Blume (female tree) in Congo and a group of imprecisely identified Eucalyptus hybrids (male tree) from a Brazilian arboretum (Delwaulle, 1988).
|
The main characteristics of the silviculture were similar in all stands, and growth of these stands was not severely affected by any biotic factors. Growth curves produced throughout the stand rotation showed that the hypothesis that age is the only difference between stands of the chronosequence can be accepted (Laclau et al., 2000). However, the stand density was higher in the chronosequence of Kissoko than in the plantation of Kondi (Table 2). In all stands, fertilizer was applied initially and weeding during the first 2 years of growth led to the lack of an understorey. In Congo, stands are usually clearfelled when they are 7 years old, but harvesting can take place between 6 and 9 years according to the pulpwood market.
Stand biomass and nutrient content evaluation
The methods used to determine stand biomass and nutrient content have been described previously (Laclau et al., 2000). In brief, above-ground biomass and nutrient contents were estimated by sampling 12 trees with a range of basal areas (determined by an inventory) in all the stands of the chronosequence. In the Kondi stand, 12 and ten trees were sampled at ages 5·3 and 8·3 years, respectively. The trees were separated into the following compartments: leaves, living branches, dead branches, stemwood and stembark. Stems were separated into two groups based on stem diameter, one group with diameter >7 cm at the thinner end and the other with diameter > 2 cm at the thinner end. Diameters, lengths and weights were measured in the field. Sub-samples were taken from all the compartments for humidity measurements and chemical analysis. Biomass and nutrient content tables were established for each component between the ages of 1 and 7 years in the chronosequence of Kissoko, and between 6 and 9 years in the Kondi stand, as polynomials of circumference at breast height and age.
Below-ground biomass and nutrient content were assessed from tables established by sampling three trees in each of the 1-, 4- and 7-year-old stands of the chrono sequence (Laclau et al., 2000).
Nutrient returns to the soil
Litterfall and forest floor.
Litterfall was collected monthly for 1 year during the first, second, third, sixth and eighth years of growth using ten litter-traps (75 cm x 75 cm) randomly situated in the stands in the chronosequence of Kissoko. Litterfall at intermediary ages was estimated by linear interpolation between the nearest ages sampled. In the Kondi stand, litterfall was collected monthly between the ages of 6 and 9 years, using 15 litter-traps identical to those used at Kissoko.
Forest floor material was collected at the end of the rainy season during 2 successive years in ten randomly chosen quadrats (0·25 m2) in three stands of the chronosequence of Kissoko. The sampling ages were 1·5 and 2·5 years for the younger stand, 3·5 and 4·5 years for the intermediary stand, and 7·5 and 8·5 years for the older stand. Litter samples were separated into four fractions by hand-picking: leaves, twigs, bark and miscellaneous (mainly fruits). Samples were dried at 65 °C until a constant weight was attained before being ground and mixed.
Crown leaching.
Crown leaching was estimated at the end of stand rotation from bulk precipitation, throughfall and stemflow measurements made over 3 years at Kondi, using the equation:
Cj = Tj – Dj – Dj
where Cj is the amount of the element j leached from the canopy, Tj is the flux of j beneath the canopy (throughfall + stemflow), Wj is wet deposition of j measured in an adjacent open area and Dj is dry deposition of j in the canopy. Na+ was considered as a tracer of dry deposition in the canopy owing to the proximity of the sea, and the dry deposition of other elements was considered to be proportional to that of Na+ (Hansen, 1994). The methodology is described accurately in Laclau (2001) and Laclau et al. (2003). Crown leaching was estimated in the chronosequence of Kissoko, correcting the values measured at Kondi proportionally to the foliar biomass of the stand at each age.
Chemical analysis
Most plant samples were analysed in a laboratory of Pointe-Noire, Congo. Total N was analysed by acid–base volumetry after Kjeldahl mineralization (Büchi B316). Following digestion with nitric acid, P was determined colorimetrically at room temperature using Murphy and Riley reagents (ANA 8 Prolabo), Ca and Mg by atomic absorption spectrophotometry, and K by flame emission spectrophotometry (GBC 901). The other samples were analysed by a laboratory in France, where N was determined by thermal conductivity after combustion (FP-428), and P, K, Ca and Mg were determined using a sequential spectrometer ICP (JY 24) after digestion with hydrofluoric acid and double calcination. The consistency of determinations performed by the two laboratories was checked, and slight differences between laboratories were corrected by intercalibration. The ash content of all the forest floor samples was determined by combustion for 4 h at 450 °C. Values of forest floor biomass presented here were corrected to eliminate the effect of the remaining soil particles. Methods used for solution analysis are described in Laclau et al. (2003).
Main fluxes of the biological cycle
Concepts and methods defined in the literature were used (Turner and Lambert, 1983; Helmisaari, 1992; Ranger and Colin-Belgrand, 1996).
Requirements.
Total current requirements for biomass production were defined as the total amount of nutrients accumulated in the biomass produced during the current year (expressed in kg ha–1 per year). Requirements were estimated at the stage of maximum maturity of vegetative organs using the equation:
where Rj is the total annual requirement of chemical element j by the stand, Bi is the biomass of compartment i in the tree produced during the current year, Ci,j is the concentration of chemical element j in the biomass produced during the current year in compartment i and n is the number of compartments distinguished in the trees.
Bi was quantified as the increment in biomass of each compartment i (calculated from tables established for the stand at each age), plus the biomass returned to the soil in the compartment through litterfall (in the case of leaves, branches and bark). To take into account the decrease in leaf mass during senescence, the annual biomass of leaves in litterfall was multiplied by 1·05 in the calculation of the increment in biomass of the compartment ‘leaves’. The specific leaf area of living leaves was, on average, 1·05 times higher than that of dead leaves for this clone (Laclau, 2001).
For stemwood, Ci,j was the mean concentration of element j measured in the outer annual ring of four trees sampled at each age between 1 and 7 years in the chronosequence of Kissoko (Laclau et al., 2001b) and three trees in the stand of Kondi at age 7 years. For leaves and bark, Ci,j was the mean concentration of j determined for 10–12 trees sampled at each age. For branches, Ci,j was the mean concentration of j in living branches sampled on 12 trees in the 1-year-old stand of the chronosequence. For roots, Ci,j was the mean concentration of j in roots sampled in three trees, at the nearest age in the chronosequence.
The fluxes of nutrients corresponding to the turnover of fine roots were not taken into account in the present study. We considered that the release of nutrients due to their death roughly equalled the amounts taken up to produce new fine roots. This hypothesis led to underestimation of the nutrient requirements and uptake from the soil, in particular during the first 2 years of stand growth when the fine root biomass increases sharply (Bouillet et al., 2002).
Immobilization
Annual immobilization of nutrients in the biomass was defined as the amount of elements permanently fixed in the biomass produced during the current year. For a particular year and a particular stage of development, immobilization is ligneous biomass production (calculated from tables established in each stand) multiplied by a stabilized concentration, i.e. residual concentration after translocation. This stabilized concentration can be considered to be identical to the mean concentration of ligneous compartments when the proportion of tissues with non-stabilized concentrations is low (Ranger and Colin-Belgrand, 1996). In the present study, the stabilized concentration in branches, bark and large roots was considered to be identical to the mean concentration determined at each age in these compartments. In stemwood, the mean concentration determined on four trees at each age was taken into account, excluding the outer ring where most of the internal translocation occurs (Laclau et al., 2001b).
Uptake
Annual nutrient uptake from the available soil reservoir by stands was defined as the sum of nutrients bound in the biomass and of nutrients returning to the soil as litterfall or as throughfall from crown leaching:
Uj = Ij + 
fj + Lj + Fj + Cj
where Uj is uptake of chemical element j from soil during the year, Ij is immobilization of j in the stand during the same period, fj is variation in the content of j in the foliage, and Lj, Fj and Cj are the amounts of j returning to the soil with litterfall, fine root turnover and crown leaching, respectively. Fine root turnover was not taken into account in the present study (Fj = 0). Annual uptake calculated from this equation includes nutrients taken up from the mineral soil and nutrients released from litter decay.
Internal translocation
Total internal translocation from old, physiologically inactive tissues to young growing tissues corresponds to the difference between the amount of nutrients required for annual growth of the stands (requirements) and the amount of nutrients taken up from the soil during the same period (uptake). Nevertheless, in the present study, internal translocation in the aerial part of the trees was determined independently of the measurements of requirements and uptake by the stands.
Translocation from foliage.
Translocation from foliage was calculated for a particular stage of stand development as the difference between the amount of nutrients in the foliage and the amount returning to the soil as litterfall or crown leaching:
Tf,j = (L x 1·05 x Cll,j) – (L x Csl,j) – Cj
where Tf,j is the translocation of the chemical element j from foliage, L is biomass of leaves in litterfall, the factor 1·05 was introduced to take into account the decrease in leaf mass during senescence, Cll,j is the concentration of j in living leaves, Csl,j is the concentration of j in senescent leaves collected in litter-traps and Cj is the amount of j returning to the soil with crown leaching.
Translocation from branches.
The biomass of dead branches remained stable from 2 years of age in the chronosequence of Kissoko (Laclau et al., 2000). Annual translocations from this compartment were calculated for each stand, multiplying the annual biomass of branches in litterfall by the difference in concentration between young branches (the active tissues) and old branches collected in litter-traps (whose concentration had stabilized after translocation).
Translocation from stems: stemwood.
Stemwood was specifically sampled to quantify internal translocation fluxes of nutrients. Discs of wood were sampled at the stump level and every 4 m up to the top, from four trees in 1-, 2-, 3-, 4-, 5-, 6- and 7-year-old stands, corresponding to the chronosequence in Table 2, and in adjacent stands of the same clone to complete the age series (Laclau et al., 2001b). Annual rings were not visible to the naked eye on the discs of wood sampled owing to the continuous growth of tropical eucalypts during the year. Taper functions fitted to stem profiles throughout stand development, combined with annual measurements of trees, were used to locate accurately the position of annual rings in stems. The wood of each annual ring was separated in all discs, dried at 65 °C until it reached constant weight, ground and homogenized. Chemical analysis, performed individually for each ring per level and per tree sampled, was used to quantify the changes in nutrient content of the rings during stand development. Nutrient translocations in stemwood were thus calculated stepwise between two successive ages.
Translocation from stems: stembark.
The annual increment of stembark is relatively small, and separating annual increments for chemical analysis requires micro-analysis techniques. These were not used here, and annual translocation from stembark was evaluated using the equation:
Tb,j = b x (Clb,j – Csb,j) – Sj
where Tb,j is the annual translocation of element j from stembark, b is the biomass of bark in litterfall, Clb,j is the concentration of j in living bark, Csb,j is the concentration of j in senescent bark collected by litter-traps and Sj is the amount of j leached in stemflow. Sj was calculated at Kondi by multiplying the annual stemflow flux by the difference between nutrient concentration in stemflow and in throughfall (Laclau, 2001). For stands of the chronosequence at Kissoko, the mean value of Sj determined at Kondi was corrected proportionally to the stembark biomass of each stand.
Forest floor decomposition
The annual rate of litter decay was quantified by comparing the amount on the forest floor after an interval of 1 year, and taking into account the amount of litterfall during the period. The amount (A) of decomposed forest floor litter is given by the equation:
A = F1 + L1–2 – F2
where F1, F2 and L1–2 are, respectively, the amount of forest floor litter at ages 1 and 2, and the amount of litterfall between ages 1 and 2. This equation is used to calculate the amount of biomass that has decomposed as well as to assess the net amount of nutrients released during the period.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONLITERATURE CITED |
|---|
Dynamics of biomass and nutrient accumulation
The dynamics of biomass and nutrient content in the various compartments of the trees up to age 7 years have been presented previously (Laclau et al., 2000). Biomass production between ages 6 and 9 years at Kondi was higher than that measured for the chronosequence at Kissoko (Fig. 1). Nitrogen accumulation was also higher at Kondi than in the chronosequence of Kissoko at the end of stand rotation, partly owing to the higher foliar biomass. The dynamics of the incorporation of each mineral during growth of the trees clearly differed among the elements: the accumulation of N and Ca continued up to 9 years, whereas the amounts of P, K and Mg incorporated into the biomass increased only slightly at the end of the stand rotation. The amounts of N, P, K, Ca and Mg accumulated in the 9-year-old stand represented 380, 54, 83, 100 and 56 kg–1 ha–1, respectively.
|
Total nutrient requirements of the stands
Current nutrient requirements increased sharply during the early growth period to reach a maximum around 2 years of age in the chronosequence (Fig. 2A). From this age onwards, requirements for N, P and Ca remained roughly stable, whereas there was a slight decrease in the requirement for K and Mg. Higher nutrient requirements at the end of the stand rotation at Kondi were a result of a higher biomass production.
|
Leaf production represented 60–70 % of N requirements throughout the stand rotation and 35–55 % of P, K, Ca and Mg requirements regardless of stand age (Table 3). The leaf life span of, on average, 6 months for this clone (Laclau, 2001) accounted for the high nutrient requirements to maintain the living crown throughout the rotation. Although stemwood production amounted to approx. 60 % of biomass production at the end of the stand rotation, nutrient requirements for stemwood increment were low. They represented about 30 % of total requirements for N, P and K, and 15 % for Ca and Mg. Despite a biomass production approx. ten times times lower for stembark than for stemwood, requirements of P, Ca and Mg were approximately the same throughout stand rotation. Nutrient requirements for the below-ground biomass increment (excluding fine root turnover) were low throughout the stand rotation.
|
Nutrient immobilization in the ligneous components of the trees
Permanent immobilization in the perennial components of the trees (stem + branches + large roots) was much lower than current requirements, regardless of stand age (Fig. 2B). This pattern showed the importance of recycling processes in this ecosystem. Nutrients were mainly immobilized in branches during the first year of growth; the major sink then became stemwood (data not shown). Current immobilization remained stable from age 2 years up to the end of the stand rotation for N and P but decreased regularly for K, Ca and Mg.
Nutrient uptake from the soil
Nutrients taken up from the soil provided most of the annual requirements during the first year of growth (Fig. 2C). At this stage, about two-thirds of the annual N uptake and 40 % of P, K, Ca and Mg uptake were directed towards foliage biomass (Table 4). From the second year of growth onwards, immobilization in the perennial biomass and litterfall were the two main nutrient sinks. The pattern differed according to the element: (1) N content in litterfall was of the same order of magnitude as that immobilized; (2) P and K contents in litterfall were lower than those immobilized; but (3) Ca and Mg contents in litterfall were higher than those immobilized. Even for mobile elements such as K, returns to the soil by crown leaching had very low absolute values (Table 4).
|
As observed for current annual requirements, maximal uptake from the soil occurred during the second year of growth for all elements. The current uptake then remained roughly stable (for N, P and Ca) or decreased slightly (for K and Mg) until the end of the rotation.
Internal translocation
Internal translocation of nutrients in the aerial part of trees was very limited during the first year after planting but increased sharply in the second year for N, P and K (Fig. 2D). In absolute terms, translocations were in the following order: N > K > P > Mg > Ca. On average, translocation supplied 60 % of annual requirements from 4 years of age for K, and 30–40 % for N and P. Regardless of stand age, internal translocation of Mg remained low, and a trend of nutrient accumulation in senescent tissues (negative translocations) was observed for Ca. Nutrient translocations occurred essentially in the foliage but stemwood was an important source of K at the end of stand rotation (Table 5). Branches and stembark were a minor source of N, P and K but translocation in stembark represented about 2 kg ha–1 per year for Ca and 1 kg ha–1 per year for Mg at the end of the stand rotation.
|
Litterfall and forest floor decomposition
Nutrient returns to the soil from litterfall were low in the first year (<5 kg ha–1 irrespective of the element) and increased sharply during the second year (Fig. 2E). They remained roughly stable between 2 years of age and the end of the stand rotation. Senescent leaves contained more than 75 % of total returns of nutrients to the soil from litterfall, regardless of the element (Table 6). Even if dead branches and bark represent a non-negligible part of the litterfall dry matter at the end of the stand rotation (approx. 1000 and 600 kg ha–1 per year, respectively), their nutrient content was much lower than that of leaves.
|
The litter layer was missing during the first 6 months after planting in the savanna soil, but its biomass then increased steadily up to the end of the rotation (Table 7). Whereas a marked accumulation of N and Ca was observed in the forest floor material throughout the stand rotation, accumulation of P, K and Mg remained low (Fig. 2F). The amounts of nutrients released during decomposition of forest floor material were negligible up to the age of 1·5 years, and a slight net immobilization of N was even observed (Table 7). From about 2 years up to the end of the stand rotation, the amounts of nutrients released annually remained stable. They amounted to approx. 13, 2, 2, 18 and 15 kg ha–1 per year for N, P, K, Ca and Mg, respectively.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONLITERATURE CITED |
|---|
At the end of the stand rotation the biomass of leaves was higher at Kondi than at Kissoko, suggesting a better nutritional status of the Kondi stand (Beadle, 1997). Plantations are generally harvested after 7 years for this hybrid in Congo because the mean annual increment then begins to decrease. However, biomass production remained high up to 9 years at Kondi, probably owing to a nutrient availability greater than that in most of the Congolese industrial plantations. Sampling stands of the same clone grown at two sites showed that differences in growth rates between sites affected the values of the fluxes of the biological cycle, in particular for N (Fig. 2). Nevertheless, the chronosequence approach used made it possible to demonstrate clear dynamics of the nutrient fluxes throughout the stand rotation.
Dynamics of the main fluxes of the biological cycle
Despite the paucity of savanna soils and very low fertilizer inputs at planting, the biological cycle of this alien Eucalyptus hybrid was efficient enough to produce high amounts of biomass. The continuous nature of growth of the trees throughout the year in Congo did not greatly modify the dynamics of the biological cycle described for temperate species (Miller, 1981; Baker and Attiwill, 1985; Ranger and Colin-Belgrand, 1996; Ranger et al., 1997). In particular, the two first stages of stand development identified by Attiwill (1979) and Miller (1981) were observed.
Stage 1: early growth period.
This period lasts from about 2 years after planting up to canopy closure. The major proportion of net primary production is used to build the photosynthetic display of the stand. Simultaneously, rapid development of the fine root network in the soil is observed for this clone, leading to a roughly uniform distribution of fine roots to a depth of 3 m at an age of 2 years (Bouillet et al., 2002). This distribution of fine roots suggests that nutrients in deep layers of soil are taken up by the trees during this juvenile phase. High nutrient requirements from the first year after planting onwards are a result of rapid development of the canopy and a low nutrient use efficiency for biomass production at this stage. Therefore, the early growth is strongly dependent upon nutrient availability in the soil. During the second year of growth, recycling processes start with internal translocation of N, P and K during leaf senescence. Nevertheless, nutrient-conserving mechanisms are not completely installed up to canopy closure, since nutrient release during litter decay represents a small amounts of nutrients.
Stage 2: nutrient-conserving mechanisms are established.
Current nutrient requirements remain high up to the end of the stand rotation owing to: (1) the high biomass production up to an age of 9 years; and (2) the average life span of leaves of only 6 months, which leads to high nutrient requirements for the maintenance of foliar biomass.
Internal translocation from old tree components to young developing ones supplies, on average, 38, 30, 51, –13 and 6 % of N, P, K, Ca and Mg requirements, respectively, between the second year after afforestation and harvest. This mechanism leads to a significant reduction in the dependence of the stand on soil N, P and K reserves. This feature is particularly marked for K, since the amounts of this element taken up from soil reserves decrease with time and, in contrast, the proportion of the requirement provided by internal translocation of this element in the trees tends to increase with time. By contrast, the net accumulation of Ca observed in tissues during senescence shows that the stands are totally dependent on the availability of this element in the soil throughout the rotation.
Although the amounts of nutrients taken up from the soil changed only slightly throughout the stand rotation, their availability in the soil was much modified. Whereas annual burnings prevented litter from accumulating at the soil surface in native savannas, the biomass and nutrient content of the forest floor increased steadily over the Eucalyptus rotation. Several indications suggest that forest floor mineralization plays a crucial role in the nutrition of these plantations: (1) trees develop a dense root mat above the mineral soil, adherent to decomposing fragments of the forest floor (Laclau et al., 2001a); (2) the distribution of fine roots is generally concentrated in the top soil at the end of the stand rotation (Bouillet et al., 2002); (3) changes in the chemistry of solutions during their transfer through the forest floor and the soil horizons show that nutrients are taken up very quickly by the vegetation, in the forest floor and in the top soil, and nutrient losses by deep drainage at the end of the stand rotation are very low (Laclau, 2001); and (4) the marked effects of slash management on tree growth and nutrient content in young Eucalyptus plantations has been shown in a field experiment conducted by Nzila et al. (2002).
A similar pattern was observed in some Amazonian rain forests where most of the fine root biomass lies above the mineral soil, in a layer of humus and detritus. A radioisotope study (32P) showed that P was transferred from leaves to living roots through fungal hyphae (Herrera et al., 1978). This direct cycling of nutrients from fallen litter to roots is pronounced in these oligotrophic ecosystems and leads to improved mineral recovery and decreased mineral loss from the ecosystem (St John, 1983).
The amounts of P, K, Ca and Mg released during litter decay between the ages of 6·5 and 8·5 years at Kondi were close to values presented in Table 7 for the chronosequence at Kissoko (Laclau, 2001). Nevertheless, the amounts of N released were higher, on average 46 kg ha–1 per year. As the nutrient content of the forest floor was assessed more extensively at Kondi than at Kissoko, an underestimation of the amounts of N released during litter decay in Table 7 cannot be excluded. In the present study, the amounts of Ca and Mg released during litter decay supplied approx. 40 % of current requirements from the age of 2 years onwards, and therefore represent a key process in the ecosystem, increasing the efficiency of the limited pool of these nutrients to produce biomass.
Nutrients involved in forest growth
Cumulating fluxes of the biological cycle during the rotation highlight the real amounts of nutrients involved in tree nutrition and also the influence of forest stands in soil nutrient dynamics (Table 8). For a production of 92 t ha–1 ligneous aerial dry matter in the 7-year-old stand of the chronosequence, immobilization in the ligneous components of the trees amounted to 235 kg ha–1 N, 47 kg ha–1 P, 59 kg ha–1 K, 68 kg ha–1 Ca and 49 kg ha–1 Mg. Cumulative requirements throughout the stand rotation were approx. four times higher than immobilization in the ligneous components of the trees for N, K, Ca, Mg, and twice as high for P. Compared with values given in the literature for adult Eucalyptus plantations throughout the world, the P use efficiency for this clone in Congo is low (Laclau et al., 2000). This might be a result of the relatively high reserves of this nutrient in the soil of this area, confirmed by the lack of response of trees to P input (Bouillet et al., 2001).
|
Permanent immobilization of nutrients in the ligneous components of the trees much lower than total requirements has been shown for numerous forest species (e.g. Miller, 1981; Dambrine et al., 1991; Ranger and Colin-Belgrand, 1996; Ranger et al., 1997). Little immobilization of nutrients in forest crops probably led to the well-known concept of forest tree frugality, and concealed the much higher nutrient requirements of forest trees, as shown here for Eucalyptus plantations. Cumulative requirements over the stand rotation represented about 10, 8, 130, 30 and 140 % of the reserves of total N, available P and exchangeable K, Ca and Mg to a depth of 3 m, respectively (Table 1). Even if deeper roots had been found, the density of fine roots was very low beyond a depth of 3 m in the 6-year-old stand sampled at Kondi (Laclau et al., 2001b). Moreover, the reserves of nutrients actually available are probably lower because: (1) most of the N capital in the B horizon is bound to an organic matter resistant to degradation, dated from 14C measurements at 8000 years BP (Trouvé et al., 1994); and (2) reserves are calculated for a considerable volume of soil, whereas we observed that large areas of soil are not explored by fine roots (Laclau et al., 2001b). The high proportion of cumulative requirements over the stand rotation compared with available stocks in the soil clearly shows that soil fertility in these plantations is a dynamic parameter: current stand requirements are mainly provided by small amounts of nutrients circulating quickly in the ecosystem.
The distinctive features of these plantations are: (1) their fast initial growth; and (2) the short length of rotations. High nutrient requirements from the first year of growth show that the biomass production of these stands is highly dependent on the availability of nutrients in the soil after afforestation, and later on the efficiency of the biological cycle. Indeed, the relative contribution of N, P and K uptake from the soil to total requirements decreased during stand development (Table 8) as a result of increased internal translocation of these elements. For Ca and Mg, the contribution of translocation remained very low during the stand rotation. Moreover, the amounts of nutrients returning to the soil with litterfall represented an increasing proportion of total requirements with stand age. A deterioration of litter decay processes, due to unsuitable practices (burning, for example), might significantly reduce nutrient availability and depress stand growth at the end of the rotation.
Consequences of the biological cycle for soil properties
The influence of vegetation on soil pedogenesis is well documented (e.g. Duchaufour, 1972). In the case of afforestation with Eucalyptus hybrids in Congolese savanna soils, large amounts of organic matter containing nutrients taken up in the deep layers of soil are deposited at the soil surface. Other effects on soil biota (Mboukou-Kimbatsa et al., 1998) and hydrological processes, with the development of a strong water repellency at the soil surface (Laclau et al., 2001b), have been observed.
The cumulative returns to the soil of organic matter (OM) with litterfall represent approx. 30 t ha–1 during stand rotation, and logging practices bring about 10 t ha–1 of slash (bark, small branches, leaves) to the soil surface every 7 years. Moreover, it has been shown for other species that fine root turnover, which was not taken into account in this study, usually represents several tonnes per hectare and per year of OM (Hendricks et al., 1993). Organic matter is a basic component of the soil’s biological, physical and chemical fertility, particularly in tropical environments where the clay fraction usually comprises mainly kaolinite (Nambiar et al., 2000). The importance of OM in these soils is a result of its functions as (1) a reservoir of nutrients released by mineralization processes; (2) physico-chemical support for cation fixation; and (3) soil aggregation conditioning drainage and soil aeration.
In the Pointe-Noire region of Congo, the OM content in the soil is very low (about 1 % in the upper 5 cm) and large returns of OM to the soil in Eucalyptus plantation are likely to have a positive effect on soil fertility. A trend of OM increase in the topsoil was observed in old planted crops (up to 19 years old) that had never been harvested (Trouvé et al., 1994). However, a paired comparison of OM contents in the upper layer of soil (0–20 cm) between savanna and adjacent plots where Eucalyptus stands have been cultivated for 18 years suggested that C contents remain roughly stable when stands are harvested every 7 years (Bouillet et al., 2001). Activated mineralization processes of OM after harvesting might account for this pattern (Carlyle, 1993).
The cumulative returns of N, P, K, Ca and Mg to the soil with litterfall during the stand rotation represent about 240, 20, 20, 120 and 100 kg ha–1, respectively. The amounts returning to the soil with slash every 7 years are of the same order of magnitude for P and K, but are approx. three times lower for N, Ca and Mg (Laclau et al., 2000). The return to the soil surface of large amounts of nutrients taken up in different soil layers counteracts the leaching processes that transfer nutrients downwards. Stands sampled here belong to the first rotation after afforestation in a savanna soil, but repeated harvests are likely to deplete the nutrient stocks in the soil through biomass removal and loss of nutrients by leaching during the early growth of trees. Despite these large amounts of nutrients that are returned to the soil with litterfall, field trials and input–output budgets of nutrients in these plantations consistently show that Eucalyptus stands benefit from a N fertility inherited from the savanna ecosystem, and that fertilizer inputs (essentially N) will have to be increased over successive rotations (Bouillet et al., 2001; Laclau, 2001). This trend was confirmed by comparison of 31 pairs of soil samples from savanna and adjacent Eucalyptus stands. A significant decrease (P < 0·05) in N content in the upper layer of soil (0–20 cm) was observed after 18 years of cultivation, and there was a significant increase in the C : N ratio (Bouillet et al., 2001). This result is also consistent with radioisotope studies, which demonstrated that savanna OM in the topsoil is progressively replaced by Eucalyptus OM, with a higher C : N ratio (Trouvé et al., 1994).
The large amounts of nutrients released during litter decay in these plantations emphasize how quickly the biota present in this savanna soil has adapted to changes in the nature of the organic matter following the introduction of this alien hybrid. Despite the unfavourable chemical properties of Eucalyptus OM (phenolic compounds, tannins, allelopathic effect) (Bernhard-Reversat, 1999), the ability of soil biota to mineralize this OM quickly plays a fundamental role in the rapid growth of the stands. Indeed, litter decay releases a substantial proportion of nutrients required for biomass production from the age of 2 years.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONLITERATURE CITED |
|---|
The dynamics of the main fluxes of the biological cycle of nutrients have been quantified throughout the first rotation after afforestation in a Congolese savanna. These monoclonal Eucalyptus plantations are an informative model because the biases due to chronosequence approaches are limited (same plant material, soils roughly homogeneous, short rotation with little climatic variation, etc).
This study made it possible to distinguish an early growth period, during which most of the nutrient requirements are supplied by uptake from soil reserves, followed by a second phase from canopy closure, characterized by intense recycling processes within the ecosystem (Fig. 3). During this second phase, internal translocation of nutrients in trees provided about 50 % of the annual requirements of K, and 30 % of those of N and P, but represented very low amounts of Ca and Mg. Litterfall returned to the soil about a quarter of the annual amounts of P and K taken up, and half of the uptake of N, Ca and Mg. In these soils with very low reserves of available nutrients, mineralization of OM is therefore a key mechanism, providing a substantial proportion of the current requirements of trees.
|
Although Eucalyptus species are exotic in Congo, trees quickly develop an efficient growth strategy. Nutrient-conserving mechanisms lead to an intense recycling within the ecosystem and a high efficiency of small pools of nutrients circulating quickly to produce biomass. A major challenge for the sustainable management of these plantations will be to maintain these reserves of available nutrients in the soil over successive rotations, through silvicultural practices. The short-rotation silviculture carried out in this area will impose fertilizer inputs matching not only the requirements throughout the stand rotation, but also the nutrient availability in the soil, essentially linked to the organic matter dynamics.
|
ACKNOWLEDGEMENTS |
|---|
We acknowledge Jean-Claude Mazoumbou and Antoine Kinana for collecting field samples and for stem ring analysis. We thank Laurent Veysseyre (IRD) and Gisèle Heral-Llimous (CIRAD) for performing the mineral analysis. We also thank our colleagues from the breeding programme of UR2PI/CIRAD, who set up the field trials used in this age series: Jean-Marc Bouvet, Raphaël Gouma, Nicodème Kimbouma, Aubin Saya and Philippe Vigneron. We are grateful to the founders of UR2PI, Republique du Congo, CIRAD-Forêt and ECO s.a. for their financial support, as well as to the Plantation Programme of CIFOR, which provided funds for this study.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONLITERATURE CITED |
|---|
-
Attiwill PM. 1979. Nutrient cycling in a Eucalyptus obliqua (L’Hérit.) forest. III Growth, biomass, and net primary production. Australian Journal of Botany 27: 439–458. [CrossRef][Web of Science]
Baker TG, Attiwill PM. 1985. Above-ground nutrient distribution and cycling in Pinus radiata D. Don and Eucalyptus obliqua L’hérit. forests in southeastern Australia. Forest Ecology and Management 13: 41–52.
Bargali SS, Singh RP, Singh SP. 1992. Structure and function of an age series of eucalypt plantations in central Himalaya. II. Nutrient dynamics. Annals of Botany 69: 413–421.
Beadle CL. 1997. Dynamics of leaf and canopy development. In: Nambiar EKS, Brown AG, eds. Management of soil, nutrients and water in tropical plantations forests. Canberra, Australia: ACIAR Monograph 43, 169–212.
Bernhard-Reversat F. 1999. The leaching of Eucalyptus hybrids and Acacia auriculiformis leaf litter: laboratory experiments on early decomposition and ecological implications in Congolese tree plantations. Applied Soil Ecology 12: 251–261. [CrossRef][Web of Science]
Bouillet JP, Laclau JP, Arnaud M, Thongo A. 2002. Changes with age in the spatial localization of roots of Eucalyptus clone in Congo. Impact on water and nutrient uptake. Forest Ecology and Management 171: 43–57. [CrossRef][Web of Science]
Bouillet JP, Safou-Matondo R, Laclau JP, Ranger J. 2001. Changes in N-P-K fertilisation over successive rotations in Eucalypt plantations in the Congo. Proceedings of the IUFRO Congress: developing the Eucalypt of the Future. Valdivia, Chile: IUFRO.
Carlyle JC. 1993. Organic carbon in forested sandy soils: properties, processes, and the impact of forest management. New Zealand Journal of Forest Science 23: 390–402.
Dambrine E, Le Goaster S, Ranger J. 1991. Croissance et nutrition minérale d’un peuplement d’épicéa sur sol pauvre. II. Prélèvement racinaire et translocation d’éléments minéraux au cours de la croissance. Acta Oecologia 12: 791–808.
Delwaulle JC. 1988. Plantations clonales au Congo. Point des recherches sur le choix des clones dix ans après les premières plantations. Proceedings of the IUFRO Meeting: genetic improvement of tropical trees, Pattaya, Thailand. Pattaya, Thailand: IUFRO.
Duchaufour P. 1972. Processus de formation des sols. Biochimie et géochimie. Nancy: Université de Nancy, Centre de pédologie.
Gonçalves JLM, Barros NF, Nambiar EKS, Novais RF. 1997. Soil and stand management for short rotation plantations. In: Nambiar EKS, Brown AG, eds. Management of soil, nutrients and water in tropical plantations forests. Canberra, Australia; ACIAR Monograph 43: 379–417.
Hansen K. 1994. Throughfall and canopy interactions in spruce forest. Forskningsserien 8. Lyngby, Denmark: Danish Forest and Landscape Research Institute.
Helmisaari HS. 1992. Nutrient retranslocation in three Pinus sylvestris stands. Forest Ecology and Management 51: 347–367. [CrossRef][Web of Science]
Hendricks JJ, Nadelhoffer KJ, Aber JD. 1993. Assessing the role of fine roots in carbon and nutrient cycling. Tree 8: 174–178.
Herrera R, Merida T, Stark N, Jordan CF. 1978. Direct phosphorus transfer from leaf litter to roots. Naturwissenschaften 65: 208–209. [CrossRef][Web of Science]
Laclau JP. 2001. Dynamique du fonctionnement minéral d’une plantation d’Eucalyptus. Effets du reboisement sur un sol de savane du littoral congolais; conséquences pour la gestion des plantations indus trielles. PhD Thesis, Institut National Agronomique Paris-Grignon, Paris, France.
Laclau JP, Bouillet JP, Ranger J. 2000. Dynamics of biomass and nutrient accumulation in a clonal plantation of Eucalyptus in Congo. Forest Ecology and Management 128: 181–196. [CrossRef][Web of Science]
Laclau JP, Arnaud M, Bouillet JP, Ranger J. 2001a. Spatial distribution of Eucalyptus roots in a deep sandy soil in the Congo: relationships with the ability of the stand to take up water and nutrients. Tree Physiology 21: 129–136. [Abstract]
Laclau JP, Bouillet JP, Ranger J, Joffre R, Gouma R, Saya A. 2001b. Dynamics of nutrient translocation in stemwood across an age series of Eucalyptus hybrid. Annals of Botany 88: 1079–1092.
Laclau JP, Ranger J, Bouillet JP, Nzila JD, Deleporte P. 2003. Nutrient cycling in a clonal stand of Eucalyptus and an adjacent savanna ecosystem in Congo. 1. Chemical composition of rainfall, throughfall and stemflow solutions. Forest Ecology and Management 176: 105–119. [CrossRef][Web of Science]
Mboukou-Kimbatsa IMC, Bernhard-Reversat F, Loumeto JJ. 1998. Change in soil macrofauna and vegetation when fast-growing trees are planted on savanna soils. Forest Ecology and Management 110: 1–12. [CrossRef][Web of Science]
Miller HG. 1981. Nutrient cycles in forest plantations, their change with age and the consequence for fertilizer practice. Proceedings of the ‘Australian forest nutrition workshop: productivity in perpetuity’, August 1981, Canberra, Australia, 187–199.
Nambiar EKS, Tiarks A, Cossalter C, Ranger J. 2000. Site management and productivity in tropical plantation forests: a progress report. Bogor, Indonesia: Center for International Forestry Research.
Nzila JD, Bouillet JP, Laclau JP, Ranger J. 2002. Slash management effects on the tree growth and nutrient cycling in young Eucalyptus replanted sites in the Congo. Forest Ecology and Management 171: 209–221. [CrossRef][Web of Science]
Parrotta JA. 1999. Productivity, nutrient cycling, and succession in single- and mixed-species plantations of Casuarina equisetifolia, Eucalyptus robusta, and Leucaena Leucocephala in Puerto Rico. Forest Ecology and Management 124: 45–77. [CrossRef][Web of Science]
Polglase PJ, Attiwill PM. 1992. Nitrogen and phosphorus cycling in relation to stand age of Eucalyptus regnans F. Muell. I. Return from plant to soil in litterfall. Plant and Soil 142: 157–166. [CrossRef][Web of Science]
Ranger J, Colin-Belgrand M. 1996. Nutrient dynamics of chestnut tree (Castanea sativa Mill.) coppice stands. Forest Ecology and Management 86: 259–277. [CrossRef][Web of Science]
Ranger J, Marques R, Colin-Belgrand M. 1997. Nutrient dynamics during the development of a Douglas-fir (Pseudotsuga menziesii Mirb.) stand. Acta Oecologia 18: 73–90. [CrossRef]
St John TV. 1983. Response of tree roots to decomposing organic matter in two lowland amazonian rain forests. Canadian Journal of Forest Research 13: 346–349. [CrossRef]
Switzer GL, Nelson LE. 1972. Nutrient accumulation and cycling in Lobolly Pine (Pinus taeda L.) plantations ecosystems: the first twenty years. Soil Science Society of America Proceedings 36: 143–147. [Web of Science]
Trouvé C, Mariotti A, Schwartz D. 1994. Soil organic carbon dynamics under Eucalyptus and Pinus planted on savannas in the Congo. Soil Biology and Biochemistry 26: 287–295. [CrossRef]
Turner J, Lambert MJ. 1983. Nutrient cycling within a 27-year-old Eucalyptus grandis plantation in New South Wales. Forest Ecology and Management 6: 155–168.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J.-P. Laclau, J. C.R. Almeida, J. L. M. Goncalves, L. Saint-Andre, M. Ventura, J. Ranger, R. M. Moreira, and Y. Nouvellon Influence of nitrogen and potassium fertilization on leaf lifespan and allocation of above-ground growth in Eucalyptus plantations Tree Physiol, January 1, 2009; 29(1): 111 - 124. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-P. LACLAU, F. TOUTAIN, A. T. M'BOU, M. ARNAUD, R. JOFFRE, and J. RANGER The Function of the Superficial Root Mat in the Biogeochemical Cycles of Nutrients in Congolese Eucalyptus Plantations Ann. Bot., March 1, 2004; 93(3): 249 - 261. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||