Keywords

4.1 Introduction

Increased demand for forest products around the world has contributed to the growth of planted forest areas over the last few decades (FAO 2015). The annual growth rate of planted forests in the tropics has been 2.5% per year, corresponding to an increase of around 20 million hectares from 1990 to 2015 (Payn et al. 2015). In Brazil, the planted forest generates and offers a huge diversity of products, especially pulp, paper, charcoal, sawn wood, and plywood, among others. Brazil currently has 7.84 million hectares of planted trees and its planted tree industry is responsible for 91% of the wood produced for industrial purposes in the country and 6.2% of the Brazilian gross domestic product; it is one of the industries with the greatest potential to help build a green economy (IBA 2017). Most of these planted forests are monocultures of Eucalyptus spp. (72%) and Pinus spp. (20%), but other species can also attend the internal and external market, such as Acacia mangium and Acacia mearnsii that together occupy 2.0% of the planted forest area. Other species include Hevea brasiliensis (2.9%), Schizolobium amazonicun (1.1%), and Tectona grandis (1.1%) (IBA 2017).

Despite the success of the Brazilian forestry agribusiness, most of these forest plantations occupy extensive areas of marginal soils, previously managed soils with low fertility or at some stage of degradation (Gonçalves et al. 2013). The sustainability of these plantations could be compromised in the medium and long terms, because they have been managed under short rotations and a fertilization regime, often far from high nutrient exports due to timber harvests, especially for P, K, and N (Santana et al. 2008; Laclau et al. 2010; Bouillet et al. 2013; Gonçalves et al. 2013).

Soil organic matter (SOM) in tropical planted or native forests has many functions that affect soil properties and processes. It plays a role of the nutrient reserve, and energy source for animals, plants, and microorganisms since its soil minerals are not great suppliers of mineral nutrients due to the high degree of weathering (Sanchez 1979; Maquere et al. 2008; Quesada et al. 2010). In addition, SOM features the greatest cation exchange capacity (CEC) of these soils (Senesi and Loffredo 1999; Oorts et al. 2000; Motta et al. 2002), which prevents essential nutrients from being easily lost by leaching and, consequently, not utilized by forests. Instead, the SOM allows such nutrients to be exchanged easily with the soil solution and to be taken up by plants. SOM participates in soil aggregation and consequently in soil pore space formation (Tisdall and Oades 1982; Denef et al. 2001; Denef and Six 2005). With an expressive and balanced pore space (i.e., macro- and micropores) the roots exploit the soil profile much better, water infiltration is facilitated, and plants get benefit from increased soil water storage (Doran and Parkin 1994; Franzluebbers 2002).

Poorly managed planted forests stimulate decomposition of SOM and C loss. This poor management includes excessive use of machines, with high traffic intensity in soil preparation or harvesting, in addition to removing or burning crop residues from the harvested area and excessive soil disturbance during preparation, which comprises the provision of numerous soil ecosystem services and the sustainability of these forests (Chaer and Tótola 2007; Dominati et al. 2010; Gonçalves et al. 2013; Jesus et al. 2015; and Chap. 10).

The rates of SOM and litter decomposition are clearly influenced by many factors, including temperature, air humidity, soil moisture content, soil microbial community, and litter quality. The last one may be described by lignin, polyphenols, carbon, and nutrient concentrations (especially N and P) or ratios of these (Attiwill and Adams 1993; Hättenschwiler et al. 2011). The N2-fixing species often have higher N concentrations and decompose more rapidly. Thus, mixtures of Eucalyptus litter with more readily decomposable and more nutrient-rich litter may enhance the litter decomposition rates (Briones and Ineson 1996). Gartner and Cardon (2004) in a meta-analysis found synergistic effects in the majority of mixed-litter decomposition studies (47.5%), while antagonistic effects (19.1%) were less frequent. However, some expected improvements due to the introduction of N2-fixing trees in litter decomposition, such as microbial activity and nutrient release (especially N), seem to be conditioned by the structural quality of the residues, the N:P ratio of litter, and the integration with microorganisms, driven by the increase in diversity of the plant community (Forrester et al. 2006; Rachid et al. 2015; Santos et al. 2018).

This chapter presents and discusses some data about changes in key soil processes, litter decomposition, and C stabilization (i.e., humification) arising from the introduction of N2-fixing legumes in mixed Eucalyptus spp. plantations. We must emphasize the dynamics of C in tropical soils, especially Brazilian soils, and of the understanding of the main drivers of C stock in mixed-forest soil. Many researchers believe that one of the legacies of mixed plantations is the increase of soil C stocks, and at the end of this chapter, some of the practices and challenges to overcome obstacles of C storage will be introduced.

4.2 Carbon Assimilation and Partitioning in Forest Plantation

Understandings of carbon dynamics in forests involve knowledge of the biotic and abiotic factors that interfere in the growth of species that colonize a particular site, as well as where assimilated C is allocated (Cannell 1989; Davidson and Hirsch 2001). According to Cannell (1989), the CO2 conversion efficiency in forest biomass is variable and associated with factors such as earth geometry, geographic location of the plantations, CO2 diffusion rate for chloroplasts, canopy interception capacity (Io), respiration rate, and proportion of the different aboveground tree compartments.

Carbon accumulation in a forest occurs in several aboveground and belowground compartments. In the case of woody tissues such as stem, thick roots, and branches, this accumulation may last for years or decades. On the other hand, in labile tissues such as leaves, flowers, and fine roots, after the senescence C will return to the atmosphere in days or weeks via decomposition (Landsberg and Gower 1997; Fearnside 2000; Schlesinger and Lichter 2001; Nouvellon et al. 2012; Bachega et al. 2016).

In general, about half of all CO2 annually fixed in terrestrial biomass via photosynthesis (~120 Pg C y−1) is respired by plants (~60 Pg C y−1) and the other half is respired heterotrophically (Janzen 2004). In other words, if not disturbed, the C reservoirs shall remain constant in these environments, including the soil compartment. Because soil is the largest active C compartment of the terrestrial ecosystem (~1500–2000 Pg) and is in direct connection with other environmental components (i.e., atmosphere, hydrosphere, pedosphere, and biosphere), land use and climate changes have led to the loss of soil capacity to provide varied ecosystem services associated to SOM loss.

The net primary production (NPP) of forests increases with the reduction of latitude because there is a higher incidence of global radiation, higher evapotranspiration, and, consequently, higher cloud formation and rainfall in tropical regions (Ometto 1981). Reducing the vapor saturation deficit in these regions increases stomatal conductance and rate of carbon fixation of plant species (Sands and Mulligan 1990; Novais and Barros 1997; Whitehead 1998). Recent papers with Eucalyptus in Brazil have shown that NPP of the genus is determined preferentially by local water availability and, to a lesser extent, by soil fertility (Reis et al. 1985; Stape et al. 2004, 2010; Rigatto et al. 2005; Balieiro et al. 2008). The importance of water supply for Eucalyptus grandis and E. urophylla was highlighted by Stape et al. (2004), when evaluating 14 sites with a significant productivity gradient (9.7–39.1 Mg ha−1 year−1). According to the authors, the NPP of the sites with intermediate productivity (i.e., average of 16 Mg ha−1 year−1) was 46% higher than NPP of the sites with low productivity (i.e., average of 11.9 Mg ha−1 year−1), where 32% of this variation was related to the site of water supply. Although the productivity difference between the high and medium productivity sites was 72%, one-third of this difference was related to the availability of soil water.

Similarly, when measured by the average monthly increment of the stem, the productivity of eucalypt was directly related to precipitation in the Rio Doce Basin, in Minas Gerais (Souza et al. 2006). According to the authors, for each increase of 100 mm in the total precipitation within a year, there was an average increase of 0.45 m3 ha−1 month−1, while 100 mm reduction affects this increase in 0.64 m3 ha−1 month−1. Rigatto et al. (2005) also found very high correlations between the height of Pinus taeda plants and the available water in eight different forest sites. Stape et al. (2010) across a large edaphoclimatic gradient (~1000 km) and eight regions found that fertilization beyond the current operational rates did not increase the growth of clonal Eucalyptus plantations, whereas irrigation raised growth about 30% (to 30.6 Mg ha−1 year−1).

At mixed-forest plantations in Brazilian and Congolese conditions, where the set of trials with Eucalyptus and Acacia mangium (Acacia) were developed, the complementary interactions in the mixed stands of acacia and eucalypt led to significant biomass production. However, the NPP was higher than in monocultures without N fertilization only where appropriate climate conditions (i.e., temperature, high humidity, and rainfall) for Acacia and poor soil occur (Bouillet et al. 2013; Santos et al. 2016; Voigtlaender et al. 2019). Under milder and drier climate, Acacia cannot compete with Eucalyptus. For E50:A50 arrangement, and comparing six different sites (five in Brazil and one in Congo), Santos et al. (2016) after Bouillet et al. (2013) found that stemwood production only exceeded E100 ones at the Congo and Seropédica sites, although the differences in both studies were not significant at the 5% level (Santos et al. 2016).

Using the C budget approach to quantify growth, C uptake, and C partitioning in pure and mixed plantations with the same stocking density of E. grandis and A. mangium plantations, Nouvellon et al. (2012) developed an interesting work with C allocation of mixed plantation in São Paulo state. According to the results, the production in mixed plantation is lower than in Eucalyptus due to the lower gross primary production and net primary production values, as well as shifts in C allocation from above- to belowground and from growth to litter production. However, the pattern seems to be site specific due to the two contrasting tropical site environments (i.e., Brazilian and Congolese). Epron et al. (2013) compared Brazilian and Congolese site, and found that mixed-species plantations at the Brazilian site had a lower stand of wood biomass and aboveground net primary production (ANPP) without change in total belowground C fluxes (TBCF). In contrast, the mixed-species plantations overyielded the monocultures at the Congolese site, which led to higher standing wood biomass at the harvest. The NPP partitioning of the mixed plantations shifted towards aboveground growth at Kissoko (Congo site) and towards belowground growth at Itatinga (Brazil).

4.3 Litterfall and Nutrient Deposition in Pure and Mixed-Forest Plantations

As Chap. 3 explores this subject, we briefly discuss the main nutrient cycling alterations during the presence of N2-fixing trees in Eucalyptus plantations.

In general, the amount of litter deposited by a given species or forest follows the same pattern of NPP (Fraser et al. 2015). Studies of litterfall are essential for the sustainability of planted forests since litter management significantly changes the biogeochemistry of these ecosystems and has already known consequences on plant growth, soil quality, and climate (Bowen and Nambiar 1984; Bonan 2008; Ponge 2013; Berg 2018). The C accumulation pattern and partitioning in the different tissues of plants, including the litterfall per unit area, depend on the density, arrangement, and interaction between the species (Reis et al. 1985; Leite et al. 1998; Nouvellon et al. 2012; Epron et al. 2013; Laclau et al. 2013).

The amount and quality of the material deposited on forest soils are related to not only genetics (i.e., planted species), structural, and aging factors of the plantation, but also climatic variables and soil type (Bernhard Reversat 1996; Landsberg and Gower 1997; Laclau et al. 2010; Voigtlaender et al. 2019). Soil fertility and climate significantly shape these two variables (i.e., quantity and quality of leaf litter) (Reis and Barros 1990; Negi and Sharma 1996; Stape et al. 2010). In soils with low fertility, it is natural for species to use nutrients more efficiently than nutrient-richer areas (Novais and Barros 1997; Malhi et al. 2006; Laclau et al. 2010). Therefore, it is common to find more nutrients being recycled in litter in soils of better fertility or without nutritional limitation. However, if there is water deficit during forest development, the water deficit induces stomata closure and, consequently, reduction in carbon fixation; that is, the plant will not succeed to express its potential and efficiency of nutrient use (Novais and Barros 1997; Epron et al. 2009; Stape et al. 2010).

Based on some studies about N deposition via litterfall in pure and mixed plantations, it was found that the litterfall in N2-fixing leguminous plantation is 65% higher (in average) than that in the monocultures of Eucalyptus (without N). In addition, combining such legumes with Eucalyptus can promote significant increase, even 42% higher, which evidences the N input increase with the introduction of a leguminous in a mixed plantation (Fig. 4.1).

Fig. 4.1
figure 1

Scatterplot between litterfall dry mass and N aboveground litterfall (Kg ha−1) among N2-fixing, non-N2-fixing, and mixed plantations. Data from n = 53 selected studies (see Tables 4.1 and 4.2). The black line represents the mean regression line independent of plantations, that is, the overall regression line (R2 = 0.60, p < 0.001). The regression line for red (circle) for non-N2-fixing species, blue (square) for mixed plantations (E50:A50), and green (triangle) for N2-fixing species are given. The shadows represent the standard error around lines for non-N2-fixing, N2-fixing, and mixed plantations only

Full size image

The presence of A. mangium in adult mixed plantations (i.e., >5 years) of five sites in Brazil also intensified the contribution of N via litterfall (Table 4.1), although the deposition pattern was different in the sites analyzed. In four sites, Acacia showed higher deposition in comparison to Eucalyptus without N fertilization, while the other one showed similar deposition. This trend correlated with the climatic adaptation of the A. mangium.

Table 4.1 Litterfall mass and nutrient deposition associated (kg ha−1 ano−1) under pure and mixed-forest plantations of N2-fixing and non-N2-fixing trees in different countries around the world*
Full size table

In Brazilian sandy soils, the addition of 120 kg ha−1 of N via fertilizer in the Eucalyptus monoculture increased the N via litterfall around 10 kg ha−1 above the monoculture without N fertilization (final phase of the first rotation). On the other hand, in mixed plantations with half of the plant density replaced by Acacia mangium (E50:A50) and double density (E100:A100), the contributions were 110 and 79 kg ha−1 higher than in the monocultures without N fertilization, respectively. In younger plantations, the contribution of N inputs via litterfall was lower. But in the first and second rotations and up to 33 months, the contributions for pure Acacia or mixed plantations are significantly higher than in the Eucalyptus monocultures (without N fertilization) (Santos et al. 2016; Tchichelle et al. 2017).

Most studies referred in Table 4.1 do not provide information about deposition of other nutrients, so it is not possible to imply whether mixed plantations improve the contributions of these elements. On the other hand, since litterfall in mixed plantations is superior to monocultures, it is believed that the overall contribution is higher, as observed for N (Fig. 4.2a). The variability of P and N deposition data (Fig. 4.2) is mostly due to the evaluated species, age of plantations, and variations in the soil types studied that present different mineralogy and fertility, as well as different patterns of growth and accumulation in plants.

Fig. 4.2
figure 2

Box plots of N (a) and P (b) in aboveground litterfall (Kg ha−1) under N2-fixing and non-N2-fixing monocultures and mixed plantations of both. Data from n = 58 selected studies (see Tables 4.1 and 4.2). Black points represent outliers

Full size image

4.4 Litter Decomposition Under Pure and Mixed-Forest Plantation

The logic of non-N2-fixing and N2-fixing species’ mixed plantations is to promote forest sustainability through complementarity and competitive reduction interactions among species. The “extra” supply of N provided by legume trees via biological nitrogen fixation and its redistribution offer throughout the system an intensification of leaf litter decomposition, N mineralization, and transferring between fixing and non-fixing species. Collectively, they are key processes to improve the growth and yield of the mixed forests, eliminating or using less N fertilizers (Forrester et al. 2006; Kaye et al. 2000; Koutika and Richardson 2019).

In general, decomposition of forest residues is driven by the litter chemical attributes, environmental conditions, and surface area/volume ratio of the residue (Landsberg and Gower 1997; Gholz et al. 2000; Laclau et al. 2010). Although the N deposition is high in pure and mixed plantations with N2-fixing trees (Fig. 4.1), this chemical characteristic is not a guarantee of higher litter decomposition, especially under tropical conditions. For instance, A. mangium in sandy and loamy textured soils shows lower leaf and fine root decomposition in comparison to Eucalyptus, although it has high N leaf concentration (Balieiro et al. 2004; Bachega et al. 2016; Doughty et al. 2018). Besides, A. mangium has high internal cycling of P (Balieiro et al. 2004; Bachega et al. 2016; Doughty et al. 2018), providing a litter with low P concentration. Furthermore, A. mangium has a more recalcitrant leaf litter, lignin rich with low contents of nonstructural carbohydrates or low-molecular-weight phenols and P concentration (Santos et al. 2017). Thus, the microbiota faces a harsh condition to decompose its residues. The concept of “decomposer starvation,” proposed by Hättenschwiler et al. (2011), reveals a syndrome of poor C litter in tropical rainforests which could be applied to pure A. mangium plantations in oligotrophic soils (e.g., Arenosol). According to these authors “in the neotropical rainforest, natural selection favored a leaf litter trait that leads to starvation-inhibition of decomposers, thereby increasing the tree ability to compete for the uptake of highly limiting nutrients, P in particular, via mycorrhizal associations.” Other authors who have also observed a decrease of P in the soil in areas in mixed plantations claim that this decrease comes from the P uptake by Eucalyptus to maintain the N:P stoichiometry of their leaves (Koutika et al. 2014).

It is worth noting that legume trees have a different demand for P as a function of symbiosis with diazotrophic bacteria (Giller and Cadisch 1995; Vadez et al. 1995). Therefore, different strategies are required for capturing this element, such as acidification of the rhizosphere that improves the solubilization of less soluble forms of P (Raven et al. 1990) and root exudation of acid phosphatases and organic acids that leads to PO4 desorption from the soil matrix and improves P availability (Vance 2001; Venterink 2011). These strategies are paramount in tropical soils since they are mainly able to fix P by ligand exchange and occupation of P sorption sites (Bhatti et al. 1998; de Campos et al. 2016). This is possibly due to the high content of aluminum and iron oxides in the Oxisols and Ultisols typically found in moist tropical sites (Leal and Velloso 1973; Lloyd et al. 2001).

Land-use history may also affect the litter decomposition. Under Ferralsols, previously managed with Eucalyptus (for 60 years), Bachega et al. (2016) detected that early decomposition of leaves and fine roots of Acacia mangium was markedly slower than that of Eucalyptus residues, despite higher N and P concentrations in both tissues of Acacia (respectively, 1.9 and 1.5 times higher for leaves and 2.9 and 3.3 times for roots). The lower values of C:N and C:P ratios were associated to litter decomposition rates of Acacia, but the authors did not confirm the home field advantage (HFA) that states which plants create a specialization of local decomposer communities of their litter. For Acacia, authors claim that the time since the start of the first rotation was not long enough to allow the decomposers to become specialized for its residues. In contrast, several researches have demonstrated that in the short term the litter and soil bacterial and fungi communities are very specific for both (Rachid 2013; Rachid et al. 2015; Bini et al. 2013).

As was pointed out in the perspective of the HFA theory, our research group studied the litter leaf decomposition of both species (i.e., Acacia and Eucalyptus) in a sandy soil. Our team observed a distinct HFA for decomposition litter from Acacia and Eucalyptus, with Acacia decomposing in the home stands faster than under Eucalyptus stands. In contrast, litter from Eucalyptus decomposed faster under Acacia stands. Additionally, the litter of each material showed very distinct fungal communities and did not change in function of time, and the local of incubation did not influence the microbial community (Rachid 2013). Higher diversity and lower dominance of fungi were reported in litter from Acacia and mixed plantations during the decomposition period (180d), regardless of the place at which they were established (Rachid 2013). By using infrared spectroscopy, it was found that Eucalyptus litter during the decomposition under Acacia stands remain more proteinaceous material than the initial residue, and the migration of N to Eucalyptus litter may be considered. On the other hand, the Acacia litter did not change its chemical composition during the incubation (Novotny et al. 2013). These results demonstrate the importance of biological N2 fixed for the Eucalyptus litter decomposition, and suggests that the most diverse fungi community is essential for mixed litter decomposition allowing N mobility. Nonetheless, it is interesting to note that the higher P concentration in Eucalyptus litter acts simultaneously and synergistically with the decomposition process in the mixed plantation (Santos et al. 2017). In addition, the C:N ratio alone seems not to be the main predictor of soil organic and litter decomposition (Cotrufo et al. 2013; Lehmann and Kleber 2015; Berg 2018). Some predictors such as N:P, lignin:N, and lignin:P, when possible, should be analyzed collectively.

The N:P ratio of the litter trait is often cited as an essential drive of litter decomposition (Güsewell 2004; Bakker et al. 2011). However, differences in the N:P ratio observed in aboveground biomass and leaf litter usually reflect even more significant changes in the available N:P ratio of soil (Güsewell 2004). Figure 4.3 demonstrates the broad range of the N:P ratio (mass) of the litter in the papers analyzed and, consequently, about edaphic conditions of the studies. Also, there is a trend towards increasing this relation in mixed plantations: non-N2-fixing (N:P–14) < N2 fixing (N:P–25) < mixed (N:P–32). Although the contribution of N via litterfall is higher for leguminous plantations, followed by mixed plantings compared to Eucalyptus plantations, for P there is a lower recycling trend in mixed planting conditions (Fig. 4.2b). This finding corroborates the productivity of these plantations and the accumulation of P in the aboveground biomass of the species, besides the edaphic limitations in the P supply. In other words, although P is being absorbed more efficiently in these plantations, the plants begin to drain this P for internal use. This process decreases the availability of P for the soil in the medium and long terms (Sanchez 1979; Binkley et al. 2000) and requires attentive management with phosphate fertilization in future rotations.

Fig. 4.3
figure 3

Box plots of N:P relationship in aboveground litterfall (Kg ha−1) among N2-fixing, non-N2-fixing, and mixed plantations. Data from n = 58 selected studies (see Tables 4.1 and 4.2). Black points represent outliers

Full size image

Climate change and especially temperature rise may also affect the chemical characteristics of leaf litter with effects on decomposition. Trees growing in high-temperature environments have a reduction in the N content of the leaves and an increase in the rate of carboxylation, which results in increased structural and nonstructural carbon levels in the leaves (e.g., glucose, sucrose, fructose) (Güsewell 2004; Pandey et al. 2015) and a recalcitrant leaf litter. These effects can be intensified in the tropics, where the soils are acidic and poor (i.e., with minerals with high adsorption capacity of P). In addition, the presence of leguminous species with high internal P recycling, such as A. mangium (Balieiro et al. 2005; Inagaki et al. 2011; Santos et al. 2017), may impair the decomposition process with unfavorable stoichiometric ratios in the leaf litter (e.g., C:N, C:P, or N:P) (Güsewell 2004). Therefore, changes in leaf litter quality may be reflected by changes in soil enzymatic activity (Fanin and Bertrand 2016; Santos et al. 2017), stabilization of C stocks (Fisk et al. 2015; Castellano et al. 2015), emission of CO2 through microbial respiration (Zhou et al. 2013, 2015), and nutrient mineralization rates (e.g., P, N, and S) (Marklein et al. 2016; van Huysen et al. 2016). All these aspects are feedbacks of climate change and productivity of forest systems (i.e., plantations or natural forests) (Bonan 2008).

4.5 Are Soil Carbon Stocks Really Higher in Mixed Plantations than in Monocultures?

In general, soils present several mechanisms related to the protection of organic matter that is associated with chemical complexation (e.g., polymerization, humification, organic synthesis, and organo-mineral interaction) and physical protection (e.g., complexation with mineral fractions and degree of aggregation) (Feller and Beare 1997; Sollins et al. 1996; Roscoe and Machado 2002). In planted forests, crop residue management, machine traffic, planting structure, climate, and planted and understory species all were identified as determinants for nutrient dynamics and degree of C storage in soils (Bernhard Reversat 1996; Binkley et al. 2000; Kaye et al. 2000; Resh et al. 2002; Qiao et al. 2014; Jesus et al. 2015). With the possibility of managing the communities of soil bacteria and fungi in mixed Eucalyptus and A. mangium plantations (Rachid et al. 2013, 2015; Bini et al. 2013), it is also reasonable to think about changes in belowground interactions (i.e., belowground competition for nutrients and water). These consequences are still poorly studied in soil C dynamics, although Sokol and Bradford (2019) have claimed that belowground inputs provided by the microbial formation pathway form mineral-stabilized soil C are more effective than aboveground inputs, partly due to the higher efficiency of formation by the rhizosphere microbial community relative to the bulk soil community. In mixed plantations of A. mangium and Eucalyptus, there might be a more favorable environment for soil C stabilization, because we observed a complementarity in terms of nutrient deposition (e.g., N and P deposition) (Santos et al. 2017, 2018), soil bacterial and fungi communities (Rachid et al. 2013, 2015), and fine root growth (segregated) in soil profile (Silva et al. 2009; Laclau et al. 2013).

The dynamics of soil organic matter from natural and planted forests, including mixed plantations, have been studied from the viewpoint of the “priming effect” by several authors (Resh et al. 2002; Balieiro et al. 2008; Hoosbeek and Scarascia-Mugnozza 2009; Forrester et al. 2013; Koutika et al. 2014). The priming effects were defined by Kuzyakov et al. (2000) as the “strong short-term changes in the turnover of soil organic matter caused by comparatively moderate treatments of the soil.” Such changes might be the input of organic or mineral fertilizer to the soil, exudation of the organic substances by roots, and simple mechanical treatment of soil or its dry and wet cycles. Figure 4.4 presents the schematic representation of the priming effect. Under an ecological point of view, the negative priming effects have a much greater significance than positive ones, but the direction of these changes depends mostly on the nutrient status of the soil and the C:N ratio of the active SOM pool (i.e., labile organic matter) (Kuzyakov et al. 2000).

Fig. 4.4
figure 4

Schematization of the priming effect—nonadditive interactions between decomposition of the added substrate and of soil organic matter (SOM): (+) acceleration of SOM decomposition—positive effect; (−) retardation of SOM decomposition—negative priming effect (adapted from Kuzyakov et al. (2000))

Full size image

4.5.1 Soil Carbon Stocks in Pure Plantations of Eucalyptus

The impacts of Eucalyptus on soil carbon stocks vary according to many factors, including the land-use history, previous crop, post-logging residue management, climate, and spatial variability of soil attributes (Forrester et al. 2006; Chaer and Tótola 2007; Balieiro et al. 2008; Gonçalves et al. 2013). Under a broader point of view, in a recent meta-analysis, Fialho and Zinn (2014) compiled data on the organic soil C stocks using 50 observations for depths between 0 and 20 cm and 39 between 0 and 40 cm of studies in Brazil in paired plots (i.e., plantations and natural forests) aiming to evaluate the impact of native vegetation conversion on plantations of Eucalyptus. The authors verified that, on average, the net effect of the conversion is null; that is, it does not damage the original C stocks of the soil, although losses and gains are related to local site conditions (Table 4.2). The authors conclude that this null effect, even after considering the rotation time, texture, and biome, suggests that other factors may control the direction and intensity of changes in soil C stocks in Eucalyptus plantations and point to the productivity, techniques of soil preparation, soil type, and management as essential factors in this evaluation.

Table 4.2 Changes in soil C stocks (Mg ha−1) under plantations of Eucalyptus spp. in comparison to native vegetation as a function of rotation, soil texture, and biome evaluated
Full size table

Cook et al. (2016) studied the effects of Eucalyptus plantations on soil carbon stocks, 0–30 cm deep, over two decades, in 306 operational eucalypt plantations across a 1200 km gradient in Brazil. The study included two tropical states (Bahia and Espírito Santo) and one subtropical state (São Paulo), and resulted in the findings that the size and rates of change in soil C stocks were due to different factors. These factors include the history of the site, soil order, clay content, seasonal precipitation (especially dry season), and mean annual temperature. In general, across all sites, the soil C showed a slight decrease (−0.22 ± 0.05 Mg ha−1 year−1) from the original sampling that ranged in approximately 3–4 rotations, but in subtropical regions the stocks remained the same (0.06 Mg ha−1 year−1).

Maquere et al. (2008) studied the impact of different land uses (i.e., savanna, pasture, and Eucalyptus saligna plantations) and a management (i.e., 60 years under short rotation vs. 60 years under continuous growth) on soil carbon and nitrogen stocks. The authors found significant soil carbon increases (approximately 25%) with Eucalyptus under short rotation management when compared to Cerrado native vegetation, whereas soil carbon stocks in the continuous forest plantation increased by 15% in relation to Cerrado vegetation. In the same biome and in degraded pastures Lima et al. (2006) reported that afforestation of former degraded pasture land leads to increased C storage in the soil in the short term (30 years). They observed carbon sequestration rates up to 0.57 Mg C ha−1 year−1 by Eucalyptus afforestation.

Finally, it is evident in the studies that pure Eucalyptus plantations present a potential for soil C increase, especially if preceded by degraded pastures. However, local abiotic conditions and forest management seem to be the significant constraints to the effective increase of C in soil. On the other hand, the introduction of legume trees associated with diazotrophic bacteria to Eucalyptus monocultures has gained expression in recent years. The increase of soil C stocks so far had always been related to these plantations but, as will be seen below, this phenomenon cannot be generalized, mainly for the Brazilian tropics.

4.5.2 Soil C Stocks Under Mixed Plantations of N2-Fixing and Non-N2-Fixing Plantations

Several authors reported higher soil C accumulation of N2-fixing species over non-N2-fixing ones (Binkley et al. 1992; Kaye et al. 2000; Resh et al. 2002). However, when a small or no variation is detected in soil C stocks, changes in soil morphological and molecular levels of soil organic matter have also been observed (Kindel et al. 2003; Voigtlaender 2012; Koutika et al. 2014; Santana et al. 2015; Voigtlaender et al. 2019), and this may have consequences for the nutrient cycling in these plantations.

Resh et al. (2002), using isotopic techniques, detected that 55% of the soil C enriching the N2-fixing plantations came from the retention of native C, in a comparison between the C soil stocks of N2-fixing species, as Albizia falcataria, Leucaena leucocephala, and Casuarina equisetifolia, and the soil C stocks of Eucalyptus saligna in four tropical soils, two Andisols and one Vertisol and Entisol. They reported that the native C stock was derived from old sugarcane plantations and pastures with metabolic assimilation of C4, while the other 45% came from C stabilized and derived from recently allocated C3 from the trees, showing a clear negative priming effect. In addition, 62% of the explanation of the stabilization of native C (C4) occurred due to the accumulation of N in the soil of the legume trees and Casuarina plantations. Very similar results were reported by Kaye et al. (2000) in Andisols from Hawaii, where Albizia monocultures had 2.3 Mg ha−1 more N and 20 Mg ha−1 more C than Eucalyptus monocultures in the 50 cm topsoil. In Acrisols (Brown Dermasols) from Australia, Forrester et al. (2013) reported a difference greater than 15 Mg ha−1 between N2-fixing and non-N2-fixing trees in the first 30 cm topsoil. However, in Arenosols from Congo and Brazil, a lower gain in soil carbon storage was detected in the N2-fixing Acacia mangium plantations at 7 and 5 years after planting in comparison to monocultive of Eucalyptus, respectively (Koutika et al. 2014; Rocha et al. 2019). In the Pseudosamanea guachapele plantations, the C storage was even much lower than that found under Eucalyptus grandis plantations in southeastern Brazil, in the same soils (Balieiro et al. 2008), a fact that was justified by the authors by high contribution of N (~250 kg ha−1 y−1) and other nutrients cycled through litterfall, resulting in a rapid decomposition of the litter and SOM. In is worth mentioning here that the pure legume tree stands studied in Brazil (Acacia mangium and Pseudosamanea guachapele) had no similar growth in comparison to Eucalyptus, contrary to the plantations of Falcataria moluccana (Albizia falcataria) in Hawaii that present higher aboveground biomass than Eucalyptus saligna (Kaye et al. 2000).

Climate and soil characteristics can control soil C stocks, but the aboveground biomass and NPP are also drivers of C stocks in various environments (Lal 2005; Cornwell et al. 2008; Qiao et al. 2014; Nottingham et al. 2015; Lange et al. 2015). With the literature data consulted, it is not possible to state that C stocks are a function of aboveground biomass production since the soil C augmentation pattern is not very different between monocultures and mixed ones.

Under mixed plantations and Brazilian conditions, mixed plantations presented aboveground biomass higher than monocultures only in sites with poor soils (for Acacia mangium and Pseudosamanea guachapele) (Santos et al. 2017; Balieiro et al. 2010). However, despite the soil C stocks in N2-fixing plantations did not differ statistically from Eucalyptus monocultures (without N) (Balieiro et al. 2008; Rocha et al. 2019), under mixed plantations were detected higher rates of C sequestration (up to 1.44 Mg ha−1 y−1), in comparison to Eucalyptus monocultives (without N fertilization), demonstrating that for oligotrophic soils the mixed plantation is an great alternative to improve soil quality.

In Australia, Forrester et al. (2013) studied mixed plantations of Eucalyptus globulus and Acacia mearnsii. They detected a linear relationship between aboveground production and soil C, but not at an N2-fixing proportion, as found by Kaye et al. (2000), who observed higher aboveground biomass production in mixed plantations and lower in monocultures. In both papers, the increase in C soil stocks in mixed plantations is justified by the negative priming effect of the mixed litter, that is, the maintenance of the C originated from previous crops (sugarcane) and by the addition and stabilization of C derivatives from the trees.

A significant contrast found among these plantations seems to be the pattern of C allocation in the belowground biomass, because Hawaiian monocultures accumulated more C in the belowground than mixed plantations, contrary to what was observed in Australia (Forrester et al. 2013; Binkley and Ryan 1998). In Hawaii, intraspecific competition intensified the use of soil resources, while interspecific competition in Australia further pressured soil resources, culminating in belowground investments to compensate for distinct constraints. In monocultures, the accumulation patterns of the species can lead to the storage for the aboveground biomass of critical elements for soil C storage; in contrast, and because of the complementarity of niches and facilitation (in both directions), mixed plantations can more efficiently cycle nutrients and better use available water. As previously mentioned (Sokol and Bradford 2019) and pointed out by other authors (Schmidt et al. 2011; Nouvellon et al. 2012), the change in the pattern of C allocation in plantations may be due to the large C stock drive to the soil.

The differential input of N by N2-fixing species appears not to be the major reason for increasing soil C storage (Forrester et al. 2013). The complementarity of niches (above- and belowground) (Tilman 1999; Tilman et al. 2001; Forrester et al. 2006), soil microbial sharing and increased fungal richness (Rachid et al. 2013, 2015), and stimulus to grow and microbial activity (Bini et al. 2013; Pereira et al. 2017; Santos et al. 2017) seem to prove that belowground dynamics commands soil C storage (Sokol and Bradford 2019), but studies in this sense need to be performed for mixed plantations.

The difference in the magnitude of C stocks of soils has a great association with mineralogy and texture (Feller and Beare 1997; Hassink 1997), hence the variations presented in Table 4.3. In general, Andisols from Hawaii have high specific surface phyllosilicate clays and naturally higher SOM contents due to the complexes formed with Al (i.e., noncrystalline Al hydroxide and Al-insoluble organic complexes) and allophanes (Feller and Beare 1997). Even in soils with a recognized SOM stability, the biologically fixed N input was responsible for a significant increase of the C and N stocks, showing the benefit of the N2-fixing species in that environment, which presents a nonexistent water deficit (i.e., 4500 mm of annual precipitation). Thus, the feedback effect of the N input and NPP (Kaye et al. 2000), associated with the fact that N favors, in consonance with high fertility and humification of soil organic matter (Kirkby et al. 2011) justifies such results for Hawaii. On the other hand, the Ferralsols, in which the Brazilian plantations were established, are highly weathered and acidic and have low fertility, which contrast with the high physical aggregate stability that confers protection of soil C (Feller and Beare 1997; Silva et al. 2013). Thus, even under tropical conditions, plantings associated with N2-fixing species can store significant amounts of C in the soil. It is interesting to note that even under a single soil class (Ferralsols), large variations in soil C stocks occur in plantations, due to variations in soil clay content and climate (Voigtlaender 2012; Bouillet et al. 2013; Voigtlaender et al. 2019). The Arenosols of Brazil and Congo, from which the main results of this class originated, are in turn naturally lower in amounts of C due to accelerated biochemical dynamics and reduced SOM protection, as well as poor structure and aggregation (Feller and Beare 1997; Zinn et al. 2002). However, it is a fact that even among the Arenosols, small variations in the clay contents of these soils can confer significant changes in the C stocks, as well as capping of the quartz grains by kaolinite and hematite or goethite type clays that increase the formation possibilities of organo-mineral complexes and C storage of these soils (Scheidegger et al. 1993; Donagemma et al. 2008).

Table 4.3 Soil C stocks (Mg ha−1) in pure and mixed plantations of eucalyptus and N-fixing trees
Full size table

The storage of C in the soil of mixed plantations is also due to the rate of decomposition of the deposited residues, which determines how fast the C is incorporated into the mineral phase. By comparing monocultures, Kaye et al. (2000), Resh et al. (2002), Forrester et al. (2013), Koutika et al. (2014), and Pereira et al. (2017) detected greater incorporation of C in soil under N2-fixing species. However, considering the literature on mixed plantations the C stock in soil was higher but do not show differences in relation to monocultures of Eucalyptus and legume trees (Fig. 4.5) due to higher variability of data. In addition, C stocks occur preferentially in the more labile compartments of SOM fractions (Bini et al. 2013; Koutika and Mareschal 2017; Pereira et al. 2017).

Fig. 4.5
figure 5

Mean (●) and SD (whiskers) of carbon stock differences (%) of N2-fixing and mixed plantations relative to non-N2-fixing species plantations (Eucalyptus) (n = 29)

Full size image

These authors compared the C content of the microbial biomass as well as C and N of the soil organic fraction (physically fractionated between 2000 and 75 μm) and found that, in the four treatments, Eucalyptus without N fertilization, Eucalyptus + N fertilization, Eucalyptus + Acacia mangium (E50:A50), or pure A. mangium plantations, they differed and were significantly higher in plantings of Acacia or in mixed plantations, confirming positive changes in microbial indicators and increases in concentration and nutrient cycling in Ferralsols. When analyzing Table 4.3, it is possible to detect that, with the exception of allophanic soils, the storage of C in poor soils follows the N:P ratio up to a certain limit. This implies that the management of these plantations in oxidic soils with high P adsorption capacity (Lloyd et al. 2001) or sandy soil with low SOM and P reserves should require, in the short and medium terms, a special attention to the management of P fertilizer.

4.6 How to Improve the Soil C Stocks in Mixed-Species Plantations Under Tropical Conditions?

Based on the reviewed literature and data collected, it can be suggested that soil C stocks in planted forests can be increased with improved management practices to overcome environmental restraints and with a broader logic of available biological resources.

Planted forests in the tropics and mainly in Brazil occupy marginal soils, with a history of intense use. Thus, these soils present low fertility, high acidity, Al saturation, and adsorption capacity of P and S (Leal and Velloso 1973; Motta et al. 2002). However, huge differences in these attributes occur within the tropical climatic zone (Sanche 1997), and may partially justify changes in species responses under monoculture and mixed conditions described in this chapter. Although the responses of the forest plantations to liming and fertilization of the soil are small, the non-replacement of the exported nutrients leads to a decrease in the nutritional capital of the soil and the future commitment of these forests. The focus of foresters shall be on the construction of soil fertility under planted forests, with the adoption of practices and technologies that potentiate the above- and belowground growth of species. For the P case specifically, although most are in the non-labile form in soil, considered reserves exist in labile and moderately labile fractions (up to 7.7 and 15.5 Tg, respectively) (Withers et al. 2018) in areas under higher fertilizer regimes. Especially when combined with no till, correction of acidity, and adoption of the 4R principles of nutrient management (right rate, right time, right place, and right form) (INPI 2012) the stocks of soil natural capital related migth be at least maintained.

In Brazilian Eucalyptus plantations, accumulation in the aboveground biomass of N, P, K, Ca, and Mg in 100 Mg is in the range of 238–298 kg for N, 16–29 kg for P, 123–236 kg for K, 176–590 kg for Ca, and 40–92 kg for Mg. From these totals, 35% of N, 29% of P, 36% of K, and 21% of Ca and Mg are potentially exportable by wood (Santana et al. 2008). In the average 7-year rotations for Eucalyptus, Gonçalves et al. (2013) recommend up to 2 Mg ha−1 of lime, 60–80 kg ha−1 of N, 60–80 kg ha−1 of P2O5, and 140–160 kg ha−1 of K2O. For trace elements around 1–5 kg ha−1 of B are applied, depending on the local water deficit, and 1 kg ha−1 of Cu and Zn. Roughly, the natural nutritional capital of the soil (Sanche 1997; Dominati et al. 2010) seems to be compromised, with detrimental consequences to the organic and inorganic nutrient reserves of these forests.

Managing mixed plantings or rotations with legumes that associate with bacteria that fix N2 from the atmosphere will lead to substantial changes in fertilizer management given the changes in biogeochemical cycles of nutrients and differentiated nutrient exports (Chap. 3). Legumes have a differentiated demand for P in relation to non-N2-fixing species (Vadez et al. 1995; Inagaki et al. 2009, Venterink 2011). Acacia mangium, the species most studied under Brazilian conditions, presents high absorption of P in the seedling and adult phase (Inagaki et al. 2011; Santos et al. 2017), and it is very efficient in recycling internally this absorbed P, so that it deposits litter with a high N:P ratio (Inagaki et al. 2010; Santos et al. 2018). Since plants and microorganisms compete strongly for soil P, affecting key soil biological processes and growth of tropical forests (Wieder et al. 2009; Hättenschwiler et al. 2011), strategies to reduce species competition for soil P or to increase complementarity in recycling due to P deposition should be pursued jointly with the adoption of management practices that favor the permanence of crop residues on the soil and SOM.

The recommendations of crop fertilization in Brazil are based on response curves of crop production (shoot biomass or stem in the case of Eucalyptus or its productivity) in relation to the applied fertilizer dose or its availability (i.e., concentration) in the soil (Cantarutti et al. 2007). Little or no attention has been given to changes in the C allocated to roots by planted forests and their relationship to fertility management and soil C stocks. Based on current hypotheses that the belowground inputs provided by the microbial formation pathway form mineral-stabilized soil C more efficiently than aboveground inputs partly due to the greater efficiency of formation by the rhizosphere microbial community relative to the bulk soil community (Sokol and Bradford 2019), it can be considered that, by managing the soil chemical limitations (i.e., acidity, P, N, S, Zn concentration, among others), there will be higher plant growth and consequently root growth, which together will increase soil C stocks. Recent works (Kirkby et al. 2011, 2014) also demonstrate the possibility of altering the potential for soil C accumulation and sequestration of soils from the introduction of nutrients to the soil, correcting the stoichiometry among C, N, P, and S of the heavy fraction of organic matter (i.e., humidified). The formation of a “new-fine fraction of soil organic matter,” the most stable C component in soil, increased threefold by increasing the residues with supplementary nutrients, which in other words implies to say that we can manipulate the nutrients in favor of C sequestration and restoration of the fertility. In other words, although Eucalyptus or Acacia (or other leguminous species) is tolerant to acidity and high saturation of Al, it does not mean that the practice of soil fertilizing and correction shall not be stimulated. On the other hand, gains in productivity in oxidic soils under the correct management of soil fertility are always accompanied by increased biomass and microbial activity (i.e., basal respiration and enzymatic activity) and, consequently, increases in SOM levels.

4.7 Final Considerations

As shown in this chapter, it seems clear that the establishment of mixed plantations of Eucalyptus and Acacia mangium represents an alternative to increase soil C stocks in marginal lands. However, the success of these plantations is not the guarantee to increase C stocks. It is necessary to take into account the climatic and soil conditions of the site for the intercropped species, to perform conservationist practices of soil preparation, residue management, and conservative harvesting practices. Better results were obtained where exist appropriated climate condition for Acacia and oligotrophic soils (e.g. sandy soils).

Finding a new set of other species (including N2-fixing and non-N2-fixing) that have ecological and economic interest seems to be the main future challenge, although this is not such a simple task. It could begin through a list of priority legume tree species with litter traits that, together with eucalypt, can promote the increase of the global stand biomass production (and also belowground) and improvements in soil properties and functions. It is interesting to note that C sequestration is only an ecosystem service that the soil can provide and that the mix of non-N2-fixing with N2-fixing trees may provide many other ecosystem services, such as flood mitigation; greenhouse gas regulation; filtration and recycling of nutrients; and biodiversity preservation among others. In Chap. 10, these services are presented and discussed.

New researches should also elucidate under which arrangements and trees density could provide positive ecological interactions (complementarity and competitive reduction) that can increase the biomass production (above- and belowground) and accelerate litter decomposition and soil C sequestration.

Finally, due to the close relationship between the N and C cycle, it has been reported that as a result of the presence of N2-fixing trees in Eucalyptus plantations positive changes in soil N stocks are more pronounced than for C (Bernhard Reversat 1996; Voigtlaender 2012; Rachid et al. 2013; Voigtlaender et al. 2019; Rocha et al. 2019). These general results confirm that the presence of legumes in Eucalyptus plantation might contribute to reducing the need for mineral N fertilization in the long term and increasing the soil nutritional capital. New mixtures of species and arrangements of mixed plantations in combination with nutrient management could enhance C sequestration and produce a more stable organic matter (humus).