Green Manure as an Alternative for Soil Recovery in a Bauxite Mining Environment in Southeast Brazil

Abstract The objective of this study was to evaluate plant cover and dry mass produced by green manures and their influence on the chemical attributes of soil for the ecological restoration of an area around bauxite mine environment in Southeast Brazil. Soil preparation, chemical fertilization and sowing of three species of green manures were carried out: Cajanus cajan, Crotalaria juncea and Stylosanthes guianensis. In general, C. cajan and C. juncea showed the best performance in plant cover. C. cajan and S. guianensis produced more phytomass. Green manures were efficient to promote rapid plant cover and increase in soil organic matter and total-N contents in all treatments. Therefore, green manure is a viable alternative that contributes to the success of ecological restoration projects in areas in the bauxite mining environment.


INTRODUCTION AND OBJECTIVES
Bauxite is an important ore, from which aluminum, the third most abundant element on the Earth's crust (USGS, 2022).This mineral is found in high quantity and quality (>40% Al 2 O 3 ) in Brazilian reserves (ANM, 2021;USGS, 2022).For this reason, Brazil has a prominent position in the production and exploration of mineral reserves existing in its territory (MME, 2021).In this sense, mining has great relevance to the Brazilian economy, the mineral production in 2020 contributed to the Brazilian GDP with an estimated gross value of USD 43.7 billion (ANM 2021).Mineral extraction activities accounted for about 2.4% of the country's GDP and promotes regional socio-economic development by generating direct and indirect jobs (ANM, 2021;USGS, 2022).
The mineral exploration of bauxite occurs in the topsoil and superficial outcrops, ranging from one a to three months and occurs in small areas (Martins et al., 2020).Thus the recovery of mined areas and areas under the influence of bauxite mining is relatively rapid (Fonseca, 2021).Generally, after bauxite exploration, mining companies carry out topographic reconfiguration, transposition of the topsoil that was removed and stored before the activities, soil preparation, fertilization, and fencing of the area (Balestrin et al., 2019;Melo & Sánchez, 2020;Balestrin et al., 2020;Martins et al., 2020).In addition, to ensure sustainability, mining companies carry out restoration actions both in mined areas and in compensation areas, aiming at environmental recovery, either through active restoration techniques such as seedling planting, direct seeding, green manure and nucleation or by natural regeneration (Martins, 2018;Martins et al.,2020;Cosimo et al., 2021;Fonseca, 2021;Martins et al., 2022).
In this context, plant cover is essential to mitigate the impacts of water erosivity and soil erodibility, especially in areas under the influence of mining (Ferreira et al., 2021).The main negative impacts of bauxite mining are the removal of topsoil and exposure of soil layers devoid of organic matter and nutrients (Anache et al., 2017;Hou et al., 2021).Soil degradation causes disruption of ecosystem functioning and decline of biodiversity worldwide, especially in tropical regions (Butchart et al., 2010;Kaiser-Bunbury et al., 2017).In this sense, as each area presents different levels of impacts, it is essential to adapt ecological restoration projects to the specific needs of each environment (Stuble et al., 2017;Martins, 2018;Villa et al., 2022).Thus, it is important that the restoration techniques also accelerate plant cover in an economically viable way and in the shortest possible time (Gann et al., 2019;Ribeiro et al 2019;Martins et al., 2022).
Green manure is a potential alternative for ecological restoration of ecosystems, given the numerous physical, chemical, and biological benefits in soil properties (Martins et al., 2015;Vásquez-Castro et al., 2020) and its contribution to climate change mitigation through carbon sequestration (Delgado et al., 2021).The ecological restoration techniques after mining disturbance are essential to ecosystem functioning recovery and stability (i.e., soil fertility, erosion mitigation, plant invasion reduction).Thus, green manures stand out for promoting rapid plant cover and protecting the soil from invasive plants (Reis et al., 2019;Melander et al., 2020;Santana et al., 2020), as well as attenuating erosion processes and increasing aggregate stability, ensuring improvements in soil water infiltration (Bonini & Alves, 2011;Ma et al., 2021).In addition, green manures can carry out symbiotic relationships with nitrogen-fixing bacteria and, therefore, can increase this nutrient in the soil, thus allowing a reduction in expenses with nitrogen fertilization.Furthermore, due to significant amounts of phytomass, green manures increase the soil organic matter content (Gao et al., 2018;Silva et al., 2022).Consequently, there is a reduction in acidity, increased cation exchange capacity, and improvements in soil microclimate and nutrient cycling (Isernhagen et al., 2014).In addition, green manures can establish interactions with fauna, facilitating the succession process and contributing to increased diversity (Marshall & Lynch, 2020).
Therefore, this study aimed to evaluate the plant cover and dry mass produced by three species of green manures and their influence on the chemical attributes changes in an area around bauxite mine environment.Thus, the hypothesis of this study is that green manure can contribute to the success of restoration projects and contribute to the sustainability of bauxite mining.

Field procedures
The three green manure species sown in January 2017 were Cajanus cajan L. Millsp.(Pigeon pea), Crotalaria juncea L. (Sunn hemp), and Stylosanthes guianensis Aubl.Sw. (Stylo).The view of the soil surface after the removal of the built structures and the earthworks, the advancement of the cover and the management of the sown green manures can be seen in Figure 2.
Ten 1 m x 1 m plots were allocated to evaluate the cover potential of each green manure species after 18 months of sowing.Thus, three treatments, with ten repetitions each, distributed in a completely randomized design, were tested.Treatment T1 -C.cajan (CAJ); o T2 -C.juncea (CRO) and T3 -S.guianensis (STY).
The plant cover rate was obtained through the visual evaluation of the plots through the percentage scale in which the grade 0 corresponded to the absence of vegetation cover and the grade 100 corresponds to the complete plant cover.In addition, phytomass production was evaluated by collecting all aboveground plant material using a 0.25 m² template in each plot.The harvested material was stored in plastic bags identified which were transported to the Forest Restoration Laboratory (LARF/UFV), where it was dried in an oven at 72ºC for 48 hours and the dry weight was measured.
For the soil chemical analysis, soil samples (0-20 cm depth) were collected with a Dutch auger from 30 random sampling points within each treatment.These samples were combined into one composite sample per treatment.To evaluate and compare the chemical properties under the three green manures, soil samples were collected at 0, 18, and 24 months after sowing.
The composite samples were sent to the Soil Laboratory of the Federal University of Viçosa for proper analyses.Subsequently, the results were interpreted according to the Recommendation Guide for Correctives and Fertilizers of Minas Gerais (Alvarez et al., 1999) and compared with soil analyses before implementing restoration techniques.

Data processing and analyses
To compare mean values of plant cover and dry weight obtained in the three treatments one-way ANOVA was used followed by Tukey test for independent samples (p<0.05)(Crawley, 2013).The R software version 3.5.3(R Core Team, 2019) was used for the statistical analyses.

Soil Analyses
According to the Recommendation Guide for Correctives and Fertilizers of Minas Gerais (Alvarez et al., 1999), the soil samples collected before the soil preparation and fertilization 5 -12 techniques (month 0) indicated an average acidification level (pH=5.39),very low organic matter content (OM=0.25 dag/ kg), low base sum (SB=0.82cmolc/dm 3 ), low cation exchange capacity (CEC=2.32cmolc/dm 3 ) and low base saturation (V=35.3%).Thus, to evaluate, relating, and comparing the soil's chemical properties under the three green manures at different times (0, 18 and 24 months after sowing), Table 2 was prepared.
Liming effectively raised soil pH and increased the concentration of calcium (Ca 2+ ) and magnesium (Mg 2+ ) nutrients.Phosphating increased phosphorus (P) concentration in the soil.The sum of bases (SB) and base saturation (V%) increased after the actions taken, while the exchangeable acidity decreased.There was an increase in total-N of the soil in all treatments after 18 months of sowing the green manures, in the soil grown with Crotalaria juncea and Cajanus cajan there was an increase of 245% and in the soil grown with Stylosanthes guianensis a rise of 177%, about the values obtained before the implementation of the restoration project, reflecting the contribution of the three green manures through biological nitrogen fixation.
Accordingly, there was an improvement in the other chemical properties of the soil, according to criteria proposed by Alvarez et al., (1999), presented in Table 3.In the soil grown with C. juncea there was a medium level of acidification (pH=6.97),low organic matter content (OM=1.88dag/ kg), very high sum of bases (SB=6.21cmolc/dm 3 ), medium capacity for cation exchange (CEC=6.81cmolc/dm 3 ) and a very high base saturation (V=91.2%).Although organic matter receives a low rating, there was an increase of 652% compared to the previous analysis.
After 24 months of green manure sowing, there was an increase in the contents of nutrients N, P, K, Ca 2+ and Mg 2+ and P-rem in all treatments.The decomposition of the phytomass of green manures promoted an increase in pH, by promoting complexation of H + and Al 3+ with compounds from the vegetable residue, leaving Ca 2+, Mg 2+ and K + freer in solution.This process resulted in an increase in CEC saturation by these basic reaction cations.
There was a weak alkalinity level in all treatments and a very high classification in the properties: phosphorus, calcium, magnesium, SB, V%, and P-rem.As for CEC, all green manures showed medium classification.The soil-grown with C. juncea, had high organic matter content, while soils under stylo and pigeon pea had medium OM content (Table 4).
To favor the growth of planted tree species, green manure management was carried out by mowing without incorporation 18 months after sowing.Consequently, 24 months after sowing, an increase in total-N and organic matter contents is observed.As for the organic matter contents, there was a significant increase in the soil grown with C. juncea (1936%) concerning the initial value obtained in the exposed soil, followed by the soil grown with ground under C. cajan (1204%) and in the soil grown with S. guianensis (1072%).The same pattern was observed in the total-N contents of the soil; there was an increase in all treatments, especially for the soil grown with C. juncea, with an increase of 914% with the initial situation of exposed soil, followed by S. guianensis with an increase in 536% and C. cajan with a 532% increase.

Plant cover
The analysis of variance showed differences between treatments, both for plant cover and the dry weight (p<0.05).The highest percentage of plant cover was observed in Cajanus cajan, with an average of 89%, followed by Crotalaria juncea with 75%.Stylosanthes guianensis had the lowest percentage of cover, with an average of 59.5%, being statistically lower than C. cajan and similar to C. juncea by Tukey's test (Figure 3).

Dry weight of phytomass
The species that produced the highest values of dry phytomass were C. cajan with an average of 853.90 g.m -2 and S. guianensis with 716.18 g.m -2 .C. juncea had a lower dry weight, equivalent to 402.26 g.m -2 , being considered statistically lower than C. cajan, but similar to S. guianensis by Tukey test p≤0.05(Figure 4).

Soil Analyses
Soils in tropical regions are generally weathered, acidic and have low natural fertility (Neto et al., 2021).In this context, liming is a practice commonly used to neutralize soil acidity, provide calcium and magnesium, reduce the toxicity of some elements, increase the cation exchange capacity and, consequently, increase the availability of nutrients for plants (Carmeis-Filho et al., 2017;Bossolani et al., 2020).In this sense, soil preparation techniques, liming, and phosphate fertilization are important for soil recovery and green manure can contribute to this process.The contributions of green manure in increasing soil pH were described by Neto et al. (2021) in a study carried out in Yellow-Red Latosol.In addition, Nascimento et al. (2019) stated that mechanical soil scarification increased the dry mass yield of four cover crops, including Cajanus cajan and Crotalaria juncea.Thus, the phytomass decomposition green manures this study may explain the increase in pH (Table 2), mainly due to the contribution of organic matter.
In this perspective, improving the quality and quantity of soil organic matter (SOM) in tropical soils is essential to increase ecosystem sustainability (Carmeis-Filho et al., 2017).
In the present study, the three species of green manures contributed to the increase in SOM contents.This result corroborates experiments conducted in different regions and with different species of green manures, in which they observed an increase in SOM levels and CEC (Sharma et al., 2017;Oliveira et al., 2017;Gao et al., 2018;Liu et al., 2020).The decomposition of organic matter is a complex and very important ecological process as it directly influences the soil microbiota and nutritional dynamics, mainly carbon and nitrogen (Soares et al., 2020).Given this relevance, the relationship between SOM and nitrogen is the subject of studies in post-bauxite mining areas (Banning et al., 2008) and in other tropical regions (Hicks et al., 2019;Sellan et al., 2020).
Nitrogen is the most dynamic nutrient in tropical soils, found in mineral, organic, ionic and gaseous forms (Amazonas et al., 2011;Kumari & Maiti, 2022).These constant transformations occur quickly in the soil influenced by numerous factors such as soil moisture, temperature, pH, microorganisms, fertility, C/N ratio of residues, N-fixing plants, soil preparation and management, among others (Figueiredo et al., 2019;Rastetter et al., 2020;Rocha et al., 2021).Given this complexity, no methodology addresses all these factors and provides a standard nitrogen availability index (Prezotti & Guarçoni, 2013;Figueiredo et al., 2019).In green manure, leguminous species are the main nitrogen fixators through the symbiotic relationship with soil bacteria (Rhizobium and Bradyrhizobium) (Muller et al., 2021).
Biological N fixation performed by green manures can also provide N for non-legume species.Thus, it is possible that the tree species planted in the area have absorbed part of the N and incorporated it into its biomass, considering that N is one of the nutrients that most limits the growth of tree species in tropical regions (Amazonas et al., 2011;Bossolani et al., 2020).The use of N by planted tree species was also verified by Silva et al. (2016) in studies conducted in the region and in a post-mining bauxite area.The authors have reported that planting seedlings intercropped with C. cajan resulted in a greater increase in basal area, lower mortality, rapid soil recovery, improvements in fertility and greater nitrogen availability due to biological fixation.Saldanha et al. (2017) have highlighted that green manure with C. cajan provides physical benefits in soil conservation.However, it cannot supply the nitrogen demand of the main crop.Thus, considering that biological nitrogen fixation is a complex process (Borges et al., 2016) and that the regions with a predominance of Latosol naturally present soils with low fertility, it is recommended that ecological restoration projects carry out complementary nitrogen fertilization, management of green manures and monitoring of soil fertility periodically (Soares et al., 2020).
The total-N contents after 2 years of sowing of green manures (Cro=0.223dag/kg; Sty=0.140dag/kg; Caj=0.139dag/kg) corroborate what was verified by Balestrin (2018), in an area after 6 years of restoration through conduction of natural regeneration (0.130 dag/kg) and with Balestrin et al. (2019), in an area after 14 years of restoration by planting seedlings (0.139 dag/kg), both studies conducted in areas after bauxite mining by Companhia Brasileira de Alumínio in the same region.However, green manures were efficient in recovering total-N contents in soil in a shorter restoration time than other mined areas.As a result, the total-N contents of the soil increased considerably after 24 months of sowing the green manures (Cro:914%; Sty:536%; Caj:532%) compared to the initial values of the same soil.The superior results observed indicate that C. juncea is a key species for soil recovery in ecological restoration projects in the region and confirms recent studies that highlighted the potential of C. juncea to incorporate nitrogen and improve soil fertility (Barbosa et al., 2020;Yao et al., 2021;Ma et al., 2021), increasing biomass (Subaedah et al., 2016) and restoring topsoil in bauxite mines (Narayanan et al., 2021).
In this sense, the management of green manures after 24 months of sowing has proved to be efficient in the incorporation of organic matter in the soil, considering that this technique favored the decomposition of the phytomass of the green manures and consequently promoted an increase in the fertility of the soil (Silva et al., 2022).In the soil-plant system, Nitrogen sho higher turnover by plant residues or phytomass (Hungria et al., 2015).For example, the phytomass incorporation using green manures can promotes biological nitrogen fixation and increase the total-N levels in the soil by decomposition (Hungria et al., 2015).Thus, green manure is a promising alternative for soil recovery and conservation in impacted areas because of physical, chemical, and biological improvements in soil attributes and ecosystem services.In addition, recent studies indicate that cover crops can contribute to climate change mitigation due to plant cover, erosion reduction and organic carbon fixation through the incorporation of organic matter (Yao et al., 2019;Delgado et al., 2021).It should be noted that the effects provided by green manure vary according to the species used, soil microorganisms, soil and climatic conditions, management and the ecological adaptation of green manure to local conditions (Alcântara et al., 2000).Therefore, the peculiarities of each environment and the intrinsic characteristics of each species are crucial for planning green manure as an alternative for soil recovery in ecological restoration projects.

Plant cover
The fast covering of the soil promoted by green manures ensured greater protection of the soil against solar radiation and the impact of raindrops, thus minimizing erosion processes.Furthermore, this fast covering of the soil can contribute to reducing the infestation of undesirable plants, as indicated by studies in different tropical regions with the species C. cajan (Silva et al., 2015;Nascimento et al., 2019;Souza et al., 2021).Cajanus cajan showed the best results in land cover and dry weight.Thus, it is a key crucial species to maximize the benefits of green manures and be a viable option in ecological restoration projects in areas with exposed soil (Martins, 2018).The plant cover rate promoted by C. cajan in this study (89.0%) was higher than that indicated by Pacheco et al. (2017) for the same species (74.9%) after 156 days after sowing.Furthermore, the authors observed similar cover values similar followed in species of the genus Stylosanthes (55.3%) and lower values in the genus Crotalaria (41.6%-55.6%).This contrast reinforces the need for this study since the results are closely related to edaphic conditions, rainfall, management, experimental conditions and evaluation period.
The effects of green manures C. cajan, C. juncea and S. guianensis on plant cover, soil protection and improvement of physical properties are described in several studies (Long et al., 2017;Martins, 2018;Melander et al.,2020;Soares et al., 2021;Silva et al., 2021;Neto et al., 2021).These authors highlighted that C. cajan, C. juncea and S. guianensis promoted increased fertility and improvements in the aggregate stability index and resistance to root penetration.In addition, these authors reinforce the benefits of intercropping between legumes and grasses, aiming at enhancing the effects on soil recovery.9 -12

Dry weight of phytomass
As for dry weight phytomass, the highest production by C. cajan (853.90 g.m -2 ) verified in this study, corroborates Cavalcante et al. (2012), when compared to six other cover crops, including two species of the genus Crotalaria (C.juncea and C. spectabilis Roth).The production of C. cajan described by the authors (870 g.m -2 ) was similar to this study, although the production of crotalaria was lower (250-300 g.m -2 ) and the material was collected at 129 days.Furthermore, superior results were also found in other Brazilian regions as observed by Carneiro et al. (2008); Xavier et al. (2017) and Nascimento et al. (2019).On the other hand, in the same forest typology of the study area and in Yellow-Red Latosol, lower dry mass values of C. cajan (301 g.m -2 ) were found (Mendonça et al., 2017).As for S. guianensis, lower values (339.18g.m -2 ) were found in research conducted in a degraded area (Godoi et al., 2008).Given these variations, it is necessary to evaluate green manures in different conditions and environments (Alcântara et al., 2000).

CONCLUSIONS
The green manures used in this study were efficient in recovering soil fertility.The increase in nutrient contents, the improvement in soil chemical parameters and the deposition of significant amounts of phytomass indicate that green manuring is a technique capable to promote rapid plant cover and assist in ecological restoration projects in mined areas.
C. cajan and C. juncea stood out for covering the soil, while C. cajan and S. guianensis were more efficient in producing dry mass.All species positively influenced the organic matter content and contributed to the increase in the total-N content of the soil, due to biological nitrogen fixation.C. juncea was the most promising species in recovering soil total-N levels.
Thus, this study adds novelties for revegetation of areas in mining environment, especially for soil reconstruction and is important to demonstrate the sustainability of bauxite mining in the region.Future research may evaluate the effects of green manures in the control of invasive grasses and mitigation of climate change.

Figure 2 .
Figure 2. Initial soil surface condition, growth and management of green manures.(A: exposed soil surface -initial situation; B: Green manures after 6 months of sowing; C: Green manures after 12 months of sowing; D: Green manures, after 18 months of sowing; E: Management of green manures to favor the development of planted seedlings; F: Area after 4 years of implementation of restoration techniques).

Table 1 .
Plant species list planted in the study area.

Table 2 .
Chemical analyses of soil from different treatments in the 0-20cm layer at different times (0, 18 and 24 months after sowing).

Table 3 .
Classification of chemical properties of the soil, 18 months after sowing green manures, according to the recommendations for the use of correctives and fertilizers in Minas Gerais.VL= Very low; L=Low; M=Medium; H= High; VH= Very high.CRO (Crotalaria juncea); STY (Stylosanthes guianensis); CAJ (Cajanus cajan).

Table 4 .
Classification of chemical properties of the soil, 24 months after sowing green manures, according to the recommendations for the use of correctives and fertilizers in Minas Gerais.VL= Very low; L=Low; M=Medium; H= High; VH= Very high.