RADICULAR BIOMASS AND ORGANIC CARBON OF THE SOIL IN FOREST FORMATIONS IN THE SOUTHERN AMAZONIAN MESOREGION

The soils of the Amazon region, despite being under one of the densest forests in the world, are mostly characterized by the low availability of nutrients, with litter being the main route of nutrient entry. The objective of this study was to quantify the biomass of fi ne roots in the dry and rainy seasons of the year, including the organic carbon of the soil, and to compare the results in diff erent study environments. The study was carried out in environments of native forest and reforestation aged over 20 years, located in the municipality of Humaitá – AM state. To assess the root biomass, collections were carried out in two periods of the year: dry and rainy seasons. In each of the study areas, fi ve trenches, 0.40 m deep by 0.40 m wide, were dug manually at depths of 0-5, 5-15, and 15-30 cm. For the organic carbon analysis, soil samples were collected in the form of clods at the same depths. The production of root biomass in the native forest environment occurred more intensely in the rainy season, reaching values of 8.19 t. ha, greater than 3.57 t. ha found in reforestation. The density as a function of the soil volume showed that the highest concentration is found in the fi rst 5 centimeters of depth, diff ering signifi cantly in the 5-15 and 15-30 cm layers for native forest area. The organic carbon of the soil showed signifi cance between the dry and rainy seasons for the natural forest environments and reforestation with genipap.


1.INTRODUCTION
The tropical forests play an important role in the storage and absorption of carbon from the atmosphere, as well as in climate change on a global scale. The survival of this ecosystem and its productivity is mainly attributed to its high plant diversity, composed of native species adapted to the climatic and nutritional conditions of the soil, which, in turn, developed effi cient mechanisms in nutrient cycling over time (Jordan, 1985;Mendes, 2018).
The impacts resulting from forest clearing imply higher levels of erosion and soil compaction, promoting the exhaustion of nutrients from forest ecosystems (Fearnside, 2006). Roots are important in sustaining trees, and fi ne roots are one of the main means of capturing soil resources, and their length and number are indicators of nutrient absorption capacity (Freitas et al., 2008).
Carbon partitioning and its dynamics in plant components can also serve to increase the effi ciency of the cycling process and improve carbon sequestration models (Block et al., 2006). Therefore, knowing the biomass production, as well as the density of the root system and the stock of organic carbon in the soil of forest ecosystems is fundamental for decision-making in order to mitigate possible impacts arising from human activities.
The production of fi ne roots can be aff ected by several factors, such as, by the type of vegetation formation and by species diversity (Finér et al., 2011;Brassard et al., 2013), by rainfall seasonality (Lima et al., 2012), and soil characteristics (Wright et al., 2011).
Root biomass plays an eff ective role in the processes that occur in the soil once its rapid renewal signifi cantly contributes to the addition of organic compounds to the soil, having an important regulatory role in the carbon and nitrogen cycle in forest ecosystems (Menezes et al., 2010). According to Gill and Jackson (2000), the behavior of fi ne roots in relation to temperature and regional precipitation can help identify species that are more sensitive to climate change.
Given the importance of the root system for the maintenance and for the perpetuation of forest species and, above all, given the scarcity of studies focused on the Amazon region, the objective of this study was to quantify and compare, in two periods of the year, the production of root biomass and the soil organic carbon in reforestation areas and native forest in the southern Amazonian mesoregion.

2.MATERIAL AND METHODS
The study was carried out in areas of native forest and reforestation, located in the municipality of Humaitá-AM state, southern Amazonian mesoregion. The municipality of Humaitá is located about 200 km north of the capital of Rondônia state, Porto Velho, and 675 km south of Manaus, capital of Amazonas state. The study areas were divided into four distinct parcels, one of native forest, one of reforestation with Teak (Tectona grandis L.), one of reforestation with Jenipapo (Genipa americana L.), and another of reforestation considered mixed, composed by the species Mahogany (Swietenia macrophylla King.), Andiroba (Carapa guianensis Aubl.), Jenipapo (Genipa americana L.), Teak (Tectona grandis L.) and Sumauma (Ceiba pentandra).
The climate in the region is of the Am type, tropical rainy (monsoon-type rain), according to the Köppen classifi cation, with average air temperature ranging between 25 and 27°C, relative humidity between 85% and 90% and average annual precipitation of 2,500 mm (Alvares et al., 2014). The region is characterized by two defi ned seasons, the rainy season, which occurs from October to April, concentrating the greatest volume of rain in the year, and the dry season, which occurs from May to September, with months presenting rainfall below 60 mm (Almeida et al., 2015).
Regarding the description of the study areas, the native forest environment is characterized by containing an enormous diversity of fl ora and fauna species, typical of Amazonian environments, maintained by the Brazilian Army (EB), with no presence of anthropic activities. The reforestation environment was implemented in early 2000, replacing a pasture area. In this area, an average of 1.5 tons of limestone per hectare were incorporated, before planting forest species. After approximately three years of reforestation implementation, cattle were introduced, again and sporadically, to control grass regrowth.
For the evaluation of fi ne root biomass (≤ 2 mm) samples were taken in two periods of the year: dry (June 2018) and rainy (January 2019). The sampling was an adaptation of the monolith method described in Bohm (1979). For each of the study areas, fi ve trenches were manually excavated, with dimensions of 0.40 m in depth and 0.40 m in width.
The trenches were opened by adopting a distance of approximately 0.80 m from the base of the plant, collecting samples with the aid of a steel cylinder 10 cm high and 7 cm in diameter, at depths of 0-5 cm, 5-15cm and 15-30cm, with three repetitions per depth, per area. The samples were stored in plastic bags, identifi ed, and taken to the soil laboratory of the Institute of Education, Agriculture and Environment of Humaitá (IEAA).
In separating the roots from the soil, the samples were subjected to manual washing in running water until total removal of the soil, using a set of overlapping sieves with meshes of 2mm, 1mm and 0.500mm, respectively. After washing the eff ective roots (≤ 2 mm), they were taken to a forced air circulation oven for 72 hours. After drying, with the aid of tweezers, they were weighed on an analytical balance of 0.0001 g precision. This expression. (1) was used to calculate dry root density (DRS): Eq.1 Where: DRS = dry root density, in g dm -3 ; MS = root dry mass, in grams; VM = volume of collected monolith, in dm -3 .
To estimate the Fine Root Biomass ≤ 2 mm (BRF) of each study environment, in a given spatial area and over a given period of time, the production of fi ne roots was calculated, disregarding the soil density by expression (2), according to Vogt et al. (1998) and used by Mendes (2018).
Soil sample collections for further analysis of organic carbon were carried out at the same time and periods of the year as the root biomass. In each study area, in the four plots, soil samples with preserved structure in clumps were collected at three diff erent depths, 0-5, 5-15, 15-30 cm. After the soil was shadedried and sieved in a 2 mm sieve, Air Dry Earth (TFSA), chemical were performed.
To determine the pH in water, the soil: the water ratio of 1:2.5 was used, shaking for 1 minute and leaving it to rest for 60 minutes, then reading the pH in a pH meter.
The soil organic carbon (COS) was determined by the Walkley-Black method, modifi ed by Yeomans and Bremner (1988) Where: Corg = Soil organic carbon concentration, in g kg -1 ; DV = Total volume of potassium dichromate solution added in the sample digestion, in ml; VA = Volume of the ammoniacal ferrous sulfate solution used in the sample titration, in ml; VB = Volume of the ammoniacal ferrous sulphate solution used in the titration of the heated blank, in ml; Value 0.003 = Milliequivalents of carbon mass (atomic weight/valence -12/4, divided by 1000); Value 10 = Transformation from % to g kg -1 ; M = Mass of the soil sample, in g.
The data were tabulated and submitted to the normality test of error distributions by the Shapiro-Wilk test and Levene's homogeneity test. Given the requirements for analysis of variance (ANOVA), for signifi cance (p<0.05), the Tukey 5% test was used as a posteriori test in comparisons of means between the respective environments. To compare the means between the collection periods, the T test of paired samples was applied, using the computer application SPSS statistics 23.3.

Climate conditions
In the rainy period (October to April), with April considered the transition month for the dry period, the volume of rainfall occurred in a regular and well distributed way, with an average of approximately 400 mm for the month of January, sustaining one of the outstanding characteristics, not only in the region, but in the Amazon forest as a whole, the regularity of rainfall and high temperatures for most of the year (Figure 1).
In the dry period (May to September), the greatest water defi cit occurred in June, with an average of 3.6 mm, coinciding with the lowest monthly average temperature of 25°C, a fact that is correlated with cold fronts from other regions, producing cold, a very common phenomenon in the region. Another result observed was the precipitation in April with an average of 260 mm, considered atypical for the transition month.

Biomass dynamics
The results obtained corresponding to the dry period (June), at a depth of 0-5 cm, were not signifi cant, it is possible to observe in Table 1 that the values of 3.13 t. ha -1 for NF; 2.78 t. ha -1 for JRE; 1.45 t. ha -1 for MRE and 0.94 t. ha -1 for TRE, they resemble a 5% signifi cance level.
At the depth of 5-15 cm, still for the month of June, the NF environment showed signifi cance for the other environments, with an average value of 1.44 t. ha -1 and coeffi cient of variation around 35.45%, supporting the accuracy of the data. At a depth of 15-30 cm, the production of fi ne root biomass was similar to the previous depth, with emphasis again on the NF environment, with signifi cance for the other environments, with an average value of 1.24 t. ha -1 .
The results corresponding to the collection in January showed a behavior similar to the month of June, with higher biomass production in the 0-5 cm depth, with an emphasis on the NF environment, diff ering signifi cantly from the other environments in the 5-layer layer. 15 with an average of 2.05 t. ha -1 and 2.34 t. ha -1 in the 15-30 cm layer, in the same month.
The production of fi ne root biomass occurred more intensely in the rainy season ( Figure 2), with the exception of RJ which occurred in the dry season, with the NF environment being the one that most contributed to the production of biomass with 8.19 t. ha -1 in the rainy season. The biomass concentration at diff erent depths was somewhat less discrepant in the native forest environment. However, in the reforestation areas the biomass production occurred more strongly in the 0-5 cm layer compared to the forest area native, reaching the highest percentage of 72.18% for JRE in the rainy season. In a work carried out by Witschoreck et al. (2003), under similar conditions found that 72% of root biomass is An explanation for the lower percentages in the deeper layers in reforestation environments may be related to the history of these environments, attributing part of these results to disturbances suff ered by the trampling of animals that constantly grazed in these areas, hindering the distribution of roots in the deeper layers. Another explanation may be correlated with age (≤ 20 years) and characteristics of planted species (homogeneous population).
Regarding the density of fi ne roots, it was found that the lowest values were found in reforestation environments, both for dry and rainy periods, with the exception of RJ as (Table 2), showing signifi cant a (p<0, 05) at a depth of 5-15 and 15-30 cm for the NF environment in the dry period, with an average of 3.11 g-dm -3 and 2.68 g-dm -3 , respectively.
The density of fi ne roots ( Figure 3) presents a behavior very similar to that of biomass, changing according to the depth of the soil. This behavior can be explained by the high concentration of litter on the surface, favoring greater water retention and availability of oxygen and nutrients resulting from the decomposition.
Another likely explanation would be due to soil compaction, as the fi rst layers exert pressure on the lower layers, making the root distribution diffi cult. Thus, it was proved in the fi eld that the dynamics in the nutritional and productive behavior of native species occurred more intensely at an equilibrium level, very likely due to the non-occurrence of anthropic actions.

3.3.Carbon in the soil
The results of the analyzes showed that the concentrations of organic carbon (Corg) in the dry period did not diff er between the layers of 0-5 and 5-15 cm, in the diff erent environments, (Figure 4). At a depth of 15-30 cm for the same period, the JRE environment showed a signifi cant diff erence to the other environments with a value of 3.22 g kg -1 , this being the lowest concentration in the dry period.
This fact supports the observations of this study, where the JRE, TRE and MRE areas suff ered disturbances due to another economic activity (livestock), which may have infl uenced the results at depths of 15-30 cm, proving to be signifi cant (p<0, 05).
The results of the analyzes in the rainy season (January) showed signifi cant diff erences in the comparison between the areas of native forest and reforestation for a depth of 15-30 cm, with values of 12.41 g kg -1 for NF, 8.97 g kg -1 for JRE and 8.88 g kg -1 Table 1 -Root biomass in diff erent environments and depth of study, Humaitá-AM. Tabela 1 -Biomassa de raízes nos diferentes ambientes e profundidade de estudo, Humaitá-AM.
Means of fi ve repetitions Values followed by the same lowercase letter in the column do not diff er by Tukey test at 5% and uppercase in the row, do not diff er by T test of paired samples; TRE = reforestation with teak; JRE = reforestation with genipap; MRE = silvopastoral with mixed species; NF = native forest. Médias de cinco repetições. Valores seguidos da mesma letra minúscula na coluna não diferem pelo teste de Tukey a 5 % e maiúscula na linha, não diferem pelo teste de T de amostras pareadas; RET = refl orestamento com teca; REJ = refl orestamento com jenipapo; REM = silvipastoril com espécies mistas; FN = fl oresta nativa.

Áreas de estudo
Biomass ( for TRE, as shown in Figure 3. This may be correlated with the characteristic of less dense vegetation cover in natural forests, whereas in homogeneous plantations the soil is more exposed to climatic infl uences such as high temperature and high rainfall at this time of year in the region.
In the comparison between the dry and rainy periods, it was found that there was a signifi cant diff erence only for the depth of 5-15 cm for the JRE environment and for the depth of 15-30 cm for NF, with higher concentrations in the rainy period. This fact may be associated with greater microbial activity, Table 2 -Density of roots in diff erents in diff erent study environments, Humaitá -AM. Tabela 2 -Densidade de raízes nos diferentes ambientes de estudo, Humaitá -AM.
Five-repeat means Values followed by the same lowercase letter in the column do not diff er by the Tukey test at 5% and uppercase in the row, do not diff er by the T test of paired samples; TRE = reforestation with teak; JRE = reforestation with genipap; MRE = silvopastoral with mixed species; NF = native forest. Médias de cinco repetições. Valores seguidos da mesma letra minúscula na coluna não diferem pelo teste de Tukey a 5 % e maiúscula na linha, não diferem pelo teste de T de amostras pareadas; RET = refl orestamento com teca; REJ = refl orestamento com jenipapo; REM = silvipastoril com espécies mistas; FN = fl oresta nativa.

4.DISCUSSION
In the comparison between the dry and rainy season, there were no signifi cant diff erences in the biomass for the diff erent collection depths and diff erent study environments, showing a slight increase in the biomass production in the rainy season, except for JRE, as shown in Figure 2. Precipitation is one of the climatic variables that most infl uence the production of biomass in tropical forests (Green et al., 2005). Similar results were also observed by Metcalfe et al. (2008), in the Amazon rainforest.
A possible explanation for the higher values that occurred in the dry period in JRE, including all depths, is that the water defi cit associated with lower concentrations of organic carbon in the soil, has stimulated the root production. Fine roots induce the production of more biomass when there are fewer resources available below the soil surface (Markesteijn and Poorter, 2009).
In the present study, it was observed that the more the depth of the soil increases, the number of fi ne roots decreases, as shown in Figure 3. According to Finér et al. (2011), the root biomass decreases exponentially from upper to lower soil layers in diff erent forest biomes, alternating its behavior only in the total rooting depth. This behavior was also observed in a study of fi ne roots in young Eucalyptus dunnii Maiden trees by Dick and Schumacher (2019), corroborating the results presented in this study.
The density of fi ne roots (Figure 3) can be a factor related to the characteristic of the plant genotype, as well as to the nutritional behavior, productive   potential and capacity to adapt to environmental stress conditions (Martins et al., 2004).
Similar results regarding the percentage of root biomass found in Eucalyptus cloeziana F. Muell plantations, with almost 50% of the biomass in the fi rst 5 centimeters (Navroski et al., 2010), corroborate the results obtained in this study, where 66% of the biomass root of JRE, are concentrated in the fi rst 5 centimeters ( Table 2). The highest concentration of fi ne root biomass in a study with Araucaria angustifólia were found in the fi rst 20 cm of soil depth, making up a total of 53.4% of the total biomass (Schumacher et al., 2004).
Organic carbon contents have been used as indicators of soil alterations, being lower in conditions of environmental disturbance (Mendes, 2018). Results found by Gatto et al. (2010), in the order of 22.44 g kg -1 , in Eucalyptus plantations are similar to the values obtained in MRE in the order of 18.92 g kg -1 , found in the rainy season. In the comparison between the two periods (dry) and (rainy), the behavior of organic carbon was similar to that of root biomass, with no signifi cant diff erences, except for JRE, which presented signifi cant concentrations at depths of 5-15 and 15-30 cm, in the comparison between dry and rainy season. This fact may have a strong correlation with the rugged relief associated with the high precipitation indices at this time of year. The higher concentrations of organic carbon found in the surface layers are explained by the fact that the soil surface suff ers greater interference from organic matter from the fall of leaves and branches, promoting more intense nutrient cycling processes (Vital et al., 2004).
Native forest soils showed greater acidity, perhaps due to the degradation process of organic matter and its rapid mineralization, resulting in greater natural soil acidifi cation in these environments. The effi ciency of organic compounds increases the availability of nutrients in the soil, it is assumed that many of these acids are degraded within a few days after the release of plant residues (Costa et al., 2013).
Approximate pH values in native forest areas of 4.30 and pasture of 4.40 were found by Oliveira et al. (2015), corroborating the values found in this study. The higher pH values of the reforestation areas may have a correlation with management practices, notably the practice of liming, and soil preparation may have contributed to higher percentages of fi ne roots in the fi rst layers of the soil.

5.CONCLUSIONS
The production of root biomass, adding the three study depths, was greater in the native forest environment, occurring more intensely in the rainy season of the year, reaching values of 8.19 t. ha -1 , decreasing as depth increases. However, the highest percentages, corresponding to the total produced, occurred in reforestation environments for a depth of 0-5 cm, reaching values of 72.18% for MRE in the rainy season.
The behavior of root density as a function of soil volume showed that the highest concentration of roots is found in the fi rst 5 centimeters of depth, not diff erentiating between environments and periods of the year. For the other depths it was signifi cant for the native forest environment, with higher values in the rainy season. Soil organic carbon (Corg) showed signifi cance between the dry and rainy season for the NF and JRE environments.
It is assumed that, considering the results obtained, edaphoclimatic factors such as precipitation, soil and characteristics of forest species infl uenced both the production of fi ne root biomass and the estimation of soil organic carbon. We can also conclude from the results obtained that reforestation can be an alternative to improve soil quality in areas degraded by activities that constantly turn the soil over and make them susceptible to various factors of soil degradation.

REFERENCES
Almeida EJ, Luizão F, Rodrigues DJ. Litterfall production in intact and selectively logged forests in southern of Amazonia as a function of basal area of vegetation and plant density. Acta Amazônica.