Distribution of available nitrogen forms in soil under Quilombola management

: Management of Quilombola systems are primitive agricultural systems based on the ancestral knowledge of Afro-Brazilian enslaved people . Here, the aim was to understand the impact of these primitive farming methods on the distribution of available nitrogen (N) forms in the soil profile of two Brazilian Cerrado phytophysiognomies. The soil was sampled in Cerradão (high Cerrado) and Cerrado Stricto sensu (low Cerrado) at six soil depths (0-10, 10-20, 20-30, 30-40, 40-50, and 50-60 cm). The following management systems were considered: pasture (PP1 and PP2), maize cultivation (M1 and M2), citrus-cassava intercropping (T1), and citrus monoculture (T2). In addition, the soil was sampled in the native area of Cerradão (NC1) and Cerrado Stricto sensu (NC2). Three N forms were determined: i) available nitrogen (Av-N), ii) ammonium (NH 4+ -N) and iii) nitrate (NO 3– -N) contents. The Av-N content decreased with increasing soil depth only in NC1 and NC2. The NO 3– -N content was similar at all soil depths for maize and pasture, while the content decreased at soil depth for NC1, NC2, and T1. NH 4+ -N was similar in M2 and PP2, but it increased in T2, ranging from 6.17 mg kg –1 to 17.54 mg kg –1 . Overall, the dynamics of available N forms varied according to the Cerrado phytophysiognomy and the management systems and NO 3– -N was the most constant N form in the soil profile. Therefore, although the management of Quilombola systems is less intensive, they negatively affect the dynamics and N availability, mainly where management is less conservative, that is, in maize and citrus monocultures.


Introduction
The Brazilian Cerrado is a Savannah-like region with a relatively dry climate covering 2 million km 2 , representing 23 % of Brazil's total agricultural area, and is considered a hot spot for biodiversity (Ratter et al., 1997).Lately, the Cerrado has become the most important agricultural area for the Brazilian agribusiness, with significant technological development and high productivity for the national and international food markets (Colli et al., 2020).The Brazilian Cerrado presents different phytophysiognomies, usually comprised of weathered ancient soils.The weathering process has impoverished the soil chemically regarding nutrient contents and acidity, with high aluminum (Al) saturation.However, physically, Cerrado soils present high aggregation and stability (Dias et al., 2019;Silva et al., 2019).Woodland savannah (also referred to as typical savannah or Cerrado Stricto sensu) is the most prevalent phytophysiognomy in the Cerrado, presenting deeper soil covered by many tortuous trees, with thick barks and leathery leaves (Eiten, 1972).The second most prevalent phytophysiognomy is the Cerradão (high savannah), which shows a predominance of forestlike vegetation characterized by an almost closed canopy where the soil is not very deep nor very fertile (Ratter et al., 1997;Felfili and Fagg, 2007).
Many years before the agribusiness started to cultivate the Cerrado area, people of African ancestry, generally descendants of olden-time enslaved people, which succeeded in escaping from their masters by hiding in Cerrado areas far from cities or busy rural areas.There, they congregated and established small villages, which are called Quilombos.Their traditional agricultural systems, based on ancestral knowledge, may cause little interference with the original soil properties (Nascimento et al., 2017;Silva et al., 2019), at least when considering the traditional agriculture that uses many inputs.Furthermore, even in areas of traditional agriculture, such as Quilombos and indigenous communities, landuse change (LUC) may cause impacts on soil quality indicators, such as changes in carbon (C) and nitrogen (N) stocks (Kohler and Brondizio, 2017;Ramos et al., 2022).Thus, although many works show the success of conservative agricultural management systems in restoring soil health, little is known about this condition in communities where technical assistance is still neglected, as in Quilombola communities.Therefore, this research is considered groundbreaking to unravel the impact of Quilombola management systems on N distribution in the soil profile.It is important to highlight that N is one of the most limiting nutrients for crop production.

Site description and soil sampling
In this study, the distribution of available N forms in soil profiles was evaluated in different Quilombola

Soils and Plant Nutrition
Research article systems in Brazilian Cerrado phytophysiognomies management systems (pasture, maize, and citrus) in two Cerrado phytophysiognomies (Cerradão and Stricto sensu) (Figure 1).The sites are in Quilombo Mesquita in the municipality of Cidade Ocidental (Central portion of Goiás State) and Federal District of Brazil (16°04'41" S, 47°52'05" W, altitude 1014 m) as described in Silva et al. (2019) and Ramos et al. (2022).According to Köppen classification, the climate is Aw, with dry winters and rainy summers.The mean annual temperature is 21 °C and rainfall is 1,500 mm (Alvares et al., 2013).The soil was classified as Rhodic Hapludox (Soil Survey Staff, 2010) and the physical and chemical soil properties are presented in Table 1.
The historical management of the sampled sites is shown in Table 2. Briefly, in HC, the M1 area has been managed for 30 years, with the first 20 years based only on the rice/bean/maize crop rotation and the last ten years based on synthetic inputs.The PP1 area has not been managed in the last 15 years and is naturally invaded by grasses without cattle grazing.The T1 area has been cultivated in the last five years with citrus intercropped with cassava.For 21 years, the T1 area was cultivated with grasses and fertilized with cattle manure in continuous grazing.
In LC, the M2 area has been managed for 15 years under grain cultivation with conventional practices of soil tillage.The PP2 area was naturally invaded by grasses without cattle grazing.The T2 area has been cultivated in the last five years with citrus monoculture    and fertilized with organic compost.Both native areas (NC1 and NC2) were used as references due to the absence of anthropic exploitation or interference.
The soil was sampled in Apr 2014 at six depths (0-10, 10-20, 20-30, 30-40, 40-50 and 50-60 cm).In reference areas (NC1 and NC2) and pasture (PP1 and PP2), a diagonal line was projected and the soil was sampled at every 50 m.In maize cultivation (M1 and M2), the soil was sampled in the plant rows and inter rows, while in citrus plantations (T1 and T2), soil samples were collected in the canopy projection of the citrus trees (Ramos et al., 2022).
Five random points were collected for each management system at each depth, representing the repetitions.We collected 240 composite samples to evaluate eight management systems, six soil depths and five replications.Soil samples were sieved through a 2.0 mm diameter sieve and stored at 4 °C.

Available nitrogen forms (Av-N)
Available N forms include organic components of low molecular weight, easily absorbed by plants, as amino acids and amino sugars besides the ammonium and nitrate (Oliveira, 1989).Several chemical extractors were used experimentally to estimate N availability in the soil, which generally shows a high correlation with N uptake by plants (Stanford, 1982;Bremner and Breitenbeck, 1983;Oliveira, 1989;Meneghin et al., 2008).
Test tubes with 2 g of soil, 0.2 g of magnesium oxide (MgO) and 0.1 g of Devarda alloy were submitted to steam distillation (Kjeldahl) using 25 mL of sodium phosphate-borate (Na 3 PO 4 /borax -pH 11.2 buffer solution).The distillate was collected in a flask containing 10 mL of 0.05 N HCl, until completing 35 mL.In this process, the heating implies N volatilization in the NH 3 form (stemming from amino sugars and N hydrolyses from amines and amides) (Oliveira, 1989).The amount of extracted N was determined by colorimetric spectrophotometry at 440 nm using 1 mL of Nessler's reagent (Meneghin et al., 2008).The readings were compared to the values of a standard solution containing 0, 15, 30, 45, and 60 μg mL -1 N (Coser et al., 2016).We determined ammonium and nitrate contents starting with a 15 g soil sample and applying the extraction method using 2 mol L -1 KCl (Bremner and Keeney, 1965).Again, the ammonium was obtained by steam distillation, collected in an indicator solution of boric acid and then determined by titration with acid of standard normality.Afterward, MgO and Devarda alloy were added to reduce nitrate and nitrite to ammonium (Bremner and Keeney, 1965).In summary, 15 g of soil were mixed with 50 mL of 2 mol L -1 KCl.After decanting the soil, the supernatant was filtered and 10 mL of the extract were transferred to test tubes, adding 2 g of

Maize cultivation M2 2 ha
Area managed for 15 years under grain cultivation with conventional practices of soil tillage.In the last ten years, the area has been annually cultivated only with maize.The soil remains permanently uncovered.
Citrus monoculture T2 0.5 ha Area managed for five years with tangerine (Citrus reticulata L.) cultivation with conventional minimum soil tillage.Citrus is not intercropped with any other crop and the soil remains always uncovered.Weeds in the area were removed and the waste was piled near the crop.
MgO and submitted to steam distillation (Kjeldahl).The distillate was collected in a flask containing 10 mL of 2 % H 3 BO 3 until it reached around 30 mL to determine ammonium by titration with 0.05 N H 2 SO 4 (standard acid).In the same test tube used in the first distillation to determine ammonium, we added 0.2 g of Devarda alloy, and we used the second distillation and another titration with the standard acid aforementioned to determine the nitrate concentration.

Data analyses
One-way ANOVA followed by Tukey's post hoc tests (p < 0.05) were used to assess soil N distribution differences between the Quilombola management systems.This approach is only possible when the sites have the same topography and soil type, and the edaphic or climatic conditions differ only in terms of land-use (Merloti et al., 2019).Therefore, this approach was adequate for our study.However, before conducting the ANOVA, the normality and variance homoscedasticity were tested using the Kolmogorov-Smirnov and Hartley tests, respectively.
The principal component analysis (PCA) was used as a data-reduction tool to select the N form most correlated with each Cerrado phytophysiognomy (Ramette, 2007).For the assumption of multivariate normality, the data were transformed to a log of (C +1), where (1) is a constant value to avoid negative results and C is the value of the measured variable (Legendre and Legendre, 1998).We compared the Quilombola management systems with the reference areas (NC1 and NC2), within each Cerrado vegetation to explore the N form dynamics, using an effect size analysis, which considers a 95 % confidential interval with 1000 bootstrap repetitions (Goedhart, 2016).For this analysis, we used the RStudio Team version 1.2.1335 (R Core Team, 2019).

Results
In general, there was a distinction between high Cerrado (HC) (Cerradão) and low Cerrado (LC) (Stricto sensu) according to the spatial ordering of the soil N forms.Even though there is partial overlapping in the confidential ellipses, the PCA (Figure 2) shows that only a minority of the samples shared the same composition.The available N (Av-N) correlated more intensively with the HC, while ammonium and nitrate correlated well with the LC (Figure 2).Overall, the Quilombola management systems located under HC presented a higher cation exchange capacity (CEC) (3.6 cmol c kg -1 ) when compared to LC (1.1 cmol c kg -1 ).
In the effect size analysis, we compared the management systems (maize, pasture and citrus cultivation) to the native Cerrado (NC1 or NC2) to show these management systems' positive or adverse effects.According to the management system, we found significant changes in N forms of the Cerrado phytophysiognomy (LC and HC).In the HC, T1 and PP1 were the most contrasting (Figures 3A and 3C) in terms of Av-N and nitrate (NO 3 --N).In the LC, however, there was no pasture management effect (PP2) for all N forms (Figures 3D, 3E, and 3F).On the other hand, comparing with NC2, T2 and M2 showed the most contrasting effects on Av-N and NH 4 + -N contents (Figures 3D and  3E).
Concurrently, there was a positive effect for the T2 system for NH 4 + -N and a negative one for NO 3 --N (Figures 3E and 3F).M2 caused a negative effect for Av-N (Figure 3D) and NH 4 + -N (Figure 3E), but not for NO 3 --N (Figure 3F).Similar results were detected for HC at PP1, which caused a negative effect on Av-N (Figure 3A) and NO 3 --N (Figure 3C), but without an effect on NH 4 + -N (Figure 3B).The systems T1 and M1 affected NH 4 + -N positively (Figure 3B).In the soil profile, there were no differences (p > 0.05) in Av-N among soil layers in all Quilombola management systems (M1, M2, PP1, PP2, T1, and T2).However, in the reference sites (NC1 and NC2), there was a decrease in Av-N with increasing soil depths, with the 0-10 cm layer presenting the highest Av-N levels (Table 3).In HC, there was a trend to decrease the NH 4 + -N to the depth of 20-30 cm in M1 and T1, but below that depth, this trend did not occur (Table 3).
On the other hand, in the LC, NH     6.17 mg kg -1 (0-10 cm) to 17.54 mg kg -1 (50-60 cm).At 0-10 cm soil depth, there was no difference between the management systems (M2, PP2, and T2) and NC2.However, the T2 system showed the highest NH 4 + -N in deeper layers of all management systems and native area (NC2) (Table 3).
Regardless of the Cerrado phytophysiognomy, maize (M1 and M2) and pasture (PP1 and PP2) presented constant NO 3 --N throughout all the layers.However, in the native Cerrado (NC1 and NC2) and intercropped citrus (T1), there was a decrease in NO 3 --N, mainly at the depths of 40-50 and 50-60 cm (Table 3).In the LC, there was no depth effect on NO 3 --N, neither within nor between management systems.However, NO 3 --N decreased at soil depth in NC2, mainly at 30-60 cm (Table 3).

Discussion
Besides the effects of the Cerrado phytophysiognomy on the distribution of available N forms in the soil profile, we also evaluated the magnitude of influence of the management of Quilombola systems compared to the native area.Overall, the Quilombola management systems under different Cerrado phytophysiognomy (LC and HC) showed apparent differences, which is somewhat expected, as they present contrasting characteristics, such as time of LUC, vegetation type, and farming methods.In addition, N distribution in the soil profile is affected by other factors, such as organic nutrient concentration, soil pH, precipitation, and biological N fixation (Mengel et al., 2001).
More than 90 % of the soil N is present in the organic form and some organic forms (e.g., amino acids, proteins, and amino sugars) are readily available for plants and microorganisms (Mechthild and Doris, 2010;Czaban et al., 2016;Stevens, 2019).Even though the soil has a considerable amount of N in organic forms, mineralization (a process of converting organic compounds into readily available forms) occurs slowly and in small amounts (1 to 3 %) during a crop cycle (Carneiro et al., 2013).Hence, it is essential to understand the dynamics of inorganic N forms in the soil, as they are often the preferred N source in nature and agriculture.Inorganic N forms consist mainly of ammonium (NH 4 + -N) and nitrate (NO 3 --N), which are readily available for uptake by plants or soil microbiota (Li et al., 2014).
For all Quilombola management systems, Av-N homogenization was observed in the soil profile while high contents were observed in the top layers of natural areas.Natural areas commonly accumulate organic matter in the topsoil, which could increase of the Av-N content.In contrast, areas that have undergone LUC (e.g., Cerradão in pastures or monocultures) tend to decrease the N content in the soil profile (Kizilkaya and Dengiz, 2010;Groppo et al., 2015;Borges et al., 2019;López-Poma et al., 2020;Costa et al., 2020).Maize cultivation under the LC presented the lowest available N content compared to other systems.This finding may be related to the farming practices used, since it is one of the most intensive tillage managements compared to the other systems studied (Green et al., 2007).NO 3 --N is well recognized as soluble and mobile and it is therefore readily leached into the soil profile (Jadoski et al., 2010), mainly in soils where positive soil charges occur, as in the charge-dependent ones (Alcântara and Camargo, 2005).In this investigation, NO 3 --N was the most constant along the deeper layers in pasture and maize cultivation, regardless of the Cerrado phytophysiognomy, but remained under the top layers in natural areas of Cerrado (NC1 and NC2) and in the citrus-intercropped system (T1).High contents of NO 3 ---N and NH 4 + -N have been observed in the superficial layer of natural areas of Cerrado when compared to pasture (López-Poma et al., 2020).These authors also argue that the conversion of native areas (Cerradão) to pasture resulted in negative impacts on soil N content.On the other hand, similar contents of NO 3 --N and NH 4 + -N between native areas of Cerrado and pasture in different soil layers have been reported (Frazão et al., 2010).
The available N forms are easily interchangeable, especially for NH 4 + -N and NO 3 --N.Thus, it is sometimes advisable to repeat the experiment for two or more years, as different climatic conditions, for example, may significantly influence it.However, in the present study, we only compared different sites cultivated with varied crops, located very close to each other and exposed to the same climatic interferences.
We consider this investigation pioneering because it shows the distribution of the available N forms in soil under different Quilombola management systems, where the agricultural practice is less intensive than in modern agriculture.Overall, we noticed that the dynamics of available N forms varied according to the Cerrado phytophysiognomy, with the prevalence of available N forms in the high Cerrado (Cerradão).In contrast, in the low Cerrado (Stricto sensu), the inorganic N forms (NH 4 + -N and NO 3 --N) were more prevalent.In addition, although the Quilombola family farmers practice less intensive tillage, there was a noticeable impact on the available N dynamics, especially where the management is less conservative, as in maize cultivation, with excessive tillage.On the other hand, the T1, planted with a consortium of legumes, represents an excellent N management while generating income for the producer.
not decrease along the soil profile in M2 or PP2.T2 presented an increase in NH 4 + -N in the soil profile, ranging from
NC1 and NC2 = native Cerrado; M1 and M2 = maize cultivation; PP1and PP2 = pasture; T1 = citrus -cassava intercropping and T2 = citrus monoculture.Lowercase letters compare the different management systems of a same phytophysiognomy within the same layer at 5 % significance by the Tukey test (p ≤ 0.05) and uppercase letters compare the layers within the same management system (p ≤ 0.05).

Table 2 -
Ramos et al. (2022)the study sites.Source: Adapted fromRamos et al. (2022).Webster, fertilized with cattle manure and the pasture maintained with continuous cattle grazing.Pasture removed and for four years, the area cultivated with tangerine (Citrus reticulata L.).Since 2016, the area has been intercropped with cassava (Manihot esculenta L.).Mulching preserved in the area composed of fruit remains, leaves, and weeds to ensure that the soil is permanently covered.