HEAVY METALS DISTRIBUTION IN BURNT SOILS OF A RAIN FOREST AREA IN RORAIMA, BRAZIL.

Marcelo T. Nascimento; Cristina Ma M. Souza; Luiz E. O. C. de Aragão; Dora Ma Villela; Carlos E. de Rezende & Marcelo, G. de Almeida (Laboratório de Ciências Ambientais, Universidade Estadual do Norte Fluminense, Av. Alberto Lamego, 2000, CEP: 28015-620, RJ, Brazil, cristal@cbb.uenf.br)

 

Abstract

            A study on heavy metals distribution in burnt soils was carried out, in a forest area in Roraima State one month after the worst fire in the Amazon region. Burn forest areas has been an increasing activity in the Amazon area as consequence of deforestation processes for human occupation. The results showed that total concentrations of Al, Fe, Mn, Cr, Zn, Cu and Ni, were in natural levels, but Pb, Cd and Hg values were higher. The sequential extraction did not show any differences between the metals distribution in burnt and unburnt areas.

 

Introduction

            Land use activities have been enhanced an intensity human occupation in rural areas. In Brazil, some regions still occupied with indigenous inhabitants have been devastated and intensely burnt. Fire has became more usual in Amazon, especially  by small farmers to clean the soil for selective wood exploiting, putting in danger adjacent areas. Nowadays, this practice is extending over the whole in Brazilian Amazonia, from Rondonia to Maranhão.

            Just in the last two decades several studies on natural and anthropogenic heavy metal emission have been done. There has been a growing recognition of the need to quantify natural sources and cycling regional processes of metals (Lacerda & Salomons, 1991). This study aimed to compare the heavy metal contents and their availability in burnt and unburnt soils on Amazon area (Roraima state).

           

Methodology

            The Roraima state is characterized by a tropical wet and sub-equatorial climate, with temperature between 200C and 380C and annual average precipitation between 1000mm and 2250 mm. “El Niño” phenomenon in 1997 changed the rainfall regime causing a very strong dry season.

            The study area was located in the “Trairão” Village in Roraima State, north of Brazil. It’s economy is mostly related with agriculture and cattle breeding, with a increasing perspectives for immigrants. The severity of the fire in this region was due to “El Niño” event that prolonged the dry season associated with low relative air humidity and absence care to prevent the fire (Nascimento et al., 2000). Four transects of 1Km long each were set up (November, 1999) in the area. Transects varied from undamage to severely damage (understory totally burnt) rainforest. Four plots (10m x 50m) in each transect was established with intervals of  250m.

            Bulked soil samples were collected in each area, in the first 10 cm with plastic trowel and packed in plastic bag until the laboratory. The samples were homogenized, considering grain size smaller than 2mm, oven dried (400 C) and weighted for chemical analysis. To determine the granulometry distribution inside the plots, an aliquot (100g) of the total soil was used (ABNT/NBR 7181, 1984).

            The extraction method for total metals was performed using 1g ± 0.001g of dried samples placed in teflon bomb with acid mixture HF/HNO3 in an oven (1100C) during 12h (Krause et al., 1995). For mercury, the samples were wet digested using an oxidant mixture HCl/HNO3, (3:1) and KMnO4 5% (Bastos et al., 1998). To certify the quality control an internal soil sample supplied by Biophysics Institute in Rio de Janeiro (Radioisotope Laboratory) was performed. Total carbon was determined using CHNS analyser (Perkin Elmer).

            Sequential extraction based on Ure et al., (1993) was performed, allowing the metals speciation in four geochemical phases: exchangeable, reducible, oxidazable and residual. For all studied metals more than 80% was recovery, with variation coefficients for analytical replicates bellow than 10%. All metal determinations were performed by ICP - AES (Varian Liberty Series II), using a VGA accessory for Hg.

 

Results And Discussion

            The average granulometry distribution of the soil samples showed that the soils area are sand (table 1) with low variation coefficient. This fraction is mostly represented by fine sand fraction. Coarse material was not found in the analysed samples.

 

Table 1: Granulometry Distribution (%) in Trairão Soils and Roraima Sampling Point Coordinates. Average and variation coefficient (%).

 

Transects

Granulometry Distribution

Coordinates

 

Stone

Sand

Silt /Clay

 

T1

-----------

82 ± 5

18 ± 2

30 33’ 10” N

610 49’ 39” W

Azimute

2900

T2

----------

79 ± 2

21 ± 3

30 33’ 12” N

610 54’48” W

30 33’ 22” N

610 55’20” W

T3

----------

77 ± 2

23 ± 3

30 38’ 04” N

610 51’19” W

30 37’43” N

610 50’48” W

T4

----------

84 ± 4

16 ± 3

30 37’ 26” N

610 50’20”W

30 37’ 09” N

610 52’ 07” W

 

            The average concentrations for all metals in the samples are presented in table 2. In general the heavy metal concentrations showed the following distribution: Al >Fe > Mn > Cr > Zn > Pb > Cu > Ni > Cd > Hg.

            The magnitude of the concentrations obtained for mostly metals are in agreement with other areas (Azcue, 1987; Houtmeyers, 1985) and below that considered for world average, except Pb, Cd and Hg in transect 2, enriched up to five times for Cd and two times for Hg and Pb. It is noteworth that this transect was not burnt at all, this transect represents the reference background for the area. All the concentrations measured along this transect was higher than the others, including total carbon (%).

            Although the fire varied in intensity along the transects (Nascimento et al., 2000), it was not  verified differences among the burnt areas. However, the transect  3 showed the lowest concentrations for some metals such as  Al   (4361 mg.g -1),      Fe  (3932

mg.g -1 ), Cr (6.0 mg.g -1 ), Cu (2.2 mg.g -1 ) and Cd (0.21 mg.g -1 ). In fact, this was the transect that had its understory totally destroyed by the fire.     

The high variation coefficient found in all transects (burnt and unburnt areas), showed the high environmental variability that occurs in the soils

 

Table 2: Average, Variation Coefficient (%) and Concentration Ranges of Heavy Metals in Soils (mg.g -1 ).

 

Metals

Transects

world average*  (mg.g -1 )

 

 

T1

T2

T3

T4

 

Al

4666 ± 40

(2798 - 6051)

13452 ± 55

(4194 – 22197)

4361 ± 91

1860 - 10305

7612 ± 28

4921 - 10207

---------

Fe

5507 ± 35

(3498 - 5670)

7252 ± 97

(1798 – 17187)

3932 ± 50

(2032 - 6702)

5101 ± 31

(3780 - 7229)

38000

Mn

88.9 ± 51

(42.2 - 135)

707 ± 82

(273 - 1545)

109 ± 90

(55 - 148)

182 ± 68

(71 - 346)

850

Cr

11.1 ± 104

(3.2 - 28.4)

26.0 ± 66

(4.7 – 46.9)

6.0 ± 34

(3.7 - 8.5)

10.6 ± 33

(5.8 - 14.0)

100

Zn

8.1 ± 33

(5.8 - 11.9)

22.7 ± 56

(10.1 - 40.4)

10.9 ± 43

(4.3 - 15.4)

14.1 ± 21

(11.2 - 18.1)

50

Cu

2.8 ± 49

(1.5 - 4.7)

8.7 ± 80

(2.9 – 18.8)

2.2 ± 36

1.3 - 3.1

3.0 ± 38

(1.9 - 4.6)

20

Pb

3.3 ± 41

(1.4 - 4.5)

19.0 ± 115

(5.0 – 51.7)

4.5 ± 31

(2.5 - 5.5)

6.3 ± 35

(4.2 - 9.1)

10

Ni

6.1 ± 101

(2.2 - 15.2)

6.3 ± 75

(2.1 – 12.6)

1.3 ± 60

(0.81 - 2.4)

0.75 ± 20

(0.58 - 0.94)

40

Cd

0.33 ± 67

(0.08 - 0.62)

0.33 ± 82

(0.08 - 0.69)

 

<DL - 0.63)

0.36 ± 76

(0.15 - 0.75)

0.06

Hg

0.03 ±36

(0.01 - 0.04)

0.06 ± 38

(0.03 - 0.08)

0.04 ± 59

(0.01 - 0.06)

0.03 ± 16

(0.03 - 0.04)

0.03

C %

0.90 - 1.3

1.1 - 5.9

1.9 - 3.1

1.2 - 2.3

-----

C/N

15 - 18

12 - 13

14 - 18

11 - 15

-----

                <DL= lower than detection limit of the instrumental analysis * Fórstner & Wittmann, 1984

 

                The results obtained from sequential extraction are presented in table3. The transects 2 (unburnt) and 3 (burnt) were choose. All metals in natural area (T2)  presented a predominant distribution in residual phase (higher than 70%), except Mn (47%). This association with residual phases indicates natural sources for these elements (Förstner & Wittmann, 1984).

           

Table 3: Percentage (%) distribution of  heavy metals in geochemical phases of  soil.

 

Phases

Metals

 

Al

Fe

Mn

Cr

Zn

Cu

Pb

Ni

Cd

 

Exc

0.5

--------

1

-------

--------

24

--------

8

-------

T3

Red

1

3

39

-------

2

18

13

6

-------

 

Oxi

0.5

--------

10

16

18

19

--------

17

-------

 

Res

98

97

50

84

80

39

87

69

100

 

Exc

0.1

---------

15

--------

1

14

-------

7

-------

T2

Red

0.5

1

35

--------

2

7

13

9

4

 

Oxi

0.4

1

3

21

6

7

-------

14

-------

 

Res

99

98

47

79

91

72

87

70

96

                Exc = exchangeable; Red = reducible; Oxi = oxidazable; Res = residual

 

Manganese showed  similar percentages between the transects, with a good relation in reducible phase, representative by manganese and iron oxides. The elements Cr (16 and 21%), Zn (18 and 6%), Cu (19 and 6 %) and Ni (17 and 14%) had some association with oxidazable phase in both transects  (T3 and T2, respectively), suggesting the chemical affinity to form organic matter complexes.

                The results found from sequential extraction did not show a relationship between the burnt areas and the metal phase associations, since the distribution observed  was similar between the two transects. Although Mn, Cu and Pb showed the highest differences between unburnt and burnt areas, it was difficult to confirm their lost, since the environmental variability is high inside the transect.

The higher concentrations for Cd, Hg and Pb compared with the world average values, suggest the need of further studies for these metals in the area. The dispersion mechanisms specially for Hg, are probably related to anthropogenic sources such as gold mining activities, spread out in Amazon region in the 80´s. The Hg released to the atmosphere during the burn amalgam may be transported to the study area, that is located  near to the “Santa Rosa” gold mining. Since Hg could remain in the soil for more than a hundred years (Lacerda & Salomons, 1998), it is very probably that this element have already been dispersed all over the region.

 

Acknowledgment: to Guttemberg de Oliveira head of Maracá ecological field station (IBAMA) for the loggistic support  and IBAMA (DF) for finantial support to field experiments.

 

References

 

ABNT/NBR 7181, 1984. Análise Granulométrica (Método de Ensaio), dezembro de 1984, 120p.

Azcue, (1987). Estudo da Poluição Ambiental por Metais Pesaods no Sistema Rio Paraíba do Sul – Rio Guandú (RPS – RG) através da Metodologia de Abordagem pelos Parâmetros críticos. Master Dissertation, IBCCF, Universidade Federal do Rio de Janeiro.

Bastos WR., Malm O, Pfeiffer W, Clearly D, 1998. Ciência e Cultura, 50(4): 255 - 260.

Förstner U, Wittmann GTW, 1984. Metal Pollution in the Aquatic Environment. Springer Verlag, Berlin, Heidelberg, 486p.

Houtmeyers J, Vanheuged P, Steegmans R; Demuyuck M, Beckers B, 1985. International Conference Heavy Metals in the Environment, Vol. 2: 581 – 583.

Krause P, Erbslöh B, Niedergesäb R, Pepelnik R, Prange A, 1995. Fresenius Journal of Analytical Chemistry, 353: 3 - 11.

Lacerda LD, Salomons W, 1998. Mercury in the Amazon: A Chemical Time Bomb? Dutch Ministry of Housing, Physical Planning and Environment, 46p.

Nascimento MT, Oliveira filho AT, Fontes MAL, França JT, Hay J, Gribel R, 2000. Ciência Hoje, vol 27 (157): 40 – 43.

Pfeiffer WC, Lacerda LD, 1988. Environ Technol Lett 9: 325 – 330.

Ure AM, Quevauviller Ph, Muntau H,  Griepink B, 1993.. Journal Environ Anal. Chem., 51: 135 - 151.