Evaluation of the Siltation of River Taquari, Pantanal, Brazil, through 210Pb Geochronology of Floodplain Lake Sediments

Journal of the Brazilian Chemical Society

Comissão Nacional de Energia Nuclear, Rio de Janeiro, Brazil¹  Galdino, Sergio¹  Guimarães, Jean R. D.¹  Carvalho, Zenildo L.¹  Padovani, Carlos R.¹  Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil¹  Vieira, Luis M.¹  Godoy, José M.¹  Pereira, José C. A.¹  EMBRAPA, Corumbá, Brazil¹  Pontifícia Universidade do Rio de Janeiro, Rio de Janeiro, Brazil¹ 

关键词: sediment dating;    210Pb;    Brazil;    Hg;    erosion  ;   

DOI  :  10.1590/S0103-50532002000100011

学科分类：化学（综合）

来源: SciELO

【 摘 要 】

The Taquari River basin (TRB) drains an area of 28,000 km2 in the plateau adjacent to the so-called Pantanal, a large floodplain area in midwest Brazil. It is characterized by sedimentary soils, primarily sandy soils, with areas of accentuated declivities and high erosive potential. These characteristics of TRB, associated with its intensive use and the mishandling of the soil by agricultural activities during the last 25 years have intensified erosion processes. This is evidenced by the presence of recent gullies and by the siltation of rivers of the TRB. The intensification of these processes causes the instability of the riverbed, and consequent flooding of large areas, leading to environmental, social and economic problems. These processes have already been reported by several authors.1-9 Direct estimates of sediment production related to agricultural activity are difficult to obtain for such a large and diversified area as the TRB, and no data are available at present. Historical records of suspended sediment loads for the Taquari River are scarce, although attempts have been made to compare such information.8 The lack of former studies for the characterization of this environment, prior to the expansion of agricultural activities in the plateau of the upper Taquari River region, makes it difficult to evaluate and compare past and present environmental impacts. Such evaluation can however be undertaken using a methodology that provides information on past sedimentation rates. Usually, 210Pb sediment dating is validated using data on 137Cs in fallout from atomic tests or from accidents such as Chernobyl. However, both contribute very little to the pool of 137Cs in the Southern Hemisphere, thereby limiting the use of this type of data for dating purposes.10 In addition to 137Cs, other indicators such as pollen or heavy metals can be applied.10-11 In the Amazon and Pantanal regions, the sedimentary records of Hg offer potential opportunity for recent sediment dating, due to the gold rush that took place in Brazil and other neighboring countries in the 1980's.12 Small scale primary gold mining, using Hg amalgamation for gold recovery, is a practice that started as early as the 18th century, in areas such as Poconé, on the border of the Pantanal and the surrounding lateritic soil plateaus.13 However, as in other gold mining areas, the link between Hg emissions to the environment from gold mining and Hg in fish could not be demonstrated, and the latter were extremely low even in piscivorous fish species sampled near Poconé.14 In contrast, high natural Hg levels have been found in most tropical soils and soil erosion was shown to be the main Hg source to aquatic systems in different drainage basins in the Amazon. Roulet et al15 and. Lechler at al16 have shown that while lateritic and other forest soils along the Madeira River were enriched in Hg, recent floodplain soils along the Jamari River were only slightly enriched in Hg and the channel bed sediments were very low in Hg. Hylander et al13 have carried out a survey of total Hg in surface sediments from floodplain lakes covering the north of the Pantanal, near Cáceres and Barão de Melgaço, and, at the South, in the region near the confluence of the Cuiabá and Paraguay rivers. They have found a mean value of 33.2 ng g-1 (first quartile 18.4 and third quartile 46.8 ng g-1). The Hg content of fine bottom sediments from water courses in the immediate vicinity of present or former gold mining sites was higher than in the lake sediments. The results reported by them, for both lakes and river sediments, were lower than those previously reported17-20 for surface sediments. Leady and Gottgens21 observed similar values for Hg in lake sediments from a non-contaminated reference area in Northern Pantanal (29.1±0.7 ng g-1). The lakes included in the present study are situated in the middle Taquari drainage basin area. Gold mining areas are absent in the upper and middle Taquari drainage basin, but the middle Taquari is considered a deposition region for the lateritic soil eroded from the upper basin, a process enhanced by the agricultural development in the Taquari river highlands. The upper land is considered a natural source of Hg to the Pantanal area13 and therefore if an increase in the mass sedimentation rates has occurred in recent decades, due to an increase in soil erosion at the upper TRB, an increase in the Hg flux to the sediment column must have ocurred . This Hg flux should correlate more closely with the mass sedimentation rate than with the Hg concentration in sediment layers of floodplain lakes from the middle Taquari region. On the other hand, these Hg concentrations in sediments should be relatively constant.  Experimental The lakes studied in this work are located close to the city of Coxim, at coordinates 180 24' 10'' S and 540 58' 46'' W (Lake 1), 180 21' 57'' S and 540 59' 38'' W (Lake 2) and 180 21' 57'' S and 550 00' 22'' W (Lake 3), Figure 1. These lakes are oxbow lakes, formed by abandoned river meanders. They are presently separated from the river and receive river water mainly during the flood season, when large amounts of sediment in suspension are transported by the river. The lakes were selected using aerial photographs from 1966, at the 1:60,000 scale and Landsat-TM satellite images (bands 3, 4 and 5 composition), at the 1:100,000 scale, from 1995. Since the aerial photographs from 1966 were taken during the drought period of 1958-1972, the chosen lakes were those older than 30 years and with a probability of receiving water even during that dry period. A 20kg gravity corer with PVC tubes of 100 cm length and 5.0 cm internal diameter was used to collect sediment cores from study lakes. A first sampling campaign was carried out during October 96, and another one year later. Three cores were collected from Lake 1, three from Lake 2 and one from Lake 3. The cores were sectioned into 2cm slices, and the central part of each slice was reserved for mercury determination. Each slice was dried at 105°C, ground, a 5 gram aliquot taken and the excess 210Pb content was determined by HBr leaching as described by Godoy et al22. A low-background proportional counter (EG&G Prof. Berthold LB 750) was used and the detection limit for 5 g dry material and 24,000 s counting time was 3 mBq g-1. The results were evaluated according to Appleby and Oldfield.23 For the total Hg determination, sediment were dried at 50oC to constant weight and samples of up to 2 g were digested for 5 min in a hot bath at 60oC, in 5 mL of Milli-Q water and 5 mL of HCl:HNO3 (3:1). After addition of 50 mL of Milli-Q water and 15 mL of 5 % KMnO4, samples were further digested for 30 min at 60oC and left overnight at room temperature. A few drops of 12 % NH4OCl were added, the extracts were filtered through Whatman 42 paper and analyzed by CVAAS. Analytical accuracy was verified by frequent interlaboratorial comparisons and regular analysis of NIES Pond sediment certified reference material.  Results and Discussion Figures 2a-2c show the changes on the mass sedimentation rates (mg cm-2 year -1) found in Lake 1. Sampling point 1 was located closer to the river than the sampling point 2, and point 2 was sampled again during the 1997 sampling campaign (Figure 2c). All the profiles show a similar picture, with a peak during the 1970s to 1980s coinciding with the end of the dry period of 1958-1972. Preliminary evaluation of the existing rainfall data for the plateau area indicated that the rainfall erosivity was larger during the period from 1974 to 1994 than that from 1965 to 1973.8 In the 1970s, agricultural activities on the plateau underwent a marked expansion, intensifying the erosion processes, over the last 25 years. The values of the mass sedimentation rate corresponding to the present decade also show a period of higher sedimentation rates at present. For all the three cores, the mean mass sedimentation rates for the 1990s are significantly higher (p<0.001) than those for the 1920-1950 period: (395±44) and (141±44) mg cm-2 year -1 for the sampling point 1, (261±54) and (125±17) mg cm-2 year -1 and (458±85) and (143±27) mg cm-2 year -1 for the first and second cores of sampling point 2. A comparison between the results of these cores is shown in Figure 3, illustrating the similarity of the two cores taken at point 2, one during 1996 and the other during 1997.     Similar results were observed in Lake 2 for the sampling points 1 and 2 (Figures 4a and 4b). Both points are located in the inlet channel of the lake, one at its left and the other at its right side, almost one in front of the other. The increase in mass sedimentation rates, observed since the late 1970s, was confirmed and the recent rates are substantially greater than those for the 1920-1950 period: (511±76) and (105±21) mg cm-2 year -1 and (525±81) and (93±18) mg cm-2 year -1 for the sampling points 1 and 2, respectively. The close concordance between the results of both sampling points is illustrated in Figure 4c.   Forsberg et al24 have reported an analogous

【 授权许可】

Unknown