Environmental Research Section A 88, 134}139 (2002)

doi:10.1006/enrs.2001.4312, available online at http://www.idealibrary.com on

Dichlorodiphenyltrichloroethane in Soil, River Sediment, and Fish

in the Amazon in Brazil
Joao P. M. Torres,*  Wolfgang C. Pfeiffer,* Steve Markowitz,- Ronald Pause,? Olaf Malm,*
and Jan JapengaA
*LaboratoH rio de RadioisoH topos Eduardo Penna Franca, Instituto de BnH ofnH sica Carlos Chagas Filho, UFRJ, 21949 Rio de Janeiro, Brazil;
-Center for the Biology of Natural Systems, Queens College, Flushing, New York 11367; ?Wadsworth Center for Laboratory Research, N.Y.
State Health Department, Albany, New York 12201-0509; and AALTERRA, Wageningen, The Netherlands

Received January 22, 2001

According to Deane and colleagues (Deane et al.,

Dichlorodiphenyltrichloroethane (DDT), its main
metabolites, and other organochlorines were ana-
lyzed in soils (n ⴝ 6), Buvial sediments (n ⴝ 14), and
Ash (n ⴝ 10) that were collected in several areas of
the Amazon region in Brazil. The samples were ana-
lyzed by capillary column gas chromatography
coupled to electron capture detection. DDT resi-
dues were present in most of the collected sedi-
ments in concentrations of approximately 10 to
100
g/kg (ppb, dry weight). Some urban top soils
were found to have more than 1 mg/kg (ppm). In Ash,
as much as 0.5 mg/kg of total DDT (wet weight) was
found in the edible parts. The presence of p,p-DDT
in most of the samples reBects the use of this insecti-
cide against vectors of malaria between 1946 and
1993, which has led to its ubiquitous presence in the
environment of the Brazilian Amazon.  2002 Elsevier
Science (USA)
Key Words: DDT; soils; sediment; bioaccumula-
tion; Ash species.
1947; cited in Wagley, 1953), one of the 7rst uses of
DDT in Brazil in the Amazonian region was in the
control of the vector of malaria in the small town of
Breves in ParaH State, near the delta of the Amazon
river in 1946. Subsequent use of DDT in this region
was widespread not only for malarial control but also
for structural pest control of other insects.
The Ministry of Health of Brazil last purchased
DDT in 1992 (3000 tons). The Ministry reported that
it ended the use of DDT for the control of the vectors
of malaria in 1992 (Oliveira Filho, 1997), but still
permitted its use in the control of leishmaniasis,
which is also endemic in the Amazon region and
other parts of Brazil.
The purpose of this study was to evaluate the
residues of DDT in urban soils, river sediments, and
7sh samples collected at two important goldmining
areas of the Brazilian Amazon, the Madeira River
and the TapajoH s River watersheds (Fig. 1).

INTRODUCTION
In developed countries, mass production of dich-
lorodiphenyltrichloroethane (DDT) began in the
early 1940s. DDT was heavily used in agriculture for
the subsequent 25 to 30 years. By the 1960s, it was
observed that DDT was lethal to 7sh, accumulated
in the food chain, and damaged the eggs of avian
predators such as the peregrine falcon, bald eagle,
and Californian pelicans (Risebrough et al., 1967).
Consequently, DDT was banned for agricultural
uses in most countries beginning in the 1970s (Cur-
tis and Lines, 2000).
1
To whom correspondence should be addressed. Fax: (55-21)-
280-8193. E-mail: jptorres@biof.ufrj.br.
134
MATERIALS AND METHODS
Study Areas
TapajoH s River. The TapajoH s River is situated in
the south of ParaH State, Brazil and is one of the main
tributaries of the Amazon River. The Rato River is
a tributary of the TapajoH s River and ends about
150 km upstream from the city of Itaituba. The
TapajoH s river basin is one of the principal goldmin-
ing areas of the Brazilian Amazon region. It is an
endemic area for malaria and yellow fever, diseases
that are transmitted to humans by mosquitoes
(Anopheles sp. and Aedes sp.). Until 1993, this vill-
age was sprayed with DDT at least two times per
year.

0013-9351/02 $35.00
 2002 Elsevier Science (USA)
All rights reserved.
 DDT IN THE AMAZON IN BRAZIL
FIG. 1. Study areas in the Amazon region in Brazil.
Madeira River. The Madeira River is one of the
most important tributaries of the Right Bank of the
Amazon River. As with the Amazon River itself, its
riverheads are located in the Andean Mountains
near La Paz, Bolivia. The Madeira is formed by the
con8uence of the Beni and the Guapore Rivers and
forms the natural border between Brazil and
Bolivia. The middle portion of the Madeira River is
located at Rondonia State, one of the centers of de-
forestation and malaria incidence in the 1980s. The
cleared areas are currently used for cattle grazing
and for soybean plantation. Porto Velho, which is
located upstream of the sampled area, is one of the
main goldmining centers of the Brazilian Amazon
Region.
Sampling and Analysis
The sampling at the Madeira River occurred in
1997, during a 7-day trip between Manaus, the capi-
tal of the Amazonas State, and Porto Velho, the
capital of Rondonia State. River sediments (n"7)
and village top soils (n"3) were collected from the
most important settlements along the Madeira
River. The soils (n"3) and sediments (n"7) col-
135
lected at the Rato and TapajoH s Rivers were taken in
1994. The analyzed 7sh (n"10), all piscivorous spe-
cies, were collected in the Madeira River basin
(n"7) and the TapajoH s River basin (n"3) between
1994 and 1995, transported in ice boxes (4@C) to
Porto Velho, and shipped frozen (!15@C) to Rio de
Janeiro. The edible parts (muscle tissue) of the 7sh
specimens were freeze-dried (1 g of 7sh"0.2 g of
freeze-dried 7sh, for most of our samples) at the
Radioisotopes Laboratory of the Biophysics Institute
of the Federal University of Rio de Janeiro.
Approximately 500 g of river sediments (0 to 5 cm
depth) and topsoil samples (0 to 5 cm depth) were
collected using a metal scoop attached at the end of
a 4-m aluminum rod. They were stored in acetone-
washed glass jars and maintained at 4@C until analy-
sis. All solvents, which were residue-analysis
quality, and other reagents were obtained from
Merck Co. (Daramstad, Germany). All glassware
was rinsed three times with pesticide-grade acetone
prior to analysis.
Samples were analyzed after continuous sohxlet
extraction (Japenga et al., 1987), using a modi7ed
hot soxhlet for about 2 h (2 ml of 25% cyclo-hexane in
n-hexane). Using a Kuderna}Danish device, the
136 TORRES ET AL.

extract was concentrated to 0.5 ml and redissolved in
2 ml of isooctane. Two to 7ve grams of wet sediment,
soil, or freeze-dried 7sh were placed in the extractor
tube over a 7ne layer of hexane-washed silica gel.
For soils and sediments the percentage water and
the organic content were determined gravimetri-
cally (at 110 and 550@C, respectively). The lipid con-
tent of the 7sh samples was also measured
gravimetrically after gentle evaporation of duplicate
extracts obtained in the same manner.
To remove potentially interfering substances such
as elemental sulfur and humic materials, the isooc-
tane extract was passed through a 7-cm narrow-bore
(1-cm) glass column containing 7 g of Na2SO3}
NaOH}Al2SO4 (11% H2O) (0.6 g}0.07 g}5.53 g}0.77
H2O) (Jensen, 1977; Japenga et al., 1988). Fifteen
milliliters of hexane was used to elute the pesticides
from the cleanup column.
For 7sh samples, the extracted lipids were treated
by two different methods. The 7rst method was
saponi7cation with potassium hydroxide (KOH) in
ethanol. This procedure converts most of the DDT to
DDE and was found to have other limitations also
(e.g., high blank values). In addition, an antifoaming
solution made with NaCl and phosphoric acid was
needed (11.7 g of NaCl in 1 liter of ortophosphoric
acid). The second procedure uses only sulfuric acid
(H2SO4 ) in a single step. This cleanup does not de-
stroy or convert DDT to other metabolites, and the
blanks were found to be superior (FAO/SIDA, 1983;
Berdie and Grimalt, 1998).
Prior to injection, the obtained extract was gently
evaporated down to 0.5 ml in a stream of N2 , and
0<5 ml of internal standard was added for quanti7ca-
tion (octachloronaphtalene; Aldrich Chem. Co.).
A Shimadzu GC-14B with a 63Ni electron capture
detector (ECD) equipped with fused silica capillary
columns (CBP1 and CBP5;5% phenyl-con7rmation
column) and automatic injector (Shimadzu AOC-17)
was used in the analysis. The ECD detection limit
for DDT and related compounds is approximately
0.01 ppb.
To achieve quality control, the aliquots of the
freeze-dried 7sh were reanalyzed at the Wadsworth
Center for Laboratory Research of the New
York State Health Department in Albany, New
York, using an EPA method with soxhlet extraction
and gel permeation chromatographic cleanup
(EPA, 1980).
The Brazilian Radioisotopes Laboratory partici-
pates in the International Atomic Energy Agency
intercomparison exercises for micropollutants in en-
vironmental matrices, using reference materials of
sediment and 7sh.
RESULTS
Soil and Sediment Samples
TapajoH s and Rato Rivers. Residues of DDT and
its metabolites in three soil samples from the Rato
village ranged from 281 to 1224 ng/g (dry weight)
(Table 1). The concentrations of DDT and DDT
metabolites decreased from the village toward the
river. All metabolites of DDT were also found in all
seven sediment samples taken from the small tribu-
taries of the Rato and the TapajoH s Rivers, ranging
from 3.2 to 61.5 ng/g of total DDT (Table 1). The
sediment sample concentrations of DDT metabolites
were much lower than those found in the topsoil. The
DDT concentration at the main channel of the
TapajoH s River was the lowest (total DDT"3.2 ng/g).

TABLE 1
DDT and Metabolites [ng/g d.w.] in Soil and Sediment Samples Collected at the TapajoH s River Basin, Brazil

Soil samples o,p-DDE p,p-DDE p,p-DDD o,p-DDT p,p-DDT Sum of DDT

Rato village ; ; 302 113 809 1224

Main street (toward the river) ; ; 32 68 446 546
Rato River banks 6 ; 187 8 80 281

Sediment samples o,p-DDE p,p-DDE p,p-DDD o,p-DDT p,p-DDT Sum of DDT

Igarape do Lontra 1.9 4.4 8.8 6.4 27.5 49

Igarape Parai 1.5 5 1.2 2.2 10.1 20
Igarape do Rato 1.6 7 2.9 1.1 13.8 26.4
Morro Grande 1.4 3 0.8 0.8 2.8 8.8
Boca do Rato 3.2 7.5 1.6 1.8 14.7 28.8
Mangabal Island 3.5 20.4 5.6 3.7 28.3 61.5
TapajoH s River 0.1 0.7 0.1 0.2 2.1 3.2
 DDT IN THE AMAZON IN BRAZIL 137

TABLE 2
DDT and Metabolites [ng/g d.w.] in Soils and Sediments from the Madeira River Basin, Amazon and Rondonia States

Urbon top soils o,p-DDE p,p-DDE p,p-DDD o,p-DDT p,p-DDT Sum of DDT

Nova Olinda do Norte 54 ; 18 3 41 116

Novo Aripuana 56 3 4 8 52 123
Manicore 2 ; 0.2 0.2 1.6 4

Sediments of the Madeira River o,p-DDE p,p-DDE p,p-DDD o,p-DDT p,p-DDT Sum of DDT

Nova Olinda do Norte 0.4 0.2 0.2 0.06 ; 0.86

Borba 0.06 ; ; ; ; 0.06
Aripuana River 0.02 ; ; ; 0.5 0.52
Boca do Matupiry 0.04 ; ; ; ; 0.04
Fazendinha 0.01 ; ; ; ; 0.01
Humaita 0.01 ; ; ; ; 0.01
Sa o Carlos 0.2 0.4 ; 0.2 0.3 1.1

Madeira River. All three top soils collected from Madeira River were considerably lower than the con-
the villages along the Madeira River were positive for centrations found near the TapajoH s and Rato Rivers.
DDT and some of its metabolites, ranging from 4 to
123 ng/g of total DDT (Table 2). Very low concentra-
Fish Samples
tions of DDT metabolites were found in the bottom
sediments (n"7) of the Madeira River (0.01 to Table 3 provides the weight, length, and percent-
1.1 ng/g of total DDT) (Table 2). The concentrations of age fat for the different species of 7sh used in the
DDT metabolites from the soils and sediments of the
TABLE 4
DDT and Metabolites in Fishes from the Amazon in Braz-
TABLE 3 il [ng/g w.w.], Results from the Radiosiotopes Laboratory,
Piscivorous Fish and Alligator from the Amazon Rio de Janeiro, Brazil
Analyzed for DDT Residues (1994+1995)
H2SO4 Cleanup
Fish species % Fat Weight Length Sample code
Sample code p,p-DDE p,p-DDD p,p-DDT Sum of DDT
Jacare (Cayman
jacare) 2.2 11.5 kg 1.43 m RAPX 1690 RAPX 1690 6 5 32 43
Filhote (Brachyplatys- ROPX 2463 139 8 389 536
toma vaillanti) 13.5 4 kg ; ROPX 2463 SAPX 1579 3 1 7 11
Tucunare (Cichla ROPX 2467 16 9 16 41
ocellaris) 3.7 570 g 33 cm SAPX 1579 ROPX 2460 1 11 15 27
Surubim (Pseudop- AFPX 2599 24 8 209 241
latystoma ROPX 2449 7 2 12 21
fasciatum) 2.8 7 kg ; ROPX 2467 ROPX 2452 13 8 20 41
Surubim (Pseudop- ROPX 2456 2 1 4 7
latystoma ROPX 2473 2 1 20 23
fasciatum) 2.4 ; ; ROPX 2460
Pescada (Plagioscion KOH Cleanup
squamosissimus) 13.9 5 kg ; AFPX 2599
Surubim (Pseudop- RAPX 1690 22 0 0 22
latystoma ROPX 2463 110 0 0 110
fasciatum) 4 3 kg ; ROPX 2449 SAPX 1579 7 0 4 11
Filhote (Brachyplatys- ROPX 2467 5 0 0 5
toma vaillanti) 10.2 12 kg ; ROPX 2452 ROPX 2460 17 1 0 18
Surubim (Pseudop- AFPX 2599 65 2 0 67
latystoma ROPX 2449 6 1 0 7
fasciatum) 2.4 5.6 kg ; ROPX 2456 ROPX 2452 6 0 0 6
Traira (Hoplias ROPX 2456 2 0 0 2
malabaricus) 1.6 775 g 33.5 cm ROPX 2473 ROPX 2473 0 0 0 0
138 TORRES ET AL.

TABLE 5 a densely populated region. Since DDT is better

DDT and Metabolites in Fishes from the Amazon in metabolized under low Eh (redox) potential (low oxy-
Brazil [ng/g w.w.], Results from the Wadsworth Center, gen levels), the presence of many by-products in
Albany, New York
sediments, especially in the TapajoH s River basin,
GPC Cleanup was expected (Parr and Smith, 1974). The low levels
of DDT and its metabolites observed at one cross
Sample code p,p-DDE p,p-DDD p,p-DDT Sum of DDT section of the TapajoH s River channel are probably
due to the high hydrological regimen of this river
RAPX 1690 10 30 76 116
and its remote location from any upstream villages
ROPX 2463 150 160 90 400
SAPX 1579 (8 (8 13 29 where DDT was used.
ROPX 2467 (15 (15 20 50 In 7sh samples, the presence of DDT in higher
ROPX 2460 (10 13 13 36 concentrations than its metabolites is an indication
AFPX 2599 27 19 54 100 of recent application of the pesticide and re8ects the
ROPX 2449 (20 (20 20 60
persistence of the residues from the last period of
ROPX 2452 10 (10 48 68
ROPX 2456 (15 (15 (15 (45 use. This DDT is probably derived from local sour-
ROPX 2473 (10 (10 (10 (30 ces, i.e., weathering of contaminated soil particles
after heavy rains. In comparing the two methods of
extraction, we found that the alkaline reactions of
the saponi7cation method converts most of the DDT
analysis. Results of the DDT analysis of these 7sh into DDE. This procedure is less optimal for ana-
samples are presented in Table 4 (Radioisotopes La- lyzing 7sh from a recently contaminated area such
boratory, Brazil) and in Table 5 (Wadsworth Center, as the Brazilian Amazon. Therefore, the acid
New York State). DDT was present in all of the cleanup is the preferred extraction method, though
analyzed samples. The minimum value was approx- its disadvantage is that organochlorine pesticides,
imately 30 ng/g of total DDT, and the maximum was such as the cyclodienes (e.g., aldrin, endrin, and
500 ng/g. There was wide variation in the DDT con- dieldrin), are lost.
centrations among the 7sh, with certain species (i.e.,
Filhote and Pescada) having 5 to 10 times the DDT CONCLUSION
concentration of some of the other species. There was
a good correlation between the laboratories in Brazil The Ministry of Health used to spray DDT on
and the United States (rp"0.94; r2"0.89, n"10). a semiannual basis inside houses in parts of the
All 7sh samples were analyzed in duplicates at the Amazon Basin. Current environmental levels of
Brazilian laboratory. DDT and its metabolites re8ect the cumulative ac-
tion of a variety of different processes: integration of
DISCUSSION residues from repeated applications, particulate
transport, volatilization, and biodegradation.
The DDT levels found in soil samples in this study This work described the presence of DDT residues
were low compared to historical data obtained from in environmental media from areas located in the
important agricultural areas in Brazil in the 1980s southern Amazon Region in Brazil. Since 1985, agri-
(200}500
g/g), where large amounts of DDT were cultural activities in Brazil have been unable to use
used (Souza et al., 1988). The presence of relatively DDT. Therefore, the presence of this compound in
high DDD levels may be an indication of biodegrada- higher levels than its main metabolites suggests
tion. There is a clear decreasing pattern of concen- that current contamination is derived from the more
trations from the village towards the Rato River. recent use of this organochlorine pesticide against
That pattern, at least in part, may be also explained disease vectors, especially malaria.
by higher evaporation rates of some DDT metab-
olites and enhanced photooxidation that may occur ACKNOWLEDGMENTS
near the deforested riverbanks.
The observed DDT levels in sediments were simi- The authors are indebted to Dr. Suz Claudio and Dr. Mary Wolf
lar to those reported by Souza and co-workers (1988) from The Mount Sinai School of Medicine. This research was
partially funded by the Brazilian Research Council (CNPq-
in sediments from one reservoir in Sa o Paulo State,
PRONEX 877) and by the Biophysics Institute of the Federal
Brazil (1.5 to 7.4 ng/g) (Celeste and Caceres, 1988). University of Rio de Janeiro. The Chromatographic Lab was built
Japenga et al. (1988) found higher concentrations, during the Joint Research Project (C11*-CT93-0055) funded by
up to 150 ng/g, in coastal areas near Rio de Janeiro, the Commission of the European Communities. Dr. Torres is
 DDT IN THE AMAZON IN BRAZIL
a Selikoff Fellow at the Mount Sinai School of Medicine and
Queens College in New York and is supported in part by Grant
1 D43 TW00640 from the Fogarty International Center of the
National Institutes of Health. The authors are indebted to Rod-
rigo Ornellas Meire and Lilian Pereira Amaral Moreira for tech-
nical assistance. We also acknowledge the Geochemical
Department of the Universidade Federal Fluminense (Dr. Luiz
Drude de Lacerda), University of Nevada (Drs. Jerry Mitchell and
Paul Lechter), and the U.S. Forest Service (Dr. John Gutierrez)
for their assistance during the 1997 Madeira River sampling.
REFERENCES
Celeste, M. F., and CaH ceres, O. (1988). ResmH duos de praguicidas
organoclorados no sedimento a represa do Lobo (Broa), Sa o
Carlos, SP. Cie9 nc. Cult. 40, 900}905.
Curtis, C. F., and Lines, J. D. (2000). Should DDT be banned by
international treaty? Parasitol. Today 16, 2000.
FAO/SIDA. (1983). ‘‘Manual of Methods in Aquatic Environment
Research.’’ Fisheries Technical Paper No. 12. Rome.
Japenga, J., Wagenaar, W. J., Smedes, F., and Salomons, W.
(1987). A new, rapid clean-up procedure for the simultaneous
determination of different groups of organic miropollutants in
sediments: Application for two European estuarine sediment
studies. Environ. Technol. Lett. 8, 9}20.
Japenga, J., Wagenaar, W. J., Salomons, W., Lacerda, L. D.,
Patchineelam, S. R., and Leitao-Filho, C. M. (1988). Organic
139
micropollutants in the Rio de Janeiro coastal areas in, Brazil.
Sci. Total Environ. 75, 249}259.
Jensen, S., Renberg, L., and Reutegardh, L. (1977). Residue anal-
ysis of sediment and sewage sludge for organochlorines in the
presence of elemental sulfur. Anal. Chem. 49, 316}318.
Oliveira Filho, A. M. (1997). ‘‘General Overview on Vector Control
in Relation to the Organic Insecticide Pollution in Brazil.’’
International Workshop on Organic Micropollutants in the En-
vironment, IBCCF-UFRJ. Nota No. 1097 AB-Dlo, Ministry of
Agriculture, The Netherlands.
Parr, J. F., and Smith, S. (1974). Degradation of DDT in an
Everglades muck as affected by lime, ferrous iron and
anaerobiosis. Soil Sci. 118, 45}52.
Risebrough, R., Menzel, D. B., Martin, D. J., and Olcott, H. S.
(1967). DDT residues in Paci7c sea birds: A persistent insecti-
cide in marine food chains. Nature 216, 589}591.
Setterberg, F., and Shavelson, L. (1993). ‘‘Toxic Nation.’’ Wiley,
New York.
Simonitch, S. L., and Hites, R. A. (1995). Global distribution
of persistent organochlorine compounds. Science 269,
1951}1954.
Souza, N. E., Rubira, A. F., Matsushita, M., and Tanamati, A.
1988. ResmH duos de pesticidas organoclorados em amostras
ambientais (aH guas e solos) do municmH pio de MaringaH . Arquiv.
Biol. Tecnol. 31, 587}594.
Wagley, C. (1953). ‘‘Amazon Town: A Study of Man in the
Tropics.’’ Macmillan Co., New York.