The ceramic artifacts in archaeological black earth (terra preta) from ...

375 VOL. 34(3) 2004: 375 - 386

The ceramic artifacts in archaeological black earth(terra preta) from Lower Amazon region, Brazil:chemistry and geochemical evolution.

Marcondes Lima da COSTA1 *, Dirse Clara KERN2, Alice Helena Eleotério PINTO1, Jorge Raimundo daTrindade SOUZA 3

ABSTRACTThis paper carried out a chemical investigation of archaeological ceramic artifacts found in archaeological sites with BlackEarth (ABE) in the Lower Amazon Region at Cachoeira-Porteira, State of Pará, Brazil. The ceramic artifacts, mostly of daily use,belong to Konduri culture (from 900 to 400 years BP). They are constituted of SiO2, Al2O3, Fe2O3, Na2O and P2O5; SiO2 andAl

2O

3 together add up to 80 % and indicate influence of acid rocks, transformed into clay minerals basically kaolinite. The

relative high contents of P2O

5 (2.37 % in average) come out as (Al,Fe)-phosphate, an uncommon fact in primitive red ceramics,

but found in some roman and egyptian archaeological sites. The contents of the trace elements are similar or below theEarth’s crust average. This chemical composition (except P

2O

5) detaches saprolite material derived acid igneous rocks or

sedimentary ones as the main raw material of the ceramics. The contents of K, Na and Ca represent the feldspars and rockfragments possibly introduced into saprolitic groundmass, indicated by mineralogical studies. The presence of cauixi andcariapé as well as quartz sand was confirmed by optical microscope, SEM analyses and by the high silica contents of ceramicfragments. Phosphorus was possibly incorporated into groundmass during cooking of foods, and ABE soil profile formationdeveloped on yellow Latosols. The raw materials and its tempers (cauixi, or cariapé, feldspar, crushed rocks, old ceramicartifacts and quartz fragments) are found close to the sites and therefore and certainly came from them.

KEY WORDSTerra preta, Black earth, Lower Amazon, Soil, Chemistry, Phosphorus, Geochemical Evolution

Artefatos cerâmicos em sítios arqueológicos com terrapreta na região do baixo Amazonas, Brasil: composiçãoquímica e evolução geoquímica.

RESUMONeste trabalho realizou-se a caracterização química de fragmentos de artefatos cerâmicos encontrados em sítios arqueológicoscom terra preta no Baixo Amazonas (Cachoeira-Porteira, Pará, Brasil), representativos da cultura Konduri (de 900 a 400anos AP). Esses fragmentos são constituídos de SiO

2, Al

2O

3, Fe

2O

3, Na

2O e P

2O

5, sendo que SiO

2 e Al

2O

3, juntos, perfazem mais

de 80 % em peso. Os teores de P2O

5 são relativamente elevados (2,37 % em média) sob a forma de (Al,Fe)-fosfatos, incomuns

em cerâmicas vermelhas primitivas, mas encontrados em algumas cerâmicas arqueológicas egípcias e romanas. Asconcentrações dos elementos traços são comparáveis ou mesmo inferiores ao nível crustal, embora a composição total sejapróxima a mesma. A composição química (exceto P

2O

5 ) em conjunto com os dados mineralógicos e texturais indicam

material saprolítico derivado de rochas ígneas félsicas ou rochas sedimentares como matéria-prima das cerâmicas. Osteores de K, Ca e Na mostram que os feldspatos e fragmentos de rochas foram adicionados ao material argiloso, comosugerido pela mineralogia. Os altos teores de sílica respondem pela presença de cauixi, cariapé e/ou areias quartzosas.Fósforo deve ter sido incorporadoà matriz argilosa da cerâmica, quando do cozimento de alimentos nos vasos cerâmicos, eainda, em parte, durante a formação do perfil de solo tipo ABE sobre Latossolos Amarelos. A matéria prima e os temperos(cauixi, cariapé, rochas trituradas e fragmentos de vasos cerâmicos descartados) encontram-se disponíveis próximos aossítios até a atualidade, e portanto foram a área fonte dos mesmos para a confecção dos artefatos cerâmicos..

PALAVRAS-CHAVETerra preta, Baixo Amazonas, Solos, Composição química, Fósforo, Evolução geoquímica.

1 Geosciences Center/UFPa, 66075-110 Belém-PA, Brazil, P.O . Box 1611 Fax: (091) 211 1609/211 1428, Phone: (091) 211 1428/ 249 5028, E-mail: [email protected] Museu Paraense Emílio Goeldi/MPEG, Belém-PA, Brazil.3 Chemistry Department/UFPa, 66075-110 Belém-PA, Brazil

376 VOL. 34(3) 2004: 375 - 386 • COSTA et al.

THE CERAMIC ARTIFACTS IN ARCHAEOLOGICAL BLACK EARTH (TERRA PRETA) FROMLOWER AMAZON REGION, BRAZIL: CHEMISTRY AND GEOCHEMICAL EVOLUTION

INTRODUCTION

After Costa et al., (2004) the black earth soils, calledby natives Indian black earth (Terra Preta de Índio inPortuguese) or by archaeologists as archaeologicalblack each (ABE) are very commonly found in thelandscape of the Amazon region and described bynaturalists, archaeologists and pedologists (Smith, 1879;Ranzani et al., 1962; Sombroek, 1966; Balee, 1986; Kern& Kampf, 1989, 1990). Pedological studies andmorphological description and classification of theirceramics artifacts as important tools for understandingthe peopling of the Amazon region were carried out(Ranzani et al., 1962; Falesi, 1974; Simoes, 1982; Edenet al., 1984; Kern (1988), Kern & Kampf (1989 E 1990),Kern et al., (1997). Multi-element geochemistry wasconducted at Caxiuana bay by Kern (1996), Kern & Costa,(1997) and Costa & Kern (1999). Although very abundantin the ABE sites, the fragments of ceramics artifacts havenot yet been investigated. The first attempt was realizedby Costa et al., (2004), who studied the mineralogy ofthe fragments of ceramic artifacts from Cachoeira-Porteira. The present work carried out a chemicalcharacterization of these materials, which represent theKonduri style culture (Nimuendaju, 1948), in order tocomplete the previous investigation of Costa et al.,(2004) and to contribute to understand the pre-historicpeopling of the Amazon region. The results obtainedunti l now show that the black earth may be apedological, chemical and mineralogical modificationof preexisting soils by pre-historic human (Indian)act iv i t ies and sett l ing in the later past t ime.Mineralogical and geochemical techniques have beensuccessfully applied in the last decades to studyarchaeological artifacts in order to improve theinterpretation of the archaeological data (Noll, 1978;Zaun, 1982; Letsch & Noll, 1983; Momsen, 1986,Freestone & Middleton, 1987; Redmont, 1996; Strazicich,1999; Costa et al. 1991; Costa & Kern, 1996 ).

Along the Trombetas river and its tributaries, there areseveral archaeological sites (Ranzani et al., 1962; Kern,1988). Close to the village Cachoeira-Porteira (OriximináMunicipality, state Pará), the investigated area on theTrombetas river, occu sever site of ABE (Fig.1).

The actual climate is rainy equatorial, with tropical forest.The drainage system is dense, the valleys are overflowed,and normally have a rocky bottom, and under theseconditions water falls and/or rapids are present.

The soils are composed mainly by yellow to red Latosols,yellowish brown Podzols and Structured Red Earth (TerraRoxa Estruturada). The ABE are classified as anthropogenicsoils (Ranzani et al., 1962, Kern, 1988; Kern, 1996),developed primordially over the llow Latosols andPetroplinthosol (Kern, 1988: Kern & Kaempf, 1989).

Paleozoic to Mesozoic sedimentary rocks (sandstonesand claystones) of Lower Amazon basin covering the

proterozoic volcanic acid rocks of Iricoumé Formationand the Mapuera Granite constitute the main geologicalunits of the region and are crosscut by basic dikes.Laterite profiles caped by Latosols are also common, aswell as holocene clays, sands and pebbles on theriverbanks and lakes (Costa et al., 2004).

MATERIALS AND METHODS

Sampling

The fragments of ceramic artifacts of ABEs with distincttempers investigated here were collected by MPEG (MuseuParaense Emilio Goeldi) in 1988, with the participation of co-author Dirse Clara Kern in antecipation of the lake projectedfor hydroelectric power of Cachoeira-Porteira. For the presentwork 51 samples of ceramic fragments covering all the types oftempers were taken from the sites 1 (PA-OR-73) and 2 (PA-TR-2) (Fig.1). After Meggers & Evans (1970) tempers are exoticmaterial intentionally put in the ceramics during theirconfection. The identified tempers are mainly ceramicfragments with cariapé; ceramics with cauxi; ceramics withsand; ceramics with sand plus feldspar; and ceramics with oldercrushed ceramics. The sampling was performed along standardtrenches along the soil horizons Ap, A1, A2 and A3, the horizonsthat carry ceramic fragments. The mineralogical study of thesematerials was carried out by Costa et al (2004).

Chemical Analyses

Whole chemical analyses of 22 samples of ceramicfragments with distinct tempers were performed by XRFspectrometry, atomic absorption spectrometry and wetchemical methods in the Centro de Geociências fromUniversidade Federal do Pará-UFPa. XRF, OES (opticalemission spectrograph), and ICP-AES were also performedin four selected samples for major and trace elements atLakefield-Geosol (Belo Horizonte, Brazil). These four sampleswere also investigated by Scanning Electron Microscope withEnergy Dispersive System at the laboratory facilities of theCompanhia Brasileira de Metalurgia e Mineração – CBMM,in São Paulo, in order to get images and chemical analyses.

RESULTS AND DISCUSSION

Chemical Composition

To help understanding the chemical treatment anddiscussion of chemical analyses it should be very importantto summarize the physical composition of the ceramicfragments after each temper described by Costa et al. (2004).Quartz and burned clay-derived mineral (originallyinterpreted to be kaolinite) are the more abundant minerals.Feldspars (albite and microcline) are frequent in ceramic

377 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


fragments with sand and feldspar tempers. Illite, hematite,maghemite, goethite, and anatase are found in low quantitiesfor all ceramic fragment tempers. Although in less quantities(3 to 11 wt.%) amorphous and microcrystalline aluminumphosphates are found all over the ceramic fragments. Cauxiand cariapé are the organic tempers found in several ceramicfragments, and after their presence are the ceramic fragmentsclassified into cauixi or cariapé temper.

The results of the chemical analyses (Table 1) show smallvariation among the average of the different tempers. Thisis clearly demonstrated by diagrams SiO

2- Al

2O

3 - K

2O (Fig.

2) and Fe2O

3 - P

2O

5 – TiO

2 (Fig. 3), where most samples fall

close together, the ceramic fragments with sand+feldspartemper being less similar.

SiO2 values vary from 63.5 to 70.8 wt. % (the average of each

ceramic temper), and those from Al2O

3from 14.7 to 17.4 wt %, and togethercomprise more than 80 wt.% of wholeceramic chemical composition. Fe

2O

3 is

the third most abundant element andreaches 5.79 Wt. % on average. The alkalielements (as K

2O, Na

2O, CaO and MgO)

constitute together less than 3 wt. %. Onthe other side come out the relative highcontents of P

2O

5, 2.37 wt. % in average.

The H2O+ values reach 7.54 wt. % on

average, which is still high for burnedceramic material (mainly clay material).The Al

2O

3 and H

2O+ contents explain the

great abundance of clay-derived minerals..The relative high contents of H

2O+

confirm the abundance of partialdehydroxylation of clay-material derivedfrom kaolinite as main mineral of theceramics and the neoformation ofkaolinite (Costa et al., 2004). It means thatthe fragments reabsorved water along itsdiscarded and incorporation into ABE-soilformation. The extremely high SiO2contents correspond, besides clay-derivedmaterial and the abundance of quartz, assand grains and rock fragments, to thepresence of cauixi (Tubella reticulata andParnula betesil), and cariapé(Bignomiacea, Moquilea, Licania utilis,and Turiuva; Hilbert, 1955). Cauixi andcariapé are the SiO

2-bearing organic

materials (Costa et al., 1991, 2004). Theiron contents represent hematite andgoethite, and some maghemite, mineralsalso identified in the studied ceramicfragments. K

2O, Na

2O and CaO built the

feldspars (microcline and albite) andtogether with MgO indicate the presenceof illite in the raw material. Thepredominance of SiO

2, Al

2O

3 and H

2O+

plus Fe2O

3 (together they make up more

than 93 Wt. % of whole sample) confirm clay-derived minerals,quartz and some iron oxyhydroxides as the main minerals ofceramic fragments . The chemical and mineralogical data, aswell as textural aspects, leave one to conclude that the main rawmaterial for ceramic elaboration comes from fine-grained clay-quartz rich material (for example saprolites, claystones, shales,and so on). The P

2O

5 contents however are relatively high for

clay material normally used for ceramic purpose. They composebasically the amorphous to criptocrystalline (Al,Fe)-phosphate.The H

2O+ contents are still high (5.6 to 8.9 Wt %) showing the

rehydration of the ceramic vessel after theeir discharging andthe formation of the soil with black earth, the formation of ABE,and as consequence the neoformation of kaolinite and goethiteafter Costa et al. (2004).

Figure 1 – Location map of two sites of archaeological black earth (ABE) investigated atCachoeira-Porteira, Lower Amazon region (after COSTA et al., 2004).

378 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


Tabl

e 1

- Who

le c

hem

ical

ana

lyse

s of

AB

E-ce

ram

ics

afte

r th

eir

tem

pers

(W

t.%).

OiS2

OiT2

lA

2O3

eF2O

3OeF

OgM

OaC

aN

2OK

2OOa

BP 2O

5On

MOnZ

H2O

+O2

H-

OC

2

IXIU

AC

1091.37

97.048.51

74.4-

82.114.0

-16.1

-74.0

4.9392.242

48.196.0

-

9187.16

19.048.81

60.4-

97.023.0

-13.1

-20.3

3.1338.413

87.0199.3

-

*91

08.0678.0

03.8104.6

37.093.0

12.091.0

02.198.0

01.3-

-98.9

-03.0

3250.76

09.058.61

36.4-

85.083.0

-48.0

-98.1

1.2523.732

62.859.3

-

2379.46

20.135.71

49.3-

85.024.0

-62.1

-64.2

1.5133.502

31.958.3

-

5321.27

96.093.31

98.3-

95.052.0

-59.0

-75.2

9.6512.712

19.664.3

-

EG

AREVA

56.6668.0

97.6175.4

37.007.0

33.091.0

02.198.0

52.20.933

4.34208.7

71.303.0

ÉPAI

RA

C

4062.26

89.069.61

95.7-

67.034.0

-93.0

-30.3

6.5023.143

35.942.4

-

*40

81.7569.0

03.8104.6

03.175.0

35.064.0

74.041.0

03.3-

-18.9

-04.0

8016.06

20.143.71

99.7-

18.014.0

-66.0

-50.3

4.8613.672

70.991.3

-

4157.06

20.152.81

81.8-

57.004.0

-97.0

-62.3

7.0414.762

60.821.3

-

5211.66

48.099.51

68.6-

74.063.0

-96.0

-94.2

1.5026.502

07.700.4

-

0481.56

59.006.51

77.6-

17.083.0

-54.0

-28.2

1.7511.741

26.875.3

-

1472.86

47.093.31

77.5-

03.003.0

-31.0

-51.3

8.2021.05

00.919.2

-

9462.36

20.124.61

25.8-

03.082.0

-54.0

-30.2

8.396.331

74.980.3

-

EG

AREVA

59.2649.0

35.6162.7

03.185.0

93.064.0

05.041.0

25.23.351

0.30219.8

44.304.0

DN

AS

5022.27

18.062.31

18.4-

95.063.0

-95.0

-90.2

6.3028.612

56.603.3

-

2135.27

57.088.21

06.4-

27.053.0

-46.0

-91.2

1.4210.452

96.488.2

-

6151.86

68.032.61

76.3-

13.033.0

-92.1

-18.1

1.7129.531

88.753.2

-

2292.96

89.095.61

88.2-

61.062.0

-34.0

-80.3

7.6627.951

93.754.3

-

EG

AREVA

97.0758.0

47.4199.3

-44.0

23.0-

47.0-

92.21.302

6.19156.6

99.2-

RAPS

DLEF+

DN

AS

2091.26

38.034.71

66.7-

07.077.0

-24.1

-25.2

0.2719.142

94.611.3

-

8108.16

87.049.71

39.6-

03.148.0

-19.1

-41.1

7.2051.342

23.694.3

-

379 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


OiS2

OiT2

lA

2O3

eF2O

3OeF

OgM

OaC

aN

2OK

2OOa

BP 2O

5On

MOnZ

H2O

+O2

H-

OC

2

RAPS

DLEF+

DN

AS

4210.96

17.066.51

47.4-

08.055.0

-18.1

-41.1

5.2028.843

68.447.2

-

*42

09.7666.0

06.5102.3

03.174.0

44.007.1

07.141.0

02.1-

-99.4

-02.0

8242.16

48.082.02

07.5-

36.042.1

-85.1

-21.2

8.1023.461

92.524.2

-

EG

ARE

VA

34.4667.0

83.7156.5

03.187.0

77.007.1

86.141.0

26.17.962

5.94295.5

49.202.0

DN

AS+

ÉPAI

RA

C

7073.66

98.078.51

58.8-

18.034.0

-27.0

-35.2

4.3138.262

21.703.3

-

5195.56

68.026.51

28.6-

36.033.0

-84.0

-08.2

8.9125.632

33.805.3

-

*51

06.4697.0

04.5102.5

01.114.0

12.024.0

54.067.0

03.3-

-70.8

-03.0

EG

ARE

VA

25.5658.0

36.5192.6

01.126.0

23.024.0

55.067.0

88.26.662

6.94248.7

04.303.0

SREP

MET

LLA

EG

ARE

VA

55.5668.0

73.6197.5

11.136.0

34.096.0

09.084.0

73.22.452

7.48145.7

12.303.0

TSUR

CS'

HTRAE02.55

06.103.51

08.208.5

02.508.8

09.209.2

50.003.0

0002

Identification of the main raw materialafter chemical composition

The diagram SiO2-Al

2O

3-10(K

2O) (Fig.2) show that

ceramics have a single field composition, nearest to the fieldof granitic saprolite and then to shale and the Earth’s crustaverage. On the other hand the average of each ceramicfragment temper studied here are aligned in a single fieldparallel to the P

2O

5-Fe

2O

3-edge of the diagram Fe

2O

3-TiO

2-

P2O

5 (Fig. 3), far from the Earth’s crust composition,

saprolite derived from granite and shale. This deviation iscaused by the anomalous contents of P

2O

5 and indicates

that P2O

5 was not a part of the original material but was

probably incorporated after the elaboration of the ceramicvessels, certainly during the their daily use and/or afterwardthe discard of them.

The contents of trace elements are similarly distributedinside of different ceramic tempers in the same way as themajor elements. It means that the ceramic vesselsindependent of ceramic tempers were made of quite similarraw materials. If these elements are compared to theiraverage in the Earth’s Crust they show the following pattern:a) elements with contents clearly above the Earth’s crustlevel: Zn, Ba, Pb, Zr, Y, Mo, Sc and Sn e; b) contents similarlyto the Earth’s crust average : Cu, B, La, Ce, Nd and Eu; c)contents below the Earth’s crust average : Cl, Co, Ga, V, Y,Mn, Sm, Gd, Dy, Ho, Er, Yb and Lu).

The organic foreign materials, cauixi and cariapé, display avery low content of trace elements, specially cariapé (Table 2).Only Ba reaches 152 ppm in cariapé, although in the ceramicit can reach up to 0.89 wt. % of BaO. This BaO values is toohigh even for feldspar-bearing rocks. Locally Ba, sometimesMn, are strong concentrated in the matrix as showed previouslyby SEM/EDS analyses. They form their own mineral, interpretedto be Mn-oxyhydroxides(Costa et al., 2004). Barium,phosphorus and even Pb seem to be the anomalous elementsfound in the ceramic fragments of Cachoeira-Porteira. Theanomalous values of Ba were frequently found in the ceramicfragments with cauixi and cariapé+sand temper.

In general the contents of trace elements found in theceramic fragments of Cachoeira-Porteira, can be wellcorrelated to saprolites derived either from granites/rhyolitesor to granodiorites/dacites. They can be slightly correlated tothe Earth’s Crust average as well. The higher contents of Zr,Sc and Sn might have been promoted by weathering processesforming the saprolite, confirmed by the decreasing of Cl, Co,Cu, V, Mn and heavy rare earth elements (Sm, Gd, Dy, Ho, Er,Yb and Lu) as demonstrated before. The contents of the rareearth elements (REE) fall in the domain of the Earth’s crustabundance and granitic rocks and its weathered products.

The distribution curves of REE normalized tochondrites (Fig. 4) are quite similar to each other anddisplay a negative Eu-anomaly and depletion toward HREE.This pattern falls close to granite-derived saprolite and isonly partially comparable with the Earth’s crust and graniteaverage. Cauxi and cariapé present a distinct pattern withco

nt. t

able

1

380 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


The raw material of the ABE-ceramics

Petrographical and chemical studies have been used byRedmount (1996) and Strazicich (1998) to identify the rawmaterial and the source of ceramics from Egypt and Mexico,respectively. The chemical composition data obtained hereand the textural and the mineralogical presented by Costa etal., (2004) allow us to propose the following materials asraw materials for the ceramics of Cachoeira-Porteira:

1) Clay materials (mainly kaolinite) from the saprolitehorizons or mottled zone of weathered fine-grained acidrocks, for example rhyolites and/or fine-grained granites oreven clayey sedimentary rocks (shale). The high contentsof Al

20

3 and SiO

2 are compatible to kaolinite-rich saprolite

or mottled zone derived from that kind of rocks;2) Slightly coarse crushed igneous rocks, e.g.

granites, granodiorites, rhyolites and quartz of veins; theyexplain the presence of unweathered feldspars androundless quartz grains (Costa et al., , 2004) and therelatively high content of K, Na and Ca for the inferredsaprolite. The addition of quartz grains , cauixi andcariapé has only promoted a slight unbalancing in SiO

2relatively to Al

20

3 turned the composition a little far from

Earth’s Crust level (Fig. 2). In contrast, the feldspar grains

very low concentration of REE and therefore do notcontribute to relative high concentration of REE found inthe ceramic fragments.

Unlikely the saprolite formation can not promote thestrong increasing of the contents of elements like Ba, Znand Pb (expecting a weathering of acid igneous rocks orcommon clay sedimentary rocks). These elements similarlyto phosphorus, La, Ce, Nd and Mn as earlier described,they (mainly Ba and Pb) were probably introduced afterceramic elaboration.

The major and trace elements indicate that the main rawmaterials used for the elaboration of the ceramic artifacts havea composition close do saprolite derived from acid rocks (fine-grained granite/rhyolite) or shales and claystones. However,elements like P, Ba and Pb (Zn) and their minerals present asforeign components of the raw materials. They may have beenpossibly incorporated only after the elaboration of the ceramicvessels, probably during cooking and other daily uses.

THE ORIGIN OF CERAMIC RAW MATERIALAND THEIR TRANSFORMATION DURING USEAND DISCARD (AS ABE COMPONENT)

Figure 3 - Chemical composition of the ceramics on the diagramFe

2O

3-TiO

2-P

2O

5.

Figure 2 - Chemical composition of the ceramics on the diagramSiO

2-10K

2O-Al

2O

3 compared to the composition of Earth’s crust

abundance, granite, granodiorite, shale and granitic saprolite

381 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


obviously kept some Na2O, CaO, and K

2O inside the clayey

matrix, which can be observed in the chemical analysesof these ceramics (Table 1);

3) Cauixi and/or cariapé. They cannot be foundnaturally in saprolite or mottled zone. Cauixi can beobserved in some young lake sediments and watercolumn e.g. Caxiuanã bay in Eastern Amazon; cauixi arefound as cm-long aggregates fixed in the trees on themargin of rivers and lakes with clear or black water, aswell as dsitributed on the surface of bottom sedimentsof lakes and some black water rivers. Cariapé is a treeskin found in the neighboring forest. Both materials richon SiO

2 are common in the Amazon region and in the

area studied here.4) Older discarded ceramics. They possibly represent

discarded ceramic vessels, that were intentionally finecrushed for reuse. The fine fragments obtained can be wellcharacterized under optical microscope;

The texture of ceramics shows that the clay-derivedmatrix was the basic raw material into which crushed rocks

and/or crushed minerals (feldspars and quartz), quartz sand,cauixi and/or cariapé and/or older ceramics wereintentionally added. After the archaeologists the last fourmain components (crushed rocks, and minerals; quartzsand; cauixi und cariapé; and old ceramics) intentionallyadded are denominated tempers (Meggers & Evans, 1970)and used to classify the archaeological ceramic fragments.

On the other hand there is no evidence of phosphorus-and barium-lead-bearing raw material .

The Source Area of Raw Material

Weathered and unweathered rhyolites, granitic andgranodioritic rocks occur quite near to the sites 1 and 2(COSTA et al., 2004). Cauixi is found to the present day inmany lakes and rivers near archaeological sites and in manyplaces of the Amazon region. Cariapé can be found in thesurrounding rain forest (Fig. 5) as well. The region is alsorich in archaeological sites with ABE and potteries, that canbe recycled.

Table 2 - Trace element contents (in ppm) of ABE-ceramics after their tempers, and of cauixi and cariapé.

SELPMAS lC oC uC eB B rZ aG oM V cS Y nS bP nZ rS

SCIMAREC

40-PCC 65 9 27 2< 01< 083 5< 251 71 05 5< 54 281 031

51-PCC 4 8 64 2< 01< 573 5< 611 71 84 5< 54 231 68

91-PCC 84 8 04 2< 01< 573 5< 251 61 86 5< 06 821 24

42-PCC 92 01 63 2< 01< 085 5< 081 31 24 5< 04 231 004

EGAREVA 43 9 94 2< 01< 824 5< 051 61 25 5< 84 441 561

IXIUAC 12 - 91 72 5< 5 021 5< 21 - 22 002< -

ÉPAIRAC 01 - 91 01< 5< 5< 36 5< 01< - 01< 002< -

S'HTRAE*TSURC

031 22 05 3 01 561 81 5.1 011 02 53 5.2 5.21 07 573

SELPMAS aB rC nM iN aL eC dN mS uE dG yD oH rE bY uL

SCIMAREC

40-PCC 04.54 25.59 70.73 514.5 070.1 637.3 768.2 575.0 365.1 316.1 52.0

51-PCC 25.93 62.38 69.92 550.5 429.0 385.3 191.3 556.0 248.1 029.1 52.0

91-PCC 31.23 02.96 69.62 418.4 958.0 833.3 530.3 426.0 477.1 868.1 42.0

42-PCC 28.32 27.15 45.22 749.3 790.0 927.2 433.2 884.0 624.1 114.1 02.0

EGAREVA 22.53 48.47 31.92 018.4 049.0 053.3 758.2 585.0 156.1 017.1 42.0

IXIUAC 44 81 85 91 549.5 02.11 77.3 083.0 801.0 523.0 292.0 050.0 031.0 271.0 30.0

ÉPAIRAC 251 22 52 71 458.9 55.81 82.3 006.0 760.0 013.0 532.0 240.0 011.0 821.0 20.0

S'HTRAE*TSURC

005 001 000,1 57 53 07 03 7 2.1 7 6 5.1 5.3 5.3 6.0

382 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


Origin of Phosphorus and Phosphates andOther Elements like Mn and Ba.

Phosphorus is one of the few chemical elements, whichare completely anomalous in the investigated ABE-ceramicartifacts. The contents of P

2O

5 range from 1.7 to 2.8 wt. %

(average of each tempers) and are represented basically byvariscite-strengite and amorphous phosphates; apatite andrhabdophane are rare. Freestone & Middleton (1987)reported contents of P

2O

5 in similar range within the clay

matrix in some Far East ceramics. There the clay source andlocal soils contain less than 0.5 wt. %. The local soils ofCahoeira-Porteira also show low P

2O

5 values, less than 0.036

wt. % (Kern, 1988). In the modern ceramic artifacts, calciumphosphates are added to act as flux and to vitrify the finalceramic product. However, the ancient ceramics with highP

2O

5 content are not vitrified, so in this case, P

2O

5 was not

used for this purpose (Freestone & Middleton, 1987).Freestone & Middleton conclude that P

2O

5 content is a result

of environmental contamination. The mineral, a calciumphosphate but not apatite, is very fine-grained. In Cachoeira-Porteira, the ceramic is not vitrified and the phosphoruscontents in the ceramic fragments do not correlate withCaO, under relatively low contents. Phosphorus forms (Al-Fe)-phosphates alone.

The ceramic fragments represent potteries of daily use,for cooking and stocking foods. The diets of the ancientIndians were similar to that of todays, basically of animaland vegetal origin. This diet includes meals make up ofbones, liver, crumb, bowels, etc. as well as many differenttypes of roots such as manioc, potatoes, yam, fruits of palmtrees, different chestnuts (Brazil nut, and so on). Most ofthese foods may contain a high content of P

2O

5 as well as

Mn, Ba, Mg, Zn, (Pb) (Scholz, 1980). The meals with bonesand fishes, as well as nuts, were certainly the most importantsource of phosphorus. During long and repeatedly cookingand stocking of food there was transference of foodcomponents into ceramic vessels, that were even porous.During the cooking with much boiling water, whichcorresponds to a true high temperature water solution(hydrothermal solution), rich in the above mentionedelements released from the foods (Fig.6), like phosphorus,they go into the solution (broth or soup), and react withthe clay-derived groundmass, disordered or amorphous.Disordered or amorphous clay or even crystalline clayminerals and iron oxyhydroxides react with phosphorusforming aluminous and iron phosphates such as variscite-strengite, crandallite group, wavellite, and so on (Kittricket al., 1955; HSU, 1968; Kafkafi et al., 1967; Yariv et al.,1979 and Costa, 1982) . In the archaeological ceramicfragments of Cachoeira-Porteira on average 94% of thephosphorus is absorbed (Table 3) in the mineral structure,confirming the presence of the phosphate phases. Thekinetic of the phosphates formation increases underhydrothermal condition, and the cooking of food may beassigned as hydrothermal environment. The reactions forphosphate formation can be written as follows.

Figure 4 - REE concentrations normalized to chondrites andcompared to Earth’ crust abundance, granite, granitic saprolite,and cauixi and cariapé. (1)(2) KRAUSKOPF (1982); (3) DUDDY(1980). Chondrite REE were normalized after abundances ofchondrites in WAKITA et al. (1971).

Figure 5 – Simplified geological cross-section showing the different sources of raw materials of archaeological potteries from Cachoeira-Porteira. After COSTA et al., (2002).

383 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


Al2Si

2O

5(OH)

4(s) + 2H

3PO

4(aq) + 3H

2O ] 2AlPO

4.2H

2O

+ 2Si(OH)4(aq)

or SiO2 .H

2O

(s) (1) Kaolinite (clay) in boiling

water variscite opal (ceramic body) water (broth orsoup) (end-use of ceramics) Fe

2O

3 + 2 H

3PO

4(aq) + H

2O

] 2FePO4 .2H

2O

(s) (2) hematite strengite.

H3PO

4 comes possibly from the desagregation and

decomposition of organic material mainly bones andfishes during cooking, besides vegetal diet. Theanomalous concentrations of Ba, Zn and Pb (also Zr, Scand Mo ) in whole ceramics and of Ba and Mn in someplaces of clay-derived matrix might have come fromanimal and vegetal foods and precipitated as (Ba,Mn)-oxyhydroxides, similarly to phosphates. This makespossible the potteries contamination by foods duringcooking (Fig.6). This process might be responsible forpromoting the formation of phosphates, manganese andbarium oxides as mineral equivalents. The short timeof the reactions (no older than the culture itself, 900 to400 years BP) and low temperature (cookingtemperature) can explain the low crystallinity and thevery fine-grained nature of these neoformed minerals.I t is possible that this kind of react ions (andcontamination) may continue after the discarding of theceramics (wastes) and during their incorporation intoanthropic soil profile developed at the time of thesettlement and after the abandonment of it (Fig.6), the

formation of the black earth (ABE). The formation ofphosphates, and locally of Ba, Mn and REE mineralsequivalents suggest that the pre-historic settlements atCachoeira-Porteira were occupied by gathering andhunting people who stayed for a long time in the site,and possibly came back to the site several times..

The Geochemical Evolutionof ABE-Ceramics

900 to 400 years BP the pre-historic peoples lived insmall groups along the margins of Trombetas/Nhamundárivers and their lakes. For their daily necessities oftransport, water keeping and food cooking, they mademany ceramic artifacts (potteries). The majority of theartifacts were slightly painted and burned. The rawmaterials for making the potteries came certainly fromneighboring areas close to sites and include clay materialfrom a saprolite or mottled zone derived from rhyolite/granite or from shales. Other materials are cauixi, cariapéand crushed rocks. All these materials are abundant inthe area. This is the motive we assumed that the pre-historic peoples collected the raw materials around thearea, close to the sites. The ceramic matrix is composedof Si, Al, Fe, Ti and Mg (Fig. 6), which shows its still clayeycomposition. To improve the plasticity properties of this

SELPMAS )mpp(SUROHPSOHPELBASOPSIDLATOT

)mpp(SUROHPSOHPSUROHPSOHPELBASOPSIDFO%

SUROHPSOHPLATOTOT

52-PCC 893 677,01 7.3

62-PCC 085 142,71 4.3

72-PCC 0012 699,81 1.11

82-PCC 248 896,7 9.01

92-PCC 153 337,51 2.2

03-PCC 323 094,01 2.2

13-PCC 981 556,41 3.1

23-PCC 418 206,9 5.8

33-PCC 192 206,7 8.3

43-PCC 552 762,9 7.2

53-PCC 3571 641,01 3.71

63-PCC 567 918,5 1.31

73-PCC 004 397,31 9.2

83-PCC 182 360,9 1.3

93-PCC 473 798,6 4.5

04-PCC 415 836,11 4.4

EGAREVA 936 212,11 0.6

Table 3 - Disposable and total phosphorus in ABE-ceramics.

384 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


matrix the Indians intentionallyadded either cauxi, cariapé, quartzsand or crushed rocks and feldspars,that promoted a SiO

2-enrichment, sn

important chemical aspect of thetempers. The addition of feldspars,which obviously contributes with K,Na and Ca into raw material improvesthe firing temperature.

The firing of the potteries occurin open atmosphere (Fig.6), possiblyin the same primitive way as theactual Indians and caboclos still do.The firing temperature didn’t exceed600 ºC, as demonstrated by partialdehydroxylation of clay material andthe formation of maghemite. Thisphase formed the ceramicminerals: dehydroxylation of claygiving rise to burned clay, maghemiteand recrystallization of anatase.Maghemite promotes the slightlybrown to red color of the potteries.

After much firing potteries wereused for daily purposes as cooking andconserving foods. At this phase they getin contact with meals, fishes, roots andso on, in boiling water, becomecontaminated and form minerals likealuminum phosphates and Ba-Mn oxyhydroxides mineralequivalents (post-ceramic minerals), which may containthe elements Mg, Ca, Ba, Zn, Pb, Y, and so on.

Finally the potteries after several uses become old, breakdown and are discarded together with many other organic(vegetal and animal) material waste. Gradually the biologicaland chemical-weathering agents which are very active underthe humid and tropical climate in the region for the last 10.000years, attack them. After the people leave the site the rainforest envelopes the area and the weathering becomes moreactive and develops the soil profiles incorporating the greatorganic matter coming from pre-historic people waste.Chemical elements such P, Mg, Ca, Mn, Ba, Zn, Pb, etc. arefixed partly and concentrated in the organic humus of ABEsoils and possibly is partly absorved in the ceramic fragments,contributing to formation of phosphates and Mnoxyhydroxides in less extension. During this phase the clay-derived materials and hematite/maghemite rehydrated andform kaolinite and goethite, respectively (post-ceramicminerals or pedogenic minerals)(Fig.6). The concentrationof ceramic fragments in the A horizons shows that the ABEdevelopment is restrict to the top of pre-existing latosols.The high content of organic matter of the A horizons explainsits gray color, caused the deferrugination the outer skin ofthe ceramic fragments, which become yellowish gray.

ACKNOWLEDGMENTS

We would like to thank the full assistance of the CompanhiaBrasileira de Metalurgia e Mineração - CBMM in conductingthe SEM/EDS analyses in their laboratory, especially thecontribution of Dr. Bruno Riffel and Dr. Abrahão Issa Filho. Thefinancial support of Brazilian National Program for Developmentof the Sciences and Technology (PADCT/FINEP) and BrazilianNational Council for Sciences and Technology (CNPq). We aregrateful to Anselmo José Monteiro dos Santos, who made thefirst drawings of this paper and to Elias Leão Moraes for hiscontribution during the elaboration of the computer drawings.The authors would like to thank the two anonymous reviewersfor their very important suggestions for manuscript corrections.

LITERATURE CITED

Baleé, W. 1986 The culture of Amazonian forest. In: Posey, D.A.Baleé, W. (eds.). Resource management in Amazon:indigenous and flak strategies. Advance in Economic Botany,7, New York, NY Botanical Garden.

Baleé, W. 1989 Cultura na Vegetação da Amazônia Brasileira. Bol. Mus.Para. Emílio Goeldi, Coleção Eduardo Galvão. p. 95 - 109.

Costa, M.L. 1982 Petrologisch-geochemische Untersuchungen zurGenese der Bauxite und Phosphat-Laterite der RegionGurupi (Ost-Amazonien). PhD Thesis. Erlangen-NuernbergUniversity. Germany.,189 p.

Figure 6 - The geochemical pathway (evolution) of ceramic fragments (potteries) found

385 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


Costa, M.L. 1991 Aspectos geológicos dos lateritos da Amazônia.Revista Brasileira de Geociências., 21 (2):146 160.

Costa, M.L. 1997 Lateritisation as a major process of ore depositformation in the Amazon region. Exploration Mining Geology,6(1): 79-104.

Costa, M.L., Kern, D.C., Pinto, A.H.E., Souza, J.R.T., 2004, Theceramic artifacts in archaeological black earth from LowerAmazon Region, Brazil: Mineralogy. Acta Amazonica 34(2): 165-178.

-Costa, M.L.; Kern, D.C.; Souza, J.R.T.; Pinto, A.H.E. 1991 Amineralogia e a geoquímica na cerâmica arqueológica deOriximiná, PA. In: Proceedings of the 3rd BrazilianGeochemical Congress, São Paulo, SBGq, 1: 1-3.

Costa, M.L.; Kern, D.C. 1999 Geochemical signatures of tropicalsoils with archaeological black earth in the Amazon. Journalof Geochemical Exploration, 66(1/2), 369-385.

Costa, M.L.; Moraes, E.L. 1998 Mineralogy, geochemistry andgenesis of kaolins from the Amazon region. MineraliumDeposita, 33 (3):283-297.

Dixon, J.B.,1989, Kaolin and serpentine group minerals. In: J.B.Dixon; S.B. Weed – ed. Minerals in soil environments. SoilSciences Society of America, 2nd. Ed., Madison, p. 467-525.

Duddy, I.R., 1980 Distribution and fractionation of rare-earthelements and other elements in a weathering profile. ChemicalGeology, 30:363-444.

Eden, M.J.; Bray, W.; Herrera,L.; Mcevan, C. 1984 Terra Preta Soilsand their archaeological context in the Caquetá Basin ofSoutheast Colombia. American Antiquity, 49(1)125-140.

Falesi, I. 1974 Soils of Brazilian Amazon. In: Ch.Wagley. ed. Manin the Amazon. Gainesville, p. 201-29.

Freestone, F.C.; Middleton, A.P. 1987 Mineralogical applicationsof the analytical SEM in archaeology. Mineralogical Magazine,51:21-31.

Hilbert, P.P. 1955 A cerâmica arqueológica da região de Oriximiná.Belém, Instituto de Antropologia e Etnologia do Pará. p.76.

Hsu, P. 1968 Interaction between aluminum and phosphate inaqueous solution. Adv. Chem., 73: 115-127.

Kafkafi, U., Posner, A.M.; Quirk, J.P. 1967 Desorption of phosphatefrom kaolinite. Soil Sciences Society of America, 31:348-353.

Kern, D.C. 1988 Caracterização pedológica de solos com terraarqueológica na região de Oriximiná-PA. Porto Alegre. MSc Thesis.Soils Department, Universidade Federal do Rio Grande do Sul. 231p.

Kern, D.C. 1996 Geoquímica e pedogeoquímica em SítiosArqueológicos com Terra Preta na Floresta Nacional deCaxiuanã (Portel-PA). Doctor Thesis, Centro de Geociências.Universidade Federal do Pará. 124 p.

Kern D.C.; Kämpfn, N. 1989 O Efeito de Antigos AssentamentosIndígenas na Formação de Solos com Terra Preta Arqueológicana Região de Oriximiná-Pa. Revista Braileira de Ciências doSolo, 13:219-25.

Kern, D.C.; Kämpf, N., 1990 Características Físicas e Morfológicas dosSolos com TPA e sua importância para os estudos Arqueológicos.Santa Cruz do Sul - RS. Newsletter CEPA, 17(20): 277-85.

Kern, D.; Costa, M.L. 1997 Os Solos Antrópicos. In: P. L.B. Lisboa(Org.) - Caxiuanã. MCT/CNPq, Museu Paraense Emílio Goeldi-MPEG, Belém-PA, p. 105-119.

Kittrick, J.A.; Jackson, M.L. 1955 Rate of phosphate react ion with soilminerals and electron microscope. Observations on the reactionmechanism. Soil Sciences Society of America, 19: 292-295.

Krauskopf, K.B. 1982 Introduction to geochemistry. McGraw-Hill,Auckland, p. 617.

Letsch, J.; Noll, W. 1983 Mineralogie und Technik der frühenKeramiken Thessaliens. Neues Jahrbuch MineralogieAbhandlung, 147(2):109-146.

Meggers, B.J.; Evans, C. 1970 Como interpretar a linguagem dacerâmica. In: Manual para arqueólogos. SmithsonianInstitution. Washington, p. 111.

Melatti, J.C. 1972 Indios do Brasil. Etnologia Brasileira. 2a ed.Brasília. p.236.

Momsen, H. 1986 Archaeometrie: Neuere naturwissenschaftlicheMethoden und Erfolge in der Archaeologie. B.G. TeubnerStuttgart, p. 304.

Nimuendaju, C. (1948) Os Tapajós. Museu Paraense Emílio GoeldiNewsletter, 10:93-106.

Noll, W., 1978 Mineralogie und Technik der bemalten KeramikenAltägyptens. Neues Jahrbuch, Mineralogie-Abhandlung,133(3): 227-290.

Ranzani,G.; Kinjo,T.; Freire,O. 1962 Ocorrência de “PlaggenEpipedon” no Brasil. Boletim Técnico- Científico da EscolaSuperior de Agricultura “Luiz de Queiroz” 5:1-11.

Redmount,C.A . 1996 Major and trace element analysis of modernEgyptian pottery. Journal of Archaeological Science, 23: 741-762.

Roosevelt, A.C.; Costa. M.L; Lopes Machado, C.; Michab, M.;Mercier, N.; Valladas, H.; Feathers, J.; Barnett, W.; Imazio daSilveira, M.; Herderson, A.; Silva, J.; Chernoff, B.; Reese, D.S.;Holman, J.A.; Toth, N.; Schick, K. 1996 Paleo-Indian CaveDwellers in the Amazon the Peopling of the Americas. Science.272: 373-384.

Scholz, H. 1980 Mineralstoffe und Spurenelemente - nötig fuerunserer Gesundheit. Paracelsus Verlag, p. 192.

Simões, M.F. 1982 A Pré-História da Bacia Amazônica: Uma tentativade reconstituição. In: Cultura Indígena, textos e catálogo.Semana do Índio, Museu Goeldi, Belém, p:5-21.

Sjoberg, A. 1976 Phosphate Analysis of Anthropic Soil. Journal ofField Archaeology, 3:448-454.

Smith, H.H. 1879 The amazons and the coast. Charles Scribner’ssons, New York, 644p.

Smith, N.J.H. 1980 Anthrosols and Human Carrying Capacity inAmazonia, In: Annals of the Association of AmericanGeographers, 70(4):553-66.

Sombroek, W. 1966 Amazon soils: A Reconnaissance of the Soils ofthe Brazilian Amazon Region. Wageningen, Center forAgricultural Publications and Documentation. 292 p.

Strazicich, N.M. 1998 Clay sources, pottery production, andregional economy in Chalchihnites, Mexico, A . D. 200-900.Latin American Antiquity, 9(3): 259-274.

386 VOL. 34(3) 2004: 375 - 386 • COSTA et al.


Yariv, S.,; Cross, H. 1979 Geochemistry of colloid systems. Springer-Verlag, Berlin, 450 p.

Zaun, P. E. 1982 Über Bodenlagerung auf antike Keramik. Mineralneu-und -rueckbildungen als moegliche Grundlagen fuer neueDatierungshilfen. Ein Beitrag zur archaeometrischen Forschung.Neues Jahrbuch fuer Mineralogie, Monatsheft, 3: 106-118.

Wada, K. 1989 Allophane and imogolite. In: J.B. Dixon; S.B. Weed– ed. Minerals in soil environments. Soil Sciences Society ofAmerica, Madison, 2nd. Ed., p. 1051-1087.

Wakita, H., Rey, P. Schimitt, R.A.1971.Abundances of the 14 rareearth elements and 12 other trace elements in Apollo 12samples: five igneous and one breccia rocks and four soils.Proc. Second Lunar Sci. Conf., 2, The M.I.T. Press, pp.1319-29.

RECEBIDO EM 05/07/2002ACEITO EM 23/03/2004

The ceramic artifacts in archaeological black earth (terra preta) from ...

Documents

Transcript of The ceramic artifacts in archaeological black earth (terra preta) from ...