Rare earth elements marking soil development at the Geoglyphs of Acre (Brazil).

Rare earth elements marking soil development at the Geoglyphs of Acre (Brazil) Gianni Gallello1*, Mirco Ramacciotti1, Ivandra Rapanelli2, Sonia López-Melón1, Fernando Ferreira3, Jacó César Piccoli3, Ángel Morales-Rubio4, Laura Bartual Moreno4, Adriana SerraMeléndez1, Carlos Arnau-Félix1, Agustín Diez-Castillo1. 1 Department of Prehistory, Archaeology and Ancient History, University of Valencia. Valencia, Spain 2 Xapecó Arqueologia. Chapecó, Brazil 3 Centre of Philosophy and Human Sciences, Federal University of Acre. Rio Branco, Brazil 4 Department of Analytical Chemistry, University of Valencia. Valencia, Spain Corrisponding*: [email protected] Abstract Earthen structures were built by ancient populations in very different contexts and chronologies. In the western Brazilian Amazon region, in the state of Acre, several earthen structures called “Geoglyphs” have been found. The questions involving the use, extension and geographic distribution of these earthen structures are still completely open. The first excavation carried out by our team (2022) in the Reserva Extrativista Chico Mendes (RESEX), showed that the archaeological materials is limited to some few ceramic shreds, lithic materials and palaeo-biological remains. The aim of this study is to test for the first time the potential of rare earth elements to provide proxies for identifying layer development and distinguish human related activities in soil sample sections from two excavated Geoglyphs located in RESEX. Forty-five samples of sediment were collected from the Antonio y Felipe and Pedro y Nelida Geoglyphs. Rare earth elements concentrations together with major elements and cross-referenced with the archaeological record show that the geochemical information can potentially become a complementary tool to better understand the human contribution to the stratigraphic formation in these contexts. Keywords: Rare earth elements, Geoglyphs, chemical elements, earthen structures, soils 1. Introduction Earthen structures were built by ancient populations in very different contexts and chronologies. Also, the Amazon area is characterised by the presence of earthworks initially named “Geoglyphs” by the researchers due to the apparent resemblance to the Peruvian Geoglyphs of Nazca, to attract the attention of scientists on the importance of this discovery. Earthen structures have been identified in a wide geographical area, showing the ability of ancient populations to manipulate their surroundings and significantly modify the natural environment and create a long-term landscape. The Geoglyphs are especially found in the Western and Southwestern Amazon (Southeast and Southwest of the current Brazilian Amazonas, Acre and Rondônia) and in the north-western area of Bolivia in current territory of Beni, which covers approximately 600 kilometres in length (Fig. 1). This work is focused on Acre State region, where more than four hundred earthen structure sites have been identified in the deforested areas (Erickson et al. 2008; Schaan et al. 2012) and our case studies are in Reserva Extrativista Chico Mendes (RESEX) area, a Brazilian national reserve where several earthen structures (Geoglyphs) have been found, some of which are dated between 2500 BC and 1400 AD. Earthen structures were built by European Neolithic and Copper Age populations in very different contexts and chronologies from the 6th millennium BC to the 2nd millennium BC. The hypotheses developed are multiple: some support the idea that the shortage of stone in the environment could have given rise to the construction of these ditches; others suggest that the construction was of social significance, in memory of certain people or events of collective importance; or perhaps a manifestation of respect and/or gratitude to their divinities or ancestors. Other hypotheses are related to astronomical concepts, physical and/or symbolic delimitation, or function such as settlements, ceremonial centres, megaxylic graves, and social power management; or simply for drainage, channels of irrigation or animal farming; or some that suggest a defensive purpose (Bernabeu et al. 2006; Erickson et al. 2008; Shaan et al. 2012). The aim of this study is to test for the first time the potential of rare earth elements (REE) to provide proxies for identifying layer development and distinguish human related activities in sediment sample sections from two recently excavated Geoglyphs located in RESEX. Chemical processes are known to modify REE ratios in sediments. REE analyses have been recently employed to help in the identification of anthropogenic deposits on archaeological sites by comparing the REE values from known anthropogenic stratigraphic units to layers with no evidence of human influence (Gallello et al. 2017; 2019a; 2019b; 2021; 2022; Selvaraj et al. 2024). However, in the investigated Geoglyphs the archaeological record is poorly represented and the natural chemical processes are predominant, therefore, to better understand the strata formation in the studied environment, REE parameters were calculated and cross-referenced with the obtained major element and trace element data. REE concentrations obtained from the analysed sediments samples were normalised, and ratios such as Lan/Ybn, Lan/Smn and Smn/Ybn used to determine relative REE enrichment or depletion. Ce and Eu anomalies were calculated to evaluate differences between measured concentrations of Ce or Eu and the concentration expected of neighbouring REE. Correlation between REE and other elements were also tested to identify significant drivers (i.e., elements) that are affecting the REE chemistry in the different materials. The interpretation of these results could potentially provide important information about the consistency of the mechanisms influencing REE chemistry in the studied sections. Furthermore, in this first test the data obtained from the excavated profiles may allow us to understand if REE are able to shed light on the function of the studied earthen structures, developing hypothesis about possible identified natural and anthropogenic layers. To do so, principal component analysis (PCA) was employed to explore the geochemical datasets, reducing the number of variables. Finally, a methodological approach based on REE chemistry processes and their drivers (minor and major elements) was developed for the interpretation of possible natural or anthropogenic activities carried out in each studied earthen structure and also to identify the nature of the changing patterns. To conclude, the obtained soil data were cross-referenced with the scarce archaeological record composed by sherds, lithic materials and palaeo-biological discovered during fieldworks. 1.2 Previous studies and developed hypotheses In the Brazilian territory of Acre, where most of the earth deposits are located within structures delimited by ditches, the soils are typical of environments with high rainfall and climatic conditions of tropical and subtropical areas. They are generally very deep and have been identified as "argisols, cambisols, luvisols, gleysols, latosols, vertisols, plinthosols" and "neosoles" (Bardales et al. 2010). The pH of the soils in the area is generally acid, rich in silicon, aluminium and iron oxides (Wadt 2002). A preliminary study in the Acre territory highlighted that the deposits of the earthen structures were located in the most ancient areas, where developed soils and degraded latosols and argisoils are found (Carmo 2012). The morphological description of the soils in the Fazenda Atlântica and Cícero Cara de Pau deposits permitted palaeosol identification with anomalous organic carbon levels, changes in the texture of the natural characteristics and the presence of archaeological materials in the surface and subsurface horizons (Carmo 2012). The soils of the Severino Calazans and Jacó Sá geoglyphs were extremely low in phosphorus. Samples collected in a dark brown soil layer presented slightly higher concentration of phosphorus, 18 mg/kg, compared to other samples collected in reddish layers. Carmo (2012) concluded in her study that the occupation and use of structures had not been of sufficient duration to produce high amount of phosphorus in the soil in these environmental conditions. These environments are characterised by rapid nutrient cycles in plants, poor weather conditions and soil leaching. Alternatively, the nature of their use has been unable to cause significant accumulation and subsequent decomposition of the organic materials. During the last decades, archaeo-botanical studies in the Acre region have shed new light on the chrono-cultural context of the earthen structures. Studies carried out by Ranzi and Aguiar (2004) imply the possibility that these were built before the formation of the forest and that in the past Amazon consisted of savannah. Carmo (2012) found phytoliths in ceramic fragments, supporting the hypothesis that the ditches were constructed in a palaeoenvironment of grass and non-forest vegetation, very different from the current scenario. The analysed material demonstrated low levels of "biogenic silica", as shown by the phytolith morphotype identified (bastonete). These morphotypes are taxonomically related to the Poaceae family, which are phytolith-producing grasses of wide geographical distribution. These data may support the hypothesis that this Amazon region was previously dominated by a relatively open landscape such as pampa or savannah, with herbaceous plants, especially grasses. Some authors (Lui and Molina 2009) affirm that the region had a “domestication of vegetation”, causing significant changes in the species distribution. The important changes in the ecology of the landscape included economic reorientations that were focused on the exploitation of specific plants and animals, including palms, fruit trees, edible roots and aquatic fauna. In addition, this domestication also reveals a heterogeneity associated with use of plants (edible, medicinal, ritual and for manufacture), found mainly along the main rivers of the region and on the margins of their floodplains. Other evidence for this diverse exploitation is found in the ethnographic writings of Cristóbal Acuña (1641), describing a very varied fauna and flora, emphasizing a great variety of animals and alimentary plants (bananas, guavas, pineapple, maize, cocoa, Brazilian chestnuts), roots of much sustenance such as yucca, sweet potato, palms of various genera, coconuts, dates as well as medicinal plants. All this suggests that the prehistoric native exercised a powerful and continuous influence, from the beginning of the Holocene, right through to the present day. This conclusion supports the hypothesis that part of what today we see as a "primary" forest, was actually an anthropogenic landscape, the result of conscious or unconscious human activity management (human manipulation of organic and non-organic components of the environment) for thousands of years. Supporting this hypothesis, Watling et al. (2018) in their recent study suggest that ancestors of the Geoglyphs builders were responsible for an intensification of landscape management around 4000 cal BP and a new type of inceptive niche construction in the region. They proposed that, as a result of the Geoglyph construction practices, a landscape with resources such as palms was created, and consequently a greater resource concentration in these areas commenced with a combination of crop cultivation and forest management. This hypothesis seems to be supported by Iriarte et al. (2020) who carried out the LIDAR survey observing distinctive architectural features. Their results suggest the possibility that after the Geoglyph culture the Amazonian Acre region was not abandoned, but a new village network system flourished. Sunaluoma et al. (2021), employing remotely sensed imagery and aerial vehicle surveys, obtained data suggesting that the roads connecting particular sites were articulate across various sites, highlighting that in Acre terrestrial movement responded to the archaeological evidence of villages interconnected by road networks. Souza (2022) studying the Acre geoglyph site of Tequinho suggests that this was a cultural and religious centre, its roads and paths being central to ritualistic ceremonies as well as landscape markers and boundaries. However, against this cultural evolution interpretations, Pärssinen et al. (2020) recently provided new evidence of anthropogenic ash and charcoal accumulation in the Acre area dating back to c. 10 000 cal BP and carbon isotope (δ13C) values suggested no important changes in climate and vegetation during Holocene. While Della Libera et al. (2022) also employing δ13C record indicates an overall tendency towards a more humid tropical forest only during the last millennia. Nacimiento et al. (2022) collected palaeoecological records from lake sediments containing charcoal and from pollen analyses to understand how human land-use affected vegetation during the early to mid-Holocene in the Western Amazon. The authors observed that significant vegetation changes were only found centuries to millennia after the human presence, suggesting that the earliest occupants of Amazonia exerted no more than a gradual change on the forest. The early human occupations across Amazonia occurred mostly in areas bordering savannahs, where fires were a part of the natural vegetation dynamics. In these regions natural fires were common and could not be distinguished from anthropogenic fires based solely on charcoal counts. Then, there are others who support a symbiosis between humans and environment. Ponte and Szlafsztein (2022), investigating the dynamics and periodisation of special social events, assumed that over time these human groups in Western Amazonia, including the Acre region, become familiar with the natural characteristics of the region to better understand the dynamics of its ecosystems, consequently selecting plant species and ecosystems with greater utilitarian potential for their livelihood. Analyses of the dynamics of socio-spatial events demonstrate that humans have always established relationships with nature through their organisational and productive capacities, the latter conditioned by the available technology; and as humanity interacted with Amazonian natural environment their knowledge increased, enhancing conditions that facilitated socio-productive stabilisation. Maezumi et al. (2022) suggests that polyculture agroforestry and cultural burning persisted within this system during the Holocene. Likewise, Silva (2019) investigated by geomorphological techniques the construction techniques of these monuments showing that the base was made by mound, where material with clay and iron nodules were employed, probably to the reinforce the structure, opposite the upper layers were made on coarser material (silt and sand).It is therefore apparent that the evidence presented by previous studies show diverse hypotheses which are in some ways contradictory regarding the development and function of the earthen structures. This inconclusive situation demonstrates the necessity for more archaeological data and the application of additional methods. 2. Materials and Methods Soil samples were collected from the Geoglyphs of Antonio y Felipe (AyF) and of Pedro y Nelida (PyN), as aforementioned, located in the RESEX area (Fig. 1). The earthen structures were discovered by Agustín Diez-Castillo and collaborators during the Hispanic-Brazilian research collaboration that have been developed since 2017 (Diez Castillo et al. 2017; Rampanelli et al. 2017). Both earthen structures have been selected for this study because their formal features are characteristic of the area, and due to their optimal state of conservation and to logistical factors, such as the facility for obtaining permits, being within a federal reserve. AyF and PyN are located under the forest and are protected under the jurisdiction of RESEX. This is an area of high ecological value owned by the Brazilian state and in which traditional uses are allowed to both indigenous communities and traditional settlers (aka serengeiros) allowing the establishment of a long-term archaeological intervention strategy not conditioned by landowners roles. The first survey carried out showed that the material in these Geoglyphs is very limited (Rampanelli et al. 2017), although some interesting ceramic assemblages were recovered, as well as lithic materials and palaeo-biological remains during fieldwork (August 2022). 2.1 Sampling Sampling data are reported in the Table 1. Forty-five samples of sediment were collected from the geoglyphs of Antonio y Felipe (AyF) and of Pedro y Nelida (PyN). Concerning AyF, ten samples come from a cross-section in the bottom of the ditch (AyF2.01-10), characterised by a reddish and silty-sandy soils, and eight from a cross-section from an excavation in inner area (AyF3.01-08), characterised by a reddish and silty-sandy soils with laterites. Regarding PyN, the samples come from an excavation area in the upper inner area of the Geoglyph: seventeen from one of the main cross-sections (PyN2.01-17), characterised by a reddish and silty-sandy sediments with presence of ceramic shreds and carbons, and ten from a survey pit in the centre of the same area (PyN1.01-10), in which the sediment are clayey-sandy with presence of laterites and carbons. 2.2 Multielement analysis Samples were air-dried, sieved to 2 mm, grounded and homogenised by an agate mortar. Portable X-ray fluorescence (pXRF) analysis was performed with a S1 Titan portable energy dispersive pXRF spectrometer by Bruker equipped with an Rh X-ray tube and X-Flash® SDD detector. Geochem-trace calibration was used to obtain of Al, Si, K, Ca, Ti, Fe, Rb and Zr concentrations. Trace elements analyses were performed using a NexION 2000 inductively coupled plasma mass spectrometer (ICP-MS) by Perkin Elmer. Samples were digested by aqua regia, adding 1.35 mL of HCl and 0.45 mL of HNO3 to 0.15 g of powdered sample in a glass tube, subsequently heated in a boiling water bath for about 40 min. Concentrations Li, Be, V, Cr, Mn, Co, Ni, Zn, Sr, Mo, Cd, Ba, Tl, Pb, Bi, U, REE (La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu), Sc and Y were determined in the solution. Details on the analytical method can be found in Gallello et al. (2022). NIM-GBW07408 Soil certified reference material was analysed by both pXRF and ICP-MS to test the accuracy and precision of the analyses. 2.3 Data analysis Data analysis was carried out by R (version: 4.4.1; R Core Team, 2024) and the following R packages were employed: factoextra (version: 1.0.7; Kassambara and Mundt, 2020), corrplot (version: 0.92; Wei and Simko, 2021), ggplot2 (version: 3.5.1; Wickham, 2016). Data were explored by principal component analysis (PCA) which was carried out using all the samples as observations and all the elemental concentrations as variables. Variables were z-scored prior to the analysis. 2.4 Rare earth elements (REE) parameters Several ratios are calculated for Ce and Eu anomalies, resulting from differential reactions associated with reduction-oxidation (Ce3+ or 4+ and Eu2+ or 3+) using formulae from Gallello et al. (2021), detailed below. Anomalies are considered positive or negative depending on whether the values are above or below 1, respectively. Finally, fractionations (Gallello et al. 2019) of light REE (LREE: La to Nd), medium REE (MREE: Sm to Ho) and heavy REE (HREE: Er to Lu) can be evaluated. In order to calculate these ratios, normalized concentrations need to be use. These concentrations are written as “Xn” in the formulae, corresponding to the normalised concentration of an element X by Post Archean Australian Shale (PAAS) concentrations reported by Taylor and McLennan (1985) as follows. [𝑋𝑛] = [𝑋]𝑟𝑎𝑤 𝑃𝐴𝐴𝑆𝑋 Where [X]raw corresponds to the concentration of the element X after ICP-MS analysis and PAASX correspond to the PAAS value of the element X. Those normalised concentrations are then used to calculate anomalies and fractionation with the formulae as follow. Anomalies from Gallello et al. (2021): 𝐶𝑒𝑛 𝐶𝑒∗ 𝐸𝑢𝑛 𝐸𝑢∗ 𝐶𝑒∗=𝐿𝑎𝑛 2+𝑃𝑟 𝑛 2 and 𝐸𝑢∗=𝑆𝑚𝑛 2+𝐺𝑑𝑛 2 Fractionations: 𝐿𝑅𝐸𝐸 𝐻𝑅𝐸𝐸 =𝐿𝑎𝑛 𝑌𝑏𝑛 ; 𝐿𝑅𝐸𝐸 𝑀𝑅𝐸𝐸 =𝐿𝑎𝑛 𝑆𝑚𝑛; 𝑀𝑅𝐸𝐸 𝐻𝑅𝐸𝐸 =𝑆𝑚𝑛 𝑌𝑏𝑛 3. Results and discussion 3.1 Data exploration Major, minor and trace elements concentrations, and REE, including Sc and Y, REE ratios and Ce, Eu anomalies parameters are in the supplementary materials (Annex 1, Tables A1.1 and A1.2). Principal component analysis was carried out to explore chemical data (Fig. 2). As can be observed in the scores plot (Fig.2), samples from AyF and from PyN can be distinguished: most of the samples from PyN have higher scores in PC1 positive direction and lower ones in PC2 negative direction than those from AyF (except for some samples from PyN2). It is worth noting that, considering the two Geoglyphs, the samples collected from different cross-sections fall in different areas of the diagram. Indeed, samples from AyF2 (bottom of the ditch) have higher PC1 in negative direction and PC2 in positive direction than those from AyF3 (inner area), and samples from PyN2 (upper inner area) has higher PC2 scores and in most of the cases higher PC1 scores than PyN1 (upper inner area – survey pit). Loadings plots for PC1 (Fig.2b) and for PC2 (Fig.2c) indicate the main features of the different classes of samples. Samples from PyN show higher levels of Al, Ti, Zr and Mo than AyF, while AyF ones show higher concentrations of K, Rb, Be, Co, Ni, Zn, Cd and Ba. For the other elements, the dissimilarities are fuzzier and involve inner differences between the cross-sections of the two geoglyphs. Concerning major elements levels, we can observe that cross-section PyN2 has higher Si levels than the others and low Ca ones, which is in most of the cases below the limit of detection. On the other hand, PyN1 has the highest Fe concentrations. As regards trace elements, PyN1 has the highest concentrations for V, Cr and Bi, and the lowest for Mn, while PyN2 has low amounts of Pb. Total REE levels are lower in AyF2 and PyN2 then in the other column of their respective Geoglyph, though for PyN differences are more blurred. Loadings for PC2 suggest that scores in this dimension are influenced also by REE fractionation. First, PyN is characterised by strong positive Ce anomalies, especially for low strata of PyN2 (PyN2.02-07), compared to AyF. Light REE are enriched over HREE, but this enrichment is less intense in AyF3, while MREE enrichment over HREE is more intense in PyN than in AyF. On the other hand, LREE are not enriched compared to MREE in all the columns and Eu anomalies are not present. Correlations among elemental concentrations and REE parameters in the four cross-sections are shown in Fig.3. Drivers for REE levels seems to variate slightly from a Geoglyph to the others, and between columns of the same Geoglyph. Total amount of REE (TREE) is positively correlated with Fe in all the columns and with Sr in AyF2-3 and in PyN1. For PyN2 and AyF2, it shows a moderate Wei, T., and Simko, V. 2021. R package 'corrplot': Visualization of a Correlation Matrix (Version 0.92). https://github.com/taiyun/corrplot Wickham, H. 2016. ggplot2: Elegant Graphics for Data Analysis. Houston: Springer. https://doi.org/10.1007/978-3-319-24277-4 Table and Figures Captions Table 1. Sampling data. Fig. 1. Area of study and sampling section location for the Geoglyphs AyF and PyN. Fig. 2. Samples/scores plot (a) of PCA and variables/loadings plots for PC1 (b) and PC2 (c). Fig. 3. Pearson correlation coefficients (r) for REE parameters and elemental concentrations. Nonsignificant coefficients (p > 0.05) are barred. Fig. 4. Distributions of elemental concentrations and REE parameters in the four columns (outliers were not considered). Sample Geoglyph Location Height Sample Geoglyph Location Height AyF2.01 Antonio y Felipe Bottom of the ditch 297.40 PyN1.06 Pedro y Nelida Upper inner area (survey pit) 98.00 AyF2.02 Antonio y Felipe Bottom of the ditch 297.50 PyN1.07 Pedro y Nelida Upper inner area (survey pit) 98.10 AyF2.03 Antonio y Felipe Bottom of the ditch 297.60 PyN1.08 Pedro y Nelida Upper inner area (survey pit) 98.20 AyF2.04 Antonio y Felipe Bottom of the ditch 297.70 PyN1.09 Pedro y Nelida Upper inner area (survey pit) 98.30 AyF2.05 Antonio y Felipe Bottom of the ditch 297.80 PyN1.10 Pedro y Nelida Upper inner area (survey pit) 98.40 AyF2.06 Antonio y Felipe Bottom of the ditch 297.90 PyN2.17 Pedro y Nelida Upper inner area 98.40 AyF2.07 Antonio y Felipe Bottom of the ditch 298.00 PyN2.16 Pedro y Nelida Upper inner area 98.45 AyF2.08 Antonio y Felipe Bottom of the ditch 298.10 PyN2.15 Pedro y Nelida Upper inner area 98.50 AyF2.09 Antonio y Felipe Bottom of the ditch 298.20 PyN2.14 Pedro y Nelida Upper inner area 98.55 AyF2.10 Antonio y Felipe Bottom of the ditch 298.30 PyN2.13 Pedro y Nelida Upper inner area 98.60 AyF3.01 Antonio y Felipe Inner area 298.30 PyN2.12 Pedro y Nelida Upper inner area 98.65 AyF3.02 Antonio y Felipe Inner area 298.40 PyN2.11 Pedro y Nelida Upper inner area 98.70 AyF3.03 Antonio y Felipe Inner area 298.50 PyN2.10 Pedro y Nelida Upper inner area 98.75 AyF3.04 Antonio y Felipe Inner area 298.60 PyN2.09 Pedro y Nelida Upper inner area 98.80 AyF3.05 Antonio y Felipe Inner area 298.70 PyN2.08 Pedro y Nelida Upper inner area 98.85 AyF3.06 Antonio y Felipe Inner area 298.80 PyN2.07 Pedro y Nelida Upper inner area 98.90 AyF3.07 Antonio y Felipe Inner area 298.90 PyN2.06 Pedro y Nelida Upper inner area 98.95 AyF3.08 Antonio y Felipe Inner area 299.00 PyN2.05 Pedro y Nelida Upper inner area 99.00 PyN1.01 Pedro y Nelida Upper inner area (survey pit) 97.20 PyN2.04 Pedro y Nelida Upper inner area 99.05 PyN1.02 Pedro y Nelida Upper inner area (survey pit) 97.50 PyN2.03 Pedro y Nelida Upper inner area 99.10 PyN1.03 Pedro y Nelida Upper inner area (survey pit) 97.60 PyN2.02 Pedro y Nelida Upper inner area 99.15 PyN1.04 Pedro y Nelida Upper inner area (survey pit) 97.70 PyN2.01 Pedro y Nelida Upper inner area 99.20 PyN1.05 Pedro y Nelida Upper inner area (survey pit) 97.90 Table 1. Collected samples and provenance. Fig. 1 Fig. 2 Fig. 3 Fig. 4 ANNEX Samples Al Si KCa Ti Fe Rb Zr Li Be V Cr Mn Co Ni Zn Sr Mo Cd Ba Tl Pb Bi U AyF2.01 11.72 21.88 1.15 0.02 0.61 5.03 52 307 1.39 0.613 52.4 21.3 368 4.31 8.87 40.2 7.22 0.403 0.029 31.0 0.075 15.8 0.110 5.67 AyF2.02 11.24 22.79 1.08 0.02 0.59 4.60 51 317 1.05 0.574 45.2 14.6 229 3.43 6.76 6.00 5.85 0.301 0.018 17.7 0.055 12.0 0.097 4.68 AyF2.03 11.26 22.63 1.06 0.01 0.60 4.64 49 318 1.05 0.533 49.2 23.1 258 3.44 6.70 4.52 5.81 0.347 0.022 23.0 0.060 13.1 0.110 4.75 AyF2.04 8.74 20.73 0.94 <LD 0.57 4.56 47 328 0.984 0.392 42.8 17.4 266 3.36 6.09 90.7 5.83 0.407 0.017 19.5 0.058 11.0 0.081 4.47 AyF2.05 11.18 25.01 1.02 0.01 0.61 4.54 50 366 0.899 0.479 43.1 15.0 259 3.73 6.26 33.5 5.36 0.220 0.012 19.0 0.061 11.5 0.083 4.49 AyF2.06 9.42 21.53 0.97 0.01 0.61 4.32 55 353 0.952 0.539 48.9 16.5 336 4.38 7.03 83.2 5.98 0.265 0.021 22.4 0.071 11.8 0.097 4.95 AyF2.07 10.10 23.09 1.01 <LD 0.60 4.63 55 389 0.815 0.406 46.2 23.4 469 5.42 6.22 55.1 5.60 0.255 0.029 17.8 0.063 12.8 0.092 4.41 AyF2.08 13.62 27.62 1.09 0.02 0.62 4.33 51 384 0.920 0.409 43.0 12.5 613 5.36 6.63 43.4 6.15 0.433 0.030 23.4 0.085 13.7 0.089 4.26 AyF2.09 8.80 21.37 0.96 0.01 0.59 4.13 51 342 0.933 0.454 42.1 14.4 682 5.23 6.32 49.3 5.96 0.382 0.073 21.7 0.070 13.6 0.082 4.24 AyF2.10 10.15 24.75 1.00 0.01 0.61 4.06 49 344 1.02 0.396 43.4 13.2 772 5.17 6.68 37.4 6.49 0.509 0.038 25.4 0.076 14.1 0.079 4.47 Mean 10.6 23 1.03 0.014 0.601 4.5 51 345 51 0.48 46 17 425 4.4 6.8 44 6.0 0.35 0.029 22 0.067 12.9 0.092 4.6 s 1.5 2 0.07 0.002 0.015 0.3 2 28 2 0.08 3 4 198 0.9 0.8 28 0.5 0.09 0.017 4 0.010 1.4 0.011 0.4 AyF3.01 14.63 22.02 1.30 0.02 0.61 5.64 59 303 1.91 0.655 51.8 17.8 183 3.73 8.49 42.4 7.98 0.353 0.020 36.8 0.082 13.6 0.106 6.03 AyF3.02 15.10 23.45 1.25 0.01 0.60 5.60 66 310 1.79 0.607 56.9 21.2 184 3.64 8.11 47.7 8.02 0.606 0.031 35.8 0.086 13.7 0.160 5.95 AyF3.03 12.30 20.12 1.21 0.01 0.59 5.56 63 363 1.99 0.616 53.4 17.1 183 3.61 8.31 50.9 8.48 0.305 0.025 36.2 0.080 13.6 0.126 5.81 AyF3.04 13.78 22.38 1.14 0.02 0.59 5.70 65 357 2.07 0.569 57.0 28.3 181 3.52 8.20 40.9 7.71 0.180 0.021 34.0 0.082 13.8 0.110 6.10 AyF3.05 13.66 21.76 1.15 0.01 0.60 6.02 64 357 2.61 0.605 62.2 28.1 241 3.96 8.87 28.8 10.2 0.414 0.030 49.6 0.110 16.8 0.125 6.51 AyF3.06 10.20 18.69 1.04 0.01 0.56 5.93 63 357 1.97 0.463 56.9 33.3 193 3.19 7.29 33.5 8.11 0.239 0.039 40.4 0.083 13.9 0.122 5.48 AyF3.07 10.30 19.28 1.00 0.01 0.57 5.71 64 320 2.16 0.545 60.3 31.4 222 3.47 7.87 38.6 9.38 0.441 0.034 44.6 0.102 15.0 0.134 5.76 AyF3.08 12.15 23.43 1.10 0.01 0.61 5.00 58 270 1.54 0.443 48.4 18.4 219 3.54 6.80 46.5 7.14 0.270 0.032 33.9 0.079 12.1 0.102 5.30 Mean 12.8 21.4 1.15 0.015 0.590 5.6 63 330 2.0 0.56 56 24 201 3.6 8.0 41 8.4 0.35 0.029 39 0.088 14.1 0.123 5.9 s 1.9 1.8 0.10 0.003 0.018 0.3 3 34 0.3 0.08 4 7 23 0.2 0.7 7 1.0 0.13 0.007 6 0.012 1.4 0.018 0.4 PyN1.01 15.93 20.71 0.74 0.01 0.71 8.66 43 528 3.56 0.272 144 39.6 160 1.76 5.53 13.7 7.24 0.453 <LD 12.9 0.098 15.6 0.227 5.15 PyN1.02 14.81 18.68 0.76 0.01 0.73 9.44 47 454 3.15 0.242 167 66.9 162 1.83 5.04 18.0 8.18 0.500 0.022 17.1 0.099 18.1 0.214 6.06 PyN1.03 16.50 19.44 0.75 0.01 0.71 12.20 43 436 3.32 0.312 272 116.9 152 1.98 5.95 26.7 10.8 1.617 0.025 18.8 0.108 25.2 0.393 8.09 PyN1.04 15.48 20.51 0.82 <LD 0.77 8.30 45 498 2.49 0.222 160 65.9 151 1.73 4.97 18.9 7.49 1.008 0.018 14.2 0.091 16.4 0.249 5.59 PyN1.05 19.43 26.38 0.84 0.01 0.74 6.81 41 557 2.39 0.199 143 63.9 156 1.63 4.50 18.9 7.02 0.626 0.024 14.0 0.087 14.8 0.212 4.74 PyN1.06 14.89 21.90 0.80 <LD 0.81 7.04 43 539 2.41 0.197 152 65.8 167 1.72 4.94 20.5 7.16 1.103 0.014 16.3 0.090 14.6 0.248 5.01 PyN1.07 15.11 21.92 0.80 <LD 0.75 7.15 43 498 2.09 0.185 122 56.6 150 1.55 4.06 5.09 6.15 0.843 0.012 15.8 0.081 12.8 0.198 4.24 PyN1.08 15.30 22.49 0.79 <LD 0.78 7.31 44 562 2.07 0.233 170 72.3 131 1.47 4.28 15.6 7.91 0.732 0.017 19.7 0.078 16.0 0.214 4.96 PyN1.09 18.79 27.21 0.82 0.04 0.74 8.32 40 512 1.74 0.230 161 82.9 130 1.24 3.52 2.95 7.33 0.889 0.013 17.2 0.073 15.1 0.213 4.41 PyN1.10 12.75 21.83 0.74 0.02 0.77 4.77 38 575 1.81 0.111 59.9 20.0 142 1.52 3.72 10.7 4.68 0.661 <LD 15.5 0.069 8.24 0.136 3.67 Mean 16 22 0.78 0.017 0.75 8 43 516 2.5 0.22 155 65 150 1.6 4.6 15 7.4 0.8 0.018 16 0.087 16 0.23 5.2 s 2 3 0.04 0.011 0.03 2 3 46 0.6 0.05 52 25 13 0.2 0.8 7 1.6 0.3 0.005 2 0.012 4 0.07 1.2 PyN2.01 10.20 38.80 0.54 0.15 0.76 3.19 17 678 0.449 0.052 47.4 25.9 513 1.67 2.80 <LD 7.37 0.740 0.029 23.0 0.032 8.04 0.078 1.57 PyN2.02 11.44 37.42 0.62 0.05 0.90 2.95 22 828 0.428 0.097 44.8 19.8 493 1.83 2.64 <LD 5.61 0.771 0.014 16.2 0.032 8.27 0.083 1.90 PyN2.03 14.48 41.35 0.69 0.04 0.97 2.94 20 804 0.526 0.065 44.7 18.6 537 2.12 2.89 <LD 5.53 0.878 0.016 18.1 0.036 8.97 0.104 2.28 PyN2.04 11.10 36.46 0.63 0.02 0.98 2.79 22 933 0.596 0.071 44.1 15.7 517 2.35 3.19 <LD 4.94 0.644 0.014 16.6 0.044 9.55 0.084 2.65 PyN2.05 13.28 36.90 0.67 0.01 0.95 3.25 23 858 0.613 0.083 46.9 19.5 413 2.06 2.81 <LD 4.87 0.711 0.020 15.3 0.045 9.73 0.095 2.58 PyN2.06 10.97 34.99 0.59 <LD 0.88 3.08 25 890 0.817 0.103 54.2 18.4 358 2.35 3.42 <LD 4.98 0.924 0.015 13.6 0.053 9.67 0.124 3.16 PyN2.07 11.34 30.59 0.69 <LD 0.96 3.49 27 856 0.819 0.085 58.4 28.6 253 1.90 2.91 <LD 4.66 0.823 0.013 13.2 0.049 10.3 0.124 3.09 PyN2.08 12.14 27.12 0.72 <LD 0.94 3.88 29 734 1.25 0.110 58.2 21.9 226 1.91 3.71 3.88 4.95 0.923 0.014 16.6 0.064 9.33 0.134 3.37 PyN2.09 14.82 33.32 0.64 <LD 0.83 4.06 32 763 1.49 0.154 61.2 23.1 226 1.99 3.87 <LD 5.14 0.609 0.017 20.8 0.075 10.2 0.125 4.05 PyN2.10 12.92 28.53 0.67 <LD 0.78 4.11 33 646 1.66 0.172 60.9 21.6 278 2.07 4.22 28.2 5.19 0.530 0.020 23.3 0.078 11.2 0.124 4.34 PyN2.11 14.74 28.69 0.70 <LD 0.77 4.16 36 567 1.81 0.164 57.5 16.6 475 2.10 4.28 <LD 5.00 0.524 0.019 23.6 0.283 11.7 0.148 4.40 PyN2.12 14.97 32.09 0.67 <LD 0.73 4.05 33 584 1.74 0.173 67.4 19.9 416 2.10 4.10 12.6 5.48 0.984 0.029 22.0 0.118 11.3 0.161 4.19 PyN2.13 15.28 27.56 0.74 <LD 0.80 3.83 33 604 1.21 0.079 51.2 16.1 222 1.64 3.24 11.9 4.31 0.430 0.012 14.1 0.081 7.95 0.116 3.16 PyN2.14 12.00 25.75 0.69 0.01 0.75 3.94 34 610 1.36 0.128 58.0 17.2 203 1.80 3.76 <LD 4.93 0.707 0.013 18.2 0.079 8.49 0.139 3.57 PyN2.15 15.01 25.05 0.79 <LD 0.81 4.44 36 633 1.59 0.144 66.2 23.1 178 1.66 3.76 <LD 5.29 0.561 0.011 17.5 0.079 9.40 0.143 3.76 PyN2.16 15.91 26.74 0.70 <LD 0.76 4.72 38 586 1.63 0.146 82.6 27.1 189 1.74 4.06 <LD 5.67 0.768 0.018 19.0 0.077 10.6 0.174 4.24 PyN2.17 14.09 29.06 0.69 <LD 0.76 4.65 37 622 1.71 0.158 103 29.9 179 1.72 3.99 <LD 5.68 0.973 0.011 17.4 0.083 10.3 0.187 4.00 Mean 13.2 32 0.67 0.05 0.84 3.7 29 717 1.2 0.12 59 21 334 1.9 3.5 14 5.3 0.74 0.017 18 0.08 9.7 0.13 3.3 s 1.8 5 0.06 0.05 0.09 0.6 7 123 0.5 0.04 15 4 136 0.2 0.6 10 0.7 0.17 0.005 3 0.06 1.1 0.03 0.9 Table A1.1. Concentrations of major element and trace alements expressed as percentage ( from Al to Fe) or mg/kg (from Rb to U) Sample La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Sc Y TREE Cen/Ce* Eun/Eu* Lan/YbnLan/SmnSmn/Ybn AyF2.01 21.2 59.4 5.20 20.7 4.00 0.833 3.29 0.417 2.26 0.373 0.951 0.121 0.720 0.097 6.88 9.09 120 1.30 1.07 2.17 0.78 2.78 AyF2.02 16.6 49.5 4.07 16.3 3.12 0.646 2.63 0.335 1.81 0.306 0.788 0.096 0.578 0.080 5.81 7.46 97 1.38 1.05 2.12 0.79 2.69 AyF2.03 15.9 49.0 3.91 15.4 2.97 0.635 2.56 0.325 1.78 0.292 0.734 0.096 0.566 0.078 5.98 7.13 94 1.43 1.07 2.07 0.79 2.63 AyF2.04 14.8 46.2 3.62 14.1 2.81 0.563 2.23 0.294 1.53 0.255 0.629 0.079 0.493 0.065 5.22 6.01 88 1.45 1.05 2.21 0.77 2.85 AyF2.05 13.3 42.8 3.26 12.7 2.52 0.530 2.19 0.277 1.53 0.251 0.634 0.085 0.524 0.070 5.18 6.26 81 1.49 1.05 1.87 0.78 2.40 AyF2.06 14.2 40.3 3.38 13.2 2.63 0.571 2.26 0.287 1.58 0.267 0.659 0.086 0.537 0.071 5.37 6.62 80 1.34 1.09 1.94 0.79 2.45 AyF2.07 12.3 35.6 2.98 11.6 2.27 0.467 1.97 0.250 1.37 0.240 0.603 0.080 0.463 0.060 5.06 5.81 70 1.35 1.03 1.96 0.80 2.44 AyF2.08 14.8 40.0 3.46 13.5 2.55 0.556 2.16 0.277 1.52 0.250 0.647 0.083 0.477 0.066 4.75 6.21 80 1.29 1.11 2.28 0.85 2.68 AyF2.09 14.1 41.7 3.33 13.0 2.39 0.494 2.06 0.255 1.37 0.230 0.571 0.071 0.449 0.059 4.92 5.65 80 1.40 1.04 2.31 0.87 2.66 AyF2.10 14.3 45.4 3.30 12.5 2.40 0.487 1.92 0.253 1.28 0.214 0.530 0.069 0.437 0.056 5.20 5.16 83 1.52 1.06 2.40 0.87 2.75 Mean 15 45 3.7 14 2.8 0.58 2.3 0.30 1.6 0.27 0.67 0.087 0.52 0.070 5.4 6.5 87 1.39 1.06 2.13 0.81 2.63 s 2 7 0.6 3 0.5 0.11 0.4 0.05 0.3 0.05 0.12 0.015 0.08 0.012 0.6 1.1 14 0.08 0.02 0.17 0.04 0.15 AyF3.01 20.5 52.0 5.23 20.9 4.14 0.874 3.62 0.467 2.48 0.442 1.12 0.146 0.854 0.119 7.19 10.8 113 1.15 1.05 1.77 0.73 2.42 AyF3.02 19.7 50.6 5.10 20.2 4.10 0.869 3.52 0.443 2.45 0.424 1.08 0.139 0.837 0.116 7.10 10.6 110 1.16 1.07 1.73 0.71 2.45 AyF3.03 20.6 51.5 5.29 21.4 4.17 0.897 3.66 0.474 2.56 0.436 1.10 0.143 0.842 0.116 7.06 10.8 113 1.14 1.07 1.80 0.73 2.48 AyF3.04 19.9 49.7 5.06 20.7 4.06 0.869 3.58 0.457 2.49 0.433 1.09 0.143 0.870 0.122 7.53 10.8 109 1.14 1.06 1.69 0.72 2.34 AyF3.05 22.8 58.6 5.95 23.8 4.69 1.026 4.00 0.510 2.75 0.476 1.19 0.155 0.921 0.129 8.00 11.2 127 1.16 1.10 1.82 0.72 2.55 AyF3.06 17.9 53.4 4.62 18.8 3.81 0.793 3.28 0.417 2.24 0.382 0.994 0.127 0.782 0.107 7.19 9.03 108 1.35 1.05 1.69 0.69 2.43 AyF3.07 18.8 57.9 4.93 19.8 3.89 0.838 3.37 0.446 2.27 0.381 0.956 0.129 0.770 0.107 7.58 8.84 115 1.38 1.08 1.80 0.71 2.53 AyF3.08 15.3 52.9 3.96 16.1 3.30 0.697 2.81 0.371 2.01 0.339 0.839 0.113 0.688 0.098 6.55 8.10 100 1.56 1.07 1.64 0.68 2.40 Mean 19 53 5.0 20 4.0 0.86 3.5 0.45 2.4 0.41 1.05 0.14 0.82 0.114 7.3 10.0 112 1.25 1.07 1.74 0.71 2.45 s 2 3 0.6 2 0.4 0.09 0.3 0.04 0.2 0.04 0.11 0.01 0.07 0.010 0.4 1.2 8 0.16 0.02 0.07 0.02 0.07 PyN1.01 12.6 66.3 3.54 14.3 2.66 0.484 1.76 0.195 1.00 0.166 0.437 0.061 0.358 0.052 11.1 3.70 104 2.27 1.04 2.59 0.70 3.72 PyN1.02 12.4 62.0 3.65 14.7 2.77 0.506 1.84 0.202 1.05 0.165 0.443 0.061 0.386 0.053 10.8 3.63 100 2.11 1.04 2.36 0.66 3.59 PyN1.03 13.9 72.1 4.04 16.6 3.10 0.595 1.99 0.225 1.14 0.184 0.487 0.066 0.436 0.057 13.8 4.08 115 2.20 1.11 2.35 0.66 3.55 PyN1.04 11.5 61.9 3.46 14.5 2.75 0.520 1.83 0.203 1.04 0.165 0.434 0.061 0.369 0.055 10.1 3.38 99 2.24 1.07 2.30 0.62 3.73 PyN1.05 10.1 56.3 3.18 13.2 2.54 0.461 1.66 0.176 0.954 0.148 0.378 0.052 0.346 0.051 9.23 2.96 90 2.26 1.04 2.16 0.59 3.68 PyN1.06 9.65 60.7 3.10 12.8 2.67 0.491 1.66 0.191 0.961 0.151 0.402 0.055 0.375 0.051 9.55 2.98 93 2.52 1.08 1.90 0.53 3.57 PyN1.07 8.01 56.4 2.65 11.1 2.25 0.421 1.45 0.159 0.861 0.134 0.343 0.049 0.307 0.042 8.50 2.51 84 2.77 1.08 1.92 0.53 3.65 PyN1.08 8.42 59.0 2.76 11.5 2.40 0.431 1.55 0.176 0.895 0.141 0.369 0.052 0.326 0.046 9.37 2.64 88 2.78 1.03 1.90 0.52 3.68 PyN1.09 7.65 51.9 2.57 10.7 2.33 0.439 1.49 0.166 0.835 0.132 0.345 0.045 0.330 0.045 8.26 2.44 79 2.65 1.09 1.71 0.48 3.52 PyN1.10 7.03 50.4 2.28 9.60 1.98 0.374 1.29 0.147 0.731 0.114 0.298 0.041 0.283 0.037 7.04 2.20 75 2.86 1.08 1.83 0.52 3.50 Mean 10 60 3.1 12.9 2.5 0.47 1.7 0.18 0.95 0.15 0.39 0.054 0.35 0.049 9.8 3.1 93 2.47 1.06 2.10 0.58 3.62 s 2 6 0.6 2.2 0.3 0.06 0.2 0.02 0.12 0.02 0.06 0.008 0.04 0.006 1.9 0.6 12 0.28 0.03 0.29 0.07 0.08 PyN2.01 2.06 13.6 0.537 2.14 0.446 0.075 0.287 0.035 0.179 0.030 0.081 0.012 0.075 0.009 2.33 0.608 20 2.97 0.97 2.03 0.68 2.99 PyN2.02 1.88 16.2 0.439 1.70 0.350 0.072 0.265 0.031 0.171 0.029 0.074 0.010 0.077 0.011 2.45 0.530 21 4.11 1.10 1.79 0.79 2.26 PyN2.03 2.31 21.1 0.566 2.29 0.445 0.086 0.337 0.041 0.225 0.037 0.093 0.014 0.077 0.012 3.10 0.681 28 4.25 1.03 2.20 0.76 2.88 PyN2.04 3.03 30.4 0.789 3.11 0.639 0.118 0.446 0.054 0.274 0.045 0.112 0.018 0.107 0.013 3.75 0.864 39 4.51 1.03 2.08 0.70 2.98 PyN2.05 3.48 34.1 0.924 3.49 0.718 0.140 0.539 0.060 0.333 0.055 0.129 0.018 0.113 0.017 4.15 1.04 44 4.36 1.05 2.26 0.71 3.17 PyN2.06 5.29 43.5 1.47 5.76 1.14 0.211 0.829 0.093 0.483 0.074 0.190 0.026 0.175 0.023 5.17 1.55 59 3.57 1.01 2.23 0.68 3.26 PyN2.07 5.34 41.8 1.51 5.95 1.21 0.220 0.836 0.093 0.512 0.079 0.201 0.027 0.180 0.025 5.12 1.60 58 3.37 1.02 2.18 0.65 3.36 PyN2.08 7.54 50.9 2.14 8.59 1.69 0.335 1.17 0.136 0.706 0.117 0.299 0.039 0.251 0.037 6.15 2.33 74 2.90 1.11 2.21 0.66 3.37 PyN2.09 9.66 60.5 2.89 11.7 2.29 0.450 1.60 0.176 0.931 0.148 0.380 0.054 0.321 0.046 7.44 3.04 91 2.61 1.09 2.22 0.62 3.56 PyN2.10 11.6 68.5 3.65 15.2 3.02 0.574 1.99 0.224 1.18 0.181 0.463 0.062 0.406 0.055 8.16 3.75 107 2.40 1.09 2.10 0.56 3.72 PyN2.11 14.0 77.0 4.78 19.6 4.11 0.739 2.60 0.296 1.50 0.236 0.591 0.079 0.498 0.077 9.61 4.57 126 2.13 1.04 2.07 0.50 4.13 PyN2.12 12.7 67.6 4.21 17.8 3.54 0.643 2.29 0.252 1.31 0.202 0.553 0.068 0.455 0.061 8.22 4.06 112 2.09 1.04 2.06 0.53 3.89 PyN2.13 9.62 49.4 3.21 13.4 2.67 0.494 1.68 0.184 0.928 0.154 0.385 0.055 0.358 0.047 6.51 2.94 83 2.01 1.08 1.98 0.53 3.73 PyN2.14 9.60 49.0 3.17 13.5 2.66 0.513 1.69 0.177 0.928 0.147 0.385 0.051 0.327 0.048 7.14 2.87 82 2.01 1.12 2.16 0.53 4.06 PyN2.15 9.58 50.1 3.17 13.2 2.65 0.495 1.71 0.188 0.929 0.146 0.398 0.051 0.342 0.053 7.46 2.92 83 2.06 1.07 2.06 0.53 3.88 PyN2.16 9.99 58.0 3.44 14.5 2.94 0.558 1.92 0.220 1.08 0.171 0.451 0.060 0.412 0.053 8.14 3.17 94 2.23 1.09 1.79 0.50 3.57 PyN2.17 7.96 53.3 2.62 11.0 2.32 0.433 1.49 0.171 0.839 0.135 0.350 0.050 0.323 0.048 7.64 2.58 81 2.65 1.08 1.82 0.51 3.59 Mean 7 46 2.3 10 1.9 0.4 1.3 0.14 0.7 0.12 0.30 0.04 0.26 0.04 6.0 2.3 71 2.95 1.06 2.07 0.62 3.44 s 4 18 1.4 6 1.2 0.2 0.7 0.08 0.4 0.06 0.17 0.02 0.14 0.02 2.2 1.3 32 0.90 0.04 0.15 0.10 0.48 Table A1.2. Concentrations of REE, expressed as mg/kg, and REE parameters. Note: TREE = REE sum (from La to Yb).