Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
I
filtering (such as geology, soil, or climate) and
ncreasing evidence suggests that the mod- accessible forests (2, 3). At the basin scale, the biotic interactions (such as animal seed dispersal
ern floristic composition and structure of magnitude to which pre-Columbian peoples trans- and predation) drive differences among species
Amazonian forests have been influenced by formed forests is still unclear (4, 5). Humans assemblages across ecological gradients. For ex-
past human activity (1). Seasonal forests and transformed forests in many ways, through plant ample, effective seed dispersal of large-seeded
river margins are thought to have been mod- cultivation (preceded by cutting and burning), tree species decreases in heavily hunted forests
ified more intensively than were wetter and less seed dispersal and propagation, and in situ tend- because of the depletion of large vertebrates (20).
ing of useful resources, such as domesticated Composition and dominance patterns of plant
plants (6, 7). assemblages in Amazonian forests differ from
*Corresponding author. Email: carollevis@gmail.com and
carolina.levis@wur.nl (C.L.); hans.tersteege@naturalis.nl (H.t.S.)
Domestication of plant populations is a result one phytogeographical region to another (17, 19);
The full list of author affiliations is available in the supplementary of the human capacity to overcome selective vary along spatial and temporal gradients of rain-
materials. pressures of the environment by creating land- fall (19, 21, 22), terrain water saturation (23), and
soil fertility (19); and may be the result of dis- base S1). We found that 20 of these 85 domes- anthropogenic soils in Amazonia (26). Our anal-
persal limitation (20). We evaluated whether the ticated species are hyperdominants: five times yses also accounted for the effects of different
plant domestication process acted together with higher than the number of hyperdominant spe- geological regions of Amazonia and for four local
evolutionary and environmental processes to cies expected by chance. environmental conditions: soil cation exchange
determine the ecological patterns documented We then tested whether forests closer to ar- capacity (CEC), soil pH, rainfall seasonality, and
in Amazonian forests. chaeological sites and rivers have higher abun- height above the nearest drainage (HAND; a
Using 1170 forest plots of the Amazon Tree dance and richness of domesticated species. proxy for water-table depth). These variables were
Diversity Network (ATDN), ter Steege and co- Forest composition was evaluated in association selected because they influence forest composi-
authors (17) identified 4962 species, estimated that with numerous types of archaeological sites, in- tion in Amazonia (19, 2123) and are available
about 16,000 woody species occur in Amazonia, cluding pre-Columbian habitation sites (with for basin-wide analysis.
and showed that only 227 hyperdominant species and without anthropogenic soils), earthworks We found a significantly higher abundance
dominate Amazonian forests. We used 1091 ATDN (mounds, causeways, raised fields, and terraces), and richness (in absolute and relative terms) of
plots located in nonflooded lowland Amazonian rock art (paintings and petroglyphs), and iden- domesticated species in southwestern Amazo-
forests to provide a list of domesticated species tified eco-archaeological regions (fig. S2) (1, 24). nian forests, followed by northwestern, southern,
on the basis of evidence of at least incipient do- We included eco-archaeological regions in the and eastern forests, and the lowest values in the
mestication processes in Amazonia and else- analysis because they indicate environmental Guiana Shield (Fig. 2 and fig. S3). The total num-
where in the Americas. We identified 85 woody settings, with large and abundant pre-Columbian ber of individuals of domesticated species per
species with populations incipiently, semi-, or earthworks (25). We also used margins of nav- hectare (abundance) ranged from 0 to 292, and
fully domesticated by pre-Columbian peoples igable rivers as proxies for pre-Columbian set- the total number of domesticated species (rich-
Fig. 1. Distribution maps of five domesticated hyperdominant species cies has not been recorded. Shading shows the interpolated distribution of
in Amazonian forests and their probable origins of domestication. (A to each species by use of loess spatial interpolation (17). The range of relative
E) Distribution maps were estimated for five domesticated species that are abundance in plots (RelAb) and the loess spatial interpolation in individual
hyperdominants: (A) Bertholletia excelsa, (B) Inga ynga, (C) Pourouma grid cells (fit) are reported in percentage above each map. Maps were
cecropiifolia, (D) Pouteria caimito, and (E) Theobroma cacao. The origin of created with custom R scripts. Amazonia was divided into six geological
domestication is shown by the symbol +++ for known origin and by the regions (NWA, northwestern Amazonia; SWA, southwestern Amazonia; SA,
symbol ++ for hypothetical origin (13, 42). Sizes of black dots indicate the southern Amazonia; CA, central Amazonia; GS, Guiana Shield; and EA, eastern
relative abundance of the domesticated species in plots where the species Amazonia). [Base map source (country.shp, rivers.shp): Esri (www.esri.
has been recorded. Red dots indicate plots where each domesticated spe- com/data/basemaps, copyright Esri, DeLorme Publishing Company.]
ranged from 0 to 61%, and the relative richness fig. S5). In four of the six Amazonian regions, in the Guiana Shield, the opposite was observed
ranged from 0 to 19%. Forests with a diverse as- the relative and absolute richness of domesti- for the distance from archaeological sites within
semblage of domesticated species tended to have cated species decreased with distance from ar- this region. One possible explanation is insuffi-
a high abundance of these species (fig. S4). The chaeological sites or rivers, and in three of these cient information about the distribution of ar-
abundance of all domesticated species was, how- four regions the relative and absolute abundance chaeological sites along tributary rivers, so moving
ever, mostly due to 20 hyperdominant species. of domesticated species also decreased with dis- away from a known archaeological site may in-
Domesticated hyperdominant species were more tance from archaeological sites or rivers. These crease the proximity to other sites that have not
widespread across Amazonian forests than were results reveal that forests closer to archaeological been mapped yet. Archaeological surveys into
nondomesticated hyperdominant species. We sites or rivers within these regions harbor a interfluves of major rivers in Central Amazonia
found that 70% of the 20 domesticated hyperdom- richer and larger assemblage of domesticated documented numerous anthropogenic soils along
inant species studied here occur in all Amazonian species than forests elsewhere. The relative abun- tributary rivers, showing that these areas were
regions (database S1) versus only 47% of the 207 dance of domesticated hyperdominant species also densely occupied (27).
nondomesticated hyperdominant species (17). also decreased with distance from archaeolog- The map showing the density of archaeo-
Most of the domesticated species that are hyper- ical sites (Fig. 4). In contrast, we tested whether logical sites in 1-grid cells (areas of ~110 km2)
dominant have incipiently domesticated pop- nondomesticated hyperdominant species in three indicated large areas of Amazonia without any
ulations, rather than fully domesticated ones. control groups were negatively affected by the archaeological site (fig. S6) and revealed that
This finding suggests that humans were prob- distance from archaeological sites, and we did some plots with high values of the relative abun-
ably managing hyperdominant species in forests not find a significant negative relation for any dance of domesticated species are located in
instead of investing their efforts to fully domes- control group (Fig. 4). Additionally, nondomes- these grid cells, most likely reflecting the lack
Table 1. Mean, median, minimum, and maximum values of all human and environmental variables used in the multiple regression models. Values
were calculated at the Amazonia-wide level (All) and region level (NWA, northwestern Amazonia; SWA, southwestern Amazonia; SA, southern Amazonia; CA,
central Amazonia; GS, Guiana Shield; and EA, eastern Amazonia).
Region (number Values Distance to archaeological Distance to Cation exchange pH N dry months HAND
of plots) sites (km) main rivers (km) capacity (cmol/kg)
help predict the occurrence of archaeological of archaeological sites along tributaries in inter- explained most (up to 30%) of the variation in
sites in Amazonian forests. Guiana Shield plots, fluvial areas is critical. the relative abundance and richness of domes-
for example, with an average of 30% of indi- Environmental conditions also controlled the ticated species in Amazonian regions (Fig. 5),
viduals of domesticated species located close to abundance and richness of domesticated species whereas the proxies for past human impacts
river margins, but more than 120 km away from (Fig. 3 and fig. S6) and may have influenced explained up to 20%. Approximately 70% of the
an archaeological site, can be used to test this where and how humans shaped forests through variation remains unexplained by either human
hypothesis and indicate that a widespread survey time. We found that environmental conditions or environmental factors in most of the regions.
Fig. 2. Spatial variation of 85 domesticated species across Amazonia. from 0 to 292 individuals of domesticated species per 1 ha and 0 to 61% of the
(A to D) Maps showing (A) the spatial variation of the total number of individuals total number of individuals, and the observed values of absolute richness (C) and
of domesticated species (abundance) per hectare (ha), (B) the relative abun- relative richness (D), ranging from 0 to 19 domesticated species per plot and 0
dance of domesticated species, (C) the total number of domesticated species to 19% of the total number of species. The white-green background shows the
(richness) per plot, and (D) the relative richness of domesticated species in interpolation of the observed values (in percent) in each plot modeled as a
lowland plots in six geological regions of Amazonia (NWA, northwestern function of latitude and longitude on a 1-grid cell scale by use of loess
Amazonia; SWA, southwestern Amazonia; SA, southern Amazonia; CA, central spatial interpolation (17). Maps were created with custom R scripts. [Base
Amazonia; GS, Guiana Shield; and EA, eastern Amazonia). Black circles show the map source (country.shp, rivers.shp): Esri (www.esri.com/data/basemaps,
observed values of absolute abundance (A) and relative abundance (B), ranging copyright Esri, DeLorme Publishing Company)].
Fig. 4. The relative abundance of hyperdominant species as a function number of dry months, and HAND). Circle size represents the relative con-
of human and environmental variables. (A to D) Standardized regression tribution of the predictors, shown by standardized coefficients at the Amazonia-
coefficients for (A) the relative abundance of 20 domesticated species that are wide level (All) and region-level regression models (NWA, northwestern
hyperdominants, (B) the relative abundance of 20 nondomesticated species Amazonia; SWA, southwestern Amazonia; SA, southern Amazonia; CA, central
that are hyperdominants and primarily dispersed by primates, (C) the relative Amazonia; GS, Guiana Shield; and EA, eastern Amazonia). Red circles indicate
abundance of 20 nondomesticated species that are hyperdominants and not negative effects, and blue circles indicate positive effects. Standardized
dispersed by primates, and (D) the relative abundance of 20 nondomesticated coefficients are presented only for significant relations analyzed in the models
species that are hyperdominants selected at random, as a function of human (P 0.05). Adjusted R2 and significant codes (***P 0.001; **P 0.01; *P
variables (distance to archaeological sites and eco-archaeological regions, and 0.05; ns, P > 0.05) are presented for the effect of regions at the Amazonia-wide
distance to navigable rivers) and environmental variables (soil CEC, soil pH, level (All) and all regression models.
The data available for this broad-scale analysis mesticated species on soils with lower pH. Plots vironmental setting had wide geographical dis-
modern forests, despite the fact that the location 23. J. Schietti et al., Plant Ecol. Divers. 7, 241253 8), CNPqSWE (207400/2014-8), CNPq Universal (307807-2009-6),
of many archaeological sites is unknown. Almost (2014). CNPq Universal (479599/2008-4), CNPq Universal 458210/2014-5, and
24. Materials and methods as well as supplementary text are CNPq Universal 303851/2015-5; Colciencias; Fundao de Amparo
one fourth of all domesticated species are hyper- available as supplementary materials. Pesquisa do Estado do Amazonas (FAPEAM) projects with Fundao de
dominant, and besides their socioeconomic im- 25. U. Lombardo, E. Canal-Beeby, H. Veit, Geogr. Helv. 66, 173182 Amparo Pesquisa do Estado de So Paulo (09/53369-6 and 465/
portance, they can also help unravel the human (2011). 2010) and PRONEX-FAPEAM (1600/2006); Gordon and Betty Moore
history of Amazonian forests, which is largely 26. C. H. McMichael et al., Proc. R. Soc. London Ser. B 281, Foundation; Guyana Forestry Commission; Investissement dAvenir grant
20132475 (2014). of the French LAgence Nationale de la Recherche (ANR) (Centre dtude
overlooked by ecological studies. Detecting the 27. C. Levis et al., What do we know about the distribution of de la Biodiversit Amazonienne: ANR-10-LABX-0025); Instituto
widespread effect of ancient societies in modern Amazonian Dark Earth along tributary rivers in Central Amazonia? Venezolano de Investigaciones Cientficas of Venezuela; Lincoln Park Zoo;
forests not only strengthens efforts to conserve in Antes de OrellanaActas del 3er Encuentro Internacional de Margaret Mee Amazon Trust; Margot Marsh Foundation; Marie
domesticated and useful wild-plant populations, Arqueologa Amaznica (IFEA, ed. 1, 2014), pp. 305312. SklodowskaCurie/European Unions Horizon 2020 (706011); Ministrio
28. D. R. Piperno, Curr. Anthropol. 52, S453S470 (2011). da Cincia, Tecnologia e Inovao (MCTI)Museu Paraense Emlio
which is of critical importance for modern food 29. A. C. Roosevelt, Anthropocene 4, 6987 (2013). GoeldiProc. 407232/2013-3PVE-MEC/MCTI/CAPES/CNPq; Miquel
security (41), but also strongly refutes ideas of 30. M. Arroyo-Kalin, Quat. Int. 249, 418 (2012). fonds; Netherlands Foundation for the Advancement of Tropical
Amazonian forests being untouched by man. 31. E. G. Neves, J. B. Petersen, R. N. Bartone, C. A. da Silva, in Research WOTRO: grants WB85-335 and W84-581; Nuffic; Primate
Domestication shapes Amazonian forests. Amazonian Dark Earths, J. Lehmann, D.C. Kern, B. Glaser, Conservation; Stichting het van Eeden-fonds; Shell Prospecting and
W. I. Woods, Eds. (Springer, 2003), pp. 2950. Development of Peru; Tropenbos International; UniAndes; Variety Woods
32. R. S. Walker, L. A. Ribeiro, Proc. R. Soc. London Ser. B 278, Guyana; U.S. National Science Foundation Projects (DEB-0918591, DEB-
RE FE RENCES AND N OT ES 25622567 (2011). 1258112, and DEB-1556338) and Wenner-Gren Foundation; Venezuela
1. C. R. Clement et al., Proc. R. Soc. London Ser. B 282, 20150813 33. E. J. M. dos Santos, A. L. S. da Silva, P. D. Ewerton, National Council for Scientific Research and Technology (CONICIT);
(2015). L. Y. Takeshita, M. H. T. Maia, Bol. Mus. Para. Goeldi. Cincias Wageningen University (Interdisciplinary Research and Education Fund
2. M. B. Bush et al., J. Biogeogr. 42, 22772288 (2015). Humanas 10, 217228 (2015). Terra Preta program and FOREFRONT program); Aarhus University;
34. F. E. Mayle, M. J. Power, Philos. R. Trans. Soc. London B Biol. Wake Forest University; and WWF-Guianas and grants to RAINFOR from
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week
in December, by the American Association for the Advancement of Science, 1200 New York
Avenue NW, Washington, DC 20005. Copyright 2016 by the American Association for the
Advancement of Science; all rights reserved. The title Science is a registered trademark of AAAS.
Editor's Summary
Article Tools Visit the online version of this article to access the personalization and
article tools:
http://science.sciencemag.org/content/355/6328/925
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week
in December, by the American Association for the Advancement of Science, 1200 New York
Avenue NW, Washington, DC 20005. Copyright 2016 by the American Association for the
Advancement of Science; all rights reserved. The title Science is a registered trademark of AAAS.