Finding forest management in prehistoric Amazonia
Introduction
The Amazon Basin, an area approximately the size of the continental United States, is presently home to the largest contiguous expanse of Neotropical forest and much of the New World’s biological diversity (Fig. 1). The cultural complexity of pre-Columbian societies in Amazonia, together with their relationships with the natural environment, are enduring questions. Archaeologists have abandoned the older idea that prehistoric Amazonia supported mainly small and highly mobile human populations who exerted little influence on the environment (Meggers, 1954, Meggers, 1971). Evidence accumulated over the last 10-20 years makes it clear that, by 3000–500 calendar years ago (hereafter, cal BP), several Amazonian regions were occupied by large and culturally-complex pre-Columbian human societies who transformed their natural environments in sometimes profound ways. The evidence includes landscape alterations in the form of raised agricultural fields, fish weirs, mound settlements, roads, geometric earthworks called geoglyphs, and the presence of highly modified anthropic soils called terra pretas or “Amazonian Dark Earths” (Fig. 1) (e.g., Arroyo-Kalin, 2010; Blatrix et al., 2018; De Souza et al., 2018; Denevan, 2011; Erickson, 2000, Erickson, 2008; Heckenberger et al., 2003; Lombardo et al., 2013; Neves et al., 2004; Neves and Petersen, 2006; Pärssinen et al., 2009; Roosevelt, 1991; Schaan et al., 2012; Watling et al., 2017; Whitney et al., 2013; Woods et al., 2012). These features are documented largely along major rivers and their tributaries, in seasonal savannas/areas of poor drainage, or in highly seasonal interfluvial (terra firme) forests.
These findings have led some investigators to hypothesize, largely by extrapolating from available data from fluvial zones and studies of modern floristic composition of selected forests, that anthropogenic landscapes were widespread across Amazonia at the end of the prehistoric era (e.g., Balée, 2010, Balée, 2013; Clement et al., 2015; Dull et al., 2010; Erickson, 2008; Heckenberger, 2013; Heckenberger et al., 2003; Levis et al., 2012, Levis et al., 2017; Watling et al., 2017). An initial formulation of this view of Amazonia as a “domesticated” or “manufactured” landscape called for widespread forest clearance with fire for agriculture across most of the Basin (e.g., see W. Woods cited in Romero, 2012). Scholars have proposed that prehistoric fires and forest clearing were on such a massive scale that post-Columbian reforestation was a significant contributor to decreasing atmospheric CO2 levels and the onset of the ‘Little Ice Age’ (Dull et al., 2010; Nevle et al., 2011; Koch et al., 2019). However, neither lake and terrestrial soil paleo-vegetation and fire histories, nor peatland and ice core carbon source studies, nor correlations of climate and fire data, so far support widespread forest clearing or fire (e.g., Bush et al., 2007, Bush et al., 2016; Griscom and Ashton, 2006; Kelly et al., 2018; Maezumi et al., 2018a, Maezumi et al., 2018b; McMichael et al., 2012a, McMichael et al., 2012b; Piperno, 2011; Piperno et al., 2015; Piperno and Becker, 1996; Power et al., 2013; Stocker et al., 2017; Urrego et al., 2013; Watling et al., 2017).
Several investigators now instead hypothesize that widespread forms of agroforestry with planted, orchard-like formations or other forest management strategies involving the care and possible enrichment of several dozens of native species. Such proposed activities are thought to have led to human-created “hyperdominance” (after ter Steege et al., 2013) of some and forests with higher diversity than before human modification (e.g. Balée, 2010, Balée, 2013; Clement et al., 2015; Erickson, 2008; Levis et al., 2012, Levis et al., 2017, Levis et al., 2018; Watling et al., 2017). These arboreal formations are said to be still prominent and detectable in the modern vegetation (e.g., Levis et al., 2012, Levis et al., 2017). At this point, the evidence for these views relies largely on inventories of modern forest composition that are said to predict what were pre-Columbian activities (e.g., Levis et al., 2012, Levis et al., 2017), though modern vegetation surveys are typically conducted in areas with high densities of known archaeological sites (McMichael et al., 2017).
An alternative viewpoint is that, although the cultural practices depicted above no doubt existed in some Amazonian forests, much more paleo-environmental and archaeological research is required, particularly in the vast and understudied interfluves, for conclusions regarding the spatial and temporal scale, characteristics, and degree of human influence (e.g., Bush et al., 2015; McMichael et al., 2015a, McMichael et al., 2017; Piperno et al., 2015). We consider that modern floristic analyses are insufficient as documentation of the dynamics of the prehistoric past. Reasons include: 1) a lack of consideration of possible historical period disturbances, which were profound in some regions and could be mistaken for prehistoric legacies, 2) the paucity of information--due to sparse archaeobotanical data and spotty early historic-period records (see Miller and Nair, 2006)--concerning which and how many of the numerous tree species whose distributions and abundance are assumed to be modern relicts of past forest management were actively managed in prehistory, 3) a poor understanding of the natural distributions and abundance of many species in different habitats and climates through time, making uncertain whether their present abundance and distribution relates more to human activity, natural habitat preferences (e.g., swamps), or climate change during the past few thousand years, and 4) that differences among the forests in rainfall and its seasonality, soil quality, floristic composition, and distance to water bodies including lakes may have influenced human habitation.
For example, examining fossil pollen in 13 lake records from western Amazonia, Bush and McMichael (2016) found that increased abundance during the last 3000 years of a now-hyperdominant palm, Iriartea deltoidei Ruiz & Pav., likely related to increased precipitation rather than human influence. Another example is that high densities in certain areas today of the babaçu palm (Attalea speciose Mart. ex Spreng.), used heavily by Balée (1989) for estimating the extent of prehistoric human disturbance, probably resulted from its invasion following swidden cultivation and cattle ranching during the past 500 years (Forline, 2008). Similarly, recent studies of Brazil nut (Bertholletia excelsa Humb. & Bonpl.) stands in mature terra firme forests along the Madeira River in Brazil, thought by some to be a legacy of prehistoric nut management, date from the rubber boom (Scoles and Gribel, 2011). In a region of southeastern Peru the Brazil nut occurs in low densities indicative of limited ancient human influence there (Porcher et al., 2018).
Therefore, among the major research questions requiring further study are the spatial scale, degree, and persistence through time of various types of management of perennial tree species. How to best document these activities, however, is less clear. This uncertainty is due in part to the fact that they could be ‘silent’, since some major economic tree fruit and nut species produce few to no recognizable phytoliths and pollen (Piperno, 2006; Piperno et al., 2015). Examples include Brazil nut, rubber (Hevea brasiliensis Müll.Arg.), and cacao (Theobroma cacao L.). Table 1 lists other species considered useful and potentially enriched by humans in prehistory (largely after Levis et al., 2012, Levis et al., 2017) that will not provide useful phytolith records. On the other hand, critical and readily detected economic taxa include palms (Arecaceae family), which are prolific phytolith producers in all structures of the plant. Also importantly, because palms always produce abundant and diagnostic phytoliths negative evidence can be brought to bear, since their absence or rare frequencies through time indicates they were not growing nearby the locales and/or were not being enriched by human populations, also suggesting no human settlement nearby (e.g., McMichael et al., 2015a; Piperno et al., 2015).
Given the growing importance of considering possible forest exploitation that did not involve cultivation of annual plants and significant clearings for agricultural plots, paleo-ecologists should seek to develop new approaches for detecting and documenting those practices. This paper describes one such beginning effort through re-examinations and expansions of modern phytolith reference collections and re-analyses of soil phytoliths sampled from underneath Amazonian forests. In particular, it addresses the following research questions: 1) Can some economically important palm genera be identified and quantified in soil profiles, such that hypotheses discussed above about prehistoric influences on their present distributions and abundances can be empirically evaluated, and 2) Can prehistoric management of Amazonian forests overall be evaluated through analysis of non-palm woody phytoliths, such that changes to forest structure and diversity hypothesized by some scholars, discussed above, are elucidated.
Section snippets
Derivation of studied samples
We re-examined a sub-set of soil cores retrieved in 2008–2009 from directly underneath present-day terra firme and fluvial (river bluff) forest of western and central Amazonia (Fig. 1). A total of 109 sites distributed over seven regions provided the cores. The regions are (Fig. 1): 1) Porto Velho to Manaus (PV-M), a long 450-km interfluvial transect running between these two cities in the central Amazon Basin (ten out of 13 terra firme sites originally sampled were studied), 2) Los Amigos
Porto Velho to Manaus
The PV-M transect runs through well-drained terra firme forest about 20 to 60 km west of the Madeira River. Terra preta and archaeological sites concentrated along the river are from 20 to 80 km east of some of the sites sampled. We sampled all PV-M sites at 20-cm intervals, including for phytolith dating. The letters A and B for site numbers in the table for PV-M and other regions examined denote different cores randomly collected within a 100 m radius of the selected GPS point for each site.
Discussion
Forest management in the past, as in the present, likely involved activities on different intensity scales from planting, to removing non-useful species, to pruning, to encouraging or sparing useful species. The subtle characteristics of some of these strategies pose challenges for paleoecological detection. This study tested approaches designed to detect the presence and management of three palm species of considerable economic importance. The approaches employed here also examined at a finer
Conclusions
This study permits answers or preliminary answers to the research questions posed. First, this and other research (Morcote-Rios and Bernal, 2001) shows that it is possible to identify the important palm genera Oenocarpus and Euterpe based on phytolith morphology and size characteristics, and to follow their presence and abundance through time in soil profiles. This capacity allows robust empirical study of hypotheses concerning prehistoric human influences on their present distributions and
Acknowledgements
Supported by the NSF Ecology Program, the Smithsonian Institution including a Scholarly Studies award, the Smithsonian Tropical Research Institute, Smithsonian National Museum of Natural History (NMNH) including a Small Grant award, the Florida Institute of Technology, and the Wenner-Gren Foundation for Anthropological Research. We thank the NMNH and Missouri Botanical Garden for permission to sample their herbarium palm collections. We would also like to thank Eduardo Neves for providing the
References (75)
- et al.
A 6900-year history of landscape modification by humans in lowland Amazonia
Quat. Sci. Rev.
(2016) - et al.
Wood and bark phytoliths of West African woody plants
Quat. Int.
(2017) - et al.
Differentiation of neotropical ecosystems by modern soil phytolith assemblages and its implications for alaeoenvironmental and archaeological reconstructions
Rev. Palaeobot. Palyn.
(2013) - et al.
Earth system impacts of the European arrival and great dying in the Americas after 1492
Quat. Sci. Rev.
(2019) Human environment interactions in pre-Columbian Amazonia: The case of the Llanos de Moxos, Bolivia
Quat. Int.
(2013)- et al.
Fire- and distance-dependent recruitment of the Brazil nut in the Peruvian Amazon
For. Ecol. Manag.
(2018) - et al.
Finding the anthropocene in tropical forests
Anthropocene
(2018) - et al.
Differentiation of Neotropical ecosystems by modern soil phytolith assemblages and its implications for palaeoenvironmental and archaeological reconstructions II: Southwestern Amazonian forests
Rev. Paleobot. Paly.
(2016) - et al.
Pre-Columbian landscape impact and agriculture in the Monumental Mound region of the Llanos de Moxos, lowland Bolivia
Quat. Res.
(2013) - et al.
The polychrome tradition at the upper Madeira River
The Amazonian formative: crop domestication and anthropogenic soils
Diversity
The culture of Amazonian forests
Contingent diversity on anthropic landscapes
Diversity
Cultural Forests of the Amazon: A Historical Ecology of People and Their Landscapes
Tropical rainforest palm communities in Madre de Dios in Amazonian Peru
Rev. Peruana Biol
The unique functioning of a pre-Columbian Amazonian floodplain fishery
Sci. Rep.
Holocene variability of an Amazonian hyperdominant
J. Ecol.
Holocene fire and occupation in Amazonia: records from two lake districts
Philos. Trans. R. Soc. B
Anthropogenic influence on Amazonian forests in prehistory: an ecological perspective
J. Biogeogr.
Forest spatial heterogeneity and palm richness, abundance and community composition in Terra Firme forest, Central Amazon
Rev. Bras. Bot.
The domestication of Amazonia before European conquest
Proc. R. Soc. B
Pre-Columbian earth-builders settled along the entire southern rim of the Amazon
Nat. Comm.
The pristine myth revisited
Geogr. Rev.
The Columbian encounter and the little ice age: abrupt land use change, fire, and greenhouse forcing
Ann. Assoc. Am. Geogr.
Artificial landscape-scale fishery in the Bolivian Amazon
Nature
Amazonia: the historical ecology of a domesticated landscape
Putting history back into historical ecology: some perspectives on the recent human ecology of the Brazilian Amazon
J. Ecol. Anthropol.
Some Description of the Río Los Amigos, Madre de Dios, Peru
A self-perpetuating bamboo disturbance cycle in a neotropical forest
J. Trop. Ecol.
Who is Amazonia? The ‘salt of the matter’ for indigenous sustainability
Environ. Res. Lett.
Amazonia 1492: pristine forest or cultural parkland?
Science
The Palms of the Amazon
The palm community in a forest of Central Amazonia, Brazil
Biotropica
Continuous human presence without extensive reductions in forest cover over the past 2500 years in an aseasonal Amazonian rainforest
J. Quat. Sci.
Historical human footprint on modern tree species composition in the Purus-Madeira Interfluve, central Amazonia
PLoS One
Persistent effects of pre-Columbian plant domestication on Amazonian forest composition
Science
How people domesticated Amazonian forests
Front. Ecol. Evol.
Cited by (29)
Production of phytoliths in woody plants of the Atlantic Forest in islands of the Paraná River, Brazil
2024, Quaternary InternationalSpatial and temporal abilities of proxies used to detect pre-Columbian Indigenous human activity in Amazonian ecosystems
2023, Quaternary Science ReviewsSedimentological perspective on phytolith analysis in palaeoecological reconstruction
2023, Earth-Science ReviewsThe variability of Amazonian palm phytoliths
2022, Review of Palaeobotany and Palynology