The Neolithic Revolution narrative associates early-mid Holocene domestications with the development of agriculture that fueled the rise of late Holocene civilizations. This narrative continues to be influential, even though it has been deconstructed by archaeologists and geneticists in its homeland. To further disentangle domestication from reliance on food production systems, such as agriculture, we revisit definitions of domestication and food production systems, review the late Pleistocene–early Holocene archaeobotanical record, and quantify the use, management and domestication of Neotropical plants to provide insights about the past. Neotropical plant domestication relies on common human behaviors (selection, accumulation and caring) within agroecological systems that focus on individual plants, rather than populations—as is typical of agriculture. The early archaeobotanical record includes numerous perennial and annual species, many of which later became domesticated. Some of this evidence identifies dispersal with probable cultivation, suggesting incipient domestication by 10,000 years ago. Since the Pleistocene, more than 6500, 1206 and 6261 native plant species have been used in Mesoamerica, the Central Andes and lowland South America, respectively. At least 1555, 428 and 742 are managed outside and inside food production systems, and at least 1148, 428 and 600 are cultivated, respectively, suggesting at least incipient domestication. Full native domesticates are more numerous in Mesoamerica (251) than the Andes (124) and the lowlands (45). This synthesis reveals that domestication is more common in the Neotropics than previously recognized and started much earlier than reliance on food production systems. Hundreds of ethnic groups had, and some still have, alternative strategies that do involve domestication, although they do not rely principally on food production systems, such as agriculture.

The productivity of rainforests growing on highly-weathered tropical soils is expected to be limited by phosphorus (P) availability 1 . Yet, controlled fertilisation experiments have failed to demonstrate a dominant role for P in controlling tropical forest net primary productivity (NPP). Recent syntheses have demonstrated that responses to N addition are as large as to P 2 , and adaptations to low P availability appear to allow NPP to be maintained across major soil P gradients 3 . Thus, the extent to which P availability limits tropical forest productivity is highly uncertain. The majority of the Amazonia, however, is characterised by soils even more depleted in P than where most tropical fertilisation experiments have previously taken place 2 . Thus, we established the first P, nitrogen (N), and base cation addition experiment in an old growth Amazon rainforest, with the site's low soil P content representative of ~60% of the basin. Here we show that NPP increased exclusively with P addition. After 2 years, strong responses were observed in fine root (+29%) and canopy productivity (+19%), but not stem growth. The direct evidence of P limitation of NPP suggests that P availability may restrict Amazon forest responses to CO2 fertilisation 4 , with major implications for future carbon sequestration and forest resilience to climate change.

Plants have been used in Amazonian forests for millennia and some of these plants are disproportionally abundant (hyperdominant). At local scales, people generally use the most abundant plants, which may be abundant as the result of management of indigenous peoples and local communities. However, it is unknown whether plant use is also associated with abundance at larger scales. We used the population sizes of 4,454 arboreal species (trees and palms) estimated from 1946 forest plots and compiled information about uses from 29 Amazonian ethnobotany books and articles published between 1926 and 2013 to investigate the relationship between species usefulness and their population sizes, and how this relationship is influenced by the degree of domestication of arboreal species across Amazonia. We found that half of the arboreal species (2,253) are useful to humans, which represents 84% of the estimated individuals in Amazonian forests. Useful species have mean populations sizes six times larger than non-useful species, and their abundance is related with the probability of usefulness. Incipiently domesticated species are the most abundant. Population size was weakly related to specific uses, but strongly related with the multiplicity of uses. This study highlights the enormous usefulness of Amazonian arboreal species for local peoples. Our findings support the hypothesis that the most abundant plant species have a greater chance to be useful at both local and larger scales, and suggest that although people use the most abundant plants, indigenous people and local communities have contributed to plant abundance through long-term management.

No abstract

No abstract