Research during the late 20th and early 21st centuries found that traces of human intervention in vegetation in Southeast Asian and Australasian forests started extremely early, quite probably close to the first colonization of the region by modern people around or before 50,000 years ago. It also identified what may be insubstantial evidence for the translocation of economically important plants during the latest Pleistocene and Early Holocene. These activities may reflect early experiments with plants which evolved into agroforestry. Early in the Holocene, land management/food procurement systems, in which trees were a very significant component, seem to have developed over very extensive areas, often underpinned by dispersal of starchy plants, some of which seem to show domesticated morphologies, although the evidence for this is still relatively insubstantial. These land management/food procurement systems might be regarded as a sort of precursor to agroforestry. Similar systems were reported historically during early Western contact, and some agroforest systems survive to this day, although they are threatened in many places by expansion of other types of land use. The wide range of recorded agroforestry makes categorizing impacts problematical, but widespread disruption of vegetational succession across the region during the Holocene can perhaps be ascribed to agroforestry or similar land-management systems, and in more recent times impacts on biodiversity and geomorphological systems can be distinguished. Impacts of these early interventions in forests seem to have been variable and locally contingent, but what seem to have been agroforestry systems have persisted for millennia, suggesting that some may offer long-term sustainability.
Long-term dynamics of cocoa agroforestry systems established on lands previously occupied by savannah or forests
