Towards understanding human-environment feedback loops: the Atacama Desert case

Abstract

The overall trajectory for the human-environment interaction has been punctuated by demographic boom-and-bust cycles, phases of growth/overshooting as well as of expansion/contraction in productivity. Although this pattern has been explained in terms of an interplay between population growth, social upscaling, ecosystem engineering and climate variability, the evoked demographic-resource-complexity mechanisms have not been empirically tested. By integrating proxy data for population sizes, palaeoclimate and internal societal factors into empirical modelling approaches from the population dynamic theory, we evaluated how endogenous (population sizes, warfare and social upscaling) and exogenous (climate) variables module the dynamic in past agrarian societies. We focused on the inland Atacama Desert, where populations developed agriculture activities by engineering arid and semi-arid landscapes during the last 2000 years. Our modelling approach indicates that these populations experienced a boom-and-bust dynamic over the last millennia, which was coupled to structure feedback between population sizes, hydroclimate, social upscaling, warfare and ecosystem engineering. Thus, the human-environment loop appears closely linked with cooperation, competition, limiting resources and the ability of problem-solving. This article is part of the theme issue 'Evolution and sustainability: gathering the strands for an Anthropocene synthesis'.

Keywords: anthroecology; boom-and-bust cycles; ecosystem engineering; population dynamic theory; social upscaling; warfare.

Figure 1.

Study area and smoothed time series. (a) Spatial distribution of archaeological sites that provide radiocarbon dates used in our palaeodemographic reconstruction for the inland Atacama Desert. (b) Time series for past population levels inferred from the normalized (solid curve) and unnormalized (dotted curve) summed probability distributions (SPDs) of calibrated ¹⁴C-dates. (c) Time series for regional conflict incidence. The warfare index was calculated from bioanthropological data for interpersonal violence [23,24]. (d) Proxy for regional hydroclimate conditions based on lithic concentrations in the SO147–106KL marine core [25]. (e) Time series for the palaeopollution index as a proxy for social upscaling. This index builds on enrichment factors for lead and mercury from Illimani ice-cores [26,27].

Figure 2.

Comparison of observed unnormalized SPD data (blue curves) and predicted (red curves) population dynamic derived from the A₁ model (table 1) that includes a pure endogenous dynamic (a); the effect of warfare and hydroclimate on parameter c (Model B₁; b); and the effect of warfare and hydroclimate on parameter c and social upscaling on parameter w (Model C₁; c). Panel d incorporates the interplay of warfare and hydroclimate on parameter c, and social upscaling and warfare on parameter w (Model D₁).

Figure 3.

Comparison of observed normalized SPD data (blue curve) and predicted (red curves) population dynamic derived from the A₂ model (table 1) that includes a pure endogenous dynamic (a); the effect of warfare and hydroclimate on parameter c (Model B₂; b); and the effect of warfare and hydroclimate on parameter c and social upscaling on parameter w (Model C₂; c). Panel d incorporates the interplay of warfare and hydroclimate on parameter c, and social upscaling and warfare on parameter w (Model D₂).