Human subsistence and signatures of selection on chemosensory genes.

Abstract

Chemosensation (olfaction, taste) is essential for detecting and assessing foods, such that dietary shifts elicit evolutionary changes in vertebrate chemosensory genes. The transition from hunting and gathering to agriculture dramatically altered how humans acquire food. Recent genetic and linguistic studies suggest agriculture may have precipitated olfactory degeneration. Here, we explore the effects of subsistence behaviors on olfactory (OR) and taste (TASR) receptor genes among rainforest foragers and neighboring agriculturalists in Africa and Southeast Asia. We analyze 378 functional OR and 26 functional TASR genes in 133 individuals across populations in Uganda (Twa, Sua, BaKiga) and the Philippines (Agta, Mamanwa, Manobo) with differing subsistence histories. We find no evidence of relaxed selection on chemosensory genes in agricultural populations. However, we identify subsistence-related signatures of local adaptation on chemosensory genes within each geographic region. Our results highlight the importance of culture, subsistence economy, and drift in human chemosensory perception.

Introduction

Transitions from hunting and gathering to agricultural food production were formative events during human history, beginning 11.5 kya and occurring independently in multiple geographic regions^1,2. Agricultural foodways have fueled profound changes––to human environments, behavior, health, and social structure^3,4,5, but also dietary ecology, resulting in diminished food diversity and shifts in the types and proportions of nutrients consumed^5,6,7. Mounting evidence from studies of modern and ancient human populations point to attendant selective pressures on the sequences and diversity of genes underlying digestion, metabolism and growth^{8,9,10,11,12,13}, sometimes convergently between agricultural origin centers^7,8,10,11. But the effects of agriculturalization on the human sensorium are understudied despite the essential role of chemosensation (olfaction, taste) during food detection and assessment (e.g., identifying fruit ripeness, the presence of toxins, or spoilage)^{7,14,15,16,17,18,19,20}.

Shifts in feeding behavior appear to drive changes in the olfactory receptor (OR) and taste receptor (TASR) genes of some species^{21,22,23,24,25}, which raises the possibility of similar differentiation between human populations. Transitions from forest-foraging to agriculturalism are expected to exert selective pressure on human chemosensory genes for at least two reasons. First, farming clears land in a way that reduces humidity and increases temperatures at ground-level, which increases convective air turbulence and hinders chemotaxis (detection and movement toward odorants)^26,27. Second, agriculturalists are more sedentary and reliant on a subset of available foods, mostly domesticates⁶, which limits their exposure to available chemosensory stimuli in the natural environment. Farming may therefore accompany a relaxation of natural selection, leading to impoverished chemosensory gene repertoires. Conversely, rainforest foragers are expected to retain greater numbers of functional OR genes for enhanced odor detection and discrimination^16,28,29. Exemplifying this pattern are language studies––the diversity of olfactory words and expressions is striking in the languages of some rainforest foragers³⁰, as are their abilities to name odors relative to nearby agriculturalists speaking closely-related languages^31,32. Taste sensitivities can also differ between rainforest foragers and neighboring farmers³³, and there is growing evidence of diet-driven haplotype variability at local scales^8,10,11,34. Yet when comparative studies of TASR genes have included foraging and farming peoples, the focus has been limited to single genes or a few gene regions^{15,29,35,36,37}, rather than a wider genomic-scale approach. Further, there was little effort to compare adjacent populations of foragers and farmers.

We have conducted recurring community-based participatory research since 2008. Our work in Uganda and the Philippines has been focused on Indigenous knowledge systems, belonging, and restorative justice among neighboring communities with different subsistence economies and histories (Fig. 1). This background motivated us to sequence the chemosensory gene repertoires (378 OR genes, 2 TAS1R genes, 24 TAS2R genes) of 133 individuals across six populations. Four of the populations – the Agta, Mamanwa, Sua, and Twa – turned increasingly, but at different moments and to varying degrees, toward agricultural foodways during the last century (Supplementary Note 1). We will refer to these populations as ‘traditional foragers’ to distinguish them from the long-term agriculturalists in our study, the BaKiga and Manobo (Supplementary Note 1). Although the word ‘traditional’ risks accusations of essentialism—or freezing people in “a simulacrum of pastness,” as Rifkin³⁸ put it—it best represents the lived experiences or collective memories of most study participants. This geographic-neighbors study design in Africa and Asia has two advantages. It accounts for higher levels of genetic diversity within Africans relative to non-Africans^29,39 and provides independent tests on the effects of parallel transitions toward agriculturalism.

Here we ask two complementary questions. First, does agriculturalism relax the selective constraints acting on chemosensory genes as evidenced by relatively higher frequencies of pseudogenes? And second, do signatures of positive selection or local adaptation exist as a function of subsistence history in one or both geographic regions, as determined using cross-population population differentiation based tests? Our overarching objective is to improve our understanding of human sensory evolution and perception.