Isotopic evidence of high reliance on plant food among Later Stone Age hunter-gatherers at Taforalt, Morocco

Abstract

The transition from hunting-gathering to agriculture stands as one of the most important dietary revolutions in human history. Yet, due to a scarcity of well-preserved human remains from Pleistocene sites, little is known about the dietary practices of pre-agricultural human groups. Here we present the isotopic evidence of pronounced plant reliance among Late Stone Age hunter-gatherers from North Africa (15,000-13,000 cal BP), predating the advent of agriculture by several millennia. Employing a comprehensive multi-isotopic approach, we conducted zinc (δ⁶⁶Zn) and strontium (⁸⁷Sr/⁸⁶Sr) analysis on dental enamel, bulk carbon (δ¹³C) and nitrogen (δ¹⁵N) and sulfur (δ³⁴S) isotope analysis on dentin and bone collagen, and single amino acid analysis on human and faunal remains from Taforalt (Morocco). Our results unequivocally demonstrate a substantial plant-based component in the diets of these hunter-gatherers. This distinct dietary pattern challenges the prevailing notion of high reliance on animal proteins among pre-agricultural human groups. It also raises intriguing questions surrounding the absence of agricultural development in North Africa during the early Holocene. This study underscores the importance of investigating dietary practices during the transition to agriculture and provides insights into the complexities of human subsistence strategies across different regions.

Fig. 1. Location of the Taforalt site in Morocco and the other sites mentioned in the text.

The circles indicate Iberomaurusian sites, the squares indicate European Upper Palaeolithic sites, the triangle indicates the Natufian site and the star indicates the Neolithic site in the Levant.

Fig. 2. Isotopic ratios of various elements from the human and faunal teeth/bone of Taforalt.

a, Carbon (δ¹³C) and nitrogen (δ¹⁵N) isotopic ratios from bulk collagen of dentine and bone samples. Each point corresponds to a sample; samples from the same individual are connected with a line. The typical analytical error is 0.1‰ for the two isotope systems. VPDB, Vienna PeeDee Belemnite; AIR, atmospheric N₂; Ind, individual. b, Zinc (δ⁶⁶Zn) and strontium (⁸⁷Sr/⁸⁶Sr) isotopic ratios from enamel bioapatite. Each point corresponds to a sample; samples from the same individual are connected with a line. The typical analytical error is 0.05‰ for δ⁶⁶Zn and 7 × 10⁻⁶ for ⁸⁷Sr/⁸⁶Sr. c, Carbon (δ¹³C) and nitrogen (δ¹⁵N) isotopic ratios from bulk collagen of dentine and bone samples with associated 95% confidence ellipses. Each point corresponds to the average value of all samples coming from a single individual (n_individual = 33; 44 samples in total); the error bars give the standard deviation for all the values from the same individual. d, Zinc (δ⁶⁶Zn) and strontium (⁸⁷Sr/⁸⁶Sr) isotope ratios from enamel bioapatite with associated 95% confidence ellipses. Each point corresponds to the average value of all samples coming from a single individual (n_individual = 33; 41 samples in total); the error bars give the standard deviation for all the values from the same individual.

Fig. 3. Zinc and nitrogen isotope values versus TP.

a, Zinc (δ⁶⁶Zn) isotope values versus the TP obtained from single amino acids (Supplementary Information Section 4). b, Nitrogen (δ¹⁵N) isotope values from bulk collagen versus the TP obtained from single amino acids. The TP was estimated from δ¹⁵N_Phe and δ¹⁵N_Glu values (Supplementary Information Section 2). Samples from the same human individual are connected with a line.

Fig. 4. Measured δ¹⁵N_Phe and δ¹⁵N_Glu values on human and faunal collagen from Taforalt.

a, Values according to the formation time of the sample. b, Values according to the species. The dashed black lines indicate approximately the theoretical TP of herbivores (TP = 2) and carnivores (TP = 3). The dashed grey line is the intermediate (TP = 2.5). Samples from the same human individual are connected with a line. RB, rib bone; LB, long bone; dm2, deciduous second molar.

Extended Data Fig. 1. Zinc isotope values in tooth enamel of various faunal species and the humans analyzed in this study (n_individual= 33, 41 samples in total).

Each point represents the mean value of a tooth. Boxes correspond to the median (centre line) and the first and third quartiles, while whiskers indicate the minimum and maximum values.

Extended Data Fig. 2. Carbon isotope values in amino acids of the humans and fauna from Taforalt.

PCA plot of the δ¹³C_Ala, δ¹³C_Asp, δ¹³C_Glx, δ¹³C_Gly, δ¹³C_Gly-Phe, δ¹³C_Lys, δ¹³C_Phe, δ¹³C_Pro, δ¹³C_val, δ¹³C_Val-Phe measurement with bulk δ¹³C and δ¹⁵N values from Taforalt and published data^,–.

Extended Data Fig. 3. Average δ⁶⁶Zn values per dental specimen of different individuals analyzed in this study.

Different parts of the tooth crowns were sampled so each point represents an average zinc isotope value of different samplings taken from the same tooth. Error bars are the standard deviation of different samples from the same dental specimen. (n_individual= 17, 45 samples in total).