Neanderthal Hunting: Understanding Spatiotemporal Variation in Zooarchaeological Assemblages

This repository contains the dataset, processing workflows, and analytical code for evaluating spatial and temporal variations in Neanderthal (Homo neanderthalensis) foraging and prey choice strategies across Eurasia (200 ka to 20 ka BP).

This work was conducted as part of the STAMP project at the Globe Institute (University of Copenhagen) and the NeanderEdge project at Aarhus University.

---

📌 Project Overview & Objectives

Understanding Paleolithic hunting strategies often involves testing whether foragers acted as megafauna specialists or flexible opportunistic hunters. This project evaluates Neanderthal prey choice by applying the Prey Choice Model (PCM) from Behavioral Ecology to a large zooarchaeological dataset.

Key Objectives:

1. Database Construction & Standardization: Process and standardize taxonomic data from Paleolithic sites across Eurasia dating between 200 ka and 20 ka BP.
2. Prey Ranking: Classify target genera into profitability rank categories (High, Medium, Low/Megafauna) based on Post-Encounter Return Rates (PERR).
3. Temporal Analysis: Evaluate shifts in prey proportions across archaeological periods (Middle vs. Upper Paleolithic) and binned temporal sequences using LOESS smoothing and Generalized Additive Models (GAMs).
4. Statistical Testing: Test PCM predictions regarding diet expansion using Generalized Linear Models (GLMs).
5. Ecological Integration: Pair zooarchaeological prey abundances with species-specific habitat suitability models derived from Ecological Niche Models (ENMs), specifically for reindeer (Rangifer tarandus).

---

Dataset Summary

The primary dataset is derived from the Role of Culture in Early Expansions of Humans Out of Africa Database (ROAD).

• Initial Dataset: 9,422 zooarchaeological assemblages across Western Eurasia.
• Filtering Criteria: Restricting observations to contexts with absolute dates (radiometric, luminescence, amino acid, ESR) and valid NISP counts for class Mammalia.
• Taxonomic Standardization: Focused on 21 target genera. Proportional abundances were calculated by excluding rodent remains (Rodentia) to eliminate background noise: $$\text{Standardized Proportion} = \frac{\text{Genus NISP}}{\text{Total Mammalia NISP} - \text{Rodentia NISP}}$$

Prey Rank Classification (12 Target Taxa):

Rank Category	Genera	Notes / Description
High	Rangifer, Sus, Dama, Cervus	Highest post-encounter return rates (most profitable)
Medium	Alces, Equus, Bison, Saiga	Intermediate returns
Low / Megafauna	Mammuthus, Coelodonta, Megaloceros, Palaeoloxodon	Lower returns or significantly higher handling costs

---