Comparing machine learning isoscapes of 87Sr/86Sr ratios of plants on the island of Sardinia - Implications for the use of isoscapes for assessing the provenance of biological specimens

Aug 4, 2026

Upload by Skyler

Strontium isotope analysis evaluates the provenance and mobility of biological specimens by comparing an observed 87Sr/86Sr ratio with a strontium isoscape, which represents expected ratios across a landscape. Machine learning isoscapes produced using random forests are widely used following prior work, yet their accuracy—especially in geologically complex regions requiring local training—has been insufficiently tested. This study compares two published Sardinia ML isoscapes with new empirical ZANBA data, creating a third combined map and evaluating predictive performance.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages16

Reading Time40 min

AccessFree to download

Highlights

Abstract

1. Introduction

Isotopic baseline maps and isoscapes

Methods for creating isoscapes

Science of the Total Environment 989 (2025) 179880 Contents lists available at ScienceDirect Science of the Total Environment journal [homepage: www.elsevier.com/locate/scitotenv](homepage: www.elsevier.com/locate/scitotenv) | Comparing machine learning isoscapes of 87Sr/86Sr ratios of plants on the island of Sardinia: Implications for the use of isoscapes for assessing the provenance of biological specimens
Emily Holt a,*, Federico Luglib,c, Davide Schirrud, Melania Gigante e, Katie Faillace a, Marc-Alban Millet a, Morten Andersen a, Richard Madgwick a
a School of History, Archaeology and Religion, Cardiff University, John Percival Building, Colum Drive, Cardiff CF10 3EU, United Kingdom b Institut für Geowissenschaften, Goethe University Frankfurt, Altenh¨oferallee 1, Frankfurt am Main 60438, Germany
c Department of Chemical and Geological Sciences, University of Modena and Reggio Emilia, Via Giuseppe Campi 103, Modena 41125, Italy d Dipartimento di Scienze dell’Antichit`a, Universit`a degli Studi di Roma “La Sapienza”, Via dei Volsci 122, 00185 Roma, Italy e Department of Cultural Heritage, University of Padua, Piazza Capitaniato 7, 35139 Padova, Italy | | | --- | --- | H I G H L I G H T S A R T I C L E I N F O Editor: Bo Gao Keywords: Strontium isotopes Isoscapes Machine Learning Provenance studies Biosphere Mediterranean isotopic landscapes G R A P H I C A L A B S T R A C T | | | --- | | A B S T R A C T | Strontium isotope analysis is widely used to evaluate the provenance and mobility of biological specimens. Frequently applied in archaeology, palaeontology, ecology, forensics, and food science, strontium isotope analysis compares the 87Sr/86Sr ratio of a specimen against a strontium isoscape – a representation of expected 87Sr/86Sr ratios across a landscape – to identify areas that are more and/or less likely to be the source of the specimen. Strontium isoscapes are built using different methods, but all approaches start with empirical 87Sr/86Sr ratios sampled from areas with known coordinates and use them to assign likely 87Sr/86Sr ratios to unknown areas. Following the publication of Bataille et al., 2018 and Bataille et al., 2020, machine learning using a random forest algorithm has become a common method of producing strontium isoscapes. Despite the recognition that this method requires training with local ratios, especially in geologically complex regions, very little work has evaluated machine learning isoscapes’ accuracy. This study compares and evaluates two previously published machine learning isoscapes of Sardinia against new empirical data provided by the project ZANBA. * Corresponding author. E-mail [address:](address: HoltE@cardiff.ac.uk)[ HoltE@cardiff.ac.uk](address: HoltE@cardiff.ac.uk) (E. Holt). [https://doi.org/10.1016/j.scitotenv.2025.179880](https://doi.org/10.1016/j.scitotenv.2025.179880) Received 8 June 2024; Received in revised form 8 June 2025; Accepted 8 June 2025 Available online 14 June 2025 0048-9697/© 2025 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license ([http://creativecommons.org/licenses/by/4.0/](http://creativecommons.org/licenses/by/4.0/)). E. Holt et al. Science of the Total Environment 989 (2025) 179880 The ZANBA data is then used to create a third machine learning map of Sardinia, which is tested against previously published empirical data. The three isoscapes show different levels of predictive accuracy, with more primary data points leading to more correct predictions. However, a densely sampled landscape did not create anisoscape that gave substantially more accurate predictions than a moderately densely sampled landscape when tested against primary data from outside the original sampling areas. Areas of an isoscape with low root mean squared error (RMSE), which is often interpreted as indicating accuracy, did not necessarily give more correct predictions. Finally, a machine learning

FAQs

什么是strontium isoscape（锶同位素景观图）以及它如何用于判断来源？

锶同位素景观图用于表达不同地理区域的预期87Sr/86Sr比值。通过将样本测得的比值与景观图进行对比，可以推断样本来源区域更可能或更不可能的范围。

为什么需要评估机器学习（random forest）生成的isoscape准确性？

虽然随机森林机器学习能常用来生成isoscape，但该方法通常需要使用本地训练数据，尤其在地质复杂地区。现有研究对其预测准确度的评估仍然不足，因此需要针对具体区域进行验证。

本研究如何使用ZANBA新数据来改进并检验撒丁岛的isoscape？

研究将ZANBA项目提供的新实测数据与两套先前发表的机器学习isoscape进行对比，并基于新旧数据构建第三张机器学习地图。随后通过与既有实测数据的检验来评估不同地图的预测表现与误差特征。

Aug 4, 2026

Upload by Skyler