Human modification of land cover alters net primary productivity, species richness and their relationship

Shuyu Deng*, Colin M. Beale, Philip J. Platts, Chris D. Thomas

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


Aim: Humans have altered ecosystem productivity and biodiversity worldwide by changing land-cover types and management. High local species richness is commonly found in geographic areas and ecosystems with high net primary productivity (NPP), but the long-term effects of modification on productivity and biodiversity change, and particularly on the relationship between the two, are poorly understood. Here we evaluate whether human modification tends to increase biodiversity in low-productivity ecosystems (where human management is likely to increase productivity) and decrease biodiversity in ecosystems that were originally high-productive. Location: Global. Time Period: 2001–2013. Major Taxa Studied: Plants, mammals, birds, hexapods, arachnids, other terrestrial invertebrates, amphibians, reptiles and fungi. Methods: We assembled a large worldwide dataset of NPP and associated species richness from MODIS land cover and NPP products and the PREDICTS biodiversity database, involving 11,849 sites. This enabled comparisons of species richness and NPP differences between samples of relatively natural and human-modified vegetation within the same geographic regions, considering 102 types of land-cover transitions. Results: (1) Modification from non-forest vegetation to forests on average increased NPP by 7.76%, and vice versa. (2) Human modification of less productive areas tended to increase NPP and vice versa, with stronger effects of major modification. (3) However, site-level species richness decreases were associated with nearly all land-cover transitions from relatively natural to modified vegetation. (4) Despite expectations, we found no significant relationship between species richness differences (between relatively natural and modified vegetation) and productivity differences, when considering conversions across land-cover types and excluding human-appropriated productivity in croplands. Main Conclusions: Human modification tends to increase NPP in low-productivity ecosystems, but species richness declines are associated with most major human-induced land-cover changes. Therefore, increasing or decreasing NPP did not generate corresponding increases or decreases in species richness, contrary to expectations of the general species richness-productivity relationship.

Original languageEnglish
Pages (from-to)385-399
Number of pages15
JournalGlobal ecology and biogeography
Issue number3
Early online date4 Dec 2023
Publication statusPublished - 13 Feb 2024

Bibliographical note

Funding Information:
The research was funded by a China Scholarship Council (202004910318) scholarship and supported by a Leverhulme Trust Research Centre grant (RC‐2018‐021)—the Leverhulme Centre for Anthropocene Biodiversity. We thank Lizzie Wandrag for commenting on the manuscript and contributing to the design. The raw species data were downloaded from the PREDICTS database (2016 release) produced by Lawrence H et al. (2016). Net primary productivity maps and land use maps downloaded from MODIS products—MOD17A3HGF and MCD12Q1. We thank all the contributors to the databases.

Publisher Copyright:
© 2023 The Authors. Global Ecology and Biogeography published by John Wiley & Sons Ltd.


  • biodiversity change
  • global change
  • human-associated drivers
  • land-cover change
  • PREDICTS project

Cite this