Volume 111, Issue 6 p. 1308-1326
RESEARCH ARTICLE
Open Access

Evenness mediates the global relationship between forest productivity and richness

Iris Hordijk

Corresponding Author

Iris Hordijk

Institute of Integrative Biology, ETH Zurich (Swiss Federal Institute of Technology), Universitätsstrasse 16, 8092 Zurich, Switzerland

Correspondence

Iris Hordijk

Email: [email protected]

Search for more papers by this author
Daniel S. Maynard

Daniel S. Maynard

Institute of Integrative Biology, ETH Zurich (Swiss Federal Institute of Technology), Universitätsstrasse 16, 8092 Zurich, Switzerland

Search for more papers by this author
Simon P. Hart

Simon P. Hart

School of Biological Sciences, University of Queensland, Brisbane, Australia

Search for more papers by this author
Mo Lidong

Mo Lidong

Institute of Integrative Biology, ETH Zurich (Swiss Federal Institute of Technology), Universitätsstrasse 16, 8092 Zurich, Switzerland

Search for more papers by this author
Hans ter Steege

Hans ter Steege

Naturalis Biodiversity Centre, Leiden, The Netherlands

Systems Ecology, Free University Amsterdam, Amsterdam, The Netherlands

Search for more papers by this author
Jingjing Liang

Jingjing Liang

Department of Forestry and Natural Resources, Purdue University, West Lafayette, Indiana, USA

Search for more papers by this author
Sergio de-Miguel

Sergio de-Miguel

Department of Crop and Forest Sciences, University of Lleida, Lleida, Spain

Joint Research Unit CTFC—AGROTECNIO—CERCA, Solsona, Spain

Search for more papers by this author
Gert-Jan Nabuurs

Gert-Jan Nabuurs

Wageningen University and Research, Wageningen, The Netherlands

Search for more papers by this author
Peter B. Reich

Peter B. Reich

Department of Forest Resources, University of Minnesota, St Paul, Minnesota, USA

Hawkesbury Institute for the Environment, Western Sydney University, Penrith, New South Wales, Australia

Search for more papers by this author
Meinrad Abegg

Meinrad Abegg

Swiss Federal Institute for Forest, Snow and Landscape Research, WSL, Birmensdorf, Switzerland

Search for more papers by this author
C. Yves Adou Yao

C. Yves Adou Yao

UFR Biosciences, University Félix Houphouët-Boigny, Abidjan, Côte d'Ivoire

Search for more papers by this author
Giorgio Alberti

Giorgio Alberti

Faculty of Science and Technology, Free University of Bolzano, Bolzano, Italy

Department of Agricultural, Food, Environmental and Animal Sciences, University of Udine, Udine, Italy

Search for more papers by this author
Angelica M. Almeyda Zambrano

Angelica M. Almeyda Zambrano

Spatial Ecology and Conservation Laboratory, Center for Latin American Studies, University of Florida, Gainesville, Florida, 32611 USA

Search for more papers by this author
Braulio V. Alvarado

Braulio V. Alvarado

Forestry School, Tecnológico de Costa Rica TEC, Cartago, Costa Rica

Search for more papers by this author
Alvarez-Davila Esteban

Alvarez-Davila Esteban

Fundacion ConVida, Universidad Nacional Abierta y a Distancia, UNAD, Medellin, Colombia

Search for more papers by this author
Patricia Alvarez-Loayza

Patricia Alvarez-Loayza

Field Museum of Natural History, Chicago, Illinois, USA

Search for more papers by this author
Luciana F. Alves

Luciana F. Alves

Center for Tropical Research, Institute of the Environment and Sustainability, UCLA, Los Angeles, California, USA

Search for more papers by this author
Christian Ammer

Christian Ammer

Silviculture and Forest Ecology of the Temperate Zones, University of Göttingen, Göttingen, Germany

Search for more papers by this author
Clara Antón-Fernández

Clara Antón-Fernández

Division of Forest and Forest Resources, Norwegian Institute of Bioeconomy Research (NIBIO), Ås, Norway

Search for more papers by this author
Alejandro Araujo-Murakami

Alejandro Araujo-Murakami

Museo de Historia natural Noel kempff Mercado, Santa Cruz, Bolivia

Search for more papers by this author
Luzmila Arroyo

Luzmila Arroyo

Museo de Historia natural Noel kempff Mercado, Santa Cruz, Bolivia

Search for more papers by this author
Valerio Avitabile

Valerio Avitabile

European Commission, Joint Research Centre, Ispra, Italy

Search for more papers by this author
Gerardo A. Aymard C

Gerardo A. Aymard C

UNELLEZ-Guanare, Programa de Ciencias del Agro y el Mar, Herbario Universitario (PORT), Portuguesa, Venezuela Compensation International S. A. Ci Progress-GreenLife, Bogotá, Distrito Capital, Colombia

Search for more papers by this author
Timothy Baker

Timothy Baker

School of Geography, University of Leeds, Leeds, UK

Search for more papers by this author
Radomir Bałazy

Radomir Bałazy

Department of Geomatics, Forest Research Institute, Raszyn, Poland

Search for more papers by this author
Olaf Banki

Olaf Banki

Systems Ecology, Free University Amsterdam, Amsterdam, The Netherlands

Search for more papers by this author
Jorcely Barroso

Jorcely Barroso

Centro Multidisciplinar, Universidade Federal do Acre, Rio Branco, Brazil

Search for more papers by this author
Meredith L. Bastian

Meredith L. Bastian

Proceedings of the National Academy of Sciences, Washington, DC, USA

Department of Evolutionary Anthropology, Duke University, Durham, North Carolina, 27708 USA

Search for more papers by this author
Jean-Francois Bastin

Jean-Francois Bastin

Gembloux Agro Bio-Tech, University of Liege, Liege, Belgium

Search for more papers by this author
Luca Birigazzi

Luca Birigazzi

United Nation Framework Convention on Climate Change, Bonn, Germany

Search for more papers by this author
Philippe Birnbaum

Philippe Birnbaum

Cirad, UMR-AMAP, CNRS, INRA, IRD, Université de Montpellier, Montpellier, France

Search for more papers by this author
Robert Bitariho

Robert Bitariho

Institute of Tropical Forest Conservation, Mbarara University of Sciences and Technology, Mbarara, Uganda

Search for more papers by this author
Pascal Boeckx

Pascal Boeckx

Isotope Bioscience Laboratory—ISOFYS, Ghent University, Ghent, Belgium

Search for more papers by this author
Frans Bongers

Frans Bongers

Wageningen University and Research, Wageningen, The Netherlands

Search for more papers by this author
Olivier Bouriaud

Olivier Bouriaud

Integrated Center for Research, Development and Innovation in Advanced Materials, Nanotechnologies, and Distributed Systems for Fabrication and Control (MANSiD), Stefan cel Mare University of Suceava, Suceava, Romania

Search for more papers by this author
Pedro H. S. Brancalion

Pedro H. S. Brancalion

Department of Forest Sciences, Luiz de Queiroz College of Agriculture, University of São Paulo, Piracicaba, Brazil

Search for more papers by this author
Susanne Brandl

Susanne Brandl

Bavarian State Institute of Forestry, Freising, Germany

Search for more papers by this author
Roel Brienen

Roel Brienen

School of Geography, University of Leeds, Leeds, UK

Search for more papers by this author
Eben N. Broadbent

Eben N. Broadbent

Spatial Ecology and Conservation Laboratory, School of Forest, Fisheries, and Geomatics Sciences, University of Florida, Gainesville, Florida, 32611 USA

Search for more papers by this author
Helge Bruelheide

Helge Bruelheide

Institute of Biology, Geobotany and Botanical Garden, Martin Luther University Halle-Wittenberg, Halle-Wittenberg, Germany

German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig, Leipzig, Germany

Search for more papers by this author
Filippo Bussotti

Filippo Bussotti

Department of Agriculture, Food, Environment and Forest (DAGRI), University of Firenze, Florence, Italy

Search for more papers by this author
Roberto Cazzolla Gatti

Roberto Cazzolla Gatti

Biological Institute, Tomsk State University, Tomsk, Russia

Search for more papers by this author
Ricardo G. César

Ricardo G. César

Department of Forest Sciences, Luiz de Queiroz College of Agriculture, University of São Paulo, Piracicaba, Brazil

Search for more papers by this author
Goran Cesljar

Goran Cesljar

Department of Spatial Regulation, GIS and Forest Policy, Institute of Forestry, Belgrade, Serbia

Search for more papers by this author
Robin Chazdon

Robin Chazdon

Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, Connecticut, USA

Tropical Forests and People Research Centre, University of the Sunshine Coast, Maroochydore, Queensland, Australia

Search for more papers by this author
Han Y. H. Chen

Han Y. H. Chen

Faculty of Natural Resources Management, Lakehead University, Thunder Bay, Ontario, Canada

Search for more papers by this author
Chelsea Chisholm

Chelsea Chisholm

Institute of Integrative Biology, ETH Zurich (Swiss Federal Institute of Technology), Universitätsstrasse 16, 8092 Zurich, Switzerland

Search for more papers by this author
Emil Cienciala

Emil Cienciala

IFER—Institute of Forest Ecosystem Research, Jilove u Prahy, Czech Republic

Global Change Research Institute CAS, Brno, Czech Republic

Search for more papers by this author
Connie J. Clark

Connie J. Clark

Nicholas School of the Environment, Duke University, Durham, North Carolina, USA

Search for more papers by this author
David B. Clark

David B. Clark

Department of Biology, University of Missouri-St Louis, St Louis, Missouri, USA

Search for more papers by this author
Gabriel Colletta

Gabriel Colletta

Programa de Pós-graduação em Biologia Vegetal, Instituto de Biologia, Universidade Estadual de Campinas, Campinas, Brazil

Search for more papers by this author
David Coomes

David Coomes

Department of Plant Sciences and Conservation Research Institute, University of Cambridge, Cambridge, UK

Search for more papers by this author
Fernando Cornejo Valverde

Fernando Cornejo Valverde

Andes to Amazon Biodiversity Program, Madre de Dios, Peru

Search for more papers by this author
Jose J. Corral-Rivas

Jose J. Corral-Rivas

Facultad de Ciencias Forestales, Universidad Juárez del Estado de Durango, Durango, Mexico

Search for more papers by this author
Philip Crim

Philip Crim

Department of Physical and Biological Sciences, The College of Saint Rose, Albany, New York, USA

Department of Biology, West Virginia University, Morgantown, West Virginia, USA

Search for more papers by this author
Jonathan Cumming

Jonathan Cumming

Department of Biology, West Virginia University, Morgantown, West Virginia, USA

Search for more papers by this author
Selvadurai Dayanandan

Selvadurai Dayanandan

Biology Department, Centre for Structural and Functional Genomics, Concordia University, Montreal, Quebec, Canada

Search for more papers by this author
André L. de Gasper

André L. de Gasper

Natural Science Department, Universidade Regional de Blumenau, Blumenau, Brazil

Search for more papers by this author
Mathieu Decuyper

Mathieu Decuyper

Wageningen University and Research, Wageningen, The Netherlands

World Agroforestry (ICRAF), P.O. Box 30677 00100 Nairobi, Kenya

Search for more papers by this author
Géraldine Derroire

Géraldine Derroire

Cirad, UMR EcoFoG (AgroParistech, CNRS, INRAE, Université des Antilles, Université de la Guyane), Campus Agronomique, Kourou, French Guiana

Search for more papers by this author
Ben DeVries

Ben DeVries

Department of Geographical Sciences, University of Maryland, College Park, Maryland, USA

Search for more papers by this author
Ilija Djordjevic

Ilija Djordjevic

Institute of Forestry, Belgrade, Serbia

Search for more papers by this author
Amaral Iêda

Amaral Iêda

National Institute of Amazonian Research, Manaus, Brazil

Search for more papers by this author
Aurélie Dourdain

Aurélie Dourdain

Cirad, UMR EcoFoG (AgroParistech, CNRS, INRAE, Université des Antilles, Université de la Guyane), Campus Agronomique, Kourou, French Guiana

Search for more papers by this author
Engone Obiang Nestor Laurier

Engone Obiang Nestor Laurier

IRET, Herbier National du Gabon (CENAREST), Libreville, Gabon

Search for more papers by this author
Brian Enquist

Brian Enquist

Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, Arizona, USA

Queensland Herbarium, Department of Environment and Science, Toowong, Queensland, Australia

Search for more papers by this author
Teresa Eyre

Teresa Eyre

Queensland Herbarium, Department of Environment and Science, Toowong, Queensland, Australia

Search for more papers by this author
Adandé Belarmain Fandohan

Adandé Belarmain Fandohan

Ecole de Foresterie et Ingénierie du Bois, Université Nationale d'Agriculture, Ketou, Benin

Search for more papers by this author
Tom M. Fayle

Tom M. Fayle

Biology Centre of the Czech Academy of Sciences, Institute of Entomology, Ceske Budejovice, Czech Republic and Institute for Tropical Biology and Conservation, Universiti Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia

Search for more papers by this author
Leandro V. Ferreira

Leandro V. Ferreira

Museu Paraense Emílio Goeldi. Coordenação de Ciências da Terra e Ecologia, Belém, Pará, Brazil

Search for more papers by this author
Ted R. Feldpausch

Ted R. Feldpausch

Geography, College of Life and Environmental Sciences, University of Exeter, Exeter, UK

Search for more papers by this author
Leena Finér

Leena Finér

Natural Resources Institute Finland (Luke), Joensuu, Finland

Search for more papers by this author
Markus Fischer

Markus Fischer

Institute of Plant Sciences, University of Bern, Bern, Switzerland

Search for more papers by this author
Christine Fletcher

Christine Fletcher

Forest Research Institute Malaysia, Kuala Lumpur, Malaysia

Search for more papers by this author
Lorenzo Frizzera

Lorenzo Frizzera

Department of Sustainable Agro-Ecosystems and Bioresources, Research and Innovation Center, Fondazione Edmund Mach, San Michele all'Adige, Italy

Search for more papers by this author
Javier G. P. Gamarra

Javier G. P. Gamarra

Forestry Division, Food and Agriculture Organization of the United Nations, Rome, Italy

Search for more papers by this author
Damiano Gianelle

Damiano Gianelle

Department of Sustainable Agro-Ecosystems and Bioresources, Research and Innovation Center, Fondazione Edmund Mach, San Michele all'Adige, Italy

Search for more papers by this author
Henry B. Glick

Henry B. Glick

School of Forestry and Environmental Studies, Yale University, New Haven, Connecticut, USA

Search for more papers by this author
David Harris

David Harris

Royal Botanic Garden Edinburgh, Edinburgh, UK

Search for more papers by this author
Andrew Hector

Andrew Hector

Department of Plant Sciences, University of Oxford, Oxford, UK

Search for more papers by this author
Andreas Hemp

Andreas Hemp

Department of Plant Systematics, University of Bayreuth, Bayreuth, Germany

Search for more papers by this author
Geerten Hengeveld

Geerten Hengeveld

Wageningen University and Research, Wageningen, The Netherlands

Search for more papers by this author
Bruno Hérault

Bruno Hérault

Cirad, UPR Forêts et Sociétés, University of Montpellier, Montpellier, France

Department of Forestry and Environment, National Polytechnic Institute (INP-HB), Yamoussoukro, Côte d'Ivoire

Search for more papers by this author
John Herbohn

John Herbohn

Tropical Forests and People Research Centre, University of the Sunshine Coast, Maroochydore, Queensland, Australia

Search for more papers by this author
Annika Hillers

Annika Hillers

Centre for Conservation Science, The Royal Society for the Protection of Birds, Sandy, UK

Wild Chimpanzee Foundation, Liberia Office, Monrovia, Liberia

Search for more papers by this author
Eurídice N. Honorio Coronado

Eurídice N. Honorio Coronado

Instituto de Investigaciones de la Amazonía Peruana, Iquitos, Peru

Search for more papers by this author
Cang Hui

Cang Hui

Department of Mathematical Sciences, Centre for Invasion Biology, Stellenbosch University, Stellenbosch, South Africa

Theoretical Ecology Unit, African Institute for Mathematical Sciences, Cape Town, South Africa

Search for more papers by this author
Hyunkook Cho

Hyunkook Cho

Division of Forest Resources Information, Korea Forest Promotion Institute, Seoul, South Korea

Search for more papers by this author
Thomas Ibanez

Thomas Ibanez

Institut Agronomique néo-Calédonien (IAC), Equipe Sol & Végétation (SolVeg), Nouméa, New Caledonia

Search for more papers by this author
Il Bin Jung

Il Bin Jung

Division of Forest Resources Information, Korea Forest Promotion Institute, Seoul, South Korea

Search for more papers by this author
Nobuo Imai

Nobuo Imai

Department of Forest Science, Tokyo University of Agriculture, Tokyo, Japan

Search for more papers by this author
Andrzej M. Jagodzinski

Andrzej M. Jagodzinski

Institute of Dendrology, Polish Academy of Sciences, Kórnik, Poland

Department of Game Management and Forest Protection, Poznań University of Life Sciences, Poznań, Poland

Search for more papers by this author
Bogdan Jaroszewicz

Bogdan Jaroszewicz

Faculty of Biology, Białowieża Geobotanical Station, University of Warsaw, Białowieża, Poland

Search for more papers by this author
Vivian Johanssen

Vivian Johanssen

Department of Geosciences and Natural Resource Management, University of Copenhagen, Copenhagen, Denmark

Search for more papers by this author
Carlos A. Joly

Carlos A. Joly

Department of Plant Biology, Institute of Biology, University of Campinas, UNICAMP, Campinas, Brazil

Search for more papers by this author
Tommaso Jucker

Tommaso Jucker

School of Biological Sciences, University of Bristol, Bristol, UK

Search for more papers by this author
Viktor Karminov

Viktor Karminov

Forestry Faculty, Bauman Moscow State Technical University, Mytischi, Russia

Search for more papers by this author
Kuswata Kartawinata

Kuswata Kartawinata

Field Museum of Natural History, Chicago, Illinois, USA

Search for more papers by this author
Elizabeth Kearsley

Elizabeth Kearsley

CAVElab-Computational and Applied Vegetation Ecology, Department of Environment, Ghent University, Ghent, Belgium

Search for more papers by this author
David Kenfack

David Kenfack

CTFS-ForestGEO, Smithsonian Tropical Research Institute, Balboa, Panama

Search for more papers by this author
Deborah Kennard

Deborah Kennard

Department of Physical and Environmental Sciences, Colorado Mesa University, Grand Junction, Colorado, USA

Search for more papers by this author
Sebastian Kepfer-Rojas

Sebastian Kepfer-Rojas

Department of Geosciences and Natural Resource Management, University of Copenhagen, Copenhagen, Denmark

Search for more papers by this author
Gunnar Keppel

Gunnar Keppel

UniSA STEM and Future Industries Institute, University of South Australia, Adelaide, South Australia, Australia

Search for more papers by this author
Mohammed Latif Khan

Mohammed Latif Khan

Department of Botany, Dr Harisingh Gour Vishwavidyalaya (A Central University), Sagar, 470003 MP, India

Search for more papers by this author
Timothy Killeen

Timothy Killeen

Museo de Historia natural Noel kempff Mercado, Santa Cruz, Bolivia

Search for more papers by this author
Hyun Seok Kim

Hyun Seok Kim

Department of Agriculture, Forestry and Bioresources, Seoul National University, Seoul, South Korea

Interdisciplinary Program in Agricultural and Forest Meteorology, Seoul National University, Seoul, South Korea

National Center for Agro Meteorology, Seoul, South Korea

Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, South Korea

Search for more papers by this author
Kanehiro Kitayama

Kanehiro Kitayama

Graduate School of Agriculture, Kyoto University, Kyoto, Japan

Search for more papers by this author
Michael Köhl

Michael Köhl

Institute for World Forestry, University of Hamburg, Hamburg, Germany

Search for more papers by this author
Henn Korjus

Henn Korjus

Institute of Forestry and Rural Engineering, Estonian University of Life Sciences, Tartu, Estonia

Search for more papers by this author
Florian Kraxner

Florian Kraxner

Ecosystems Services and Management, International Institute for Applied Systems Analysis, Laxenburg, Austria

Search for more papers by this author
Diana Laarmann

Diana Laarmann

Department of Geography, University College London, London, UK

Search for more papers by this author
Mait Lang

Mait Lang

Institute of Forestry and Rural Engineering, Estonian University of Life Sciences, Tartu, Estonia

Search for more papers by this author
Simon Lewis

Simon Lewis

School of Geography, University of Leeds, Leeds, UK

Department of Geography, University College London, London, UK

Search for more papers by this author
Huicui Lu

Huicui Lu

Faculty of Forestry, Qingdao Agricultural University, Qingdao, China

Search for more papers by this author
Natalia Lukina

Natalia Lukina

Center for Forest Ecology and Productivity, Russian Academy of Sciences, Moscow, Russia

Search for more papers by this author
Brian Maitner

Brian Maitner

Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, Arizona, USA

Search for more papers by this author
Yadvinder Malhi

Yadvinder Malhi

School of Geography, University of Oxford, Oxford, UK

Search for more papers by this author
Eric Marcon

Eric Marcon

UMR EcoFoG, AgroParisTech, Kourou, France

Search for more papers by this author
Beatriz Schwantes Marimon

Beatriz Schwantes Marimon

Departamento de Ciências Biológicas, Universidade do Estado de Mato Grosso, Nova Xavantina, Brazil

Search for more papers by this author
Ben Hur Marimon-Junior

Ben Hur Marimon-Junior

Departamento de Ciências Biológicas, Universidade do Estado de Mato Grosso, Nova Xavantina, Brazil

Search for more papers by this author
Andrew Robert Marshall

Andrew Robert Marshall

Tropical Forests and People Research Centre, University of the Sunshine Coast, Maroochydore, Queensland, Australia

Flamingo Land Ltd, Kirby Misperton, UK

Department of Environment & Geography, University of York, York, UK

Search for more papers by this author
Emanuel Martin

Emanuel Martin

Department of Wildlife Management, College of African Wildlife Management, Mweka, Tanzania

Search for more papers by this author
Olga Martynenko

Olga Martynenko

Forestry Faculty, Bauman Moscow State Technical University, Mytischi, Russia

Search for more papers by this author
Jorge A. Meave

Jorge A. Meave

Departamento de Ecología y Recursos Naturales, Facultad de Ciencias, Universidad Nacional Autónoma de México, Mexico City, Mexico

Search for more papers by this author
Omar Melo-Cruz

Omar Melo-Cruz

Universidad del Tolima, Ibagué, Colombia

Search for more papers by this author
Casimiro Mendoza

Casimiro Mendoza

Colegio de Profesionales Forestales de Cochabamba, Cochabamba, Bolivia

Search for more papers by this author
Cory Merow

Cory Merow

Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, Connecticut, USA

Search for more papers by this author
Stanislaw Miscicki

Stanislaw Miscicki

Warsaw University of Life Sciences, Department of Forest Management, Dendrometry and Forest Economics, Warsaw, Poland

Search for more papers by this author
Abel Monteagudo Mendoza

Abel Monteagudo Mendoza

Jardín Botánico de Missouri, Oxapampa, Peru

Universidad Nacional de San Antonio Abad del Cusco, Cusco, Peru

Search for more papers by this author
Vanessa Moreno

Vanessa Moreno

Department of Forest Sciences, Luiz de Queiroz College of Agriculture, University of São Paulo, Piracicaba, Brazil

Search for more papers by this author
Sharif A. Mukul

Sharif A. Mukul

Tropical Forests and People Research Centre, University of the Sunshine Coast, Maroochydore, Queensland, Australia

Centre for Research on Land-use Sustainability, Dhaka, Bangladesh

Search for more papers by this author
Philip Mundhenk

Philip Mundhenk

Institute for World Forestry, University of Hamburg, Hamburg, Germany

Search for more papers by this author
Maria G. Nava-Miranda

Maria G. Nava-Miranda

Instituto de Silvicultura e Industria de la Madera, Universidad Juárez del Estado de Durango, Durango, Mexico

Search for more papers by this author
David Neill

David Neill

Universidad Estatal Amazónica, Puyo, Pastaza, Ecuador

Search for more papers by this author
Victor Neldner

Victor Neldner

The Santa Fe Institute, Santa Fe, New Mexico, USA

Search for more papers by this author
Radovan Nevenic

Radovan Nevenic

Institute of Forestry, Belgrade, Serbia

Search for more papers by this author
Michael Ngugi

Michael Ngugi

Queensland Herbarium, Department of Environment and Science, Toowong, Queensland, Australia

Search for more papers by this author
Pascal A. Niklaus

Pascal A. Niklaus

Department of Evolutionary Biology and Environmental Studies, University of Zürich, Zürich, Switzerland

Search for more papers by this author
Jacek Oleksyn

Jacek Oleksyn

Institute of Dendrology, Polish Academy of Sciences, Kórnik, Poland

Search for more papers by this author
Petr Ontikov

Petr Ontikov

Forestry Faculty, Bauman Moscow State Technical University, Mytischi, Russia

Search for more papers by this author
Edgar Ortiz-Malavasi

Edgar Ortiz-Malavasi

Forestry School, Tecnológico de Costa Rica TEC, Cartago, Costa Rica

Search for more papers by this author
Yude Pan

Yude Pan

Climate, Fire, and Carbon Cycle Sciences, USDA Forest Service, Durham, North Carolina, USA

Search for more papers by this author
Alain Paquette

Alain Paquette

Centre for Forest Research, Université du Québec à Montréal, Montréal, Quebec, Canada

Search for more papers by this author
Alexander Parada-Gutierrez

Alexander Parada-Gutierrez

Museo de Historia natural Noel kempff Mercado, Santa Cruz, Bolivia

Search for more papers by this author
Elena Parfenova

Elena Parfenova

V. N. Sukachev Institute of Forest, FRC KSC, Siberian Branch of the Russian Academy of Sciences, Krasnoyarsk, Russia

Search for more papers by this author
Minjee Park

Minjee Park

Department of Forestry and Natural Resources, Purdue University, West Lafayette, Indiana, USA

Department of Agriculture, Forestry and Bioresources, Seoul National University, Seoul, South Korea

Search for more papers by this author
Marc Parren

Marc Parren

Forest Ecology and Forest Management Group, Wageningen University & Research, Wageningen, The Netherlands

Search for more papers by this author
Narayanaswamy Parthasarathy

Narayanaswamy Parthasarathy

Department of Ecology and Environmental Sciences, Pondicherry University, Puducherry, India

Search for more papers by this author
Pablo L. Peri

Pablo L. Peri

Instituto Nacional de Tecnología Agropecuaria (INTA), Universidad Nacional de la Patagonia Austral (UNPA), Consejo Nacional de Investigaciones Científicas y Tecnicas (CONICET), Rio Gallegos, Argentina

Search for more papers by this author
Sebastian Pfautsch

Sebastian Pfautsch

School of Social Sciences (Urban Studies), Western Sydney University, Penrith, New South Wales, Australia

Search for more papers by this author
Oliver L. Phillips

Oliver L. Phillips

School of Geography, University of Leeds, Leeds, UK

Search for more papers by this author
Nicolas Picard

Nicolas Picard

GIP ECOFOR, Paris, France

Search for more papers by this author
Maria Teresa Piedade

Maria Teresa Piedade

Instituto Nacional de Pesquisas da Amazônia, Manaus, Brazil

Search for more papers by this author
Daniel Piotto

Daniel Piotto

Laboratório de Dendrologia e Silvicultura Tropical, Centro de Formação em Ciências Agroflorestais, Universidade Federal do Sul da Bahia, Itabuna, Brazil

Search for more papers by this author
Nigel C. A. Pitman

Nigel C. A. Pitman

Field Museum of Natural History, Chicago, Illinois, USA

Search for more papers by this author
Irina Polo

Irina Polo

Jardín Botánico de Medellín, Medellín, Colombia

Search for more papers by this author
Lourens Poorter

Lourens Poorter

Wageningen University and Research, Wageningen, The Netherlands

Search for more papers by this author
Axel Dalberg Poulsen

Axel Dalberg Poulsen

Royal Botanic Garden Edinburgh, Edinburgh, UK

Search for more papers by this author
John R. Poulsen

John R. Poulsen

Nicholas School of the Environment, Duke University, Durham, North Carolina, USA

Search for more papers by this author
Hans Pretzsch

Hans Pretzsch

Chair for Forest Growth and Yield Science, TUM School for Life Sciences, Technical University of Munich, Munich, Germany

Search for more papers by this author
Freddy Ramirez Arevalo

Freddy Ramirez Arevalo

Universidad Nacional de la Amazonía Peruana, Iquitos, Peru

Search for more papers by this author
Zorayda Restrepo-Correa

Zorayda Restrepo-Correa

Servicios Ecosistémicos y Cambio Climático (SECC), Fundación Con Vida & Corporación COL-TREE, Medellín, Colombia

Search for more papers by this author
Mirco Rodeghiero

Mirco Rodeghiero

Department of Sustainable Agro-Ecosystems and Bioresources, Research and Innovation Center, Fondazione Edmund Mach, San Michele all'Adige, Italy

Centro Agricoltura, Alimenti, Ambiente, University of Trento, San Michele all'Adige, Italy

Search for more papers by this author
Samir Rolim

Samir Rolim

Laboratório de Dendrologia e Silvicultura Tropical, Centro de Formação em Ciências Agroflorestais, Universidade Federal do Sul da Bahia, Itabuna, Brazil

Search for more papers by this author
Anand Roopsind

Anand Roopsind

Department of Biological Sciences, Boise State University, Boise, Idaho, USA

Search for more papers by this author
Francesco Rovero

Francesco Rovero

Department of Biology, University of Florence, Florence, Italy

Tropical Biodiversity, MUSE—Museo delle Scienze, Trento, Italy

Search for more papers by this author
Ervan Rutishauser

Ervan Rutishauser

Info Flora, Geneva, Switzerland

Search for more papers by this author
Purabi Saikia

Purabi Saikia

Department of Environmental Sciences, Central University of Jharkhand, Ranchi, India

Search for more papers by this author
Christian Salas-Eljatib

Christian Salas-Eljatib

Centro de Modelación y Monitoreo de Ecosistemas, Universidad Mayor, Santiago, Chile

Vicerrectoría de Investigación y Postgrado, Universidad de La Frontera, Temuco, Chile

Departamento de Silvicultura y Conservación de la Naturaleza, Universidad de Chile, Santiago, Chile

Search for more papers by this author
Peter Schall

Peter Schall

Silviculture and Forest Ecology of the Temperate Zones, University of Göttingen, Göttingen, Germany

Search for more papers by this author
Dmitry Schepaschenko

Dmitry Schepaschenko

Ecosystems Services and Management, International Institute for Applied Systems Analysis, Laxenburg, Austria

Search for more papers by this author
Michael Scherer-Lorenzen

Michael Scherer-Lorenzen

Faculty of Biology, Geobotany, University of Freiburg, Freiburg im Breisgau, Germany

Search for more papers by this author
Bernhard Schmid

Bernhard Schmid

Department of Geography, Remote Sensing Laboratories, University of Zurich, Zurich, Switzerland

Search for more papers by this author
Jochen Schöngart

Jochen Schöngart

Instituto Nacional de Pesquisas da Amazônia, Manaus, Brazil

Search for more papers by this author
Eric B. Searle

Eric B. Searle

Centre for Forest Research, Université du Québec à Montréal, Montréal, Quebec, Canada

Search for more papers by this author
Vladimír Šebeň

Vladimír Šebeň

National Forest Centre, Forest Research Institute Zvolen, Zvolen, Slovakia

Search for more papers by this author
Josep M. Serra-Diaz

Josep M. Serra-Diaz

Université de Lorraine, AgroParisTech, Inra, Silva, Nancy, France

Department of Biology, Center for Biodiversity Dynamics in a Changing World (BIOCHANGE), Aarhus University, Aarhus, Denmark

Search for more papers by this author
Douglas Sheil

Douglas Sheil

Wageningen University and Research, Wageningen, The Netherlands

Faculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, Ås, Norway

Search for more papers by this author
Anatoly Shvidenko

Anatoly Shvidenko

Ecosystems Services and Management, International Institute for Applied Systems Analysis, Laxenburg, Austria

Search for more papers by this author
Javier Silva-Espejo

Javier Silva-Espejo

Departamento de Biología, Universidad de la Serena, La Serena, Chile

Search for more papers by this author
Marcos Silveira

Marcos Silveira

Centro de Ciências Biológicas e da Natureza, Universidade Federal do Acre, Rio Branco, Acre, Brazil

Search for more papers by this author
James Singh

James Singh

Guyana Forestry Commission, Georgetown, French Guiana

Search for more papers by this author
Plinio Sist

Plinio Sist

Cirad, UPR Forêts et Sociétés, University of Montpellier, Montpellier, France

Search for more papers by this author
Ferry Slik

Ferry Slik

Environmental and Life Sciences, Faculty of Science, Universiti Brunei Darussalam, Bandar Seri Begawan, Brunei Darussalam

Search for more papers by this author
Bonaventure Sonké

Bonaventure Sonké

Plant Systematic and Ecology Laboratory, Department of Biology, Higher Teachers' Training College, University of Yaoundé I, Yaoundé, Cameroon

Search for more papers by this author
Alexandre F. Souza

Alexandre F. Souza

Departamento de Ecologia, Universidade Federal do Rio Grande do Norte, Natal, Brazil

Search for more papers by this author
Krzysztof Stereńczak

Krzysztof Stereńczak

School of Geography, University of Leeds, Leeds, UK

Search for more papers by this author
Jens-Christian Svenning

Jens-Christian Svenning

Université de Lorraine, AgroParisTech, Inra, Silva, Nancy, France

Section for Ecoinformatics & Biodiversity, Department of Biology, Aarhus University, Aarhus, Denmark

Search for more papers by this author
Miroslav Svoboda

Miroslav Svoboda

Faculty of Forestry and Wood Sciences, Czech University of Life Sciences, Prague, Czech Republic

Search for more papers by this author
Ben Swanepoel

Ben Swanepoel

Wildlife Conservation Society, Bronx, New York, USA

Search for more papers by this author
Natalia Targhetta

Natalia Targhetta

GIP ECOFOR, Paris, France

Search for more papers by this author
Nadja Tchebakova

Nadja Tchebakova

Centre for Forest Research, Université du Québec à Montréal, Montréal, Quebec, Canada

Search for more papers by this author
Raquel Thomas

Raquel Thomas

Iwokrama International Centre for Rainforest Conservation and Development (IIC), Georgetown, French Guiana

Search for more papers by this author
Elena Tikhonova

Elena Tikhonova

Faculty of Forestry, Qingdao Agricultural University, Qingdao, China

Search for more papers by this author
Peter Umunay

Peter Umunay

Forestry Division, Food and Agriculture Organization of the United Nations, Rome, Italy

Search for more papers by this author
Vladimir Usoltsev

Vladimir Usoltsev

Botanical Garden of Ural Branch of Russian Academy of Sciences, Ural State Forest Engineering University, Ekaterinburg, Russia

Search for more papers by this author
Renato Valencia

Renato Valencia

Pontificia Universidad Católica del Ecuador, Quito, Ecuador

Search for more papers by this author
Fernando Valladares

Fernando Valladares

LINCGlobal, Museo Nacional de Ciencias Naturales, CSIC, Madrid, Spain

Search for more papers by this author
Fons van der Plas

Fons van der Plas

Plant Ecology and Nature Conservation Group, Wageningen University, P.O. Box 47, Wageningen, 6700 AA The Netherlands

Search for more papers by this author
Do Van Tran

Do Van Tran

Silviculture Research Institute, Vietnamese Academy of Forest Sciences, Hanoi, Vietnam

Search for more papers by this author
Michael E. Van Nuland

Michael E. Van Nuland

Department of Biology, Stanford University, Stanford, California, USA

Search for more papers by this author
Rodolfo Vasquez Martinez

Rodolfo Vasquez Martinez

Warsaw University of Life Sciences, Department of Forest Management, Dendrometry and Forest Economics, Warsaw, Poland

Search for more papers by this author
Hans Verbeeck

Hans Verbeeck

Forestry Faculty, Bauman Moscow State Technical University, Mytischi, Russia

Search for more papers by this author
Helder Viana

Helder Viana

Centre for the Research and Technology of Agro-Environmental and Biological Sciences, CITAB, University of Trás-os-Montes and Alto Douro, UTAD, Vila Real, Portugal

Department of Ecology and Sustainable Agriculture, Agricultural High School of Polytechnic Institute of Viseu, Portugal and Centre for the Research and Technology of Agro-Environmental and Biological Sciences, CITAB, University of Trás-os-Montes and Alto Douro, 5001-801 Vila Real, Portugal

Search for more papers by this author
Alexander C. Vibrans

Alexander C. Vibrans

Biology Department, Centre for Structural and Functional Genomics, Concordia University, Montreal, Quebec, Canada

Department of Forest Engineering Universidade Regional de Blumenau, 89031 Blumenau, Santa Catarina, Brazil

Search for more papers by this author
Simone Vieira

Simone Vieira

Environmental Studies and Research Center, University of Campinas, UNICAMP, Campinas, Brazil

Search for more papers by this author
Klaus von Gadow

Klaus von Gadow

Department of Forest and Wood Science, University of Stellenbosch, Stellenbosch, South Africa

Search for more papers by this author
Hua-Feng Wang

Hua-Feng Wang

Key Laboratory of Tropical Biological Resources, Ministry of Education, School of Life and Pharmaceutical Sciences, Hainan University, Haikou, China

Search for more papers by this author
James Watson

James Watson

Division of Forestry and Natural Resources, West Virginia University, Morgantown, West Virginia, USA

Search for more papers by this author
Gijsbert D. A. Werner

Gijsbert D. A. Werner

Department of Zoology, University of Oxford, Oxford, UK

Search for more papers by this author
Susan K. Wiser

Susan K. Wiser

Manaaki Whenua–Landcare Research, Lincoln, New Zealand

Search for more papers by this author
Florian Wittmann

Florian Wittmann

Department of Wetland Ecology, Institute for Geography and Geoecology, Karlsruhe Institute for Technology, Karlsruhe, Germany

Search for more papers by this author
Verginia Wortel

Verginia Wortel

Centre for Agricultural Research in Suriname (CELOS), Paramaribo, Suriname

Search for more papers by this author
Roderick Zagt

Roderick Zagt

Tropenbos International, Wageningen, The Netherlands

Search for more papers by this author
Tomasz Zawila-Niedzwiecki

Tomasz Zawila-Niedzwiecki

Polish State Forests, Coordination Center for Environmental Projects, Warsaw, Poland

Search for more papers by this author
Chunyu Zhang

Chunyu Zhang

Research Center of Forest Management Engineering of State Forestry and Grassland Administration, Beijing Forestry University, Beijing, China

Search for more papers by this author
Xiuhai Zhao

Xiuhai Zhao

Research Center of Forest Management Engineering of State Forestry and Grassland Administration, Beijing Forestry University, Beijing, China

Search for more papers by this author
Mo Zhou

Mo Zhou

Systems Ecology, Free University Amsterdam, Amsterdam, The Netherlands

Search for more papers by this author
Zhi-Xin Zhu

Zhi-Xin Zhu

Key Laboratory of Tropical Biological Resources, Ministry of Education, School of Life and Pharmaceutical Sciences, Hainan University, Haikou, China

Search for more papers by this author
Irie Casimir Zo-Bi

Irie Casimir Zo-Bi

Cirad, UPR Forêts et Sociétés, University of Montpellier, Montpellier, France

Search for more papers by this author
Thomas W. Crowther

Thomas W. Crowther

Institute of Integrative Biology, ETH Zurich (Swiss Federal Institute of Technology), Universitätsstrasse 16, 8092 Zurich, Switzerland

Search for more papers by this author
First published: 02 May 2023
Citations: 2
Handling Editor: Ellen Simms

[Correction added on 10 May 2023, after first online publication: Author name Hyun Seok Kimhas been corrected.]

[Correction added on 28 July 2023, after first online publication: Author name Stanislaw Miscicki has been corrected.]

Abstract

1. Biodiversity is an important component of natural ecosystems, with higher species richness often correlating with an increase in ecosystem productivity. Yet, this relationship varies substantially across environments, typically becoming less pronounced at high levels of species richness. However, species richness alone cannot reflect all important properties of a community, including community evenness, which may mediate the relationship between biodiversity and productivity. If the evenness of a community correlates negatively with richness across forests globally, then a greater number of species may not always increase overall diversity and productivity of the system. Theoretical work and local empirical studies have shown that the effect of evenness on ecosystem functioning may be especially strong at high richness levels, yet the consistency of this remains untested at a global scale.

2. Here, we used a dataset of forests from across the globe, which includes composition, biomass accumulation and net primary productivity, to explore whether productivity correlates with community evenness and richness in a way that evenness appears to buffer the effect of richness. Specifically, we evaluated whether low levels of evenness in speciose communities correlate with the attenuation of the richness–productivity relationship.

3. We found that tree species richness and evenness are negatively correlated across forests globally, with highly speciose forests typically comprising a few dominant and many rare species. Furthermore, we found that the correlation between diversity and productivity changes with evenness: at low richness, uneven communities are more productive, while at high richness, even communities are more productive.

4. Synthesis. Collectively, these results demonstrate that evenness is an integral component of the relationship between biodiversity and productivity, and that the attenuating effect of richness on forest productivity might be partly explained by low evenness in speciose communities. Productivity generally increases with species richness, until reduced evenness limits the overall increases in community diversity. Our research suggests that evenness is a fundamental component of biodiversity–ecosystem function relationships, and is of critical importance for guiding conservation and sustainable ecosystem management decisions.

1 INTRODUCTION

In this era of diminishing biodiversity, understanding how changes in plant biodiversity will impact the functioning of ecosystems is critical (Ceballos et al., 2015; Isbell et al., 2017). Many studies show that the productivity of ecosystems typically increases with a greater number of species (Balvanera et al., 2006; Cardinale et al., 2007; Grace et al., 2016; Hooper et al., 2005). However, this positive relationship between productivity and diversity generally saturates and decays at high levels of species richness (Brun et al., 2019; Fei et al., 2018; Fraser et al., 2015; Liang et al., 2016; Schmid, 2002). Despite the consistency of this pattern across forests globally (Liang et al., 2016), we still lack a comprehensive understanding of the ecological relationships driving this attenuating effect of richness, which limits our capacity to relate diversity with productivity across the globe (Cardinale et al., 2012; Fraser et al., 2015).

The species redundancy hypothesis has been proposed to explain the saturating shape of the relationship between richness and ecosystem productivity (Cardinale et al., 2011; Loreau & Hector, 2001). This theory posits that, as the number of species in a region increases, functional redundancy increases and the proportional impact of any single species on ecosystem functioning declines (Cardinale et al., 2011; Gitay et al., 1996). However, another possible explanation for the diminishing effect of diversity on productivity is that species richness only captures one aspect of community structure, and does not reflect the true ‘diversity’ of a community (Caswell, 1976; Stirling & Wilsey, 2001) (Figure 1). By reflecting the homogeneity of species abundances within the community, evenness serves as the other central component of diversity (Jost, 2010; Peet, 1974; Tuomisto, 2012). The relationship between community productivity and evenness might differ from the relationship between productivity and richness.

Details are in the caption following the image
Four hypotheses describing how evenness interacts with the relationship between richness and productivity. The four levels of evenness are indicated with different colours from low evenness (yellow) to high evenness (red), and lower colour intensity indicates a lower data density. The dashed black line indicates the average relationship across the system. In all hypotheses, it is assumed that species richness has a positive effect on productivity (Balvanera et al., 2006; Cardinale et al., 2007; Hooper et al., 2005). If community evenness interacts with the effect of species richness on productivity, then (1) the effect of species richness on productivity depends on the evenness of the community, and (2) richness is correlated with evenness across the system (either positively or negatively). If there is no interaction between richness and evenness (Hypothesis A) or if there is no correlation between evenness and richness (Hypothesis B) (Ma, 2005), then the average effect (dashed line) of richness on productivity will neither attenuate nor increase at high richness levels. In such instances, the observed decrease in productivity at high richness levels is more likely a by-product of other ecological processes (e.g. functional redundancy). If, however, there is a significant interaction between richness and evenness, such that uneven communities have lower productivity at high richness (Hypothesis C), then a negative correlation between richness and evenness (Cook & Graham, 1996; Hanlin et al., 2000; Symonds & Johnson, 2008) would lead to an attenuation in productivity at high richness level: the marginal trend (dashed line) first tracks the high evenness isocline at low richness levels but then bends down towards the low evenness isoclines at high richness. Conversely, if uneven communities exhibit higher productivity at high richness levels (Hypothesis D), then a positive correlation between richness and evenness (Cotgreave & Harvey, 1994; Manier & Hobbs, 2006; Tramer, 1969) would explain this reduction in productivity: the marginal trend first tracks the low evenness isocline at low richness levels but then bends down towards the high evenness isoclines at high richness.

Broad-scale analyses suggest that speciose communities tend to be dominated by a few species, with a long tail of rare species (i.e. are highly uneven) (ter Steege et al., 2013). If such a negative relationship between richness and evenness holds across forests globally, then (i) speciose communities might tend to be dominated by a small number of species that have a disproportionate influence on ecosystem functioning, and (ii) the increasing richness of speciose communities might come at the expense of evenness. This trade-off might ultimately limit both functional diversity and ecosystem productivity.

Of course, a negative relationship between richness and evenness would only be relevant if both components of diversity influence the productivity of the community. Experimental studies provide clear evidence that community evenness has a direct effect on productivity (Kirwan et al., 2007; Sonkoly et al., 2019; Yan et al., 2021). Furthermore, if uneven communities tend to be dominated by one or a few species, the mass-ratio hypothesis suggests that increasing the number of rare species should have a relatively minimal impact on productivity (Grime, 1998; Lembrechts et al., 2018; Loreau, 1998). In contrast, in highly even communities, introducing a new species that is relatively abundant could substantially alter overall community productivity, for example through niche complementarity (see Figure 1) (Lembrechts et al., 2018; Nijs & Roy, 2000; Niklaus et al., 2017). Still, it remains unclear whether these effects of evenness mediate the effects of richness on productivity within complex natural communities and across biomes, because we do not know whether (i) the apparent negative relationship between richness and evenness is consistent across forests globally (Soininen et al., 2012; Zhang, John, et al., 2012) and (ii) there is an interactive effect of evenness and richness on productivity across a range of environmental conditions. Disentangling these relationships between richness, evenness and productivity is critical for understanding the mechanisms governing the diversity–productivity relationship across broad spatial scales (Figure 1).

Here, we used approximately 1 million forest inventory plots (sourced from the Global Forest Biodiversity Initiative database) to explore whether community evenness affects the relationship between richness and productivity across global forests. If evenness restricts the positive effect of richness on ecosystem productivity, then two things must hold: (i) richness trades off with evenness across forests globally, such that speciose communities tend to be dominated by a few species and (ii) there is an interactive effect of both richness and evenness on the productivity of forest ecosystems. Specifically, we tested the hypothesis that (i) high species richness generally comes at the expense of evenness, as the greater number of species within a community corresponds to a greater proportion of locally rare species, and (ii) that richness positively correlates with productivity in highly even, species-poor communities, but at the highest levels of species richness this positive relationship will break down due to an intrinsic reduction in evenness (Figure 1, Hyp. C).

2 MATERIALS AND METHODS

To evaluate the relationship between evenness and richness, we incorporated all 1,011,027 forest inventory plots from the Global Forest Biodiversity Initiative (GFBI database, 2021, including data from Condit et al., 2019a, 2019b). Forest plot size ranged from 0.0008 to 2.0 ha. Because evenness and richness inherently vary with scale (Gleason, 1922; Wilson et al., 1999), plot sizes smaller than 0.02 and larger than 1.5 ha were excluded to ensure comparable results. Indeed, the effect of plot size on richness is especially strong when plots are smaller than 0.02 ha (r = 0.39, r2 = 0.15, p < 0.001) or larger than 1.5 ha (r = −0.84, r2 = 0.71, p < 0.001). Elimination of these two plot size groups resulted in a filtered dataset with weak correlations between plot size and richness (r = 0.13, r2 = 0.02, p < 0.001), and plot size and evenness (r = −0.09, r2 = 0.008, p < 0.001). Although there is a range in plot sizes, 75% of the plots have a size between 0.02 and 0.06 ha. Quality controls of tree density estimations were conducted, and we removed plots with tree densities that fell outside the median ± 2.5 times the median absolute deviation, a moderately conservative threshold, within each biome (0.8% of total plots) (Leys et al., 2013). Additionally, we removed plots with unlikely biomass accumulation and productivity values (for details, see Section 2.2). The final dataset (Figure 2) comprised 896,276 forest plots, containing information on tree species richness and abundance. Of these, 367,565 plots contained diameter at breast height (DBH) information of individual trees which allowed us to estimate biomass. For both datasets, the plots were measured once between 1980 and 2017, with 2002 as the mean measurement year. The mean age of the forest where the plots are established is 52 years, with a standard deviation of 15 years (estimated with the global forest age map; Poulter et al., 2019).

Details are in the caption following the image
(a) Location of the Global Forest Biodiversity Initiative (GFBI) plots used in this study, where the density of forest plots is indicated from low density (blue) to high density (red). (b) The distribution of evenness and richness in the boreal, temperate and tropical biomes. An evenness value of one resembles either a monospecific stand or an even abundance of species. The tail of richness values of the tropical biomes extends to 380 species (not shown in the graph). The majority of our dataset is composed of secondary forests (mean age is 52 years), and especially the monospecific and relatively species-poor stands were affected by human activity in some degree.

Richness and evenness were calculated for each plot, enabling us to evaluate the nature of the relationship between these two components of diversity at biome and global scales. To examine the effect of evenness and richness on biomass accumulation and productivity, we fit linear models that simultaneously accounted for evenness and richness while controlling for a wide range of environmental variables, including temperature, precipitation and soil characteristics.

We estimated forest productivity using two complementary approaches. First, we used biomass estimates, calculated using regional allometric equations that included information about DBH and species information from each individual tree, and divided this information by forest age (Poulter et al., 2019) to approximate stand-level biomass accumulation over time. Second, we estimated net primary productivity (NPP) for each plot using MODIS satellite data (Running et al., 2011), thus providing both ground-sourced and remotely sensed estimates of productivity. Satellite-derived NPP estimates largely agreed with ground measured temporal data for the plots where this information was available (r = 0.65, p < 0.001), see Section 2.3 for more details.

2.1 Examining the relationship between species richness and evenness

Richness was defined as the total number of tree species per forest plot. We estimated Hill's evenness at the same plot level using the following equation:
Evenness = exp H / S , (1)
where H is Shannon's entropy, S is the species richness and exp(H) can be figuratively interpreted as an approximation of the number of relatively abundant species (Hill, 1973; Jost, 2010). Therefore, Hill's evenness can roughly be seen as the proportion of species that dominate the community in terms of abundance. Evenness values range from close to zero, when the community is dominated by only a few species, to one, where all species in the community have the same number of individuals.

There are many evenness indices available in the ecological literature (Tuomisto, 2012), and some commonly used indices are mathematically restricted by richness (including Pielou's J), reflect particularly dominance in the community (e.g. Simpson's evenness) or express evenness as the ratio between diversity and richness (including Hill's evenness). We also included the standardized Hill's evenness index in our supplementary analyses, as Hill's evenness reports an evenness of one in monospecific sites, which is avoided when using standardized Hill's evenness. As we explored the relationship between richness and evenness, each commonly used and distinct evenness index evaluated (standardized Hill's evenness, Pielou's J, Simpson's evenness, Evar, PIE, Eq) revealed similar overall trends (Figure S2) (Tuomisto, 2012). Given this overall consistency across different indices, including standardized Hill's evenness, we focus on the well-known Hill's evenness for the majority of our analyses because (i) it reflects evenness according to the current understanding of evenness as the ratio between diversity and richness (Tuomisto, 2012) and therefore (ii) consists of the mathematically robust components of Shannon's entropy and richness (Hill, 1973).

In addition to analysing the relationship between evenness and richness at a global scale and for the major forest biomes (boreal, temperate and tropical biomes), for every World Wildlife Fund (WWF) biome we selected a number of plots that is proportional to the forested area within that biome (Table S2; Olson et al., 2001). By taking this subset of the data, we reduced the sampling bias by assuring that the results are more representative for either the forests globally or the boreal, temperate or tropical biomes. Regarding the global dataset, the majority of the plots fall within the temperate broadleaf and mixed forest biome and a smaller proportion of the data represents the tropical moist forest, but with subsetting the data we create a more representative global dataset. The same procedure is followed when subsetting biomes to represent the boreal, temperate and tropical biomes. The proportions of forested area for every biome were calculated in Google Earth Engine by overlaying the WWF biomes with a global map of existing forest cover (Hansen et al., 2013). Areas with more than 10% canopy closure for vegetation taller than 5 m were defined as forests (FAO, 2000). To avoid heteroscedasticity due to the skewed nature of the species richness distribution, richness was log transformed, and the relationship was evaluated with a Pearson correlation (see Figure 3). To show that the relationship between evenness and richness is neither a mathematical artefact of the Hill's index nor dependent on the evenness index used, we visualized the results of a null model (Figure S1) and the relationship for commonly used evenness indices (Figure S2). The null model was created by evaluating the relationship between evenness and richness of 10.000 random data subsets, formed by drawing species according to a multinomial distribution from the dataset (see Figure S1 for a more detailed explanation). Additionally, to evaluate the effect of monospecific stands on the relationship between evenness and richness in forests globally, we compared the correlation coefficients of the relationship with and without monospecific stands.

Details are in the caption following the image
The relationship between evenness and logged species richness, where the data density is indicated from low density (blue) to high density (red). The Pearson correlation is highly significant for all relationships, we therefore describe the adjusted r2 as a measure of effect size. (a) The global relationship between evenness and richness (N = 20,272, r = −0.52, r2 = 0.28). (b) The relationship between evenness and richness for the boreal (N = 61,712, r = −0.42, r2 = 0.18), temperate (N = 374,142, r = −0.53, r2 = 0.28), and tropical (N = 9570, r = −0.48, r2 = 0.23) biomes.

2.2 Biomass estimation

To estimate the above-ground biomass of each tree in extratropical biomes, we used 430 species-specific DBH-based allometric equations obtained from the GlobAllomeTree database to estimate the above-ground biomass of each tree (Henry et al., 2013). These allometric equations use a common logarithmic equation for estimating above-ground biomass from DBH measures (Jenkins et al., 2003):
Biomass = e β 0 + β 1 × ln DBH , (2)
where biomass is the total above-ground biomass (kg dry weight), DBH is the measured diameter at breast height (cm), ln is the logarithm to the base e (2.718), and ß0 and ß1 are free parameters governing the effect of DBH on above-ground biomass. Following Jenkins et al. (2003), we applied back calculation to generate a pseudo dataset for biomass changes along DBH gradients based on each of the 430 allometric equations. To generate the pseudo data, we applied the following rules: (1) for a DBH between 5 and 25 cm, each centimetre was assigned a corresponding pseudo biomass value; (2) for a DBH between 25 and 100 cm, every 5 cm were assigned a corresponding value, (3) for a DBH between 100 and 300 cm (maximum DBH), every 10 cm were assigned a corresponding value. Consequently, we trained biome-specific allometric equations for each biome (varying in the β 0 and β 1 parameter estimates) based on the pseudo-DBH and biomass dataset (Olson et al., 2001) (Table S1; Figure S3).

Biomass estimations for the tropical biomes followed the allometric equations for pantropical regions from Chave et al. (2014), which are available through the r package ‘biomass’ (Réjou-Méchain et al., 2017). These equations require information on wood density, and we compiled species-specific wood density estimates from the Global Wood Density Database (Chave et al., 2009) and the BAAD database (Falster et al., 2015). To match the binomial species names between the GFBI and the wood density databases, we standardized species binomials using the Taxonomic Name Resolution Service platform (Boyle et al., 2013).

After computing the above-ground biomass for all 24 million individuals in our dataset, plot-level biomass values were obtained by summing up the biomass of all individuals in the respective plot. Biomass densities (t/ha) of each plot were obtained by dividing the total above-ground biomass (t) by the plot size (ha).

2.3 Biomass accumulation rates and productivity data

We selected approximately 95% of the data by excluding plots that had biomass values greater than two times the standard deviation above and below the mean plot-level biomass per biome to filter potential outliers in biomass values due to errors in measurement or data management. Additionally, we excluded plots with biomass values higher than 1 million kg/ha, as these values were likely overestimating biomass due to the presence of a large tree in the plot (Bastin et al., 2018; Chave et al., 2004; Slik et al., 2013). This filtering procedure resulted in 367,565 plots that were used in the final analyses (see Table S2 for an overview of the number of plots for the biomass and productivity analyses). All biomass values were divided by estimated forest age (Poulter et al., 2019), to control for differences in forest developmental stage that cause striking differences in forest biomass accumulation (Peichl & Arain, 2006; Poorter et al., 2016). As such, final biomass estimates reflect the mean annual biomass accumulation over time.

Biomass accumulation over time is not a precise indicator of forest productivity, as it cannot account for all of the ecological dynamics and disturbances that occurred during forest development, but it is a useful proxy for overall cumulative growth at the ecosystem level. However, to attempt to account for the annual variation in forest productivity, we used productivity estimates from an independent satellite-derived product. Specifically, we estimated NPP for every plot location from MODIS satellite images, using Google Earth Engine (Gorelick et al., 2017; Running et al., 2011) to supplement the forest biomass analysis. NPP is calculated as absorbed fraction of photosynthetically active radiation, which is a combination of leaf area index and fraction of photosynthetically active radiation, while taking temperature and water stress into account as well (Running et al., 2011). We calculated the mean productivity between 2000 (first year of NPP data availability) and 2009 (third quartile of forest age in our data), to obtain a robust NPP value and coincide the average NPP measurement year with the median forest age in our dataset. In the temperate biomes—the only region where sufficient DBH information was available across multiple years in the GFBI database—the ground-measured temporal changes in biomass were fairly well correlated with the satellite-derived productivity data (r = 0.65, p < 0.001). The lack of forest plots measured multiple times in the boreal and tropical biomes, precluded the estimation of the accuracy of the NPP data from these regions. We conducted all analyses across both mean annual biomass accumulation and NPP to explore the unifying trends that emerge across both approaches.

2.4 Evaluating the effect of evenness and richness on biomass accumulation and productivity

We used linear models to assess the effect of evenness, richness and their interaction, on NPP and biomass accumulation. The data met the linear model assumptions and evaluation of the model fit (Q–Q plots, distribution of residuals) suggested that linear models were applicable. To account for the fact that biomass accumulation does not necessarily vary linearly with time or successional state, we included time as a covariate in the model of biomass accumulation rate, which is equivalent to allowing the relationship between stand-level biomass and time to be quadratic (Poulter et al., 2019). Additionally, we controlled for the potentially confounding effects of climatic drivers and other environmental influences (Ali et al., 2019), including mean annual temperature, isothermality, annual precipitation, variation in seasonal precipitation (Hijmans et al., 2005), and soil carbon, sand content and pH in the upper 15 cm of the soil as covariates, each extracted from global maps at a 30-arc second resolution (Batjes et al., 2017; Ribeiro et al., 2018). These climate variables capture both the mean and variation in temperature and rainfall, while the edaphic variables include the most important soil drivers of biomass and productivity (Ali et al., 2019). Human impact was also considered in the linear models by including estimates of the percentage of human development, calculated as the percentage urban areas and managed vegetation per square kilometre (Tuanmu & Jetz, 2014), and population density as the number of people per square kilometre (Center for International Earth Science Information Network—CIESIN—Columbia University, 2016). The density of trees was included as an independent variable in the analyses, as this can affect both evenness (Wilson et al., 1999) and richness values (Lomolino, 2000). Additionally, ‘biome’ was included as a factor in the global model, and we accounted for plot size in all models (Poulter et al., 2019). To explore the possible effect of plot size, we analysed the relationship between evenness and richness and the effect of both on the two measures of productivity for small, medium and large plot sizes for the dataset globally (Figure S5). Additionally, we performed a sensitivity analysis on species richness, evenness and their interaction to plot size when predicting NPP or biomass accumulation rates in the global dataset (Figure S6A) and on biome level (Figure S6B–D).

Collinearity between variables in our model was evaluated using variance inflation factors (VIF). Although evenness and richness were negatively correlated, on average, there was substantial variation in this trend (Figure 3), such that the VIF values between evenness and richness were less than 3.0 in all models, and both were therefore included in every model (Becker et al., 2015). The model was implemented at the global scale, in which every biome was proportionally represented according to the extent of forested area within that biome (as described above), and at the biome level. At the biome level, the boreal, temperate and tropical biomes were represented by their largest forested ecosystems, respectively, the boreal forest, temperate broadleaf and mixed forest, and tropical moist forest. To visualize how richness and evenness interacted to affect productivity, we plotted the relationship between richness and productivity at four different levels of evenness (<0.4, 0.4–0.6, 0.6–0.8, and 0.8–1). The lowest evenness level <0.4 was chosen as this represents the lower end of the evenness values in the dataset, while including sufficient observations in this category. The subsequent categories were defined as an increase of 20% in the evenness values. For each evenness level, productivity was predicted with every model covariate (e.g. climate, soil, human impact) set at its median value. Thus, each predicted productivity is marginal to the aggregate trend at the global scale and for each biome. To examine whether the relationship between richness, evenness and productivity varied with evenness index, we evaluated the global trend for six commonly used evenness indices (Figure S7). We used a bootstrapping approach to incorporate uncertainty in the biomass calculations and satellite-derived NPP (Figure S8).

To quantify the relative importance of evenness, richness, their interaction, environmental variables and human impact, we used the scaled calc.relimp function in R (Grömping, 2006). This function evaluates the contribution of each independent variable to the variation explained, by averaging the contribution of each independent variable to the r2 in terms of its sum of squares across all possible fitting sequences. The statistical analyses were performed in R version 3.5.1 (R Core Team, 2018). The code used to perform the statistical analyses can be found at Github, following this link: tinyurl.com/3vvf52v9.

3 RESULTS

There was a consistent, negative correlation between evenness and richness at the global scale, which was apparent across all forest biomes (Figure 3a–d), scales (Figure S4) and plot sizes (Figure S5). This negative correlation was also found using several other commonly used evenness indices, including the normalized Hill's evenness index, Simpson's evenness, Evar index and PIE (Figure S2), suggesting that the relationship is robust to the evenness index used. Moreover, the magnitude of the negative correlation was more negative than expected at random, suggesting that the relationship between evenness and richness is not a mathematical artefact of the evenness metrics used (Figure S1). Instead, it demonstrates that highly speciose, highly even forest communities are less rare in nature than would be expected by chance (Figure S1), suggesting that biotic processes may play a role in shaping the relationship between richness and evenness. Additionally, excluding monospecific stands did not change the sign of the correlation, although the relationship was stronger when monospecific stands were included as they force the correlation through one (r = −0.08 for forests globally, r = −0.07 excluding monospecific stands). Within forests, species-poor communities tend to have a relatively uniform distribution of abundances, while among speciose communities, having more species is associated with increasingly uneven abundance distributions caused by a few dominant and many rare species (Figure 3a–d). Temperate and boreal biomes exhibited a saturating relationship, with evenness never extending below E = 0.75, on average, even at the highest richness levels.

However, after accounting for underlying environmental variation, we identified a significant positive relationship between richness and both biomass accumulation and productivity (0.06 < ß < 0.48, p < 0.01), supporting a recent global analysis (Liang et al., 2016) (but see also (Sheil & Bongers, 2020)). However, the relationship varied among regions, as temperate forests exhibited a marginally negative relationship between richness and productivity (ß = −0.09, p < 0.05) (Figure 4). Given the observed negative relationship between richness and evenness, increasing evenness was generally associated with lower biomass accumulation. However, this pattern was not observed in tropical forests (ß = 0.04, p < 0.01), where there was a positive relationship between evenness and productivity (0.004 < ß < 0.014, p < 0.01) (Figure 4).

Details are in the caption following the image
(a) Positive (green) and negative (blue) regression coefficients, and (b) variable importance of evenness, richness, the interaction of evenness and richness and climate, soil and human impact variables on biomass and productivity. Only the results for the boreal, temperate, tropical moist forest and all the biomes globally are visualized, and variables causing multicollinearity are taken out (see Section 2). The open circles in (a) indicate non-significant coefficients, while the filled circles indicate significant coefficients. The adjusted r2 values of the linear models are displayed in (a).

In combination, both the evenness and richness of forest communities were more strongly related to plot-level biomass accumulation than to productivity (25% vs. 19% of variance explained in the global model) (Figure 4). Among boreal forests, which are species poor, richness was more strongly related to both biomass accumulation and productivity than was evenness (variance explained by richnessBiomass = 10.1, evennessBiomass = 6.8 and richnessNPP = 14.8, evennessNPP = 0.4) (Figure 5b,f). In contrast, among moist tropical forests, which are species rich, productivity was similarly related to either richness or evenness (variance explained by richnessNPP = 1.8, evennessNPP = 1.6) (Figure 5d,h).

Details are in the caption following the image
The hypothesized effect of different levels of evenness on the relationship between species richness and mean annual biomass accumulation (a–d) or productivity (h–e). The graphs visualize predicted values based on the results of a linear model, with the covariates held constant (see methods), and as a cut-off point the third quantile of the biomass and NPP values to avoid overfitting. The data are projected on the graph (black line), and the 95% upper and lower confidence intervals are visualized in grey. At the right side of every graph the scaled variable importance, according to a linear model including covariates, of richness (green), evenness (dark blue), and the interaction between richness and evenness (light blue) is visualized. In the graphs at the left side, the global effect is visualized, while at the right side the data are split among boreal, temperate and moist tropical forests. The uncertainty of the biomass calculations and estimated productivity are visualized in Figure S7.

Experimentally increasing species richness increases biomass accumulation (Balvanera et al., 2006; Cardinale et al., 2007; Grace et al., 2016; Hooper et al., 2005), but observational data suggest that the effect becomes weaker or even reverses at high species richness (Brun et al., 2019; Fei et al., 2018; Fraser et al., 2015; Liang et al., 2016; Schmid, 2002). We hypothesized that if increasing community evenness causes productivity to decline, then the attenuating effect of richness on productivity could be explained by a negative correlation between species richness and community evenness (Figure 1). To explore if productivity data from the world's forests might be explained by such a hypothesis, we modelled the data with a regression featuring the hypothesized interaction between richness and evenness (Figure 5). The proposed interaction between richness and evenness is statistically significant in every model (p < 0.05), although the strength of the interaction varies. The strongest interactive effect was observed on biomass accumulation in the moist tropical forest, which are highly speciose (ß = 0.14) (Figure 5d). A weaker interaction was observed in temperate and boreal forests (ß = 0.03 and ß = −0.06, respectively), which contained fewer species (Figure 5b,c). In the tropics, biomass accumulation is predicted to vary by 2.5-fold between high- and low-evenness communities (6000 vs. 15,000 kg/ha year−1 at maximum richness) (Figure 5c), whereas this range is much lower in the temperate (4700–5000 kg/ha year−1 at maximum richness) (Figure 5c) and boreal systems (850–1150 kg/ha year−1 at maximum richness) (Figure 5b). Moreover, we find the same trend for productivity (Figure 5f–h), across different evenness indices (Figure S7), plot sizes (Figure S5) and when incorporating uncertainty of the biomass and productivity values (Figure S8). Additionally, the sensitivity analyses show that there is no consistent effect or no effect of plot size on the relationship of evenness, richness and their interaction on both measures of productivity (Figure S6).

4 DISCUSSION

Among forests across the globe, the positive correlation between species richness and ecosystem productivity appears to attenuate in the most speciose communities (Liang et al., 2016). Our analysis examines whether the evenness of plant communities might contribute to this attenuation. We observed a consistent negative correlation between richness and evenness across forests, globally (Figure 3), whereby highly specious communities exhibited relatively low levels of evenness. We also detected interactive effects of the correlation of richness and evenness with ecosystem productivity, which lends support to the hypothesis that evenness might mediate the impacts of richness on forest productivity. In short, when the number of species is relatively low, species richness correlates positively with productivity. However, in the most speciose communities, the long tail of rare species makes them highly uneven, which may limit the impact of diversity on productivity. As such, in these highly uneven communities, a greater number of species does not necessarily correlate with greater productivity.

Our results suggest that, when there are more species present, they are more likely to have an uneven distribution. However, this relationship between richness and evenness was significantly more negative for our observational data than would be expected under null-model predictions (Figure S1), highlighting this pattern is at least partially a by-product of ecological and evolutionary processes. The observed negative relationship between evenness and richness observed across forests globally (Figure 3) may potentially arise if species-rich ecosystems contain relatively fewer dominant, but more rare species than species-poor systems. We focus on the Hill's evenness index because it is largely uncorrelated with richness (Hill, 1973), to ensure that the observed negative correlation between richness and evenness is not merely a mathematical artefact of the evenness index used (Figure S1). Nevertheless, our results are robust to a range of common evenness indices (Figures S2 and S7), demonstrating the generality of this relationship and of the resulting global trends. The drivers of this trend are likely to include a range of ecological coexistence mechanisms. A possible explanation for this negative correlation could be that in cold boreal forests, there are only a few species present, and intraspecific competition is expected to be higher than the interspecific competition (Aguiar et al., 2001). However, in warm, moist environments, such as moist tropical forests, there are many species present. Higher species richness is likely to give rise to substantial interspecific competition, as well as scope for additional ecological mechanisms such as the Janzen Connell effect to influence the abundance of species (Connell, 1971; Janzen, 1970). As species have different competitive abilities and strategies, high levels of asymmetric competition allow relatively few species to become dominant, with the majority of the species being either abundant or rare (McGill et al., 2007).

The strength of the negative correlation between evenness and productivity varies between biomes (Figure 5) (Sonkoly et al., 2019; Zhang, Chen, & Reich, 2012), which may be indicative of a greater effect of niche partitioning in more speciose and even forests. In biomes with few species, increasing the richness of species tends to enhance resource partitioning and productivity (Figure 5b,c) (Isbell et al., 2009; van Ruijven & Berendse, 2005). In contrast, in biomes with many species, evenness of those species is relatively low as there are many rare species, and so increasing the evenness might increase ecosystem productivity (Kirwan et al., 2007; Zhang, John, et al., 2012; Zhang, Chen, & Reich, 2012) (Figure 5d,h). Interestingly, a different trend was observed for productivity in temperate forests, where at high evenness productivity peaks at low richness, and at low evenness productivity peaks at high richness (Figure 5c). This could possibly be explained by the overarching importance of environmental drivers for productivity in this biome, with the combined effects of richness and evenness explaining only 2% of the total explained variance (Figure 4b). Conversely, this biome contains the greatest proportion of forest plots with high human activity (Figure 4b), such that this trend may reflect different management practices across biomes.

Our results suggest that it may not only be the redundancy of the species that drives the attenuating effect of the relationship between richness and ecosystem productivity (see Cardinale et al., 2011), but also the low abundances (and corresponding minimal contribution to productivity) of the rare species that flattens the relationship between richness and productivity at high richness levels. Our hypothesis predicts that, within speciose communities, adding new species at extremely low abundances will have little impact on the overall productivity, relative to the effect on productivity of adding new species to less speciose communities. We stress that this finding does not discount the importance of rare species, which contribute significantly to productivity through positive complementarity effects in many ecological communities (Loreau & Hector, 2001; Sonkoly et al., 2019). Indeed, rare species contribute to ecosystem multifunctionality, which can have very important indirect effects on productivity (e.g. being critical for pollination) (Dee et al., 2019; Lyons et al., 2005), or sustain productivity over the long term when they become more abundant with (environmental) change (Loreau & Hector, 2001; Loreau et al., 2003; Yachi & Loreau, 1999). However, our analysis supports the idea that—in line with the mass-ratio hypothesis—rare species tend to contribute less to productivity than do dominant species. If this hypothesis were correct, then a higher relative abundance of rarer species (i.e. increasing evenness) would increase the biodiversity and productivity within the system.

The two measures of productivity used here—biomass accumulation and satellite-derived NPP—each have their unique drawbacks and challenges (Sheil & Bongers, 2020), such as relying on coarse estimates of forest age, being susceptible to mismatches between plot size and satellite resolution, and uncertainty in the calculations. For the temperate region, satellite-derived NPP and productivity calculated from multiple times measured ground-sourced data were well correlated (r = 0.65); however, we could not estimate the accuracy for the other biomes due to limited data. Despite the uncertainty in the biomass and productivity calculations, the main results were robust when considering the uncertainty within these two estimates of productivity individually (Figure S8). We chose to consider these metrics in tandem specifically to minimize the data limitations of each, and ensure that our results are qualitatively robust to the choice of productivity metric. Additionally, in our dataset, we have considerable variability in plot sizes, which can affect both species richness and evenness values, and subsequently the importance of niche and neutral processes (Gleason, 1922; Viana & Chase, 2019; Wilson et al., 1999). Yet, sensitivity analyses show that the differences in plot size do not change the main results (Figure S5), with no consistent biases across plot sizes (Figure S6). However, we do detect considerable noise across the range of plot sizes, partially due to unbalanced sample sizes in relation to plot size in between biomes.

Although our work establishes baseline empirical trends, it is important to highlight that these results rely on correlative trends using observational data from a compilation of different databases and sources. Future experimental research will be needed to test the hypotheses we have presented and to explore direct causal relationships among evenness, richness and productivity. Long-term experimental studies with various plot sizes and successional stages will be critical for disentangling the relative importance of the different processes underpinning diversity and function. By being able to manipulate species richness, evenness and functional redundancy, while also obtaining direct temporal measurements of productivity, experimental studies will be critical for identifying mechanistic drivers that are difficult to assess from broad-scale observational and statistical approaches (Paquette et al., 2018). Nevertheless, our analysis is consistent with the findings from a wide range of biodiversity–ecosystem function experiments, observational and modelling studies, highlighting that the positive effect of species richness on forest productivity declined in the most speciose communities (Cardinale et al., 2007; Hooper et al., 2005). In addition, by exploring how these trends vary across biomes, this analysis can help to provide the context for the highly variable relationships between richness and productivity across the globe.

5 CONCLUSIONS

An ever-growing body of evidence suggests that, across forests globally, plant species richness correlates with ecosystem productivity of plant communities (Liang et al., 2016; Luo et al., 2019). Our results support previous studies, showing that richness correlates positively with ecosystem productivity, particularly in communities with few species. However, as the number of species increases, the relationship between richness and productivity attenuates. We observed that, as species richness increases, the evenness of those communities tends to decline, which may potentially limit the influence of richness on ecosystem productivity in the most speciose communities. Because communities with many species tend to be dominated by relatively few species, the evenness of those communities tends to be relatively low. This negative correlation between richness and evenness may partially contribute to the attenuating effect of species richness on ecosystem productivity observed in highly diverse communities (Liang et al., 2016). In addition, this apparent trade-off between richness and evenness may explain some of the idiosyncrasies observed in previous biodiversity–productivity analyses, as the slope of the richness–productivity relationship will vary considerably across time and space due to the confounding effect of evenness. These trends have direct implications for ecosystem management practices by showing where community productivity is most dependent on richness or evenness. Ultimately, the interacting effects of richness and evenness help shape our understanding of the biodiversity–productivity relationship, identifying core relationships that link community structure to the functioning of forest ecosystems worldwide.

AUTHOR CONTRIBUTIONS

Iris Hordijk, Tom W. Crowther, Daniel S. Maynard and Simon P. Hart conceived of the study. Iris Hordijk extracted and analysed the data, and drafted the manuscript with assistance from Dan Maynard and Tom Crowther. Biomass calculation was carried out by Lidong Mo. Members of the GFBI consortium (all authors not mentioned previously) provided data for the analysis. All authors assisted with revisions and gave final approval for publication.

ACKNOWLEDGEMENTS

This research has been funded by a grant from DOB Ecology. Swiss National Science Foundation, Ambizione grant #PZ00P3_193612 to DSM. JCS considers this work a contribution to his VILLUM Investigator project ‘Biodiversity Dynamics in a Changing World’ funded by VILLUM FONDEN (grant 16549). The GFBI data from New Zealand were drawn from the Natural Forest plot data collected between January 2009 and March 2014 by the LUCAS programme for the New Zealand Ministry for the Environment and sourced from the New Zealand National Vegetation Survey Databank’. Instituto de Conservação da Natureza. FCT—UIDB/04033/2020. Open access funding provided by Eidgenossische Technische Hochschule Zurich.

    CONFLICT OF INTEREST STATEMENT

    The authors declare no competing interests.

    PEER REVIEW

    The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/1365-2745.14098.

    DATA AVAILABILITY STATEMENT

    The plot-level evenness data are stored in the Research Data Repository of the Research Collection of ETH Zürich http://doi.org/10.3929/ethz-b-000597256. The GFBI database is 3rd party data, and is publicly available upon request at https://gfbinitiative.net/data/.