NATURE CLIMATE CHANGE | VOL 3 | MARCH 2013 | www.nature.com/natureclimatechange 187 S oils can act as a source or sink for the three principal greenhouse gases (GHGs). Approximately 20% of global CO 2 emissions originate from soils 1 , as well as roughly one third of global CH 4 emissions and two thirds of N 2 O emissions 2 . he production of GHGs in soils is ultimately the result of a variety of biotic processes: CO 2 is emitted through soil respiration (root, microbial and faunal respiration) 1 ; CH 4 through methanogenesis 3 ; and N 2 O through a combination of microbial processes, mostly nitriication, denitrii- cation and nitriier-denitriication 4,5 . All of these GHG-producing processes are controlled by substrate availability (for example, min- eral nitrogen (N) and labile carbon (C) for N 2 O), as well as soil physico-chemical factors (such as soil moisture, temperature and difusivity) that ultimately determine microbial activity. Although earthworms hardly produce any GHGs themselves, they may sig- niicantly afect substrate availability and soil physico-chemical characteristics and thereby indirectly afect emissions. Earthworms are soil ecosystem engineers, as they modify soil structure and interact with microbes through their feeding, bur- rowing and casting activities 6,7 . hey are typically subdivided in three functional groups, based on their feeding and burrow- ing behaviour: (1) anecic species, which feed on fresh litter from the soil surface and pull it deep into the soil in permanent bur- rows; (2) epigeic species, which are surface-dwellers that also feed on fresh surface litter and do not make permanent burrows; and (3) endogeic species, which live and feed on mineral soil and asso- ciated organic matter below the surface 8 . In the earthworm gut, conditions are ideal for denitrifying bacteria as it is essentially an anaerobic microsite where the local enrichment of mineral N, available C and favourable moisture lev- els all stimulate denitriier activity 9 . hese optimal N 2 O-producing conditions are extended into the soil volume that is directly inlu- enced by earthworm activity: casts, mucus and burrow walls. As a result, N 2 O emissions from casts and burrow walls can be up to three times greater than from bulk soil 10 . Earthworms also afect the production and emission of N 2 O and CO 2 indirectly by incorporating plant residues and mixing the soil, by stimulat- ing soil aggregation and by changing soil moisture dynamics and gas difusivity 11–14 . Greenhouse-gas emissions from soils increased by earthworms Ingrid M. Lubbers 1 *, Kees Jan van Groenigen 2 , Steven J. Fonte 3 , Johan Six 4 , Lijbert Brussaard 1 and Jan Willem van Groenigen 1 Earthworms play an essential part in determining the greenhouse-gas balance of soils worldwide, and their inluence is expected to grow over the next decades. They are thought to stimulate carbon sequestration in soil aggregates, but also to increase emis- sions of the main greenhouse gases carbon dioxide and nitrous oxide. Hence, it remains highly controversial whether earth- worms predominantly afect soils to act as a net source or sink of greenhouse gases. Here, we provide a quantitative review of the overall efect of earthworms on the soil greenhouse-gas balance. Our results suggest that although earthworms are largely beneicial to soil fertility, they increase net soil greenhouse-gas emissions. By stimulating the decomposition of plant material, earthworms can increase the availability of plant nutrients 15 . Beside this well- known positive efect on soil fertility, it is also oten suggested that earthworms induce long-term stabilization of soil C by protecting C in microaggregates formed in large macroaggregates 16,17 . his has led to repeated suggestions that earthworms promote soil C stor- age 18 and hence reduce net CO 2 emissions. his possible contribu- tion of earthworms to long-term C stabilization seems to be in sharp contrast with the short-term earthworm-induced emissions of CO 2 and N 2 O (Box 1). Over the next few decades, earthworm presence is likely to increase in ecosystems worldwide. For example, large parts of North American forest soils are now being invaded by earthworms for the irst time since the last glaciation 19 . Earthworm abundance and importance in agroecosystems will also steadily increase over the coming decades. Higher inputs of organic fertilizers will be applied to agricultural soils to feed the world’s growing population 20 , providing food for earthworms. Earthworm activity is likely to be stimulated by the increasing worldwide shit from conventional land-management practices to zero- or conservation tillage. Both types of tillage reduce soil disturbance, which can be beneicial to earthworms 21 . For example, adoption of no tillage has resulted in two- to nine-fold increases in earthworm density, as well as in shits in earthworm species composition (for example, a relative increase in the number of anecic earthworms) 22 . More land will be culti- vated, resulting in possible losses in earthworm diversity: increases are likely to occur in earthworm biomass under managed pasture; efects are unclear under arable land 23 . However, no consensus has been reached on how this expected increase in earthworm abundance will impact the GHG balance of soils. herefore, we used meta-analysis to synthesize the efect of earthworm presence on soil organic carbon (SOC) content and luxes of CO 2 and N 2 O from soils. We did not consider impacts of earthworms on CH 4 emissions as the anaerobic conditions that are conducive to signiicant methane emissions are generally not asso- ciated with earthworm habitats; as a consequence, very few (see Bradley et al. 24 , and references therein) suitable published studies were found. In total, we collated 237 observations from 57 published 1 Department of Soil Quality, Wageningen University, PO Box 47, 6700AA Wageningen, the Netherlands. 2 Department of Biological Sciences, Northern Arizona University, Flagstaf, Arizona 86011, USA. 3 Tropical Soil Biology and Fertility Program (Latin American and Caribbean Region) International Center for Tropical Agriculture (CIAT), Unidad Suelos, Apartado Aereo 6713, Cali, Colombia. 4 Department of Plant Sciences, University of California, Davis, California 95616, USA. *e-mail: ingrid.lubbers@wur.nl REVIEW ARTICLE PUBLISHED ONLINE: 3 FEBRUARY 2013 | DOI: 10.1038/NCLIMATE1692 © 2013 Macmillan Publishers Limited. All rights reserved