Earthworms contribute significantly to global food production

The study addresses how much earthworms, as key soil ecosystem engineers, contribute to global agricultural production. While agricultural intensification has boosted yields, it has also led to biodiversity loss, pollution, and climate impacts, underscoring the need for agroecological approaches. Soil health and biodiversity, particularly earthworms, influence plant growth through effects on soil structure, water capture, organic matter cycling, nutrient availability, plant hormone production, and crop immune responses. Despite their recognized importance, the global-scale contribution of earthworms to agricultural productivity had not been quantified. This paper aims to estimate earthworm impacts on the productivity of major cereal and legume crops by combining earthworm-yield responses from a meta-analysis with global maps of earthworm abundance, soil properties, management inputs, and crop yields and areas.

Prior work has documented earthworms’ roles in improving soil structure, water dynamics, nutrient cycling, and plant health, and has linked soil biodiversity to ecosystem services supporting agriculture. A key meta-analysis (van Groenigen et al., 2014) found that earthworms generally increase plant production, with effects varying by crop type, soil pH and texture, nitrogen inputs, and earthworm abundance. Additional studies report earthworm-mediated regulation of plant productivity in specific contexts and interactions with other soil biota. However, before this study, these effects had not been scaled to quantify contributions to global agricultural production.

The analysis integrates global spatial datasets with effect sizes from a meta-analysis to estimate the earthworm contribution to crop yields.

• Data layers: Global crop yields and harvested areas at 5 arcminute resolution from Monfreda et al. (2008) for selected crops: cereals (wheat, rice, maize, barley) and leguminous annuals (grain and forage legumes). Non-cereal grasses were excluded due to lack of corresponding global crop layers. Soil pH and texture classes were from SoilGrids (top 0–100 cm weighted average). Crop-specific nitrogen application rates were adapted from Mueller et al. (2012) to classify areas as low (≤30 kg N ha−1 yr−1) or high (>30 kg N ha−1 yr−1). Global earthworm distribution and abundance were from Phillips et al. (2019), upscaled to 5 arcminutes (Matlab R2020a). Analyses were conducted in R (v4.1.3).
• Effect size framework: Starting from a mean 23.3% increase in aboveground plant biomass attributable to earthworms across 58 studies (462 datapoints) in the meta-analysis, the authors identified key moderators with distinct category effects and available global layers: crop type, soil pH, soil texture, N application rate, and earthworm abundance.
• Weighted coefficients: For each factor, category-specific coefficients were derived from meta-analytic effect sizes weighted by sample sizes (see Table 1 examples: cereals 31.41% effect; legumes 9.19%; soil pH >7.0: 13.63% vs ≤7.0: ~33%; sandy soils 9.64% vs clayey 46.79%; low N 19.00% vs high N 8.51%). Earthworm abundance was modeled with a continuous nonlinear power function providing best fit to meta-analysis data: coefficient = 0.1032 × abundance (individuals m−2)^0.400.
• Scaling equation: For each grid cell i and crop k, the earthworm effect on yield E_ik was calculated as E_ik = c_i p_i t_i n_i a_i^0.400 × 23.3%, where c, p, t, n are the category coefficients for crop type, pH, texture, and N rate, and a is earthworm abundance. This approach yields near-zero effects at very low abundances, considered conservative.
• Aggregation: Maps of E were generated per crop within harvested areas. Absolute yield effects were computed by applying E to crop yield layers and multiplying by harvested area to obtain production attributable to earthworms, summed across grid cells. Regional summaries used UN SDG regional groupings. Cells lacking earthworm data were omitted.

• Global contribution: Across major cereals and legumes, earthworms account for ~5.4% of total production.
• Cereals: Estimated contribution is 6.45% of global production (~128 million metric tons of grain; wheat, rice, maize, barley).
• Legumes: Estimated contribution is 2.3% of global production (~16 million metric tons; includes soybean, dry beans, peas, chickpeas/garbanzos, lentils, alfalfa, clover and other pulses). The lower effect relative to cereals is consistent with legumes’ N fixation reducing benefit from earthworm-facilitated N mineralization.
• Combined grains and legumes: Over 140 million metric tons of annual production attributable to earthworm activity (consistent with 128 + 16 ≈ 144 MMT).
• Regional relative effects (approximate % of total production attributable to earthworms): • Sub-Saharan Africa: cereals ~10–11%; legumes ~3–3.2%. • Latin America & Caribbean: cereals ~8%; legumes ~3–3.1%. • Eastern/South-Eastern Asia: cereals ~7.4–8%; legumes ~2%. • Europe: cereals ~7.4–8%; legumes ~3%. • Northern America: cereals ~5%; legumes ~1%. • Central Asia: cereals ~5%; legumes ~1%. • Northern Africa & Middle East: cereals ~4%; legumes ~1%. • Australia & Oceania: cereals ~4%; legumes ~1%.
• Regional absolute cereal production increases (millions of metric tons per year attributable to earthworms): Eastern/South-Eastern Asia ~40; Europe ~40; Northern America ~20; Latin America & Caribbean ~8; Central Asia ~10; Sub-Saharan Africa ~2; Northern Africa & Middle East ~2; Australia & Oceania ~1.
• Drivers of regional variation: Higher relative effects in the global South are linked to lower soil pH, higher clay content, and lower fertilizer inputs—conditions under which earthworm benefits to plant growth are stronger. Higher-than-average effects in Europe relate to higher predicted earthworm abundance; in South-Eastern Asia to lower soil pH.

The findings quantitatively demonstrate that earthworms are meaningful contributors to global food production, especially for cereals and in regions with soil and management conditions that amplify earthworm benefits. By linking meta-analytic effect sizes to spatially explicit data on soils, management, and earthworm abundance, the study shows both relative and absolute contributions, highlighting that even modest absolute gains in regions like Sub-Saharan Africa can be significant for food security. The larger response in cereals relative to legumes aligns with biological mechanisms: legumes’ ability to fix nitrogen reduces reliance on mineralized nitrogen made available by earthworm activity. Regional patterns reflect interacting drivers—earthworm abundance, soil pH and texture, and fertilizer practices—supporting the view that agroecological management that fosters earthworm populations could enhance yields where potential is greatest. Overall, the results support the hypothesis that soil biota, notably earthworms, are important drivers of agricultural productivity and relevant to sustainability goals.

This study provides the first global quantification of a beneficial soil organism’s contribution to agricultural production, showing substantial impacts of earthworms on cereal and legume yields. The results suggest that fostering soil biological communities can contribute to sustainable intensification and food security. The authors caution against introducing earthworms to regions where they are not native due to ecological risks and instead recommend investing in research and agroecological practices that enhance entire soil biotic communities (e.g., no-till, organic inputs/returns). Future research should assess contributions of other soil organisms, refine global biodiversity and abundance maps (especially in under-sampled regions), and evaluate interactions among soil biota, soils, climate, and management to better inform policy and practice.

• Experimental basis: Many underlying studies used mesocosms with earthworm additions or exclusions and sometimes employed high, potentially unrealistic earthworm densities, possibly overestimating effects. Experiments were generally short-term, not capturing longer-term, indirect benefits (e.g., erosion control, water dynamics).
• Modeling assumptions: The analysis assumed additive effects across factors (crop type, pH, texture, N rate, abundance), without accounting for interactions among drivers.
• Data constraints and biases: The global earthworm abundance map is heavily biased toward the global North (Europe and eastern North America). Abundance appears low in many parts of the global South, particularly the tropics, which could either indicate potential for management gains or reflect underestimation due to sparse sampling. Combining global maps of differing resolutions introduces additional uncertainty. Cells lacking earthworm data were omitted, potentially biasing estimates.