Global earthworm distribution and activity windows based on soil hydromechanical constraints

The study investigates how soil hydromechanical conditions, together with earthworm biomechanical limits, govern global earthworm occurrence and seasonal activity. Earthworms are key ecosystem engineers whose burrowing modifies soil structure, hydrology, and fertility, yet their ability to penetrate and expand cavities in soil depends on moisture- and texture-controlled soil strength relative to their hydroskeletal pressure. The authors hypothesize that soil physical constraints, modulated by climate, are primary determinants of earthworm habitat suitability and seasonal activity windows worldwide. The objectives are: (i) to model soil hydromechanical conditions and derive temporal windows of potential earthworm burrowing activity, (ii) to delineate geographic regions where earthworm activity is mechanically prohibited, and (iii) to compare predicted regions with global earthworm presence data. The work places ecological patterns of earthworm distributions into a mechanistic biophysical and climatic context.

Prior work has established earthworms as ecosystem engineers that enhance soil structure, water flow, aeration, aggregation, and crop yields, and can mitigate soil compaction. Regional distribution patterns and seasonal activity have been documented, and sensitivities to temperature, soil compaction, and moisture have been reported. Recent biomechanical studies quantified earthworm penetration and cavity expansion pressures, linking soil strength with water content and texture, and suggested hydromechanical constraints as key physical bounds on bioturbation. Global datasets of soil properties (SoilGrids) and climate-driven soil moisture (reanalyses) enable spatially explicit modeling. Additional ecological and edaphic constraints include freezing temperatures, soil acidity (low pH), and high sand content. Correlative studies of earthworm distribution provide valuable benchmarks but may conflate correlated variables; mechanistic niche modeling can disentangle causal drivers.

• Biophysical model: Earthworm burrowing is represented as radial cavity expansion in elasto-viscoplastic soil. The minimal radial stress at the cavity wall for expansion is linked to soil shear strength s_L and geometry: σ_r(R_c) = s_L, yielding the limiting cavity expansion pressure P_L = s_L(1 + 2 ln(R_c/r_c)). Soil mechanical properties (shear strength s_L and shear modulus G) are parameterized as functions of soil texture (summed silt+clay fraction n) and volumetric water content θ using power-law relations based on rheological descriptions of wet soils. Earthworm maximal hydroskeletal pressure is taken as P_m ≈ 200 kPa. Activity is mechanically impeded when P_L(θ, n) ≥ P_m.
• Data sources and processing: Global soil texture from SoilGrids (0–5 cm) and monthly soil moisture from ERA5-Land (0–7 cm) were harmonized to a 0.1° × 0.1° grid (~10 km). Permafrost-affected regions (permafrost zonation index > 0.1) were excluded. For each grid cell and each month (1981–2019), P_L was computed. Ensemble annual summaries used harmonic averages of P_L. Temporal activity windows were defined as sequences of consecutive months with P_L < 200 kPa; at least two consecutive favorable months per year were required to permit habitation (to allow one reproductive cycle). Additional exclusion masks were applied where mean annual temperature (MAT) < 0 °C, soil pH < 4.5, or sand content > 80%.
• Evaluation and comparisons: Modeled hospitable regions were compared with independent earthworm occurrence datasets (GBIF presence-only records across 10 families; n = 7346), regional distribution maps for Australia and North America, and with precipitation climatology (e.g., 400 mm yr−1 MAP contour). Agreement between modeled hospitable areas and auxiliary constraint masks was quantified using Jaccard indices. Sensitivity analyses included random resampling of occurrences weighted by observation density to assess hit-rate robustness. Seasonal dynamics were evaluated by comparing monthly P_L to a 2002–2008 monthly earthworm abundance time series from the New Forest, UK, and by relating P_L seasonality to daily precipitation (MSWEP, smoothed with a 30-day rolling average) at exemplar grassland and desert sites.

• Global hospitable zones: Regions with harmonic mean P_L < 200 kPa (after excluding permafrost) delineate extensive hospitable areas, particularly outside arid interiors and high latitudes. Approximately half of the terrestrial surface below ~60°N supports potential activity.
• Agreement with observations: Modeled hospitable vs. inhospitable classification matched 86% of 0.1° grid cells with reported earthworm presence (n = 7346). False negatives (~13%) often occurred in local moist niches (e.g., river corridors) not resolved at ~10 km.
• Auxiliary constraints overlap: P_200 overlapped with 60% of regions limited by subzero MAT, 90% with the two-month activity requirement, 60% with low soil pH, and 70% with high sand content (Jaccard-based comparisons), indicating soil hydromechanics as the dominant constraint while corroborating roles for temperature, acidity, and texture.
• Regional patterns: In Australia, hospitable zones align with coastal, wetter regions; the 400 mm yr−1 MAP contour demarcates limits well. In North America, the model predicts suitable conditions from the east coast through the Midwest with sharp declines toward arid central-west regions, consistent with published distributions.
• Seasonality and activity windows: Temporal variability of P_L tracks precipitation seasonality. At the New Forest, UK (2002–2008), peaks in P_L coincide with troughs in earthworm abundance; when P_L approached 200 kPa (2003–2004; 2006–2007), abundance approached zero. A representative grassland site exhibited up to ~4.5 consecutive favorable months per year, whereas a desert site exhibited none.
• Latitudinal insights: Median P_L across latitudes highlights persistent inhospitable arid bands (notably 20–30°N). Habitat fragmentation (count of habitable land fragments within latitudinal bands) correlates with higher reported species richness, suggesting that spatially patchy hospitable conditions may foster diversity over long timescales.

The findings support the hypothesis that soil hydromechanical conditions, driven by climate-regulated moisture and soil texture, are primary mechanistic determinants of earthworm habitat suitability and seasonal activity. Incorporating additional constraints (subzero MAT, low pH, high sand, permafrost) refines predictions but does not supplant hydromechanical dominance. The strong correspondence with independent occurrence records and seasonal abundance dynamics indicates that the mechanistic model captures both spatial and temporal controls. Discrepancies arise largely from spatial resolution limits and local microhabitats (e.g., riparian zones) not resolved by global datasets, as well as potential smoothing in digital soil maps. The framework elucidates barriers to migration (e.g., arid central North America) and the role of seasonal windows in shaping life-history strategies (e.g., dormancy). Compared to correlative approaches, the mechanistic model clarifies causal pathways, avoids confounding among climate and soil variables, and is amenable to scenario testing within climate and land-use models. The results also imply coupled benefits for plant roots, as soil moisture ranges permitting earthworm burrowing overlap with favorable mechanical conditions for root growth and with patterns of gross primary production.

This study introduces a global, mechanistic, biophysical framework linking earthworm hydroskeletal limits to soil hydromechanics to map habitat suitability and seasonal activity windows at ~10 km resolution. The approach reproduces major biogeographic patterns, explains regional distributions in terms of physical constraints, and captures seasonal dynamics aligned with observed abundance. The model can inform assessments of ecological connectivity, potential migration routes and barriers, and support sustainable land-use decisions (e.g., regions amenable to bioturbation aiding no-till practices). Future work should incorporate energetic and resource constraints (e.g., soil organic carbon, plant-derived particulate organic matter, GPP) to predict abundance and community composition, refine mechanical thresholds for diverse taxa (including large tropical species), improve representation of soil variability and microhabitats, and conduct targeted experiments across moisture and compaction gradients to validate and calibrate parameters. Integration into climate models could enable projections of shifting earthworm habitats under changing precipitation and temperature regimes.

• Spatial resolution (~0.1°) and smoothing in global soil datasets (SoilGrids) limit representation of fine-scale niches (e.g., river corridors), contributing to false negatives.
• The maximal hydroskeletal pressure (≈200 kPa) is based on measurements for temperate species; larger tropical/anecic species may exert different pressures, introducing uncertainty in thresholding.
• Additional constraints (low pH, high sand) are applied via coarse thresholds and may proxy for other unmodeled factors (e.g., flooding, redox), while permafrost and low-temperature mechanical effects are simplified.
• Occurrence data exhibit strong geographic sampling bias (notably Europe), affecting sensitivity assessments despite resampling.
• The model focuses on physical and some chemical constraints and does not mechanistically include resource availability (soil organic carbon/POM), biotic interactions, or land management impacts beyond their covariation with hospitable hydromechanical conditions.
• Assumed minimum activity window (≥2 consecutive months) may vary with temperature and species-specific life histories.