The role of connectivity in understanding grassland-shrubland regime shifts: Impacts of aridity, grazing and wind dynamics

Regime shifts from grassland to shrubland are a key feature of dryland degradation, yet the role of connectivity in understanding these transitions remains poorly quantified. Using a numerical model, we simulate vegetation change from 1895 to 2022 under two climate endmembers bracketing the regional aridity gradient: a dry extreme (Southwest Arizona) and a wet extreme (Northern Plateau New Mexico). Under near-natural grazing (1 g /m²/year), grass biomass remained stable with negligible shrub presence. Moderate grazing (30%) initiated shrub emergence by 1930 in dry climates, while wetter systems largely retained grass (>20 g /m²). High grazing (60%) drove abrupt, near-total grass loss, with shrub biomass exceeding 40 g /m² in dry and ∼60 g /m² in wet climates. Connectivity metrics strongly predicted shrub expansion, with structural and functional connectivity of water and nitrogen correlating up to 0.96 with shrub biomass, providing early-warning signals 4 to 16 years before vegetation shifts. By moving beyond static biomass measures and incorporating connectivity dynamics, this study advances predictive understanding of grass-shrub regime shifts and offers a scalable approach for dryland management under intensifying global change.