Regenerative agriculture improves productivity and profitability while reducing greenhouse gas emissions on Australian sheep farms.

Regenerative agriculture can be deconstructed into several constituent practices, including adaptive multi-paddock (AMP) grazing, improved biodiversity, silvopasture, and minimizing cultivation and synthetic fertilizer inputs. Here, using farms across a rainfall gradient, we examined how three constituents-pasture species composition, antecedent soil organic carbon (SOC) and AMP grazing-influenced SOC accrual, greenhouse gas (GHG) emissions, production and profitability. Whole-farm stocking rate and rainfall exerted a stronger influence on pasture production, SOC, GHG emissions and profit than pasture diversity or grazing management. Production was more strongly associated with individual pasture species, rather than species diversity per se. Notwithstanding carbon removals through increased SOC stocks, enteric methane remained the dominant source of farm GHG emissions. Low-intensity grazing with short rest periods was generally more profitable, whereas AMP grazing promoted greater pasture growth, SOC accrual and emissions abatement; AMP also performed more favourably when emissions, profit and productivity were considered together. Persistent trade-offs between economic and environmental outcomes indicate that grazing regimes delivering the greatest SOC accrual and GHG mitigation are not necessarily the most profitable, reinforcing the need to rationalize objectives when designing resilient, practical and low-emissions farming systems.