Rewilding for biodiversity offsets: A case study of passive ecological restoration on lowland agricultural land for Biodiversity Net Gain in England

England is a country with ambitious targets for habitat restoration and increased woodland cover, along with new Biodiversity Net Gain (BNG) regulations requiring most new development projects to increase overall biodiversity by 10 % (measured via the statutory Defra Biodiversity Metric). Typically this involves intensively managed conservation or restoration – but could habitat rewilding based on passive restoration be used to increase biodiversity at lower cost? We analysed the potential of passive lowland agricultural rewilding in England to fulfil the requirements of BNG policy. We considered arable land cover, deer browsing pressure and broadleaved woodland cover as our variables affecting ‘rewilding potential’ and quantified the resulting potential habitat gains using the Biodiversity Metric. We found the likely outcome is mainly habitat restored to poor or moderate condition, and that the southeast part of England has the best rewilding potential, with the eastern side having more potential than the western part of the country. The maximum possible biodiversity units that could hypothetically be generated for different woodland habitat type options varied between 6.0 million and 22.3 million units, in the (albeit highly improbable, and undesirable) case that all arable lowland in England were rewilded. The estimated annual need is currently around 39,000 biodiversity units, which means rewilding a cumulative 0.27–0.90 % of agricultural land back to woodlands starting one year in advance of development could compensate for annual development impacts. A key challenge to this approach is that planners would have to embrace long timescales and uncertainty about the ecological trajectories of habitat offsets.