Hedgerow study using artificial intelligence to help National Trust bring back its ‘blooming boundaries’

Hedgerow study using artificial intelligence to help National Trust bring back its ‘blooming boundaries’ - The National Trust

The National Trust on Thursday 27 April published research into the changes in hedgerows and density on its land in England and Wales since the start of the 20th Century, which will help inform its ambition for future conservation projects and aid plans to establish 4 million blossoming trees by 2030.

Results revealed during the conservation charity’s first ever Blossom Week – a week designed to celebrate the beauty of the wide variety of tree and hedgerow blossom – found that across England and Wales in 1900 there were just under 10,000km of boundaries on land now cared for by the National Trust.

sunsetting over a field, the foreground is filled with a white blossom covered hawthorn hedge — Hawthorn in flower at Stockbridge Down, Hampshire (© National Trust Images/John Miller)

Over the last 100 years, one third (33 per cent), nearly 3,300km, of the boundaries on land now cared for by the National Trust has disappeared, compared to 50 per cent of hedgerow loss across the rest of the country – suggesting that, while still a significant decline, National Trust places have fared better than the wider countryside.

This loss on land cared for by the conservation charity, is likely due to farmers following government policy, or because the loss occurred prior to acquisition.

A regional analysis shows that this loss appears to be fairly consistent across all regions and countries. The one key exception appears to be in the East of England where nearly half (48 per cent), over 400km, of boundaries recorded in the 1900 appear to have been lost.

Explaining how Artificial Intelligence (AI) aided the research, Professor Matthew Heard, Head of Environmental Research and Data said: “Historic mapping provides a rich point-in-time snapshot of our landscapes – but this information is effectively ‘locked up’ in the images meaning that previously the analysis of the data has been a time consuming, manual process. However, the development of AI approaches presents an opportunity to ‘detect’ features of interest with much greater speed and at greater scales than ever before. This research will help us to build an understanding of the role hedges played in past landscapes, looking at national and regional variations and the extent and survival of hedgerows over time as well as the implications of this for landscape character, biodiversity, habitat connectivity, pollinator food sources and access to blossom for people.”