Search

The Welsh Government (WG) uses data on hedges and field boundaries for a variety of purposes including scheme delivery, environmental monitoring and regulatory compliance. Hedge data are currently acquired mainly through a combination of aerial photography and field visits. Considerable cost savings may be possible, if optimising the use of satellite data enables the number of field visits to be reduced. This project explored the potential for high resolution satellite data to provide accurate spatial data on hedge location and length. A number of methods of hedge mapping were tested, including manually digitising hedges and more automated methods using Skysat Imagery Products data from Planet Labs Inc.1 and LiDAR. The key findings of the project were: A. Aerial photography was a better source of data for manually digitising hedges than the Planet Skysat data available for this project. This is because the aerial photography has higher spatial resolution (25cm compared to ~50cm) and the Planet Skysat data used in this project was collected in winter and was badly affected by shadows. Note, for the purposes of this project, only Planet Skysat data collected in the winter were available, however, it is highly likely that multi-temporal Planet Skysat data would improve results further. B. For two of the three sites, the spatial accuracy of the Planet Skysat data was too low to map hedges, without spending additional effort manually geo-correcting images. C. Automated methods using LiDAR show promise, but there are issues with producing a final ‘clean’ vector data set that require additional work. LiDAR-based methods could be deployed in a number of ways depending on requirements. D. Automated methods using LiDAR and aerial photography could be deployed in a number of ways depending on whether the aim is to measure some attributes of hedge condition, or hedge location and length. E. Methods using automated detection of new hedges using LiDAR data, followed up by manual

Undertaking a field survey, such as the Countryside Survey (Carey et al., 2008) or the Welsh Glastir Monitoring and Evaluation Programme (GMEP) (Emmett and GMEP team, 2014), is a relatively expensive and time consuming way of collecting habitat data in comparison with remotely sensed techniques. In order to assess the information gained from a field survey in relation to the information that can be gained from aerial photography, a short project has been undertaken with the following objectives: • To measure the time taken to survey a 1km square using aerial photography (for a range of different and UK representative landscape types) • To measure the accuracy and level of detail of data derived using this method relative to data collected using field survey • To provide an idea of time costs associated with each of the methods • To determine the extent to which Priority Habitats can be assessed using remotely sensed methods in addition to Broad Habitats.

Context: Historical maps of land use/land cover (LULC) enable detection of landscape changes, and help to assess drivers and potential future trajectories. However, historical maps are often limited in their spatial and temporal coverage. There is a need to develop and test methods to improve re-construction of historical landscape change. Objectives: To implement a modelling method to accurately identify key land use changes over a rural landscape at multiple time points. Methods: We used existing LULC maps at two time points for 1930 and 2015, along with a habitat time-series dataset, to construct two new, modelled LULC maps for Dorset in 1950 and 1980 to produce a four-step time-series. We used the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) Scenario Generator tool to model new LULC maps. Results: The modelled 1950 and 1980 LULC maps were cross-validated against habitat survey data and demonstrated a high level of accuracy (87% and 84%, respectively) and low levels of model uncertainty. The LULC time-series revealed the timing of LULC changes in detail, with the greatest losses in neutral and calcareous grassland having occurred by 1950, the period when arable land expanded the most, whilst the expansion in agriculturally-improved grassland was greatest over the period 1950–1980. Conclusions: We show that the modelling approach is a viable methodology for re-constructing historical landscapes. The time-series output can be useful for assessing patterns and changes in the landscape, such as fragmentation and ecosystem service delivery, which is important for informing future land management and conservation strategies.

© 2019, The Author(s). Context: Studies evaluating biodiversity loss and altered ecosystem services have tended to examine changes over the last few decades, despite the fact that land use change and its negative impacts have been occurring over a much longer period. Examining past land use change, particularly over the long-term and multiple time periods, is essential for understanding how rates and drivers of change have varied historically. Objectives: To quantify and assess patterns of change in semi-natural habitats across a rural landscape at five time points between 1930 and 2015. Methods: We determined the habitat cover at over 3700 sites across the county of Dorset, southern England in 1930, 1950, 1980, 1990 and 2015, using historical vegetation surveys, re-surveys, historical maps and other contemporary spatial data. Results: Considerable declines in semi-natural habitats occurred across the Dorset landscape between 1930 and 2015. This trend was non-linear for the majority of semi-natural habitats, with the greatest losses occurring between 1950 and 1980. This period coincides with the largest gains to arable and improved grassland, reflecting agricultural expansion after the Second World War. Although the loss of semi-natural habitats declined after this period, largely because there were very few sites left to convert, there were still a number of habitats lost within the last 25 years. Conclusions: The findings illustrate a long history of habitat loss in the UK, and are important for planning landscape management and ameliorative actions, such as restoration. Our analysis also highlights the role of statutory protection in retaining semi-natural habitats, suggesting the need for continued protection of important habitats.