Radley Large Wood, nestled in the picturesque Oxfordshire countryside near the village of Radley, is a serene and enchanting woodland known for its rich biodiversity and tranquil atmosphere. A haven for dog walkers, this woodland is dotted with ancient oak, beech, and ash trees, creating a diverse canopy that shelters a variety of flora and fauna. The area has a decent amount of well-maintained footpaths that weave through the woods, although some difficulty getting there in the first place if you don't live locally. Radley Wood is a cherished site for the local community, and was recently bought in 2022 by St Hilda's College.
In recent months, Radley Large Wood has been in the local news. Ash Dieback, is a fungal pathogen posing a serious threat to woodlands and ecosystems across the UK, including places like Radley Wood in Oxfordshire. It is a real hot button issue, with many campaigners saying most foresters are going too far by clear felling Ash trees. As the disease leads to the gradual dieback of ash trees, it weakens their structural integrity and creates hazards as branches or entire trees become prone to falling. This poses significant safety risks along public pathways, roads, and recreational areas, necessitating costly removal and maintenance efforts.
The decline of ash trees is also causing a ripple effect on biodiversity, as many species of birds, insects, and fungi rely on these trees for food and shelter. The loss of mature ash trees is altering woodland habitats, reducing canopy cover, and disrupting the ecological balance, leaving gaps that are often difficult to restore naturally. Without intervention, the decline could lead to long-term habitat fragmentation, affecting the health of surrounding ecosystems. So we thought this would be a great location for a case study.
We deployed 5 sensors for two weeks, two Wildlife Acoustics Song Meter Micro 2's, and three secret pre-production prototypes, soon to be available on the market.
A - Central location, protected from the wind, but close to logging activity.
B - On the edge of the Ash Clear felling site. In a patch of willow, and next to a path and river.
C - On a main entrance/exit to the site, next to an open point of the river. Open fields on two sides, woodland on the other two sides.
D - Another main entrance to the site. Became quite busy with machinery during the recordings.
E - A small bridge over the stream, on a busy path, and next to a large field.
Initial results showed that Site D was hit hard with wind from 2 major storms, and was particularly disturbed by logging movements over the recording period. E and B also had some issues with wind, being in more open spaces, but still created valuable information.
With the 5 detectors, over two weeks, we had over 100GB of data to process. Using off the shelf AI technology we identified over 150,000 individual calls over 34 confirmed different species. The top three being Blue Tits, Great Tits and Long-tailed Tits. This was confirmed by three in person recordings. Other interesting birds included Green Woodpeckers, Nuthatches, and Short-toed Treecreeper. With the occasional Song Thrush, and Marsh Tit making an appearance too.
Another key part of the assessment is not only looking at amount of recordings for a single species, but also what time of day were those birds most active. As you can see in the graph below we took the three most active birds and compared time of day for a single recorder. Blue tits are most active between 8 and 9 (coincidently the busiest time of day on average for this area). The Long Tailed Tit had a gentle rise and fall throughout the day, with a peak at 12am. Then you have the great Tit which is most active in the early afternoon, with a dip at midday. Ecologists will often look at data like this when looking at particular species assessments.
Having multiple sensors, in a range of habitats, means we can make assessments on how different birds react in those environments. For instance sensors C, D and E are on the busy path through the woodland. Equally, there is a small brook running down this side of the wood. Robins were 30% more prevalent on these sensors, compared to the other two. Equally, sensors E, and D were much more prone to issues with wind noise, especially site D. These two sites had a 25% dip in recordings compared to the other sensors. One interesting thing we noticed was that overall, the busiest time of day for bird calls appeared to be between 8 and 9am.
If you are interested in an assessment of your site, similar to this, get in contact with us at smithrobotics.co.uk/contact. This does not replace the work of ecologists, but can aid in working towards your company ESG goals. We can provide hardware, and software support. As well as aiding in deployment, or making connections with other partners if required.
Comments