30 by 30 Big Biodiversity Layer Development

General Big Biodiversity Layer methodology

The BBL will initially focus on the six “important for biodiversity” values separately and produce a heat map for each of these. To do this, combinations of species and habitats (selected for their relevance to each criteria), will be weighted against the different values and mapped with points/polygon data. For example, species and habitats which can be considered rare, threatened or endangered will fit the first biodiversity values and can be weighted according to red lists and nationally scarce/nationally rare lists. Combinations of different datasets will be used, largely looking at data from NBN Atlas for species, and datasets like HabMoS for mapped habitats. 

Scotland will be split into a 1km x 1km grid; with each square summing up weightings from any species or habitats present within it, working on a presence/absence basis, and producing a “biodiversity value”. This will create a heat map, with the higher weighted grid squares indicating higher biodiversity values. We are working iteratively, layers produced sequentially and published as they become available. 

For a final summary layer, grid squares within each individual layer which indicate areas that are “important” for that biodiversity value will be identified and receive a new weighting of “1”. Adding up these 1’s, combining the six layers into one, will produce a summary layer which identifies “hotter” grid squares as those areas which are important for multiple biodiversity values. With the complete set of layers combined, we will have a layer broadly summarising the levels of biodiversity across Scotland – the Big Biodiversity Layer! 

Big Biodiversity Layer use

With the completed BBL, it will be easier to identify the most biodiverse areas in Scotland. The visualisations will be simple, intuitive and thus accessible to a wider audience. 

It will be possible to break down the completed BBL into its constituent heat maps produced for different biodiversity values, allowing for further exploration of areas. Some areas may not come out as very hot in the summary layer, but could be highly important for just one biodiversity value, and this will allow such areas to be easily identified. These areas will still have potential for OECM status if they do not come out as very hot in the overall BBL, and may be important in ensuring the overall suite delivers for biodiversity in the broadest way possible.

The BBL is going to be one tool of several that helps identify areas suitable for OECM status/help identify areas which should be explored further. Where someone wishes to take positive action confirming that high biodiversity shown in the BBL is present on the ground (in the area) remains important. This tool will allow users to search for areas of high biodiversity value, and it will also benefit the bottom-up approach, as it will allow landowners to assess their own land for areas with OECM potential. 

Timeline

As the separate heat map layers are developed they will be published for initial sense checking and feedback, and they can begin to inform searches for potential OECM areas.

Feedback will help us to ensure that the methodology is working to identify biodiversity as the tool develops, so changes can be made as/when required. 

Engagement

Engagement has been taking place across various organisations and stakeholder groups, as well as internally in NatureScot, to inform people of the ongoing work. Through this engagement we are gaining feedback and suggestions, and speaking with people about similar work and work which could benefit/be benefitted by the BBL. 

We are also searching for data which we can use to inform the tool which may not have been previously known to, or available to, NatureScot.

1. Background

Creating the Big Biodiversity Layer (BBL) tool involves developing one heat map for each of the OECM biodiversity values, initially producing a total of 6 heat maps. For each biodiversity value, relevant species and habitat data, as well as any other relevant data, will be identified. A weighting system will then be attributed to the data being input to that layer, corresponding to how much they contribute to that particular biodiversity value. The data will be mapped in a 1km x 1km grid across Scotland. Data which occurs in a grid square will be added up by the weightings attributed to them, providing a "biodiversity value" for each individual grid square. This will work by adding any occurrence of a species/habitat/other once for a grid square in which it occurs, and no detractor will be applied to squares in which they do not occur. These layers are being created iteratively, repeating this process for each biodiversity value to identify, weight and map data to find the best areas in Scotland for each value.

The BBL is going to be one tool which helps to inform decisions around 30 by 30, along with other tools and ground truthing of BBL outputs. It will provide a starting point, indicative of where we should explore for land which could potentially contribute to 30 by 30. This is caveated by data availability. If a BBL map shows an area of land to be contributing to a biodiversity value, then this is because the data is available for this land, and at some point the area has had attributes (species/habitats/other relevant data) which make it good for said biodiversity value. This indicates that areas could be good for this biodiversity value, and that an area should then be explored to realise this potential. On the flipside of this are areas which do not come out as strong for a given biodiversity value. This would indicate that there is no data to suggest that this area contributes to the biodiversity value, but doesn't mean that it has no biodiversity value. There could be good biodiversity in the area, but a lack of surveys/visitors recording data, and thus it doesn't come out as high value on the BBL. These caveats should be kept in mind when using the BBL, and it should be used as a tool to inform decisions along with other evidence, and not alone.

This methodology outlines the creation of a heat map for the Big Biodiversity Layer One, corresponding with OECM Biodiversity Value 1 “rare, threatened or endangered species and habitats and the ecosystems that support them”. This work was carried out in QGIS.

2. Datasets used

• The NBN Atlas was used for the bulk of most species’ records in the tool.
• Full lists of species and habitats used in the tool, along with corresponding weightings assigned to them (link to HTML coming soon) 

Further data sets for species/habitats records used were:

• Habitat Map of Scotland (HabMoS) dataset
• Habitat Map of Scotland (HabMoS) EUNIS base layer
• Raised Bogs EUNIS Scotland 2024 dataset
• NESBReC Integrated Habitat System survey data
• SEPA rivers and lochs electrofishing data
• National Electrofishing Programme Scotland (NEPS) fish count data

Context data layers used were:

• The 2023 MLWS boundary from Ordnance Survey (OS)
• OS Boundary-Line 

3. Identifying relevant species and habitats, and weighting them

Biodiversity Value One in the IUCN criteria for OECMs suggests use of IUCN red lists, or equivalents. We interpreted this to mean any species on a Red List classed as vulnerable, endangered, or critically endangered, or listed as nationally rare or scarce. Where possible we relied on GB red lists that used the IUCN ranking (least concern, near threatened, vulnerable, endangered, critically endangered, extinct in wild, extinct), and included all species from vulnerable to critically endangered. Extinct species were excluded so not to include data from the past which would skew the scores/heat map. Where an IUCN list was not available, or we were advised to use a different list, we used accepted alternatives, e.g. the Birds of Conservation Concern list, and compared them with the international IUCN red list. For some taxa, many species are also assessed for nationally rare/nationally scarce. For these, the red list status was prioritised over the nationally rare/scarce status. 

1. Weighting

For Biodiversity Value One, each species was given a score based on its IUCN or rare/scarce ranking. The rare/scarce status was only used if the red list ranking was lower than vulnerable – red list rankings score more highly than rare/scarce. This scoring was derived from the SSSI scoring, also used in projects such as Rivers for Conservation.

Birds of Conservation Concern (BoCC) uses a traffic light rating, with scores assigned to the categories included Red = 12, Amber = 9. The most recent review was used (BOCC5, December 2021).

2. Mammals

Mammals were chosen from the Red List of British Mammals, produced by the Mammal Society. The Scotland-specific list was used to ensure we only included mammals found in Scotland. The list was last updated in 2020. 

Data source(s): NBN Atlas, downloaded on 16/09/24

Additional sensitive species data sourced from corresponding advisers within NatureScot

3. Herptiles

Herptiles were chosen from the IUCN Red List assessment of amphibians and reptiles at Great Britain and country scale, produced by Amphibian and Reptile Conservation for Natural England. We only included species in the Scottish column, using their Scottish status. The list was published in 2021. 

Data source(s): NBN Atlas, downloaded on 16/09/2024

4. Birds

Birds have both a GB Red List and a different assessment called Birds of Conservation Concern (BoCC), which is split into red and amber categories. We combined the GB Red List with the BoCC list, as there are birds on the BoCC list that are not on the GB Red List. This means not all birds have an IUCN category. Both lists are GB wide, so we noted non-Scottish birds by hand, and birds with no Scottish records will be excluded once the data are mapped. The BoCC list was given priority, based on advice from the bird advisers at NatureScot, weighting species based on their BoCC status over their GB Red List status. Bird species which were on the GB Red List, but were not included on the BoCC list, were then added in and weighted based on their red list status. The BoCC list is maintained by the BTO, along with other NGOs and nature agencies, including RSPB, JNCC, Natural England and NatureScot. The most recent BoCC list, when compiling lists and collecting species data, was from 2021. 

Data source(s): NBN Atlas, sent by NBN as too much data to download online, received on 04/10/24

5. Fish

Fish were chosen from a 2023 study that assessed the status of freshwater fish in Britain. As with other taxa, we excluded all species not found in Scotland, leaving only five fish species. The fish assessment was funded by Natural England.  

Data source(s): NBN Atlas, downloaded on 16/09/24

Additional data was sourced from SEPA, who supplied electrofishing survey data for fish species 

National Electrofishing Programme for Scotland (NEPS) 

6. Non-vascular plants (bryophytes)

Bryophytes were chosen from the IUCN Red List of bryophytes of Britain, 2023

We were advised by our species specialist to only use record data going back 50 years. We also included Nationally Scare and Nationally Rare species, both found in the red list linked above, and endemic species, found in the SSSI guidelines for bryophytes.

Data source(s): NBN Atlas, downloaded on 20/09/24

Additional sensitive species data sourced from the corresponding adviser within NatureScot.

7. Fungi

Fungi were chosen through a combination of red lists and SSSI guidelines, as no individual list was deemed to be comprehensive enough alone. The lists consisted of:

• The Red List of Fungi for Great Britain: Boletaceae (Species Status No.14), 2013, produced by JNCC
• The Red List of Threatened British Fungi, 2006, produced by the British Mycological Society
• Four red lists produced by the Fungus Conservation Trust
• SSSI Guidelines – Chapter 14 – Non-lichenised Fungi, 2018

Like with bryophytes, we were advised to use data going back 50 years and included Nationally Scarce, Nationally Rare, and Endemic species. 

Data source(s): NBN Atlas, downloaded on 23/09/24

8. Lichens

Lichens were chosen from the Conservation Evaluation of British Lichens and Lichenicolous Fungi (Species Status No. 13), 2012. Again, advice from internal species specialist was to use data spanning back 50 years and classifications Nationally Rare and Nationally Scarce species were included. The red list linked above also provided information on endemic species which were included in the data used for this taxa.

Data source(s): NBN Atlas, downloaded on 23/09/24

9. Vascular plants

Vascular plants were chosen from a GB IUCN assessment of the status of vascular plants. As usual, species were included if they are found in Scotland and fall between VU-CR categories. This list was developed by a variety of partners, including Natural England, NatureScot, and BSBI. It was revised in March 2024.

Data source(s): NBN Atlas, downloaded on 17/09/24

Additional sensitive species data sourced from corresponding advisers within NatureScot.

10. Invertebrates

As the invertebrate group is so large, our invertebrate choices come from a variety of different lists, and do not encompass all invertebrates in Scotland. We included every invertebrate group with a status review, as published by Natural England, plus the Butterfly Conservation Trust ‘Red List’. These differed in what they included but where possible we included only species found in Scotland. Many of the invertebrate lists also included a category for nationally rare or nationally scarce, which we also included. The lists of invertebrate categories and their associated lists are below.

Data source(s): NBN Atlas, sent by NBN as too much data to download online, received on 08/10/24

Additional sensitive species data sourced from corresponding advisers within NatureScot

11. Habitats

Habitats were chosen from the European Red List of Habitats: 

As usual, habitats were included if they are found in Scotland and fall between VU-CR categories. This work was completed by the European Environment Agency for the European Union and surrounding areas, and was completed in 2021. 

Additionally Habitats Directive Annex I habitats were included. These were cross-checked using the manual of terrestrial EUNIS habitats in Scotland.

Annex I do not clearly fit in to any of the initial weighting categories, so we decided that they would all be weighted 3, except for Machair weighted as a 6 due to Scotland’s global importance for the habitat type, and woodland types which can be classified as Atlantic Rainforest (H9180, H91A0, H91C0, H91D0, H91E0), which were weighted as 6 when within the Scottish Rainforest zone (“Woodland Trust – West Coast Rainforest” on geo.view) and a 3 when outside of this zone due, once again, to Scotland’s international importance for these habitat types. These proposed weightings were suggested to relevant habitats advisors in NatureScot and no concerns were raised.

Data source(s): HabMoS 

NESBReC IHS 

Raised Bogs EUNIS Scotland 2024 

4. Processing data

• Relevant habitats data was extracted from layers and species were downloaded from NBN/extracted from datasets.
• Species point data was clipped to MLWS to remove any points which were off the coast or sitting outside of Scotland’s land border with England. This retained only point data within the terrestrial extent of Scotland.
• Rare, threatened, and endangered habitats were picked out of the relevant habitat layer, and a new layer created with only the relevant habitat types. This was done by SQL query for the polygon layer, listing the habitat names which should be kept, and then creating a new layer from this selection.
• For the raster data, Raster Calculator was used to specify the rare/threatened/endangered habitats which we wanted to retain.
• The attribute table was dissolved to combine all polygons/points by habitat/species name. This would allow each habitat/species type to be counted only once for each grid square, later in the methodology.
• For data layers with potential overlaps/duplications of data, mostly for the HabMoS dataset and the HabMoS base layer, “symmetrical differences” was used to cut out overlapping data. The preferred data layer would be given priority in this process, and the other layer would be clipped to only retain data which falls outside of the extent of the priority layer.
• Attribute tables were matched up with the red lists, adding a “Weight” column to attribute a weight to each habitat/species

  1. Spatial weightings within layers

• Grid squares could have multiple polygons/points within and overlapping one another. We added these up to get overall weightings for habitats/taxa across the layer before converting into a raster.
• To do this, a grid was first created across Scotland, allowing everything to match to the same extent later (named “Scotland_Grid”).
• Using “Join Attributes by Location (Summary)”, polygons/points sitting within each grid square of the grid polygon layer had their “weight” summed up, specified in the parameters. It is important to make sure you use the “(Summary)” version, or it will not add the attributes together (the non-summary version just chooses one weighting from one of the habitats/species present). 

Parameters for Join attributes by location (summary):

• Scotland_Grid for the first input layer (it works the wrong way round if you put it as the second layer, so it is important to put the layers in correctly)
• Where the features: Intersect
• Dissolved points/polygons as the second layer
• Fields to summarise: weight
• Summaries to calculate: sum
• Save to file and run
• This creates a new polygon layer, which looks the same as the grid layer, but for squares which had points/polygons within they now have summed up weightings. It is important to make sure that all habitats/species are dissolved to group all of the same habitat/species types, or there will be double counting within grid cells. We want a presence/absence count so this allows for that.
• This polygon grid layer could now be turned into a raster using “Rasterise (Vector to Raster). 

Parameters for rasterise:

• Input layer the output from the previous step
• Field to use for burn-in value: weight_sum
• Output raster size units: Georeferenced units
• Width: 1000, Height: 1000
• Output extent: Use Scotland_Grid for this (EPSG: 27700)
• Additional command-line parameters: -at
• Save to file and run

5. Combine

• Raster’s were then combined and summed, to give the weightings for all habitat layers combined. All rasters were created to the extent of the Scotland grid, to ensure that all layers will match up when finally combined.
• Rasters were reweighted to allow each species taxa grouping to contribute equally to the final species raster. The birds taxa layer was the most overwhelming, so all other taxa rasters were upscaled to match the upper limit of weighting for this. The habitats layer was then upscaled to match the combined species raster upper weighting limit, so both species and habitats contributed equally to the final combined BBL1 layer. 

Parameters for rescale raster:

• Input raster: your chosen species or habitats raster layer
• Band number: Band 1 (Gray) – this will be the default
• New minimum value: Put the same value as the current minimum for that layer
• New maximum value: Put the value which you wish to rescale the raster to
• New NODATA value: Not required
• Save to file and run

6. Heat map

• To make the raster appear as a heat map, the render type was changed to palleted/unique, using a suitable colour scale to display the higher numbers, the “hot” spots, in the dark end of the scale, and the lower values in the lighter end of the scale.

5. Mapping BBL2 methodology

Using the dissolved layers from the habitat stock take, for each habitat type, weightings were attributed to the layers, corresponding with those listed above. Each layer was then divided into a 1km x 1km grid across Scotland using “intersection”. This split the habitat types into smaller polygons within each grid square. The weight attribute was kept from the habitat layer, and for each of the new polygons created in this way, an area attribute was calculated in hectares. 

A new attribute was now created using the area and weight attributes. Area and weight were multiplied to give a value to each habitat polygon within each grid square. This accounted for the assigned weighting and the area covered by habitat types within grid squares. These new weightings were then attributed to the grid squares, by joining the attributes from the habitat polygons to the Scotland grid layer using “Join attributes by location (summary)”. For this step, it was specified to join attributes where the grid squares “contain” the polygons from the habitat layer, and to “sum” the field with the area*weight multiplier. 

The layer produced by joining attributes to the Scotland grid gave a polygon grid where each grid square had a value corresponding to the habitat type it was produced for. This grid could finally be turned into a raster using “rasterise (vector to raster)” to produce a heat map for the habitat type. Once these steps were performed for all habitat types, and a heat map was created for each type, they could be combined using “cell statistics” with “sum” to add all values together and produce a heat map summarising all habitat types. This produced the final raster for BBL2.