Development of a combined marine and terrestrial biodiversity indicator: research

6. Review of suitable biodiversity data for Scotland

105. A list of data sources that may be available to be used to create a biodiversity indicator for Scotland can be found in the accompanying spreadsheet "Biodiversity_indicator_data_review.xlsx". This spreadsheet summarises details including dataset name, taxonomic group covered, number of species covered, start year, current end year, type of metric, whether it is an ongoing project and if data are freely available. The datasets have been split into two types; those that are species-based and measure trends in either abundance or occupancy, and other relevant non-species metrics.

106. Datasets documenting trends in either abundance or occupancy within Scotland's terrestrial, freshwater and marine environment are currently available for approximately 2400 species; however, most of these trends (>80%) measure change in occupancy rather than abundance. These datasets can be split into the following main groups:

• Abundance-based data generated by well-established national monitoring schemes. These could be considered as the highest quality data, as they are already published, usually annually, at the spatial resolution required, and would therefore require minimum additional effort to incorporate into any biodiversity indicator. Examples include the BTO/JNCC/RSPB Breeding Bird Survey, BC/CEH/UK/BTO/JNCC Butterfly Monitoring Scheme and the Bat Conservation Trust's National Bat Monitoring Programme. At a Scotland resolution, these cover approximately 380 species of bird, mammal, butterflies and moths.
• Occupancy-based metrics using more ad hoc presence-only biological records. Following several years of development and refinement of methods to produce such trends at the UK scale (e.g. Outhwaite et al. 2018), draft trends for the UK's four constituent countries have been produced by the UK Centre for Ecology and Hydrology (UKCEH; previously the CEH) (Outhwaite et al. 2019). Whilst the production of such trends at a smaller (national-level) spatial scale means a reduction in sample size and a consequent reduction in the number of species for which sufficiently robust trends can be derived, this work presents a step change in the ability to report upon Scotland's biodiversity. UK occupancy trends are available for a wider range of taxonomic groups, potentially encompassing over 4000 species of invertebrates, 1200 vascular plants and 1600 bryophytes and lichens. For Scotland 1,761 species have sufficient data to be included in an indicator (although 1,578 have been used in the draft indicator presented by this report, as abundance-based trends have been selected for moth species in preference to occupancy trends for species for which both existed).
• Datasets that are regularly published but for which further manipulation would be required for use in a Scotland biodiversity indicator, for example, due to the data not being available at the correct spatial resolution. This is the case with data from the Rare Breeding Bird Panel (approximately 50 species), which is currently only published at a UK-level. Similarly, very little of the marine data is currently published at other spatial resolutions; for example, OSPAR region or individual seas.
• The application of Bayesian occupancy modelling to vascular plant datasets (provided by the Botanical Society of the British Isles (BSBI) and held by the BRC) has not generated suitable annual trends: consultation with BSBI identified issues with temporal bias in recording effort that modelling has been unable to account for satisfactorily. For the State of Nature 2019 report (Hayhow et al. 2019) an alternative approach using the Frescalo method (Hill 2011) was used to create trends at the UK and country scales. However, this relied on pooling records into just five time periods between 1930 and 2018 of which only two lie within the period from 1994 onwards proposed for this combined indicator. This in effect provides a straight line trend between two estimates for all vascular plant species, without annual updates, which we feel is not appropriate for the main purposes of the indicator. The National Plant Monitoring Scheme (NPMS), initiated in 2015 by the Botanical Society of the British Isles, CEH, JNCC and Plantlife, will produce annual trends in abundance for a suite of widespread vascular plant species at a UK and, where possible, country scale. These trends could be integrated into future iterations of the combined indicator.
• Datasets based on intermittent, non-annual, often species-specific, surveys (approximately 20 species). Generally, these cover bird and mammal species not included in national monitoring schemes, often due to their small population or range size; for example, the approximately 17 bird species covered by the Statutory Conservation Agency and RSPB Annual Breeding Bird Scheme (SCARABBS), otter (SNH 2015) and seal surveys (SMRU 2015). Several weaknesses apply to these datasets, including that some surveys haven't been repeated for many years (e.g. the last national merlin survey was in 2008 (Ewing et al. 2011)), and technical issues must be surmounted when combining these intermittent data into an annual abundance-based indicator.

107. The situation within the marine environment is different. Indicators are very widely used in assessment of the state of marine ecosystems (e.g. under MSFD), but these indicators tend to be different from terrestrial-focused assessments, in that most are not based on species data, and when they are they tend to be based on data on species composites rather than individual species. While some continuous species data are available that could be incorporated into an overarching abundance/occupancy-based Scottish indicator (for example, from the International Bottom Trawl Survey (IBTS)) the amount available is limited. Further manipulation and analysis of datasets, such as Marine Biodiversity and Climate Change Project (MarClim), Seasearch and Ocean Biogeographic Information System (OBIS), would be required to derive species-level trends in occurrence and/or abundance from these additional sources.

108. Single-species indices from many of the potential datasets, particularly the national monitoring schemes, are freely available either through online downloads or data requests, although prior engagement with data stakeholders will assist with the provision of data; the engagement process (see Section 7) conducted as part of this project has secured the relevant permissions for the inclusion of data within the draft indicator and although further permission will be required for use in a published NPF indicator, we firmly anticipate this will be the case. A data collation process would need to be established before the indicator is updated, and this will be more complex for data sources away the established data channels.

109. Another factor to consider is how many years it takes from data collection to publication (the data lag), which is inherent within products derived from biological data, particularly as a large proportion of field data collection is by volunteers working within citizen science projects. As many of our national monitoring schemes are already used for national indicators the data lag is often minimal (<2 years). Issues arise when trying to incorporate additional data sources away from these schemes and how and when they are integrated into any indicator. For example, data from the Rare Breeding Birds Panel and the Seabird Monitoring Programme tend to be slightly older, with currently (September 2019) published trends terminating in 2016 and 2015 respectively, while the trends derived from occupancy modelling end in 2014, 2015 or 2016, but in a few cases in 2011, depending on taxonomic group. As noted above, incorporating species-specific surveys can be problematic as some are relatively out of date and repeated irregularly.

110. Twenty-eight relevant non-species metrics were identified. These come from a variety of sources including Scottish Natural Heritage, JNCC, Scottish Environment Protection Agency, CEH, Forestry Commission, Defra and Scottish Water, and cover many of the topics discussed in section 4.5, such as protected sites/habitat extent and condition, pollution, and invasive species. Much of the data are freely available; however, six of the metrics have been discontinued. Like the species data, they have various data lags associated with them, with end dates varying from 2013 to 2018.

6.1 Summary

111. Our review of biodiversity data for Scotland found a considerable volume of species' data suitable for inclusion in a combined terrestrial and marine indicator. Robust abundance trends are available for 380 species of birds, mammals, butterflies and moths, and trends in occupancy (distribution) are available for an additional 1,578 species across a much broader taxonomic range including bryophytes, lichens and invertebrates. Issues around the availability of these data are discussed.

112. There are many gaps in data availability; for example, at present vascular plants are not included. Furthermore, we have far less data for marine species, and much of that which has been collected has not yet been analysed and the resources required to do so were beyond the capacity of this current project.