Marine mammal abundance and distribution data form an important part of assessments to estimate the potential effects of proposed offshore wind developments. Therefore, there is a need to ensure that abundance and distribution data are collected and analysed to ensure robust estimates to inform the planning, consenting and licensing processes. There are several sources of information that can contribute to estimating the abundance and distribution of marine mammals. Digital aerial surveys (DAS) and static passive acoustic monitoring (PAM) are two data collection modes, which have been developed relatively recently, compared to standard visual aerial and ship-based surveys used within the UK. Both digital aerial and static PAM surveys can collect data on fine temporal and spatial scales, though they have their strengths and limitations. Aerial surveys typically provide better spatial survey coverage than static acoustic recorders, while acoustic recorders generally provide improved temporal coverage. The overarching goals of this project were to: (1) produce a modelling framework integrating DAS and PAM data; (2) produce a test case study on harbour porpoise to validate the methods; and (3) provide recommendations on standards for static PAM and DAS data collection.
\r\n\r\nThis project ran from January 2023 \u2013 February 2024 and a series of technical meetings were held by the project team to review data integration methods, available software to assist data integration and survey design considerations and recommendations. A dataset from the Moray Firth, Scotland, was prepared and analysed for the case study. The main deliverables were this final report and accompanying analysis code.
\r\n\r\nMultiple methods to integrate DAS and PAM data were assessed. The selected method for the case study used a Bayesian model to calibrate the PAM data using absolute densities derived from the DAS data. This method allows uncertainty to be propagated in all elements of the density estimators for both the DAS and PAM data. Currently available tools to assist with DAS and PAM data integration were also assessed. R-based MRSea software was used for the spatial modelling components of the case study analysis, and the project also identified how other R-based tools can be used for survey design (dssd/dsims) and to assess power to detect changes in density/abundance (MRSeaPower/AVADECAF). All assessed software packages have potential for extensions, which would ultimately aid data integration, though these were outside the scope of this project. Discussions about software concluded that clear documentation and long-term support are key features of any software used for analysis so should be considered a priority in any future software development.
\r\n\r\nThe case study used PAM and DAS data in the Moray Firth from August and September 2010 to assess the distribution and abundance of harbour porpoise (Phocoena phocoena). The analysis demonstrated the use of Bayesian data integration following methods in Jacobson et al. (2017). A parameter combining detection probability of harbour porpoise clicks and probability of clicking was estimated, with associated uncertainty. The estimation of this parameter enabled absolute density to be estimated from the PAM data, including during time periods where no DAS were flown. This is the primary benefit of implementing this method: long-term time series of PAM data can be used to estimate absolute densities, assuming that the parameters estimated from the combined DAS and PAM data are representative across the time periods analysed. Density surfaces were also estimated from the calibrated PAM data, showing spatial changes in absolute density. This approach is likely to be of most practical use in applications to support offshore wind development, where DAS data are able to provide an estimate of the absolute density of cetaceans (albeit with some limitations) and the PAM data play a supporting role by collecting continuous data across the time period of interest. Relevant code and data used for the case study are available on GitHub. Finally, survey design discussions within the project team highlighted key topics such as (1) defining project goals, (2) addressing survey design principles such as replication and coverage at the survey design stage using available software tools, (3) using best available data regarding required parameters for density/abundance estimation and (4) specific considerations relating to DAS-PAM data integration. Future survey design-related research steps were also outlined, focusing on assessing how the number of PAM instruments and DAS flights influence precision and accuracy in the resulting calibrated PAM data.
\r\n\r\nIn conclusion, the following survey design recommendations were suggested:
\r\n\r\nFinally, the following future research directions across the project were identified and summarised:
\r\n\r\nMarine mammal abundance and distribution data form an important part of assessments to estimate the potential effects of proposed offshore wind developments. Therefore, there is a need to ensure that abundance and distribution data are collected and analysed to ensure robust estimates to inform the planning, consenting and licensing processes. There are several sources of information that can contribute to estimating the abundance and distribution of marine mammals. Digital aerial surveys (DAS) and static passive acoustic monitoring (PAM) are two data collection modes, which have been developed relatively recently, compared to standard visual aerial and ship-based surveys used within the UK. Both digital aerial and static\u00a0PAM\u00a0surveys can collect data on fine temporal and spatial scales, though they have their strengths and limitations. Aerial surveys typically provide better spatial survey coverage than static acoustic recorders, while acoustic recorders generally provide improved temporal coverage. The overarching goals of this project were to: (1) produce a modelling framework integrating\u00a0DAS\u00a0and\u00a0PAM\u00a0data; (2) produce a test case study on harbour porpoise to validate the methods; and (3) provide recommendations on standards for static\u00a0PAM\u00a0and\u00a0DAS\u00a0data collection.
\nThis project ran from January 2023 \u2013 February 2024 and a series of technical meetings were held by the project team to review data integration methods, available software to assist data integration and survey design considerations and recommendations. A dataset from the Moray Firth, Scotland, was prepared and analysed for the case study. The main deliverables were this final report and accompanying analysis code.
\nMultiple methods to integrate\u00a0DAS\u00a0and\u00a0PAM\u00a0data were assessed. The selected method for the case study used a Bayesian model to calibrate the\u00a0PAM\u00a0data using absolute densities derived from the\u00a0DAS\u00a0data. This method allows uncertainty to be propagated in all elements of the density estimators for both the\u00a0DAS\u00a0and\u00a0PAM\u00a0data. Currently available tools to assist with\u00a0DAS\u00a0and\u00a0PAM\u00a0data integration were also assessed. R-based\u00a0MRSea\u00a0software was used for the spatial modelling components of the case study analysis, and the project also identified how other R-based tools can be used for survey design (dssd/dsims) and to assess power to detect changes in density/abundance (MRSeaPower/AVADECAF). All assessed software packages have potential for extensions, which would ultimately aid data integration, though these were outside the scope of this project. Discussions about software concluded that clear documentation and long-term support are key features of any software used for analysis so should be considered a priority in any future software development.
\nThe case study used\u00a0PAM\u00a0and\u00a0DAS\u00a0data in the Moray Firth from August and September 2010 to assess the distribution and abundance of harbour porpoise (Phocoena phocoena). The analysis demonstrated the use of Bayesian data integration following methods in Jacobson et al. (2017). A parameter combining detection probability of harbour porpoise clicks and probability of clicking was estimated, with associated uncertainty. The estimation of this parameter enabled absolute density to be estimated from the\u00a0PAM\u00a0data, including during time periods where no\u00a0DAS\u00a0were flown. This is the primary benefit of implementing this method: long-term time series of\u00a0PAM\u00a0data can be used to estimate absolute densities, assuming that the parameters estimated from the combined\u00a0DAS\u00a0and\u00a0PAM\u00a0data are representative across the time periods analysed. Density surfaces were also estimated from the calibrated\u00a0PAM\u00a0data, showing spatial changes in absolute density. This approach is likely to be of most practical use in applications to support offshore wind development, where\u00a0DAS\u00a0data are able to provide an estimate of the absolute density of cetaceans (albeit with some limitations) and the\u00a0PAM\u00a0data play a supporting role by collecting continuous data across the time period of interest.\u00a0Relevant code and data used for the case study are available on GitHub.\u00a0Finally, survey design discussions within the project team highlighted key topics such as (1) defining project goals, (2) addressing survey design principles such as replication and coverage at the survey design stage using available software tools, (3) using best available data regarding required parameters for density/abundance estimation and (4) specific considerations relating to\u00a0DAS-PAM\u00a0data integration. Future survey design-related research steps were also outlined, focusing on assessing how the number of\u00a0PAM\u00a0instruments and\u00a0DAS\u00a0flights influence precision and accuracy in the resulting calibrated\u00a0PAM\u00a0data.
\nIn conclusion, the following survey design recommendations were suggested:
\nFinally, the following future research directions across the project were identified and summarised:
\n