Abstract
Grey and harbour seal usage maps around the British Isles were produced at a spatial resolution of 5 km x 5 km. A paper was published describing how usage maps can be used and interpreted, the caveats and limitations, and methodology used to produce them, which is available to download:
- Jones, E.L., McConnell, B.J., Smout, S., Hammond, P.S., Duck, C.D., Morris, C.D., Thompson, D.,Russell, D.J.F., Vincent, C., Cronin, M., Sharples, R.J. & Matthiopoulos, J. (2015). Patterns of space usein sympatric marine colonial predators reveal scales of spatial partitioning. Marine Ecology ProgressSeries, 534, 235–249. doi:10.3354/meps11370
Usage maps have been made available to download in GeoTiff and shapefile formats. These are provided in a Universal Transverse Mercator 30⁰ N World Geodetic System 1984 datum (UTM30N WGS84) projection. This note details how terrestrial count data were used to scale the usage maps up to population levels, and how the maps should be interpreted. Readers should first be familiar with Jones et al. (2015) to contextualise detailsdescribed below.Aerial surveys count grey and harbour seals along a coastline on an annual basis during the harbour seal moult in August. Animals are counted and their locations are recorded to an accuracy of 50m. Not all the coastline is surveyed in any one year. The locations of survey effort are recorded. Usage maps use these historical count data to scale each onshore 5 km x 5 km grid cell in which seals are hauled out to a local population estimatefor all usage associated with that onshore cell. The count data used in this analysis were coded in such a way that a missing value for an onshore grid cell can either mean the cell was not surveyed that year or it was surveyed but no animals were observed. Currently, the algorithm used to generate usage maps treats every missing value as though no survey effort was expended, meaning that observations where the count was 0 werenot included. This may bias population estimates in instances where missing values in the count data represent no animals being observed (i.e. the count for the onshore grid cell was observed to be 0 at some point). This may particularly affect areas where the local population is fluctuating or decreasing (where 0 counts are more likely). Note that this misinterpretation is limited to usage map generation and does not affect annual population assessments to Special Committee on Seals (SCOS). Work is currently being undertaken in the Sea Mammal Research Unit, St Andrews to explicitly synthesise count and effort data, which will be rolled out in the next planned release of the usage maps. To gauge the potential impact of using non-0 count data in the usage maps, testing was carried out in two areas.
For the test case areas, variation in the population estimates of both species when using non-0 count and 0 count data ranged from -14.8% to +1.5% over the regions of Orkney and the Western Isles. The population estimate will also change with the inclusion of 0 count data when a recent count is 0; the estimate will beextrapolated to the current year (i.e. 2013), and may be higher or lower than when non-0 count data are used.
- Jones, E.L., McConnell, B.J., Smout, S., Hammond, P.S., Duck, C.D., Morris, C.D., Thompson, D.,Russell, D.J.F., Vincent, C., Cronin, M., Sharples, R.J. & Matthiopoulos, J. (2015). Patterns of space usein sympatric marine colonial predators reveal scales of spatial partitioning. Marine Ecology ProgressSeries, 534, 235–249. doi:10.3354/meps11370
Usage maps have been made available to download in GeoTiff and shapefile formats. These are provided in a Universal Transverse Mercator 30⁰ N World Geodetic System 1984 datum (UTM30N WGS84) projection. This note details how terrestrial count data were used to scale the usage maps up to population levels, and how the maps should be interpreted. Readers should first be familiar with Jones et al. (2015) to contextualise detailsdescribed below.Aerial surveys count grey and harbour seals along a coastline on an annual basis during the harbour seal moult in August. Animals are counted and their locations are recorded to an accuracy of 50m. Not all the coastline is surveyed in any one year. The locations of survey effort are recorded. Usage maps use these historical count data to scale each onshore 5 km x 5 km grid cell in which seals are hauled out to a local population estimatefor all usage associated with that onshore cell. The count data used in this analysis were coded in such a way that a missing value for an onshore grid cell can either mean the cell was not surveyed that year or it was surveyed but no animals were observed. Currently, the algorithm used to generate usage maps treats every missing value as though no survey effort was expended, meaning that observations where the count was 0 werenot included. This may bias population estimates in instances where missing values in the count data represent no animals being observed (i.e. the count for the onshore grid cell was observed to be 0 at some point). This may particularly affect areas where the local population is fluctuating or decreasing (where 0 counts are more likely). Note that this misinterpretation is limited to usage map generation and does not affect annual population assessments to Special Committee on Seals (SCOS). Work is currently being undertaken in the Sea Mammal Research Unit, St Andrews to explicitly synthesise count and effort data, which will be rolled out in the next planned release of the usage maps. To gauge the potential impact of using non-0 count data in the usage maps, testing was carried out in two areas.
For the test case areas, variation in the population estimates of both species when using non-0 count and 0 count data ranged from -14.8% to +1.5% over the regions of Orkney and the Western Isles. The population estimate will also change with the inclusion of 0 count data when a recent count is 0; the estimate will beextrapolated to the current year (i.e. 2013), and may be higher or lower than when non-0 count data are used.
| Original language | English |
|---|---|
| Publisher | Marine Scotland |
| Commissioning body | Marine Scotland |
| Number of pages | 11 |
| Publication status | Published - 1 Oct 2016 |
