The assessment is based on the numbers of breeding pairs of selected waterbird species, counted in breeding colonies or in monitoring plots. Site level raw data are used for each species to calculate the annual indices and trends. The national monitoring programmes provide the breeding bird monitoring data. Each site level data for each species consists of site code, coordinates of the site, year of survey, recorded abundance and the units in which the abundance is expressed (mostly pairs). There is a separate entry for each year the site was visited. Each site is assigned a code indicating to which country and assessment unit it belongs.
To calculate the yearly indices and trends, the TRIM framework and "rtrim" package for the R statistical software is used. Models explaining the observed abundance by site effects and year effects while accounting for serial correlation and overdispersion in the data are built for each species. The method is based on loglinear Poisson regression and is able to impute the missing observations (ter Braak et al. 1994, van Strien et al. 2001, 2004). For more details of the procedure, see also http://www.ebcc.info/trim.html and https://www.cbs.nl/en-gb/society/nature-and-environment/indices-and-trends--trim--/ . The method produces yearly indices and linear trend estimates (the slope of the regression line through the logarithm of the indices). The year 1991 or the start year of the time series (if later) is used as the point of reference (when the index is 1), but the results are then scaled to a reference period (i.e. the average index values from 1991-2000 are scaled to 1).
The multiplicative overall slope estimate in TRIM is converted into one of the following categories. The category depends on the overall slope as well as its 95% confidence interval (= slope +/- 1.96 times the standard error of the slope) (Pannekoek & van Strien 2001):
All analyses are conducted on the level of species. Though in some species diverging trends are observed in different parts of the Baltic Sea, those differences are owing to two subspecies only in lesser black-backed gull and black guillemot. Lesser black-backed gull is represented by Larus fuscus intermedius in Kattegat, Belt Group and Bornhom Group, but by L. f. fuscus in Gotland Group, Åland Group, Gulf of Finland and Bothnian Group. Black guillemots breeding in Kattegat and Belt Group belong to the subspecies Cepphus grylle arcticus, those from further east in the Baltic Sea to C. g. grylle.
For the parameter breeding success of Baltic waterbirds, no assessment protocol currently exists.
The indicator is in a state allowing evaluation of the status of breeding waterbirds in the entire Baltic based on population sizes. Development is needed to include breeding success as an additional criterion to assess the status of breeding waterbirds. The assessment of population sizes would gain from the establishment of species-specific reference periods, which would allow to compare recent population sizes to pristine conditions.
The assessment units are defined in the HELCOM Monitoring and Assessment Strategy Annex 4.
The assessment was conducted at two spatial scales, the entire Baltic Sea (HELCOM assessment unit scale 1) and seven subdivisions of the Baltic Sea, which were defined as aggregations of up to four of the 17 sub-basins (HELCOM assessment unit scale 2) following recommendation by OSPAR/HELCOM/ICES (2017, 2018) (Assessment units figure 1). Several waterbird species (terns in particular) are known to switch between breeding colonies from year to year, possibly even at distances involving switches between sub-basins, leading to the estimate that HELCOM assessment unit scale 2 is not an appropriate scale. Further, the use of the seven subdivisions shall make it easier to localize problems and to implement necessary regional or local measures to improve the status. These smaller scale assessments are better suited to reflect the conditions of a given part of the Baltic Sea rather than downscaling the results from the entire Baltic Sea to everywhere. In addition, subdivision assessments serve better the national reporting according to Article 8 of MSFD, because there is much less influence from other parts of the Baltic on the national assessments. The seven subdivisions are defined as follows:
Assessment units figure 1. Grouping of 17 sub-basins (HELCOM assessment unit scale 2) to seven subdivisions as spatial units for breeding waterbird abundance evaluations as recommended by OSPAR/HELCOM/ICES (2018). The left figure shows the entire subdivision coloured, and the right figure shows the coastal areas, as used in the current assessment, coloured by the seven subdivisions.