The main objective of this long-term study (1978-2016) was to find the underlying factors behind the declining trends of eider Somateria mollissima in the Baltic/Wadden Sea. Specifically, we aimed at quantifying the bottom-up effect of nutrients, through mussel stocks, on reproduction and abundance of eider, and the top-down effects caused by white-tailed eagle Haliaeetus albicilla predation. Bottom-up effects increase marine primary productivity with subsequent effects on food availability for a major mussel predator. Top-down effects may also regulate eider populations because during incubation female eiders are vulnerable to predation by eagles.

Our structural equation modelling explained a large percentage of the variance in eider abundance. We conclude that the Baltic/Wadden Sea eider population was regulated directly by white-tailed sea-eagle predation on incubating females and indirectly by the amount of nutrients in seawater affecting both mussel stocks and the breeding success of eiders, reflecting density dependence. These findings may explain the decreasing trend in the Baltic/Wadden Sea eider population during the last decades as an additive effect of top-down and bottom-up factors, and likely as an interaction between them.

Eider numbers

Due to lack of long-term estimates of the Baltic/Wadden Sea eider population the Danish eider bag during 1960-2003 was used as a proxy for population size. After 2003, the open season was shortened, and thus not comparable to population size. The eider bag increased from 93,000 individuals in 1960 to a maximum of 190,300 in 1982, followed by a decrease to 75,200 in 2003. It is mandatory for hunters in Denmark to report their bag after each hunting season, and in most years, more than 85% of all hunters report the bag. Detailed information includes assessment of reliability by Strandgaard and Asferg (Strandgaard & Asferg 1980) and by Christensen and Hounisen (Christensen & Hounisen 2014). Significant positive correlations between the Danish eider bag and counts of eiders in Finland, Sweden and at migration sites is given in ESM (S1).

Breeding success of eider

The ratio of juveniles to the number of adult females in wing samples reported by hunters under the Danish Wings Survey scheme was used as estimates of breeding success (Clausager 2004; Christensen 2013). We included wings from adult females and juveniles shot during the first two months (1 October - 30 November) of the open season during 1982-2003, which provided an average (SE) of 514 (44) wings per year, all of which were sexed and aged (Boyd., Harrison & Allison 1975). Comparison of the breeding success index and fledging success (log₁₀ transformed) measured in the large eider colony at Tvärminne, Finland show a significant, positive relationship (linear regression: F = 4.88, df = 1,18, p = 0.041, estimate (SE) = 0.2801 (0.1268), data from Markus Öst for 1990-2014 (no data available for 1992-1993)).

Monitoring of white-tailed sea-eagles

During May to June all suitable sites were monitored to assess occupancy and breeding success by a team of experienced volunteers. Due to building of more nests in a territory search for alternative nests were done. However, logistic constraints restricted it in some territories. A nest was considered occupied when it was decorated with fresh twigs. Of the occupied nests, those with at least one fully-grown young was considered successful. In total 150 territories of white-tailed sea-eagles were monitored during 1978-2016 at Åland, Finland. The territories have been surveyed for 1-48 years, and 28 territories have been followed for more than 20 years.

Fertilizer and outlet of nitrogen

Fertilizers are used by farmers to stimulate growth of the crops. However, parts of the fertilizers are not used by plants, but are transported via rivers to coastal waters, where they stimulate plankton growth. Thus, fertilizer and the amount of outlet is two measurements of nitrogen supply to the environment (earth and marine water).  An index of fertilizer (1950 = 100) used by Danish farmers during 1920-1998 is reported by Duus and Zinglersen (Duus & Zinglersen 2000) and for 1990-2016 by Blicher-Mathiesen et al. (Blicher-Mathiesen et al. 2019). The two data sets were calibrated for 1998 by estimating the means of the two data sets during the overlap period (1990-1998, correlation between the two data set was r² = 0.96), and multiplying the values from Blicher-Mathiesen et al. (Blicher-Mathiesen et al. 2019) by the ratio. The index of fertilizer increased from 44 units in 1920 to 385 units in 1991 subsequently decreasing to 272 units in 2015. The annual outlet of total-N (t) via rivers to coastal waters in Denmark was used as an indicator of nutrients in the marine environment at the winter grounds as given by Conley et al. (Conley et al. 2007). Total outlet of total-N increased from 54,415 t in 1960 to 124,311 in 1982, and decreased afterwards to 46,250 t in 2003.

Estimation of mussel stocks

The total annual stocks of blue mussels Mytilus edulis have been estimated (metric t) for the intertidal zone in the Danish (1986-2007) and Schleswig-Holstein (1988-2015, missing values in 1994-1997) parts of the Wadden Sea (Kristensen, Borgstrøm & Laursen 2006; Büttger, Wittes & Nehls 2012). There was a positive correlation between the annual mussel stocks in the two parts of the Wadden Sea (r = 0.92). The data suggest that mussel stocks develop synchronously across years and in large marine areas (Nehls et al. 2009). The mussel stocks increased from 64.100 t in 1986 to 171.800 t in 1994 after which it decreased to 43.350 t in 2013. Further information on estimation of mussel stock is given in ESM (S1).

The dataset contains the metadata in the second excel sheet, describing each field. 

The dataset contains the metadata in the second excel sheet, describing each field. 

Additional information about the dataset could be asked to the corresponding author: E-mail: anders.moller@u-psud.fr