The environmental status evaluation for the coastal areas of the Baltic Sea using the productivity of the white-tailed sea eagle considers three parameters: productivity, nestling brood size and breeding success. For six of the 11 assessment units (55%, see Key message figure 1 and Results figure 1), good status was achieved for all three parameters in the evaluated period 2011-2016. The failure to achieve good status was mainly caused by nestling brood size not reaching the threshold values (Key message figure 1 and Results figure 1) in four regions: Finland – Gulf of Bothnia, Germany, Latvia, and Sweden – Gulf of Bothnia. Breeding success reached the threshold in all regions except the Finnish Åland/Åbo region (the Archipelago Sea). Productivity failed the threshold value when evaluated on its own only in Germany.
Results figure 1. Estimates for white-tailed sea eagle productivity (top), brood size (middle) and breeding success (bottom) for the Holas II-evaluation period 2011-2016. Confidence intervals were estimated by a non-parametric bootstrap applied to the pooled data for each individual region 2011-2016. Reference levels based on "historical" data from Sweden is drawn as horizontal lines, where the lower limit of the 95% confidence interval (drawn as red dashed lines) equals the threshold value for the evaluation. Point estimates of variables not reaching Gold (due to the "one-out, all-out"-nature of this indicator).
The evaluation of mean annual productivity during 2011-2016 indicates that the threshold value is achieved in all studied areas (Key message figure 1 and Results figure 1), with the exception of Germany. Note that the full evaluation of the indicator must also consider the brood size and breeding success variables.
The time series since the 1970s (Results figure 1), available for some countries, indicates a great increase in productivity since the mid-1980s, and the threshold value was reached or nearly reached mainly during the last 10 years. In several of the studied areas the increase in productivity has stabilized at the lower end of the estimated reference level. See for instance time series for Germany and Sweden (Results figure 2).
Results figure 2. Mean annual productivity (number of nestlings per checked occupied territory) of coastal subpopulations of white-tailed sea eagles around the Baltic Sea. Trend lines (grey smoothers) are drawn for series where a significant time trend was identified by fitting a General Additive Model (GAM) to the data set. Periods of significant change were found by calculating the rate of change (and corresponding confidence intervals) in the fitted GAM-smoother. Bold blue parts illustrate significant increases in the time trend, whereas bold red lines illustrate significant decreases.
A pre-1950 reference level (middle horizontal line) with 95% confidence intervals for breeding success (according to Helander 2003a) is given in each graph. The threshold value is the lower boundary of the confidence interval (red dashed horizontal line).
Nestling brood size reached the threshold value in 2011-2016 in six regions, but four regions (Finland – Gulf of Bothnia, Germany, Latvia and Sweden – Gulf of Bothnia) also failed the threshold value.
Nestling brood size has improved strongly since the 1970s but has not reached the threshold value in all parts of the Baltic Sea. Brood sizes began increasing in the studied areas from the 1980s (Results figure 3), roughly in synchrony with the increase in breeding success (Results figure 4). This is inherent with an improvement in the hatching success of the eggs, affecting both indicator parameters in parallel. Brood size in the Baltic Proper reached the lower end of the pre-1950 reference level in the late 1990s (Results figure 3).
Results figure 3. Mean brood size (number of nestlings per successfully breeding pair) of coastal white-tailed sea eagle subpopulations around the Baltic Sea. Trend lines (grey smoothers) are drawn for series where a significant time trend was identified by fitting a General Additive Model (GAM) to the data set. Periods of significant change were found by calculating the rate of change (and corresponding confidence intervals) in the fitted GAM-smoother. Bold blue parts illustrate significant increases in the time trend, whereas bold red lines illustrate significant decreases.
A pre-1950 reference level (middle grey line) with 95% confidence limits according to Helander (2003a) is given in each graph. The threshold value is the lower boundary of the confidence interval (red dashed horizontal line).
The overall positive long-term trend for regions with the longest data series is obvious, but the times series might also reflect on-going short-term changes. Although difficult to detect if located towards the end of the time series, some regions (Results figure 3) show short-term decreases in brood sizes during the evaluation period. This should be taken into consideration, as the failure to achieve threshold values and good status might indicate a downwards change again, after a peak at the turn of the century (see for instance Sweden – Baltic Proper and Finland – Gulf of Bothnia, Results figure 3). This development should be followed closely, as alternative explanations do exist. It could for example be explained by hazardous substances or, perhaps more likely, by density-dependent effects. Sea eagle populations might have reached levels where demography and population process are increasingly affected by population density, which might mask effects of hazardous substances. This interplay between ecotoxicology and population ecology should be evaluated further among HELCOM-countries.
However, some results from recent decades likely suggests an impact of hazardous substances! The smaller average nestling brood size on the Swedish coast of the Bothnian Sea during 2000-2009 is due to a significantly higher frequency of nests with young that also contained dead eggs: 7.1% as compared to 2.9% in the Baltic Proper (n = 461 and 932, respectively). This may imply an influence of hazardous substances on the hatching success in the Gulf of Bothnia. This case also indicates that nestling brood size is a more sensitive indicator, specifically for hazardous substances, than productivity.
The breeding success of white-tailed sea eagles reached the threshold in nearly all areas along the Baltic Sea during 2011-2016 (Results figure 1 and Results figure 4), with the exception of the Finnish Archipelago Sea. Retrospective studies have shown that breeding success along the whole Swedish Baltic Sea coast decreased from an average of 72% in the early 1950s, down to 47% between 1954–1963, and 22% between 1966-1982 (Thresholds figure 1) (Helander 1985).
Breeding success then increased significantly from the early 1980s (Germany, Sweden) and generally reached good status by the mid- to late 1990s (Results figure 4). The development in the southern Baltic Sea (Germany) is similar to that in the central parts (Sweden and Finland). Impacts of intraspecific competition in areas with a high density of breeding pairs in Mecklenburg-Western Pomerania have been discussed as a possible reason for lower breeding success in recent years (Hauff 2009; Heuck & Albrecht 2012; Heuck et al. 2017). In densely populated areas also in Sweden and Finland, fatal territorial fights have been recorded more frequently in recent years. Intraspecific competition in densely populated areas could potentially cause decreases in breeding success, but it should be noted that evidence of decreases in breeding success is not very obvious in the data evaluated here (Results figure 4). Rather, it seems that breeding success in general is increasing or stable in most regions. Also note that the reference level is based on data from a more sparse population during the first half of the 20th century, where intra-specific competition likely had a negligible effect.
Results figure 4. Breeding success in % (proportion of successfully reproducing out of all checked territorial pairs) of coastal white-tailed sea eagle subpopulations around the Baltic Sea. Trend lines are drawn for series where a significant trend, was identified by fitting a Generalized Additive Model (GAM) with binomial error structure to the data. See Figure 2 for additional explanations.
A pre-1954 reference level (middle grey line) with 95% confidence limits according to Helander (2003a) is given in each graph. The threshold value is the lower boundary of the confidence interval (red dashed horizontal line).
The confidence of the indicator status evaluation is considered to be high. Annual data is currently available from nine countries, covering almost the entire Baltic Sea coastal area. White-tailed sea eagle reproduction has been monitored on an annual basis around the Baltic Sea for decades and the available historical data for all three evaluated parameters is considered to increase the overall confidence of the indicator evaluation.
There is no bias in the spatial distribution of the data. The parameters are robust and the comparability of data from different areas is high. Annual sample sizes are big for countries with long stretches of coastline and are adequate for other countries based on averages for 5-10 year periods. The national monitoring is generally focused on the whole population, and subsets of available data is used for HELCOM-purposes.
Sample sizes do however vary in relation to population numbers and size of the assessment units, which also affects confidence of the indicator in statistical terms. There is considerable uncertainty in parameter estimates for some regions (see Results figure 1, particularly in southeastern Baltic Sea).