Good status in terms of white-tailed sea eagle reproduction is evaluated using the parameter 'productivity' and the two supporting variables 'brood size' and 'breeding success'. For an assessment unit to be evaluated as having achieved the threshold values (i.e. good status), all three parameters have to achieve their respective threshold values. The threshold values for each parameters are based on an acceptable deviation from the target level determined during a reference period.
The reference levels are based on actual reference status data collected from the Swedish Baltic Sea coast. Breeding success data cover the period 1915-1953 and nestling brood size includes data from the period 1858-1950. The target level for productivity is based on the combined data for breeding success and nestling brood size. It should be noted that the population in those times was much smaller than today and was most probably under no influence of density-dependent effects.
Due to the lack of reference data from other parts of the Baltic Sea, the same reference level has been tentatively used for the entire Baltic Sea coastal zone. Where possible, the applicability of the reference level and the resulting threshold values should be validated using data from other parts of the Baltic Sea.
The reference level for breeding success has been determined based on data from the period 1915-1953 (n=43 years). The data has been assembled as series of records over time periods of 3-10 years in succession from eight white-tailed sea eagle territories. The mean value of successful nests was 72%, and the 95% confidence interval ranges from 59% to 86% based on binomial distribution (Thresholds figure 1).
Thresholds figure 1. Breeding success (% reproducing pairs) in the white-tailed sea eagle population on the Swedish Baltic Sea coast. The upper dot and the blue line indicate the background reference mean value with 95% confidence interval (grey) for a time period 1915-1953, based on data from eight eagle territories (n=43). The lower dot indicates a mean value for the time period 1954-1963 based on data from 14 territories (n=68).
The reference level for brood size has been determined based on data on white-tailed sea eagle nestling brood size retrieved from banding records and from literature, comprising a total of 91 broods from the period 1858-1950. The recorded brood size can only result in a discrete number (1, 2 or 3 nestlings). Up to 1950, the arithmetic mean for nestling brood size was 1.84. The distribution for samples taken from such a population cannot be expected to be normally distributed. In order to investigate the true sample distribution, for estimation of a confidence interval around the mean value for brood size, samples of 25 individual brood sizes were randomly taken from the population using the 1858 – 1950 dataset (Thresholds figure 2). This was repeated 1,000 times (bootstrapping). The estimated sample distribution deviates significantly from the normal distribution (p<0.03). An estimated 95% confidence interval for a sample size of 25 was between 1.64 and 2.04.
Thresholds figure 2. Mean brood size (number of nestlings per successfully breeding pair) of white-tailed sea eagle on the Swedish Baltic Sea coast 1854-2013. Sample size for each time period is given in brackets. A reference level (solid black line) with 95% confidence limits (grey area) is based on data from 1854-1950 (blue bars) according to Helander (2003a).
The reference level for productivity has been derived by combining the reference levels for brood size and breeding success. This gives a reference level for mean productivity of 1.84 x 0.72 = 1.32, with confidence limits from 1.64 x 0.59 = 0.97 up to 2.04 x 0.86 = 1.75. This estimate of confidence interval has been used in previous assessments. A more stringent estimate based on frequency distributions was derived from the dataset for nestling brood size (n = 91, all successful breeding attempts) with the addition of 35 'fictive' unsuccessful breeding attempts, based on the mean value of 72% breeding success in the population. The 95% confidence interval around the mean value of 1.32 was estimated with the same method as for nestling brood size above (bootstrapping) and is from 1.15 to 1.50. This confidence interval is built from a population that was probably under no influence from density dependent mechanisms. Under current conditions, it might be more appropriate to apply the wider interval given above, and setting the lower end at 0.97.
The target used to determine whether good status is achieved or not is set to the lower 95% confidence limit of the observations during the reference period. The data for the three parameters are presented as time trends. Observations should be measured as averages for a recent five to 10 year-period (depending on sample sizes). The current evaluation is based on an 8-year assessment period.
The threshold value to achieve good status for productivity is 0.97 nestlings.
The threshold value to achieve good status for brood size is 1.64 nestlings.
The threshold value to achieve good status for breeding success is 0.59 (59%).
Thresholds figure 3. The threshold values are based on an acceptable deviation from the target level determined during a reference period.
The confidence in the threshold values, based on the reference levels, is considered to be high as it has been determined based on carefully selected actual observations from the time period 1854-1953.