Results and Confidence

The evaluation of zooplankton mean size and total stock (MSTS) for the period 2011-2016 indicates that in the Bothnian Bay, the Bothnian Sea and the Gdansk Basin, the MSTS values are above the threshold values indicating good status. By contrast, in the Åland Sea, Gulf of Finland and Western Gotland Basin, the MSTS values are significantly below the threshold values, which implies that good status has not been achieved. The details for each of the evaluated sub-basins are presented below.

In the Bothnian Bay (Results figure 1a), MSTS has not changed considerably over the time period for which data are available (1979 – 2016). Although in some years values were below the threshold, these occasional deviations were not significantly different from the threshold value as indicated by CuSum analysis, implying that these fluctuations are stochastic and that MSTS reflects a good status.

In the Bothnian Sea (Results figure 1b) MSTS also suggests good food web structure, with no indication of the status decline over the assessed period. In the Åland Sea (Results figure 1c), starting from 1996, zooplankton mean size stayed significantly below the threshold and the total biomass values were often below the threshold values. Although the total biomass occurring during 2011-2016 are not significantly below the threshold, the mean size is, which implies that that the good status has not been achieved in this sub-basin.

In the Gulf of Finland (Results figure 1d), the values of the mean size indicate that the system was not in good status from 2001 onwards. Also, the biomass failed the threshold during the same years on multiple occasions, albeit not significantly. Thus, MSTS indicates that in 2011-2016, zooplankton community is not in good status.

In the Western Gotland Basin (Results figure 1e), the MSTS indicates that the system is not in good status since 1998, although some signs of recovery – at least in the coastal station Askö (monitoring station B1) – appear after 2007. Nevertheless, during the assessment period 2011-2016, zooplankton community is not in good status.

In the Gdansk Basin (Results figure 1f), the MSTS values indicate that the system is good status, with no deviations from its reference state for the last 30 years with regard to both mean size and biomass values.

 

 Zooplankton results figure 1.png

Result figure 1. Assessment results on the performance of MSTS indicator, which integrates mean size (Y axis) and total biomass of zooplankton (X axis). Blue and red lines show threshold values for the total biomass and mean size, respectively. Green shaded quartile indicates good status. Observations in good and not in good status are shown as green and red years, respectively. Stars indicate the assessment period years (2011 to 2016) with blue and red symbols (stars) for the observations that are in good and not in good status, respectively. Note that some years falling below the threshold values were assigned as being in good status, because these values were not judged as significantly different from the threshold value according to the CuSum analysis, which is based on the cumulative summing of the persistent deviations from the reference mean.

 

The difference in the MSTS components between the reference conditions and the assessment period varied from -34% to +75% for the mean zooplankter size (µg ind-1) and from -42% to +42% for the total biomass (mg m-3) among the sub-basins (Results figure 2). Prominent decreases in both body size and total biomass of zooplankton were observed in the Åland Sea, Gulf of Finland and Western Gotland Basin, where size and total biomass decreased by 39% and 38%, respectively, from the reference period to the assessment period (2011-2016). Similar changes occurred in the Bornholm Basin (preliminary assessment) where mean size and biomass decreased by 20% and 39%, respectively.

Contrary to all other sub-basins, both mean size and biomass have increased in the Bothnian Sea from the reference period to the assessment period (Results figure 2). The increase observed in the Bothnian Sea is related to an increased population size of the large-bodied copepod Limnocalanus macrurus. This species, which is a glacial relict in the Baltic Sea, responded positively to the low salinity conditions during the last decade, which improved herring feeding conditions (Rajasilta et al. 2014) as well as MSTS values in this sub-basin. In the other sub-basins, species that contributed to the detected changes in the MSTS components varied. However, regardless of the variability among the species and species groups contributing to general declines in body size and biomass values among the sub-basins, an increase in proportion of small-sized taxa and groups was observed in all assessment units (except the Bothnian Sea). In the Gulf of Finland, the change is largely attributed to a decline in the biomass of large cladocerans. In the Western Gotland Basin and the Bornholm Basin, the decline in mean size and total biomass is mostly due to declining copepod populations and thus shifting size spectra and biomass of the zooplankton communities.


Zooplankton results figure 2.png

Results figure 2. Pair-wise comparisons between the MSTS values observed during the assessment period (2011-2015 for GoF and ÅS (n = 5), and 2011-2016 for BB, BS, WGB and GB (n = 6), where n is the number of years included in the analysis) and the reference period (RefCon; n varied from 8 to 17) for mean zooplankton size (MeanSize; A) and total zooplankton biomass (Total biomass; B) in the Bothnian Bay (BB), Bothnian Sea (BS), Åland Sea (ÅS), Gulf of Finland (GoF), Western Gotland Basin (WGB) and Gdansk Basin (GB). The basin-specific data were compared using unpaired t-test with Welch correction and statistically significant differences (p < 0.05) are indicated with percent change and red asterisk. For the nearly significant difference in ÅS, the p value is shown. Percentage values indicate change (positive or negative) in the value observed for the assessment period relative to the reference period. Data are shown as means and standard deviations for the untransformed data; the statistical comparisons were done using Box-Cox transformed values that were normally distributed.

Future work

At present, the MSTS indicator has not been evaluated for all open sea assessment units in the Baltic Sea where zooplankton monitoring is conducted. The applicability of the indicator and the determination of relevant threshold values are still needed in the northern Baltic Proper, much of the eastern, south-eastern- and southern Baltic Sea before evaluation for these areas can be conducted.

Temperature- and salinity-induced MSTS responses also need to be further evaluated and, if relevant and significant, they need to be accounted for in the indicator-based assessment of the pelagic food webs.

In order to assess the status of the food webs in the Baltic Sea, further development of the interpretation of the indicator results in relation to other assessment results is needed. A full assessment of pelagic food webs is still to be developed, and the outcome of the MSTS-based assessment needs to be considered in conjunction with other food web indicators. The interpretation of MSTS should also be integrated with the results of the eutrophication status assessment.

 

Confidence of the indicator status evaluation

The overall confidence of the evaluation varies from low to high between the assessment units.

Confidence of the evaluation accuracy depends on the time series length and between-year variability during the reference period. Also, different number of stations per assessment unit contributes to the between-year data variability. The data availability is the main reason for the variation in the confidence across the assessment units.

Zooplankton monitoring stations are generally found in every Baltic Sea sub-basin, and suitable monitoring data series are available for relatively long (>18 years) time periods from most of the sub-basins. A similar confidence in the evaluation (moderate to high) is expected for the most assessed basins with fairly similar length of the data sets and similar number of observations (number of data points per basin and per year). However, in case of low observation frequency (for example, Gdansk Basin, where only August data were used from a single station each year), the confidence is low.

The accuracy component of the confidence is considered to be high. This confidence classification is due to: (1) the CuSum technique that is used to determine whether the observed value reflects good status or not is considered to be a very sensitive method for detecting persistent small changes (Lucas 1982),

(2) the lower bound of 99% confidence interval around the baseline (reference condition) was used as threshold, thus minimizing the risk of false negatives (i.e., assigning not good status to an observation that is in fact reflecting good status), and

(3) using a pre-cautionary principle by selecting the higher value after comparing threshold values obtained for RefConFish and RefConChl for each part of the indicator (i.e. MeanSize and total biomass).