Results and Confidence

This indicator is a HELCOM supplementary indicator and is applicable in assessment units shared by Finland and Sweden.


This supplementary indicator on malformed embryos of amphipods has been applied in sub-basins shared by Finland and Sweden (the Quark and the Bothnian Sea), the Northern Baltic proper and Western Gotland Basin. A preliminary assessment was also conducted in the Swedish coastal waters of the Bornholm Basin. Although during 2011-2016 most of the observations are within the BAC values, i.e. achieve the threshold value, the deviations in both frequency of embryo malformations and frequency of females carrying malformed embryos were apparent in at least some areas of all assessment units except the Quark (Results and confidence figure 1). For comparative purposes, the indicator evaluation for Hanö Bight (Bornholm, Swedish waters; this area is known as a disturbed system, with strong indications for contaminant-driven changes) was conducted, and the results indicate high reproductive and developmental toxicity in the amphipod populations inhabiting this area (Results figure 1). For the assessment of different basins on the annual basis, a 50% rule has been applied which means that:

  • each station per year was assessed individually;
  • all stations grouped by regions were evaluated, and the region was considered as being in good status when >50% of the stations within the region were in good status;
  • all regions within a basin were evaluated, and the basin was considered as being in good status when >50% of the regions within the basin were in good status;
  • the whole six-year assessment period was considered as being in good status if >3 years were in good status.

Assessment for the period 2011 to 2016 concluded that the Quark and the Western Gotland Basin have achieved good status, whereas Bothnian Sea and Northern Baltic Proper were not in good status.

The highest levels of the reproductive aberrations (frequency of females with aberrant embryos >0.4 and frequency of malformed embryos >0.06) that were observed in the evaluated units during 2011-2016 are sufficient to cause increased fluctuations in the amphipod stocks, decline in the abundance and increased probability of population decline and extinction, as suggested by population modelling of M. affinis (Reutgard 2015). It is unlikely that the entire Baltic population can go extinct, but decreases in local population abundance may become detrimental for the productivity of benthic communities now dominated by M. affinis, benthic-pelagic coupling and energy fluxes in the soft-bottom sediments.

Results figure 1.png

Results figure 1. Assessment results on the performance of embryo malformations indicator), which integrates frequency of malformed embryos (Y axis) and frequency of females carrying more than one malformed embryo (X axis) in the population of Monoporeia affinis in 2011-2016. Blue and red lines show threshold values for the frequency of malformed embryos and frequency of females carrying more than one malformed embryo, respectively. Green shaded quartile indicates good status. Thus, observations in good status are those located in the green shaded area, and not in good status are outside of it. Each data point represents a surveyed region with several sampling stations: the Quark (1 region, 4 stations), the Bothnian Sea (BS; 4 regions, 13 stations), the Northern Baltic Proper (NBP; 1 region, 5 stations), the Western Gotland Basin (WGB; 3 regions, 13 stations), and Bornholm basin (single station in Hanö Bight); see Assessment Protocol figure 1 for sampling coverage in each assessment unit.


The relative importance of specific malformation types to the indicator-based status assessment and differences among the basins has been evaluated (Results figure 2). It was apparent that the frequencies of different aberration types vary among the basins, with arrested development being more common in the Western Gotland Basin, whereas membrane damage and complex developmental malformations of the body (limbs, eye and midgut) were more common in the Bothnian Sea and the Northern Baltic Proper. Since these malformation types were found to be specific for certain groups of contaminants (Löf et al. 2015), it would be relevant to explore these associations and integrate these data with assessment based on the contaminant analyses.


Results figure 2.png

Results figure 2. Between-group Principal Component Analysis (bgPCA) evaluating relative contributions of specific malformation types (AD: arrested development, MD: membrane-damaged embryos, M: complex developmental pathologies, malformed limbs, eye and midgut) and proportion of females with more than one malformed embryo (F) to the basin-specific status of reproductive aberrations and the separation between the assessment units. Each data point is the combined value for a region, in each region, several stations were surveyed; see Assessment Protocol figure 1.


Confidence of the indicator status evaluation

The confidence of the indicator is high in the Bothnian Sea and Western Gotland Basin since more than 20 years of data following the same methodology for embryo analysis were used for setting the threshold values, and the 2011-2016 assessment is based on the data collected from multiple stations that were surveyed on annual basis within these assessment units. The confidence of the indicator is moderate in the Quark and the Northern Baltic Proper, because the spatial coverage and number of stations were somewhat lower (Results figure 1), although, considering a smaller total area of the Quark, this sampling coverage is comparable to that in the Bothnian Sea. Another area of concern with respect to the uncertainty assessment is the varying availability of the field-collected females for embryo analysis, and, thus possible inequality of the sample size between the stations/assessment units. The statistical analyses underlying indicator evaluation employ bootstrapping to generate distributions for identical sample size and derive an estimate of the confidence interval. To decrease statistical uncertainty of the analysis, several (4-8) sampling stations per assessment unit are recommended.


Future work

At present, the embryo malformation indicator has not been evaluated for all assessment units in the Baltic Sea, partly due to the lack of monitoring in these areas. The validation of the applicability of the indicator and the determination of the threshold values are needed in the Åland Sea as well as much of the eastern, south-eastern, and southern Baltic Sea before the evaluation for these areas can be conducted.

In Monoporeia affinis populations exposed to contaminated sediments in situ, frequencies of different aberration types were found to be a function of the contaminant type. In particular, occurrence of females with embryo limb malformations was strongly related to elevated concentrations of Cd and PCBs, while females with membrane-damaged embryos occurred at high PAH concentrations. Also, frequency of embryos with arrested development was positively related to levels of PAHs and metals. Thus, these reproductive aberrations can serve as contaminant-specific indicators of PCB, PAH and heavy metal exposure in biological effect monitoring. In the assessment for 2011-2016, it was found that specific aberration types contribute differently to the reproductive toxicity in the evaluated assessment units (Results figure 2).

The integration of the contaminant-specific assessment and relative frequencies for specific embryo malformations (Results figure 2) would be necessary to establish a coherent assessment for biological effects of contaminants in the Baltic Sea. Depending on the outcome of the integrated analysis of the chemical and biological data, the next step might include development of thresholds for specific malformation types. The latter would provide several additional dimensions to the multimetric embryo malformation indicator to increase its specificity toward particular contaminant types, while keeping the same concept.