Thresholds and Status evaluation

This indicator and its threshold values are yet to be commonly agreed in HELCOM. 
The indictor is included as a test indicator for the purposes of the 'State of the Baltic Sea' report, and the results are to be considered as intermediate.


The concept for evaluating good environmental status using the succession of dominant groups in the phytoplankton community is structured around a reference status succession and the acceptable deviation from that pattern. The indicator evaluates the coincidence of seasonal succession of dominating phytoplankton groups over an assessment period (commonly 5−6 years) using regionally established reference seasonal growth curves and wet weight biomass data. The indictor result value is based on the number of data points falling within the acceptable deviation range set for each monthly point of the reference growth curve and expressed as the percentage to the total number of data points. This result value is then compared to regionally relevant threshold values established to represent acceptable levels of variation. Strong deviations from the reference growth curves will result in failure to meet the thresholds set for acceptable variation, indicating impairment of the environmental status and a failure to meet good status (Thresholds figure 1).


Seasonal phytoplancton Thresholds figure 1.png

Thresholds figure 1. Good status is achieved when the indicator result (number of data points that fall within the established acceptable variation range) is above the regionally defined threshold value.

 

The specific regional threshold values used in this test indicator are presented in Thresholds table 1. The threshold values are still preliminary and subjects of possible change for the next assessment period. The final evaluation is based on the average score of single dominant groups. This indicator may also be used as background data for the development of a modified lifeform approach in the monitoring and environmental assessments in the HELCOM area. Lifeform approach has been considered to be taken into use in the MSFD assessments by OSPAR (Tett et al. 2008, Gowen et al. 2011).

 

Thresholds table 1. Reference periods and threshold values (percentage of data points falling within the acceptable deviation range set for each monthly point) for selected assessment units in the Baltic Sea area.

Assessment unitReference periodThreshold value
Bay of Mecklenburg*2006-20150.71
Arkona Basin2005-20140.70
Bornholm Basin2002-20110.60
Gdansk Basin1995-2000, 2010-20150.58
Eastern Gotland Basin2005-20140.74
Gulf of Riga2000-2005, 2009-20150.69
Northern Baltic Proper (NBP)1997-20110.69
NBP Swedish coastal waters1997-20120.69
Gulf of Finland Estonian coastal waters (western part)1993-20040.65
Gulf of Finland Estonian coastal waters (eastern part)2001-20150.62
Gulf of Riga Estonian coastal waters1997-20100.67
Gulf of Riga Latvian coastal waters1999-2002, 2008-20150.61
Eastern Gotland Basin Lithuanian coastal waters2006-20150.64

*data sets from coastal and open sea waters combined.

 

Background information on deriving the threshold values

The term 'Good status' has, however, to be taken with care as the first eutrophication affected changes in ecosystems emerged already in the mid-1950s in the Baltic Sea (Andersen et al., 2015). In the few basins, regular phytoplankton datasets date back to the mid-1980s only, mostly the observations start from the 1990s. This means that most areas of the Baltic Sea have been heavily influenced by anthropogenic pressures prior to the initiation of regular monitoring and it may thus be difficult to determine the reference conditions for the succession based on pristine environmental conditions. Further analysis with data from mostly offshore areas seemed to indicate that in several cases, the deviations from the long-term mean reference growth curves have become less frequent during the last decade than in the 1990s and the early 2000s. This may infer an improvement in the current environmental status.

Due to the lack of confirmed high status waterbodies or historical datasets, the reference seasonal growth curves have been set through observations made after the 1980s and the threshold between good and not-good status is based on expert judgement. To define unit-specific reference conditions, the periods of stability in long-time biomass data were ascertained. This approach was tested in two ways: 1) by calculating 5-year moving averages of standard deviations in yearly total biomass values (Thresholds figure 2), and 2)by using a multiplicative decomposition model, where the values are seasonally adjusted and trends can be seen more clearly (Thresholds figure 3). The two methods resulted in partly different periods of stability, which may be used as reference periods in the evaluation of environmental status. Analysis with datasets from other sea areas will be continued.

The threshold values based on calculations with data points representing reference periods varied from 0.58 (Gdansk Basin) to 0.74 (Eastern Gotland Basin). The average value for open sea areas was 0.67 and 0.65 for coastal waters. Offshore communities might have more coherent responses across the sea than coastal communities that tend to be more isolated and may therefore show little coherence within and among regions (Griffiths et al., 2015). This also means that phytoplankton community structure and timely performance of dominant groups are more predictable in the areas with stable hydrological conditions (e.g. no major freshwater discharges and turbulent mixing).  Further work is required for the testing and expansion of suitable values to cover more assessment units within the Baltic Sea.

 

Seasonal phytoplancton Thresholds figure 2.png

Thresholds figure 2. Selection of reference period by calculting 5-years moving averages of standard deviations in yearly total biomass (µg l-1) values. The period with lowest variability is indicated between red bars. Test data are based on the Northern Baltic Proper.

 

 Seasonal phytoplancton Thresholds figure 3.png.jpg

Thresholds figure 3. Selection of reference period by decomposition of multiplicative time series. Observed − raw fluctuations in total biomass; seasonal – seasonality effect; random – irregular component in dataset. Trend shows seasonally adjusted periods of stability (within red bars, possible extension of the period is marked with green bar), where the fluctuations are smoothed. This example is based on the data collected from the Northern Baltic Proper.