Page 171 - Freshwater-Biology-and-Ecology-Handbook

Convert SD to percentile determining class interval
y = 68.423x
12 Percentile defining
2
R = 0.9465
percentile defining G/M boundary 0 0 0.05 0.1 0.15 0.2 Linear (percentile
10
8
G/M boundary
6
4
defining G/M
boundary)
2

SD of EQls at RIVPACS’ reference sites

Figure 3.21
Trend line for determining the whole number percentile for determining status class intervals for
an index from the standard deviation of O/E values at RIVPACS III+ GB and NI reference sites

The mathematical calculation of initial class boundaries following 5M principles ensured that the WFD
class boundaries were equivalent to the GQA (General Quality Assessment) class boundaries that
preceded them. The environment protection agencies, through UK TAG, amended some of the initial
boundaries for NTaxa to reduce the number of sites that this index downgraded in Scotland. This
process of initial mathematical calculation followed by amendment to reflect operational experience
or requirements has been used to set the UK’s official river invertebrate classification boundaries since
1990. The move from BMWP Ntaxa to WHPT Ntaxa caused a slight tightening of standards in Scotland
because SEPA continued to apply bias values for BMWP NTaxa to WHPT NTaxa.
These class boundaries expressed as observed/expected (O/E) ratios are also known as EQIs
(Environmental Quality Indices). For WFD, we must express them as Environmental Quality ratios
(EQRs), which is O/reference-value. The predicted values need to be adjusted (Section 3.3.1) and
converted to reference values (Section 3.3.2).

value of metric observed in sample(s)
EQR =
value of metric at reference condition

A WFD reference value is the value of an index that we would expect at that type of site in its near
natural, but not pristine, reference state (see Chapter 1 Section 5.7 Reference Conditions).
Reference state is somewhere in High status. Reference values are usually type-specific and based
on the average value of an index calculated from WFD reference sites belonging to the same type. In
the UK, we predict site-specific reference values based on weighted averages of RIVPACS reference
sites in different end groups (see Section 3.2). RIVPACS end groups are essentially abiotic river
types defined by the biological communities that they support, and the weighted average reflects the
fact that the types are not discrete.

Before these class boundaries could become official, they had to be agreed by all UK environmental
protection agencies. This was done through WFD UK TAG. It was at this stage that minor changes
were made to the boundaries for WHPT NTaxa, based on practical experience. The boundaries then
had to be intercalibrated, to ensure that the High and Good status classes covered the same range of
quality as in other Member States (see Chapter 1 Section 5.8 Intercalibration).

Freshwater Biology and Ecology Handbook | 171
–