Home About Browse Search
Svenska


Northern Rivers GIG Phytobenthos Intercalibration Exercise

Kahlert, Maria and Gonczi, Mikaela and Mykrä, Heikki and Vuori, Kari-Matti and Jarlman, Amelie and Kennedy, Bryan and Bennett, Cathy and Miettinen, Juha and Kelly, Martyn (2007). Northern Rivers GIG Phytobenthos Intercalibration Exercise. ..

[img]
Preview
PDF
380kB

Abstract

General issues associated with phytobenthos intercalibration exercises are addressed in the report on the CB GIG intercalibration exercise. The conclusions and recommendations listed in that report are all equally valid for the N GIG exercise. This section highlights a few points that are unique to the N GIG exercise. The CB GIG exercise involved 12 Member States; whilst the N GIG exercise is much smaller, with just four participants. An important implication is that the exercise has lower statistical power and it is not always clear if those MS that fall outside the ‘acceptable band’ do so because there are issues that those MS need to address or because the ‘acceptable band’ is itself based on a small (and potentially atypical sample). On the other hand, however, the ‘acceptable band’ should not be equated with ‘best practice’. MS that comply with the minimum requirements of the exercise are included in the acceptable band and the position of this band, therefore, reflects the consensus of those. This must affect how results from N GIG and other smaller intercalibration exercises are judged. In particular, a ‘Type 1 error’ (i.e. erroneous rejection of the [null] hypothesis that boundaries are the same) may lead to the conclusion that a MS needs to adjust boundaries when, in fact, the median value of the ICM (which anchors the acceptable band) is unlikely to be stable with such a small sample size. The approach adopted here was, therefore, to perform a suite of tests using different permutations of the statistical criteria and to make final judgements about the need (or otherwise) to adjust boundaries based on the weight of evidence. Whilst the CB GIG exercise evaluated two versions of the ICM (one based on the mean of component metrics, the other based on the minimum), the N GIG exercise used both versions. TISI-min favoured IE and UK, both of whose national metric was the TDI, which correlates more strongly with the nutrient-sensitive TI, whilst TISI-mean favoured FI and SE whose national metric was the IPS, which correlated more strongly with the SI. Whilst TISI-mean is not biased by a low value of one or other metric, TISI-min better embodies the ‘one out, all out’ principle used when comparing biological elements as part of status assessments. Three of the four MS taking part in this exercise were also involved in the CB GIG exercise. Boundaries calculated in this exercise are broadly consistent between the two exercises. For H/G, IE, SE and UK were all inside the acceptable band for the CB GIG exercise whilst, for N GIG, UK were inside whilst SE was above the acceptable band for TISI-min but inside for TISI-mean and IE was marginally below for TISI-mean. For G/M, UK and SE were inside the acceptable band whilst IE was above. For the N GIG exercise, IE and UK were inside the acceptable band on all occasions whilst SE was again above the acceptable band when TISI-min was used. In the case of IE, the relatively small size of the dataset plus the low number of poor quality sites may be responsible for the differences in regression equations. Whilst SE were above the acceptable band on two out of four occasions for each of H/G and G/M comparisons, it is only those MS that fall below the acceptable band that need to consider harmonisation. In this exercise, both IE and FI fell below the acceptable band on one out of four occasions, both were only marginally below the acceptable band on these occasions and we believe that there is no case for either MS to adjust their boundaries

Authors/Creators:Kahlert, Maria and Gonczi, Mikaela and Mykrä, Heikki and Vuori, Kari-Matti and Jarlman, Amelie and Kennedy, Bryan and Bennett, Cathy and Miettinen, Juha and Kelly, Martyn
Title:Northern Rivers GIG Phytobenthos Intercalibration Exercise
Year of publishing :2007
Volume:June 2007
Page range:1-31
Number of Pages:31
Place of Publication:.
Publisher:.
Associated Programs and Other Stakeholders:SLU - Environmental assessment > Programme Lakes and watercourses
Language:English
Publication Type:Other
Article category:Popular
Version:Published version
Full Text Status:Public
Agris subject categories.:M Aquatic sciences and fisheries > M01 Fisheries and aquaculture - General aspects
Subjects:(A) Swedish standard research categories 2011 > 4 Agricultural Sciences > 405 Other Agricultural Sciences > Environmental Sciences related to Agriculture and Land-use
(A) Swedish standard research categories 2011 > 4 Agricultural Sciences > 401 Agricultural, Forestry and Fisheries > Fish and Aquacultural Science
Keywords:phytobenthos, rivers
URN:NBN:urn:nbn:se:slu:epsilon-e-2692
Permanent URL:
http://urn.kb.se/resolve?urn=urn:nbn:se:slu:epsilon-e-2692
ID Code:12244
Faculty:NL - Faculty of Natural Resources and Agricultural Sciences (until 2013)
Department:(NL, NJ) > Dept. of Environmental Assessment
Deposited By: SLUpub Connector
Deposited On:09 Jun 2015 13:19
Metadata Last Modified:08 Mar 2016 11:49

Repository Staff Only: item control page

Downloads

Downloads per year (since September 2012)

View more statistics

Downloads
Hits