Interlaboratory Comparison of Nutrient
Seawater Samples in the German Bight
Sieglinde Weigelt-Krenz*; Rita Kramer*, Thomas Raabe**
* Bundesamt für Seeschifffahrt und Hydrographie (Federal Maritime and Hydrographie Agency); Wüstland 2; 22589 Hamburg; Germany
** AquaEcology; Marie-Curie-Straße 1; 26129 Oldenburg, Germany
BUNDESAMT FÜR
SEESCHIFFFAHRT
UND
HYDROGRAPHIE
Abstract
In 2007 and 2008, an intercomparison exercise for nutrient analysis of sea
water samples was carried out by six independent German laboratories. Main
objective was to compare different conservation techniques (deep freezing or
poisoning with mercury chloride, HgCI 2 ) as well as the different analytical
methods used by the investigators. The samples were analysed for dissolved
nitrate+nitrite, nitrite, ortho-phosphate, silicate and ammonium.
The sea water samples were taken in the German Bight during two monitoring
cruises in March 2007 and March 2008. The samples were divided into 250-ml-
aliquots, of which one was directly measured onboard the research vessel with
SFA (segmented flow analysis). The other aliquots were simultaneously
distributed among the participating laboratories, half of them being conserved
by deep-freezing at -18 °C, the other half being poisoned by HgCI 2 (stored at 4-
S'C), respectively. All samples were processed within a fixed time-frame.
The analysed data were compared and evaluated according to Gofino et al.
(2004). One laboratory failed at the analysis of phosphate and ammonium.
Apart from that the interlaboratory variability was generally less than 10 % for
nitrate+nitrite, nitrite, silicate and phosphate. For ammonium the variability
ranged between 0.5 and 45 %. This clearly revealed methodic problems for the
storage and/or analysis of ammonium samples.
Results and statistical evaluation
The results [pmol/L] of six independant laboratories (A-F) are exemplarily summed
up for nitrite, slilicate and ammonium at seven stations in the German Bight (S =
28,3-35,0). The differences are evaluated statistically according to Cofino et al. (2004).
z-score
-3,00 -2,00 -1,00 0,00 1,00 2,00
stations
Distribution pattern of DIN in the German Bight
0 50
45
H40
® 35
30
25
20
15
10
5
2.5
m
Winter
average
(2004-2008)
of DIN
(dissolved
inorganic
nitrogen
compounds)
in the bottom
water of the
German
Bight.
Z-score (nitrite) at station 3 referring to the assigned value (AV); AV = 0,88 pmol/L
z-score
-3,00 -2,00 -1,00 0,00 1,00 2,00
1 2 3 4 5 6 7
stations
Water sampling and sample processing
7 samples
from the
German Bight
monitoring cruises
March 2007/2008
SFA
filtration
0.4um
Polycarbonate filter
deep
freezing
at -18°C
*
poisoning
with
HgCI 2
(4 - 8°C)
onboard
analysis
simultaneous analysis in July 2007/2008
Z-score (silicate) at station 1 referring to the assigned value (AV); AV= 31,40 pmol/L
z-score
-3,00 -2,00 -1,00 0,00 1,00 2,00
Z-score (ammonium) at station 1 referring to the assigned value (AV); AV= 3,68 pmol/L
(without laboratory C)
Conclusions
• There are comparable results for onboard measurements as well as for
analyses of deep frozen and HgCI 2 preserved samples.
• The z-scores for almost all institutes are between -2 and +2.
• The comparability of North Sea water analysis methods for dissolved
nutrients is assured within the participating laboratories.
The results confirm that a common standard for nutrient analysis exists among
German laboratories and thus, German monitoring tasks are fulfilled in an
effectual way. However, some additional work on the storage and analysis of
ammonium has to be done.
One aim for the near future will be to extend this intercomparison project to
the international community in order to ensure comparability of nutrient
results within the whole OSPAR region.
Literature: d.e. Wells, W.R Cofino and J.A. Scurfield (2004). The Application of the Cofino Model to evaluate Laboratory
Performance Study Data using the Bandwith Estimator. Fisheries Research Services Collaborative Report No 04/04
www.bsh.de
