1. TITLE SLIDE, AUTHORS
2. PARTICIPANTS.

Work was done by the WDNR and USGS as a pilot study

funded by three agencies as part of a larger effort by the

Intergovernmental Task Force on Monitoring Water Quality

3. OBJECTIVE

The primary objective was to evaluate differences in water

quality monitoring results caused by differences in sample

collection methods

4. OTHER COMPARABILITY ISSUES

• Differences in sampling method variability
• Differences in sample processing and preservation, and
• laboratory differences

1. COMPARISON OF TWO SAMPLE-COLLECTION METHODS

Two water-quality sampling methods were evaluated

• Flow-integrated (EWI)—USGS programs (for streams)
• Grab(single or composite)—WDNR (for streams)
• Lake sampling methods were similar

1. SAMPLING TEAMS

Sampling was done by three teams representing three monitoring programs:

NAWQA, NASQAN, WDNR Ambient Monitoring Program

2. SAMPLING SITES/CONDITIONS

• 3 Streams – representative of the variety of conditions in Wisconsin and producing a range of concentrations (Milwaukee River, Manitowoc River, and Wolf River—New London)

• two base flow samples and two high-flow events
• three consecutive rounds of concurrent samples at base flow to evaluate variability within each team

• 1 Lake – sampled two times (Little Green Lake)
• The study was designed to evaluate sample-collection methods

1. CONSTITUENTS

For streams, 4 constituents were selected as surrogates to represent the behavior of broader groups

Streams Lakes

Diss. Cl (similar methods) Total P

Total P (different digestion for streams,

similar to lakes) Diss. Ortho P

Diss. Ortho P (similar method) Chlorophyll a

Susp. Sed./TSS

Again, the intent was to evaluate collection methods rather than water-quality conditions at the sites

2. LABORATORIES

WSLH

USGS NWQL, Iowa District Sediment Lab

3. SAMPLE SPLITTING SCHEME—In order to do the desired tests, the following was used:

• Concurrent samples collected by the 3 sampling teams
• Each sample was split between labs to evaluate the effects of collection method independently using the results from each lab.
• Each lab sample was further split into triplicates to evaluate laboratory precision.
• An additional sample was split to evaluate field v. lab filtration.
• Results in a total of 63 samples per site.

1. Decaport cone splitter was used to split samples
2. These are the bottle requirements for one round of sampling at the 3 stream sites. With the sediment samples this came to about 200 bottles.
3. SAMPLING METHODS

USGS uses cross-sectionally integrated, equal width increment/equal transit rate, flow weighted sampling with an isokinetic sampler (DH-81, D-77). WDNR uses grab sampling (dip sample or Van-Dorn sampler).

13b. USGS field filters samples, DNR samples are filtered at lab.

4. RESULTS—COMPARISON AMONG MONITORING PROGRAMS AND SAMPLE COLLECTION METHODS

15.

1. No signif. differences among programs for TP or diss. Cl
2. Significant differences occurred for:

- DOP, for lake and low-flow samples analyzed at NWQL (USGS>WDNR)

(2.9 ug/L difference was smaller than the st’d dev. Of 3.6 ug/L among lab replicates)

- TSS/SS – not comparable methods, separate analyses were done for each lab.

Flow-integrated samples> grab samples for high flows.

Pattern was similar at low flow but not significant.

- Chlorophyll a – only for samples analyzed at WSLH (WDNR> USGS)

Diff=4.2 ug/L

16. TSS/SS concentrations. USGS flow-integrated> DNR grab for each lab.

SS>TSS

17. VARIABILITY WITHIN MONITORING PROGRAMS

• There were no significant differences among consecutive samples for TP, DOP, Cl collected during low-flow conditions.
• Sampling results within teams were therefore largely repeatable.

18. DIFFERENCES BETWEEN SAMPLES FILTERED IN THE FIELD AND IN THE LAB (for DOP and chlor. a)

- Highly signif. difference for chlor. a (lab filtered>field filtered) 3.3 ug/L difference

- No signif. difference for DOP

19. COMPARISON BETWEEN LABORATORIES

Highly significant differences were found for TP, DOP, Cl. (WSLH>NWQL)

Although lab methods are different,

Susp Sed>TSS

Chlor. a (WSLH)> chlor. a (NWQL)

Trichromatic fluorometric

Spectrophotometer

20. DIFFERENCES BETWEEN LABORATORIES

Observed max Mean diff.(mg/L)

Total P 0.30 0.025

Diss. Ortho P 0.14 0.005

Diss. Cl 60 1.0

Differences varied among dates or flow, so can not use a simple correction factor.

Differences are combined effects of lab performance and sample processing and handling-(relative differences).

Differences are greater than those caused by sample-collection method.

21. CONCLUSIONS

• Flow-integrated versus grab-sample collection methods affect monitoring results for some constituents (TSS/SS at high flow, chlor. a, DOP not expected)
• Laboratory differences were highly significant even for comparable analytical methods (TP, DOP, Cl)
• Differences in results between laboratories and sampling methods may limit the combining of some data for certain data analyses, such as for trends. Great care must be taken to know the data when exchanging data.