A Citizen's Guide to Understanding and Monitoring Lakes and Streams

Chapter 4 - From the Field to the Lab

Now that you’ve read about the different parameters and why we monitor them, it’s time for the "hands-on" information needed to collect and analyze the samples. Proper sample collection and analysis are absolutely vital to the success of your monitoring program, whether it’s simple or sophisticated. You just can’t get reliable data with poor technique. This chapter begins by describing what makes good data and explains the language of quality assurance and quality control (QA/QC). Next, it presents basic ground rules for sampling and step-by-step instructions for collection and measurement for each parameter.

What Makes Good Data?

Good data are data you can feel confident about – confident that the measurements made really do reflect the true conditions in the lake or stream, and confident that you have collected enough data over a long enough period to adequately characterize your environment.

Period of Record

The period of record is the length of time over which you collect data. If the data had been collected every 2 weeks for 10 years, you would have a good period of record. Sudden or gradual increases in pollutant levels could be reasonably attributed to some event or change within the system. Conversely, if data had been collected sporadically for 1 year, it would be difficult to say whether changes in pollutant levels signified an unusual occurrence or just normal seasonal change. In fact, an unusually large change measured on one sampling date would likely cause as much in the way of suspicions about equipment or laboratory problems as concerns about water quality. (Of course, if you had a good QA/QC plan those suspicions would be hard to justify.)

A long period of record can sometimes compensate for lower level sampling and analysis techniques. Even if there is a lot of "noise" in your data, water quality trends may be identifiable when you have a long period of record.

Quality of Each Piece of Data

It is not enough to know you used state-of-the-art equipment and analysis techniques. Equipment breaks down and people make mistakes. Checks are needed within your monitoring program to catch potential problems. These checks are referred to as a quality assurance and quality control (QA/QC) plan.

QA/QC is the practice of making sure that collection and analysis techniques provide precise (i.e., repeatable) and accurate (i.e., correct) information. Some QA/QC procedures apply to field operations and many apply to laboratory procedures.

Before you read on, it is important that you understand what QA/QC is all about. Imagine you have used the best possible field collection techniques and then sent your samples to a lab for phosphorus testing. What you don’t know is whether the machine the lab uses to measure phosphorus will be working properly at the time your samples are measured. You also don’t know how well the sample you collected represents average conditions in the lake or stream at the time of collection. By collecting a few extra samples, namely lab replicates and field replicates, you can answer these questions.

A field replicate is used to measure natural or field variability. The water you collect in one spot in a lake may be slightly different from the water a few feet away. If you collect two samples of stream water from the exact same spot by dipping the first bottle and capping it, and then dipping the second bottle, a good deal of water will have flowed past in the short time between dipping the samples. The second sample actually represents an entirely different "slug" of water. The two samples, called field replicates, will often be similar, but not always. Field replicates help you estimate the magnitude of this natural variation.

A lab replicate is used to assess analytical precision – the ability of the equipment, technique, and technician to come up with the exact same value in subsequent measurements on the same sample. For this reason, it is essential that the lab replicates are true replicates of the same sample. For example, if you fill a bucket with stream water, collect two phosphorus samples from the water in the bucket, and have them both analyzed at the same time, the samples are lab replicates. Laboratory replicates are usually submitted in such a way that the lab technician running the analysis does not know that the sample is a replicate. "Blind" replicates eliminate the possibility for bias from the laboratory in reporting their results.

The lab and field replicates are ways of determining how precise your results are. It also is important to know how accurate the measurements are. Accuracy is measured in the lab through the calibration standards. These are samples prepared from distilled-deionized "pure" water that contain a known concentration of a specific substance or will produce a know instrument response.

Other QA/QC terms you may hear are field blanks, lab blanks, and spiked samples, all of which are additional checks on the accuracy and precision of results. Unless you become involved in data analysis, the main QA elements in which you will be interested concern field sampling procedures and field equipment checks and calibrations.

The complexity of the QA/QC plan determines how much confidence you and others will have in your results. Certain objectives require greater accuracy and precision than others. If you were cutting down a dead snag in the middle of the forest, you would decide approximately where the snag should fall and then cut away. If the same snag were in your front lawn, with the house, high-voltage electric lines, and the family car nearby, you would make sure that the tree fell exactly where you wanted it too. You might even hire a professional to cut it down. The same holds for water quality monitoring – there are times when a rough estimate will do and times when you need to be exact.

QA/QC guidelines need to be set for each project regardless of the intended use of the data. If you do not include any QA/QC checks, make a statement to that effect in the monitoring plan. This will ensure that people reviewing the data will know how to categorize the information when comparing it with other data.

Perhaps one of the most notable differences between a "seasoned" water quality specialist and a beginner is the willingness of the former to throw out questionable samples and data. If there is a reason to suspect a sample wasn’t collected or analyzed properly, it is best to discard the data and not let it "pollute" the high-quality data you may already have collected.

Next Section: Ground Rules and Sample Selection.

Return to Table of Contents | Lakes | Streams | From the Field to the Lab | Hydrology

Last updated on April 01, 2008