Environment Canada

Susan Norman
Hydrologist, Lake Tahoe Basin Management Unit

Water Quality monitoring can be an effective way to determine quantitatively the impacts of various management activities on streams. However collecting useful data requires a commitment of resources to collect adequate information on baseline conditions. Because of budget constraints, and the inadequate long-term planning it is often difficult to obtain this baseline data. In these situations managers must carefully evaluate if water quality monitoring will be the most useful and cost-effective tool for measuring impacts. The following is a discussion of some real life examples of the importance of collecting baseline data for effective analysis of water quality data.

The water quality monitoring program at the Lake Tahoe Basin Management Unit (LTBMU) has been in existence since 1980. In any given year approximately 10 to 15 stations are sampled at a frequency of 15 samples a year, and are analyzed for sediment and nutrient content. Samples are collected between late March and early October, and are collected weekly during spring runoff and monthly during baseflow. A minimum of 3 to 5 years of data is collected at each site. Statistical analysis is used to compare data collected before and after management activity, and comparisons between above and below sites, using both parametric and nonparametric tests for significance. Average annual means are also derived to compare to state water quality standards.

Background/Reference Conditions

It is important to collect data from undisturbed sites to determine what background conditions are for the water quality parameters of interest. This information can then be used to determine desired future conditions, and reference variability for specific water quality parameters. Data has been collected at several relatively undisturbed watersheds in the Lake Tahoe Basin by both the USGS and USFS. The period of record at these sites ranges from 5 to 12 years. Recent preliminary analysis of data collected at these "undisturbed" watersheds have resulted in two significant findings.

1) Baseline water quality data can vary widely from watershed to watershed and is probably related to watershed characteristics such as geology, soils, and geomorphology.

2) State standards set for total phosphorus in the Lake Tahoe Basin are frequently exceeded at these sites (50 to 100% of the time, depending on the site.) Standards for other parameters are not often exceeded, suggesting that the standard for total phosphrus may be too low.

Period Of Record

The most effective monitoring design usually entails establishing sites immediately above and below the management activity of concern. It is also important to establish these sites for an adequate period before the management activity begins to determine if any difference observed in parameter values between above and below sites were occuring before the mangement activity began, and to what degree. Water quality data is inherently highly variable and obtaining enough data to determine statistically significant differences may require a long period of record.

The LTBMU recently completed an anlysis of data collected for 5 years above and below a timber salvage sale. The period of record was 5 years, but only 1 year of this data was collected pre-project. Nonparametric statistical comparisons (which determines the number and rank of positive and negative differences between daily values) of above and below data indicated significant increases in nitrogen discharge for two of the years after the timber sale. However parametric anaysis did not show any significant differences in average annual means, and the difference in the average annual means observed in the one year of pre-sale data was similar to the differences observed in the post-sale data. Because only one year of pre-sale data was obtained it is difficult to determine if the timber sale resulted in any incremental increases in nitrogen over what was occurring in the watershed prior to that particular activity.

If water quality monitoring is to be used on your area I recommended you try to design your program considering the following principles: The longer the period of record the better. Try to plan monitoring far enough in advance so the adequate pre-treatment project data can be obtained. Whenever possible, establish sites that will be useful for a long period of time, even if management activity changes. The timber sale described above is over, but sites in the watershed are continuing to be monitored because new sales and watershed restoration projects are being planned in the area.

Try to obtain data from undisturbed sites that are representative of the watersheds on your forest to determine what back ground conditions are for the water quality parameters of concern. Possibly someone else has already done this, and you only need to obtain their data.

However water quality data are obtained, a good dataset can be extremely valuable in quantitatively defining reference variability and desired future conditions for stream water quality, for long term ecosystem planning and management.