Few research topics in watershed management have stimulated more debate in recent years than cumulative watershed effects (CWE) analysis. Though it exhibits great promise as a comprehensive land use planning and management tool, it is still resisted and criticized by many in the resource management field.

Critics point out that there is insufficient data to use CWE analysis as a predictive tool. They argue that there are too many variables, and that current CWE analysis methods are too simplistic.

The debate over CWE analysis has clarified the need for specific research into several key areas. I see a danger in this prolonged debate, however. The danger is that we will become paralyzed by the disagreement over CWE, and that many of us will decide that we must forestall CWE analysis until all issues are resolved through exhaustive scientific research. To draw such a conclusion would be to miss an opportunity to participate in the advancement of our science.

I believe that all watershed managers should join in the development of CWE analysis. Our professional research organizations should gather data not only from the few instrumented watersheds, but also from those of us managing timber, grazing, mining , and urbanization activities throughout the West. One major task of CWE research should be to establish a means of gathering data from both public and private watershed managers, evaluating their findings, and communicating advances and breakthroughs back to the participants.

This seems like a big task. How could we start out? What would be a minimum level of cumulative impacts analysis that would: 1) be beneficial to the land manager at the local level, and; 2) be helpful to the researchers at the state level? Here are my ideas about how to begin the CWE analysis process in a way that will benefit everyone.

Let's start small and keep it simple.

The first essential phase is to choose a way to rank our watersheds according to their natural fragility and place them in classes of low, moderate, or high sensitivity. An intensive field survey is not required. Use resources that are already available, such as a soil survey.

This leads us to the second essential phase. We must create a way to quantify the relative level of disturbance in each of our watersheds. This is an entirely different kind of inventory from phase one. It can be called a land use inventory or perhaps a "Watershed Disturbance History". The key element here is to keep track of the acreage of all past, present, and reasonably-foreseeable future management activities. A file is maintained for each watershed and updated annually as new activities are planned and old ones recover over time. In National Forests, these inventories might best be compiled by district personnel.

The Region-5 US Forest Service CWE methodology recommends using "normalized disturbance coefficients" to help add up the disturbance value of various management activities. The equivalent roaded acre (ERA) is one such standard of measure. It is merely a unit of hydrologic disturbance, presumed to be related to water quality impacts such as erosion, sediment delivery, and downstream sedimentation. If you believe the ERA is not an appropriate unit of measure, you can choose either to use it now and then refine it over time, or to develop a better way to measure watershed disturbance from past, present and future management activities. Just be sure that the method is practical. Keep it simple and consistent.

The product of phase two is similar to that of phase one in that it shows the watershed manager how their watersheds rate in terms of the relative amounts of recent or proposed disturbance. If Brown Creek is more sensitive than Stony Creek and also has more recent disturbances, then perhaps Stony Creek should be considered for the next harvest activities rather than Brown Creek.

To determine if this conclusion is reasonable, we must add phase three. We must walk the main tributaries of each watershed at least once every two years to get a rough assessment of the degree of erosion, mass wasting, and recent sedimentation. A hike down the channel can give a professional hydrologist a good idea of whether a "Threshold of Concern" (TOC) has been exceeded or not. If the channel system is in terrible shape it is probable that the TOC has been exceeded and that adverse CWE's are occurring.

Many people feel that if Stony Creek is in good shape now, it is difficult to define a firm, numerical TOC which will predict when CWE will occur in the future. I believe they are correct. In our first year of CWE analysis, we establish only an approximation of baseline data. Nevertheless, if we continue to perform the three phases of analysis every year or two, we acquire increasingly useful information for assessing CWE risks. Within a few years , we will refine our process to the point that we will be able to perform phase four , estimating the TOC for watersheds that are still in good shape. In truth, we can perform phase four now, but we should state clearly that we are making only a rough estimate in most cases.

In sum, I am asserting that we can and should begin CWE analysis now. My suggestions are based largely on the 1988 Cumulative Off- Site Watershed Effects Analysis methodology of Region-5 of the U.S. Forest Service. I used this methodology on the Eldorado Forest and found it to be workable and useful. I described the detail of this work in "Proceedings of the 1989 AWRA Symposium on Headwaters Hydrology". CWE analysis gives the land manager well-organized information so that he or she will not have to spend days or weeks wading through stand record cards or alder thickets in order to comply with legal mandates to consider CWE. We will learn by doing, and if our results can be incorporated into a statewide CWE research program, we will all benefit from the refinement of the process and the development of new and better analysis techniques.