Steelhead 101: Using snorkel surveys to estimate adult steelhead escapement

In Alaska, California, Idaho, Oregon, Science Friday, Washington by Nick Chambers

Another week, another post on how scientists and fisheries managers measure steelhead escapement. Last week, we described redd counts and why they are an important tool. This week, we review snorkel surveys.


Snorkel surveys entail divers swimming in the river and visually counting adult steelhead. Like redd counts, snorkel surveys do not cover an entire river, but rather break out the river into sections and sample some of those sections on a bi-weekly to monthly basis. Good water clarity and visibility is vital for this type of study, because it can be very difficult to see fish underwater even if visibility is only slightly diminished.


Thus snorkel surveys are most often conducted to count summer runs in late summer through early fall, when stream flows are relatively low and water clarity is high. Conversely, snorkel surveys are less common for winter runs because stream flows are higher and visibility is more variable and often worse. The late summer and early fall timing is important because that ensures as many fish as possible have entered the system and are present and able to be counted.




Snorkel surveys are generally conducted in one of two ways. In some cases managers will capture adult steelhead at collection facilities such as weirs or fish traps, and mark those fish with a visible tag that is easily seen by a surveyor. Those fish are released back into the system, and later the surveyors will go back and snorkel. They will count both marked (with the tag) and unmarked fish, and then estimate total number of fish based on the ratio of marked:unmarked steelhead. This is referred to as mark-resight, and is the most accurate method for enumerating steelhead with snorkel surveys.


In cases where capturing and marking fish is not possible, surveyors will simply count total numbers of observed steelhead. The limitation of this approach is you don’t have a way to estimate how many fish a surveyor missed because there is no data on the number of tagged fish.  This typically means there will be a greater level of uncertainty in the data, though this can be mitigated to some degree by conducting more frequent surveys.


While snorkel surveys are not as commonly used as redd counts, they do have benefits. For example, you actually get to see the fish. That allows surveyors to potentially identify sex and whether or not the fish are hatchery or wild. In addition, unlike redd counts where there is uncertainty in converting redds to fish numbers, snorkel surveys cut to the chase and actually count fish.


The primary limitation with snorkel surveys is that they can be more labor intensive than redd counts, and hence it is more difficult to cover larger areas. Redd counts, especially if conducted by helicopter, can cover large areas of stream in a short period, which is not possible with snorkel surveys. In some cases the challenge of spatial coverage during snorkel surveying has been addressed by identifying over-summering pools that steelhead rely on and focusing snorkel survey efforts on those places.



As scientific tools both snorkel surveys and redd counts have strengths and weaknesses, in addition to uncertainty. While snorkel surveys are less common than redd counts, in some watersheds managers use them on a regular basis. For instance, the Washington Department of Fish and Wildlife conducts regular snorkel surveys for summer steelhead in places like the Washougal and Wind Rivers in SW Washington. Data derived from this methodology has been collected since the mid-1980s and has allowed the Department to effectively track population trends in those rivers. Kudos to the Department for maintaining such an intensive effort.


At the end of the day, neither method is necessarily better than the other. In a perfect world fisheries managers would employ both redd counts and snorkel surveys, which would complement each other and deliver more reliable fish count estimates. But in reality agencies and managers typically do not have enough money to employ all methods, so they make the best choice possible based on the tradeoffs of each method in a given watershed.