Monitoring Juvenile Trout Numbers in Fish Creek (Harriman East)

Photo of five people electrofishing in Fish Creek.
Photo of three people electrofishing in Fish Creek.
Photo of juvenile rainbow trout being measured.

Personnel from the Henry’s Fork Foundation (HFF), U.S. Forest Service (USFS), and Brigham Young University-Idaho (BYUI) conducted an estimate of juvenile trout abundance in Fish Creek on April 15-16, as part of a long-term restoration and monitoring effort. How did our 2015 numbers compare with those in 2011, prior to major restoration of the creek?


Before answering this question, let me provide a little background. First, we are talking about the Fish Creek that flows out of Fish Pond, across the lower part of Harriman East, and into the Henry’s Fork near Wood Road 16. Fish Creek is one of only a few small tributaries to the Henry’s Fork between the Buffalo River and Riverside Campground.

Research conducted by HFF since 2008 has shown that young rainbow trout entering their first winter (what we call “age-0 fish”), migrate into these tributaries from late November through January, spend the winter there, and migrate back out during late spring and early summer. Under our collaborative strategy for managing water in Island Park Reservoir, the flow out of the reservoir actually increases during this time frame, so you would think that habitat availability in the river would also increase. That is true of Box Canyon, where rocks and logs along the banks provide the majority of habitat for young trout and higher water depth creates more habitat, but it is not true of the Harriman reach, where aquatic plants (“macrophytes”) provide the majority of the cover. Based on recent results from Zach Kuzniar’s study of adult trout habitat in the Ranch and from our water-quality monitoring network, we now know that macrophyte abundance, not river flow, is the most important determinant of river depth and habitat availability in the Ranch. So, in late fall/early winter, as the macrophytes start to die and are uprooted and/or eaten by waterfowl, river depth and habitat availability in the Harriman reach decrease, despite increased river flow. Thus, young fish move out of the river and into the small tributaries to this reach, including Fish Creek. When macrophytes start growing again in June, the young fish move back out of the tributaries and into the main river.

To help improve the quantity and quality of over-winter habitat in Fish Creek, the U.S. Forest Service led a major restoration effort on Fish Creek between Fish Pond dam and the Henry’s Fork confluence. The restoration involved narrowing the channel, planting willows along the banks, and restoring the stream’s natural sinuosity. By restoring sinuosity, the length of stream between the dam and the Henry’s Fork increased from about 2.2 km (1.4 miles) to 2.6 km (1.6 miles). The restoration work was done in the fall of 2012.

HFF, USFS and other partners conducted pre-restoration estimates of age-0 rainbow trout abundance in Fish Creek in November 2010, April 2011, and July 2011. To help assess the effectiveness of restoration in providing habitat for age-0 rainbow trout, we repeated these estimates in November 2014 and April 2015 and plan to repeat the July estimate this year as well. The 2014-2015 field work is being conducted by HFF staff, Caribou-Targhee National Forest fisheries biologist Lee Mabey and other USFS staff, and BYUI fisheries professor Dr. Eric Billman and some of his students.


During both 2010-2011 and 2014-2015, we used a method called “removal estimation” to estimate the number of age-0 rainbow trout in the stream. However, the 2010-2011 estimates were based on a sample of only three 100-meter reaches in the stream, and the 2014-2015 estimates use a sample of nine 100-meter reaches, in order to obtain higher precision. For details on sampling methods and how the mathematics work out, please see this linked document.


The results are summarized in the figure and table below, which give the stream-wide estimate of the number of age-0 rainbow trout, along with what we call the 95% confidence interval. This interval represents the uncertainty in the abundance estimate; the wider the interval, the less certain we are in our estimate. The statistical meaning of a 95% confidence interval is that if we performed the same estimation procedure repeatedly a very large number of times, 95% of the confidence intervals resulting from this estimation procedure will contain the true value of fish abundance. The document linked above provides more information on sampling and estimation.

Graph showing age-0 trout abundance versus month and year.

Date Estimated number of fish Lower 95% confidence bound Upper 95% confidence bound
Total number of age-0 rainbow trout in Fish Creek.
Nov 2010 1,596 211 7,441
Apr 2011 3,385 410 16,345
Nov 2014 190 80 300
Apr 2015 377 123 662

From either the graph or table, you can see the pattern described above, in which more fish are present in the stream at the end of winter than at the beginning of winter. You can also see that our abundance estimates are quite a bit lower in 2014-2015 than in 2010-2011.  However, you can also see that in 2014-2015, the uncertainty in our estimates is much lower than it was in 2010-2011. This is because we used a much larger sample size in 2014-2015. Again, details are given in the linked document. Notice that the confidence intervals around the 2010-2011 population estimates overlap with those of the 2014-2015 estimates, which indicates that we do not have enough statistical evidence to conclude that the true abundance in 2014-2015 is really lower than it was in 2010-2011.

Closer inspection of the raw data reveals that the high abundance estimates in 2010-2011 resulted from one sample reach located immediately downstream of the Fish Pond Dam, where a large number of fish were congregated.  For example, the per-reach estimates from the three reaches sampled in November 2010 were 8, 4, and 202 fish respectively. You can see why a high abundance in this one reach could greatly skew the extrapolation of reach-level estimates to the whole stream. It is very likely that the 2010-2011 estimates are artificially high because of the small sample size. Use of nine sample reaches in 2014-2015 greatly reduces the potential for one reach to skew the stream-wide calculation. Not only that, but we found that in 2014-2015, the reach-level abundance estimates were relatively constant across reaches, perhaps indicating that habitat quality is more uniform throughout the creek as a result of the restoration.

Implications for the Henry’s Fork Fishery

Based on our most recent estimates, it appears that around 200-500 age-0 RBT are wintering in Fish Creek and migrate back to the Henry’s Fork as age-1 fish in the late spring and early summer. At typical annual survival rates for fish in the Henry’s Fork, this would produce 80-200 fish entering the population of catchable-sized trout (nine inches or longer) the following fishing season. If, as adults, these fish all reside in the vicinity of Harriman East and Pinehaven, these 80-200 individuals could be an important component of the fishery there. On the other hand, if the Harriman East/Pinehaven fishery is supported by the larger population of fish that migrate throughout the river between Island Park Dam and Hatchery Ford, then the potential 80-250 fish added to the population from Fish Creek contributes only a few percent the total population, given that tens of thousands of young fish survive their first winter in Box Canyon and the Buffalo River. 

Regardless, HFF and its partners, including USFS and Idaho Department of Fish and Game, have devoted a large amount of time, effort, and money into increasing winter survival of young fish in all three of these locations—through restoration of habitat in Fish Creek, improvements in flow management at Island Park Dam, and restoration of fish passage to the Buffalo River. We will continue to monitor the success of all three of these projects and quantify their respective contributions to the Henry’s Fork fishery.