Forests and Water: A State of the Art Review for Colorado

• Year: 2003
• Title: Forests and Water: A State-of-the-Art Review for Colorado
• Author: Colorado State University
• Description: Colorado State University 2003
• Publications - Doc Type: Other

to remove much of the temporal variability in runoff.
The problem is that this design is very difficult to apply
in larger watersheds, as it is nearly impossible to find two
large basins where one basin can serve as an untreated
control and the other basin can be subjected to treatment.
An extensive search across several states was needed to
find a 2,200 -acre control basin and an adjacent unhar-
vested basin that could be used to test the effects of forest
management on runoff at an operational scale (Troendle
and Nankervis, 2000). The resulting Coon Creek study
did show that the hydrologic effects of forest manage-
ment on a 6.5 mil basin were directly comparable to the
results from much smaller basins (Troendle et al., 2001).
Results from even larger basins are scarce and not as de-
finitive, as one has to rely on natural disturbances in an
uncontrolled experimental design. Love (1955) claimed
to detect a two -inch increase in streamflow as a result of
extensive beetle kill in the 762 mil White River basin,
but his results were contested by Bue et al. (1955). A
later study used a different, somewhat unconventional
procedure to detect the changes in flow caused by beetle
kill in both the White River and the Yampa River, and
this claimed that water yields increased by about 10%
relative to the Elk River basin ( Bethlahmy, 1974). The
increases claimed by Love (1955) and Bethlahmy (1974)
are consistent with the results obtained on much smaller
basins, and support the view that the changes in runoff
measured on small experimental basins can be scaled up
to much larger basins.
The other factor that greatly affects the detectability of
change is the closeness of the relationship between the
control and treated basins, or the consistency of the data
before and after treatment if observations are limited to a
single basin (Loftis et al., 2001). Typically there is a rela-
tively high correlation between paired basins for annual
water yields, and a strong correlation was essential to
detecting a change in annual water yields at Coon Creek
because there were only five years of calibration data and
five years of post - treatment data (Troendle et al., 2001).
The correlation in the size of peak flows between paired
basins is usually not as strong, and this makes it more
difficult to detect the effect of a given treatment on the
size of peak flows. In the case of the Coon Creek experi-
ment, it was not possible to detect a change in the size of
the average annual maximum peak flow, even though the
snowpack data and results from other studies indicated
that there probably was an increase (Section 2.2.5). The
detection of a significant change over time at a single
stream gauging station is much more difficult because
there is almost always more unexplained variability rela-
tive to a paired- watershed design (Loftis et al., 2001).
Taken together, these issues mean that one gener-
ally should not expect to measure significant changes
in runoff in larger watersheds as a result of typical for-
est management activities. Field research, hydrologic
theory, and modeling studies all indicate that a change
in forest cover will affect streamflows, particularly at
higher elevations where there is more precipitation and
hence a greater potential to alter interception and tran-
spiration. The resulting changes in runoff will be trans-
mitted downstream, but with increasing drainage area
the changes from a given set of activities will become
proportionally smaller. The opportunity to generate a
detectable change in runoff in larger basins also may be
limited by the amount of area that can be treated within
a relatively short time period and the intensity of the ap-
plied treatments (e.g., thinning as opposed to clearcut-
ting) (Hibbert, 1979; Ponce and Meiman, 1983). High -
severity fires are the primary exception to these general
principles, as both the change in runoff and the size of
the affected area can be relatively large.
These limitations in detecting change do not mean that
changes in runoff won't occur in larger basins, or that
there is no opportunity to alter runoff through forest
management. Changes in runoff will occur with changes
in the forest canopy, and a small percentage change im-
posed on a large area can equate to a large amount of
water in absolute terms. The difficulty is that we cannot
expect to measure these changes, so the effect of changes
in forest management on runoff at lager scales will usu-
ally have to be quantified through hydrologic models,
and presumed to be present even though the changes
may not be statistically detectable at existing gauging
stations.
In conclusion, this chapter has summarized the type
and magnitude of changes in runoff that can occur as a
result of management actions and wildfires in forested
areas. The process -based understanding provided in this
chapter is necessary for predicting future changes, as
the hydrologic effect of a given action or disturbance
depends on a large number of site - specific factors and
will vary with changing climatic and site conditions. The
next chapter reviews the effects of forest management, or
the absence of forest management, on water quality. This
is followed by an assessment of the historic changes in
forest vegetation in Colorado and the estimated impact
of the changes in vegetation on runoff.
20