Water Quality and Exurbanization in Southern Appalachian Streams - River Conservation and Management - page 96

2008) or with significant urban areas within the
catchments (Sliva and Williams, 2001; Groffman
et al ., 2004). In agricultural and residential areas,
input of chemicals as fertilizer is often a major cause
of high solute levels in streams, and in urban areas
point sources may contribute to high chemical
loads. In many areas, it is difficult to separate
riparian and whole catchment effects because the
riparian corridor generally reflects the catchment
land-cover (Johnson et al ., 1997; Dodds and Oakes,
2008).
The Upper Little Tennessee River Basin differs
in several ways from basins that have been the
focus of previous studies. Except in a few areas (e.g.
Rabbit Creek), there is relatively little agricultural
fertilizer application, although fertilization of home
lawns is probably common. Also, there are no
intentional sewage inputs in any of the catchments
analysed (though there are treated sewage inputs
to the Little Tennessee River itself). Finally, in most
catchments (other than those completely forested),
there is a large difference between valley and
mountainside land cover. When land cover in the
whole catchment was compared with land cover in
just the 200 m riparian corridor along streams, they
were well correlated (riparian agricultural land-
cover and catchment agricultural land-cover, r =
0.90; riparian developed land-cover and catchment
developed land-cover, r

riparian corridor was slightly better correlated with
residential and commercial land-use (r = 0.79, land
cover log transformed) than was whole catchment
developed land-cover (r = 0.78, Figure 8.3).
These relationships are consistent with the
observation that land use is a better predictor
of sediment-associated variables (turbidity, organic
matter in particles, total dissolved phosphorus)
because sediment travels relatively short distances
and its presence during baseflow conditions
generally reflects near-stream or instream activity.
Had the streams been sampled during periods of
elevated discharge, sediment concentrations may
have reflected much more extensive areas of
disturbance. Nitrate and chemicals contributing
to specific conductance are more mobile than
sediment and come from areas throughout the
catchment (Strayer et al ., 2003) and thus are better,
or just as well-predicted, by catchment-level land-
cover data.
High variability in nitrate concentration within
the Upper Little Tennessee River Basin was
observed and found to be closely correlated with
catchment land-cover, as has been shown in
many previous studies (Herlihy et al ., 1998). Most
of the basin was logged during the period of
intense industrial logging in the early 20th century
(Mastran and Lowerre, 1983). Areas that have
not been subsequently disturbed are now very
retentive of nitrogen (Swank and Vose, 1997).
However, areas with more recent disturbances
are approaching or are in the initial stages of
nitrogen saturation (Swank and Vose, 1997), when
the capacity of catchment vegetation to retain
nitrogen has been exceeded (Aber et al ., 1989). This
study showed a good correlation between nitrate
concentration and the extent of mountainside
development. Even when land development occurs
in relatively small areas, high in the catchment
away from the riparian area and the stream
channel, and with little change to forest land-cover,
significant increases in stream nitrate concentration
are seen (Figure 8.9).
Land use refers to human activities within a
catchment and land-use change is the proximate
factor driving land-cover change (Osborne and
Wiley, 1988; Turner and Meyer, 1994), but there

0.99), but the percentage
of agricultural and developed land-cover were both
higher in the riparian corridor (11% versus 6% for
agriculture, 15% versus 11% for developed).
In the mountainous Upper Little Tennessee
River Basin, development and agriculture
have traditionally been in the valleys, and the
mountainsides have remained forested even
though they have been repeatedly logged. Thus
land-use change has been primarily a reflection
of changes within the valley, along streams. In
general, correlations of land use with riparian
corridor land-cover were slightly greater than
with catchment land-cover. For example, at the
catchment level, agricultural land-use and land-
cover were significantly correlated (Figure 8.3,
r = 0.37), but when just riparian corridor land-
cover was used, the correlation improved slightly
(r = 0.40). Similarly, developed land-cover in the

=