eLearning

resampling study included multiple soil samples/
plot (which were composited), multiple samplings
through time, replication of plots, and relatively
simple forest structure that allowed the change in N
content of the vegetation over time to be estimated
with precision. The rate of N accretion is very con-
sistent with expectations of atmospheric N deposi-
tion in this area (5-10 kg N ha-1
y-1
), and the lack of
substantial leaching losses measured in the 1990s.
No occult explanations were necessary to balance
the N budget of the Calhoun plantations, even if the
true N accretion rate was at the upper end of the
95% confidence interval.
Watersheds Near Walker Branch
The N content of vegetation and soils have been
resampled over an 11-year period at approximately
3 km from theWalker BranchWatershed in eastern
Tennessee, USA (Trettin and others 2000). Eight
replicate plots were chosen to represent the variety
of vegetation and soils in a 98-ha watershed.
Within each plot, soils were sampled at six random
locations in 1982 and in 1993. The rate of change in
N content on a within-plot basis ranged from –120
to 143 kg N ha-1
y-1
, with an overall average across
the watershed of –35 kg N ha-1
y-1
(P 5 0.15) for
the aboveground vegetation, O-horizon, and min-
eral soil to 60 cm. The 95% confidence interval
around this mean rate of change in N was –76 kg N
ha-1
y-1
to 16kgNha-1
y-1
. The careful quality
assurance procedures of this study, and the inten-
sive replication of the sampling, warrant high con-
fidence in the estimate of change in forest N con-
tent.
Postharvesting Watersheds Near Walker
Branch
Johnson and Todd (1998) measured 15-year
changes in total ecosystem N content after harvest-
ing in a mixed oak forest approximately 3 km from
the Walker Branch Watersheds, near Oak Ridge,
Tennessee, USA. Quality assurance also received a
great deal of attention in this study, including both
field sampling protocols (including the same scien-
tists performing the sampling at both times) and
archived samples. Before forest harvest, soils were
sampled from three soil cores in each of four 10 x
10–m2
subplots/stand (or two per stand in the ref-
erence stand). The N content of vegetation also was
carefully estimated. They found net increases in
forest N content of 825 and 1320 kg N/ha in the two
harvested watersheds. The combustion-furnace
method of N analysis in the later samples recovered
4% more N than the Kjeldahl method used for the
earlier samples (Johnson and Todd 1998). Includ-
ing this correction (which the authors did not in-
clude), reduced the estimate of N increase in the
mineral soil by approximately 120 kg N/ha and the
overall forest N increase to approximately 800 and
1200 kg N/ha, for annual changes of 53 and 80 kg
Nha-1
y-1
. The rate of N deposition in these forests
was expected to be on the order of 10 kg N ha-1
y-1
,
leaving an occult N input estimate of 43 to 70 kg N
ha-1
y-1
. Statistical comparisons were made only on
the basis of individual horizons (not on the whole
soil or whole forest). In one harvest unit, the
changes in N were significant in all three horizons
(A, E, and Bt). In the other unit, changes were
significant only in the E and Bt horizons. The au-
thors remain skeptical of the apparent occult N
input but are confident in the overall quality assur-
ance of the project (D. Johnson, personal commu-
nication). In addition, the authors are at a loss to
explain the large net N accretion in the harvested
watersheds compared with the Walker Branch con-
trol watershed described above (average loss of -35
kgNha-1
y-1
, not significant).
Coweeta Watershed #2
Knoepp and Swank (1997) summarized 16 years of
repeated soil sampling in an intensively studied
watershed at the Coweeta Hydrologic Laboratory in
western North Carolina, USA. Although this study
included other watersheds, we focus only on wa-
tershed #2 because the sampling depths in the other
unmanipulated watersheds changed over time
(based on different interpretations of boundaries
between horizons), so the observed N declines
might have resulted from changes in depths sam-