eLearning

y-1
does not suggest any occult N input, but the
strength of this unreplicated chronosequence (with
just one soil horizon) is weak.
In Puerto Rico, Garcia-Montiel (1996) examined
N accretion in a two replicate, four-stage sequences
of forest development after abandonment of pas-
tures. Each site was sampled with 18 soil cores to a
depth of 15 cm. The average range of N accretion
over 601 years was 0.3 kg N ha-1
y-1
, which was
not significant (P , 0.9). The 95% confidence in-
terval on the rate of N accretion ranged from –5.7 to
16.2, providing strong evidence of a lack of sub-
stantial occult N input.
Jenny (1980) reported on a chronosequence
study (original data from Dormar and Lutwick
1966) of tree invasion of grasslands in Alberta, Can-
ada. The sequence included current grassland, a
45-year-old forest of balsam poplar (Populus balsam-
ifera L.), an 85-year-old forest of aspen (Populus
tremuloides Michx.), and a 150-year-old forest of
Douglas-fir. Soils differed greatly along this se-
quence, from a Mollisol in the grassland to an
Alfisol under Douglas-fir. If the differences across
these unreplicated sites truly represented 150 years
of soil development, then 90 kg N ha-1
y-1
were lost
(a total loss of over 13,000 kg N/ha). If this amount
of N were lost as soil leachate with 50 cm of annual
water loss, the average N concentration in soil
leachate would have been 18 mg N/L for 150 years;
we know of no forest that has ever sustained such
an extreme concentration of nitrate over more than
a century. The rate of loss of calcium1magnesium1
potassium also would have been staggeringly high,
approximately 3 kmolc ha-1
y-1
. We conclude that
this study warrants low confidence because the
change in soil orders was probably too dramatic to
have occurred in just 150 years, and that the im-
plied rates of loss for N and cations greatly exceed
expected rates, particularly for a dry climate.
In southern Sweden, Alriksson and Olsson
(1995) examined soils under 18 Norway spruce
stands (in three age classes) planted on former ag-
ricultural fields. The O-horizon plus 0–20-cm min-
eral soil contained 3160 kg N/ha in 20-year sites,
3400 kg N/ha in 40-year sites, and 3130 kg N/ha in
55-year sites. The average rate of change from year
20 to year 40 was an apparent net gain of approx-
imately 12 kg N ha-1
y-1
(correction for horizon
depths for the decline in bulk density would de-
crease this estimate somewhat). From year 40 to 55,
the N content appeared to decline by 18 kg N ha-1
y-1
. We suspect that the simplest interpretation
would be that little change in soil N occurred in
these stands, and soil N content versus stand age
shows no significant trend with age. The rate of N