eLearning

(with six to nine replicate pits/mudflow), and they
estimated a rate of N accretion of approximately 20
kgNha-1
y-1
, assuming that the mudflows had 0 N
when soil development begins (no value for fresh
mudflow material was given). If the 300 years date
is correct for mudflow B, the data of Sollins and
others (1983) indicate a rate of net increase from
the youngest flow to the second youngest flow of
just5kgNha-1
y-1
. The high rate of 63 kg N ha-1
y-1
from Dickson and Crocker (1953) is clearly errone-
ous based on the wrong age for the mudflow, the
single sample of the mudflow, and the subjective
choice of this single sample beneath a pine tree with
the maximum O-horizon depth. The intermediate
estimates of 14–20 kg N ha-1
y-1
accretion from
Glauser (1967) and Sollins and others (1983) as-
sumed that the mudflow had no N at the time of
deposit. The N concentration in the 30–70-cm
depth of the 60-year-old mudflow was 0.08 g/kg; if
this N represented the N present in the mudflow at
the time of deposition, then the N accretion rate
only would be 6 kg N ha-1
y-1
. This wouldmatch the
accretion rate for age 60–300 years of 5 kg N ha-1
y-1
(Sollins and others 1983). We conclude that this
set of unreplicated chronosequence studies pro-
vides moderately strong evidence of N accretion on
the order of 5–10 kg N ha-1
y-1
, but the evidence for
higher rates is weak.
In Australia, Turvey and Smethurst (1988) col-
lected samples of three stands at three ages, provid-
ing a replicated, 8-year chronosequence under ra-
diata pine (Pinus radiata D. Don). Soil N increased
significantly over time in the top 40 cm of soil, at a
rate of 285 kg N ha-1
y-1
. However, resampling the
chronosequence 6 years later showed no change in
soil N within each stand. The authors attributed the
apparent (and false) N accretion in the chronose-
quence to a fertility gradient across the sites that
coincided with plantation ages.
Aguilera and others (1993) analyzed a single soil
horizon (0–10-cm depth) in an age sequence Dou-
glas-fir (Pseudotsuga menziesii (Mirb.) Franco) stands
aged 2, 11, 36, 130, and 450 years. The differences
among stands were tested with analyses of vari-
ance, but given that time is a continuous variable
(and the lack of replication of stand ages), we used
a regression approach to test for changes in soil N in
relation to chronosequence age. On average, N in-
creased by approximately 2–3 kg N ha-1
y-1
through
the sequence, but the trend was not significantly
different from 0 (P 5 0.12). The trend depended
heavily on the high N in the old-growth soil; omit-
ting this stand dropped the significance level to 0.66
and gave a net loss of N over time rather than an
increase. A rate of N accretion of just 2–3 kg N ha-1