Biondini, M.E., C.D. Bonham, and E.F. Redente

Secondary Successional Patterns in a Sagebrush (Artemisia tridentata) Community as They
Relate to Soil Disturbance and Soil Biological Activity
Author(s): Mario E. Biondini, Charles D. Bonham, Edward F. Redente
Source: Vegetatio, Vol. 60, No. 1 (Mar. 15, 1985), pp. 25-36
Published by: Springer
Stable URL: http://www.jstor.org/stable/20146195
Accessed: 06/04/2010 10:04
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=springer.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
Springer is collaborating with JSTOR to digitize, preserve and extend access to Vegetatio.
http://www.jstor.org
successional
Secondary
Mario
of Range
Artemisia
Keywords:
sion
and soil biological
Charles D. Bonham,
E. Biondini,
Department
in a sagebrush
patterns
relate to soil disturbance
Science,
as they
community
activity*
& Edward F. R?dente*****
Colorado
tridentata,
tridentata)
(Artemisia
State
University,
community,
Sagebrush
Fort
CO
Collins,
Soil biological
80523,
activity,
U.S.A.
Soil disturbance,
Succes
Abstract
and soil biological activity were studied
The relationship between secondary succession, soil disturbance,
on a sagebrush community
in
the
Piceance Basin of northwestern Colorado,
U.S.A.
(Artemisia tridentata)
:
were
Four levels of disturbance
removed and as much topsoil as
imposed. 1 the vegetation was mechanically
removed and the topsoil scarified to a depth of 30 cm; 3:
possible was left; 2: the vegetation was mechanically
topsoil and subsoil were removed to a depth of 1m, mixed and replaced; 4: topsoil and subsoil were removed
to a depth of 2 m and replaced in a reverse order. Plant species composition,
and phosphatase
dehydrogenase
enzymatic
activity, mycorrhizae
infection
potentials,
and percent
organic matter
were
the variables meas
ured.
succession. Treatments
Treatment 4 drastically altered the pattern of vegetation
2, 3, and 4 started with
Salsola ib?rica as the dominant species but six years later, 3 and to lesser extent 2 changed in the direction of
of 1, dominated by perennial grasses and perennial forbs. Treatment 4 developed a
the species composition
shrub dominated community. The rate of succession was not decreased by the increased levels of disturbance.
infection potential (MIP) increased with the change
Both dehydrogenase
enzyme activity and mycorrhizae
from Salsola ib?rica to a vegetation dominated by either perennial grasses and forbs or shrubs. The intensity
in 2, 3, and 4 reduced drastically dehydrogenase
of disturbance
activity and MIP, but in six years they
to 1. Phosphatase
recovered to levels comparable
enzyme activity and organic matter were unrelated to
to
treatment
and
time elapsed. In both cases a significant decrease was
but
related
species composition
observed throughout the six-year period.
Introduction
Succession
plant ecology
caused
succession
1938).
theory has played a central role in
for more than 80 years. Early views
to be defined
as a community
or
species replacement driven exclusively by autogenic
modification
environmental
(Weaver & Clements,
1982).
Most
* Nomenclature
follows
** To whom all
should be addressed.
correspondence
*** This
study was funded by the United States Department
No. DE-AS02-76EV04018.
Energy under Contract
60, 25-36 (1985).
Vegetatio
1985. Dr W. Junk Publishers,
?
Dordrecht.
Printed
of
in the Netherlands.
Most
recently
new
theories
have
been
devel
oped that relate succession to tolerance and inhibi
and
tion factors, species life-history characteristics
&
1977;
(Connell
processes
Slatyer,
population
1974;
1973; Egler,
1976; Horn,
Drury & Nisbet,
Peet & Christensen,
1980; van H?lst,
1978; Grime,
1980; Matthew & Vankat,
1979; Noble & Slatyer,
studies have been confined to
succession
the vegetation part of the ecosystem. Only recently
between plant succession and
the interrelationship
soil biological
activity has begun to be studied.
Data that relates plant succession to nutrient reten
26
turno
1975), biological
matteraccumulation
1982), organic
ver(Titlyanova,
1981), nitrification
(Aweto,
potentials (Rice, 1974;
Robertson & Vitousek,
1982), and soil enzyme ac
& Reiners,
tion (Vitousek
1: the
Treatment
2:
Treatment
the
noted
succession,
lack
of
position during succession.
for this study.
Three hypotheses were developed
of
1. A severe soil disturbance,
by an alteration
the
was
vegetation
re
mechanically
and the topsoil scarified to a
of
30 cm.
depth
3: topsoil and subsoil (C horizon) were
removed to a depth of 1m. The mate
rial was mixed together and replaced.
4: two layers of 1 m of soil were re
in a reverse
and replaced
moved
order with the second layer placed on
Treatment
informa
to
moved
1977; Rossetal.,
1982) are
(Rice & Mallik,
a
of
review
literature
in
Parkinson
available.
( 1979),
tion which relates plant succession to type and lev
els of disturbance and soil biological activity.
to address two
The present study was designed
1
to
how
various forms
main objectives: ( ) determine
can
affect the soil
and intensities of soil disturbance
succes
rate
of vegetation
activity and
biological
sion; and (2) to determine the degree of relationship
activities and species com
between soil biological
re
mechanically
with minimal disturbance
topsoil (A and B horizons).
tivities
reclamation
was
vegetation
moved
Treatment
the
The
block
surface.
was arranged
in a randomized
two replications
per treatment
in total). The plots were 6 X 8 m with a
experiment
design with
(eight plots
1.5-m
buffer
zone
between
was
measured
variable
The
plots.
plant
vegetation
cover.
canopy
The
alters
physical and soil biological characteristics
the direction of secondary succession.
retard the rate of
2. Increased levels of disturbance
the level of soil
successional
reducing
change by
sampled once a year, at the end of the
season, with ten 0.25 m2 (25 X 100 cm)
growing
had been randomly
permanent
quadrats which
locted within each plot (Redente et al, 1982). Cover
biological activity.
3. There is a relationship between species composi
tion and levels of soil biological activity.
values
Materials
and methods
The
site was
plots were
were
Measurements
study
in the
located
Piceance
was
Basin
soil
of 2200 m.
at an elevation
Colorado
Sagebrush-grassland
A.
the dominant
S tipa
hymenoides,
nea
were
Koeleria
riparium,
major
vegetation
coma
understory
cristata,
ta, and
Oryzopsis
Sphaeralcea
species
cocc?
0.5
mmhos/cm,
soil
matter,
organic
potential
enzymatic activi
phosphatase
infection potential (MIP) were
soil biological activity (Kleinen
al, 1982). Organic matter was
in the present study as a general index of the
nitrate-nitrogen
was
reserve
co-factors.
(Harrington,
1964). Soil texture ranged from loam to clay loam
30-60 cm
with the combined A and B horizons
was
electrical
The
(EC)
8.0,
conductivity
pH
deep.
averaged
of
com
species
relative cover).
status.
nutrient
Soil
activities
enzymatic
are a far better index of soil biological activity than
counts since little is known about the
microbial
activities of individual microbial
species (Kupre
en
vich & Shcherbakova,
1966). Dehydrogenase
in the oxidation of carbohydrates
zymes participate
as
and require the presence of NAD and NADP
tridentata
and Artemisia
type before disturbance
cover.
of
the
60-80%
canopy
Agropyron
comprised
smithii,
to calculate
utilized
and
dehydrogenase
and
ty,
mycorrhizal
utilized as indices of
al, 1982; Reeves et
chosen
of northwest
then
(as a percent
position
5
in soils
Dehydrogenases
are
only
in
found
and as such
micro-organisms
functioning
their level of activity can be used as an index of the
(Sku
capacity to process carbon by the microflora
in
of
The
enzymes
1978).
phosphatase
majority
jins,
the soil are contributed by soil heterotrophic micro
intact
ppm (water extract), and phosphorus was 2.3 ppm
bicarbonate
extract) in the first 15 cm
(ammonium
is 250-300 mm, ap
of soil. Annual precipitation
snow (Redente et
as
one
half
received
proximately
al, 1984).
The study was initiated in the summer of 1976.
Treatments
consisted of four levels of soil disturb
level,
1978). Their activity
zymes (Speir & Ross,
of free
then, can be an indicator of the availability
carbon and nu
(not part of the plant material)
trients in the soil. It has been shown that mycorrhiz
of many cli
al fungi are crucial in the functioning
ance:
max
organisms
species
even
though
in a variety
some
can
of ecosystems.
exist
Their
as
free
presence
en
27
or
absence
can
then
a determinant
be
factor
in the
control of successional patterns (Langford & Buell,
1969; Reeves et al, 1979, 1982).
soil samples from a depth of 5-10 cm were
collected from each plot at the time of the
vegetation
sampling. Soil samples were stored in
Three
randomly
double-wrapped
plastic bags at room temperature
until analyzed
in the laboratory. Dehydrogenase
and phosphatase
enzymatic activities and soil or
were
matter
ganic
& Klein
measured
to Hersman
according
presented
(1979). The dehydrogenase
activity values
in this study represent the 'potential' ac
tivities,
i.e.,
croflora
to process
the maximum
of
capacity
carbon.
The
the
were
assays
mi
soil
carried
out with
the addition of 0.5 ml of a 1% glucose
solution
in place of distilled water. Mycorrhizal
as percentage
infectivity of the soil was measured
infection (mycorrhizal
infection potential or MIP)
in corn bioassay plants as described by Moorman &
Reeves
(1979).
Multi-response
(MRPP)
statistical
permutational
procedures
et al, 1981a) were used for the
rate of
analysis of species composition,
(Mielke
successional
(as measured
change
by
com
species
and soil characteristics.
(See
changes),
on the MRPP
for
details
'Appendix'
technique.)
between vegetation and soil biologi
Relationships
cal activity were analyzed with canonical correla
regression
procedures.
Results
successional
Vegetation
The
main
patterns
patterns
of
Agropyron
riparium,
A.
smithii,
in
succession
vegetation
this study were given by the changes
of: ( 1) perennial grasses (the dominant
through time
species were:
Koeleria
cristata,
and Stipa comata), (2) per
Oryzopsis hymenoides,
ennial forbs (the dominant species were: Sphaeral
cea
cocc?nea,
Erigeron
engelmanii,
Phlox
longifo
lia, Senecio
and Trifolium gymno
multilobatus,
(3) annual forbs (the dominant
carpon),
species
was: Salsola ib?rica), and (4) shrubs (the dominant
species
were:
Artemisia
C.
nauseousus,
viscidiflorus,
tridentata,
and
Chrysothamnus
saro
Gutierrezia
thrae).
The
grasses
general
was
an
pattern
increase
followed
in percent
by
perennial
relative
cover
perennial
grass
and an inverse relationship
and
composition
the
sever
ity of the treatment (Table 1). The grass PRC in
1 increased from 49.04%
Treatment
in 1977 to
62.15% in 1982. The vegetation of Treatments 2 and
3 began with a very low grass PRC but made a
substantial gain in the six-year period. In contrast,
the vegetation of Treatment 4 had a low grass PRC
the
throughout
nent increased
1982.
six-year
period.
from 0.04%
Its grass
compo
in 1977 to only 5.44% in
The general pattern followed by perennial forbs
was (a) an increase in PRC as time
elapsed, and (b)
an inverse relationship
between
forb
perennial
PRC and the severity of the treatment (Table 1).
Perennial
forb PRC
in Treatment
1was virtually
in the six-year period. The vegetation of
unchanged
2 and 3 started with a low perennial
Treatments
forb component but as time elapsed they changed in
the direction of Treatment
1. Perennial forb PRC
also increased with
time in Treatment
4 even
though it remained below the level of the other
three
treatments.
ib?rica was a major
Salsola
position
tion and multiple
(PRC) as time elapsed
between
steps
of
succession.
species
Its contribution
in the initial
to species
com
3 and 4 was very high in the
position of Treatments
first year of succession
ib?rica
(Table 1). Salsola
PRC decreased sharply with time in all treatments
in contrast to grasses and perennial forbs. The only
treatment with a sizable Salsola ib?rica component
after six years was Treatment
3 with a PRC of
21.04%.
Shrubs were the main group of species to differ
entiate Treatments
1, 2, and 3 from 4. Shrub PRC
on Treatments
1 and 3 never surpassed
10.0%
the six-year period (Table 1). Treat
throughout
ment 2 showed a steady increase in shrub PRC
(Table
74.86%
4.
1). The biggest
increase, from 5.40% to
in shrub PRC was observed
in Treatment
The major
trends in the PRC of the species
groups described above support the hypothesis that
high levels of soil disturbance can alter the direction
of secondary succession (as defined by
species com
The
was
position).
hypothesis
formally tested with
multi-response
permutational
procedures (M RPP).
The PRC of perennial grasses, perennial forbs, an
nual forbs, and shrubs were used as the multivariate
observation
which characterized
the species com
position
of
each
treatment.
Treatments
were
ana
28
O oo en ir>
O sO oo r)
O O CN?'
?, os
en ? O
r^ os o
T?;O ?
? ? ?
en o
oo r
CN 00
O ?
t* ?' ? t*
rf O O
o
o
CNO O O O
o
o
o
Tf O
o
o
83
O O O O O O o
o
SO
so O O
?'
r-- O
en Os O
?
t?- ?
sO?
?
?
?
?
?
en
?
CN tJ
?
?
o
p
o
r- Tt m r
p
en?oo ?n
o o ? ?'<
ON OS
<N
o
so so un rj- m
o
o
o
t?- o
O
en o
en O O
en o
o
?
Os r
CNO O O O O
?
o
o
o
o
o
CN
CN r? CNO Tf o?i ?n
un ?OO CN O
o ?' o ?'
Os' CN
?
OO
CN?
en
r- o
en un ?? cn ?
?
CN CN
CN r- CN?? O ? Tf
^J- o? m
SO OO?
?'
O
o
o
?
o
o
r- os
o
o
o
cmm
o
o
o
?
?
t?- ?. o
r
un en
o ?'<
os
o
?;
??' O O O
en
?' un r-. es un
?- Tf CN r-? ?o
es un
CNo
o
?? O
CN
O O O ?
-*i-m m
u->O
CNjo
oo
o
o
o
o
o
SO Tf Os
00 O
en O O O
CN
o
o
o
o
o
so O O O
en
O O ?
O
CN
CNO
un CN CN oo
O O SO CN
K -J \0 -^
sO O
"o
sO -_ OOo
O CN?
*r!
??
o
?
00 O
so OOOs en Tf O
^
?
?
o o ? ?
O O Os Os
?o un
o o
00
en O
CN CN un
?
?
Os cn so en sOo
Tf CN Tf Tf O
?
CN so CN?
CN?? p O T?;
O Tt O O Tf
CN Tt
CN
CN
s:
o
s
o
-a
IV.
c
o
i
o
3
:?5
>
53
O
o
*
00 a
II
-Ci
a
c/5
g X X ^
&
=
-g ^
"^
? A-a
OnOh
? c
toc g
s
?1
O .3 O-
?
Co kt Q^ co
'S
C
"
o.
g
a
|S
o o
?5 ?
<sj3
"S^
111
l?o
c
"5
3 1s
i
*: $:
-S
>,
<*.
c?_
? ^ sj U ^ h
cd E
2 s?
29
lyzed at two points in time: ( 1) one year into succes
sion, and (2) six years into succession. One year
into succession, Treatments
2, 3 and 4 were shown
to be different (p = 0.035) from Treatment
1. No
was
found among
significant difference (psO.ll)
ed how important the rate of change of a particular
group of species was to the overall species composi
The sum of these
tion of the entire community.
can
as
an
be used
indicator of the
then,
products,
rate of species composition
change through time for
1
Treatments 2,3 and 4. Six years later, Treatments
=*
were
and 3
similar (p
0.090). Treat
marginally
ment 2 was different from Treatments
1 and 3 (p =
was
different from 2 (p =
0.042) while Treatment 4
index for the rate of succession within each treat
ment (Table 2). The values calculated according to
3 above were used to test the hypothesis
that in
0.03). The hypothesis
creased
can
disturbance
that increased
alter
the
of
pattern
the
levels of soil
six
the
however,
later,
years
extent
lesser
became
ences were
rates of
found among the successional
2 and 3 (p s 0.18), but they were lower
4 (p s 0.07). Under the
than that of Treatment
the hypothesis was rejected.
circumstances
Soil biological
grass-forb
rate
of
successional
vegetation
grasses,
perennial
and
test the hypothesis of a relationship between
and soil biological activity, a
species composition
canonical correlation analysis was performed. The
main groups of plants that defined vegetation
trends (as emerged from the previous section) were
run
were
shrubs
annual
forbs,
perennial
on each
treatment.
forbs,
In all
of
the
regression analyses only the linear term was signifi
cant (p < 0.05 was the criteria used). This resulted
then in a simple straight line regression; (2) on each
treatment a ratio was calculated between the high
est PRC achieved during the six-year period by
each of the group of species mentioned
above and
the sum of these values; (3) the absolute value of the
total grasses, perennial forbs, annual forbs (Salsola
ib?rica), and total shrubs. Their PRC were the vege
and phos
tation variables used. Dehydrogenase
M
and
IP,
percent organ
phatase enzyme activities,
ic matter were utilized as measurements
of soil
biological activity.
The analysis rendered
only one significant (p =
0.001) canonical variable with a canonical correla
tion of 0.806. The coefficients
for the vegetation
slope of the calculated regressions (see 1above) was
an indication of the PRC rate of change through
time
for
activity
To
was
change
in the following way: (1) stepwise poly
regressions between time and the PRC of
calculated
nomial
the
different
of
group
species.
tiplied by the PRC ratios (see 2 above)
Table 2. Treatments
mean
values
value
highest composition
of succession.
Perrenial
When
mul
for Salsola
for: ( 1) the slope of the linear regression
by each group of species throughout
Perennial
forbs
were
ib?rica, -0.54
-0.84
for
total
grasses,
1
Treatment
2
p
b
p
b
0.56
2.94
0.30
-0.51
Salsola
0.21
5.14
0.22
3.00
7.44
0.15
0.06
Treatment
3
0.26
Treatment
4
0.03
1.12
= ratio between
the highest composition
p
b= slope of the linear regression
between
value achieved
species
p
and time, (2) the ratio between the
species composition
the six-year period and the sum of these values and (3) the rate
Shrubs
Rate
of
ib?rica
b
b
p
0.12
-2.39
0.36
-13.41
0.215.27
7.67
4.60
0.54
-12.70
0.050.28
9.50
2.21
0.50
-15.50
0.41
by each group of species
and time.
composition
0.85
for total shrubs and 0.09
between
achieved
grasses
variable
canonical
they indicat
-succession
Treatment
an
communities.
dominated
The
2
Treatment
as
used
Treatments
species
of Treatments
2, 3 and 4 was statisti
composition
cally different and (2) Treatment 4 became a shrub
dominated community while Treatments
1,3 and to
a
was
levels of soil disturbance
retard the rate of
change. The analysis showed that the
rate of Treatment
1 was lower than
successional
that of Treatments
2, 3 and 4 (p == 0.04). No differ
cession was accepted because of the following: (1)
the species composition
of Treatments
2, 3 and 4
was not statistically different in the early stages of
succession;
sum
This
community.
successional
suc
secondary
entire
0.02
=
Xp* \b |
2.11-0.14
11.96
12.82
and the sum of these values
for all groups
of species.
30
forbs. The coefficients
for perennial
logical
canonical
activity
40
for the soil bio
were
variable
-0.85
that total perennial grasses, total shrubs,
Salsola
ib?rica, dehydrogenase
enzymatic activity
for
and M IP were the main variables responsible
indicates
the overall
relationship between vegetation and soil
biological
activity. To further study this relation
ship regression analyses were run between each soil
measurement
and
of
each
the
<
O 20
M
O
16
E
o
o oer
12
o o>
> =t 8
x
1, 2 and to a certain extent 3) or a
(Treatment 4). The
community
did not differ significantly
(/; ==
in
in
year 6
0.15)
dehydrogenase
enzymatic activity
of
of succession
of
the
the
soil
regardless
severity
disturbance
(Table 3).
36 |_
Activity
64
56
20
40
values
pendently
of whether
I'6 12
?
8
32
40
?2
80
88
96
GRASSES AND SHRUBS (%)
on
MIP
Fig.
in relation
tial (MIP)
grasses and shrubs.
and mycorrhi/al
activity
to the species
composition
infection
poten
of perennial
or
values
Phosphatase
ganic
sition.
succession
was
vegetation
a
inde
changing
com
shrub-dominated
after
six
did not have
of
years
succes
They
activities
enzymatic
were
matter
were,
enzymatic
not
however,
percent
or
compo
correlated
negatively
in the succession
activity
and
to vegetation
related
to
(Fig. 3a). Phospha
from
decreased
an
average
'
! in
(across all treatments) of 147.65 pg PNP g hr
'
1977 to 56.22 pg PNP g hr
in 1982. Likewise,
organic
I. Dehydrogenase
_L_
88 96
72 80
severity of the treatment
The
effect
tase
16 24
the
a grass-forb-
time elapsed
_l_
48 56 64
64
with
increased
CL
32 ?
24
S
-I_L.
48 56
sion. In 1982 MIP values for Treatments
1, 2 and 4
were not significantly different (p = 0.20) (Table 3).
3 was significantly different from Treat
Treatment
ments
4 (p = 0.05).
and
1,2
J28
'o*24
gisl
40
infection poten
activity and mycorrhizal
to the composition
of Salsola
ib?rica.
The MIP
an
72
3?
T
32
ib?rica. InTreatments
3 and 4, where Salso
la ib?rica comprised more than 90% of the species
MIP values were 13% and 11.5% in
composition,
the first year of succession
2
(Table 3). Treatment
1 had a
had a MIP value of 44.5% while Treatment
1and 2 were significant
value of 55.0%. Treatments
3 and 4 (p = 0.032).
ly different from Treatments
munity.
80
Dehydrogenase
Q.
32 i
24
16
infection potential followed a pat
Mycorrhizal
tern similar to dehydrogenase
activity. It has a sig
nificant relationship with both perennial grasses
and shrubs and Salsola ib?rica (Figs. 1and 2). Low
levels of MIP coincided with the dominance
of
toward
o MIP
^sV
40
Salsola
nity (Treatments
shrub-dominated
four treatments
40
56
48
0> 24
Fig. 2. Dehydrogenase
tial (MIP)
in relation
commu
grass-forb-dominated
72
64
SALSOLA IB?RICA (%)
served as succession
and perennial
proceeded
grasses or shrubs became established. This increase
was independent
was
of whether
the vegetation
a
Activity
a 0.01
16 24
activities
coincided with stands domi
drogenase
nated by Salsola
ib?rica. Treatments
2, 3 and 4
started with dehydrogenase
activities of 8.45, 3.82,
1
and 4.25 pg formazan g ' 24 h 'while Treatment
had values of 12.16 pg formazan g l 24 h ' (Table
3). An increase in dehydrogenase
activity was ob
toward
c 32
\j 28
variables.
vegetation
enzyme activity was positively
Dehydrogenase
to
of perennial grasses
related
the establishment
and shrubs and negatively related to Salsola ib?rica
(Figs. 1 and 2). Low values of dehy
composition
changing
Dehydrogenase
o
EP
activity, 0.89 for M IP, 0.12 for
dehydrogenase
activity and 0.05 for soil organic mat
phosphatase
ter. The structure of the first canonical
variable
80
o MIP
36 |_
for
matter
decreased
from
an
average
(across
all treatments) of 2.22% g 'dry soil in 1977to0.93%
'
g dry soil in 1982. The severity of the treatment
proved to have a significant and lasting effect on
O
as'
?
O
as
?
p
SO
?
?
Os
sd
?
r
?
*r^
03
'Ja
?
as'
as
a,
o
fi
O.
N
\3
? a,
.S -3
g _
o L *> a
^5
I" fi 3
?o ^ ? ?o L -c
* S
o
CO
O ?
-S * Je t ?
.s-ioj.?'i.l.-s
32
nial
CL
Z
?.
to the establishment
related
both
as
species
a group
rather
cies or life forms
(as shown
were
in both
values
similar
dominated
matter,
were
to
unrelated
compo
species
on treatment
and time.
The general pattern of succession for Treatments
of an initial stage in which
2, 3 and 4 consisted
ib?rica was the dominant
Salsola
species followed
a
by
YEAR OF SUCCESSION
>
shift
200
the
throughout
small
nial
Z
CL
O? 80
genase
six-year
between
trade-off
160
^
i
O?
levels
06
24
SOIL ORGANICMATTER (%)
matter
activity
enzymatic
and phosphatase
Six
parameters.
and
activity.
years
after
disturbance,
3 and 4 had lower phosphatase
enzy
matic activity than 1 and 2 (p == 0.05) (Table 3).
1 and 2 was consis
Organic matter in Treatments
4
in
in 1977 (p =s0.05)
than
3
both
and
tently higher
and 1982 (p ==0.052) (Table 3). Phosphatase
enzy
Treatments
activity and percent organic matter were line
correlated
arly
(Fig. 3b).
The general hypothesis of the existence of a rela
tionship between vegetation and soil biological ac
but as the detailed analyses
tivity was accepted,
matic
shown,
not
all
the
parameters
ilar pattern. Dehydrogenase
activity
followed
with
and
peren
processes,
belowground
dehydro
activity and MIP increased their
the
a sim
and MIP were
as
decreased
advance
time
of
succession
were
and
in the soil) sharply
This
advanced.
as
that
indication
succession
decline
may
the
advanced
be an
nutrient
in
flow became tight (more nutrients
immobilized
the plant biomass) and that the grasses and forbs
that dominated Treatments
1, 2 and 3 or the shrubs
that
two
enzymatic
bility of free nutrient and carbon
soil organic matter
and soil organic
(a) Average
activity
phosphatase
dynamics,
between
matter as related to successional
time; (b) relationship
3. Phosphatase
a
only
showing
ib?rica
to the shift in dominance
from Salsola
correlated
ib?rica to perennial grasses and shrubs (Figs. 1and
2). Phosphatase
enzymatic activity and soil organic
matter (which can give an indication of the availa
54.6X
p~ 0.01
-L_
18
1.2
period,
Salsola
grasses.
Considering
y = 15.3+
r2 =0.60
these
com
grass-forb-dominated
invasion by Salsola
ib?rica. The species composi
1 remained relatively unchanged
tion of Treatment
>
soil organic
a
toward
2 and 3 and a shrub-domi
in Treatments
munity
in 4. In Treatment
nated community
1, the lower
level of soil disturbance allowed only for a reduced
240
have
shrub
and
activity and organic
Discussion
</)
?
<
X
CL
(/)
O
X
CL
Fig.
spe
grass-forb-
sition but rather dependent
o
<
LU
peren
individual
by the fact that their
plots). Phosphatase
however,
of native
than
these
exploit
4 were
Treatment
dominated
conditions.
We
more
speculate
to
able
that
the
ca
species to exploit
pacity of the latter successional
conditions of low nutrient availability may be relat
ed to a successional
shift in the microflora
composi
tion
from
on free nutrients
ganisms
plant dependent micro-organisms
in the rhizosphere.
Both
corrhizal
capacity
rhizosphere
fungi
of plants
micro-or
heterotrophic
predominantly
that depend
micro-organisms
have
been
to acquire
shown
nutrients
in the soil to
which
as well
function
as my
to increase
under
the
condi
tions of nutrient stress (Alexander,
1977; St. John
This
& Coleman,
the
diversion of
1983).
requires
33
fixed
toward
carbon
occur
erally
of
conditions
as
matter
organic
rhizo
a
nutrient
abundance
of
consequence
tied
previously
of
material
that
(Gorhman et al, 1979). Therefore
it is not a profitable strategy
under these conditions
for plants to divert part of their fixed carbon to
maintain
an
ulation.
With
more
extensive
the
nutrients
and
rials
forms,
less,
resulting
microflora
rhizosphere
of
advance
in plant mate
immobilized
become
in our
are
view,
in a nutrient
pop
however,
succession,
stress
for
the
plants.
this point, we theorize, it becomes a profitable
strategy for plants to divert part of their fixed car
in order
bon to maintain a rhizosphere population
to increase their capacity to acquire nutrients. It is
within this general context that we view our hypo
At
shift in the microflora
thesized
one
predominantly
comprised
ganisms
to one with
zosphere
micro-organisms.
a higher
composition
of
free
from
soil micro-or
of rhi
composition
Microbial
activity in the free soil and the availa
a
source are correlated with the
of
carbon
bility
level of phosphatase
enzymatic activity (Speir &
A
nutrient
Ross, 1978).
cycle and a reduction
tight
would
in free soil heterotrophic
micro-organisms
be consistent with the observed decline in phospha
tase activity and organic matter. Cundell
(1977)
found
evidence
that
in semi-arid
areas
the
sphere of grasses and shrubs is a very favorable
for micro-organism
development.
The
observed
rhizo
place
in
creases
in dehydrogenase
activity and MIP could
then be explained by an increase in the rhizosphere
micro-organisms
induced
by an
improvement
in the
rhizosphere environment with the shift from annual
forbs (Salsola ib?rica) to native perennial grasses
and shrubs. It could also explain the relationship
found between dehydrogenase
activity and MIP
and
species
composition.
One factor that was not anticipated was the rapid
recovery of dehydrogenase
activity and mycorrhi
zae infection potential under conditions of extreme
soil disturbance
(such as Treatment 4). This result
the
of a
may explain
rejection of the hypothesis
in vegetation
rate with an
successional
soil disturbance. This recovery in dehy
drogenase activity and MIP was particularly unex
4 where the horizons were
in Treatment
pected
reversed and the C horizon became the topsoil. This
decrease
increased
other
the
and
could
hand,
dominance
poor
more
to
adapted
soils
re
been
in
of shrubs
of
performance
nial grasses and forbs. Shrubs, with
are
have
peren
their deep root
a
with
coarse
structure and precipitation
out of the growing season
that takes place either
or consists of large but
infrequent events (Neils & Tueller, 1971). These two
conditions were met in the study. The reversed ho
rizons resulted in a new 'topsoiF with a very rocky
surface. Fifty percent of the rainfall in the Piceance
occurs in the winter. The very low level of
in these
forb establishment
grass and perennial
could
have
the
also
enhanced
1)
(Table
plots
prob
in
ability of shrub establishment
by a reduction
Basin
in mineral
available
treatment
system,
we theorize
the
for the ultimate
sponsible
this
gen
breakdown
in plant
up
on
operation,
In the initial stages of secondary
sphere population.
succession,
a
of
the maintenance
The
competition.
could
have
of
establishment
created
adequate
shrubs
then
condi
rhizosphere
tions for microbial and mycorrhizae
development.
This could explain
in part the rapid recovery of
to
dehydrogenase
activity and MIP
(according
Reeves et al (1979) the four shrubs in question
(Table 1) are mycorrhizal).
Another
enough
corrhizae
four
years
result was
unexpected
viable
of
spores
were
still present
without
an
the fact
vesicular-arbuscular
that
my
to reinfest plants after
host.
adequate
re
Recent
search by Schmidt & Reeves (in press) advances the
that Salsola ib?rica, even though not a
proposition
mycorrhizal
plant,
can
create
conditions
around
the roots (such as some carbohydrate
exudates) to
to
allow the fungi spores
remain viable until an
adequate host develops.
en
In the present experiment
dehydrogenase
zyme activity potentials after six years of succession
'
averaged 25.99 pg formazan g^ soil 24 hr while
native undisturbed
vegetation had values of 19.05
!
'
pg formazan g soil 24 hr (Kleins al, 1982). This
was
consistent
with
difference
Titlyanova's
(1982)
in the Siberian
succession
analysis of vegetation
steppe. She found that the activity of the microbi
ocenosis
was
maximized
stages and declined
at
the
as vegetation
intermediate
approached
serai
cli
max.
One aspect of the results was in direct contradic
tion to most of the literature. Organic matter has
been widely reported to increase with succession
1965; Zedler & Zedler, 1969;
(Bard, 1952; Davison,
Chertov & Razunovskii,
1980; Aweto,
1981; Shav
kat et al,
however, a
1982). In this experiment,
decline in soil organic matter was observed
in all
34
treatments
ter
to
parable
levels of soil organic mat
(Table 2). The
1 and
in Treatments
the
ones
2 after
six
in
observed
years
the
were
native
com
undis
turbed vegetation,
1.12% (Klein p/ al, 1982). After
six years soil organic matter was still below the
in Treatments
3
native vegetation
levels, however
for this result could
and 4. A plausible explanation
of an equilibrium
be a lag in the re-establishment
inputs and outputs in the carbon cycle. We
theorize that the release of C02 via the decomposi
in the soil by
tion of organic matter (incorporated
the disturbance) was higher than the corresponding
between
inputs of new organic
in a negative
resulted
matter
ference
can be made.
as
and
occurred
a reduc
such,
through
time.
are distributed
as the multivariate
the two subgroups
normal
with equal variances
and covariances.
Since these
distribution
are seldom met in practice,
to consider
it is desirable
conditions
alternative
One
way
procedures.
to test differences
1J=
||.V,
.Vj||=[
explanation
is expanded
of
the statistical
upon
MRPP,
technique,
that which was developed
from
et al. (1981a).
of MRPP
concepts
general
a comparison
between
considering
groups of objects (A and B) where
that
from
Mklkc
The
have
been
can
best
be
two mutually
two measured
illustrated
are
A,./)2]1
exclusive
responses
(x\
in the two sub
object
how these responses could be represent
the responses
of the
diagram where
obtained
groups. Figure 4 shows
ed in a two-dimensional
from
each
as crosses
A are plotted
in subgroup
and the
objects
in subgroup B are plotted as circles.
responses of the four objects
three
i
l
I
5+-4
4
i
i
?
I
l
4
4
?
--
--M
24
?
-t-4
I
I
?
in Table
listed
4 and ordered
from
the visual
the
impression
of a common
points
than the distances
between
points
lowest
average
subgroup
of different
Thus,
?A
=
Table
points
either
circles,
is
(l/3)
v
AA/=
A
1.6095
(2)
4. Ordered
distances
between
all 21 pairs of the seven
in Figure 4 where distances
between
in
points
A or subgroup
B are indicated
subgroup
by crosses or
shown
respectively.
Rank
Points
Distance
l
,A3
I
I
IB2
B3
|
14-Y
I
I
B,B2 1
B2B3 2
A,A2
i
I
lB4
I
I
H-1-1
I'
the points of the two subgroups
Fig. 4. Scatter diagram
showing
responses X\ and
(A and B) plotted as a function of the measured
Xi.
to
of cluster
this clustering
is
way to consider
tendency
of the between-point
for each
distances
for the three distances
of subgroup
A, the
1.000 (o)
1.000 (o)
B3B4 3
-*--4
1
3+
(1)
4--4--*
l
4f-~
i
2
A natural
subgroups.
to form an average
subgroup.
by
sub
is by first
the groups
in the diagram.
between
all distinct
The seven
pairs of points
=
points of Figure 4 imply there are (72) 21 distinct pairs of points
21 distances
must
and consequently
be computed.
21
These
tend to be smaller
and xi)
^
(.Y,,
i= I
highest value. Table 4 confirms
between
ing since the distances
Appendix
follows
between
the distances
examining
distance
The
involve the use
approach would
T2 test which has the disadvantage
that the response measurements
of
A classical
of the two-sampled
Hotelling
the assumption
of requiring
by the plants. This
balance,
tion of soil organic matter
a visual impression
Although
suggests that the subgroups A and
a more rigorous and objective
B are separated,
characterization
or in
is needed before a quantitative
of this separation
evaluation
1.000(0)
1.414 (x)
1.414 (x)
4
A2A3 5
6
A2B,
7
A3B3
1.414
1.414
B|B3 8
B2B4 9
B,B310
11
A2B3
12
A3B1
13
A2B2
'
14A?B2
15
A3B4
B|B416
17
A,B,
1.414(o)
1.414(0)
2.000 (x)
2.000
2.000
2.236
2.236
2.236
2.236
2.828
18
A2B4
19
A,B4
3.000
20
A,B2
21
A,B4
3.606
3.162
4.123
(o)
35
and for the six distances
?b
=
(1/6)2
B
a7</=
A measure
or statistic
the points
of subgroups
given
?=
B the average
of subgroup
1.3441.
is
approximate
tion in order
(3)
the separation
between
describes
A and B is the simple weighted
mean
which
by
1.4578.
(3/7)?A4(4/7)?B=
(4)
values of ? would
indicate a tendency for clustering while
indicate a lack of clustering. The prob
larger values of ? would
=
the observed
lem is to determine whether
statistic(<5
1.4578) for
this particular
(A and B) is unusual with respect to
partition
Small
that could
possible partitions with the same size structure
have been made with these seven objects. Now yVobjects can be
into twosubgroups(A
and B) with fixed numbers of
partitioned
of ? from a continuous
the distribution
distribu
a p value for an observed
value of ? is smaller than some prescribed
level of significance.
As shown by M ielke et al. ( 1981 a), the distribution
of the statistic
? can be approximated
type III distribution.
by a Pearson
to determine
whether
we
In this appendix
The procedure
of a two group
have given an example
can be extended
to multigroup
analysis
of a one-way
and to a two-way
MANOVA)
analysis.
(the equivalent
of a factorial MANOVA)
factorial
(the equivalent
analysis
for MRPP
(Mielke et al., 1981b, 1982). Computer
programs
at Colorado
have been developed
and are available
analysis
State
University.
other
points
a?a and a7?, respectively,
in precisely
References
A weto,
(5)
M=M/(nA\nti\)
Ecol.
35 for this example,
35 values of ? can be
ways. Since
all the possible 35 partitions. These 35
obtained
by enumerating
values of ? are listed in Table 5 and ordered from the lowest to
=
statistic
1.4578)
(?
highest value. We see that the observed
obtained
for the realized partition
is
and
unusual
indeed
(A
B)
since each of the remaining
have occurred with
then the observed
Thus,
we would
34 values is greater.
If all partitions
(the null hypothesis),
equal chance
level or /; value is 1/35 = 0.0286.
significance
the
realized
(A and B) as
accept
partition
at the 2.8% level of significance.
significant
When M is large (e.g., M = 1.55 X 108 when N = 30 and nA =
to generate
the discrete probabili
15), it is obviously
impractical
of 6 illustrated
to
in Table 5. It is then necessary
ty distribution
Table
5. Ordered
1981. Secondary
succession
in southwestern
Nigeria: 2. Soil
69: 609 614.
of New
1952. Secondary
succesion on the Piedmont
Ecol. Monogr.
22: 195 216.
S. M., 1980. Ecological
trends in
Chertov, O. G. & Razumovskii,
Zh. Obshch.
Biol. 41: 386-396.
soil forming processes.
Bard, G. E.,
M =
could
and soil fertility resto
J.
fertility restoration.
A. O.,
ration
values
of the seven
of ? for all 35 partitions
(A and B) having
points shown in Figure 4 into two subgroups
fixed sizes nA = 3 and /7b = 4
Rank
Value
1.4578
1
Rank
19
Value
2.1381
2
1.5421
20
3
1.6939
21
2.1591
4
1.7505
22
2.1646
1.8389
5
23
2.1709
1.8547
6
24
7
1.8935
25
2.1769
2.1891
2.1480
Jersey.
R. O., 1977. Mechanisms
J. H.&Slatyer,
of succession
and the role in community
in natural communities
stability
Am. Nat.
Ill: 1119 1144.
and organization.
Connell,
Cundell,
tation
A. M., 1977. The role of microorganisms
in the revege
J.
of strip-mined
land in the western
United States.
30: 299-305.
Manage.
in
R. L., 1965. An experimental
study of succession
D.
J.
S.
In:
D.
S.
the Transvaal
Davis,
DeMeillon,
Highveld.
Range
Davidson,
& M. Valk(eds.),
Studies
Ecological
pp. 113 125. W. Junk, The Hague.
I. C. T., 1973. Succession.
W. H. & Nisbet,
Drury,
Egler,
Univ.
Harv.
Arbor,
F. E.,
1976. Nature
Comm.
Mismanagement.
527 pp.
J. Arnold
54: 331-368.
of Vegetation,
Cons. Assoc,
Its Management
of the Plants
soil used
in land reclamation
and
Workers,
Processes.
of Colorado.
111.666 pp.
1979. Evaluation
L. E. & Klein, D. A.,
Hersman,
shale effects on the microbiological
and
Conn.
Bridgewater,
Method
for Research
Fisher, R. A., 1958. Statistical
13th ed. Hafner, New York, N.Y.
175 pp.
J. P., 1979. Plant Strategies
and Vegetation
Grime,
John Wiley & Sons. 222 pp.
H. D., 1964. Manual
Harrington,
Press Inc., Chicago,
Swallow
2.1740
in Southern
Harrington
Africa,
of retorted
characteristics
revegetation
The
soil
of surface
processes.
J.
1.9898
8
26
1.9915
9
27
1.9988
10
28
2.2025
11
2.0060
29
2.2169
12
2.0157
30
2.2258
2.0176
13
31
2.2280
of secondary
Ann.
succession.
Syst. 5: 25-37.
of vegetation
H?lst, R. van., 1978. On the dynamics
patterns of
environmental
and vegetation
38: 63-76.
changes. Vegetatio
14
2.0522
32
2.2470
Klein,
15
2.0575
33
2.2518
2.0829
16
34
2.2812
17
2.0944
35
2.2935
18 2.1158
2.1939
Qual. 8: 524 529.
H. S., 1974. The ecology
Environ.
Horn,
Rev.
Ecol.
D. L. & Metzger, W.,
D. A., Sorensen,
1982. Soil micro
of retorted shale reclamation.
and management
organisms
In: E. F. R?dente & C. W. Cook(eds.),
Studies
Revegetation
on Oil Shale
Related
DOE/EV/04018-6.
Fort Collins.
Disturbances
Dept.
Range
in Colorado,
Sei., Colorado
pp. 27-44.
State Univ.,
36
V. F. & Shcherbakova,
T. A.,
1966. Soil
the Russian by the Indian National
Kuprevich,
from
(translated
ic Documentation
Center,
National
Agriculture).
Delhi,
Technical
in Ecological
Press, New York,
Research,
demic
N.Y.
Interpretation,
identity
In: J. B. Cragg(ed.),
Vol.
6, pp. 83-135.
sonal
in monthly
changes
Weather
Rev.
mean
sea-level
pressure
Sei., Colorado
P. W. &
for
Iyer, H. K., 1982. Permutation
techniques
data
from
block
randomized
multi-response
Commun.
Meth.
Statis.-Theor.
11: 1427-1437.
experiments.
T. & Reeves,
F. B., 1979. The role of endomycorrhi
Moorman,
analyzing
in severely
13.
Rice,
Rice,
Press,
tials
171. Intermt.
For. Range
1972.
Tech.
INT-I,
Rep.
I. R. & Slatyer,
Noble,
predict successional
recurrent
Parkinson,
Exp.
R. O.,
changes
M. K. Wali
use of vital attributes
in plant communities
43: 5-21.
Vegetatio
1979. Microbes,
mycorrhizaeand
and Coal Resource
(ed.), Ecology
Vol.
Peet,
D. L.,
simulation
129
to
subject
to
spoil.
In:
mine
Development,
York, N.Y.
Press, New
2, pp. 634 642. Pergamon
R. K. & Christensen,
N. L., 1980. Succession:
tion process. Vegetatio
43: 131 140.
Phillips,
ter
A popula
1978. Polynomial
ordination:
Field and compu
37:
testing of a new method.
Vegetatio
140.
E. F., Doerr, T. B., Grygiel,
C. E. & Biondini,
M. E.,
R?dente,
on disturbed
1984. Vegetation
and succession
establishment
Colorado.
Reel. & Reveg.
soils in northwestern
Res. 3:
153-165.
R?dente,
Stark,
376
J. M.&
soil material
and
C. E., Allerdings,
E.,
B., Grygiel,
M. E., 1982. Effect of plant species,
and cultural practices upon plant establishment
succession.
Am.
J. Bot. 66:
and
1977. Causes
in soil during
P. M.,
of decreases
old-field
1982. Nitrification
succession.
secondary
in
succes
poten
62:
Ecology
D. J., Speir, T. W., T?te, K. R., Cains, A., Meyrick,
K. F.
E. A., 1982. Restoration
of pasture after topsoil
removal: Effects on soil carbon and nitrogen mineralization,
chem.
biomass
14:475
and
enzyme
activities.
Soil.
Biol.
Bio
481.
S. & Reeves,
F. B., 1982. Effect of non-mycorrhizal
on vesicular
pioneer plant Salsola kali L. (Chenopodiaceae)
J. Bot.
arbuscular
(VAM)
(in
mycorrhizal
fungi. Am.
Schmidt,
Shavkat,
S., Khan,
T.
Biondini,
In: E. F. R?dente
& C. W.
J., 1978. History
Skujins,
G. Burns (ed.), Soil
of abiotic
Enzymes,
1982. The
Karachi,
vegetation
Pakistan.
Pak.
soil enzyme research.
I 49. Academic
pp.
In: R.
Press,
Press, London,
England.
A. A.,
1982. Ecosystem
succession
and biological
Titlyanova,
turnover. Vegetatio
50: 43 51.
P. M. & Reiners, W. A., 1975. Ecosystem
succession
Vitousek,
and nutrient retention: A hypothesis.
BioScience25:376
381.
F. E., 1938. Plant Ecology. McGraw
J. E. & Clements,
York. 601 pp.
of species and their
J. & Zedler,
Zedler,
P., 1969. Association
to microtopography
within old fields. Ecology
relationship
Weaver,
50: 432
442.
Cook
Range
J. A.,
hills around
London,
England.
and sulfatase.
Speir, T. W.& Ross, D. J., 1978. Soil phosphatase
In: R. G. Burns (ed.), Soil Enzymes,
pp. 197 250. Academic
(eds.),
Disturbances
in
Studies on Oil Shale Related
Revegetation
Colorado,
pp. 1 25. DOE/ EV/0401806.
Dept.
State Univ., Fort Collins.
Colorado
S. D. & Gadir,
of calcareous
dynamics
J. Bot. 13: 17-38.
Hill, New
E. F., Doerr,
environment.
press).
1980. The
disturbances.
D.,
Dept.
Ross,
microbial
and Utilization,
pp. 165
For. Serv. Gen.
Stn., USDA
Disturb
N.Y.
York,
in primary
386.
Am. J. Bot. 66: 14 18.
domycorrhizae
populations.
E. W. & Tueller,
P. T., 1971. Special approaches
to studies
of competition
and succession
in shrub communities.
In:
Biology
New
G. P. & Vitousek,
Robertson,
& Pausier,
Their
vs natural
disturbed
E. L. & Mallik,
M. A. B.,
residual carbohydrate
activity
58: 1297 1309.
sion. Ecology
in revegetation
in the semiarid west:
II. A
practices
to determine
on en
the effect of land disturbance
bioassay
Shrubs,
Related
DOE/EV/04018-6.
Univ., Fort Collins.
E. L., 1974. Inhibition
In
of nitrification
by vegetation;
creases
succession
and pronounced
inhibition
during
by
climax ecosystems.
In: E. L. Rice (ed.), Allelopathy.
Aca
zae
Wildland
pp.
State
6
demic
P. W., Berry, K. J., Brockwell,
P. J. & Williams,
J. S.,
tests based on multi-re
1981b. A class of non-parametric
68: 720-724.
Biometrika
sponse permutation
procedures.
Neil,
Range
Shale
45-65.
F. B., Wagner,
T. & Kiel, J., 1979. The
D., Moorman,
role of endomycorrhizae
in the
in revegetation
practices
semiarid West:
I. A comparison
of incidence of mycorrhizae
patterns.
Mielke,
Mielke,
Revegetation
in Colorado,
Reeves,
126.
109: 120
on Oil
Studies
(eds.),
ances
Aca
H. H & Vankat,
J. L., 1982. Species
removal from a
Matthew,
first year old field plant community.
63: 705-711.
Ecology
P. W., Berry, K. J. & Brier, G. W.,
198la. Application
of
Mielke,
sea
for examining
procedures
multi-response
permutation
Mon.
F. B., Reinsvold,
J. & Park, A.,
1982.
R., Saboloni,
of mycorrhizal
in regulating
disturbed
Importance
fungi
soils and retorted shale.
In: E. F. R?dente & C. W. Cook
Reeves,
for the U.S. Department
of
Information
Service,
Va. 392 pp.
Springfield,
A. N. & Buell, K. F., 1969.
Langford,
and stability
in the plant association.
Advances
Enzymes
Scientif
Sei.,
Accepted
12.10.1984.