Ecology, 80(5), 1999, pp. 1623–1631
! 1999 by the Ecological Society of America
LINKS BETWEEN MICROBIAL POPULATION DYNAMICS AND NITROGEN
AVAILABILITY IN AN ALPINE ECOSYSTEM
DAVID A. LIPSON, STEVEN K. SCHMIDT,
AND
RUSSELL K. MONSON
Department of Environmental, Population, and Organismic Biology, University of Colorado,
Boulder, Colorado 80309-0334 USA
Abstract. Past studies of plant–microbe interactions in the alpine nitrogen cycle have
revealed a seasonal separation of N use, with plants absorbing N primarily during the
summer months and microbes immobilizing N primarily during the autumn months. On
the basis of these studies, it has been concluded that competition for N between plants and
microbes is minimized along this seasonal gradient. In this study, we examined more deeply
the links between microbial population dynamics and plant N availability in an alpine dry
meadow. We conducted a year-round field study and performed experiments on isolated
soil microorganisms. Based on previous work in this ecosystem, we hypothesized that
microbial biomass would decline before the plant growing season and would release N that
would become available to plants. Microbial biomass was highest when soils were cold,
in autumn, winter, and early spring. During this time, N was immobilized in microbial
biomass. After snow melt in spring, microbial biomass decreased. A peak in the soil protein
concentration was seen at this time, followed by peaks in soil amino acid and ammonium
concentrations in late June. Soil protease rates were initially high after snow melt, decreased
to below detection limits by midsummer, and partially recovered by late summer. Proteolytic
activity in soil was saturated early in the growing season and became protein limited later
in the summer. We concluded that the key event controlling N availability to alpine plants
occurs after snow melt, when protein is released from the winter microbial biomass. This
protein pulse provides substrate for soil proteases, which supply plants with amino acids
during the growing season. On average, microbial biomass was lower in the summer than
at other times, although the biomass fluctuated widely during the summer. Within the
summer months, maximum numbers of amino-acid-degrading microorganisms and the maximum amount of microbial biomass coincided with the peak in soil amino acids, when
plants are most active. All bacterial strains isolated from this summer community had the
ability to grow rapidly on low concentrations of amino acids and to degrade protein. This
explains the previously observed result that the soil microbial biomass can compete strongly
with plants for organic N, despite the seasonal offset of maximum plant and microbial N
uptake.
Key words: alpine tundra; Kobresia myosuroides; microbial population dynamics; N uptake,
seasonal; nitrogen availability; organic nitrogen; soil microorganisms; soil protease activity; substrate-induced growth response; temporal partitioning.
INTRODUCTION
Soil microbial population dynamics have important
implications for N availability to plants in many ecosystems. For example, microbial biomass often assimilates N at times when plants are inactive, releasing the
N to plants some time later when plants are active and
microbial biomass turns over (Singh et al. 1989, Zak
et al. 1990). In the Colorado alpine, the link between
seasonal fluctuations in microbial biomass and N uptake by plants is still mysterious. Recent work (Fisk
and Schmidt 1996, Jaeger et al. 1999) shows that a
temporal partitioning of N between plants and microbes
exists in this ecosystem, wherein plants take up N in
the summer and microbes immobilize N maximally in
the autumn. This manifests as an increase in microbial
Manuscript received 3 April 1998; revised 1 July 1998;
accepted 14 July 1998.
biomass N in the fall, and implies that a release of N
from microbial biomass must occur sometime before
the plant growing season. Microbial biomass N was
observed to decline in the later stages of snow melt in
the spring, but no flush of inorganic N was seen in the
soil (Brooks et al. 1998). Brooks et al. (1996) found
that net N mineralization occurred in soil under the
snowpack during the spring; however, net immobilization of N was observed after snow melt. It is still
unknown whether a seasonal change in microbial population size causes a flush of available N to plants.
These reported patterns of seasonal N uptake by alpine plants and microbes have not yet been reconciled
with recent evidence for organic N utilization by alpine
plants. Net N mineralization can account for only a
fraction of annual plant N uptake in this ecosystem
(Bowman et al. 1993, Fisk and Schmidt 1995). The
uptake of amino acids may explain the balance of this
1623
DAVID A. LIPSON ET AL.
1624
N (Raab et al. 1996, Lipson and Monson 1998). However, we lack a description of the seasonal fluxes of
amino acids in the soil. Previous work shows that Kobresia myosuroides (Cyperaceae), the dominant plant
species of alpine dry meadows in the Rocky Mountains,
takes up most of its required N early in the summer,
when soils have warmed but are still moist (Theodose
et al. 1996, Jaeger et al. 1999). Also, K. myosuroides
competes best with microbes for amino acids under
these conditions (Lipson and Monson 1998). When
taken together, the seasonal studies and observations
of plant organic N utilization suggest the following
scenario. (1) Microbial populations decline before the
plant growing season and release organic N. (2) This
organic pulse provides N for plants. (3) N is immobilized into the soil microbial biomass after plants senesce in the fall, and is retained there throughout the
winter. To test these hypotheses, we conducted a field
study of changes in soil microbial biomass over the
year, and the flux rate of amino acids in soil (proteolysis
rate) during the plant growing season. To understand
the role of the summer microbial community in contributing to proteolysis of soil protein and its potential
to compete with plants for amino acids, we measured
population sizes of microorganisms capable of growing
with amino acids as their sole C source, and studied
the amino acid uptake and protease activities of microbes isolated using a dilution culture method. Ultimately, we hoped to understand the links between microbial population dynamics and N availability for alpine plants.
MATERIALS
AND
METHODS
Study sites and sampling regime
The study site is located at the Niwot Ridge Long
Term Ecological Research (LTER) site in the Front
Range of the Colorado Rocky Mountains, United States
(40!03" N, 105!35" W). The site is dominated by Kobresia myosuroides (Vill.) Paol. and Fiori, and the soil
is classified as a pergellic cryumbrept.
On sampling dates during the first year of the study
(1996–1997), at least three soil cores (10 cm deep, 5
cm diameter) were collected randomly across the landscape at each site. The study was extended into 1997–
1998, but because of the sensitive nature of the alpine
habitat, a less intensive regime was used. Smaller cores
(1 cm diameter) were taken across the landscape and
composited, so as to minimize disturbance to the site
while providing a single, spatially averaged sample.
Because they required a smaller amount of soil, protein
measurements in 1997 were performed on the smaller
cores before compositing, whereas substrate-induced
growth response (SIGR) measurements were performed on the composited samples. On all sampling
dates between November and April, soil was frozen
solid. On these days, soil was excavated using a trowel,
and was allowed to thaw at 3!C overnight before pro-
Ecology, Vol. 80, No. 5
cessing. Soils were coarsely sieved (mesh size 4.75
mm) to remove rocks and plant material. All population
measurements were initiated within 2 d of collection,
during which time soils were kept refrigerated.
Microbial population measurements
We chose the substrate-induced growth response
(SIGR) method as our primary measure of soil microbial biomass because it is free of some of the uncertainties inherent in other methods. When the chloroform fumigation-extraction (FE) method (Brookes et
al. 1985) is used, uncertainty arises from variations in
the effectiveness of the fumigation and extraction, and
because no guarantee may be made as to the viability
of the measured biomass. The substrate-induced respiration (SIR) method (Anderson and Domsch 1978)
provides an index of potentially active biomass, but
must be calibrated for each soil to yield an actual quantity of microbial biomass. The SIGR method is similar
to the SIR method, but utilizes the full kinetics of respiration over time to produce a measurement of microbial biomass that is essentially independent of soil
type (Schmidt 1992, Colores et al. 1996). This is the
first test of this method in a ecological field study, and
so SIR and FE were measured on a subset of samples
to validate the SIGR method.
For the SIR and SIGR measurements, 10–15 g dry
mass equivalent of each soil was placed in a biometer
flask with 1 mL NaOH (1 mol/L) in the side arm to
trap CO2 (Colores et al. 1996). Amounts of substrate
previously determined to induce maximal respiration
in these soils (2 mg glutamate-C or glucose-C/g) were
added to the flasks, with enough uniformly 14C-labeled
substrate (Sigma Biochemical, St. Louis, Missouri) to
yield 2500 Bq (150 000 dpm) per flask, and enough
deionized water to bring the soils to 60% of field capacity. The base trap was removed and replaced at various time points and radioactivity was measured by
liquid scintillation after mixing with Scintiverse II scintillation cocktail. Flasks were incubated at 22!C during
the growing season, and at 3!C during the colder
months. Respiration data were analyzed using Microsoft Excel, essentially as described by Colores et al.
(1996). The SIGR estimate of population size is calculated from the initial respiration rate, the maximum
exponential growth rate, and the growth yield. The
growth yield was calculated by the formula: Yc # (C
added $ C recovered)/C added, where ‘‘C recovered’’
is the quantity of CO2-C evolved before the soil respiration returns to its basal rate.
Fumigation-extraction measurements were performed as described by Brookes et al. (1985). Total N
in extracts was converted to NH4% using a modified
micro-Kjeldahl technique. Mixtures with 10 mL of extract, 1 mL concentrated H2SO4, and &0.5 g of catalyst
were placed in digestion tubes and heated for 1 h at
180!C, and then at 350!C until digestion was complete.
Samples were diluted to 15 mL with deionized water,
July 1999
MICROBIAL DYNAMICS AND N AVAILABILITY
and NH4% was analyzed by the phenolate-hypochlorite
method (U.S. EPA 1983) using a plate reader (Molecular Devices Corporation, Milpitas, California) to measure absorbance at 650 nm. Total organic C in extracts
was measured using a Mn(III)-reduction assay (Bartlett
and Ross 1988).
Proteolytic measurements
Past studies of protease activities in soils have always used high temperatures and excess protein to provide an index of potential activity (e.g., Watanabe and
Hayano 1995). In this study, we incubated soils at natural temperatures without exogenous protein to estimate actual rates of amino acid production in soils.
Protease activities were measured using a modification
of the method of Watanabe and Hayano (1995). The
rate at which amino acids were released from soil protein (‘‘native’’ protease rates) was measured by incubating soil with sodium citrate buffer (50 mmol/L pH
5.2) at a ratio of 15 mL/g soil. Toluene (0.4 mL) was
added to inhibit microbial uptake of amino acids. The
incubations were performed at 5!C for 6 h. Preliminary
tests showed that activities were linear over at least a
6-h period. On a subset of dates, protease measurements were also performed at higher temperatures (up
to 22!C). Initial and final aliquots were removed and
the reaction was halted with trichloroacetic/acetic acid
solution as described by Watanabe and Hayano (1995).
Total amino acids were measured using a modified ninhydrin method (Lipson and Monson 1998). Potential
protease activity was measured as described just above,
with the substitution of half the volume of the buffer
with 0.6% casein.
Soil C and N pool measurements
Within 48 h of collection, soils were extracted with
deionized water for amino acids (Kielland 1996), with
0.1 mol/L NaHCO3 for soluble proteins (Ladd and Paul
1973), and with 0.5 mol/L K2SO4 for the fumigationextraction measurements, described previously. Water
was chosen for an extractant for amino acids because
it allows concentration without precipitation problems
and produces good recoveries of neutral and acidic amino acids from soils (Paul and Schmidt 1960), and is
therefore a reasonable but conservative index of soil
amino acid concentration. Bicarbonate was chosen to
extract protein because it is a mild alkaline solvent that
does not lyse cells but solubilizes protein (Ladd and
Paul 1973). Solvent was added in a ratio of 5 mL/g
soil, and the mixture was shaken for 30 min and filtered
on Whatman number 3 filter paper. Extracts were frozen
until analysis. Amino acids and ammonium concentrations in water extracts were measured as described earlier (Lipson and Monson 1998). Ammonium was also
measured in K2SO4 extracts. Protein in NaHCO3 extracts was measured using the Bradford assay (Bradford
1976).
1625
MPN and isolation of microorganisms
Microbial isolates were obtained using a dilution culture method (Button et al. 1993). The advantage of this
method is that it allows organisms to be isolated that
might otherwise be outcompeted in laboratory enrichments with rich media. Relatively low concentrations
of substrate were supplied (1 mmol/L) so that only
ecologically relevant amino acid degraders would be
isolated and so that oligotrophs would not be overwhelmed. The mineral salt solution contained 0.14 g
K2HPO4, 0.02 g KH2PO4, 1 g MgSO4·7H2O, 0.10 g
NH4NO3, and 0.05 g CaCl2, per liter H2O. To this was
added 1 mL soil extract (10:1 volume : mass, sterile
filtered), and 0.1 mL yeast extract (1% mass : volume,
sterile filtered). Potassium glutamate or glycine (Sigma
Biochemical) was added to this media to obtain final
concentrations of 1 mmol/L. Soil (10 g dry mass equivalent) was blended in sterile MgSO4·7H2O solution (1
g/L) for 2 min, allowed to settle for 1 min, then blended
again for 2 min. A dilution series was prepared in
MgSO4 solution, and eight replicates of 0.05 mL from
each dilution were pipetted into 0.15 mL of media in
96-well plates. Most probable numbers of cells (MPN)
were calculated from the number of positive wells in
each dilution, using a computer program written in Borland Turbo C. This method yielded only bacteria. On
one date, direct counts of active bacteria were performed by the Microbial Counting Service at the Oregon State University, using fluorescene diacetate as a
viable stain (Bottomley 1994). Cells were taken from
wells in the limiting dilutions and purified by streaking
on agar plates containing the liquid media previously
described and 1.5% bacto agar.
Amino acid growth and uptake experiments on
bacterial isolates
The presence or absence of proteolytic potential was
determined by testing the ability of the purified strains
to liquefy gelatin (40 g gelatin, 40 g tryptic soy broth,
17 g tryptone, 3 g Soytone, 5 g NaCl, 2.5 g K2HPO4
per liter; Gelatin Media DifCo, Detroit, Michigan). Isolates were incubated at 22!C, and positive results were
recorded after chilling the media to 5!C for 30 min.
Growth rates of the isolates on seven amino acids were
measured by inoculating the media just described with
each isolate in a 96-well plate. Turbidity was measured
at 595 nm on a plate reader and the data were fit with
nonlinear regression to obtain an exponential growth
rate. Because glycine (gly) is a potentially important
amino acid for Kobresia myosuroides (Raab et al.
1996), the kinetics of glycine uptake by glycine dilution culture isolates were studied by measuring disappearance of glycine from solution. Flasks with 200
mL sterile 1 mmol/L potassium phosphate buffer (pH
6.0), with 200 'mol/L gly labeled with 14C-gly to
&1667 Bq (100 000 dpm) per flask. The flasks were
inoculated with centrifuged and rinsed aliquots from
DAVID A. LIPSON ET AL.
1626
cultures of each isolate growing in glycine media. Aliquots were removed regularly and filtered (0.2 'm pore
size, Millipore, Bedford, Massachusetts). Samples
were counted by liquid scintillation after the addition
of a drop of H2SO4 to remove dissolved CO2. Concentration vs. time data were converted to rate vs. concentration and the half-saturation constant (KM) for glycine was fit using nonlinear regression (MODEL procedure, SAS). To get the best fit, data from all isolates
were combined after uptake rates were normalized to
the initial uptake rate, when concentration change was
approximately linear.
Ecology, Vol. 80, No. 5
TABLE 1. Pearson correlation coefficients (r) between measures of microbial population size (gcs SIR, glucose substrate-induced respiration; glu SIGR, glutamate substrateinduced growth response; FE-C, fumigation-extraction carbon; FE-N, fumigation–extraction nitrogen).
Measure
Measure
gcs SIR
FE-C
FE-N
glu SIGR
gcs SIR
FE-C
0.785
0.834
0.933
0.895
0.906
0.866
Notes: The sample sizes (n) range from 22 to 28. All relationships are significant at the P # 0.0001 level.
Statistical analyses
RESULTS
Seasonal changes in microbial biomass were tested
by regressing microbial biomass indices on sampling
date, using time as a continuous variable. This approach
was used because (1) our hypotheses concerned
changes over time in general, rather then pairwise comparisons between specific dates, and (2) this approach
is simpler and more powerful in this case, because it
uses a statistical model with fewer parameters. To test
whether biomass was higher, on average, in the colder
months than during the growing season, we used an
ANOVA, with sampling date as a categorical variable.
Only the 1996–1997 data were used in this ANOVA.
Pairwise differences between dates in protease activity,
MPN, and soil chemical pools were tested using Tukey’s multiple-range test. Because we wished to know
whether the extractable soil protein concentration was
highest on the days in 1996 and 1997 immediately after
the post-snow melt decline of biomass, we contrasted
these two days against the 10 others in the full ANOVA.
The Bonferroni inequality was applied to the critical
P value to adjust for this post hoc test. The ratio of
native to potential protease activity was too variable
to reveal significant differences using date as a categorical variable in an ANOVA, but declined significantly when the ratio was regressed on time as a continuous variable. In this case, the ratio on each date
was also tested against the null hypothesis that the
mean ratio was not different from 1.0, using the TEST
option of the REG procedure of SAS programming
language (SAS Institute 1990). Each data set was tested
for heteroscedasticity by plotting error against predicted values. Because errors were generally proportional to means, most of the analyses required transformation of the data. When unequal variances were
detected, data were log- or power-transformed before
analysis (Judd and McClelland 1989). For clarity, these
data are shown on a linear scale, before transformation.
On 7 May 1996, an outlier was omitted. This was because the sample was inadvertantly wetted with snow
to near-saturation, above the optimal moisture content
for SIR, and was unusually low. The outlier was omitted from analysis (this did not alter the outcome), but
is shown in Fig. 2.
Seasonal changes in microbial populations and soil
N pools
To validate the SIGR method, gcs SIR (glucose substrate-induced respiration) and FE measurements were
taken on a subset of samples. All these measurements
were highly correlated (Table 1), and showed the same
general seasonal patterns (Fig. 1). Biomass decreased
initially after snow melt, fluctuated dynamically
throughout the growing season, and generally increased
through the fall, winter, and spring. Although biomass
was highly temporally variable in the growing season,
it was more spatially variable during the winter. The
results of the ANOVA (F # 6.87; df # 8, 17; P (
0.001) indicate that the glu SIGR (glutamate substrateinduced growth response) was higher, on average, in
the colder months than during the growing season (F
# 17.73; df # 1, 17; P ( 0.001).
The decline of biomass during the first month after
snow melt was observed in the spring of all three years
(Fig. 2). The biomass (SIGR) increased between midsummer and the following spring in both 1996–1997
and 1997–1998 (Fig. 3A). Also over this interval in
1996–1997, FE-C and FE-N increased (Fig. 3B), and
the C:N ratio of the chloroform-labile fraction of the
microbial biomass decreased (Fig. 3C).
In 1996 and 1997, there were peaks in soil NaHCO3extractable protein (Fig. 4A) that coincided with the decreases in microbial biomass after snow melt. Immediately after this in 1996, water-extractable amino acids and
K2SO4-extractable NH4% peaked (Fig. 4B,C). (Amino acids and ammonium were not measured in 1997.)
During May 1996, when extractable protein levels
were at their peak, native protease activities were also
maximal (Fig. 5A). Soil protein levels were sufficient
to saturate soil protease activities in the early part of the
growing season, but the ratio of native to potential protease activity decreased throughout the summer, indicating the onset of substrate limitation (Fig 5B). Protease
activity dropped to below detection limits in late July,
but was again measurable in late summer (Fig. 5A).
During June, when extractable amino acid pools
were maximal, populations (most probable numbers,
MPN) capable of growth on glycine or glutamate were
July 1999
MICROBIAL DYNAMICS AND N AVAILABILITY
1627
MPN were 5% of the value of the direct counts. Of the
bacterial isolates obtained in the summer from glutamate dilution culture, those five that were tested grew
well on all of the seven amino acids, except for one
isolate that grew on all but glycine (Table 2). A halfsaturation constant (KM) for glycine uptake of 17.4 )
3.5 'mol/L was fit from data of six isolates from glycine dilution culture (Fig. 6). All of the eight isolates
tested had the ability to liquefy gelatin.
DISCUSSION
Microbial biomass (SIGR, SIR, and FE) reached its
peak during the spring and declined after the snowpack
retreated. Biomass was highly variable during the plant
growing season and increased in the fall. These results
are consistent with all of the previously observed dynamics of FE-N in this system (Fisk and Schmidt 1995,
Brooks et al. 1998, Jaeger et al. 1999). This corroboration is not surprising, given the high correlations
among SIGR, SIR, FE-C, and FE-N found in the present
study.
Although interpretation of temporal fluctuations in
the microbial biomass during the growing season is
beyond the scope of this paper, one consistent and highly interpretable feature is the decline after snow melt.
FIG. 1. Seasonal changes of microbial biomass indices in
1996–1997: (A) glutamate substrate-induced growth response, glu SIGR; (B) glucose and glutamate substrate-induced respiration, gcs and glu SIR; and (C) fumigation-extraction carbon and nitrogen FE-C and FE-N. Values are
means ) 1 SE.
also high (Fig. 5B). On the date that glu MPN peaked,
microbial biomass also reached its maximum value for
the growing season (Fig. 1). MPN dropped sharply by
late July and partially recovered in the early fall.
Amino acid growth and uptake experiments on
bacterial isolates
A comparison of MPN and direct counts of active
bacteria on a single day in the fall showed that glu
FIG. 2. Decline of microbial biomass during the first
month after snow melt in 1996, 1997, and 1998, as indicated
by (A) SIGR and (B) SIR, both measured at 22!C. The site
was free from snow on 7 May 1996, 4 June 1997, and 26
May 1998. The asterisk indicates an outlier that was accidentally saturated with snow during collection.
1628
DAVID A. LIPSON ET AL.
Ecology, Vol. 80, No. 5
FIG. 3. Changes in microbial biomass and C:N ratio following the end of the growing season: (A) glu SIGR in 1996–
1997 and 1997–1998, carried out at 22!C up to November
and at 3!C after November; (B) FE-C and FE-N in 1996–
1997; and (C) C:N ratio of the chloroform-labile fraction in
1996–1997.
A pulse of extractable soluble protein was simultaneously observed in the soil, and amino acid and ammonium levels peaked shortly after. Measurable proteolysis of soil protein was detected during this time.
This period is typically when N uptake by plants is
maximal (Theodose et al. 1996, Jaeger et al. 1999).
These patterns are all consistent with our original hypothesis that microbial biomass decreases at the start
of the growing season and releases plant-available N.
FIG. 4. Seasonal changes in (A) soil NaHCO3-extractable
protein; (B) H2O-extractable amino acids; and (C) K2SO4-extractable NH4%. The site was free from snow on 7 May in 1996
and on 4 June in 1997. Dates with the same letter are not
significantly different by Tukey’s multiple-range test. The two
dates marked with an asterisk (in panel A) were significantly
different (P # 0.0001) from the rest in an ANOVA, using the
Bonferroni adjustment (critical P # 0.004) for a post hoc contrast (overall ANOVA: F # 7.26; df # 11, 25; P # 0.0001).
Protein and amino acid data were power-transformed, and ammonium data were log-transformed before analysis.
July 1999
MICROBIAL DYNAMICS AND N AVAILABILITY
FIG. 5. (A) Rates of proteolysis of native soil protein at
5!C in 1996. (B) Log of most probable numbers (MPN) of
bacteria capable of growth on the amino acids glutamate or
glycine. (C) Ratio of native to potential protease rates. For
(A) and (B), within each time series, dates with the same
letter are not significantly different. For (C), the ratio declines
significantly over time, as indicated, although individual dates
do not differ significantly. The probability value (P) for the
test that the average ratio is different from 1.0 on that date
is given next to each point.
1629
The most reasonable explanation for the high level of
extractable soil protein after the biomass decline is that
it was released from microbial biomass. The fact that
the ratio of native to potential proteolytic activity decreased during the summer is evidence that protease
activity relies on the early-season pulse of protein. The
abundance of substrate (protein), enzymatic activity
(proteolysis), and product (amino acids) clearly shows
a burst of available N in the late spring and early summer, linked to the decline of the soil microbial biomass
after snow melt. The accumulation of amino acids in
the soil solution occurred while plants were active and
microbial populations capable of growing on amino
acids were high. Therefore, it is unlikely that the high
soil amino acid concentrations were due to a decreased
sink strength for amino acids in the soil.
The conclusion of a past study (Jaeger et al. 1999),
that plants and soil microbes are most active in N uptake at different times of the year, is confirmed by the
present work. Microbial biomass (SIGR) was higher
during the colder months than during the growing season. However, this fact does not imply that all competition between plants and microbes is avoided. Microbial biomass exhibited transient peaks during the
summer, and microbial populations capable of growth
on amino acids coincided with the flush of available N
after snow melt. All strains cultured during the growing
season produced proteases and grew rapidly on low
concentrations of amino acids. Based on their proteolytic abilities and rapid growth rates on low concentrations of amino acids, these organisms could act both
as competitors for and suppliers of amino acids and
inorganic N. It is frequently the case that bacteria isolated in the laboratory bear little relation to those active
in the field (Amann et al. 1995). However, our culturing
techniques appear to be relatively effective; most probable number (MPN) measurements using these culturing techniques are 5% as high as direct counts of active
bacteria. This ratio is an order of magnitude higher
than those reported for most soils, 0.25–0.3% (Amann
et al. 1995). Because these isolates were obtained in
dilution culture, at least we are confident that the isolates studied were random representatives of the &108
cells/g of soil measured by MPN. The large populations
of amino acid-degrading microbes in the summer explain the results of a previous study in which it was
found that soil microbes competed strongly for glycine
with Kobresia myosuroides (Lipson and Monson
1998). What mechanisms allow plants to coexist with
these microbes is an important question for future
study.
The bacterial isolates obtained in this study probably
contribute to proteolytic activity in the soil. However,
mycorrhizal fungi might also be responsible for some
of the soil protease activity. K. myosuroides associates
with ectomycorrhizal fungi (Gardes and Dahlberg
1996), and ectomycorrhizae can transfer N from com-
DAVID A. LIPSON ET AL.
1630
Ecology, Vol. 80, No. 5
TABLE 2. Generation times at 22!C for bacteria isolated from glutamate dilution culture growing on seven amino acids. The
values are means of five isolates, except for glycine (for which n # 4).
Mean generation time (h)
1 SE
ala
arg
glu
gly
pro
trp
tyr
10.2
1.01
11.85
1.05
9.12
1.40
21.41
5.68
11.88
1.53
12.39
1.48
14.51
2.91
Note: The seven amino acids were alanine, arginine, glutamic acid, glycine, proline, tryptophan, and tyrosine.
plex organic forms directly to plants (Stribley and Read
1980).
Microbial biomass increased throughout the fall and
winter. The increase in microbial biomass N and the
decrease in the C/N ratio of the microbial biomass over
this interval is consistent with the hypothesis that microbes immobilize N after plants senesce in the fall and
retain it through the winter. Microbial activity at subzero temperatures has been reported in the alpine
(Brooks et al. 1996) and in the arctic (Clein and Schimel 1995). The finding that microbial biomass increases during the fall and winter is consistent with previous
results that showed this alpine microbial community to
be resistant to freeze–thaw events (Lipson and Monson,
1998). Because of this property of the alpine soil microbial community, it is unlikely that the decrease in
biomass after snow melt is due solely to freeze–thaw
events in the spring. Spatial variability in the winter
microbial biomass was notable, suggesting that the
winter microbial biomass is susceptible to some stochastic factors such as litter deposition or moisture content. Similarly, the fluctuations in biomass during the
summer could be caused by changes in soil moisture
and C availability. The lowest biomass level was observed at midsummer, when soil moisture was at a minimum (D. A. Lipson, S. K. Schmidt, and R. K. Monson,
unpublished data), suggesting that moisture could control microbial biomass in the summer through diffusive
effects on substrate availability or through direct osmotic effects (Stark and Firestone 1995). Other factors
that might prove to be important in explaining the post-
FIG. 6. Glycine uptake by six bacteria isolated in glycine
dilution culture. Data from each isolate were normalized and
combined by dividing uptake rate by the inital quasi-linear
rate. The combined data were fit by nonlinear regression to
estimate the half-saturation constant, KM, for glycine uptake.
snow melt decline include protozoal grazing or a microbial community shift due to changes in substrate
quality and temperature. Whatever the mechanism is,
this event appears to be crucial to plant N nutrition in
Colorado alpine dry meadows.
ACKNOWLEDGMENTS
This work was funded by NSF grant 9514123 to R. K.
Monson and S. K. Schmidt. Logistical support was provided
by the Niwot Ridge LTER project (NSF DEB 9211776) and
the Mountain Research Station (BIR 9115097). Thanks to Dr.
John Blair for his patient handling of this manuscript, and to
three anonymous reviewers for numerous useful observations.
LITERATURE CITED
Amann, R. I., W. Ludwig, and K.-L. Schleifer. 1995. Phylogenetic identification and in situ detection of individual
microbial cells without cultivation. Microbiological Reviews 59:343–369.
Anderson, J. P. E., and K. H. Domsch. 1978. A physiological
method for the quantitative measurement of microbial biomass in soils. Soil Biology and Biochemistry 10:215–221.
Bartlett, R. J., and D. S. Ross. 1988. Colorimetric determination of oxidizable carbon in acid soil solutions. Soil
Science Society of America Journal 52:1191–1192.
Bottomley, P. J. 1994. Light microscopic methods for studying soil microorganisms. Pages 81–104 in R. W. Weaver,
S. Angle, P. Bottomley, D. Bezdicek, S. Smith, A. Tabatabai, and A. Wollum, editors. Methods of soil analysis,
part 2. Soil Science Society of America, Madison, Wisconsin, USA.
Bowman, W. D., T. A. Theodose, J. C. Schardt, and R. T.
Conant. 1993. Constraints of nutrient availability on primary production in two alpine tundra communities. Ecology 74:2085–2097.
Bradford, M. M. 1976. A rapid and sensitive method for the
quantitation of nanogram quantities of protein using the
principle of protein-dye binding. Analytical Biochemistry
72:248–254.
Brookes, P. C., A. Landman, G. Pruden, and D. S. Jenkinson.
1985. Chloroform fumigation and the release of soil nitrogen: a rapid direct extraction method to measure microbial biomass nitrogen in soil. Soil Biology and Biochemistry 17:837–842.
Brooks, P. D., M. W. Williams, and S. K. Schmidt. 1996.
Microbial activity under alpine snowpacks, Niwot Ridge,
Colorado. Biogeochemistry 32:93–113.
Brooks, P. D., M. W. Williams, and S. K. Schmidt. 1998.
Inorganic nitrogen and microbial biomass dynamics before
and during spring snowmelt. Biogeochemistry 43:1–15.
Button, D. K., F. Schut, P. Quang, R. Martin, and B. R. Robertson. 1993. Viability and isolation of marine bacteria by
dilution culture: theory, procedures, and initial results. Applied and Environmental Microbiology 59:881–891.
Clein, J. S., and J. P. Schimel. 1995. Microbial activity of
tundra and taiga soils at sub-zero temperatures. Soil Biology and Biochemistry 27:1231–1234.
Colores, G., M. C. Fisk, and S. K. Schmidt. 1996. Estimating
the biomass of microbial functional groups using rates of
July 1999
MICROBIAL DYNAMICS AND N AVAILABILITY
growth-related soil respiration. Soil Biology and Biochemistry 28:1569–1577.
Fisk, M. C., and S. K. Schmidt. 1995. Nitrogen mineralization and microbial biomass nitrogen dynamics in three alpine tundra communities. Soil Science Society of America
Journal 59:1036–1043.
Fisk, M. C., and S. K. Schmidt. 1996. Microbial responses
to nitrogen additions in alpine tundra soils. Soil Biology
and Biochemistry 28:751–755.
Gardes, M., and A. Dahlberg. 1996. Mycorrhizal diversity
in arctic and alpine tundra: an open question. New Phytologist 133:147–157.
Jaeger, III C. H., R. K. Monson, M. C. Fisk, and S. K.
Schmidt. 1999. Seasonal partitioning of nitrogen by plants
and soil microorganisms in an alpine ecosystem. Ecology
80, in press.
Judd, C. M., and G. H. McClelland. 1989. Data analysis: a
model-comparison approach. Harcourt Brace-Jovanovich,
Orlando, Florida, USA.
Kielland, K. 1996. Landscape patterns of free amino acids
in arctic tundra soils. Biogeochemistry 31:85–98.
Ladd, J. N., and E. A. Paul. 1973. Changes in enzymatic
activity and distribution of acid-soluble, amino acid-nitrogen in soil during nitrogen immobilization and mineralization. Soil Biology and Biochemistry 5:825–840.
Lipson, D. A., and R. K. Monson. 1998. Plant–microbe competition for soil amino acids in the alpine tundra: effects
of freeze–thaw and dry–rewet events. Oecologia 113:406–
414.
Paul, E. A., and E. L. Schmidt. 1960. Extraction of free
amino acids from soil. Soil Science Society of America
Proceedings 24:195–198.
Raab, T. K., D. A. Lipson, and R. K. Monson. 1996. Nonmycorrhizal uptake of amino acids by roots of the alpine
sedge Kobresia myosuroides: implications for the alpine
nitrogen cycle. Oecologia 108:488–494.
1631
SAS Institute. 1990. SAS/SYSTAT user’s guide. Version 6,
fourth edition. SAS Institute, Cary, North Carolina, USA.
Schmidt, S. K. 1992. A substrate-induced growth response
method for estimating the biomass of microbial functional
groups in soil and aquatic systems. FEMS Microbiology
Ecology 101:197–206.
Singh, J. S., A. S. Raghubanshi, R. S. Singh, and S. C. Srivastava. 1989. Microbial biomass acts as a source of plant
nutrients in dry tropical forest and savanna. Nature 338:
499–500.
Stark, J. M., and M. K. Firestone. 1995. Mechanisms for
soil moisture effects on activity of nitrifying bacteria. Applied and Environmental Microbiology 61:218–221.
Stribley, D. P., and D. J. Read. 1980. The biology of mycorrhizae in the Ericaceae: VII. The relationship between
mycorrhizal infection and the capacity to utilize simple and
complex organic nitrogen sources. New Phytologist 86:
365–371.
Theodose, T. A., C. H. Jaeger III, W. D. Bowman, and J. C.
Schardt. 1996. Uptake and allocation of 15N in alpine
plants: implications for the importance of competitive ability in predicting community structure in a stressful environment. Oikos 75:59–66.
U.S. EPA (United States Environmental Protection Agency).
1983. Methods for chemical analysis of water and wastes.
EPA-600/4-79-02a Method 351.2, U.S.E.P.A., Cincinnati,
Ohio, USA.
Watanabe, K., and K. Hayano. 1995. Seasonal variation of
soil protease activities and their relation to proteolytic bacteria and Bacillus spp. in paddy field soil. Soil Biology and
Biochemistry 27:197–203.
Zak, D. R., P. M. Groffman, K. S. Pregitzer, S. Christensen,
and J. T. Tiedje. 1990. The vernal dam: plant–microbe
competition for nitrogen in northern hardwood forests.
Ecology 71:651–656.