1 23
Biogeochemistry
An International Journal
 
ISSN 0168-2563
 
Biogeochemistry
DOI 10.1007/s10533-013-9848-y
The response of microbial biomass and
hydrolytic enzymes to a decade of nitrogen,
phosphorus, and potassium addition in a
lowland tropical rain forest
Benjamin L.?Turner & S.?Joseph Wright
1 23
Your article is protected by copyright and all
rights are held exclusively by US Government.
This e-offprint is for personal use only
and shall not be self-archived in electronic
repositories. If you wish to self-archive your
article, please use the accepted manuscript
version for posting on your own website. You
may further deposit the accepted manuscript
version in any repository, provided it is only
made publicly available 12 months after
official publication or later and provided
acknowledgement is given to the original
source of publication and a link is inserted
to the published article on Springer's
website. The link must be accompanied by
the following text: "The final publication is
available at link.springer.com?.
The response of microbial biomass and hydrolytic enzymes
to a decade of nitrogen, phosphorus, and potassium addition
in a lowland tropical rain forest
Benjamin L. Turner ? S. Joseph Wright
Received: 1 November 2012 / Accepted: 4 April 2013
 US Government 2013
Abstract Nutrient availability is widely considered to
constrain primary productivity in lowland tropical
forests, yet there is little comparable information for
the soil microbial biomass. We assessed microbial
nutrient limitation by quantifying soil microbial biomass
and hydrolytic enzyme activities in a long-term nutrient
addition experiment in lowland tropical rain forest in
central Panama. Multiple measurements were made over
an annual cycle in plots that had received a decade of
nitrogen, phosphorus, potassium, and micronutrient
addition. Phosphorus addition increased soil microbial
carbon (13 %), nitrogen (21 %), and phosphorus
(49 %), decreased phosphatase activity by *65 % and
N-acetyl b-glucosaminidase activity by 24 %, but did
not affect b-glucosidase activity. In contrast, addition of
nitrogen, potassium, or micronutrients did not signifi-
cantly affect microbial biomass or the activity of any
enzyme. Microbial nutrients and hydrolytic enzyme
activities all declined markedly in the dry season, with
the change in microbial biomass equivalent to or greater
than the annual nutrient flux in fine litter fall. Although
multiple nutrients limit tree productivity at this site, we
conclude that phosphorus limits microbial biomass in
this strongly-weathered lowland tropical forest soil. This
finding indicates that efforts to include enzymes in
biogeochemical models must account for the dispropor-
tionate microbial investment in phosphorus acquisition
in strongly-weathered soils.
Keywords Gigante Peninsula  b-glucosidase 
Fertilization  Microbial biomass  N-acetyl
b-glucosaminidase  Panama  Phosphodiesterase 
Phosphomonoesterase
Introduction
Nutrient availability is widely considered to constrain
primary productivity in lowland tropical forests
(Vitousek 1984; Tanner et al. 1998) and is likely to
regulate the response of these ecosystems to increas-
ing atmospheric carbon dioxide concentrations
(Cleveland et al. 2011). For example, recent evidence
indicates that nutrient availability will modulate the
physiological responses of tropical tree seedlings to
future changes in climate and atmospheric chemistry
(Cernusak et al. 2007, 2009), while nitrogen, phos-
phorus, and potassium can all affect components of
plant productivity in lowland tropical forests (Wright
et al. 2011; Santiago et al. 2012).
In contrast, there is little information on the nutrient
status of soil microbial communities in tropical
forests. It is often assumed that heterotrophic soil
microbes are limited by the availability of suitable
organic carbon substrate (Wardle 1992). However,
nitrogen addition increased soil microbial biomass in
B. L. Turner (&)  S. Joseph Wright
Smithsonian Tropical Research Institute,
Apartado 0843-03092, Balboa, Ancon,
Republic of Panama
e-mail: turnerbl@si.edu
123
Biogeochemistry
DOI 10.1007/s10533-013-9848-y
Author's personal copy
lower montane tropical forests of Puerto Rico and
Panama (Li et al. 2006; Corre et al. 2010; Cusack et al.
2011), while phosphorus appears to regulate microbial
processes in lowland tropical forests (Cleveland et al.
2002; Cleveland and Townsend 2006) and plantations
(Gnankambary et al. 2008). These findings support the
assumption that nitrogen is the limiting nutrient in
montane forests and phosphorus the most limiting in
lowland forests (Vitousek 1984; Tanner et al. 1998),
but do not account for the marked variation in soil
properties, including nutrient concentrations, that
occur in tropical regions (Baillie 1996; Andersen
et al. 2010). The role of potassium, other base cations,
and micronutrients in regulating microbial processes
has not been addressed.
The nutrient status of soil microbial communities
can be inferred from measurements of hydrolytic
enzymes, because investment in enzyme synthesis is
assumed to reflect biological nutrient demand. For
example, soil phosphatase activity provides a sensitive
measure of phosphorus demand, being reduced mark-
edly by phosphorus addition (Marklein and Houlton
2012) and responding closely to changes in phospho-
rus availability along natural gradients in both tem-
perate and tropical rain forests (Olander and Vitousek
2000; Allison et al. 2007). Similarly, enzymes
involved in the acquisition of carbon (e.g., b-gluco-
sidase, involved in cellulose degradation) and nitrogen
(e.g., N-acetyl b-glucosaminidase, involved in chitin
degradation) can be used to infer microbial nutrient
demand for carbon and nitrogen, respectively (e.g.,
Sinsabaugh and Moorhead 1994).
It was recently suggested that enzymes involved in
carbon, nitrogen, and phosphorus acquisition are
synthesized by microbes in a consistent stoichiometric
ratio across ecosystems (Sinsabaugh et al. 2009).
Although this appears to be the case for freshwater
systems, it seems unlikely to occur in tropical soils,
given the markedly greater phosphatase activity
compared to other enzymes in such systems (Olander
and Vitousek 2000; Turner 2010; Turner and Romero
2010; Cusack et al. 2011; Nottingham et al. 2012).
There remains, however, a remarkable lack of data on
hydrolytic enzyme activities in lowland tropical
forests, particularly in terms of the response to
experimental nutrient manipulation at the field scale.
Here we report the response of soil microbial
biomass and hydrolytic enzyme activities to a decade
of fertilization with nitrogen, phosphorus, potassium,
and micronutrients in a lowland tropical rain forest in
central Panama. We hypothesized that addition of the
limiting nutrient would result in (1) an increase in soil
microbial biomass, as well as the concentration of the
nutrient in the microbial pool, and (2) a reduction in
the investment in hydrolytic enzymes associated with
the acquisition of the limiting nutrient. Our aim was to
use measurements of microbial biomass and hydro-
lytic enzymes to determine the microbial nutrient
status of this lowland tropical forest soil.
Methods
The Gigante fertilization experiment
The study was conducted on the Gigante Peninsula,
part of the Barro Colorado Nature Monument, Repub-
lic of Panama. The site supports mature moist semi-
deciduous rain forest at least 200 years old (Wright
et al. 2011). On nearby Barro Colorado Island, the
mean annual temperature is 26 C and annual rainfall
averages 2,600 mm, with just 10 % falling during the
4 month dry season between December and April
(Windsor 1990). Soils on Gigante Peninsula are clay-
rich Oxisols developed on Miocene basalt and are
morphologically similar to the Typic Eutrudox (AVA
and Marron soil classes) on nearby Barro Colorado
Island (Dieter et al. 2010). Soil analyses of control
plots indicates very low concentrations of readily
available phosphate (\1 mg P kg-1, extracted in
Mehlich-3 solution or by anion-exchange membranes;
Turner et al. 2013) and extractable inorganic nitrogen
(e.g., ammonium and nitrate concentrations of
*1 mg N kg-1 in samples extracted in the field;
B.L. Turner, unpublished). Potassium concentrations
(*100 mg K kg-1) are moderate for the central
Panama region and tropical forests worldwide (Bart-
hold et al. 2008; Yavitt et al. 2011). Foliar nutrient
concentrations indicate that the site is relatively rich in
phosphorus (Sayer et al. 2012), yet fine litter fall
increased by *25 % following phosphorus addition
(Wright et al. 2011; see below).
The fertilization experiment began in 1998 and is
described in detail elsewhere (Wright et al. 2011;
Yavitt et al. 2011; Sayer et al. 2012; Turner et al.
2013). The experiment is remarkable in its duration
and in the range of responses observed in terms of
forest productivity, with significant effects involving
Biogeochemistry
123
Author's personal copy
nitrogen, phosphorus, and potassium (Wright et al.
2011). Briefly, fertilizer has been added four times per
year during the wet season to 40 9 40 m plots
replicated four times for each treatment in a full
factorial design. Nitrogen is added as urea, phosphorus
as triple superphosphate, and potassium as potassium
chloride. A further treatment (four replicate plots)
includes calcium and magnesium (from dolomitic
limestone) and micronutrients. However, we assume
that calcium, magnesium, and micronutrients do not
limit the microbial community at this site, because the
dolomite/micronutrient treatment did not significantly
affect microbial biomass or enzyme activities
(p [ 0.15; results not shown). Fertilizer was added
at the following times during the study period: 13?15
September 2006, 16?18 May 2007, 4?6 July 2007,
22?24 August 2007, and 10?12 October 2007 (Fig. 1).
Summary of changes in response to nutrient
addition
Changes in soil nutrients were reported previously
(Yavitt et al. 2009; Yavitt et al. 2011; Turner et al.
2013). Phosphorus addition increased available phos-
phorus markedly from \1 to [30 mg P kg-1. Nitro-
gen addition increased soil nitrate, had no effect on
extractable ammonium, and decreased soil pH (*0.8
units in plots receiving only nitrogen), with a corre-
sponding decline in extractable soil cations. Extract-
able potassium increased in the potassium treatments
but declined in the nitrogen treatments, almost
certainly due to the acidification of soils under urea
application. There was significant seasonal variation
in soil pH and all nutrients measured, although
patterns varied among nutrients (Turner et al. 2013).
In addition to the increase in fine litter fall in
response to phosphorus addition, tree basal area growth
was influenced by a combination of nitrogen and
potassium (amelioration of a long-term decline), while
fine root growth was reduced by potassium addition
(Wright et al. 2011; Yavitt et al. 2011; Santiago et al.
2012). At the seedling stage, potassium and phospho-
rus appear to play key roles, influencing growth and
rates of herbivory (Santiago et al. 2012). There is
currently no information on the nutrient status of the
soil microbial community, although multiple nutrients
s s s s s s s s s s s s s
R
ai
nf
al
l (m
m 
d-1
)
0
50
100
150
Wet season
2006 2007
Sampling
Fertilizer
Nov Jan Mar May Jul Sep Nov
So
il w
at
er
 (g
 H
2O
 g
-
1  
dr
y 
so
il)
0.2
0.4
0.6
0.8
1.0
Control
Nitrogen
Phosphorus
N + P
a
b
Fig. 1 Daily rainfall for
Barro Colorado Island,
Panama (a), and soil water
content in experimental
plots on Gigante Peninsula
(b) between October 2006
and December 2007. The
dates of fertilizer addition
and sampling, and the
approximate extent of the
wet season are shown (From
Turner et al. 2013). Error
bars are standard errors of
four replicate plots in each
treatment
Biogeochemistry
123
Author's personal copy
influence litter decomposition (Kaspari et al. 2008) and
microbial community composition varies among
nutrient treatments (Kaspari et al. 2010). However,
determination of microbial carbon and nitrogen after
6 years of nutrient addition yielded no significant
differences among treatments (Sayer et al. 2012).
Soil sampling
Soils were sampled monthly between November 2006
and November 2007 (Fig. 1). On the third week of
every calendar month, soils were collected from 16
plots in an N 9 P factorial design (control, ?N, ?P,
?NP). Every 4 months all 36 plots were sampled,
including 32 plots in the full N 9 P 9 K factorial
design plus the four dolomite/micronutrient plots.
Soils were sampled to 10 cm depth using a 2.5 cm
diameter corer at nine systematically distributed
points in the central 20 9 20 m quadrat of each plot.
Samples from each plot were composited in the field
and returned immediately to the laboratory, where
roots, stones, and mesofauna were removed by hand.
Soils were then stored overnight at 4 C before
extraction of microbial biomass nutrients within 24 h
of sampling and hydrolytic enzyme activities within
48 h of sampling. A previous study showed no
significant effect of cold storage on soil microbial
phosphorus or hydrolytic enzyme activities for at least
2 weeks (Turner and Romero 2010).
During the study period, the dry season began at the
end of December 2006 and continued until late April
2007, although there were a few days of heavy rain in
early April (Fig. 1a; Turner et al. 2013). The maxi-
mum daily rainfall was 147 mm in November 2006.
Soil moisture was highest (*0.8 g H2O g-1 soil) in
November 2006, when sampling occurred immedi-
ately following the highest rainfall event, but was
otherwise relatively stable throughout the wet season
(0.60?0.75 g H2O g
-1 soil) (Fig. 1b). Dry season soil
moisture was lowest in March when values ranged
between 0.34 and 0.40 g H2O g
-1 soil.
Analytical procedures
Microbial biomass
Microbial biomass carbon and nitrogen were deter-
mined by chloroform fumigation and extraction in
K2SO4 by standard procedures (Brookes et al. 1985;
Vance et al. 1987). Briefly, fresh samples of fumigated
(24 h in a desiccator with ethanol-free chloroform)
and unfumigated soil were shaken in a 1:5 soil to
solution ratio with 0.5 M K2SO4 for one hour and then
centrifuged at 8,0009g for 10 min. Extracts were then
frozen at -35 C until analysis. Organic carbon and
total nitrogen were determined simultaneously on a
TOC-VCHN analyzer (Shimadzu, Columbia, MD) after
five-fold dilution of the extracts. The difference in
organic carbon and total nitrogen in fumigated and
unfumigated samples was assumed to originate from
microbial biomass. Both microbial carbon and micro-
bial nitrogen concentrations were corrected for unre-
covered biomass using a k factor of 0.45 (Jenkinson
et al. 2004).
Microbial phosphorus was determined by hexanol
fumigation and extraction with anion-exchange mem-
branes based on a method described by Kouno et al.
(1995) as modified in Turner and Romero (2010).
Briefly, anion exchange membrane strips were pre-
pared by shaking twice in 0.5 M NaHCO3. For each
sample, two portions of fresh soil (5 g on a dry-weight
basis) were weighed into 120 mL bottles with 80 mL
deionized water and five anion-exchange membrane
strips. One bottle received 1 mL of hexanol and the
samples were shaken for 24 h. The membranes were
then removed and rinsed in deionized water and the
phosphate recovered by shaking for 1 h in 50 mL of
0.25 M H2SO4, with detection by automated molyb-
date colorimetry at 880 nm using a Lachat Quikchem
8500 (Hach Ltd, Loveland, CO). Microbial phosphorus
was calculated as the difference between the fumigated
and unfumigated samples and corrected for unrecov-
ered biomass using a kp factor of 0.40 (Jenkinson et al.
2004). We did not correct for phosphate sorption
during the assay, because preliminary studies showed
that the majority of the phosphate spike was recovered
by the membranes in control soils, but that variable
results were obtained for phosphorus addition soils.
Ratios of C:N, N:P, and C:P in soil microbial biomass
are expressed on a molar basis.
Hydrolytic enzyme assays
The activities of four hydrolytic enzymes were
determined using fluorogenic substrates as described
previously (Turner 2010; Turner and Romero 2010).
The enzymes and substrates were: (i) acid
Biogeochemistry
123
Author's personal copy
phosphomonoesterase (Enzyme Commission (EC)
number 3.1.3.2) assayed with 4-methylumbelliferyl
phosphate; (ii) phosphodiesterase (EC 3.1.4.1)
assayed with bis-(4-methylumbelliferyl) phosphate;
(iii) b-glucosidase (EC 3.2.1.21) assayed with 4-meth-
ylumbelliferyl b-D-glucopyranoside; and (iv) N-acetyl
b-D-glucosaminidase (EC 3.2.1.52) assayed with
4-methylumbelliferyl N-acetyl b-D-glucosaminide.
Substrates were purchased from Glycosynth Ltd
(Warrington, UK) and were dissolved in 0.4 % meth-
ylcellosolve (2-methoxyethanol; 0.1 % final concen-
tration in the assay). For each sample, soil suspensions
were prepared in a 1:100 soil to water ratio (containing
1 mM NaN3 to inhibit microbial activity) by stirring
rapidly on a magnetic stir-plate for 15 min. Soil
suspension (50 lL) was then pipetted into wells on a
micro-well plate (16 wells per substrate) containing
100 lL of 200 lM substrate dissolved in deionized
water and 50 lL of 200 mM sodium acetate?acetic
acid buffer (pH 5.0). Final concentrations in the assay
mixture were therefore 100 lM substrate and 50 mM
buffer. Plates were incubated for 30 min at 26 C to
approximate the daytime temperature in the upper
10 cm of soil in lowland forests in central Panama
(Marthews et al. 2008). Incubation times were based on
preliminary assays to assess the linearity of the reaction
over time. The reaction was terminated by adding
50 lL of 0.5 M NaOH (final solution pH [ 11) and the
fluorescence determined immediately on a FLUOstar
Optima multi-detection plate reader (BMG Labtech,
Offenburg, Germany), with excitation at 360 nm and
emission at 460 nm. Control wells were prepared for
each substrate and contained substrate, buffer, and
1 mM NaN3 (no soil suspension). Blank wells con-
tained soil suspension and buffer only (no substrate).
Standard wells contained buffer, 1 nmol methylum-
belliferone (MU), and either soil suspension or 1 mM
NaN3 to account for the reduction of fluorescence in the
presence of soil (quenching). Standard curves showed
that fluorescence was linear to at least 2 nmol MU
under these assay conditions. All enzyme activities are
expressed as nmol MU g-1 soil (dry weight) min-1.
Statistical analysis
We performed repeated measures analyses of variance
(ANOVA) for each response variable to evaluate
temporal variation. Between-subject (or between-plot)
effects evaluate responses to treatments over the entire
experiment. Within-subject (or within-plot) effects
evaluate variation among sampling dates and interac-
tions among treatments and sampling dates. Repeated
measures ANOVA assumes compound symmetry of
the variance?covariance matrix if there are more than
two repeated measures. We therefore used the con-
servative Greenhouse-Geisser correction for viola-
tions of the compound symmetry assumption. All
analyses were performed with SYSTAT 11.0 (Rich-
mond, CA).
The NPK factorial design is replicated four times
along a 36-m north?south topographic gradient and
blocked within each replicate along the perpendicular
east?west gradient to control spatial variation in soil
properties (Wright et al. 2011). However, the NP
design does not account for blocks within replicates,
so must overcome additional uncontrolled spatial
variation. The quarterly NPK design therefore has
greater power to detect treatment effects, while the
monthly NP design has greater power to detect
seasonal effects. Throughout the results we report
treatment effects with respect to the quarterly NPK
factorial design and seasonal effects with respect to the
monthly NP factorial design. Time 9 nutrient inter-
actions were evaluated in the monthly NP factorial
design except for time 9 potassium effects, which
were evaluated in the quarterly NPK factorial design.
A one-way analysis of variance (ANOVA) was used to
contrast the micronutrient and control treatments
without blocking.
Results
Soil microbial biomass
Microbial carbon was increased significantly by
phosphorus addition (F1,18 = 8.0, p = 0.011), but
not by nitrogen or potassium addition (Fig. 2a;
Table 1, which summarizes significant treatment and
seasonal effects for all microbial and enzymatic
measurements). Mean microbial carbon concentra-
tions were 972 ? 38 mg C kg-1 in no-phosphorus
plots and 1,096 ? 36 mg C kg-1 in phosphorus addi-
tion plots, an increase of 13 % (Fig. 2a). Microbial
carbon varied seasonally (F12,108 = 18.6, p \ 0.001),
with lowest concentrations in the dry season (Fig. 2b).
In no-phosphorus plots, concentrations ranged from
751 ? 35 mg C kg-1 in the dry season (March 2007)
Biogeochemistry
123
Author's personal copy
to 1321 ? 118 in the wet season (September 2007), an
increase of 76 % (Fig. 2b).
Microbial nitrogen was increased significantly by
phosphorus addition (F1,18 = 7.7, p = 0.013), but not
by nitrogen or potassium addition (Fig. 2c). The mean
microbial nitrogen concentration was 188 ? 8 mg
N kg-1 in no-phosphorus plots and 227 ? 15 mg
N kg-1 in plots receiving phosphorus, an increase of
21 % (Fig. 2c). Microbial nitrogen varied seasonally
(F12,108 = 26.2, p \ 0.001), with lowest concentra-
tions in the dry season (Fig. 2d). In no-phosphorus
plots, microbial nitrogen concentrations ranged from
137 ? 7 mg N kg-1 in the dry season (March 2007)
to a 255 ? 11 mg N kg-1 in the wet season (July
2007), an increase of 86 %. In phosphorus addition
plots, concentrations ranged from 173 ? 9 mg N kg-1 in
the dry season (March 2007) to a 280 ? 22 mg N kg-1
in the wet season (July 2007), an increase of 62 %
(Fig. 2d).
Microbial phosphorus was increased significantly
by phosphorus addition (F1,18 = 14.5, p = 0.001), but
not by nitrogen or potassium addition (Fig. 2e). We
ignore a significant but obscure N 9 K 9 time inter-
action (not shown). The mean microbial phosphorus
concentration was 64.3 ? 3.1 mg P kg-1 in no-phos-
phorus plots and 95.9 ? 7.7 mg P kg-1 in plots
a
M
ic
ro
bi
al
 C
 (m
g C
 kg
-
1 )
600
800
1000
1200
1400
1600
b
600
800
1000
1200
1400
1600
No phosphorus
Phosphorus addition
c
M
ic
ro
bi
al
 N
 (m
g N
 kg
-
1 )
100
150
200
250
300
d
100
150
200
250
300
No phosphorus
Phosphorus addition
e
Treatment
-P +P
M
ic
ro
bi
al
 P
 (m
g P
 kg
-
1 )
0
30
60
90
120
150
f
Nov Jan Mar May Jul Sep Nov
0
30
60
90
120
150
No phosphorus
Phosphorus addition
2006 2007
Fig. 2 Microbial biomass
carbon (a, b), nitrogen (c, d),
and phosphorus (e, f),
following a decade of
fertilization of a lowland
tropical forest on Gigante
Peninsula, Panama. Graphs
on the left show the
significant effect of
phosphorus addition in the
full NPK factorial design
(means and standard errors
of 16 plots in each treatment
sampled four times between
November 2006 and
November 2007). Graphs on
the right show significant
seasonal changes in the NP
factorial design (means and
standard errors of eight
no-phosphorus and eight
phosphorus addition plots
sampled monthly between
November 2006 and
November 2007)
Biogeochemistry
123
Author's personal copy
receiving phosphorus, an increase of 49 %. Microbial
phosphorus varied seasonally (F12,108 = 13.2, p\0.001),
with lowest concentrations in the dry season (Fig. 2f).
In no-phosphorus plots, concentrations ranged from
41.1 ? 6.2 mg P kg-1 in the dry season (March 2007)
to 82.4 ? 8.6 mg P kg-1 in the late wet season
(November 2006), a 100 % increase (Fig. 2f). In
phosphorus addition plots, concentrations ranged from
65.9 ? 12.6 mg P kg-1 in the dry season (March
2007) to 132.3 ? 18.3 mg P kg-1 in the wet season
(November 2006), an increase of 101 % (Fig. 2f).
Microbial nutrient ratios
The microbial C:N ratio (molar) was increased
significantly by nitrogen addition (F1,18 = 3.8,
p = 0.014), but was not altered by phosphorus or
potassium addition (p [ 0.05) (Fig. 3a). However, a
decrease in the ratio following phosphorus addition
was marginally insignificant in the quarterly NPK
factorial design (F1,18 = 2.1, p = 0.055) (not shown).
Microbial C:N varied significantly among months
(F12,108 = 16.2, p \ 0.001), but there was no seasonal
pattern (Fig. 3a). The mean microbial C:N ratio
(molar basis) was 5.8 ? 0.2 in no-nitrogen plots and
6.2 ? 0.1 in plots receiving nitrogen.
The microbial C:P ratio (molar) was not affected by
addition of any nutrient (p [ 0.05). The mean micro-
bial C:P ratio (molar basis) across all 32 plots in the
NPK factorial design was 40.9 ? 2.2, but this con-
ceals strong seasonal variation (F12,108 = 9.1,
p \ 0.001), with lowest ratios in the late wet season
and early dry season (Fig. 3b). The monthly mean
microbial C:P ratio across all plots in the NP factorial
design ranged from 28.2 ? 1.6 in the early dry season
(January 2007) to 54.9 ? 2.8 in the early wet season
(May 2007) (Fig. 3b).
The microbial N:P ratio (molar) was not affected
by addition of any nutrient (p [ 0.05). The mean
microbial N:P ratio (molar basis) across all 32 plots in
the NPK factorial design was 6.9 ? 0.4, but this
conceals strong seasonal variation (F12,108 = 6.9,
p = 0.001), with lowest ratios in the late wet season
and early dry season (Fig. 3c). The monthly mean
Table 1 Summary of seasonal and treatment effects for soil microbial biomass and hydrolytic enzyme activities in the Gigante
fertilization experiment
Parameter Treatment (F1,18) Season (F12,108) Time 9 nutrient
interactions (F12,108)
Microbial C Increased by P (F = 8.1*) Lowest in dry season (F = 18.6***) ns
Microbial N Increased by P (F = 7.7*) Lowest in dry season (F = 26.2***) ns
Microbial P Increased by P (F = 14.5**) Lowest in dry season (F = 13.3**) ns
Microbial C:N Increased by N (F = 7.4*)a Variable (F = 16.2***) ns
Microbial C:P ns Lowest in late wet season (F = 9.1***) ns
Microbial N:P ns Lowest in late wet season (F = 6.9**) ns
Phosphomonoesterase Reduced by P (F = 129***) Lowest in dry season (F = 8.4**) Time 9 P interaction
(F = 5.7**)
Phosphodiesterase Reduced by P (F = 169***) Lowest in dry season (F = 6.5**) Time 9 P interaction
(F = 4.1*)
b-glucosidase ns Lowest in dry season (F = 8.9***) ns
N-acetyl b-glucosaminidase Reduced by P (F = 4.9*) Lowest in dry season (F = 13.9***) ns
Enzymatic C:N (bG:NAG) Increased by P (F = 10.4**) Highest in late wet season (F = 3.6*) ns
Enzymatic C:P (bG:PME) Increased by P (F = 188***) Highest in late wet season (F = 5.1**) Time 9 P interaction
(F = 5.6**)
Enzymatic N:P (NAG:PME) Increased by P (F = 110***) Lowest in dry season (F = 8.5***) Time 9 P interaction
(F = 3.5*)
Treatment effects were evaluated in the quarterly NPK factorial design, seasonal effects in the monthly NP factorial design, and
time 9 nutrient interactions in the NP design. F values are shown in parentheses, with significant effects indicated by *, **, and ***,
representing probability at the 5, 1, and 0.1 % levels, respectively. All microbial ratios are on a molar basis
ns not significant (p [ 0.05), bG b-glucosidase, NAG N-acetyl b-glucosaminidase PME phosphomonoesterase
a Marginally insignificant phosphorus effect of decreasing microbial C:N ratio (F1,18 = 4.2, p = 0.055)
Biogeochemistry
123
Author's personal copy
microbial N:P ratio across all plots in the NP factorial
design ranged from 4.7 ? 0.2 in the late wet season
(November 2007) to 8.6 ? 0.8 in the mid-wet season
(July 2007).
Hydrolytic enzyme activities
Phosphomonoesterase activity
Phosphomonoesterase activity was decreased signifi-
cantly by phosphorus addition (F1,18 = 129, p \ 0.001;
Fig. 4a), but was unaffected by nitrogen or potassium
addition (p [ 0.05). However, the effect of nitrogen was
marginally insignificant (F1,18 = 3.2, p = 0.09) and
activity was always higher where nitrogen was added
alone compared to control plots (Fig. 4b). The mean
phosphomonoesterase activity in the NPK design was
83.1 ? 3.5 nmol MU g-1 min-1 in no-phosphorus
plots and 27.0 ? 0.9 nmol MU g-1 min-1 in plots
receiving phosphorus, a decline of 68 %.
Phosphomonoesterase activity showed strong sea-
sonal variation (F12,108 = 8.4, p = 0.001), with low-
est activity in the dry season (Fig. 4b). For no-
phosphorus plots in the monthly NP factorial design,
activity ranged from 63.9 ? 4.1 nmol MU g-1 min-1
in the late dry season (April 2007) to
93.8 ? 7.9 nmol MU g-1 min-1 in the late wet sea-
son (November 2006), an increase of 47 %. For
phosphorus addition plots, activity ranged from
23.3 ? 1.5 nmol MU g-1 min-1 in the dry season
(February 2007) to 30.1 ? 2.1 nmol MU g-1 min-1
in the wet season (July 2007), an increase of 29 %.
There was a significant time 9 phosphorus interaction
(F12,108 = 5.7, p = 0.005), because the seasonal
change in phosphomonoesterase activity was much
smaller where phosphorus was added compared to no-
phosphorus plots (Fig. 4b).
Phosphodiesterase activity
Phosphodiesterase activity was decreased signifi-
cantly by phosphorus addition (F1,18 = 169.2,
p \ 0.001; Fig. 4c), but was unaffected by nitrogen
or potassium addition (p [ 0.05). The mean phospho-
diesterase activity in the NPK design was 11.97 ?
0.41 nmol MU g-1 min-1 in no-phosphorus plots and
4.14 ? 0.08 nmol MU g-1 min-1 in plots receiving
phosphorus, a decline of 65 %. There was significant
seasonal variation in phosphodiesterase (F12,108 = 6.5,
p = 0.002), with lowest activity in the dry season
(Fig. 4d). There was a significant time 9 phosphorus
interaction (F12,108 = 4.1; p = 0.019), because the
seasonal variation in phosphodiesterase activity was
much smaller where phosphorus was added compared
to no-phosphorus plots (Fig. 4b).
b-Glucosidase activity
b-Glucosidase activity was unaffected by addition of
any nutrient (p [ 0.05; Fig. 5a), but showed strong
seasonal variation (F12,108 = 8.9, p \ 0.001), with
a
M
ic
ro
bi
al
 C
:N
4
5
6
7
8
- N
+ N
b
M
ic
ro
bi
al
 C
:P
20
40
60
c
Nov Jan Mar May Jul Sep Nov
M
ic
ro
bi
al
 N
:P
4
6
8
10
2006 2007
Fig. 3 Seasonal changes in the ratios of microbial carbon to
nitrogen (a), carbon to phosphorus (b), and nitrogen to
phosphorus (c) following a decade of fertilization of a lowland
tropical forest on Gigante Peninsula, Panama. Graphs show the
significant seasonal variation in all cases, and the significant
effect of nitrogen addition on the microbial carbon to nitrogen
ratio (panel a). Values are monthly means and standard errors in
the NP factorial design (i.e. eight plots per treatment in a, 16
plots in b and c)
Biogeochemistry
123
Author's personal copy
lowest activity in the dry season (Fig. 5b). The mean b
glucosidase activity varied from 2.43 ? 0.13 nmol -
MU g-1 min-1 in the dry season (March 2007) to
3.75 ? 0.16 nmol MU g-1 min-1 in the wet season
(July 2007), an increase of 35 %.
N-acetyl b-glucosaminidase activity
The activity of N-acetyl b-glucosaminidase was
decreased significantly by phosphorus addition
(F1,18 = 4.9, p = 0.04), but was unaffected by nitro-
gen or potassium addition (Fig. 5c). The mean N-
acetyl b-glucosaminidase activity in the NPK factorial
design was 6.18 ? 0.41 nmol MU g-1 min-1 in no-
phosphorus plots and 4.71 ? 0.20 nmol MU g-1 min-1
in plots receiving phosphorus (Fig. 5c), a decline of
24 %. There was significant seasonal variation
(F12,108 = 13.9, p \ 0.001), with lowest activity in
the dry season (Fig. 5d). Mean monthly N-acetyl
b-glucosaminidase activity in no-phosphorus plots
ranged from 3.82 ? 0.53 nmol MU g-1 min-1 in the
dry season (March 2007) to 7.41 ? 1.08 nmol
MU g-1 min-1 in the late wet season (November
2006), an approximate two-fold difference. For phos-
phorus addition plots, activity ranged from 3.29 ?
0.30 nmol MU g-1 min-1 in the dry season (March
2007) to 5.98 ? 0.64 nmol MU g-1 min-1 in the wet
season (July 2007), an 82 % increase.
Stoichiometric enzyme ratios
The b-glucosidase to N-acetyl b-glucosaminidase ratio
(C:N) was increased significantly by phosphorus addi-
tion (F1,18 = 10.4, p = 0.005), but was not affected by
nitrogen or potassium addition (Fig. 6a). The mean b-
glucosidase to N-acetyl b-glucosaminidase ratio in the
a
Ph
os
ph
om
on
oe
st
er
as
e 
ac
tiv
ity
 
(nm
ol 
MU
 g-
1  
m
in
-
1 )
0
25
50
75
100
125 b
0
25
50
75
100
125
-P-N 
-P+N 
+P-N 
+P+N 
c
Treatment
-
P 
-N
-
P 
+N
+
P 
-N
+
P 
+N
Ph
os
ph
od
ie
st
er
as
e 
ac
tiv
ity
 
(nm
ol 
MU
 g-
1  
m
in
-
1 )
0
5
10
15
d
Nov Jan Mar May Jul Sep Nov
0
5
10
15
-P -N
-P +N
+P -N
+P +N
2006 2007
Fig. 4 Phosphomonoesterase activity (a, b) and phosphodi-
esterase activity (c, d) following a decade of fertilization of a
lowland tropical forest on Gigante Peninsula, Panama. Graphs
on the left show treatment effects in the full NPK factorial
design (means and standard errors of eight plots in each
treatment sampled four times between November 2006 and
November 2007). Graphs on the right show significant
seasonal changes in the NP factorial design (means and
standard errors of the four replicate plots in each treatment
sampled monthly between November 2006 and November
2007). See Table 1 for significant effects. In the units of
enzyme activity, MU = methylumbelliferone
Biogeochemistry
123
Author's personal copy
NPK factorial design was 0.57 ? 0.03 in no-phosphorus
plots and 0.75 ? 0.02 in phosphorus addition plots, an
increase of 32 % (Fig. 6a). The ratio varied seasonally
(F12,108 = 3.6, p = 0.01), being highest in the dry
season (Fig. 6b). The seasonal effect appeared strongest
for plots that did not receive phosphorus, although the
time 9 phosphorus interaction was not significant
(p = 0.22).
The b-glucosidase to phosphomonoesterase ratio
(C:P) was increased by phosphorus addition (F1,18 =
188, p \ 0.001), but was not affected by nitrogen or
potassium addition (Fig. 6c). The mean b-glucosidase
to phosphomonoesterase ratio in the NPK factorial
design was 0.039 ? 0.002 in no-phosphorus plots and
0.126 ? 0.004 in phosphorus addition plots, a greater
than three-fold increase (Fig. 6c). The ratio varied
seasonally (F12,108 = 8.5, p\0.001), being lowest in the
dry season (Fig. 6d). There was a significant time 9 phos-
phorus interaction (F12,108 = 3.5, p = 0.010), because
the seasonal variation was greater in the phosphorus
addition plots compared to the no-phosphorus plots
(Fig. 6d).
The N-acetyl b-glucosaminidase to phosphomonoes-
terase ratio (N:P) was increased by phosphorus addition
(F1,18 = 103, p \ 0.001), but was not affected by
nitrogen or potassium addition (Fig. 6e). The mean N-
acetyl b-glucosaminidase to phosphomonoesterase ratio
in the NPK factorial design was 0.073 ? 0.0053 in no-
a
CO
N
+
N
+
P
+
N
P
+
K
+
N
K
+
PK
+
N
PK
?-
gl
uc
os
id
as
e 
ac
tiv
ity
 
(nm
ol 
MU
 g-
1  
m
in
-
1 )
2.0
2.5
3.0
3.5
4.0
4.5
b
2.0
2.5
3.0
3.5
4.0
4.5
c
Treatment
-P +PN
-a
ce
ty
l-?
-g
lu
co
sa
m
in
id
as
e 
ac
tiv
ity
 
(nm
ol 
MU
 g-
1  
m
in
-
1 )
0
2
4
6
8
10
d
Nov Jan Mar May Jul Sep Nov
Nov Jan Mar May Jul Sep Nov
0
2
4
6
8
10
No phosphorus
Phosphorus addition
2006 2007
2006 2007
Fig. 5 The activity of b-glucosidase (a, b) and N-acetyl b-
glucosaminidase (c, d) following a decade of fertilization of a
lowland tropical forest on Gigante Peninsula, Panama. For b-
glucosidase, the figure shows a the absence of a significant
treatment effect in the full NPK factorial design (means and
standard errors of four replicate plots in each treatment sampled
four times between November 2006 and November 2007) and
b the significant seasonal variation in the NP factorial design
(means and standard errors of 16 plots per treatment sampled
monthly over the same period). For N-acetyl b-glucosamini-
dase, the figure shows c the significant effect of phosphorus
addition in the full NPK factorial design (means and standard
errors of sixteen replicate plots in each treatment sampled four
times between November 2006 and November 2007) and d the
significant seasonal variation in the NP factorial design (means
and standard errors of eight no-phosphorus and eight phospho-
rus addition plots sampled monthly over the same period). In the
units of enzyme activity, MU = methylumbelliferone
Biogeochemistry
123
Author's personal copy
phosphorus plots and 0.174 ? 0.005 in phosphorus
addition plots, a greater than two-fold increase (Fig. 6e).
The ratio varied seasonally (F12,108 = 5.1, p = 0.005),
being lowest in the dry season (Fig. 6f). There was a
significant time 9 phosphorus interaction (F12,108 =
5.6, p = 0.003), because the decline in the dry season
was greater in the phosphorus addition plots compared to
the no-phosphorus plots (Fig. 6f).
a
?G
: N
AG
 ra
tio
0.2
0.4
0.6
0.8
1.0
b
0.2
0.4
0.6
0.8
1.0
No phosphorus
Phosphorus addition
c
?G
: P
M
E 
ra
tio
0.00
0.05
0.10
0.15
0.20
d
0.00
0.05
0.10
0.15
0.20
No phosphorus
Phosphorus addition
e
Treatment
-P +P
N
AG
: P
M
E 
ra
tio
0.00
0.05
0.10
0.15
0.20
0.25
f
Nov Jan Mar May Jul Sep Nov
0.00
0.05
0.10
0.15
0.20
0.25
No phosphorus
Phosphorus addition
2006 2007
Fig. 6 Stoichiometric enzyme ratios, showing the ratio of
carbon to nitrogen enzymes (b-glucosidase to N-acetyl b-
glucosaminidase) (a, b), the ratio of carbon to phosphorus
enzymes (b-glucosidase to phosphomonoesterase) (c, d), and
the ratio of nitrogen to phosphorus enzymes (N-acetyl b-
glucosaminidase to phosphomonoesterase) (e, f). Graphs on the
left (a, c, e) show the significant effect of phosphorus addition in
the full NPK factorial design (means and standard errors of 16
replicate plots per treatment sampled four times between
November 2006 and November 2007). Graphs on the right
show significant seasonal changes in the NP factorial design
(means and standard errors of eight no-phosphorus and eight
phosphorus addition plots sampled monthly over the same
period). bG, b-glucosidase; NAG, N-acetyl b-glucosaminidase;
PME, phosphomonoesterase
Biogeochemistry
123
Author's personal copy
Discussion
The clear response of the soil microbial biomass and
hydrolytic enzyme activities to long-term phosphorus
addition demonstrates the significance of phosphorus
as the element limiting the microbial community in
this lowland tropical forest. Phosphorus addition
reduced soil phosphatase activities and increased
carbon, nitrogen, and phosphorus concentrations in
the microbial biomass, while the addition of nitrogen,
potassium, other base cations, and micronutrients
elicited no response. The high rates of phosphatase
activity compared to enzymes involved in carbon and
nitrogen acquisition (i.e., b-glucosidase and N-acetyl
b-glucosaminidase to phosphomonoesterase ratios
\0.1; Fig. 6) further point to the strong microbial
phosphorus limitation on Gigante Peninsula. Indeed,
the phosphomonoesterase activities reported here are
greater than the majority of the values reported in a
global compilation (which included only a single
lowland tropical forest), while activities of b-gluco-
sidase and N-acetyl b-glucosaminidase are within the
global range (Sinsabaugh et al. 2008).
These results are consistent with evidence for
limitation of litter decomposition by phosphorus and
micronutrients at this site (Kaspari et al. 2008), but
differ from evidence for nutrient limitation of plant
productivity, which responds to nitrogen, phosphorus,
and potassium addition (Wright et al. 2011). The
increased microbial biomass with phosphorus addition
could have been caused indirectly by the approxi-
mately 25 % increase in fine litter fall in the
phosphorus treatment (Wright et al. 2011), but this
seems unlikely given that doubling the litter standing
crop in an adjacent experiment did not significantly
increase microbial biomass (Sayer et al. 2012).
Collectively, results from the Gigante fertilization
experiment provide strong evidence that plants and
microbes are limited by different nutrients in this
lowland tropical forest.
Although soil microbes are usually assumed to be
limited by organic carbon availability (Wardle 1992),
several previous studies have reported phosphorus
limitation of microbial processes in tropical forests
(Cleveland et al. 2002; Cleveland & Townsend 2006;
Liu et al. 2012) and plantations (Gnankambary et al.
2008) on strongly-weathered soils. For example,
3 years of phosphorus addition promoted an increase
in soil respiration during the dry to wet season
transition in a lowland forest in Costa Rica (Cleveland
and Townsend 2006). In contrast, studies in tropical
montane forests have reported increases in microbial
biomass in response to nitrogen addition (Li et al.
2006; Corre et al. 2010; Cusack et al. 2011). This
supports the assumption that tropical montane forests
are limited by nitrogen availability (Tanner et al.
1998), although none of the three montane studies
included a phosphorus treatment, and hydrolytic
enzyme activities reported by Cusack et al. (2011)
indicate a much greater investment in phosphorus
acquisition compared to nitrogen acquisition. Given
the strong nutrient gradients that can occur in tropical
montane forests (e.g., Andersen et al. 2010), it seems
likely that phosphorus might also regulate microbial
biomass in those environments.
It was recently shown that microbial nutrient ratios
are relatively well-constrained across ecosystems
worldwide (Cleveland and Liptzin 2007). This is
consistent with the results reported here, because
despite marked seasonal variation, microbial nutrient
ratios were not affected greatly by a decade of nutrient
addition, with a significant effect only for microbial
C:N in response to nitrogen addition. Mean microbial
C:N:P ratios in the fertilization experiment across all
plots and sampling dates were 41:6:1, which is
reasonably close to the global mean for all soils of
60:7:1 (at least for N:P), but lower than the global
mean for forest soils of 74:9:1 (Cleveland and Liptzin
2007). The fact that our ratios are lower than the global
mean for forest soils is surprising given that tropical
soils are under-represented in the global data set and
suggests that temperate forests are more strongly
phosphorus limited than commonly assumed. In their
meta-analysis, Cleveland and Liptzin (2007) sug-
gested that microbial N:P might provide greater
insight into ecosystem nutrient status than foliar N:P
in tropical forests given marked variation in foliar N:P
among tropical tree species. Here, microbial N:P ratios
were lower than the global mean for forest soils and
did not vary greatly following long-term phosphorus
addition, despite considerable changes in microbial
nutrient concentrations. This suggests that microbial
nutrient ratios are well constrained for a given soil and
therefore provide limited insight into microbial nutri-
ent status.
In a meta-analysis of the response of phosphatase
activity to experimental nutrient addition, Marklein
and Houlton (2012) found that phosphorus addition
Biogeochemistry
123
Author's personal copy
strongly suppressed root and soil phosphatase activity,
while nitrogen addition enhanced phosphatase activ-
ity, although the increase was slight for soil phospha-
tase (Fig. 3b in Marklein and Houlton). Results for the
lowland tropical forest studied here support the strong
suppression of phosphatase by phosphorus addition,
but not the enhancement by nitrogen addition. This
might be due to the relatively high intrinsic nitrogen
availability at the site (Corre et al. 2010; Hietz et al.
2011).
Olander and Vitousek (2000) showed that the
activity of N-acetyl b-glucosaminidase declined as
nitrogen availability increased in tropical montane
forests along the Hawaiian Islands soil chronose-
quence. They also detected a decrease in activity
following nitrogen addition in the organic horizon of a
young, nitrogen limited site. Our finding that N-acetyl
b-glucosaminidase decreased with phosphorus addi-
tion, but not nitrogen addition, was therefore unex-
pected. We suggest two possible mechanisms. First,
the activity of N-acetyl b-glucosaminidase might
reflect factors other than nitrogen demand, such as
fungal biomass and activity (Miller et al. 1998).
Second, reduced investment in nitrogen-rich phospha-
tase enzymes under phosphorus addition would
increase cellular nitrogen and, in turn, allow microbes
to reduce investment in nitrogen-acquiring enzymes
such as N-acetyl b-glucosaminidase. This explanation
is consistent with the increase in microbial nitrogen
under phosphorus addition.
Despite a decade of phosphorus addition and a
marked increase in extractable soil phosphorus, phos-
phatase activity in the phosphorus addition treatment
remained relatively high in comparison to b-glucosi-
dase and N-acetyl b-glucosaminidase activities (e.g., a
mean b-glucosidase to phosphomonoesterase ratio of
0.13 in phosphorus addition plots). Possible explana-
tions include (i) phosphatase activity in phosphorus
addition plots is derived from stabilized extracellular
enzymes, rather than actively synthesized enzymes
(i.e. the extracellular fraction of the activity is
insensitive to phosphorus status), (ii) the phosphatase
assay includes constitutive or intracellular enzymes
involved in cellular function and not linked to
phosphorus acquisition from the ambient environment
(Allison and Vitousek 2005), and (iii) phosphatase is
involved in processes other than phosphorus acquisi-
tion (e.g., carbon acquisition; Heath 2005; Spohn and
Kuzyakov 2013). Although a number of soil microbes
synthesize phosphatase constitutively, the first possi-
bility seems likely to make the greatest contribution,
given that phosphatase is readily stabilized by sorption
onto clay surfaces or by complexation with soil
organic matter (Quiquampoix 2000; Rao et al. 2000)
and can then constitute a permanent potential activity
that is independent of the soil microbial biomass
(Burns 1982). If this is the case, it indicates that up to
one-third of the phosphatase activity in untreated soils
is associated with stabilized extracellular enzymes.
There was marked seasonal variation in all mea-
sures of microbial biomass and hydrolytic enzymes,
with lowest concentrations and activities in the dry
season. Similar seasonal variation in microbial bio-
mass has been reported previously for other tropical
forest sites (Singh et al. 1989; Ruan et al. 2004; Liu
et al. 2012) and is linked to the death of microbial cells
following desiccation (Salema et al. 1982; West et al.
1992). Drying also reduces enzyme activity (Speir &
Ross 1978), including in tropical forest soils (Turner
and Romero 2010). The death of microbial biomass in
the dry season can represent a considerable release of
nitrogen and phosphorus to the soil (Singh et al. 1989;
Srivastava 1997). Here, the dry season declines in
microbial nitrogen and phosphorus equate to the
release of 118 kg N ha-1 and 41 kg P ha-1, calcu-
lated for the upper 10 cm of soil and assuming a bulk
density of 1.0 g cm-3 (Sayer and Tanner 2010).
These amounts are similar to those added experimen-
tally as fertilizer. The amount of nitrogen released
from microbial biomass is similar to that returned
annually in fine litter fall on Gigante Peninsula
(143 kg N ha-1), but the amount of phosphorus is
several times greater (6 kg P ha-1 in annual litter fall)
(Sayer and Tanner 2010).
The seasonal changes in microbial nutrients were
accompanied by parallel changes in microbial C:P and
N:P ratios, which both declined in the late wet season
and increased again in the mid-dry season. Stoichi-
ometric enzyme ratios also varied seasonally, with
higher b-glucosidase to N-acetyl b-glucosaminidase
ratios in the dry season and higher b-glucosidase and
N-acetyl b-glucosaminidase to phosphomonoesterase
ratios in the wet season. These patterns suggest either
changes in the nutrient content of microbial cells in
response to seasonal variation in nutrient availability,
or a seasonal shift in microbial community composi-
tion (e.g., changing proportions of bacteria and fungi;
Eaton et al. 2011). Seasonal changes in the microbial
Biogeochemistry
123
Author's personal copy
community seem most likely, because the ratios varied
in a similar manner in both control and nutrient
addition plots and were therefore not a simple
response to fertilizer addition in the wet season.
Microbial carbon and nitrogen were previously
determined in the Gigante experiment after 6 years of
nutrient addition (a single sampling in the mid-wet
season), but no significant differences were detected
among treatments (Sayer et al. 2012). This means that
either additional years of treatments, or multiple
analyses over the annual cycle, were necessary to
detect the significant effect of phosphorus addition on
microbial biomass in this experiment. This might
explain why no significant differences in microbial
carbon or nitrogen were detected after 1 year of
nitrogen and phosphorus addition to an Oxisol
supporting secondary tropical forest (Davidson et al.
2004) and highlights the value of long-term ecological
experiments in tropical forests.
Given that previous measurements after 6 years of
fertilization in the Gigante experiment showed no
change in microbial carbon and nitrogen among
treatments (Sayer et al. 2012), and that the results
presented here show no effect of nitrogen addition on
the microbial biomass, we cannot explain why a
previous study during the wet season after 9 years of
fertilization detected a 46 % decline in microbial
carbon and a 29 % decline in microbial nitrogen in the
four plots of the nitrogen only treatment (Corre et al.
2010). The microbial carbon concentration reported
for control plots in that study (1.23 g C kg-1) is
similar to values reported here for wet season samples,
whereas the concentration in the nitrogen addition
plots is considerably lower and similar to our values
for dry season samples (Fig. 2). As we conducted
multiple measurements across the entire experiment
(i.e. all 32 plots in the NPK factorial design plus the
four dolomite/micronutrient plots), compared to the
limited number of samples collected on one occasion
in Corre et al. (2010), we disregard the latter result and
conclude that microbial biomass did not respond to
nitrogen addition in the experiment.
The strong limitation of microbial biomass by
phosphorus availability in this lowland tropical forest
suggests that efforts to incorporate enzymes into
biogeochemical models must account for the dispropor-
tionate investment in phosphorus acquisition by micro-
bial communities in strongly-weathered soils. Some
recent models of soil organic matter cycling explicitly
incorporate nitrogen availability, but do not include
phosphorus as a factor regulating microbial processes,
including enzyme synthesis (e.g., Schimel and Wein-
traub 2003). The high rates of phosphatase activity and
low ratios of carbon and nitrogen to phosphorus
enzymes reported here demonstrate clearly the extent
to which below-ground phosphorus limitation can
influence microbial communities. Despite this, we
caution against the assumption that microbial commu-
nities will be limited by phosphorus in all lowland
tropical forests. Although vast areas of tropical forest
occur on strongly-weathered soils, many tropical forests
are supported by fertile soils in which microbial
phosphorus limitation is unlikely. In central Panama,
for example, soils contain total phosphorus concentra-
tions between 74 and 1,650 mg P kg-1 (Turner and
Engelbrecht 2011) compared to *400 mg P kg-1 on
the Gigante Peninsula (Vincent et al. 2010).
In summary, microbial biomass in a lowland
tropical forest soil responded strongly to phosphorus
addition, but not to nitrogen or potassium addition.
Phosphorus addition increased microbial carbon,
nitrogen, and phosphorus, and caused a marked
decline in the activity of enzymes involved in the
acquisition of phosphorus from organic compounds.
Nitrogen addition did not increase investment in the
activity of any enzyme, including phosphatases,
perhaps due to the relatively high intrinsic nitrogen
availability at the site. Microbial biomass and enzyme
activities showed strong seasonal variation, with the
dry season decline in microbial nitrogen and phos-
phorus accounting for a similar or greater amount of
these nutrients than in annual fine litter fall. Although
a combination of nutrients limits plant productivity on
Gigante Peninsula, we conclude that microbial bio-
mass is constrained by phosphorus availability in this
strongly-weathered tropical forest soil.
Acknowledgments We thank Rufino Gonzalez and Omar
Hernandez for assistance in the field and Tania Romero and
Dianne de la Cruz for laboratory support.
References
Allison SD, Vitousek PM (2005) Responses of extracellular
enzymes to simple and complex nutrient inputs. Soil Biol
Biochem 37:937?944
Allison VJ, Condron LM, Peltzer DA, Richardson SJ, Turner BL
(2007) Changes in enzyme activities and soil microbial
Biogeochemistry
123
Author's personal copy
community composition along carbon and nutrient gradi-
ents at the Franz Josef chronosequence, New Zealand. Soil
Biol Biochem 39:1770?1781
Andersen KM, Turner BL, Dalling JW (2010) Soil-based habitat
partitioning in understorey palms in lower montane tropi-
cal forests. J Biogeogr 37:278?292
Baillie IC (1996) Soils of the humid tropics. In: Richards PW
(ed) The tropical rain forest: an ecological study. Cam-
bridge University Press, Cambridge, pp 256?286
Barthold FK, Stallard RF, Elsenbeer H (2008) Soil nutrient-
landscape relationships in a lowland tropical rainforest in
Panama. For Ecol Manag 255:1135?1148
Brookes PC, Landman A, Pruden G, Jenkinson DS (1985)
Chloroform fumigation and the release of soil nitrogen: a
rapid direct extraction method to measure microbial bio-
mass nitrogen in soil. Soil Biol Biochem 17:837?842
Burns RG (1982) Enzyme activity in soil: location and a possible
role in microbial ecology. Soil Biol Biochem 14:423?427
Cernusak LA, Winter K, Aranda J, Turner BL, Marshall JD
(2007) Transpiration efficiency of a tropical pioneer tree
(Ficus insipida) in relation to soil fertility. J Exp Bot
58:3549?3566
Cernusak LA, Winter K, Turner BL (2009) Physiological and
isotopic (d13C and d18O) responses of three tropical tree
species to water and nutrient availability. Plant Cell
Environ 32:1441?1455
Cleveland CC, Liptzin D (2007) C:N:P stoichiometry in soil: is
there a ??Redfield ratio?? for the microbial biomass? Bio-
geochemistry 85:235?252
Cleveland CC, Townsend AR (2006) Nutrient additions to a
tropical rain forest drive substantial soil carbon dioxide
losses to the atmosphere. Proc Natl Acad Sci USA 103:
10316?10321
Cleveland CC, Townsend AR, Schmidt SK (2002) Phosphorus
limitation of microbial processes in moist tropical forests:
evidence from short-term laboratory incubations and field
studies. Ecosystems 5:680?691
Cleveland CC, Townsend AR, Taylor P, Alvarez-Clare S,
Bustamante MMC, Chuyong G, Dobrowski SZ, Grierson
P, Harms KE, Houlton BZ, Marklein A, Parton W, Porder
S, Reed SC, Sierra CA, Silver WL, Tanner EVJ, Wieder
WR (2011) Relationships among net promary productivity,
nutrients and climate in tropical rain forest: a pan-tropical
analysis. Ecol Lett 14:939?947
Corre MD, Veldkamp E, Arnold J, Wright SJ (2010) Impact of
elevated N input on soil N cycling and losses in old-growth
lowland and montane forests in Panama. Ecology 91:1715?
1729
Cusack DF, Silver WL, Torn MS, Burton SD, Firestone MK
(2011) Changes in microbial community characteristics
and soil organic matter with nitrogen additions in two
tropical forests. Ecology 92:621?632
Davidson EA, Reis de Carvalho CJ, Vieira ICG, Figueiredo
RdO, Moutinho P, Ishida FY, Primo dos Santos MT,
Guerrero JB, Kalif K, Saba RT (2004) Nitrogen and
phosphorus limitation of biomass growth in a tropical
secondary forest. Ecol Appl 14: S150?S163
Dieter D, Elsenbeer H, Turner BL (2010) Phosphorus frac-
tionation in lowland tropical rainforest soils in central
Panama. Catena 82:118?125
Eaton WD, McDonald S, Roed M, Vandecar KL, Hauge JB,
Barry D (2011) A comparison of nutrient dynamics and
microbial community characteristics across seasons and
soil types in two different old growth forests in Costa Rica.
Trop Ecol 52:35?48
Gnankambary Z, Ilstedt U, Nyberg G, Hein V, Malmer A (2008)
Nitrogen and phosphorus limiation of soil microbial res-
piration in two tropical agroforestry parklands in the south-
Sudanese zone of Burkina Faso: the effects of tree canopy
and fertilization. Soil Biol Biochem 40:350?359
Heath RT (2005) Microbial turnover of organic phoshorus in
aquatic systems. In: Turner BL, Frossard E, Baldwin DS
(eds) Organic phosphorus in the environment. CAB Inter-
national, Wallingford, pp 185?203
Hietz P, Turner BL, Wanek W, Richter A, Nock CA, Wright SJ
(2011) Long-term change in the nitrogen cycle of tropical
forests. Science 334:664?666
Jenkinson DS, Brookes PC, Powlson DS (2004) Measuring soil
microbial biomass. Soil Biol Biochem 36:5?7
Kaspari M, Garcia MN, Harms KE, Santana M, Wright SJ,
Yavitt JB (2008) Multiple nutrients limit litterfall and
decomposition in a tropical forest. Ecol Lett 11:35?43
Kaspari M, Stevenson BS, Shik J, Kerekes JF (2010) Scaling
community structure: how bacteria, fungi, and ant taxoc-
enes differentiate along a tropical forest floor. Ecology
91:2221?2226
Kouno K, Tuchiya Y, Ando T (1995) Measurement of soil
microbial biomass phosphorus by an anion exchange
membrane method. Soil Biol Biochem 27:1353?1357
Li Y, Xu M, Zou X (2006) Effects of nutrient additions on
ecosystems carbon cycle in a Puerto Rican tropical wet
forest. Global Chang Biol 12:284?293
Liu L, Gundersen P, Zhang T, Mo J (2012) Effects of phos-
phorus addition on soil microbial biomass and community
composition in three forest types in tropical China. Soil
Biol Biochem 44:31?38
Marklein AR, Houlton BZ (2012) Nitrogen inputs accelerate
phosphorus cycling rates across a wide variety of terrestrial
ecosystems. New Phytol 193:696?704
Marthews TR, Burslem DFRP, Paton SR, Yangu?ez F, Mullins
CE (2008) Soil drying in a tropical forest: three distinct
environments controlled by gap size. Ecol Model 216:
369?384
Miller M, Paloja?rvi A, Rangger A, Reeslev M, Kj?ller A (1998)
The use of fluorogenic substrates to measure fungal pres-
ence and activity in soil. Appl Environ Microbiol 64:613?
617
Nottingham AT, Turner BL, Chamberlain PM, Stott AW,
Tanner EVJ (2012) Priming and microbial nutrient limi-
tation in lowland tropical forest soils of contrasting fertil-
ity. Biogeochemistry 111:219?237
Olander LP, Vitousek PM (2000) Regulation of soil phosphatase
and chitinase activity by N and P availability. Biogeo-
chemistry 49:175?190
Quiquampoix H (2000) Mechanisms of protein adsorption on
surfaces and consequences for extracellular enzyme
activity in soil. In: Bollag J-M, Stotzky G (eds) Soil Bio-
chem. Marcel Dekker, New York, pp 171?206
Rao MA, Violante A, Gianfreda L (2000) Interactions of acid
phosphatase with clays, organic molecules and organo-
Biogeochemistry
123
Author's personal copy
mineral complexes: kinetics and stability. Soil Biol Bio-
chem 32:1007?1014
Ruan HH, Zou XM, Scatena FN, Zimmerman JK (2004)
Asynchronous fluctuation of soil microbial biomass and
plant litterfall in a tropical wet forest. Plant Soil 60:147?
154
Salema MP, Parker CA, Kidby DK, Chatel DL, Armitage TM
(1982) Rupture of nodule bacteria on drying and rehydra-
tion. Soil Biol Biochem 14:15?22
Santiago LS, Wright SJ, Harms KE, Yavitt JB, Korine C, Garcia
MN, Turner BL (2012) Tropical tree seedling growth
responses to nitrogen, phosphorus and potassium addition.
J Ecol 100:309?316
Sayer EJ, Tanner EVJ (2010) Experimental investigation of the
importance of litterfall in lowland semi-evergreen tropical
forest nutrient cycling. J Ecol 98:1052?1062
Sayer EJ, Wright SJ, Tanner EVJ, Yavitt JB, Harms KE, Powers
JS, Kaspari M, Garcia MN, Turner BL (2012) Variable
responses of lowland tropical forest nutrient status to fer-
tilization and litter manipulation. Ecosystems 15:387?400
Schimel JP, Weintraub MN (2003) The implications of exoen-
zyme activity on microbial carbon and nitrogen limitation
in soil: a theoretical model. Soil Biol Biochem 35:549?563
Singh JS, Raghubanshi AS, Singh RS, Srivastava SC (1989)
Microbial biomass acts as a source of plant nutrients in dry
tropical forest and savanna. Nature 338:499?500
Sinsabaugh RL, Moorhead DL (1994) Resource allocation to
extracellular enzyme production: a model for nitrogen and
phosphorus control of litter decomposition. Soil Biol
Biochem 26:1305?1311
Sinsabaugh RL, Lauber CL, Weintraub MN, Ahmed B, Allison
SD, Crenshaw C, Contosta AR, Cusack D, Frey S, Gallo
ME, Gartner TB, Hobbie SE, Holland K, Keeler BL,
Powers JS, Stursova M, Takacs-Vesbach C, Waldrop MP,
Wallenstein MD, Zak DR, Zeglin LH (2008) Stoichiometry
of soil enzyme activity at global scale. Ecol Lett 11:1252?
1264
Sinsabaugh RL, Hill BH, Shah JJF (2009) Ecoenzymatic stoi-
chiometry of microbial organic nutrient acquisition in soil
and sediment. Nature 462:795?799
Speir TW, Ross DJ (1978) Soil phosphatase and sulphatase. In:
Burns RG (ed) Soil enzymes. Academic Press, New York,
pp 198?249
Spohn M, Kuzyakov Y (2013) Phosphorus mineralization can
be driven by microbial need for carbon. Soil Biol Biochem
61:69?75
Srivastava SC (1997) Microbial contribution to extractable N
and P after air-drying of dry tropical soils. Biol Fertil Soils
26:31?34
Tanner EVJ, Vitousek PM, Cuevas E (1998) Experimental
investigation of nutrient limitation of forest growth on wet
tropical mountains. Ecology 79:10?22
Turner BL (2010) Variation in pH optima of hydrolytic enzyme
activities in tropical rain forest soils. Appl Environ
Microbiol 76:6485?6493
Turner BL, Engelbrecht BMJ (2011) Soil organic phosphorus in
lowland tropical rain forests. Biogeochemistry 103:297?
315
Turner BL, Romero TE (2010) Stability of hydrolytic enzyme
activity and microbial phosphorus during storage of trop-
ical rain forest soils. Soil Biol Biochem 42:459?465
Turner BL, Yavitt JB, Harms KE, Garcia MN, Romero TE,
Wright SJ (2013) Seasonal changes and treatment effects
on soil inorganic nutrients following a decade of fertilizer
addition in a lowland tropical forest. Soil Sci Soc Am J: in
press
Vance ED, Brookes PC, Jenkinson DS (1987) An extraction
method for measuring soil microbial biomass C. Soil Biol
Biochem 19:703?707
Vincent AG, Turner BL, Tanner EVJ (2010) Soil organic
phosphorus dynamics following perturbation of litter
cycling in a tropical moist forest. Eur J Soil Sci 61:48?57
Vitousek PM (1984) Litterfall, nutrient cycling, and nutrient
limitation in tropical forests. Ecology 65:285?298
Wardle DA (1992) A comparative assessment of factors which
influence microbial biomass carbon and nitrogen levels in
soil. Biol Rev 67:321?358
West AW, Sparling GP, Feltham CW, Reynolds J (1992)
Microbial activity and survival in soils dried at different
rates. Aust J Soil Res 30:209?222
Windsor DM (1990) Climate and moisture availability in a
tropical forest, long term record for Barro Colorado Island,
Panama. Smithsonian Contributions to Earth Science 29:
1?145
Wright SJ, Yavitt JB, Wurzburger N, Turner BL, Tanner EVJ,
Sayer EJ, Santiago LS, Kaspari M, Hedin LO, Harms KE,
Garcia MN, Corre MD (2011) Potassium, phosphorus or
nitrogen limit root allocation, tree growth and litter pro-
duction in a lowland tropical forest. Ecology 92:1616?
1625
Yavitt JB, Harms KE, Garcia MN, Wright SJ, He F, Mirabello
MJ (2009) Spatial heterogeneity of soil chemical properties
in a lowland tropical moist forest, Panama. Aust J Soil Res
47:674?687
Yavitt JB, Harms KE, Garcia MN, Mirabello MJ, Wright SJ
(2011) Soil fertility and fine root dynamics in response to
4 years of nutrient (N, P, K) fertilization in a lowland
tropical moist forest, Panama. Aust Ecol 36:433?445
Biogeochemistry
123
Author's personal copy