o
nc
Ta
nin
ng S
d fo
n pe
ated
decomposition and the abundance of meso-arthropods in a moist tropical forest. Litterbags containing freshly fallen leaves of Cecropia insignis
The litter layer helps to maintain favourable conditions for
which may also affect decomposition rates (Ostertag and
Forest Ecology and ManagementHobbie, 1999; Hobbie and Vitousek, 2000; Rothstein et al.,
2004). Thus, the amount of leaf litter on the forest floor
influences factors affecting its decomposition. For example,
pulses of litter following hurricanes have been shown to
* Corresponding author at: Smithsonian Tropical Research Institute-BCI,
Unit 948, APO AA 34002, USA. Tel.: +1 507 212 8929; fax: +1 507 212 8975.
E-mail address: sayere@si.edu (E.J. Sayer).
0378-1127/$ ? see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.foreco.2006.04.007Sanford, 1986) and it is widely believed that the release of
nutrients from decaying plant matter is critical for the
maintenance of ecosystem production (Jordan, 1985; Cuevas
and Medina, 1986). Climatic conditions in the lowland wet
tropics are generally highly favourable to decomposition
(Couteaux et al., 1995) and although seasonal drought in many
tropical forests leads to the inhibition of decomposition
processes and a transient build-up of leaf-litter (e.g. Madge,
1965; Wieder and Wright, 1995), most of the litter in lowland
2006) and creating habitats for arthropods (Pearse, 1943; David
et al., 1991; Arpin et al., 1995). Litter removal and litter
addition treatments in temperate forests not only caused
substantial changes to the microclimate on the forest floor, but
also affected the number of decomposer organisms, such as
arthropods (Pearse, 1943; Poser, 1990; David et al., 1991;
Ponge et al., 1993; Arpin et al., 1995) and fungi (Tyler, 1991;
Cullings et al., 2003). Furthermore, changes in litterfall modify
the amounts of available nutrients in the soil (Sayer, 2006),represent the major pathway for the transfer of nutrients
between the plants and the soil (Swift et al., 1979; Vitousek anddecomposition by regulating the microclimate (e.g. Anderson
and Swift, 1983; Vasconcelos and Lawrence, 2005; Sayer,In many tropical forests litterfall and decompositionplaced in either (1) plots where all litterfall was removed monthly (L
); (2) plots where litterfall was doubled monthly (L+), or (3) control plots
(CT). Litter removal significantly slowed decomposition of both species and reduced the abundance of meso-arthropods on Simarouba litter. The
fungicide treatment did not reduce apparent mass loss of Cecropia leaves. The litter addition treatment accelerated the decay of birch wood,
probably because of increased nutrient availability from the extra litter; but there was no change in leaf-litter decomposition or meso-arthropod
abundance in the L+ treatment. After 68 days, the concentrations of nitrogen, phosphorus, potassium, and magnesium in partially decomposed
Cecropia litter were higher in the L+ treatment and lower in the L
 treatment. The accumulation of phosphorus and nitrogen was greater in the
litter in L+ plots and lower in the L
 plots while the release of potassium and magnesium from decomposing litter was lower in the L+ treatment
and greater in the L
 plots. Thus, differences in the quantity of litterfall affect decomposition with consequences for carbon and nutrient storage
and cycling.
# 2006 Elsevier B.V. All rights reserved.
Keywords: Decomposition; Leaf-litter; Litterbag method; Mesofauna; Nutrient dynamics; Tropical forest
1. Introduction moist and wet tropical forests decays within a year (Sampaio
et al., 1993), ensuring rapid nutrient cycling in these systems.(above and below the litter on the forest floor, and with and without fungicide) and Simarouba amara, or untreated birch wood (Betula sp.) wereEffects of litter manipulation
and meso-arthropod abunda
E.J. Sayer a,*, E.V.J.
aDepartment of Plant Sciences, Dow
bDepartment of Zoology, Downi
Received 22 April 2005; received in revise
Abstract
Differences in forest productivity due to climate change may result i
experimental litter removal and litter addition treatments, we investign early-stage decomposition
e in a tropical moist forest
nner a, A.L. Lacey b
g Street, Cambridge CB2 3EA, UK
treet, Cambridge CB2 3EJ, UK
rm 2 March 2006; accepted 3 April 2006
rmanently altered levels of litterfall and litter on the forest floor. Using
the effects of increased and decreased litterfall on early-stage litter
www.elsevier.com/locate/foreco
229 (2006) 285?293
nd Mincrease the forest floor by 20?150%. This large single input of
mostly green litter caused an acceleration of decomposition
resulting in the return to prehurricane forest floor levels within
five months (Ostertag et al., 2003).
Litterfall has been shown to increase with rising atmospheric
CO2 levels (DeLucia et al., 1999; Allen et al., 2000; Schlesinger
and Lichter, 2001; Finzi et al., 2001; Zak et al., 2003), but
decreases are also conceivable if changes in rainfall patterns
affect tree phenology. Although litter quality is one of the most
important factors in determining decomposition rates in the
tropics (Bloomfield et al., 1993; Anderson and Swift, 1983),
changes in the quantity of litter on the forest floor may also
affect decomposition rates.
In a model of litterfall and decomposition in a tropical forest,
Sampaio et al. (1993) calculated the rates of mass transfer
between fresh litter and older organic matter layers and the
turnover times of the Oi (litter), Oe (fermentation), and Oa
(humus) horizons. They then simulated the effects of increased
or decreased litterfall over time and found that after a
permanent 50% increase in litterfall, a new equilibrium of the
transfers and turnover times of organic matter would only be
reached after 4, 14, and 40 years for the Oa, Oe, and Oi
horizons, respectively. However, the decomposition rates used
in the model were kept constant; if changes in the amount of
litterfall alter decomposition rates, the turnover time of organic
matter in the forest floor may differ greatly from the values
calculated. Longer or shorter turnover times for litter in the
forest floor may have wide-reaching consequences for tropical
forest carbon and nutrient cycling.
Despite the substantial number of studies on decomposition
in a wide range of ecosystems, the influence of a sustained
change in litter quantity on decay rates has thus far been largely
neglected. In the humid tropics, micro- andmesoarthropods and
fungi are considered to be among the most important
decomposers (e.g. Zhang and Zak, 1998; Heneghan et al.,
1999; Gonzalez and Seastedt, 2001). Thus, we collected meso-
fauna from decomposing litter within large-scale litter addition
and litter removal treatments in a tropical moist forest to
investigate whether increased or decreased litterfall affects
early-stage decomposition of litter and whether this was
correlated with changes in meso-arthropod abundance. In
addition, we treated a subset of litterbags with fungicide to
determine whether reduced fungal infection affects decom-
position.
2. Materials and methods
2.1. Study site
The study site is in old-growth lowland moist tropical forest
located on the Gigante Peninsula of the Barro Colorado Nature
Monument in Panama, Central America. The soil is an oxisol
with pH ca. 5.0, a low ?available? phosphorus concentration, but
relatively high total nitrogen and exchangeable potassium,
magnesium, and calcium (Cavalier, 1992; Sayer et al., 2006).
Nearby Barro Colorado Island (ca. 5 km from the study site)
E.J. Sayer et al. / Forest Ecology a286has a mean annual rainfall of 2600 mm with a strong dryseason, typically for four months from January to April, and an
average temperature of 27 8C (Leigh, 1999).
Fifteen 45 m  45 m plots were set up in 2000. The plots
were trenched to a depth of 0.5 m to minimize nutrient- and
water import via the root/mycorrhizal network, the trenches
were double-lined with plastic and backfilled; a 7.5-m buffer
was left around the inside of the trenches to eliminate trenching
effects, resulting in a measurement plot size of 30 m  30 m.
Starting in January 2003, the litter (including branches with a
diameter <100 mm) in five plots was raked up and removed
once a month, resulting in low, but not entirely absent, litter
standing crop (L
 plots). The removed litter was immediately
added to five further plots, where it was spread out as evenly as
possible, effectively doubling the litterfall (L+ plots); five plots
were left undisturbed as controls (CT plots).
2.2. Decomposition
To investigate how changes in litterfall affect decomposition
rates, a litterbag experiment was set up in the 15 litter
manipulation plots. Leaves of the common forest species
Cecropia insignis Liebm. (henceforth referred to as Cecropia)
were used as the trees grow abundantly in the study area and the
nitrogen concentration of freshly fallen leaves is similar to the
average concentration in mixed litter collected from litter traps
in this forest (1.5%; Sayer and Tanner, unpublished data).
Freshly fallen Cecropia leaves were collected from the study
area in March and April 2003 and oven-dried at 60 8C.
Litterbags were made of 1.2-mm fiber-glass mesh and
measured 200 mm  250 mm; 160 bags received 12 g
(0.05 g; 10 g of leaf and 2 g of petiole) of Cecropia leaf-
litter and 60 bags received 10 g (0.05 g) of oven-dried craft
sticks made from untreated birch wood (henceforth referred to
as ?wood?). Craft sticks were used instead of native wood
because the samples were homogeneous in size, shape, and
composition. To determine the importance of fungal decom-
position in the litter manipulation treatments 60 Cecropia
litterbags were treated with the common broad-spectrum
fungicide chlorothalonil. Fungicide was reapplied to the treated
bags every 2 weeks. In order to avoid contaminating untreated
bags with fungicide, the treated bags were placed 3?5 m from
all other decomposition bags. The fungicide was applied to
each bag separately by lifting the bags onto a large plastic bag
(1000 mm  800 mm) and spraying the fungicide at a close
range (200?300 mm); lifting the bags reinforced the fungicide
treatment by severing any existing hyphal connections (Cuevas
and Medina, 1988).
In May 2004, four bags of wood, four untreated bags of
Cecropia leaf-litter, and four bags treated with fungicide were
placed in each of the 15 litter manipulation plots. The bags were
placed on the surface of the litter layer in the CT and L+ plots
and on the soil surface in the L
 plots. To determine whether
the position of leaf-litter in the forest floor influences
decomposition rates, four additional untreated Cecropia bags
were placed on the surface of the mineral soil in each of the the
CT and L+ plots and covered with litter. All litterbags, except
anagement 229 (2006) 285?293fungicide treated bags, were anchored to the soil surface to
nd Mavoid movement that could sever roots and fungal hyphae.
Litterbags in the CT plots were covered with natural litterfall,
and litterbags in the L+ plots by natural litterfall and by
additional litter during each raking cycle; in the L
 plots fallen
litter was carefully removed from the surface of the bags twice a
month. One bag of each subtreatment (untreated control,
fungicide treated, beneath the forest floor, and wood) was
collected at random from each plot every 21?26 days for 68
days. The collected litterbags were washed to remove soil
particles. As the bags in the litter removal plots tended to
accumulate more soil particles due to rain splashing, the
washing time for each bag was standardised by first washing the
bag with the most soil particles and timing the washing required
to remove the soil; all other bags were then washed for the same
length of time. After the third collection (68 days) the bags in
the litter removal plots contained so much soil that it was not
possible to wash the contents without losing a large proportion
of the remaining biomass. The contents of the bags were dried
to constant weight at 60 8C and weighed. Decomposition rates
were calculated as percent mass loss per day (% day
1).
The dried contents of the untreated litterbags of Cecropia
were finely ground and composite samples (one per treatment
and time) were made from the litterbags after 21 and 42 days;
these composite samples and all the bags (one per plot) from the
final litter collection (after 68 days) were analysed for nutrient
concentrations. The analyses were carried out by Waite
Analytical Services, Adelaide, Australia. Phosphorus and
cations were determined by nitric acid digestion, finished with
hydrochloric acid, and radial view inductively coupled plasma-
optical emission spectrometry (ICP-OES); total nitrogen was
determined using a Carlo Erba elemental analyser (NA1500,
series 2, Carlo Erba, Milan, Italy). Nutrient accumulation or
release was calculated as the increase or decrease in the
absolute amount of a given nutrient in the leaf-litter after 68
days.
2.3. Meso-arthropod abundance
A second litterbag experiment was set-up to assess whether
changes in litterfall affect meso-arthropod abundance on
decomposing litter. Freshly fallen leaves of Simarouba amara
Aubl., (henceforth referred to as Simarouba) were collected in
January 2003. Litterbags were made of 1.2-mm fibre-glass
mesh and measured 250 mm  250 mm; thirty bags received
10 (0.05) g of oven-dried (60 8C) Simarouba leaves. In May
2003, two bags were placed randomly on the soil surface in
each plot beneath the crown of an adult Simarouba tree and
anchored to the soil surface; the bags in the litter addition plots
were immediately covered with litter. Litterbags in the CT plots
were covered with natural litterfall and litterbags in the L+ plots
by natural litterfall and by additional litter during each raking
cycle; in the L
 plots fallen litter was carefully removed from
the surface of the bags twice a month. One bag was collected at
random from each plot after 30 and 59 days and placed in a
cloth bag to allow oxygen diffusion to the samples.
Immediately upon returning from the field the contents of
E.J. Sayer et al. / Forest Ecology athe litterbags were placed in Berlese funnels for 48 hours, afterwhich the leaf litter was desiccated and any remaining
arthropods were likely to be dead. Extracted fauna was stored
in 70% alcohol and the litter was washed, oven-dried at 60 8C,
and weighed to determine mass loss due to decomposition. In
each sample the total number of meso-arthropods (body size
0.1?2 mm; Swift et al., 1979) and the abundances of mites and
springtails were determined under a dissecting microscope.
Voucher specimens were deposited in the insect collection at
the University Museum of Zoology, Cambridge, UK.
2.4. Data analysis
The comparison of litter decomposition in the treatments
was based on the values of dry mass remaining at each
collection. Repeated measures ANOVAs were performed to
assess differences between litter manipulation treatments (CT,
L+, L
) in decomposition for each experiment separately: (i)
untreated Cecropia leaves placed above the forest floor, (ii)
untreated Cecropia leaves placed below the forest floor, (iii)
fungicide-treated Cecropia leaves, (iv) wood, (v) Simarouba
leaf litter, (vi) total meso-faunal abundance on Simarouba leaf
litter. In addition, separate paired t-tests were performed to
assess differences in decomposition between untreated Cecro-
pia leaf-litter and either fungicide-treated leaf-litter or leaf-
litter beneath the forest floor within a given litter manipulation
treatment.
Differences between treatments in the final nutrient
concentrations and the accumulation or release of nutrients
from Cecropia litter and the abundance of springtails and mites
at each collection time were compared between treatments
using separate one-way ANOVAs. Post hoc comparisons were
made using Fisher?s LSD test. Relationships between mass loss
of Simarouba leaf-litter and the total number of arthropods,
mites, and springtails were investigated using Pearson?s
correlations.
All analyses were carried out using Genstat 7.0.2 (VSN
International Ltd., Hemel Hempstead, UK).
3. Results
3.1. Decomposition
Litter removal slowed the decomposition of untreated
Cecropia leaf-litter relative to the control and litter addition
treatments (P < 0.001, rep. measures ANOVA; Fig. 1a). The
rate of decay was more than twice as high in the CT (0.22%
day
1) and the L+ plots (0.25% day
1) than in the L
 plots
(0.09% day
1). Mass-loss from the litter after 68 days was 7%
in the L
 plots, but 17% and 20% in the CT and L+ plots,
respectively. Cecropia leaf-litter treated with fungicide also
decomposed more slowly in the L
 plots compared to the other
two treatments (P < 0.001, rep. measures ANOVA; Fig. 1b);
mass-loss after 68 days was 12% in the L
 plots, 23% in the CT
plots, and 20% in the L+ plots. In all litter manipulation
treatments there was a trend towards faster decay of leaf-litter
in fungicide-treated litterbags compared to untreated litterbags,
anagement 229 (2006) 285?293 287which was marginally significant (P = 0.06, rep. measures
E.J. Sayer et al. / Forest Ecology and M288
gnis
sticANOVA). The position of the litter in the CT and L+ plots
(whether on the surface or beneath the forest floor) had no effect
on decomposition rates (Fig. 1c).
Wood decomposed faster in the L+ treatment than in the L

or the CT plots, which did not differ (P < 0.001, rep. measures
ANOVA; Fig. 1d). The decay rate was 0.48% day
1 in the L+
Fig. 1. Mass loss during 68 days of decomposition of (a) untreated Cecropia insi
(c) untreatedC. insignis leaves placed beneath the litter layer, and (d) birch wood
squares: control, triangles: litter addition, and circles: litter removal.plots, 0.36% day
1 in the CT plots, and 0.30% day
1 in the L

plots. The mass-loss of wood after 68 days was 33% in the L+
plots, compared to 25% and 20% in the CT and L
 plots,
respectively.
Simarouba leaf-litter decomposed more slowly in the L

treatment than the CT or L+ plots (P = 0.01, rep. measures
ANOVA; Fig. 2). The rate of Simarouba leaf-litter decay was
Fig. 2. Total abundance of arthropods extracted from litterbags of Simarouba
amara after 30 and 59 days (bars) and mass loss during 68 days of decom-
position of S. amara leaves (lines) in litter manipulation plots in moist tropical
forest, Panama, Central America; dark grey bars and squares: control, light grey
bars and triangles: litter addition, and white bars and circles: litter removal.0.90% day
1 in the CT and L+ plots but only 0.60% day
1 in
the L
 treatment. There was a strong time  treatment
interaction, with no difference between treatments in mass
loss of Simarouba leaf-litter after 30 days, but after 59 days the
percentage of mass lost in the L
 treatment (36%) was less than
in the CT or L+ plots (both 53%).
anagement 229 (2006) 285?293
leaves placed on the litter surface, (b) C. insignis leaves treated with fungicide,
ks in litter manipulation plots in moist tropical forest, Panama, Central America;3.2. Meso-arthropods
The total number of meso-arthropods on Simarouba leaf-
litter was lower in the L
 plots than in the other two
treatments (P = 0.02, rep. measures ANOVA; Table 1). The
abundance of springtails was marginally lower in the
L
 treatment than in the CT or L+ plots after 30 days
(P = 0.058, one-way ANOVA; Table 1) and the abundance of
mites was lower in the L
 plots after 59 days (P = 0.019, one-
way ANOVA; Table 1). The abundance of mites in June and
the abundance of springtails in July were significantly greater
in the L+ plots than in the CT plots (P = 0.41 and P = 0.42,
respectively, LSDs; Table 1).
Mass loss of Simarouba leaf-litter over 59 days was related
to both the total number of arthropods and the number of mites
in the litterbags (P = 0.003, R = 0.48 and P = 0.016, R = 0.39,
respectively, Pearson correlation; Fig. 2).
3.3. Nutrient concentrations in decomposing Cecropia
leaf-litter
The concentrations of nitrogen, phosphorus, potassium, and
magnesium in decomposing Cecropia litter at 68 days were
higher in L+ and lower in L
 plots and the net accumulation or
release of these nutrients were altered by litter addition and
litter removal (Table 2). Nitrogen accumulated in decomposing the concentration of sulphur was 33% higher in the L+ plots
E.J. Sayer et al. / Forest Ecology and Management 229 (2006) 285?293 289
Table 1
Total abundance of mesofauna, mites, and springtails extracted from decomposing Simarouba amara leaf-litter in litter manipulation treatments in Panama, Central
America, given as the number of individuals per m
2
Mites (Acari) Springtails (Collembola) Total mesofauna
June July June July June July
CT 422  67 b 886  177 b 365  109 154  28 b 1123  221 b 1539  325 b
L+ 752  239 a 598  81 b 291  90 666  10 a 1372  410 b 1971  365 b
L
 346  147 b 314  73 a 115  37* 198  86 b 774  475 a 1075  266 a
CT: control, L+: litter addition, L
: litter removal. Standard errors for N = 5 plots per treatment are given. Different letters in a column denote significant differences
between treatments at P < 0.05.
* Denotes a marginally significant difference (P = 0.058) to the other two treatments.Cecropia litter in all three treatments (Fig. 3a), but litter
addition significantly increased the net accumulation of
nitrogen relative to the L
 treatments (P = 0.036; Table 2)
and nitrogen concentrations were 10% higher in the L+ plots
compared to the L
 plots (P = 0.017; Table 2) at the end of the
experiment. Net accumulation of phosphorus occurred in all
three treatments over the 68 days of decomposition (Fig. 3b),
but the net accumulation of phosphorus in the L+ treatment was
three times as high as in the L
 treatment (P = 0.019) and
almost twice as high as in the controls (Table 2). The final
concentration of phosphorus was 40% greater in the L+ plots
compared to the L
 plots (P = 0.019; Table 2). Less potassium
was released in the L+ treatment than in the controls
(P = 0.055) or L
 treatments (P = 0.008; Fig. 3c) and the
concentration of potassium in Cecropia litter in the L+ plots
after 68 days was almost twice as high as in the controls
(P = 0.049), and three times greater than in the L
 plots
(P = 0.005; Table 2). The release of magnesium from
decomposing litter (Fig. 3d) was also reduced in the L+ plots
compared to the controls (P = 0.009) or L
 plots (P < 0.001;
Table 2); the concentration of magnesium was 25% higher in
the L+ plots and 21% lower in the L
 compared to the controls
(P = 0.002 and 0.008, respectively; Table 2). Calcium
accumulated equally in the decomposing litter in all three
treatments (Fig. 3e) and the final concentrations of calcium did
not differ between treatments (Table 2). Sodium and sulphur
were released from decomposing Cecropia litter in all three
treatments (Fig. 3f and g, respectively) but at the end of the
experiment, the concentration of sodium was 38% higher andTable 2
Concentrations and net accumulation or release of nutrients inCecropia insignis leave
in Panama, Central America
Initial concentration (mg g
1) Final concentration (mg g

CT L+
N 15.0 17.7 ab 18.2 b
P 0.25 0.37 ab 0.46 b
K 2.50 0.51 a 0.97 b
Ca 11.6 17.1 17.8
Mg 4.50 2.67 a 3.28 b
Na 1.21 0.10 0.12
S 1.88 1.55 1.65
CT: control, L+: litter addition, L
: litter removal. Different letters denote differenc
for a given nutrient at P < 0.05. Where a row has no letters, there were no signifithan in the L
 plots (P = 0.013 and 0.005, respectively;
Table 2).
4. Discussion
4.1. Decomposition of litter and abundance of mesofauna
Litter removal treatments slowed the decomposition of leaf-
litter. Although litterbags in the CT and L+ plots were covered
by litter during the experiment while the litterbags in the L

plots remained exposed, it is unlikely that the decreased
decomposition rates in litterbags in the L
 plots were due to
greater exposure. Frequent rainfall during the study period
(rainy season) prevented drying-out, especially as litterbags
usually keep litter wetter than under natural conditions (Tanner,
1981); the litterbags were generally very wet at each collection
time. Furthermore, the initial position of theCecropia litterbags
(above or beneath the forest floor) in the CTand L+ plots had no
effect on decomposition rates. Lower abundance of decom-
poser organisms and decreased nutrient availability following
litter removal are the most likely causes for slower decom-
position in the L
 plots.
Litter removal reduces the amount of substrate for
microorganisms and the habitat space for arthropods (Sayer,
2006). Litter arthropods are important for the comminution of
leaf-litter, which exposes a greater surface for attack by
microbial decomposers (Singh and Gupta, 1977; Petersen and
Luxton, 1982; Bradford et al., 2002), and meso- and
microarthropods, in particular mites and springtails, are thoughts after 68 days of decomposition in litter manipulation plots in old-growth forest
1) Net accumulation/release (mg g
1)
L
 CT L+ L

16.4 a 2.67 ab 3.22 a 1.39 b
0.31 a 0.12 ab 0.21 a 0.06 b
0.37 a 
1.99 a 
1.53 b 
2.13 a
15.6 5.53 6.16 4.04
2.42 a 
1.83 a 
1.22 b 
2.08 a
0.10 
1.11 
1.09 
1.11
1.51 
0.33 
0.23 
0.37
es between treatments in the final concentration or the net accumulation/release
cant treatment differences for that nutrient.
nd ME.J. Sayer et al. / Forest Ecology a290to play an important role in decomposition in the tropics
(Madge, 1965; Heneghan et al., 1999; Gonzalez and Seastedt,
2001). In our study, mites and springtails made up between 64%
and 76% of the total fauna in the litterbags in the CT and L+
plots and between 47% and 60% of the total in the L
 plots.
Lower abundances of meso-arthropods, especially mites, were
found in litterbags in the L
 plots; the significant relationship
between the decrease in mass of Simarouba leaf-litter with the
decreased abundance of meso-arthropods suggests that the
Fig. 3. Changes in the concentrations of (a) nitrogen, (b) potassium, (c) phosphoru
decomposition of C. insignis leaves in litter manipulation plots in moist tropical fo
circles: litter removal. Standard errors are given for the values at 68 days; values giv
each of the five plots per treatment.anagement 229 (2006) 285?293lower numbers of decomposers in the L
 plots contributed
greatly to the reduced decomposition rates of the leaf-litter.
Surprisingly, the total abundance of meso-arthropods in
the litterbags did not differ greatly between the CT and
L+ treatment, despite the increase in substrate and habitat space
with the addition of litter, and we observed only transient
increases in mite and springtail abundance in the L+ plots.
Similar results have been reported for temperate forests, where
changes in both abundance and community composition in
s, (d) magnesium, (e) calcium, (f) sodium, and (g) sulphur during 68 days of
rest, Panama, Central America; squares: control, triangles: litter addition, and
en for 21 and 42 days are composite samples made by combining samples from
litte
the
gre
D
diff
nd Maddition treatments may be due to possible adverse effects of
litter addition treatments on soil and litter fauna such as reduced
diffusion of oxygen in the denser forest floor (Judas, 1990),
increases in the numbers of predators (Uetz, 1979), and greater
amounts of phytochemicals (Sayer, 2006).
Although the initial nutrient concentration of the litter is
important in determining decay rates, nutrient availability in the
surface soil may also have some influence (Prescott et al., 1993;
Hobbie and Vitousek, 2000). The increase of nutrient
concentrations in decomposing litter generally indicates which
nutrients are limiting the activity of decomposer organisms
(Anderson et al., 1983;McGroddy et al., 2004). In our study, the
concentrations of nitrogen and phosphorus increased in
Cecropia leaf-litter during the early stages of decomposition
(Fig. 3a and b, respectively). An analysis of surface soil
samples (0?20 mm) showed that the availability of these
nutrients was lower in the L
 plots than the controls (Sayer and
Tanner, unpublished data). Thus, the lower decay rates are
related to the lower soil nutrient content in the L
 treatment.
In contrast to the patterns of decomposition for leaf-litter,
wood decomposed faster in the L+ treatment than in the CT or
L
 plots, which did not differ. As our litterbags excluded larger
saprophagous insects such as termites and beetles, fungi were
the main decomposers of woody debris in this experiment; this
result therefore suggests that fungal activity in the forest floor is
greater in the L+ plots than the other treatments. Higher nutrient
availability in the forest floor in the L+ plots appears to have
reduced the nutrient limitation of decomposition for nutrient-
poor, high-lignin substrate such as wood.
Although it has been suggested that fungi are more critical
for decomposition processes in the dry season than the rainy
season (Cornejo et al., 1994; Lodge et al., 1994), Yavitt et al.
(2004) found greater abundance of fungi on decomposing litter
during the rainy season in the same forest type on Barro
Colorado Island, 5 km away from our study site. This indicates
that fungi are important decomposers during the rainy season in
this forest. Nevertheless, contrary to expectations, the leaf-litter
in bags treated with fungicide decomposed at the same rate as,
or slightly faster than, untreated litter. It is unlikely that wetting
the leaf-litter by fungicide application had an effect on their
decomposition during the rainy season, as the litter on the
ground was generally very wet and did not dry out between
rainfall events. It is also improbable that lifting the bags
resulted in a noticeable loss of fine material from the treated
bags, as fine mesh was used for the litterbags; furthermore, the
greater losses during washing render any small losses due to
lifting and/or transportation negligible.
There remain two possible explanations for the similarity in
mass loss between treated and untreated litterbags:
(i) Untreated leaf-litter decomposed faster than fungicide-
treated litter, but fungal biomass in the untreated litterbags
compensated for the lost weight. Casual observation of the(Pose
1995r addition treatments were generally less than expected
r, 1990; David et al., 1991; Ponge et al., 1993; Arpin et al.,
; Sayer, 2006). The lack of expected increase in litter
E.J. Sayer et al. / Forest Ecology asamples supports this suggestion, as there was no visible greh litter is facilitated by the release of phosphorus from the
ater mass of older decomposing organic matter.
espite the lack of effects of the fungicide treatments, the
erences in the concentrations of potassium are suggestive offloor
fresater accumulation of phosphorus in the L+ plots compared to
controls suggests higher phosphorus availability in the forest
(Rothstein et al., 2004), as the import of phosphorus intoevidence of fungal mycelium in the treated litterbags, while
untreated bags often had a high proportion of mycelium on
the litter.
(ii) The application of fungicide reduced the competition
between fungi and bacteria and bacterial decomposition
increased (M?ller et al., 1999).
It is probable that a combination of these two factors
contributed to the lack of differences in decomposition between
fungicide-treated and untreated Cecropia litter. Similarly, the
differences in mass loss between CT and L+ plots in wood, but
not in leaf-litter, may be a consequence of greater fungal
biomass in the L+ plots. In early stages of decomposition,
fungal hyphae and mycelium can be removed from the surface
of wood samples during washing, but leaf fragments and fungal
biomass cannot be separated. In this study, an increase in fungal
biomass in the L+ plots would therefore contribute to the
measured weight remaining in leaf-litter, but not in wood.
We could find no studies on the effects of chlorothalonil on
soil fauna and did not measure meso-arthropods in the fungicide
experiment, so there is a small possibility that chlorothalonil
affected decomposition via an effect on litter fauna.
4.2. Nutrient dynamics in decomposing litter
Littermanipulationmodified the timing anddegree of nutrient
release or accumulation during early-stage decay in this
experiment, and resulted in differences between treatments in
the concentrations of nitrogen, phosphorus, potassium, and
magnesium in Cecropia litter after 68 days of decomposition.
There are generally three phases of nutrient dynamics during
decomposition: (i) initial release caused by leaching, (ii) net
accumulation (immobilization), and (iii) net release with
mineralization (Gosz et al., 1973; Swift et al., 1979; Chuyong
et al., 2002). When a nutrient is limiting to the metabolic
processes of decomposer microorganisms the initial release does
not take place and accumulation occurs until the nutrient
concentration approaches that of microbial tissue (Anderson
et al., 1983), in tropical forests this pattern is typical for
phosphorus and nitrogen (Vitousek and Sanford, 1986; Rogers,
2002). Decomposition processes are often primarily nitrogen-
limited, but other nutrients can become limitingwhen nitrogen is
in sufficient supply (Swift et al., 1979) and although nitrogen
concentrations in the Cecropia litter increased during the study
period, net accumulation was relatively low (Fig. 3a; Table 2).
Phosphorus is thought to be the main limiting nutrient in the
study area, as concentrations in the soil are low (Cavalier, 1992;
Sayer et al., 2006), and the net accumulation of phosphorus in the
decomposing litter supports this assumption (Fig. 3b; Table 2).
Gre
anagement 229 (2006) 285?293 291ater fungal activity in the L+ plots. Potassium is selectively
nd Mtaken up by fungi during the decomposition process (Cromack
et al., 1975; Tyler, 2005) and the net release of potassium was
23% less in the L+ treatment than in the controls and 28% less
than in the L
 plots. As potassium is normally readily leached
from litter (Lousier and Parkinson, 1978), the greater concentra-
tion and amount of potassium in the litter in the L+ plots may be
indicative of increased fungal biomass. Although it is possible
that litter addition reduced leaching, there were no differences
between treatments in release of sodium (Fig. 3f), which was the
element most readily leached from the litter. Washing the
litterbags likely caused leaching in particular of water-soluble
nutrients from the litter, especially potassium and sodium,
therefore comparisons of the concentrations of these nutrients
with other datasets should be treated with caution. However, in
this study, between-treatment comparisons are valid as all the
bags were washed for the same length of time.
5. Conclusions
Decreased litter standing crop slowed the decomposition of
leaf-litter by decreasing the population of decomposers and by
reducing the supply of available nutrients in the soil. Increased
litterfall appeared to have little effect on early-stage leaf-litter
decomposition, but the accelerated decomposition of wood
suggests greater availability of nutrients to decomposer organ-
isms in theL+ treatment. It is possible that greater fungal biomass
in litter in the L+ plots obscured increased mass loss due to
decompositioncompared to the controls; thismerits further study.
Litterfall in forests may be affected by climate change
through changes in precipitation and increased primary
production. Thus, although many studies have found that
differences in litter quality under elevated CO2 levels had little
effect on the decay rates of litter in forests (Hirschel et al., 1997;
Gahrooee, 1998; Finzi et al., 2001), the effects of climate
change on litter quantity may influence the decomposition of
litter and woody debris and greatly modify the transfer of
organic matter through the forest floor in tropical forests, with
consequences for the carbon and nutrient cycle.
Acknowledgments
We thank Jesus Valdez and Ruscena Wiederholt for help
with fieldwork, Dr. M. Kaspari and Mary Johnston for helpful
comments and the use of the Berlese funnels, and two
anonymous reviewers for critical comments that improved this
manuscript. This study was funded by the Mellon Foundation,
the Gates Cambridge Trust, the Smithsonian Tropical Research
Institute, a Worts? Travelling Scholarship, and a Cambridge
Philosophical Society Travel Grant. The entire study was
carried out at the Smithsonian Tropical Research Institute, to
whom we are very grateful.
References
Allen, A.S., Andrews, J.A., Finzi, A.C., Matamala, R., Richter, D.D., Schle-
singer, W.H., 2000. Effects of free-air CO2 enrichment (FACE) on below-
E.J. Sayer et al. / Forest Ecology a292ground processes in a Pinus taeda forest. Ecol. Appl. 10, 437?448.Anderson, J.M., Proctor, J., Vallack, H.W., 1983. Ecological studies in four
contrasting lowland rain forests in Gunung-Mulu National Park Sarawak. 3.
Decomposition processes and nutrient losses from leaf litter. J. Ecol. 71,
503?527.
Anderson, J.M., Swift, M.J., 1983. Decomposition in tropical forests. In:
Sutton, S.L., Whitmore, T.C., Chadwick, A.C. (Eds.), Tropical Rain
Forest: Ecology and Management. Blackwell Scientific, Oxford, pp.
287?309.
Arpin, P., Ponge, J.F., Vannier, G., 1995. Experimental modifications of litter
supplies in a forest mull and reaction of the nematode fauna. Fund. Appl.
Nematol. 18, 371?389.
Bloomfield, J., Vogt, K.A., Vogt, D.J., 1993. Decay-rate and substrate quality of
fine roots and foliage of 2 tropical tree species in the Luquillo experimental
forest. Puerto-Rico. Plant Soil 150, 233?245.
Bradford, M.A., Tordof, G.M., Eggers, T., Jones, H., Newington, J.E., 2002.
Microbiota, fauna, and mesh size interaction in litter decomposition. Oikos
99, 317?323.
Cavalier, J., 1992. Fine-root biomass and soil properties in a semi-deciduous
and a lower montane rain forest in Panama. Plant Soil 142, 187?201.
Chuyong, G.B., Newbery, D.M., Songwe, N.C., 2002. Litter breakdown and
mineralization in a central African rain forest dominated by ectomycor-
rhizal trees. Biogeochemistry 61, 73?94.
Cornejo, F.H., Varela, A., Wright, S.J., 1994. Tropical forest litter decomposi-
tion under seasonal drought: nutrient release, fungi and bacteria. Oikos 70,
183?190.
Couteaux, M.-M., Bottner, P., Berg, B., 1995. Litter decomposition, climate and
litter quality. Trends Ecol. Evol. 10, 63?66.
Cromack, K., Todd, R.L., Monk, C.D., 1975. Patterns of basidiomycete nutrient
accumulation in conifer and deciduous forest litter. Soil Biol. Biochem. 7,
265?268.
Cuevas, E., Medina, E., 1986. Nutrient dynamics within Amazonian forests. I.
Nutrient flux in fine litter fall and efficiency of nutrient utilization. Oeco-
logia 68, 466?472.
Cuevas, E., Medina, E., 1988. Nutrient dynamics within Amazonian forests. II.
Fine root growth, nutrient availability and leaf litter decomposition. Oeco-
logia 76, 222?235.
Cullings, K.W., New, M.H., Makhija, S., Parker, V.T., 2003. Effects of litter
addition on ectomycorrhizal associates of a lodgepole pine (Pinus con-
torta) stand in Yellowstone National Park. Appl. Environ. Microb. 69,
3772?3776.
David, J.-F., Ponge, J.F., Arpin, P., Vannier, G., 1991. Reactions of the
macrofauna of a forest mull to experimental perturbations of litter supply.
Oikos 61, 316?326.
DeLucia, E.H., Hamilton, J.G., Naidu, S.L., Thomas, R.B., Andrews, J.A.,
Finzi, A.C., Lavine, M., Matamala, R., Mohan, J.E., Hendrey, G.R., Schle-
singer, W.H., 1999. Net primary production of a forest ecosystem with
experimental CO2 enrichment. Science 284, 1177?1179.
Finzi, A.C., Allen, A.S., DeLucia, E.H., Ellsworth, D.S., Schlesinger, W.H.,
2001. Forest litter production, chemistry, and decomposition following 2
years of free-air CO2 enrichment. Ecology 82, 470?484.
Gahrooee, F.R., 1998. Impacts of elevated CO2 on litter quality litter decom-
posability and nitrogen turnover rate of two oak species in a Mediterranean
forest ecosystem. Glob. Change Biol. 4, 667?677.
Gonzalez, G., Seastedt, T.R., 2001. Soil fauna and plant litter decomposition in
tropical and subalpine forests. Ecology 82, 955?964.
Gosz, J.R., Likens, G.E., Bormann, F.H., 1973. Nutrient release from decom-
posing leaf and branch litter in the Hubbard Brook Forest New Hampshire.
Ecol. Monogr. 43, 173?191.
Heneghan, L., Coleman, D.C., Zou, X., Crossley, D.A., Haines, B.L., 1999. Soil
microarthropod contributions to decomposition dynamics: tropical-tempe-
rate comparisons of a single substrate. Ecology 80, 1873?1882.
Hirschel, G., Korner, C., Arnone, J.A., 1997.Will rising atmospheric CO2 affect
leaf litter quality and in situ decomposition rates in native plant
communities? Oecologia 110, 387?392.
Hobbie, S.E., Vitousek, P.M., 2000. Nutrient limitation of decomposition in
Hawaiian forests. Ecology 81, 1867?1877.
Jordan, C.F., 1985. Nutrient Cycling in Tropical Forest Ecosystems. JohnWiley
anagement 229 (2006) 285?293and Sons Ltd., Chichester.
Judas, M., 1990. The development of earthworm populations following manip-
ulation of the canopy leaf litter in a beechwood on limestone. Pedobiologia
34, 247?255.
Leigh Jr., E.G., 1999. Tropical Forest Ecology. Oxford University Press, Oxford.
Lodge, D.J., McDowell, W.H., McSwiney, C.P., 1994. The importance of
nutrient pulses in tropical forests. Trends Ecol. Evol. 9, 384?387.
Lousier, J.D., Parkinson, D., 1978. Chemical element dynamics in decomposing
leaf litter. Can. J. Bot. 56, 2795?2812.
Madge, D.S., 1965. Leaf fall and litter disappearance in a tropical forest.
Pedobiologia 5, 273?288.
McGroddy, M., Silver, W.L., Cosme de Olveira Jr., R., 2004. The effect of
phosphorus availability on decomposition dynamics in a seasonal lowland
Amazonian forest. Ecosystems 7, 172?179.
M?ller, J., Miller, M., Kj?ller, A., 1999. Fungal-bacterial interaction on beech
leaves: influence on decomposition and dissolved organic carbon quality.
Soil Biol. Biochem. 31, 367?374.
Ostertag, R., Scatena, F.N., Silver, W.L., 2003. Forest floor decomposition
following hurricane litter inputs in several Puerto Rican forests. Ecosystems
6, 261?273.
Ostertag, R., Hobbie, S.E., 1999. Early stages of root and leaf decomposition in
Hawaiian forests: effects of nutrient availability. Oecologia 121, 564?573.
Pearse, A.S., 1943. Effects of burning-over and raking-off litter on certain
animals in the Duke Forest. Am. Midl. Nat. 406?424.
Petersen, H., Luxton, M., 1982. A comparative analysis of soil fauna popula-
tions and their role in decomposition processes. Oikos 39, 287?388.
Ponge, J.F., Arpin, P., Vannier, G., 1993. Collembolan response to experimental
perturbations of litter supply in a temperate forest ecosystem Eur. J. Soil Sci.
Sampaio, E.V.D.B., Dall?Olio, A., Nunes, K.S., Delemos, E.E.P., 1993. Amodel
of litterfall, litter layer losses and mass-transfer in a humid tropical forest at
Pernambuco Brazil. J. Trop. Ecol. 9, 291?301.
Sayer, E.J., 2006. Using experimental manipulation to assess the roles of leaf
litter in the functioning of forest ecosystems. Biol. Rev. 81, 1?31.
Sayer, E.J., Tanner, E.V.J., Cheesman, A.W, 2006. Increased litterfall changes
fine root distribution in a moist tropical forest. Plant Soil 281, in press.
Schlesinger, W.H., Lichter, J., 2001. Limited carbon storage in soil and litter of
experimental forest plots under increased atmospheric CO2. Nature 411,
466?469.
Singh, J.S., Gupta, S.R., 1977. Plant decomposition and soil respiration in
terrestrial ecosystems. Bot. Rev. 43, 449?528.
Swift, M.J., Heal, O.W., Anderson, J.M., 1979. Decomposition in Terrestrial
Ecosystems. University of California Press, Berkley.
Tanner, E.V.J., 1981. The decomposition of leaf litter in Jamaican montane rain
forests. J. Ecol. 69, 263?275.
Tyler, G., 1991. Effects of litter treatments on the sporophore production of
beech forest macrofungi. Mycol. Res. 95, 1137?1139.
Tyler, G., 2005. Changes in the concentrations of major, minor and rare-earth
elements during leaf senescence and decomposition in a Fagus sylvatica
forest. Forest Ecol. Manage. 206, 167?177.
Uetz, G.W., 1979. The influence of variation in litter habitats on spider
communities. Oecologia 40, 29?42.
Vasconcelos, H.L., Lawrence, W.F., 2005. Influence of habitat, litter type, and
soil invertebrates on leaf-litter decomposition in a fragmented Amazonian
landscape. Oecologia 144, 456?462.
Vitousek, P., Sanford Jr., L., 1986. Nutrient cycling in moist tropical forest. Ann.
E.J. Sayer et al. / Forest Ecology and Management 229 (2006) 285?293 29329, 141?153.
Poser, T., 1990. The Influence of litter manipulation on the centipedes of a beech
wood. In: Minelli, A. (Ed.), Proceedings of the Seventh International
Congress on Myriapods. Brill, Leiden, pp. 235?245.
Prescott, C.E., Taylor, B.R., Parsons, W.F.J., Durall, D.M., Parkinson, D., 1993.
Nutrient release from decomposing litter in Rocky Mountain coniferous
forests: influence of nutrient availability. Can. J. For. Res. 23, 1576?1586.
Rogers, H.M., 2002. Litterfall decomposition and nutrient release in a lowland
rain forest, Morobe Province Papua New Guinea. J. Trop. Ecol. 18, 449?456.
Rothstein, D.E., Vitousek, P.M., Simmons, B.L., 2004. An exotic tree alters
decomposition and nutrient cycling in a Hawaiian montane forest. Ecosys-
tems 7, 805?814.Rev. Ecol. Syst. 17, 137?167.
Wieder, R.K., Wright, J.S., 1995. Tropical forest litter dynamics and dry season
irrigation on Barro Colorado Island, Panama. Ecolology 76, 1971?1979.
Yavitt, J.B., Wright, S.J., Wieder, R.K., 2004. Seasonal drought and dry-season
irrigation influence leaf- litter nutrients and soil enzymes in a moist, lowland
forest in Panama. Austral. Ecology 29, 177?188.
Zak, D.R., Holmes, W.E., Finzi, A.C., Norby, R.J., Schlesinger, W.H., 2003.
Soil nitrogen cycling under elevated CO2: a synthesis of forest FACE
experiments. Ecol. Appl. 13, 1508?1514.
Zhang, Q., Zak, J.C., 1998. Potential physiological activities of fungi and
bacteria in relation to plant litter decomposition along a gap size gradient in
a natural subtropical forest. Microb. Ecol. 35, 172?179.