d o i: 1 0 .1 1 1 1 /j.1 4 6 1 -0 2 4 8 .2 0 1 2 .0 1 8 3 7 .x
E cology Letters, (2012)
LETTER
The origin of litter chemical com plexity during decom position
A bstract
Kyle W ickings^*
A. S tuart G ra n d y / Sasha C. Reed^^
and
Cory C. Cleveland^
T he chemical complexity o f decom posing plant litter is a central feature shaping the terrestrial carbon (C)
cycle, b u t explanations o f the origin o f this complexity rem ain contentious. H ere, we ask; H ow does litter
chemistry change during decom position, and w hat roles do decom posers play in these changes? D uring a
long-term (730 days) litter decom position experiment, we tracked concurrent changes in decom poser com ­
m unity structure and function and litter chemistry using high-resolution molecular techniques. Contrary to
the current paradigm, we found that the chemistry o f different litter types diverged, rather than converged,
during decom position due to the activities o f decom posers. Furtherm ore, the same litter type exposed to
different decom poser comm unities exhibited striking differences in chemistry, even after > 90% mass loss.
O ur results show that during decom position, decom poser com m unity characteristics regulate changes in lit­
ter chemistry, w hich could influence the functionality o f htter-derived soil organic m atter (SOM) and the
turnover and stabilisation o f soil C.
K eywords
D ecom position, enzymes, m icroarthropods, microbial comm unities, plant litter, soil carbon sequestration,
soil organic matter.
Ecology letters (2012)
While the biotic and abiotic factors controlling litter decom position
rates have been well studied (Adair et al. 2008), the drivers o f
changes in litter chemistry during decom position — and thus the
factors that contribute to the chemical complexity o f decom posing
litter and htter-derived soil organic m atter (SOM) — rem ain unre­
solved (Poirier et al 2005; G randy & N e ff 2008; V ancam penhout
et al 2009). Nevertheless, determ ining the origin and consequences
o f this complexity is critical, as the chemical complexity o f decom ­
posing litter and SOM is a central feature shaping the function o f
sohs. F or example, litter chemistry influences SOM turnover and
carbon (C) stabhisation dynamics (Schmidt et al 2011; Schnitzer &
M onreal 2011), as well as microbial biomass and decom poser
com m unity structure (Ladd et al 1994; B aum ann et al 2009), soil
nutrient cychng, and the form ation o f soil structure (Chivenge et al
2011). Accordingly, changes in the chemistry o f htter entering soil
have the potential to significantly alter multiple ecosystem character­
istics, including organo-mineral interactions, soil C stabilisation
dynamics, and soil C O 2 emissions to the atm osphere.
While multiple drivers may be im portant in regulating organic
m atter chemistry during decom position, here we focus on a set o f
testable hypotheses addressing the m ajor controls over htter chem i­
cal transform ations (Fig. 1). First, the ‘Chemical Convergence
H ypothesis’ suggests that htter chemistry is determ ined by the stage
o f decom position and that, although htter chemistry in the early
stages o f decom position may reflect initial htter quahty, during
decom position ah htter — regardless o f differences in initial chem is­
try — whl eventuahy converge towards a common chemistry. This
hypothesis has come to represent the current paradigm o f htter
decom position and its rationale is that, regardless o f any taxonom ic
variation that may exist am ong soil microbial and faunal
comm unities, decom posers share a com m on and hm ited set o f bio­
chemical pathways and physiological abhities that should constrain
the chemical trajectory o f decom posing organic m atter (Fig. 1;
M cG hl 2007; Fierer et al 2009). Because o f these constrained bio­
logical controls, differences in chemistry during htter decom position
result from the preferential loss o f easily degradable com pounds
(e.g. starch and proteins) and the persistence and relative accumula­
tion o f com pounds that are m ore resistant to decay (e.g. lignin,
and hgnin-encrusted carbohydrates) (Berg & M cClaugherty 2008).
Thus, as chemicahy distinct htter types pass through the decom ­
poser Tunnel’ (Jenkinson & Ladd 1981; G randy & N eff 2008; Fierer
et al 2009), physiological constraints whl lead to converging chemi­
cal structures (Fig. 1). C onsistent w ith this hypothesis, many htter
bag studies have noted that initially diverse htter types possess
rem arkably simhar chemistries fohowing c. 75—80% o f mass loss
(Mehho et al 1989; Preston et al 2009; M oore et al 2011).
T he second hypothesis — the ‘Initial Litter Quahty H ypothesis’ —
posits that initial htter quahty has persistent effects on htter chemical
com position during decom position, and thus differences in initial h t­
ter chemistry persist throughout the decay sequence regardless o f
the structure o f the decom poser com m unity or the extent o f htter
processing (Fig. 1). Consistent w ith this hypothesis, Berg &
M cClaugherty (2008) showed that initial differences in htter chemis­
try, particularly nitrogen (N) and hgnin concentrations, persist from
early to late stages o f decom position. O th er w ork has also show n that
initial differences in htter chemistry may even persist once htter is
incorporated into SOM (Angers & Mehuys 1990), and recent studies
have observed that chemical differences in htter inputs are reflected in
INTRODUCTIO N
^D epartm ent o f N atu ra l Resources a n d the Environm ent, University o f N e w
^D ep artm ent o f Ecosystem a n d Conservation Sciences, University o f M o n tan a,
Hampshire, Durham , NH, USA
Missoula, M T, USA
^D ep artm ent o f Ecology and Evolutionary Biology, University o f Colorado,
* Correspondence: E-mail: k.wickings@ unh.edu
Boulder, CO, USA
^Current address: US G eological Survey, Southwest Biological Science Center,
M oab, UT, USA
© 2012 BlackweU PubUshing L td /C N R S
2 K. W ic k in g s
et al.
Letter
Hj) chemical convergence
H^: initial litter quality
Hj) decom poser control
Figure 1 H ypothesised controls o n litter chem ical c o m p o sitio n d uring decom position. L e ft panel — chem ical convergence hypothesis: due to co m m o n physiological
constraints am ong decom posers, initial differences in p la n t h tte r chem istry are expected to converge over the course o f decom position; center panel — Initial h tte r quahty
hypothesis: th e chem istry o f h tte r th ro u g h o u t d eco m p o sitio n is a fu n ction o f initial h tte r quahty; right panel — decom poser c o n tro l hypothesis: during d e com position plant
htter chem istry Is shaped by th e c o m p o sitio n an d th e m etabohc capabhities o f the decom poser com m urhty.
the chem istty o f b o th aggregated and non-aggregated SO M (FiUey
et al. 2008a; Stewart et al 2011). Taken together, this w ork dem on­
strates that initial differences in litter chemistry may persist into late
stages o f decom position w hen plant litter becomes SOM.
T he third hypothesis — the ‘D ecom poser C ontrol H ypothesis’ —
suggests that changes in litter chemistry during decom position do
n o t simply reflect the effects o f tim e (i.e. the stage o f decom posi­
tion) a n d /o r initial litter quality, b u t are also strongly influenced by
the functional attributes o f distinct decom poser comm unities
(Fig. 1). In other w ords, this hypothesis predicts that distinct
decom poser comm unities create functionally distinct decom poser
‘funnels’ (and hence diverse litter chemistries) th at persist through­
o u t decom position. I f so, changes in environm ental conditions that
modify decom poser comm unities (e.g. those associated w ith hum an
activities) could theoretically result in functionally significant
changes in the chemistry o f decom posing litter. This may be due to:
(I) differences in enzyme activities th at could result in selective
decom position o f some com pounds (Carreiro et al. 2000; GaUo et al
2005; G randy et al. 2008); (2) variation in the m etabolic capabili­
ties o f decom poser organisms (CavigeUi et al. 1995; D eleporte &
Charrier 1996; Balser & Firestone 2005); a n d /o r (3) differences in
the chemical structure o f microbial and macrobial necrom ass
(K ogel-Knabner 2002). T here are clear relationships betw een litter
chemistry, microbial comm unities and enzyme activities (Conn &
D ighton 2000; Berg & M cClaugherty 2008; Keeler et al. 2009) and
other w ork suggests that, in the short-term , decom poser com m u­
nities can influence changes in C cycling (Strickland et al. 2009) and
litter chemistry (Wickings et al 2011). However, few decom position
studies have assessed C chemistry dynamics over long-term litter
decay in relation to decom poser com m unity com position and activ­
ity, although such an effort is critical for assessing the validity o f
these three com peting hypotheses (but see Sinsabaugh et al 2002;
Snajdr et al 2011).
W ith these hypotheses providing the context, o u r overall objec­
tive was to examine the origin o f litter chemical structure during
decom position. Specifically, we decom posed plant litter w ith varying
initial quality in ecosystems w ith m anagem ent-induced differences in
decom poser com m unities to better understand shifts in plantderived organic m atter chem istry in the context o f decay stage,
© 2012 Blackwell P ublishing L td /C N R S
initial litter quality and decom poser communities. O ur results indi­
cate th at the chemistry o f p lant litter during all stages o f decom po­
sition — even over the long-term (i.e. 730 days) — reflects b o th its
initial chemical traits and the unique metabolic traits o f its decom ­
poser community, and th at distinct decom poser comm unities may
enhance chemical complexity and diversity in decom posing litter
rather than reducing it as the current paradigm suggests. By regulat­
ing the chemical structure o f plant litter inputs to soil, decom poser
com m unities have the potential to influence fundam ental soil func­
tions including the form ation and stabilisation o f SOM.
MATERIALS AND METHODS
Stud y site
T he experim ent was conducted at the W.K. Kellogg Biological
Station (KBS) L ong-T erm Ecological Research Site (LTER) in
H ickory Corners, M I, USA over three growing seasons (730 days)
from 2008 through 2010. T he experim ent was conducted in con­
ventionally tilled (CT), no-tiU (NT) and early successional old field
(OF) treatm ents (Wickings et al 2011). Treatm ents consisted o f four
replicated I ha plots arranged in a random ised com plete block
design (total o f 12 plots). M ean annual precipitation at the L T E R
site is 890 m m per year and soils are classified as K alamazoo (fineloamy) and O shtem o (coarse-loamy) mixed, mesic, Typic Hapludalfs
(Alfisols) developed o n glacial outw ash (Crum & Collins 1995).
Litter b a g e x p e r i m e n t
Litter bags were constructed from nylon m esh measuring
18 X 18 cm (1.5 m m m esh size) and were filled w ith 7 g o f either
corn o r grass litter. Grass litter (primarily Bromus inermis, Leyss) was
collected as live standing shoots (c. 50 cm tall at harvest) from old
field ecosystems at KBS during May 2008. C orn litter was collected
as standing dead biomass during Fall 2007 from large-scale experi­
m ental fields at KBS. AU litter was air dried and cut into pieces
2—4 cm in length. These litter types were chosen because they are
the dom inant plant biomass inputs in these systems and also have
different chemical com positions. Litter bags were placed in direct
Letter
contact w ith soil in aU treatm ents and secured at their corners with
7 cm stainless steel naUs o n 5 June 2008. Forty-eight Utter bags (24
corn, 24 grass) were placed into each o f 12 plots for a total o f 576
Utter bags according to Wickings et al. (2011). Litter bags were col­
lected after 6, 17, 26, 39, 73, 108, 368, 411, 452, 482, 712 and
730 days o f decom position. U pon collection, aU Utter bags were
stored at 4 °C until analysis and aU analyses were conducted within
24 h o f coUection. Litter bags overw intered in aU plots b u t were
briefly rem oved from aU sites during spring field activities, placed in
individual, open plastic bags, and stored in plastic paUs in the field.
O nce field activities were com pleted. Utter bags were returned to their
original locations foUowing the same procedures outUned previously.
Origin o f litter c h e m ic a l c o m p le x it y 3
tivorous m icroarthropods, including oribatid mites, coUembolans,
enchytraeid w orms, mUUpedes, and dipteran larvae were taUied and
are reported as cumulative num ber o f individuals g^^ dry Utter.
O nce Berlese extractions were com plete, dry Utter mass was
recorded and subsamples ic. 0.5 ^ were incinerated at 500 °C in a
muffle furnace for the determ ination o f Utter ash content. Litter
mass loss is reported here as the percentage o f ash-free dry mass
rem aining at a given coUection time. A n additional subsample (0.25
—0.5 g) o f air-dried Utter was set aside for chemical analysis.
Litter chem istry
W e analysed the activities o f three hydrolytic enzymes [P-W-acetylglucosaminidase (NAG), (3-glucosidase (BG) and acid phosphatase
(PHOS)] and phenol oxidase (POX) at every sample date o n subs­
amples (0.25-0.5 g) o f Utter using previously pubUshed m ethods
(Saiya-Cork et al 2002). Hydrolytic enzymes were assessed using
black, 96-weU plates and substrates containing the fluorescing m ole­
cule 4-methylumbeUiferone and phenol oxidase was measured using
L-3, 4-dihydroxyphenylalanine in clear, 96-weU plates. AU enzyme
activities were first calculated in units o f pm ol h^^ g^^ dry Utter
and were subsequently integrated over aU sample dates (mol g ^ Ut­
ter) foUowing Sinsabaugh et al. (2002) to derive cumulative enzyme
activities. Integrated activities are presented as m ol mg^^ Utter for
simpUcity in data presentation.
Litter chemistry was assessed using pyrolysis-gas chrom atography
and mass spectrom etry (py-GCMS) o n corn and grass Utter coUected
at 0, 108, 487 and 730 days. In addition, we also analysed grass Utter
chemical com position at 17 days w hen its mass loss was c. 30—35%,
w hich was similar to corn after 108 days (to have an extra data point
for testing Hypothesis 1). Samples were pyrolysed at 600 °C for 20 s
o n a CDS Pyroprobe 5150 pyrolyzer (CDS Analytical, Inc., O xford,
PA, USA), and then transferred to a T herm o Trace G C U ltra gas
chrom atograph (Therm o Fisher Scientific, Austin, TX, USA) and
Polaris Q mass spectrom eter (Therm o Fisher Scientific). Peaks were
analysed using A utom ated Mass Spectral D econvolution and Identi­
fication System, (AMDIS, V 2.65) and the N ational Institute o f Stan­
dards and Technology (NIST) com pound Ubrary. C om pound
abundances were determ ined relative to the total ion signal from aU
detected and identified peaks and are reported as percentages
(G randy et al. 2009; Wickings et al. 2011).
DNA e x tra ctio n a n d q u a n ti t a t i v e PCR analysis
Statistical analysis
M icrobial D N A was extracted from Utter subsamples (0.05—0.15 g)
decom posed for 0, 108, 487, and 730 days using the M oBio PowerSoU D N A kit foUowing the m anufacturer’s instructions (MoBio
Laboratories, Carlsbad, CA, USA; H ofm ockel et al 2007; M anerkar
et al. 2008). T he qPC R assays were conducted in 96-weU plates on
three analytical repUcates foUowing the procedures outUned in Fierer
et al. (2005). Standard curves were created for bacteria and fungi
using E . coli and Saccharomyces cervidae, respectively. F or aU samples,
each 25 pL qPCR reaction contained 12.5 pL o f PowerSYBR
G reen PCR M aster mix (AppUed Biosystems, Foster City, CA,
USA); 1.25 pL each o f 10 pM forw ard and reverse prim ers
(Eurofins M W G O peron, HuntsvUle, AL, USA); 5 pL o f sterile,
nuclease free water (Promega, Madison, W I, USA); and 5 pL o f
standard or environm ental D N A sample. F or bacteria, Eub338 and
Eub518 were used as forward and reverse prim ers, respectively, and
for fungi, 5.8s and IT S lf were used as forward and reverse primers,
respectively (Fierer et al. 2005). Before analysis, sample D N A was
diluted to roughly equal concentrations and aU reactions were car­
ried o u t on a StepO ne Plus qPC R system (AppUed Biosystems).
M elting curve analyses and agarose gel electrophoresis were used to
confirm that the fluorescence signal resulted from specific PCR
products and n o t ampUfication artifacts.
W e analysed data by b o th time and decay stage. First, data were
analysed using three-way analysis o f variance in SAS (Cary, N C,
USA) using time. Utter and m anagem ent type as m ain factors. T reat­
m ent means for main effects, and two- and three-way interactions
were com pared using least square means com parison. Because detritivore densities were n o t normaUy distributed they were log-trans­
form ed prior to analysis, b u t back-transform ed before presentation
here. T hese analyses provided insight into differences betw een Utter
types and m anagem ent practices over time b u t Utter chemical con­
vergence and its underlying drivers need to be examined at consis­
ten t decay stages. Thus, P earson’s correlation in SAS was used to
assess relationships betw een Utter chemistry, percent mass rem ain­
ing, and aU biological variables at time points that corresponded
w ith c. 30, 80, 95 and 97% mass loss in grass and c. 30, 70, and
80% mass loss in corn Utter. In addition, apparent enzyme efficien­
cies were calculated as the slope o f the regression betw een % mass
loss and cumulative enzyme activities and reflect the percent change
in Utter mass per accumulated enzyme activity (mol mg ^ Utter). We
also used non-m etric multidim ensional scaUng (NMS) to look for
patterns in litter chem istry by Utter type, management, and stage o f
decay and their relationship to Utter decom posers.
Extracellular e nzym es
M i c ro a rth ro p o d c o m m u n itie s a n d m a s s loss
M icroarthropods were extracted from Utter bags at every sample
date using Berlese funnels (BioQuip, Inc., Rancho D om inguez, CA,
USA) within 3—4 h o f Utter coUection (Wickings et al. 2011). D etri-
RESULTS
D e co m po sition r a te a n d litter chem istry
Initial C:N ratios were greater in corn (61) than in grass Utter (18)
{P < 0.0001; Table 1). C orn Utter also had significantly low er relative
© 2012 Blackwell Publishing L td /C N R S
4 K. W ic k in g s
et al.
Letter
T able 1 Initial an d final Utter chenUstry (carbon-C , nitro g en -N , Ugnin-L), percentage o f Utter m ass rem aining-M R EM % , and m onthly decay rate-k m o ^ (:means ± SE)
after 730 days o f c o rn and grass Utter d eco m p o sitio n in conventionaU y tUled (conventional), no--till, and old field treatm ents
C o rn
Grass
N (mg g ')
c (mg g ^ ')
Initial
441.1
C:N
L:N
7.3
60.5
4.2
% M REM
k mo ^
c (mg g ^ ')
N (mg g ')
C:N
L:N
433.2
23.7
18.3
2.5
% M REM
k mo ^
Final
C onventional
155.5 (23.7)
6.9 (0.8)
22.4 (1.1)
3.0 (1.4)
2.5 (1.4)
0.46 (0.06)
28.0 (2.1)
19.3 (3.9)
95.7 (12.6)
314.1 (33.8)
13.5 (1.1)
9.6 (1.0)
0.22 (0.01)
0.19 (0.02)
7.0 (0.4)
262.0 (8.5)
12.2 (4.9)
9.7 (1.8)
14.4 (1.8)
No-tiU
22.9 (3.2)
13.8 (0.4)
3.7 (0.6)
0.7 (0.2)
0.56 (0.03)
O ld field
311.8 (49.7)
11.0 (1.3)
27.8 (1.5)
17.1 (5.8)
26.6 (2.4)
0.15 (0.01)
350.6 (15.0)
26.2 (1.6)
13.5 (0.5)
3.0 (0.3)
6.5 (2.6)
0.34 (0.05)
\
1.5 -
Lignin
n 30%
O
▲
\
0.5 -
^
F:B
^
1
70%
11
95%
Polysac.
POX
©
A
(N
c/5
Lipids
-0.5 -
k
.
Detritivore
\A
□
BG
A
Proteins
Phenols
-1.5 -
1
1
1
- 1 .6
1
0.4
t
■
t
0.4
\
1
1
1
1 .6
A x is 1 (5 % )
Figure 2 N o n -m etric m ultidim ensional scaling o rd in atio n o f litter chem istry by
decay stage, litter type an d m an ag em en t (also see Fig. SI). O p e n symbols
indicate 30% , bne fiU indicates 70%
(corn only), black indicates 80% and
stippled fiU indicates 95% (grass only) m ass loss. Symbols d en o te m anagem ent
treatm ents. Circles — conventionally tilled, squares — no-tiU, triangles — old field.
Prim ary (chemical) and secondary (biological) variables correlated w ith ordination
axes are p lo tte d o v er o rdinated samples (gray arrow s). D istances betw een sample
points reflect variation in fitter chem istry (see Fig. SI for significant m anagem ent
and fitter type effects o n chem istry).
am ounts o f proteins (0.5 vs. 11%) and a higher abundance o f N -bearing com pounds o f non-protein origin and polysaccharides than grass
(10 vs. 4% and 26 vs. 12%; Fig. SI Supporting Inform ation). Grass
litter exhibited greater total mass loss than corn over the three grow ­
ing seasons (c 3 vs. 20% mass remaining, respectively; P < 0.0001;
Table 1). M anagem ent also exerted m odest b u t significant effects on
decom position o f bo th litter types as total mass loss was significantly
greater in CT systems than in O F in corn and greater in b o th agricul­
tural systems than in O F in grass, (P < 0.05; Table 1).
D uring decom position, grass litter maintained greater abundances
o f lipids, phenols, N -bearing com pounds, and proteins and lower
abundances o f lignin and polysaccharides than corn litter (Fig. 2, see
Fig. SI). Initial differences in C:N con ten t betw een litter types also
persisted over the course o f the experim ent {P < 0.0001; Table 1).
© 2012 BlackweU PubUshing L td /C N R S
Litter chemical com position changed substantially over the course o f
three seasons and differed am ong the three m anagem ent systems.
Relative to the initial litter, the abundances o f lipids and N -bearing
com pounds declined in corn litter during years 1 and 2. Grass litter
showed a general increase in polysaccharides over time and a marginaUy significant increase from time 0 to year 2 (P = 0.058). T otal litter
C and N con ten t were b o th significantly greater in O F systems than
in agricultural systems (P < 0.0001); however, C:N ratios were n o t
affected by m anagem ent during decom position (Table 1). A t
17 days, the relative abundance o f phenols and com pounds o f
unknow n origin in grass litter differed significantly am ong m anage­
m ent systems (OF > N T , O F > CT = NT). Additionally, at the end
o f year one lipids and N -bearing com pounds were relatively m ore
abundant in grass litter decom posed in O F than in agricultural sys­
tems (Fig. 2, Fig. SI). T he relative abundance o f N -bearing com ­
pounds rem ained higher in O F than CT systems in year 2, b u t this
effect did n o t persist into year 3. Polysaccharide relative abundance
was higher in CT systems than in O F during year 2 and higher
in year 3 than b o th N T and O F systems. Lipid relative abundance in
corn litter also differed significantly during year 3 and was higher in
N T than in CT systems (Fig. 51).
Enzymes a n d d e c o m p o s e r com m u n itie s
T he activities o f all three hydrolytic enzymes were highest in year 1
(Fig. 3). Activities were similar betw een years 2 and 3, although they
were greater in year 3 than year 2 for acid phosphatase, and greater
in year 2 than 3 for phenol oxidase (Fig. 3, Table 51 5upporting
Inform ation). Activities were also greater in grass than in corn w hen
averaged across all dates (e.g. P-Af-acetylglucosaminidase: 2.77 vs.
1.62; (3-glucosidase: 7.83 vs. 2.89; and acid phosphatase: 2.27 vs.
1.92 m ol mg ^ dry litter, respectively; Fig. 3, Table 51). In contrast,
phenol oxidase activity was significantly greater in corn than in grass
(2.19 vs. 1.73 m ol m g ^ dry litter, respectively). Litter type X year
interactions were present for all enzymes; P-Af-acetylglucosaminidase,
(3-glucosidase and acid phosphatase activities differed betw een corn
and grass in year 1 and phenol oxidase differed betw een corn and
grass in year 2 (Table 51). M anagem ent intensity also had significant
effects o n enzyme activities regardless o f litter type (Fig. 3, Table
51). T he activity o f P-Af-acetylglucosaminidase was greater in O F
than in N T and greater in N T than in CT. (3-glucosidase was
greater in O F and N T than in CT, and acid phosphatase activity
was greater in O F than b o th N T and CT systems. In contrast to
the response o f hydrolases, phenol oxidase activity was greater in
CT systems than in N T and OF. M anagem ent effects o n enzymes
also varied by year; while (3-glucosidase and acid phosphatase were
affected by m anagem ent in year 1 only, m anagem ent effects were
Letter
Origin o f litter c h e m ic a l c o m p le x it y 5
(3-N-acetylglucosaminidase
(3-glucosidase
14-1
12-
,0
35-1
□ --2010
zulU
□
□ - 2- 02009
09
■ - 2008
30-
----_
25201510-
5
0
J
lOn
Phenol oxidase
10
Acid phosphatase
o
6
MMlU i j j
-
1400-1
0.50n
Detritivores
1200-
I
(1)
>
"
Fungi: Bacteria
0.40-
lO O O H
800 ■
0.30-
600-
0 .2 0 -
“::iL
0 . 10-
C G
Conventional
C G
No-till
C G
Old field
0.00
C
G
C
G
Conventional No-till
C
G
Old field
Figure 3 Extracellular enz 5Ane activities (m ol m g ^ litter), d etritivore densities (individuals g ^ litter) and fungal-to-bacterial ratios (F:B) during three years o f co rn
(C) and grass
(G) litter d eco m p o sitio n in conventionally tilled
(conventional), no-tdll and old field treatm ents. F o r enzym es and detritivores, bars represent
cum ulative w ithin-year activities (enz 5mies) o r densities (detritivores) (black — 2008, gray — 2009, w hite — 2010). Fungal-to-bacterial ratios are color coded by year in
the sam e m anner.
present in P-V-acetylglucosaminidase and phenol oxidase for years
1 and 2 (Fig. 3, Table 51).
Cumulative detritivore density was greatest in year 1. T here was a
significant m anagem ent X litter type interaction and densities were
higher in CT systems than in N T or O F systems in grass litter
(Fig. 3, Table 51). Fungal-bacterial (F:B) ratios were also signifi­
cantly altered by litter and m anagem ent type. F:B ratios were consis­
tently higher in corn than in grass; however, the effect o f
m anagem ent varied by year; during year 1 F:B ratios were greater in
O F than in agricultural systems b u t in year 3 the p attern was
reversed (Fig. 3, Table 51).
In tera ctio n s b e t w e e n d e c a y rate, c hem istry a n d d e c o m p o s e r
co m m u n itie s
T he pattern and frequency o f relationships betw een mass loss, litter
chemistry, and the com position and activity o f the decom poser
com m unity differed by litter type (Table 2, Table 52 5upporting
Inform ation). C orn and grass litter mass rem aining during the
experim ent were both significantly correlated w ith integrated
enzyme activities (Table 52). Enzym e efficiencies
calculated
from the slope o f mass loss against integrated enzyme activities
(5insabaugh et al. 2002), representing the percentage o f mass loss
p er unit enzyme activity, were greater in grass litter than corn litter
(Fig. 4, Table 52). H ydrolase efficiencies were also highest in O F
treatm ents followed by N T and CT treatm ents, while phenol oxi­
dase exhibited the opposite trend (Fig. 4, Table 52). Relationships
betw een chemistry and biology in corn litter w ithin individual stages
were rare and only a single significant correlation occurred betw een
70 and 80% mass loss (Table 2). W hen all stages were grouped,
there were relationships betw een phenol oxidase activity and the rel­
ative abundance o f polysaccharides (-L) and lignin (—), as well as
w ith the ratio o f lignin plus phenols to polysaccharides plus N -bear­
ing com pounds (—) (Table 2). In contrast, many correlations
betw een litter chem istry and decom posers occurred w ithin every
stage o f grass litter decom position. F or example, the activities o f (3Af-acetylglucosaminidase and (3-glucosidase were positively corre­
lated with the relative abundance o f N -bearing com pounds both
w ithin and across decay stages. T he ratio o f phenol oxidase to total
hydrolase activity (the sum o f all three hydrolytic enzymes) was also
© 2012 Blackwell Publishing L td /C N R S
et al.
6 K. W ic k in g s
Letter
T a b le 2 Pearso n coefficients (r) fo r correlations betw een chem istry (individual classes and ratios), and biology (Individual enzym e activities and ratios, detritivore density
and F:B) w ithin and across d eco m p o sitio n stages (Full) In co rn and grass htter
C o rn
Correlates
30%
Polysaccharides
PO X
D etritivores
N -bearing
Grass
70%
Full
80%
30%
80%
95%
97%
Full
0.13
0.33
0.57
0.41
0.33
0.56
0.95
0.61
0.08
0.35
0.43
0.16
0.34
0.47
0.88
0.70
0.69
0.26
NAG
- 0 .1 2
0.33
- 0 .0 7
- 0 .0 4
0.29
0.91
0.77
- 0 .1 8
0.36
-0 .2 4
-0.09
0.74
0.89
0.76
0.01
-0 .3 2
0.55
BG
0.53
- 0 .1 8
- 0 .1 9
0.09
-0 .0 4
- 0 .7 1
- 0 .8 1
0.44
0.15
NAG
- 0 .3 8
0.12
- 0 .5 4
-0 .1 1
0.68
0.85
0.77
-0 .1 4
0.43
BG
- 0 .3 7
0.21
- 0 .4 3
- 0 .0 7
0.84
0.86
0.37
- 0 .0 4
0.39
0.28
0.67
- 0 .2 5
- 0 .6 9
- 0 .5 6
- 0 .3 9
0.12
0.37
0.08
D etritivores
0.43
Proteins
PO X
Lignin
BG
0.07
- 0 .1 4
0.39
- 0 .0 3
- 0 .7 1
0.24
0.26
0.50
0.37
PO X
- 0 .2 3
- 0 .1 0
-0 .6 2
- 0 .4 2
0.42
- 0 .1 2
-0.65
-0 .1 1
0.19
PHOS
-0 .3 4
0.24
0.57
O.Sl
0.45
0.75
0.03
- 0 .2 6
- 0 .3 7
PHOS
-0 .4 4
0.63
0.47
- 0 .0 2
0.77
0.73
0.29
-0 .2 1
-0 .6 6
P O X iH Y D R O
- 0 .0 5
0.46
- 0 .3 6
0.09
0.90
0.83
0.08
-0 .0 1
0.40
- 0 .2 6
- 0 .1 5
-0 .6 8
- 0 .3 9
0.31
0.65
- 0 .7 1
- 0 .2 0
-0 .1 1
Lipids
Phenols
Hg:N
(iig+phen:ps+N )
PO X
hg-hgnln, N -nltrogen-bearlng, P O X -p h e n o l oxidase, N A G -P-N -acetylglucosam lnldase, B G -jl-glucosidase, P H O S -acld phosphatase, H Y D R O -su m o f N A G , B G , P H O S.
*« = 9 w ithin stages, 27 an d 36 fo r co rn and grass, respectively, across stages (Fuh). Values in bo ld Indicate significant correlations (P < 0.05).
60 u
significant stage-specific correlation betw een litter biology and
chemistry in corn, while there were 25 in grass.
□ POX
g
BG
40
DISCUSSION
CT
NT
Com
CT
NT
Grass
OF
Figure 4 P h en o l oxidase (POX ) an d jl-glucosidase (BG) enzym e efficiencies (k^)
calculated o v er 730 days o f co rn and grass h tter d eco m p o sitio n in conventionahy
thled (CT), no -th l (N T), and old field (OF) treatm ents. E n zy m e efficiencies are
calculated as the slope o f th e regression betw een % m ass loss and Integrated
enzym e activity (m ol m g ^ htter). A n Increase In efficiency Indicates greater htter
mass loss p e r unit enzym e activity. T h e g rap h shows: (1) greater overah enzym e
efficiency in grass th a n c o rn htter; (2) variation betw een m anagem ent treatm ents,
particularly In grass; and 3) the ratio o f p h e n o l oxidase to hydrolase (represented
by B G ) efficiency Is higher In co rn h tte r reflecting its higher hgnin c o n te n t and
recalcitrance and higher in C T th a n N T o r O F.
positively correlated with lignin to N ratio during the first year o f
decom position, as well as w hen all stages were combined. Some
relationships changed over stages; for example, detritivore densities
were correlated w ith the relative abundance o f polysaccharides (+)
and N -bearing com pounds (—) during tw o to three different decay
stages but n o t across all the stages. In total, there was one
© 2012 Blackwell P ublishing L td /C N R S
I f the ‘Chemical Convergence H ypothesis’ is true, we w ould expect
th at the chemical com position o f corn and grass litter would
converge during decom position regardless o f differences in initial
litter quality or decom poser com m unity characteristics, and that
the chemistry o f the different litter types would be indistinguish­
able in late stages o f decom position (Tig. I). Past studies have
argued th at convergence in litter chemical com position should
occur at roughly 75—80% mass loss (MeUUo et a l 1989; Preston
et a l 2009; M oore et al 2011), yet at this stage we found that sig­
nificant chemical differences persisted (e.g. polysaccharides: 27 vs.
17%; proteins: I vs. 5%; and N -bearing com pounds: 3 vs. 6% in
corn and grass, respectively; P < 0.05 for each) (Fig. 2, Fig. SI).
M oreover, novel differences in litter chemistry also emerged during
decom position; lignin and phenols were initially similar in corn and
grass litter (Fig. SI) b u t differed substantially at 75—80% mass
rem aining (lignin: 47 vs. 28%; and phenols: 2 vs. 8% in corn and
grass, respectively; P < 0.05). Together, these results indicate that
litter chemistry before decom position influences its chemistry
throughout the decay process, supporting the ‘Initial Litter Quality
H ypothesis’ (Fig. I).
A long w ith evidence supporting the ‘Initial Litter Quality H ypoth­
esis’, we also found th at the chemistry o f a single litter type
diverged w hen it was decom posed in m anagem ent systems w ith dis­
tinct decom poser comm unities (Fig. 2). Additionally, once chem is­
tries diverged, they rem ained distinct over long-term decom position,
Letter
providing strong evidence for the ‘D ecom poser C ontrol H ypothesis’
(Fig. 1). F o r example, the relative am ounts o f phenols, lipids, poly­
saccharides and N -bearing com pounds in decom posing grass litter
varied betw een agricultural- and O F-decom posed litter, and the dif­
ferences in chemistry were associated with low er hydrolytic and
higher oxidative enzyme efficiencies, plus greater detritivore densi­
ties in agricultural systems (Figs 3, 4 and Table SI). O nly a handful
o f studies have dem onstrated that the chemistry o f a single litter
species can diverge w hen decom posed in different ecosystems or by
different decom posers (Steffen et al. 2007; FUley et al. 2008b;
W allenstein et al 2010), and ours may be the first to show that
m anagem ent-induced differences in decom poser comm unities can
alter litter chemistry over long-term decom position.
O ur data p o in t to the key role o f decom poser comm unities in
regulating changes in litter chem istry during decom position; how ­
ever, the specific effects seem to depend strongly o n initial litter
quality. Thus, the data underscore that the three hypotheses p re­
sented here (Fig. 1) are n o t mutually exclusive, as the controls on
litter chemistry during decom position clearly interact and vary
depending on litter quality. W hereas corn litter chemistry suppressed
decom poser activity and density, grass litter p rom oted active biolog­
ical com m unities that varied betw een m anagem ent practices and
exhibited strong relationships w ith litter chemistry. F or example,
grass litter polysaccharide abundance differed betw een m anagem ent
treatm ents and was related to detritivore density th roughout the
study. Furtherm ore, N -bearing com pounds differed betw een treat­
m ents in years one and two, corresponding w ith positive relation­
ships betw een P-W-acetylglucosaminidase and (3-glucosidase activities
and N abundance. However, the specific relationships betw een biol­
ogy and chemistry were dynamic and their strength and direction
fluctuated by decom position stage. F or example, while phenol abun­
dance was affected by m anagem ent and strongly correlated with
acid phosphatase at 30% mass loss (17 days in the field; Fig. SI),
the relationship appeared to change later in decom position and the
fuU m odel correlation betw een phenol abundance and acid p h o s­
phatase was negative. Such fluctuating relationships betw een biology
and chemistry may indicate that decom position is regulated by bo th
top-dow n (i.e. em ergent properties o f decom poser communities)
and bottom -up (i.e. litter quality) controls. F or instance, while acid
phosphatase activity may result in a relative accum ulation o f p h en o ­
lic com pounds early in decay, litter phenolic concentration may, over
the fuU course o f decom position, suppress phosphatase activities.
In contrast to grass litter, differences in corn litter chemistry
am ong m anagem ent practices were n o t apparent until after 730 days
o f decom position (corresponding with c. 80% mass loss), w hich was
the same time that interactions betw een corn litter chemistry and
decom posers emerged (Table 2). Relationships betw een corn litter
chemistry and biology were stronger across the entire study than
w ithin specific stages, and primarily involved either phenol oxidase
or lignin. Furtherm ore, enzyme efficiencies were markedly low er in
corn than in grass litter indicating the requirem ent for greater enzyme
activity to decom pose the same am ount o f C (Fig. 4, Table S2). T he
greater tim e required for m anagem ent effects o n corn chemistry to
emerge, the im portance o f lignin and lignin-degrading enzymes in
corn, and the lower enzyme efficiencies all suggest th at corn litter
chemistry suppressed decom poser com m unity activity, decom position
rates, and slowed the emergence o f decom poser-driven differences
in chem istry am ong m anagem ent treatm ents. These findings support
the hypothesis that decom poser com m unities may drive changes in
Origin o f litter c h e m ic a l c o m p le x it y 7
litter chemistry during decom position b u t th at their effects are
strongly influenced by initial litter quality.
W e acknowledge th at factors such as climate and soil type might
also explain m anagem ent-induced changes in litter chem istry during
decom position, b u t to minimise these effects we used a com m on
experimental site with consistent soil types and climate. Nonetheless,
nutrient content differed betw een o ur agricultural and O F systems
due to fertilisation. D uring corn-growing years, for example, soil N
availability was r. 4 X greater in CT and N T treatm ents than in O F
(21 vs. 5 pg g \ respectively) (R obertson 2009). Furtherm ore, litter
inputs were m ore diverse in the O F b u t m ore abundant in the agri­
cultural sites. Soil nutrient availability can influence microbial N
dem and and resource partitioning am ong soil biota w ith diverse stoi­
chiom etric requirem ents (Paul & Clark 1996; P arton et al. 2007), and
N availability is know n to influence hydrolase and oxidase enzyme
activities (Carreiro et al. 2000; Sinsabaugh et al. 2002; Keeler et al.
2009). C onsistent w ith this, we found that O F systems had higher
hydrolytic and lower oxidative enzyme activities and efficiencies
(Figs 3, 4 and Table S2) relative to agricultural systems. W hile differ­
ences in resource dynamics are unlikely to directly explain variations
in litter chemistry during decom position, they may have played a
role in shaping decom poser comm unities during decom position.
H ow do we reconcile our results w ith other studies that have
reported that litter chem istry converges during decom position irre­
spective o f m anagem ent type or initial litter quality? First, we sug­
gest th at the difference in initial litter quality betw een this study and
others offers one possible explanation. T he initial difference in N
concentration betw een corn and grass litter (0.7 vs. 2.4%, respec­
tively) was m ore than twice as large as that from MeUUo et al.
(1989), and the litter types used in our experim ent had lower initial
lignin to N ratios than those used in the litter bag studies by others
(e.g. MeUUo et al 1989; M oore et al 2011). AdditionaUy, we used
very different m anagem ent systems w ith know n differences in
decom poser comm unities whereas past studies have used a nar­
row er range o f systems (typicaUy forests).
It is also possible that the high-resolution molecular technique we
used to measure Utter chemistry explains w hy we observed differ­
ences in chem istry after extensive decay w here others have not.
M any o f the studies reporting convergence in the chemistry o f dif­
ferent Utter types measured broad elemental classes (e.g. total Utter
C and N) or used chemical digestion (e.g. proxim ate analysis) to
analyse changes in chem istry over tim e (e.g. MeUUo et al. 1989;
M oore et al. 2011). WhUe changes in total soU C and N concentra­
tions are im portant and informative, they teU us Uttle about quaUtative shifts w ithin these broad elemental groups. F or example, we
observed substantial differences in the chemical structure o f C-containing com pounds (e.g. Ugnin derivatives, Upids and phenols) and
N -containing com pounds (e.g. proteins and other N -bearing com ­
pounds) betw een and w ithin Utter types that could n o t be detected
using elemental and proxim ate analyses (Fig. SI). Advanced
chemical m ethods such as py-GCMS may, therefore, aUow a m ore
mechanistic understanding o f the decom position process. W e sug­
gest th at m ore widespread appUcation o f these m ethods in Utter
decom position studies could considerably enhance our understand­
ing o f Utter decom position dynamics and their relationship to
decom poser communities.
T here are som e caveats associated w ith the Utter bag m ethod
(Kampichler & Bruchner 2009; Cotrufo et al 2010) that may have
also influenced o u r results and their interpretation. F or example, Ut­
© 2012 Blackwell Publishing L td /C N R S
8 K. W ic k in g s
et al.
ter bag m easurem ents do n o t encompass organo-m ineral complexes,
including the incorporation o f partially decom posed litter (e.g. par­
ticulate SOM) into aggregates (Six et al. 2002). Aggregate incorpora­
tion o f partially decom posed litter could serve to enhance, maintain,
or even prevent further differences in litter chemistry (G randy &
N e ff 2008). Thus, we cannot confirm w hether o ur observed differ­
ences in litter chemistry w ould persist in old SO M — a contentious
aspect o f current soil organic m atter models (e.g. Six et al 2006;
G randy & N eff 2008; Schm idt et al. 2011; Schnitzer & M onreal
2011). However, other py-GCM S studies have observed that chem i­
cal variation in SO M pools remains high (e.g. G randy et al 2007),
providing evidence that diverse chemical inputs may have distinct,
long-term effects on SO M chemistry. By limiting contact w ith the
surrounding litter, litter bags may also inhibit possible cometaboHsm
o f labile and recalcitrant com pounds, w hich may be a com m on
m echanism for the decom position o f recalcitrant com pounds
(K lotzbocher et al 2011).
This study argues against the current paradigm suggesting that
chemical convergence occurs over the course o f litter decom posi­
tion, and dem onstrates that interactions am ong initial litter quality
and decom poser com m unities increase the chemical complexity o f
decom posed litter (Figs 1 and 2). U nderstanding such changes to
chemical complexity has significant implications, as the accumula­
tion o f chemically distinct organic m atter in soil w ould have num er­
ous im portant effects o n comm unity- and ecosystem-level
processes. Perhaps m ost im portantly, this chemical variation could
lead to variation in resource utilisation patterns in soil food webs,
alter aggregate form ation and associated soil characteristics including
erosivity and bulk density, and ultimately contribute to variation in
SO M stability, soil-atm osphere C O 2 exchange, and global soil
C storage.
Letter
Balser,
T.C.
&
F irestone,
M .K .
(2005).
Linking
m icrobial
com m unity
c om position and soil processes in a C alifornia annual grassland and mixedconifer forest. Biogeochemistty, 73, 395—415.
B aum ann, K ., M arschner, P., Smernik, R J. & B aldock, J.A. (2009). Residue
chem istry
and
m icrobial com m unity
structure
during
d e com position
of
eucalypt, w heat and vetch residues. Soil Biol Biochem., 41, 1966—1975.
Berg, B. & M cC laugherty, C. (2008). Plant Utter: Decomposition, H um us Formation,
Carbon Sequestration, 2nd edn. Springer-V erlag, Berhn.
Carreiro, M .M ., Sinsabaugh, R.L., R epert, D .A . & Parkhurst, D .F .
(2000).
M icrobial enz 5mie shifts explain h tte r decay responses to sim ulated nitrogen
deposition. Ecology, 81, 2359—2365.
CavigeUi, M .A., R obertson, G .P. & King, M.J. (1995). Fatty acid m ethyl ester
(FAM E) profiles as m easures o f soil m icrobial com m unity structure. Plant Soil,
170, 99-113.
Chivenge, P., V anlauw e, B., G entile, R. & Six, J. (2011). O rganic resource quaUty
influences short-term aggregate dynam ics and soil organic c arbon and nitrogen
accum ulation. Soil Biol Biochem., 43, 657—666.
C onn, C. & D ighton, J. (2000). L itter quaUty influences o n decom position
ectom ycorrhizal com m unity
structure
and
m ycorrhizal ro o t
surface
acid
ph osphatase activity. Soil Biol Biochem., 32, 489—496.
C otrufo, M .F., N gao, J., M arzaioh, F. & P ierm atteo, D . (2010). Inter-com parison
o f m ethods for quantifying above-ground leaf h tte r decom position rates. Plant
Soil, 334, 365-376.
C rum , J.R. & CoUins, H .P. (1995). Soil description: K BS soils. K e h o ^ Biological
Station L T E R . AvaUable at: w w w .lter.kbs.m su.edu/research/site-descriptionand-m aps/soil-description. L ast accessed 31 D ecem b er 2009.
D eleporte, S. & Charrier, M. (1996). C om parison o f digestive carbohydrases
b etw een tw o forest sciarid (D iptera: Sciaridae) larvae in relation to their
ecology. Pedobiologia, 40, 193—200.
Fierer, N ., Jackson, J.A ., VUgalys, R. & Jackson, R.B. (2005). A ssessm ent o f soil
m icrobial com m unity structure by use o f taxon-speciflc quantitative PC R
assays. A p p l Environ. Microbiol, 71, 4117M-120.
Fierer, N ., G randy, A.S., Six, J. & Paul, F.A . (2009). Searching for unifying
principles in soil ecology. Soil Biol Biochem., 41, 2249—2256.
Fhley, T.R., B ou tto n , T.W ., Liao, J.D ., Jastrow , J.D . & G am bhn, D .F . (2008a).
C hem ical changes in nonaggregated particulate soil organic m atter fohow ing
grassland-to-w oodland transition in a subtropical sa v a n n a ./. Ceophys. Res., 113,
ACKNOWLEDGEMENTS
W e thank Cynthia KaUenbach, Lisa Tiem ann, M arshall McDaniel,
A m anda D aly and Sarah Castle for com m ents o n earlier drafts
o f the manuscript. Funding for this w ork was provided by the
W .K. Kellogg Biological Station L ong T erm Ecological Research
Program , the Michigan Agricultural E xperim ent Station, the Univer­
sity o f N ew H am pshire Agricultural E xperim ent Station, the
N ational Science Foundation and the U nited States D ep artm en t o f
Agriculture. Any use o f trade names is for descriptive purposes only
and does n o t imply endorsem ent by the U.S. governm ent.
G 03009, doi: 10.1029/2007JG 000564.
Fhley, T.R ., M cC orm ick, M .K ., C row , S.E., Szlavecz, K ., W higham , D .F .,
J o h n so n , C.T. et a l (2008b). C om parison o f the chem ical alteration trajectory
o f L iriodendron tuhpifera L. leaf h tte r am ong forests w ith different invasive
earthw orm a c tiv ity ./. Ceophys. Res., 113, G 01027, doi:10.1029/2007JG 000542.
G aho, M .F ., L auber, C.L., Cabaniss, S.F., W aldrop, M .P., Sinsabaugh, R.L. &
Zak, D .R .
(2005). Soil organic m a tte r and htter chem istry response to
experim ental N deposition in n o rth e rn tem perate deciduous forest ecosystem s.
Clob. Change Biol, 11, 1514—1521.
G randy, A.S. & N eff, J.C. (2008). M olecular C dynam ics dow nstream : T h e
biochem ical d e com position sequence and its im pact o n soil organic m atter
structure and function. Sci Total Environ, 404, 297—307.
G randy, A.S., N eff, J.C. & W eintraub, M .N . (2007). C arb o n structure and
enzym e activities in alpine and forest ecosystem s. Soil Biol Biochem., 39, 2701—
AUTHORSHIP
2711.
G randy, A., Sinsabaugh, R., N eff, J., Stursova, M. & Zak, D . (2008). N itrogen
K W and A SG designed the experiment. K W conducted the experi­
m ent and analysed the data. SR and CC conducted quantitative
PCR analyses. AU authors contributed to interpreting the data and
writing the manuscript.
d eposition effects o n soil organic m atter chem istry are hnked to variation in
enzym es, ecosystem s and size fractions. Biogeochemistty, 91, 37M-9.
G randy, A.S., Strickland, M.S., Lauber, C.L., B radford, M .A. & Fierer, N . (2009).
T h e influence o f m icrobial com m unities, m anagem ent, and soil texture o n soil
organic m atter chem istry. Ceoderma, 150, 278—286.
H ofm ockel, K ., Zak, D .R . & B lackw ood, C.B.. (2007). D oes atm ospheric N 0 3 -
REFERENCES
d eposition alter the abundance and activity o f hgnolytic fungi in forest soils?
Ecosystems, 10, 1278—1286.
A dair, E.C ., P arton, W.J., D e l G ro sso , S.J., Silver, W .L., H arm o n , M .E ., Hall, S.A.
Jenkinson, D.S. & Ladd, J.N . (1981). M icrobial biom ass in soil: m easurem ent
et a l (2008). Sim ple th ree-p o o l m o d el accurately describes p attern s o f long­
and turnover. In: Soil Biochemistry, V o l 5 (eds Paul, F.A . & Ladd, J-N.).
term litter d eco m p o sitio n in diverse climates. Glob. Change B iol, 14, 2636—2660.
D ekker, N ew Y ork, pp. 415M 71.
A ngers, D .A . & M ehuys, G .R. (1990). Barley and alfalfa c ro p p in g effects o n
K am pichler, C. & B ruchner, A. (2009). T h e role o f m icroarthropods in terrestrial
carbohydrate contents o f a clay soil and its size fractions. Soil Biol Biochem., 22,
decom position: a meta-analysis o f 40 years o f htterbag studies. Biol. Rev., 84,
285-288.
375-389.
© 2012 B lackw eh P ubhshing L td /C N R S
Letter
Origin o f litter c h e m ic a l c o m p le x it y 9
K eeler, B.L., H obble, S.E. & K ellogg, L .E . (2009). E ffects o f long-term nitrogen
Six, J., C onant, R.T., Paul, E.A . & Paustian, K . (2002). Stabhization m echanism s
addition o n m icrobial enzym e activity in eight forested an d grassland sites:
o f soil organic m atter: im phcations fo r C -saturation o f soils. Plant Soil, 241,
im plications fo r litter an d soil organic m atter decom position. Ecosystems, 12, 1—
15.
K lotzbocher, T., Kaiser, K ., G uggenberger, G ., G atzek, C. & K albitz, K. (2011).
A new co n cep tu al m o d el for th e fate o f lignin in d ecom posing p la n t litter.
Ecology, 92, 1052-1062.
155-176.
Six, J., Frey, S.D., T hiet, R.K. & B atten, K.M . (2006). B acterial and fungal
contributions to C -sequestration in agroecosystem s. Soil Sci. Soc. Mm. J., 70,
555-569.
Snajdr, J., Cajtham l, T., V alaskova, V., M erhautova, V., Petrankova, M., Spetz,
K ogel-K nabner, I. (2002). T h e m acrom olecular organic co m p o sitio n o f p la n t and
P. et a l (2011). T ransform ation o f Quercus petraea htter: successive changes
m icrobial residue as inputs to soil organic m atter. Soil Biol. Biochem., 34, 139—162.
in htter chem istry are reflected in differential enzym e activity and changes
Ladd, J.N ., A m ato, M., Li-K ai, Y. & Schultz, J.E . (1994). D ifferential effects o f
rotation, p la n t residue an d nitro g en fertilizer o n m icrobial biom ass and
organic m atter in an A ustralian alfisol. Soil Biol Biochem., 26, 821—831.
M anerkar, M .A., Seena, S. & B arlc h er, F. (2008). Q -R T -P C R for assessing
archaea, bacteria an d fungi during leaf d e co m p o sitio n in a stream . Microb.
Ecol, 56, 467M 73.
M cG ill, W .B., (2007). T h e physiology and biochem istry o f soil organism s. In: Soil
Microbiology, Ecology and Biochemistry, 3 rd e d n (ed. Paul, E.A.). O x fo rd , U K , pp.
231-256.
M elillo, J.M ., A ber, J.D ., T.inkins, A .E ., Ricca, A., Fry, B. & N adelhoffer, K.J.
(1989). C arb o n an d nitro g en dynamics along th e decay continuum : p lan t htter
to soil organic m atter. Plant Soil, 115, 189—198.
303.
Steffen, K .T., C ajtham l, T., Snajdr, J.
degradation
of
oak
{Quercus petraea)
& B aldrian, P.
leaf
htter
by
(2007). D ifferential
htter-decom posing
basidiom ycetes. Res. Microbiol, 158, 447—455.
Stew art, C.E., N eff, J.C ., A m atangelo, K.L. & V itousek, P.M . (2011). V egetation
effects o n soil organic m atter chem istry o f aggregate fractions in a H aw ahan
forest. Ecosystems, 14, 382—397.
Strickland, M.S., L auber, C., Fierer, N . & B radford, M .A. (2009). Testing the
functional significance o f m icrobial com m unity com position. Ecolo^, 90, 441—
451.
V ancam penhout, K ., W outers, K., D e V os, B., B uurm an, P., Sw ennen, R. &
M oore, T.R ., T rofym ow , J.A ., P rescott, G.E. & Titus, B .D . (2011). N a tu re and
n u rtu re in the dynam ics o f C, N
in the m icrobial com m unity com position. F E M S Microbiol Ecol, 75, 291—
and P d u ring h tte r d eco m p osition in
C anadian forests. Plant Soil, 339, 163—175.
D eckers, J. (2009). D ifferences in chem ical com p o sitio n o f soil organic m atter
in natural ecosystem s from different chm atic regions — a pyrolysis-G C /M S
study. Soil Biol Biochem., 41, 568—579.
Parton, W., Silver, W .L., Burke, I.C., G rassens, L., H arm o n , M .E ., Currie, W.S.
et a l (2007). Glogal-scale similarities in nitro g en release pattern s d uring long­
term decom position. Science, 315, 361—364.
W ahenstein, M .D ., H ess, A., Lewis, M., Steltzer, H . & Ayres, E.
(2010).
D e co m p o se d aspen leaf htter m etabolom es differ w h en decom posed under
d ifferent tree species. Soil Biol Biochem., 42, 484—490.
Paul, E.A. & Clark, F.E. (1996). Soil Microbiology and Biochemistry. A cadem ic Press,
San D iego.
W ickings, K ., G randy, A.S., Reed, S. & C leveland, C. (2011). M anagem ent intensity
alters d e com position via biological pathw ays. Biogeochemistry, 104, 365—379.
Poirier, N ., Sohi, S.P., G aunt, J.L ., M ahieu, N ., Randah, E.W ., Pow lson, D.S.
et a l (2005). T h e chem ical c o m p o sitio n o f m easurable soil organic m atter
pools. Org Geochem., 36, 1174—1189.
P reston, C.M., N au lt, J.R. & T rofym ow , J.A . (2009). C hem ical changes during
6 years o f d e co m p o sitio n o f 11 htters in som e C anadian fo rest sites. P a rt 1.
E lem ental com position, tannins, phenohcs, and pro x im ate fractions. Ecosystems,
12, 1053-1077.
R obertson, G .P. (2009). M ain c ro p p in g system experim ent: KBS soil inorganic
nitrogen. K ehogg Biological S tation L T E R . K BS021-002.18. A vailable at:
w w w .lte r.k b s.m su .ed u /d atatab les/5 5 . L ast accessed 31 D ecem b er 2009.
Saiya-Cork, K .R., Sinsabaugh, R.L. & Zak, D .R . (2002). T h e effects o f long term
nitrogen dep o sitio n o n extracehular enzym e activity in an Mcer saccharum forest
soil. Soil Biol Biochem., 34, 1309—1315.
Schm idt, M .W .I., T o rn , M.S., A biven, S., D ittm ar, T.,
G uggenberger,
SUPPORTING INFORMATION
Additional Supporting Inform ation may be downloaded via the online
version o f this article at Wiley Onhne Library (www.ecologyletters.com).
As a service to o ur authors and readers, this journal provides sup­
porting inform ation supplied by the authors. Such materials are
peer-reviewed and may be re-organised for online delivery, b u t are
n o t copy-edited o r typeset. Technical support issues arising from
supporting inform ation (other than missing files) should be
addressed to the authors.
G.,
Janssens, LA. et al. (2011). Persistence o f soil organic m atter as an ecosystem
property. Nature, 478, 49—56.
Schnitzer, M. & M onreal, C.M. (2011). Q u o vadis soil organic m atter research?: a
biological h n k to th e chem istry o f hum ification. Mdv. Mgron., 113, 139—213.
Sinsabaugh,
R.L.,
Carreiro,
M .M .
&
R epert,
D .A .
(2002).
A h o cation
of
extracehular enzym atic activity in relation to h tter com position, N deposition,
Editor, N ancy Jo h n so n
M anuscript received 10 May 2012
First decision m ade 11 June 2012
M anuscript accepted 25 Ju n e 2012
and m ass loss. Biogeochemistry, 60, 1—24.
© 2012 Blackw eh Pubhshing L td /C N R S