Effects of a humic acid and its size-fractions on the bacterial community
of soil rhizosphere under maize (Zea mays L.)
Edoardo Puglisia, George Fragoulisa, Patrizia Ricciutib, Fabrizio Cappac, Riccardo Spaccinid,
Alessandro Piccolod, Marco Trevisana,*, Carmine Crecchiob
aIstituto di Chimica Agraria ed Ambientale, Università Cattolica del Sacro Cuore, Via Emilia Parmense 84, 29100 Piacenza, Italy
bDipartimento di Biologia e Chimica Agroforestale ed Ambientale, Università di Bari, Via Amendola 165/A, I-70126 Bari, Italy
cIstituto di Microbiologia, Università Cattolica del Sacro Cuore, Via Emilia Parmense 84, 29100 Piacenza, Italy
dDipartimento di Scienze del Suolo, della Pianta, dell’Ambiente e delle Produzioni Animali, Università di Napoli Federico II, Via Università 100, 80055 Portici, Italy
a r t i c l e i n f o
Received 19 May 2009
Received in revised form 29 July 2009
Accepted 31 July 2009
Available online 26 August 2009
a b s t r a c t
The effects of a humic acid (HA) and its size-fractions on plants carbon deposition and the structure of
microbial communities in the rhizosphere soil of maize (Zea mays L.) plants were studied. Experiments
were conducted in rhizobox systems that separate an upper soil–plant compartment from a lower com-
partment, where roots are excluded from the rhizosphere soil by a nylon membrane. The upper rhizobox
compartment received the humic additions, whereas, after roots development, the rhizosphere soil in the
lower compartment was sampled and sliced into thin layers. The lux-marked biosensor Pseudomonas flu-
orescens 10586 pUCD607 biosensor showed a significant increase in the deposition of bioavailable
sources of carbon in the rhizosphere of soils when treated with bulk HA, but no response was found
for treatments with the separated size-fractions. PCR–DGGE molecular fingerprintings revealed that
the structure of rhizosphere microbial communities was changed by all humic treatments and that the
smaller and more bioavailable size-fractions were more easily degraded by microbial activity than the
bulk HA. On the other hand, highly hydrophobic and strongly associated humic molecules in the bulk
HA required additional plant rhizodeposition before their bio-transformation could occur. This work
highlights the importance of applying advanced biological and biotechnological methods to notice
changes occurring in plant rhizodeposition and rhizosphere microbial activity. Moreover, it suggests cor-
relations between the molecular properties of humic matter and their effects on microbial communities
in the rhizosphere as mediated by root exudation.
? 2009 Elsevier Ltd. All rights reserved.
Humification of different organic materials for compost produc-
tion is an useful strategy for both waste disposal and soil fertility
increase. The evaluation of the effects of humified compost mate-
rials on soil–plant systems thus represents an important challenge
in a future scenario of increased population, larger amounts of
waste to be disposed, and progressive reduction of soil organic
matter levels due to intensive agricultural activity.
Humic acids (HA) are a major component of organic fertilizers,
and as heterogeneous molecules of different sizes that are self-or-
ganized in supramolecular conformations (Piccolo, 2002), they are
also the most reactive ones. HA effects on plant physiology are
mainly positive, and they include enhancement of biomass yields
(Ayuso et al., 1996; Arancon et al., 2006), induction of lateral roots
emergence and ATPase activity (Canellas et al., 2002), increase of
cell respiration and membrane uptake of nutrients, and exertion
of hormone-like activities (Nardi et al., 2002). Given the HA struc-
tural complexity, a structure–activity approach aimed at linking ef-
fects on plant physiology with specific humic chemical properties
is a difficult task. One way to partly reduce the HA heterogeneity
is to carry out a size-fractionation of humic matter and character-
ize the separated size-fractions by combined pyrolysis and NMR
spectroscopy (Piccolo et al., 2002). The characteristics of such more
homogenous humic molecules may be related to well defined soil–
plant process of ecological importance.
Rhizodeposition is an important ecological process, through
which plants supply microbes with readily available organic
carbon sources, increase the microbial biomass surrounding the
roots, and profoundly affect the activity and composition of micro-
bial communities. On the other hand, microbes mobilize nutrients,
primarily nitrogen, and render them available to plants, as de-
scribed by the ‘‘microbial loop” model (Bonkowski, 2004; Paterson,
2003). A multidisciplinary approach based on different physical,
chemical and molecular techniques is necessary to study plant
0045-6535/$ - see front matter ? 2009 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +39 0523599345; fax: +39 0523599217.
E-mail address: firstname.lastname@example.org (M. Trevisan).
Chemosphere 77 (2009) 829–837
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/chemosphere
rhizodeposition and its effects on soil microbial community struc-
ture. Rhizobox systems are useful tools to study root-induced
changes in soil properties with a minimum disturbance (Youssef
and Chino, 1989; Wenzel et al., 2001). In fact, a HA amendment
to the upper and physically separated rhizobox compartment al-
lows to evaluate its effects in the untreated lower rhizosphere
compartment. The amount of bioavailable C in rhizodeposits can
then be assessed by means of lux-marked biosensors such as Pseu-
domonas fluorescens pUCD607 (Standing et al., 2005; Puglisi et al.,
2008), while possible changes in the structure of microbial com-
munities are estimated by means of polymerase chain reaction
(PCR) of ribosomal DNA coupled to denaturing or temperature gra-
dient gel electrophoresis (Muyzer and Smalla, 1998).
The objective of this work was to evaluate whether a well-char-
acterized HA and its separated size-fractions had an effect on C
deposition of maize plants and on the structure of rhizosphere
microbial communities. The hypotheses were: (i) HA can affect
the amount of C deposited by plant roots; (ii) if rhizodeposition
is affected, then the structure of microbial communities is signifi-
cantly changed; (iii) it is possible to link the effects of humic mat-
ter to plant rhizosphere processes by indentifying the responsible
specific molecular properties. In order to test these hypotheses,
maize plants were grown in rhizobox systems amended in their
upper soil compartment with a soil-extracted HA and its size-frac-
tions separated by preparative high pressure size exclusion chro-
matography (HPSEC). The consequent carbon deposition in each
unamended rhizosphere lower layer was assessed with the lux-
marked P. fluorescence 10586 pUCD607 biosensor, while soil bacte-
rial communities were characterized by 16S PCR–DGGE applied to
2. Materials and methods
An orchard loamy soil (USDA) was collected near Matera, South-
East Italy. Soil samples (7 cm diameter cores) were randomly col-
lected from the surface layer (0–20 cm) at least 2 m far from trees,
sieved at 2 mm, pooled on site and stored at 4 ?C for maximum
4 weeks before starting the bulk soil microcosms and rhizobox
experiments. The soil had a loamy texture (2.47 g kg?1clay,
3.38 g kg?1silt, 4.15 g kg?1sand), a pH (H2O) of 6.7 and total or-
ganic C of 8.3 mg C g?1d.w. soil.
2.2. HA and fractionation by HPSEC
A representative soil HA was isolated from a Fulvudand soil
near Vico (Italy), by standard methods reported earlier (Piccolo
et al., 1999). The HA was titrated to pH 7.2 with a 0.5 M KOH solu-
tion in an automatic titrator (VIT 90 Videotitrator, Radiometer,
Copenhagen) under N2 atmosphere and stirring. After having
reached the constant pH 7.2, the solution containing potassium-
humate was left under titration for 2 h, filtered through a glass
microfibre filter (Whatman GF/C), and freeze-dried.
The HPSEC mobile phase, a NaCl/NaN3 (2.89 g L?1/0.3 g L?1)
solution, was used to dissolve the HA to reach a concentration of
0.6 g L?1. Preparative separation of HA was conducted with a Bio-
sep SEC-S-2000 (300 mm ? 21.2 mm internal diameter, i.d.) col-
umnpreceded bya Biosep
(78.0 mm ? 21.2 mm i.d.) by Phenomenex. A Gilson 305 pump, a
Gilson autosampler model 231, a Gilson FC205 fraction collector,
and a Gilson 116 UV detector (Gilson Inc., Middleton, WI, USA)
set at 280 nm were used to automatically isolate humic fractions
in continuous. The nominal molecular-weight range of the pre-
parative column was calibrated with polystyrene sulphonates of
SEC-S-2000 Guard Column
known molecular weights (Piccolo et al., 2002). The HA and stan-
dard solutions were injected with a Rheodyne rotary injector
equipped with a 5 mL loop and the elution run at a 1.5 mL min?1
flow rate. A Unipoint Gilson Software (Gilson Inc., Middleton, WI,
USA) was used to automatically record all chromatographic runs.
Three fractions (I–III) were collected during the HPSEC elution of
HA separation: I, between 26 and 38 min; II, between 38 and
50 min; III, between 50 and 98 min. The collected fractions were
dialyzed in dialysis tubes (1000 Da cut-off) against distilled water
until they were free of chloride, and freeze-dried. The preparative
HA elution was repeated automatically 100 times, and, for a total
injection of 300 mg, the recovery was 91, 92, and 98 mg for frac-
tions I, II, and III, respectively. All humic samples were character-
ized for their elemental content using a Fisons EA 1108
Elemental Analyzer (Fisons Instruments, Milano, Italy).
2.3. On-line thermochemolysis-GC–MS and CPMAS-NMR spectroscopy
Thermochemolysis-GC–MS and cross-polarization magic angle
spinning carbon-13 nuclear magnetic resonance spectroscopy
(CPMAS-13C NMR) analyses of the HA and its three size-fractions
were carried out as described in detail by Spaccini and Piccolo
(2007). All pyrolysis-TMAH GC–MS analyses were conducted in
triplicates. The relative abundance (%) of each compound was cal-
culated as the ratio of the area of each singular peak over the total
2.4. Plants cultivation and microtome sampling
All experiments were carried out in the rhizobox system de-
scribed by Wenzel et al. (2001). The system consists of a soil–plant
compartment (130 ? 64 ? 50 cm) with a narrow slit at its bottom
that can be vertically penetrated only by plant roots to form a root
plane. In the lower compartment (130 ? 30 ? 105 cm) such root
system is confined between a transparent acrylic window on one
side and a membrane inhibiting further root penetration on the
other side. Nylon membranes of 7-lm mesh width, which exclude
root hairs from penetrating into the rhizosphere, were used in the
experiment. The membrane separates the root plane from the rhi-
zosphere soil compartment. The soil humidity at the beginning of
the experiments was set to 50% WHC, and was then kept in the rhi-
zoboxes by means of glass fiber strips connected to a water
The following treatments were applied to the upper rhizobox
compartment as humic solutions (three replicates per each treat-
ment, each replicate being represented by a single rhizobox): (i)
control, no organic addition; (ii) 14 mg kg?1of bulk HA (HA-1);
(iii) 140 mg kg?1of the same HA (HA-10); (iv) 14 mg kg?1of frac-
tion I (FR-I); (v) 14 mg kg?1of fraction II (FR-II); and (vi)
14 mg kg?1of fraction III (FR-III). The amount of HA added to the
soil in this study accounts for the upper average found in soils as
dissolved organic matter (Lilienfein et al., 2004; Jokinen et al.,
2006), and it is similar to that applied in other HA amendment
studies (Arancon et al., 2006).
Maize (Zea mays L.) seeds were germinated in Petri dishes and
pre-grown in the upper compartment of each rhizobox (six seeds
per each box). Ten days after germination the upper and the lower
compartments of each rhizobox were connected and the roots al-
lowed penetrating through the slit. Twenty five days after germi-
nation the lower compartments, which did not receive any
amendment, were sampled using a microtome (Wenzel et al.,
2001), in order to obtain soil samples at specific distances from
the root plane (0–2, 2–4 and 6–4 mm). Soil samples from the upper
compartment of the control treatment (i.e., bulk soil without any
amendment) were also sampled. Plants were sampled from each
E. Puglisi et al./Chemosphere 77 (2009) 829–837
treatment at the same development stage when the root plane
fully developed in each rhizobox, and dry biomass values not sig-
nificantly differed among treatments. In all experiments the rhizo-
boxes were kept in a Binder KBWF 240 phytotrone (Tuttlingen,
Germany) in order to maintain the following controlled conditions:
16 h photoperiod at 120 lE s?1m2with day/night temperatures
30 ?C/20 ?C and 60% humidity.
DNA fingerprinting was also investigated in bulk soil micro-
cosms without plants following additions of humic solutions as de-
scribed for the upper rhizobox compartment. Analyses were
conducted for each microcosm made up with 100 g of soil in a
three replicate experiment. Microcosms were incubated up to
60 d at room temperature and 50% WHC and aliquots of 2 g were
sampled at 0, 10, 20, 30 and 60 d.
2.5. Bioassay protocol
Cells cultures and bioassays were conducted as described by
Yeomans et al. (1999). Cultures ofP. fluorescens pUCD607 (lux CDA-
BE from Vibrio fischerii, kanr, ampr; Hamin-Hanjani et al., 1993)
were grown in 250 mL Erlenmeyer flasks containing 100 mL of LB
broth at 25 ?C, and rotated at 200 rpm. Late exponential cells were
harvested by centrifugation (5 min at 4000 g) at OD550of 2.0, cor-
responding to a concentration of cells of around 3 ? 107mL?1
(Yeomans et al., 1999). The supernatant was decanted and the cells
resuspended in the same volume of C-free M9 minimal medium.
The process was repeated, and the final cell suspension shaken at
200 rpm in C-free M9 for 2 h at 25 ?C (starvation step). Kanamycin
was added at 50 lg mL?1to all cultures.
One milli Liter aliquot of starved cells was added to 100 mg
samples of soil collected from rhizoboxes in order to measure
in situ carbon depositions. After 20 min of contact, 800 lL of soil
and bacteria slurry was collected and measured for light emission.
2.6. Luminometry and biosensor response data
Light output was measured using a SystemSure luminometer
Model 18172 (Nova Biomedical Waltham MA, USA). In order to
compare results from different experiments, bioassays were cali-
brated on a dilution series of glucose standard solutions in the
range of concentrations between 1 and 10 mM of C content. Lumi-
nescence data were then converted from relative light units (RLU;
1 RLU equivalent to 1 mV per 10 s) to relative glucose–carbon units
(RGU), following interpolation with the calibration curve of glucose
standard solutions, and expressed in terms of lmol of glucose–C
equivalents per g of soil.
2.7. DNA extraction from soil and genetic fingerprinting
DNA fingerprinting was performed not only on soil samples col-
lected from rhizobox lower compartments, as described above, but
also on soil samples from soil microcosms amended as in rhizo-
boxes, but without maize plants (bulk). For the bulk soil experi-
ments, samples were analysed at different incubation times, in
order to assess the response of microbial communities to different
amendments at different sampling time. For DNA fingerprinting
analyses, the 2–4 and 4–6 mm layers from rhizobox samples were
pooled in a single 2–6 mm fraction.
Total DNA was extracted from 500 mg samples with a rapid and
efficient method of direct lysis. Soil samples were placed into a
bead beater vial and shaken in a Fast Prep 1 FP120 beater (BIO
101 Vista, CA, USA) at 5.5 m s?1for 30 s. Total DNA from soil sam-
ples was isolated with a fast DNA kit for soil (BIO 101 Vista, CA,
USA). Extracted DNA was analysed on 0.7% agarose gel containing
0.5 lg mL?1ethidium bromide.
Soil DNA was amplified in a PCR Sprint thermocycler (Hybaid,
Ashford, UK) with one of the universal primer sets for 16S rDNA,
968F-1401R (Heuer and Smalla, 1997), to obtain products of about
500 base pairs. The 968F primer was modified with a 40mer GC-
clamp to better separate PCR products in the DGGE gradient gels
(Muyzer et al., 1993). Each PCR mixture contained 50 ng of DNA,
1 ? reaction buffer implemented with 2.5 mM MgCl2, 50 pmol of
each primer, 0.2 mM of each dNTP, 3 U Taq-polimerase (Euroclone)
in a final volume of 50 lL. Bovine serum albumin (BSA) (4 lg) was
added to minimize any inhibition of amplification by organic com-
pounds co-extracted from soil. The PCR consisted of 3 min at 95 ?C
followed by 40 cycles, each one consisting of a denaturing step
(10 s at 95 ?C), primer annealing (20 s at 54 ?C), and an extension
step (40 s at 72 ?C). A final extension step (10 min at 72 ?C) was fi-
nally performed. Amplification products, together with a Low
Range ladder (1000–80 bp, MBI Fermentas), were checked by elec-
trophoresis on ethidium bromide stained 1.5% agarose gel run at
10 V cm?1in 0.5 ? TBE buffer.
PCR products were characterized by denaturing gradient gel
electrophoresis (DGGE) which allows the detection of DNA bands
representing dominant bacterial species in soil microbial commu-
nity (Muyzer et al., 1993). DGGE was performed with the Bio-Rad
Dcode system. PCR products, 8 lL, were loaded onto denaturing
gradients. Bacterial amplicons were separated by a 6% polyacrila-
mide gel (acrylamide: bisacrylamide, 37:1) with 45% (3.15 M
urea-18% v/v formamide) 60% (4.2 M urea-24% v/v formamide)
top filling gradient and run for 15 h at 75 V at 60 ?C in TAE buffer.
Sybr Green stained gels were recorded with a Bio-Rad Gel Doc
2000 Documentation system from Bio-Rad Laboratories (Hercules,
CA, USA), equipped with Quantity One software for comparison
and clustering profiles.
2.8. Statistical analyses
Different effects on biosensors response were studied by means
of fixed model analysis of variance (ANOVA) using SPSS software:
(1) Effect of sampling position (classification variable SAMP).
The sampling positions were: bulk 0 d (for control only), rhi-
zosphere 0–2 mm, rhizosphere 2–4 mm, rhizosphere 4–
(2) Effect of treatments (classification variable TREAT). The five
treatments were: control, HA-1, HA-10, FR-I, FR-II, FR-III.
(3) The interaction between the two previous effects (classifica-
tion variable TREAT ? SAMP).
All significant effects were confirmed by Tukey test (at P < 0.05)
for comparison of means. Genetic fingerprints were analysed using
the Quantity One software of the Bio-Rad Gel Doc image analyser
system. The similarity of the electrophoretic profiles was evaluated
by determining the Dice coefficients of similarity and by obtaining
the unweighted pair group with average linkage method (UPGMA)
3.1. Molecular characteristics of humic samples
The elemental analyses of HA and its size-separates (Table 1)
show how the compositional elements and their ratios were differ-
ently distributed when the bulk HA was separated in its size-frac-
tion during HPSEC elution. In particular, while the carbon, nitrogen,
and hydrogen content as well as the C/N ratios were reduced in the
size-separates in comparison to HA, the oxygen content was com-
plementary increased in the fractions. Moreover, the H/C and the
E. Puglisi et al./Chemosphere 77 (2009) 829–837
O/C ratios were larger in the fractions, especially in fraction III,
than in HA, thereby suggesting that the size-fractions were pro-
gressively more hydrophilic than the bulk HA. This is in line with
literature findings based on HPSEC isolations of size-fractions
(Conte et al., 2006, 2007).
The carbon distribution in humic samples was obtained by the
CPMAS-13C NMR spectra. The spectra (not shown) revealed signals
in the alkyl–C (0–50 ppm) and the N–C and O–C (50–110 ppm) re-
gions. The former region is composed by carbons in (CH2)n- and
terminal CH3groups typical of plants lipid compounds such as
waxes and aliphatic bio-polyesters. Plant woody tissues were also
indicated by the 56 ppm shoulder of methoxy groups on aromatic
rings of guaiacyl and siringyl units of lignin structures (Hatcher,
1987). The most dominant resonance in the 50–110 ppm region
is mainly assigned to monomeric units in oligo- and poly-saccha-
ridic chains of plant woody tissues (Vane et al., 2001). An intense
signal around 72 ppm corresponds to the overlapping resonances
of C2, C3 and C5 carbons in the pyranoside structure of cellulose
and hemicellulose, whereas signals at 106 ppm (sharp), 65 ppm,
and 82–85 ppm (shoulders) are assigned to the anomeric C1 car-
bon and the C6 and C4 carbons, respectively, (Atalla and Vander-
Hart, 1999). The aromatic region (110–160 ppm) revealed a
broad band around 130 ppm that is related to p-hydroxy phenyl
rings of cynnamic units in both lignin and suberin biopolymers
(Hatcher et al., 1995). A prominent signal for quaternary carbons
at 172 ppm is currently assigned to carboxyl groups.
The spectra showed a general redistribution of the different
classes of carbon components when passing from bulk HA to
size-fractions. The major feature was the progressive relative in-
crease of signals for alkyl and carbohydrate carbons with decreas-
ing size of fractions. The relative carbon distribution is also shown
in Table 1, where signal integrations are reported. The size-frac-
tions were richer in alkyl carbon (0–50 ppm) than the bulk HA,
but its content somewhat decreased with decreasing size of frac-
tions. Similarly, the carbohydrate carbon, mainly represented in
the 50–110 ppm interval, increased significantly in the size-sepa-
rates and especially in the fraction of lowest size. Conversely, the
content of aromatic carbon (100–160 ppm) and carboxyl carbon
(160–190 ppm) was lower in size-fractions than in bulk HA. This
relative distribution suggests that Fraction III contained more
Elemental analysis (%), elemental ratios, and relative distribution (%) of signal area over chemical shift regions (ppm) in13C-CPMAS-NMR spectra of humic samples.
Humic samplesC (%) N (%)H (%)O (%) C/NH/CO/C
aHydrophilic carbons/Hydrophobic carbons = [(50 ? 110) + (160 ? 230)]/[(0 ? 50) + (100 ? 160)].
Relative abundance (%) and compositionaof main termochemolysis products released from bulk HA and size-fractions.
Identified productsBulk HA Fraction I Fraction IIFraction III
Average sum43.921.1 24.114.0
Other aromatic compounds
Average sum of aromatics
Saturated fatty acids10.4 (±1.81)
Unsaturated fatty acids
a.x-Alkane dioic acids
Average sum21.3 31.943.9 42.3
aTotal range varying from Ci to Cj; compounds in parentheses are the dominant homologous.
E. Puglisi et al./Chemosphere 77 (2009) 829–837
hydrophilic carbon than the rest of humic samples, as it is also
indicated by the slightly increasing HI/HB value for this fraction.
These NMR results are in line with the findings by elemental anal-
yses, that also showed the largest oxygen content for fraction III.
The classes of compounds identified in Total Ion Chromato-
grams (TIC) derived from the on-line thermochemolysis of humic
samples, as well as their relative content, are shown in Table 2.
The majority of compounds derived from higher plants, and was
represented by lignin, waxes and aliphatic biopolymers, as already
found for other bio- and geochemical materials (Guignard et al.,
The original content of lignin products in HA was spread over
the three size-fractions by the HPSEC treatment, with the least
amount found in the smallest size-fraction III. For the rest of the
identified products, a shift of relative content was observed from
the bulk HA to the fractions. In particular, aromatic compounds
(including some polyaromatic hydrocarbons), which are not imme-
diately related to lignin, were less prominent in fraction II than in
bulk HA. However, the sum of average content of lignin plus other
aromatic compounds showed that aromaticity was larger in the
bulk HA and decreased in fractions with molecular size (Table 2).
Among the alkyl products, also the saturated fatty acids, the al-
kanes/alkenes, and the a,x-alkanedioic acids were found to be
more present in the size-fractions than in bulk HA, while the unsat-
urated fatty acids showed slightly less relative content, except for
fraction II. Sterols decreased in relative importance passing from
the bulk HA to the fractions of progressively decreasing size. Terpe-
noids were detected only in bulk HA and fraction II. The average
sum of alkyl compounds revealed a decreasing trend with molecu-
lar size of fractions (Table 2). Among the hydrophilic components
detected by on-line thermochemolysis, protein derivatives were
more important in the TIC of size-fractions, while polysaccharides
derivatives were more prominent in fractions I and III.
3.2. Response of P. fluorescens pUCD607
The linear regression obtained with glucose standards was used
to convert RLU values into RGU (relative glucose–carbon units), de-
fined as a mol of available C per g of soil. The choice of glucose as
standard is justified by the evidence provided by Yeomans et al.
(1999), who showed that in the first weeks of maize growth the
P. fluorescens pUCD607 biosensor response to root exudates was
very similar to that of a glucose standard.
Biosensor responses for each treatment at each sampling posi-
tion are reported in Fig. 1, together with the results of Tukey’s test
for comparison of means for the TREAT ? SAMP effect. The treat-
ment effect (TREAT) and the interaction between treatment and
sampling position effect (TREAT ? SAMP) were statistically signifi-
cant for the biosensor response (F values of 46.91 and 12.61,
respectively). In particular, the HA-10 treatment (Fig. 2) had the
largest biosensor response (1.39 RGU), followed by the HA-1 treat-
ment (0.88 RGU). The three HA size-fractions and control had sig-
nificantly lower RGU values (FR-I: 0.59; FR-II: 0.57; FR-III: 0.56;
control: 0.48), with no significant differences among them (Tukey’s
test for comparison of means on the TREAT effect).
In the control treatment, where no organic matter was added to
the rhizobox upper compartments, RGU units for luminescence in
rhizosphere samples were slightly larger (though not statistically
significant) than in bulk soil samples. The rhizosphere samples of
the rhizoboxes treated with HA at both 14 and 140 mg kg?1rates
had RGU values significantly larger than those treated with HA
size-fractions. Furthermore, RGU values in the first three layers
were almost twice in the HA-10 treatment than in the HA-1 treat-
ment. No differences were observed between control soil and soils
added with the three humic size-fractions.
3.3. DNA fingerprinting
Bulk soil microcosms without the presence of plants and roots
were analysed for soil bacterial community structure at different
incubation times up to 60 d. This period was more than twice the
period of soil incubation in the rhiobox.
Replicates were separately incubated, corresponding DNA was
extracted, PCR amplified and DGGE analysed immediately at the
end of each experiment. A very high degree of reproducibility
among replicates at each incubation time was found for all treat-
ments and control. In fact, the cluster analysis of the electropho-
retic profiles showed Dice indices of similarity with values
between 0.9 and 1.0, indicating that any difference found among
profiles should not be attributed to the variability among repli-
cates. Fig. 2 is an example of the reproducibility of the replicates,
as it clearly indicates that the three bulk soil microcosm replicates
amended with 14 mg (HA-1) humic acids kg?1soil and sampled
after 10 d (a), 20 d (b), 30 d (c), 60 d (d) show almost identical pro-
files at each sampling (Dice coefficients > 0.95). Moreover, in order
to evaluate the changes of microbial community during incubation
time, regardless of the treatment, control samples at different incu-
bation times were compared. A similarity index P 0.85 indicated a
remarkable stability of the microbial community in soil micro-
cosms (data not shown).
Given the large number of samples, there was no way to load all
replicates onto one gel. Although it is possible to compare results
from different gels, this approach might bear some pitfalls strictly
related to the general consideration that gel electrophoresis, in
particular under denaturing conditions, depends on experimental
variables that make it not perfectly reproducible. Thus, once we
control HA-10 HA-1FR-I FR-IIFR-III
non-amended bulk soil 0-2 mm 2-4 mm4-6 mm
Fig. 1. Response of P. fluorescens pUCD 607 in samples from rhizoboxes subjected to different treatments. Data are expressed in terms of relative carbon units (RGU), which
are lmol of carbon equivalents per g of soil (means ± standard deviations, n = 3). Results of Tukey’s test for comparison of means for the TREAT ? SAMP effect are reported as
minor letters above each bar.
E. Puglisi et al./Chemosphere 77 (2009) 829–837
verified the reproducibility of our replicates, we decided to avoid
these potential drawbacks to better detect the effects of the treat-
ments under investigation by loading onto a single summarizing
gel only one sample per treatment. This approach was performed
for bulk soil incubations as well as for rhizoboxes.
The effects of the two rates of bulk HA (140 and 14 mg) on soil
bacteria diversity are reported in Fig. 3. The rather similar Dice
coefficients for controls and amended microcosms (0.81–0.89)
indicate that the unfractionated HA did not particularly affect the
structure of bacteria in soil microcosms, as the size-fractions in-
The effect of time on the structure of the bacterial communities
was evaluated for each amendment. Concerning this, Fig. 4 reports
the cluster analysis of genetic fingerprints relative to soil micro-
cosms amended with the three HA size-fractions. Dice coefficients
of similarity indicated that controls (tosamples) clustered sepa-
rately (0.54–0.65) in comparison with the profiles corresponding
to amended microcosms incubated up to 60 d, which tightly clus-
tered together (0.84–0.92; 0.79–0.88; 0.87–0.91, for FR-I, FR-II, FR-
III, respectively). Soil bacteria were therefore affected by the humic
size-fractions, since changes in community structure occurred rap-
idly and persisted during time. However, no significant difference
was observed among the size-fractions.
The effects of rhizodeposition of maize plants grown in rhizo-
boxes treated in their upper compartments with HA and its
size-fractions were evaluated by 16S PCR–DGGE in soil samples
collected from the lower compartment of rhizoboxes, which did
not receive the HA amendments. As for the microcosm studies,
Fig. 2. 16S rDNA-DGGE electrophoretic profiles. Three bulk soil microcosm replicates amended with 14 mg (HA-1) humic acids kg?1soil. Samples collected after 10 d (a), 20 d
(b), 30 d (c), 60 d (d). See M & M for more details.
Fig. 3. Cluster analysis of 16S rDNA-DGGE electrophoretic profiles. Bulk soil microcosms amended with 140 mg (HA-10) and 14 mg (HA-1) humic acids kg?1soil; samples
collected after 0–60 d. See M & M for more details.
E. Puglisi et al./Chemosphere 77 (2009) 829–837
cluster analysis of genetic fingerprints (not shown) indicate that
reproducibility of replicates was large. Fig. 5 reports the effects of
organic amendments on bacterial fingerprinting of rhizosphere soil
samples at distances from the root plane of 0–2 and 2–6 mm. With-
in the same treatment, the community structure did not change
significantly as a consequence of the radial distance from the root
plane, although the profiles for FR-II and FR-III were slightly over-
lapped. With the only exception of the HA-10 treatment, all treated
samples clustered separately (Dice coefficients 0.69–0.37) as com-
pared to the controls, thus suggesting a significant change of soil
bacterial community structure as a consequence of humic treat-
ments mediated by plant growth and rhizodeposition.
Among the three hypotheses tested in this work, the first two,
i.e., a stimulating activity of humic acids on maize plants rhizode-
position and a consequent change in rhizosphere microbial com-
munities, were confirmed. Concomitantly, interesting suggestions
and conclusions arose from the third hypothesis regarding the
existence of a structure–activity relationship between physico-
chemical properties of humic acids and their effects on the plant-
bacteria interactions occurring in the rhizosphere. In the rhizobox
systems employed here, humic acids are added only in the upper
compartment, while measures were carried out in the lower com-
partment where no humic acids were added. Previous evidences
(Wenzel et al., 2001; Cattani et al., 2006) have shown also that
pH changes in the rizosphere samples are neglectable. The mea-
sured effects are thus mediated by plant physiology and no direct
effects of humic acids are involved.
In rhizoboxes treated with bulk HA, the biosensor response to
rhizosphere soils was in fact larger than that to the control soil,
whereas the three layers at different distances from the roots plane
had similar RGU values. A dose effect was also observed, since the
soil treated with larger HA amount (HA-10) showed a greater bio-
sensor response. The similar RGU values found for the three soil
slices at different distances may indicate that carbon-rich root exu-
dates tend to move in the soil solution by passive diffusion. In par-
ticular, this may be likely for sugars compounds, while organic
acids are usually more sorbed and thus have lower diffusion coef-
ficients (Kuzyakov et al., 2003). Only the bulk humic materials (HA-
1 and HA-10) stimulated the deposition of easily available carbon
compounds from roots to rhizosphere, whereas no differences
from control were observed in soils treated with the three size-
Application of DNA fingerprinting methods enabled the detec-
tion of diverse members of soil bacterial consortia, even including
bacteria that are not cultivable. DNA fingerprinting was also ob-
tained for bulk soil microcosms to test both the reproducibility of
this molecular approach and the stability of the microbial commu-
nity in unamended soil samples. The comparison of genetic finger-
prints between soils treated with the two HA rates failed to show
variation in the structure of bacterial communities, while changes
were induced by amendments with the separated size-fractions.
DNA fingerprinting of treated and control soil from the lower rhi-
zobox compartment, where 25-d-grown maize roots could have
mediated the effect of organic amendments on soil bacteria, re-
vealed that HA and its three size-fractions had a different influence
on the structure of rhizospheric bacterial community. In particular,
the bulk HA, which had poorly responded in bulk soil microcosms,
had larger effects in the lower rhizobox compartment through rhi-
zodeposition processes, at least the lower rate. This result is in
Fig. 4. Cluster analysis of 16S rDNA-DGGE electrophoretic profiles. Bulk soil
microcosms amended with of 14 mg of each HA size-fraction kg?1soil; samples
collected after 0–60 d. (a) FR-I, (b) FR-II, (c) FR-III.
Fig. 5. Cluster analysis of 16S rDNA-DGGE electrophoretic profiles. Rhizospheric soil samples from lower compartments of rhixoboxes amended with HA and their size-
fractions. HA-1: 14 mg HA kg?1soil; HA-10: 140 mg HA kg?1soil; FR-I: 14 mg of size-fraction I kg?1soil; FR-II: 14 mg of size-fraction II kg?1soil; FR-III: 14 mg of size-fraction
III kg?1soil. 0–2: microtome sampling of rhizospheric soil at a distance of 0–2 mm from root plane; 2–6: microtome sampling of rhizospheric soil at a distance of 2–6 mm
from root plane. See M & M for more details.
E. Puglisi et al./Chemosphere 77 (2009) 829–837
accordance with the greater biosensor response for HA-1 and HA-
10 treatments and supports the hypothesis that the increased
rhizodeposition induced by bulk HA had also a larger effect on
the rhizosphere microbial communities.
Although difficult to be correlated by statistical means, rhizode-
position and community structure results can be interpreted in the
light of pyrolsyis-GC–MS and CPMAS-NMR characteristics of HA
and its three size-fractions. These showed distinct differences be-
tween the molecular composition of HA and its three fractions (Ta-
bles 1 and 2).13C-CPAS-NMR spectra revealed a larger content of
aromatic, carboxylic and ketonic carbon compounds in HA (Table
1). Consequently, GC/MS analysis of thermolytic products showed
that the amount of lignin compounds in HA was larger than in the
three fractions. Some of the identified compounds (p-hydroxy-
phenyl, guaiacyl, syringyl and cinnamic acid) have known effects
on plant physiology (Yu et al., 2003), thus suggesting a significant
structure activity relationship between the molecular characteris-
tics of HA and the effects on roots rhizodeposition processes. Sep-
arated size-fractions were progressively more hydrophilic than the
bulk HA. The larger hydrophobic character of the latter was condu-
cive of a stronger conformational stability, with its molecular com-
ponents remaining less accessible to metabolization by soil
microorganisms. Molecular fingerprinting results for the bulk
microcosms amended with HA and its size-fractions indicated that
soil microorganisms hardly utilized the bulk HA, most probably be-
cause of its structural complexity and stability.
These evidences suggest that bulk HA is less degraded by soil
microorganisms, and hence maintains a larger stimulation on
plants rhizodeposition. Conversely, it is very likely that solubiliza-
tion, extracellular degradation, and uptake of the three size-frac-
amendments available for heterotrophic bacteria. This increased
degradation might be explained by changes in pH and solubility
of organic molecules, increased microbial activities, and coopera-
tive degradation among different bacterial species, which are phe-
nomena typically induced by plant root rhizodeposition. Therefore,
the structural and chemical complexity of added humic materials
appears to be a key factor for their utilization by heterotrophic
Previous findings have shown that the same HA and size-frac-
tions as those studied here had a stimulatory effect on maize
plants, increasing the enzyme activities related to glycolysis and
tricarboxylic acid (TCA) cycle (Nardi et al., 2007). Plants grown
for 10 d in presence of HA in the upper compartment of the rhi-
zoboxes may have released more organic carbon in the rhizo-
sphere as a consequence of the induction of such enzymatic
activities. This release was still evident in the lower compartment
of the rhizoboxes 15 d later, when plants were still exposed to HA
in the upper compartment. An explanation of this process is that
roots should first exudate sufficient organic acids to disrupt the
strong humic associations, from which bioavailable humic com-
ponents are then released and become accessible to microbes
for transformation (Piccolo et al., 1999; Piccolo, 2002). Rather
than specific differences in molecular composition between HA
and its size-fractions, it is the difference in strength of molecular
conformation that may explain the results. In fact, a greater root
input is needed for disaggregating the bulk HA, thus implying a
larger presence of bioavailable carbon that stimulated the biosen-
This study confirms the hypothesis that HA amendments affect
the rhizodeposition of maize plants, and, concomitantly, the struc-
ture of rhizosphere microbial communities. In particular, we
showed that HA with a large conformational stability increased
rhizodeposition by plants, and consequently the microbial commu-
nity structure. The main hypothesis, in accordance with the micro-
bial loop model of Bonkowsky (2005), is that in the presence of
thus makingthese humic
hardly bioavailable organic materials, plants stimulate a larger
microbial activity. This may have important consequences on the
management of soil fertility and agricultural exploitation of humi-
fied organic matter.
Our results also show the importance of applying different
methods to assess the effects of organic matter on soil rhizosphere.
Technologically advanced methods such as biosensors and DNA
fingerprinting gave more valuable results than classical approaches
based on microbial counts, which usually fail to detect significant
changes in the rhizosphere. Results obtained here by biosensors
and DNA fingerprinting are complementary. The former method
provides a quantitative assessment of carbon species readily
available to the commonP. fluorescens soil bacterium, while finger-
printing methods well distinguished among different sample treat-
ments, as a result of induced changes in the structure of soil
This research was funded by the MIUR (COFIN 2003) project
‘‘Organic fertilization role in the maximization of plant nutrition
and in the attenuation of agricultural pollution”. We acknowledge
the support of the CERMANU Interdepartmental Research Centre
(Università di Napoli Federico II) for NMR spectra. Ken Killham
and Dominic Standing are kindly acknowledged for providing the
Arancon, N.Q., Edwards, C.A., Lee, S., Byrne, R., 2006. Effects of humic acids from
vermicomposts on plant growth. Eur. J. Soil. Biol. 42, 65–69.
Atalla, R.H., VanderHart, D.L., 1999. The role of solid state 13C NMR spectroscopy in
studies of the nature of native celluloses. Solid State Nucl. Mag. Res. 15, 1–19.
Ayuso, M., Hernández, T., Garcia, C., Pascual, J.A., 1996. Stimulation of barley growth
and nutrient absorption by humic substances originating from various organic
materials. Bioresource Technol. 57, 251–257.
Bonkowski, M., 2004. Protozoa and plant growth: the microbial loop in soil
revisited. New Phytol. 162, 617–631.
Canellas, L.P., Façanha, A.O., Olivares, F.L., Façanha, A.R., 2002. Humic acids isolated
from earthworm compost enhance root elongation, lateral root emergence, and
plasma membrane H+-ATPase activity in maize roots. Plant Physiol. 130, 1951–
Cattani, I., Fragoulis, G., Bocelli, R., Capri, E., 2006. Copper bioavailability in the
rhizosphere of maize (Zea mays L.) grown in two Italian soils. Chemosphere 64,
Conte, P., Spaccini, R., Piccolo, A., 2006. Advanced CPMAS-13C NMR techniques for
molecular characterization of size-separated fractions from a soil humic acid.
Anal. Bioanal. Chem. 386, 382–390.
Conte, P., Spaccini, R., Smejkalova, D., Nebbioso, A., Piccolo, A., 2007. Spectroscopic
and conformational properties of size-fractions separated from a lignite humic
acid. Chemosphere 69, 1032–1039.
Guignard, C., Lemée, L., Amblès, A., 2005. Lipid constituents of peat humic acids and
humin. Distinction from directly extractable bitumen components using TMAH
and TEAAc thermochemolysis. Org. Geochem. 36, 287–297.
Hamin-Hanjani, S., Meikle, A., Glover, L.A., Prosser, J.I., Killham, K., 1993. Plasmid
and chromosomally encoded luminescence marker systems for detection of
Pseudomonas fluorescens in soil. Mol. Ecol. 2, 47–54.
Hatcher, P.G., 1987. Chemical structure studies of natural lignin by dipolar
dephasing solid-state13C nuclear magnetic resonance. Org. Geochem. 11, 31–
Hatcher, P.G., Nanny, M.A., Minard, R.D., Dible, S.D., Carson, D.M., 1995. Comparison
of two thermochemolytic methods for the analysis of lignin in decomposing
gimnosperm wood:the CuOoxydation
thermochemolysis with tetramethylammonium hydroxyde (TMAH). Org.
Geochem. 23, 881–888.
Heuer, H., Smalla, K., 1997. Application of denaturing gradient gel electrophoresis
and temperature gradient gel electrophoresis for studying soil microbial
communities. In: van Elsas, J.D., Trevors, J.Y., Wellington, E.M.H. (Eds.),
Modern Soil Microbiol. Marcel Dekker, New York, pp. 353–373.
Jokinen, H.K., Kiikkilä, O., Fritze, H., 2006. Exploring the mechanisms behind
elevated microbial activity after wood ash application. Soil Biol. Biochem. 38,
Kuzyakov, Y., Raskatov, A., Kaupenjohann, M., 2003. Turnover and distribution of
root exudates of Zea mays. Plant Soil 254, 317–323.
Lilienfein, J., Qualls, R.G., Uselman, S.M., Bridgham, S.D., 2004. Adsorption of
dissolved organic carbon and nitrogen in soils of a weathering chronosequence.
Soil Sci. Soc. Am. J. 68, 292–305.
method and themethod of
E. Puglisi et al./Chemosphere 77 (2009) 829–837
Muyzer, G., Smalla, K., 1998. Application of denaturing gradient gel electrophoresis
(DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial
ecology. Anton. van Lee. 73, 127–141.
Muyzer, G., de Waal, E.C., Uitterlinden, A.G., 1993. Profiling of complex microbial
populations by denaturing gradient gel electrophoresis analysis of polymerase
chain reaction-amplified gene encoding for 16S rRNA. Appl. Environ. Microb. 59,
Nardi, S., Pizzeghello, D., Muscolo, A., Vianello, A., 2002. Physiological effects of
humic substances on higher plants. Soil Biol. Biochem. 34, 1527–1536.
Nardi, S., Muscolo, A., Vaccaro, S., Baiano, S., Spaccini, R., Piccolo, A., 2007.
Relationship between molecular charachteristics of soil humic fractions and
glycolytic pathway and krebs cycle in maize seedlings. Soil Biol. Biochem. 39,
Paterson, E., 2003. Importance of rhizodeposition in the coupling of plant and
microbial productivity. Eur. J. Soil Sci. 54, 741–750.
Piccolo, A., Conte, P., Cozzolino, A., 1999. Effects of mineral and monocarboxylic
acids on the molecular association of dissolved humic substances. Eur. J. Soil Sci.
Piccolo, A., 2002. The supramolecular structure of humic substances. A novel
understanding of humus chemistry and implications in soil. Sci. Adv. Agron. 75,
Piccolo, A., Conte, P., Trivellone, E., Van Lagen, B., Buurman, P., 2002. Reduced
heterogeneity of a lignite humic acid by preparative HPSEC following
interaction with an organic acid. Characterization of size-separates by PYR-
GC-MS and1H-NMR spectroscopy. Environ. Sci. Technol. 36, 76–84.
Puglisi, E., Fragoulis, G., Del Re, A.A.M., Spaccini, R., Piccolo, A., Gigliotti, G., Said-
Pullicino, D., Trevisan, M., 2008. Carbon deposition in soil rhizosphere following
amendments with compost and its soluble fractions, as evaluated by combined
soil–plant rhizobox and reporter gene systems. Chemosphere 73, 1292–1299.
Spaccini, R., Piccolo, A., 2007. Molecular characterization of compost at increasing
stages ofmaturity. 2.Thermochemolysis-GC–MS
spectroscopy. J. Agric. Food Chem. 55, 2303–2311.
Standing, D., Rangel Castro, J.I., Prosser, J.I., Meharg, A.A., Killham, K., 2005.
Rhizosphere carbon flow – a driver of soil biodiversity? In: Bardgett, R.,
Hopkins, D., Usher, M. (Eds.), Biological Diversity and Function in Soils.
Cambridge University Press, Cambridge, pp. 154–170.
Vane, C.H., Martin, S.C., Snape, C.E., Abbott, G.D., 2001. Degradation of lignin in
wheat straw during growth of the oyster mushroom (Pleurotus ostreatus) using
off-line thermochemolysis with tetramethylammonium hydroxyde and solid
state13C NMR. J. Agric. Food Chem. 49, 2709–2716.
Wenzel, W.W., Wieshammer, G., Fitz, W.J., Puschenreiter, M., 2001. Novel rhizobox
design to assess rhizosphere characteristics at high spatial resolution. Plant Soil
Yeomans, C.V., Porteous, F., Paterson, E., Meharg, A.A., Killham, K., 1999. Assessment
of lux-marked Pseudomonas fluorescens for reporting on organic carbon
compounds. FEMS Microbiol. Lett. 176, 79–83.
Youssef, R.A., Chino, M., 1989. Root induced changes in the rhizospher of plants. I.
Changes in relation to the bulk soil. Soil Sci. Plant Nutr. 35, 461–468.
Yu, J.Q., Ye, S.F., Zhang, M.F., Hu, W.H., 2003. Effects of root exudates and aqueous
root extracts of cucumber(Cucumis
photosynthesis and antioxidant enzymes in cucumber. Biochem. Syst. Ecol.
sativus) andallelochemicals on
E. Puglisi et al./Chemosphere 77 (2009) 829–837