Date post: | 18-Mar-2018 |
Category: |
Documents |
Upload: | nguyenhanh |
View: | 221 times |
Download: | 8 times |
Top-down controls and trophic
cascades in estuarine intertidal
sediments: the role of predatory
nematodes
Sara Albuixech Martí
Academic year 2013 – 2014
Promoter: Prof. Tom Moens
Supervisors: Xiuquin Wu; Tania
Campiñas Bezerra
Ghent University
Faculty of Science
Department of Biology
Thesis submitted to obtain the degree
of European Master of Science in
Nematology
1
The trophic web, which interconnects the different living and non-living
compartments in an ecosystem via the transfer of matter and energy, is a fundamental
concept in ecology. The trophic web is characterized by multiple and complex
interactions between different levels (Van der Meer et al., 2005). Consequently its study
is difficult, and reductionist lab experiments and simulations which simplify the natural
system are essential for a better understanding of the different processes and interactions
that govern food web structure and dynamics (O'Gorman et al., 2008).
O'Gorman et al. (2008) observed that small changes in ecosystem complexity,
such as a single species addition or deletion, may propagate throughout the whole food
web. This phenomenon is called a trophic cascade, whereby for instance a change in a
predator’s abundance not only affects the population of its prey, but also those of the
prey’s prey (Gamfeldt et al., 2005; O'Gorman et al., 2008). Trophic cascades have been
widely studied in many ecosystems (Posey et al., 1995; Shurin et al., 2002; Gruner et
al., 2008; Baum & Worm, 2009), including marine systems. It has been reported that
the loss of diversity at top trophic levels can alter the classic trophic pyramid of marine
food webs and change it into a compressed version, the consequences of which are
largely unknown (O'Gorman et al., 2008). It is well known that changes in marine
predator abundance may promote far-reaching effects for ecosystem structure,
functioning, and resilience (Baum & Worm, 2009). Such cascading effects of changing
top predator density or diversity on functioning of marine ecosystems are particularly
important for their conservation and economic value (O'Gorman et al., 2008; Baum &
Worm, 2009). Trophic cascades have been less well studied in benthic systems, and
have often been less focused on the lower trophic levels. Therefore, a better knowledge
of cascades to lower trophic levels is fundamental (Baum & Worm, 2009).
In our study we have focused on nematodes, because they are the most abundant
and diverse meiofaunal taxon in marine, freshwater and terrestrial ecosystems. In terms
of abundance, an estimated three quarters, or more, of all animals on earth are
nematodes. The number of species which have actually been properly described is
limited (ca. 30,000 including all parasitic taxa). A substantial part of that diversity and
abundance is present in marine habitats. As a taxon, free-living nematodes consume a
wide range of food sources and use a variety of feeding strategies (Yeates et al., 2009;
Moens et al., 2013). Many species presumably feed on bacteria (Moens & Vincx,
1997). They are able to influence bacterial activity, either stimulatory or inhibitory (De
2
Mesel et al., 2003). Nematode effects on bacteria may be direct, i.e. through grazing. A
moderate grazing pressure by benthic invertebrates, mainly nematodes and protists, may
be directly positive for bacteria in the sediment by keeping bacterial populations in a
prolonged state of logarithmic growth (Ingham et al., 1985; Traunspurger et al., 1997).
Furthermore nematodes may indirectly stimulate the bacterial activity in the sediment
by (1) enhanced recycling of nutrients, particularly of N (Ingham et al., 1985;
Kristensen, 1988), (2) converting particulate organic matter to dissolved organic matter
(Kristensen, 1988), (3) stimulating the upward transport of reduced and the downward
transport of oxidized compounds (Kristensen, 1988), (4) secreting mucus trails which
selectively stimulates growth of certain bacteria (Riemann & Schrage, 1978; Moens et
al., 2005), (5) transporting bacterial cells, either internally or externally on the cuticle,
to unexploited microsites (Ingham et al., 1985). Consequently, the nematode activity
may have an important influence on decomposition rates and facilitate nutrient
exchange, which plays an important role in the functioning of many ecosystems
(Traunspurger et al., 1997; De Mesel et al., 2004). Nevertheless inhibitory effects on
bacteria population have also been recorded in the presence of nematodes. De Mesel et
al. (2003) studied the influence of bacterivorous nematodes on the decomposition of
cordgrass and observed that the decomposition and nutrient mineralization slowed down
in the presence of nematodes. Bacterivorous nematodes appeared to inhibit the fungial
growth and in turn fungi are important in the early stages of decomposition of
cordgrass, hence this could be a reason for the initially slower decomposition process in
the presence of nematodes (De Mesel et al., 2003).
It is also well known that meiofauna plays an important role as a trophic link
between bacteria and larger fauna (Coull, 1999). Based on long-term variability in
meio- and macrofaunal abundance, it has been suggested that predation is the main
direct interaction that affects meiofaunal communities (Coull, 1986; Coull, 1999).
Indeed, many studies of predator-prey interactions have been performed in aquatic
communities using macrofauna as predators and meiofauna as prey (Olafsson, 2003;
O'Gorman et al., 2008). However, information about meiobenthic predators controlling
other meiofauna and/or their prey is comparatively scarce (Moens et al., 1999; Moens et
al., 2000; Hamels et al., 2001; Gallucci et al., 2005; dos Santos & Moens, 2011),
despite the suggestion that even at relatively low abundance, predatory nematodes can
substantially impact abundance, species composition and/or diversity of their prey
3
assemblages (Moens et al., 2000; Gallucci et al., 2005), i.e. the predation by nematodes
may be a driving force on community structure (Gallucci et al., 2005; Yeates et al.,
2009; Moens et al., 2013).
If predatory nematodes influence their prey assemblages, they may also affect
the abundance and activity of their prey’s prey, i.e. microphytobenthos and bacteria, and
hence ultimately influence ecosystem functioning (O'Gorman et al., 2008). Such trophic
cascades occur in many ecosystems, but have not hitherto been properly investigated for
predatory nematodes, exception made for one study by Mikola & Setala (1998) in soil
decomposer food webs. These authors performed a microcosm experiment containing
different species of bacteria and fungi as the first trophic level, a bacterivorous
nematode and a fungivorous nematode species as the second level, and a predatory
nematode species as the third level. Although the experiment did not provide evidence
that predation regulated the microbial biomass and productivity, the authors
demonstrated that microbivore biomass and mineralization of carbon and nitrogen were
influenced by the predators.
Hence, microcosm experiments are a powerful tool to evaluate top-down and
bottom-up control in benthic communities (Moens et al., 2000; Worm et al., 2002;
Gallucci et al., 2005; De Mesel et al., 2006; dos Santos & Moens, 2011). We thus used
laboratory microcosm experiments under controlled conditions to study top-down
effects in the meio- and microbenthic compartments of estuarine intertidal food webs.
Our aim was to determine the effects of bacterivorous nematodes on bacteria and how
these are influenced by predacious nematodes.
We propose four fundamental hypotheses: (1) we expect that bacterivorous
nematodes affect the microbial activity in fairly short-term microscosm experiments; (2)
predators cause a direct decrease in the abundance of their prey; (3) predation effects on
prey nematode assemblages cascade down to bacteria activity by alleviating the effects
of their grazers (bacterivorous nematodes); and (4) we hypothesize that top-down
effects are more pronounced at relatively low microbial activity.
4
MATERIAL AND METHODS
Sampling site:
Sediments, bacteria and predatory nematodes for the present experiments were
collected at the Paulina intertidal flat, Westerschelde Estuary, in SW Netherlands
(Figure 1). Sandy intertidal marine sediment (upper 1 to 2 cm) was sampled during low
tide at a medium-grained station with low silt fraction. The sampling was performed
during spring and early summer 2014.
Figure 1: Location of the sampling stations in the Westerschelde estuary, SW
Netherlands (Moens & Vincx, 1998).
Nematodes:
Bacterivorous nematodes used belonged to two species of Monhysteridae,
Diplolaimelloides meyli and D. oschei, and a species of Rhabditidae, Litoditis marina,
which were all harvested from permanent monospecific, agnotobiotic cultures on agar
(Moens & Vincx, 1998). These cultures were maintained under identical temperature
(20°C) and salinity (20‰) conditions. Diplolaimelloides species are opportunistic
colonizers of various types of decaying organic matter (Warwick, 1987; Moens &
5
Vincx 2000). Likewise, Litoditis marina is associated with decomposing macroalgae in
the littoral zone of coastal and estuarine environments (Moens & Vincx 2000; De
Meester et al., 2012). The three nematode species used in this experiment belong to the
same functional group, the deposit feeders, and are assumed to feed mainly or
exclusively on bacteria (Moens & Vincx, 1997). However these species showed
differential responses to food availability; L. marina usually is found in patches
characterized by very high food density where it feeds continuously, while the two
species of Monhysteridae, D. meyli and D. oschei, prefer lower food densities (dos
Santos et al. 2008) and feed more intermittently (Moens et al., 2004). Therefore the
rhabditid species is classified as an enrichment opportunist and both monhysterid
species as general opportunists (dos Santos et al. 2008).
Enoploides longispiculosus and Adonchoaimus fuscus were used as predators,
since the predatory feeding behaviour of both species has been previously well studied
(Moens & Vincx, 1997; Moens et al., 1999; Moens et al., 2000; Hamels et al., 2001;
Gallucci et al., 2005; Moens et al., 2005; dos Santos & Moens, 2011). E.
longispiculosus feeds on nematodes, ciliates and other meio- to microsized benthic
organisms and may be able to even control prey biomass and community structure
(Moens et al., 2000; Hamels et al., 2001; Gallucci et al., 2005; dos Santos & Moens,
2011). A tracer experiment failed to detect uptake of bacterial carbon by
E.longispiculosus (Moens et al., 1999), but recent evidence suggests this species is not a
strict predator and may supplement its diet with microalgae (Moens et al., 2014).
Nevertheless, predation is considered the main feeding strategy for E.longispiculosus.
By contrast, it has been suggested that A.fuscus, like other oncholaimids, is only a
facultative predator which presents additional feeding strategies besides predation
(Moens & Vincx, 1997); scavenging of dead metazoans may be one important feeding
strategy in this species (Moens & Vincx, 1997). On the other hand, in a tracer
experiment, it ingested measurable but very small amounts of bacteria (Moens et al.,
1999).
The predator–prey combinations used in our experiments are artificial, as these
particular predator species are not usually found in the same type of sediments as these
prey species. However, at the Paulina site, they occur at very nearby stations (separated
by only a few tens of meters). This artificial assemblage was chosen for several reasons:
(1) E. longispiculosus and A. fuscus are very common marine predatory nemtodes and
6
survive well in laboratory incubations on different substrata, including agar, for periods
up to several weeks (Moens et al., 1999); (2) D.meyli, D.oschei and L.marina are easy
to culture in the lab, and their life history and, in the case of Diplolaimelloides sp., their
mutual interactions have been well studied (Moens & Vincx, 1998; Moens & Vincx,
2000; De Mesel et al., 2006; dos Santos et al., 2008; dos Santos et al., 2009; dos Santos
& Moens, 2011), moreover (3) they are suitable prey to E. longispiculosus and A. fuscus
(Moens et al., 1999; Gallucci et al., 2005; dos Santos & Moens, 2011), and (4) they
have been shown to measurably affect decomposition rates (De Mesel et al., 2003; De
Mesel et al., 2006) and bacterial assemblage composition (De Mesel et al., 2004; De
Mesel et al., 2006). Therefore despite the fact that the tested assemblage is artificial, we
believe that it is suitable for our aims.
Nematodes were harvested from the agar plates. In the case of Diplolaimelloides
sp. the agar became fully liquid, thus we were able to pour the medium over a sieve;
whereas in the case of L.marina the agar remained solid, hence it was necessary to
harvest them by sucrose washing. L. marina cultures were washed with sucrose in a
final concentration of 40% to remove agar remains and most adhering microbiota,
rinsed 4 times in artificial seawater (ASW), and eventually resuspended in it (Moens &
Vincx, 1998). We counted five aliquots from the resulting suspensions and
consequently we calculated how much suspension we needed to inoculate 500 prey
nematodes to each replicate. In the case of Diplolaimelloides sp., we added 250
individuals of D. oschei and 250 individuals of D. meyli. In both cases the resulting
nematode suspension may be concentrated to the desired density by centrifugation
(1500rpm, 10 min) or diluted by addition of ASW. The ASW used in all our tests had a
salinity of 22 ‰, which is a proper salinity for the nematodes used (Moens & Vincx,
1998; Moens & Vincx 2000).
On the other hand, the predatory nematodes E. longispiculosus and A. fuscus,
were collected from the sandy sediment from the Paulina Polder tidal flat (see Gallucci
et al., 2005, for location of stations with high abundances of these species). Nematodes
were extracted alive from fresh sediment by repeated decantation with tap water over a
125 μm mesh sieve. We collected the retained nematodes on the sieve with ASW, and
handpicked E. longispiculosus and A. fuscus under a compound microscope, specifically
50 individuals per replicate and genus. The collected predator nematodes were rinsed in
ASW to remove adhering organisms, and kept in ASW during one or two days to allow
7
voiding of gut contents. Then this ASW was distributed between all microcosms to be
sure that any bacteria introduced in a predator treatment because of its association with
the predator would also be present in non-predator treatments. Specifically 50µl of
ASW in which E. longispiculosus had been kept, were added to each microcosm
without E. longispiculosus; and 50µl of ASW in which A. fuscus had been kept, were
added to each microcosm without A. fuscus.
Preparation of a bacterial inoculum:
Bacteria were extracted from natural sediments. Sandy intertidal marine
sediment (100ml) from Paulina was rehydrated, mixed and shaken with autoclaved and
filtered (0.2 μm) artificial seawater (salinity of 22). To obtain an uniform bacteria
community in all the treatments from the same experiment, a few drops of agar from the
cultures of corresponding prey nematodes (L. marina or D. meyli and D. oschei) were
added to the resulting slurry, such that when we added these nematodes, the transfer of
bacteria specific for these cultures would not make differences between treatments
without these nematodes and associated bacteria. Then this slurry of sediment and agar
drops was filtered 5 times using 1.2µm Whatman GF/A filters to separate bacteria from
benthic microalgae and heterotrophic protists.
Extracellular Enzymatic Activities:
Two types of exo-enzymatic activity were evaluated as proxies of microbial
activity in microcosms. The exo-enzymatic activity was measured via fluorescence
using the spectrophotometer ‘VICTOR™ Multilabel Plate Reader’. In particular,
activities of ß-D-glucosidase and aminopeptidase were quantified fluorometrically using
fluorogenic analogues, respectively, L-Leucine-4-methylcoumarinyl-7-amide (Leu-
MCA) and 4-methylumbelliferone ß-D-glucopyranoside (MUF-Glu) as substrates
whose hydrolysis provides an estimate of the potential degradation rates of the more
readily available fractions of organic matter in the sediment (Hoppe, 1993; Danovaro et
al., 2001),. The enzymatic analysis was performed on sediment slurries following the
protocol of Danovaro et al., (2001). A sample (0.3g) of sediment was collected
randomly on the sediment column of each microcosm and mixed with 0.2 μm-pore-
8
filtered autoclaved ASW. The enzymatic reactions were started by adding 15μM of the
corresponding enzymatic analogue (Glu-MUF or Leu-MCA) (1mM) in each sediment
sample. A first measurement (blanks) was taken immediately after reading its
fluorescence at 355nm excitation and 460nm emission wavelength. Then incubations
were performed in the dark and at in situ temperature for 1 hour in the experiment with
L. marina and 2 hours in the experiments with Diplolaimelloides sp. The incubation
time in the dark was extended from 1 to 2 hours to allow a longer time for the
enzymatic action on the organic substrates added in the assay. After that the
fluorescence was measured again. Standards (0.1-10µM) to compare with our results
were freshly prepared diluting the stock solution (5mM) in pre-filtered and autoclaved
ASW. Finally the fluorescence of each sample was converted into nmol of MUF or
MCA released per g of sediment dry weight according to the formula:
nmol (MUF or MCA)/g*h=(((Fluoc-Fluoblk)*C+K)*v) / Ps*Tinc
where: MUF or MCA: hydrolyzed fluorogenic analogues of our target enzymes
Fluoc: Fluorescence of the sediment sample
Fluoblk: Fluorescence of the sediment blank
C: slope of the linear regression obtained by calibration curve
K: constant of the linear regression obtained by calibration curve
v: volume of the slurry in ml
Ps: sediment dry weight in grams
Tinc: incubation time in hours
Bacteria pre-tests:
The sediment used for all pre-tests and final experiments was washed with fresh
water and decanted to remove the fauna. The sediment was then dried at 180ºC for 20
hours. Afterwards it was washed again with fresh water and sieved between a 1000µm
sieve and a 32µm sieve to remove the bigger particles and the salts as much as possible.
Ultimately, it was dried again at 180ºC for 20 hours. Because these sediments are poor
in organic matter, even more so after the repeated decantation (which removes the small
silt-clay fraction to which much organic matter adsorbs), organic matter had to be added
as a substratum for bacterial growth. Consequently we tested three different substrates
to find one which would support proper microbial growth as assessed from levels of
enzymatic activity. The tested substrates were domestic sugar, rice and potato dextrose.
These pretests were made in 12-wells and 6-wells plates with dry sediment (5 grams)
and a bacteria suspension, which had been pre-incubated during 48h with the respective
9
substrates before addition to the sediment. The enzymatic activity of these microcosms
was measured after 2, 4, 6 and 8 days. Once we had decided that the rice extract was the
best substrate, additional experiments were performed comparing various rice extract
concentrations and with addition of prey and predator nematodes to the microcosms. In
addition the enzymatic activity of sediments to which no bacteria had been added was
measured as a control. This is largely for two reasons: (1) to test for background
fluorescence which can be caused, for instance, by chlorophyll pigments present in the
sediments, and (2) to test for bacterial contaminations.
Experimental designs:
Microcosms for all experiments were prepared in 6-well plates, using dry
sediment as a substrate and bacteria and nematodes (prey and predator) as living
organisms. Considering the results of the pre-tests we performed three sets of
experiments using different bacterivorous nematodes and/or different rice extract
concentrations to assess trophic cascades from predatory nematodes over bacterivores to
bacterial activity. One experiment was made with L. marina and the other two with a
combination of D. oschei and D. meyli using two different concentrations of rice extract.
Two rice extract concentrations were evaluated to compare the effect of the substrate
concentration on bacterial enzymatic activity. These rice extracts were prepared with
rice previously washed with boiled ASW and bacteria suspension in different
concentrations depending on the experiment: 1grain/ml in the experiment with L.marina
(LM) as well as in one of the experiments with Diplolaimelloides sp. (D1); and
5grains/ml in the other experiment with Diplolaimelloides sp. (D2). The bacteria
suspension was incubated with the rice during 48h at 15°C. Likewise we incubated
ASW with the corresponding concentrations of rice for the controls without bacteria of
the different experiments.
The bacteria in the suspensions tended to aggregate around rice, thus after the
incubation the resulting suspensions were shaken to release the bacteria from the rice
properly. The corresponding suspension was added to the sediment distributed in the 6-
wells plates until the sediment moistened completely and the microcosms were pre-
incubated. The pre-incubation of the bacteria in the sediment allowed build-up of the
bacterial population. Particularly in the experiment with L.marina, 3.5ml of rice extract
10
suspension was added to each replicate with 11g of dry sediment. The microcosms were
pre-incubated for 48h at 15°C. In contrast, in the experiments with Diplolaimelloides sp.
3ml of the corresponding suspension was added to each replicate with 10g of dry
sediment. The extra gram in the experiment with L.marina was kept in the freezer for
further analyses of the microbial community in posterior studies. Because enzymatic
activities in the L. marina experiments tended to be rather low, the pre-incubation time
was increased to 4 days and the incubation temperature was changed to 18ºC to raise the
bacteria activity in the experiments with Diplolaimelloides sp.
After pre-incubation, we did the first measurement (T0) of enzymatic activity of
bacteria for each sediment microcosm (a well in a 6-well plate). Subsequently, we
inoculated the corresponding species and number of nematodes to microcosms. Four
replicates were made for each treatment and control. In both set-ups the controls and
treatments were (Figure 2): rice extract without bacteria (Ar); bacteria suspension (B);
bacteria suspension + bacterial-feeding nematodes (BP); bacteria suspension + E.
longispiculosus (BE); bacteria suspension + A. fuscus (BA); bacteria suspension +
bacterial-feeding nematodes + E. longispiculosus (BPE); bacteria suspension +
bacterial-feeding nematodes + A. fuscus (BPA). Besides we conducted two extra
controls following the protocol for the experiments with Diplolaimelloides species. The
first control was made only with sediment and autoclaved ASW (A) and the second one
was made with fresh sediment (FS) with its natural fauna from Paulina Polder.
Figure 2: Experimental design.
All microcosms with nematodes were incubated for 8 days at 15°C in the case of
the experiment with L. marina and at 18ºC in the case of experiments with
Diplolaimelloides sp. After 2 (T1), 4 (T2), 6 (T3) and 8 (T4) days the enzymatic activity
11
of bacteria was measured. At the end of the experiment all the replicates were decanted
and the nematodes were collected on a sieve with a mesh size of 32µm and counted.
Data Analysis
The data of our experiments did not fit the assumptions for parametric analyses.
Therefore our data was assessed by two-way crossed and nested PERMANOVA
(Permutational Multivariate Analysis of Variance), a non-parametric, permutation-based
statistic for the analysis of multivariate data in response to treatments in an experimental
design. The PRIMER (Plymouth Routines In Multivariate Ecological Research) &
PERMANOVA version 6.1.6, PRIMER-E Ltd. (2006) software was used to evaluate the
variability between treatments of each experiment and the factors promoting this
variability, i.e the main components of variation. Analyses were based on a resemblance
matrix of Euclidean distances and permutations of residuals under a reduced model
(9999 permutations). Pair-wise comparisons were performed (9999 permutations) for
the factors of interest, i.e. treatments and time.
RESULTS:
Experiment with Litoditis marina (LM):
There is statistically significant interaction in the effects of Treatments and Time
on variability in the activity of ß-D-glucosidase (PERMANOVA TrxTi p<0.05; Table
1). This interaction means that the effects of a given factor are different when it is
considered separately within each level of the other factor. The presence of a significant
interaction generally indicates that the effects of each factor alone may not be
meaningful. Therefore our main interest is to examine the interaction effects doing pair-
wise comparisons among levels of the factor Treatment within each level of the factor
Time and vice versa.
12
Table 1: Results from the two –way crossed PERMANOVA analysis of activity ß-D-
glucosidase in response to different treatments and sampling times in the experiment
LM.
Source df SS MS Pseudo-F
P(perm)
Tr 6 5,0241E18 8,3734E17 6,0517 0,0002
Ti 4 2,7624E19 6,9059E18 49,911 0,0001
TrxTi 24 8,1517E18 3,3965E17 2,4548 0,0007
Res 105 1,4528E19 1,3837E17
Total 139 5,5328E19
The pair-wise comparisons testing the ß-D-glucosidase found significant
differences between treatments within each level of the factor Time. At T0, the control
with ASW showed an enzymatic activity significantly lower than the other treatments
(PERMANOVA p<0.05; Figure 3). Unlike what we expected, there were significant
differences between some treatments with predators and without them (PERMANOVA
p<0.05, B, BP > BA, BPE; Figure 3), the former exhibiting lower enzymatic activities
than the latter. At T1, the enzymatic activity in the treatment with E. longispiculosus
was significantly higher than in the treatment with A. fuscus (PERMANOVA p<0.05,
Figure 3). At T2, control ASW only presented significant differences with the treatment
with E. longispiculosus (PERMANOVA p<0.05, Ar < BE; Figure 3), but there were
other significant differences between treatments (PERMANOVA p<0.05, BP, BPA <
BE and B > BPA; Figure 3). At T3, the control ASW showed an enzymatic activity
significantly higher than treatments with nematodes (PERMANOVA p<0.05, Ar > BP,
BA, BPE; Figure 3), and the treatment with only bacteria displayed significant
differences with some treatments with nematodes (PERMANOVA p<0.05, B > BP, BA,
BPE, BPA; Figure 3). Ultimately in the last sampling time (T4) the only significant
differences were between the treatment with bacteria and both treatments with E.
longispiculosus (PERMANOVA p<0.05, B > BE, BPE; Figure 3).
13
Likewise the pair-wise comparisons between sampling times within each level
of the factor Treatment displayed significant differences. Thus the activity of ß-D-
glucosidase varied through time in all the treatments. Particularly, the ß-D-glucosidase
activity displayed a peak of activity at T2, which was reflected in the registered
significant differences in all the treatments (PERMANOVA p<0.05, T0 < T1, T2, T3 in
Ar; T0 < T1, T2, T3, T4 and T2 > T4 in B; T0 < T1, T2, T4 and T1, T2 > T3 in BP; T0,
T4 < T1, T2 and T2 > T1, T3 in BE; T0 < T2, T3 and T2 > T1, T3 in BA; T0 < T1, T2,
T4 and T2 > T3 in BPE; T0 < T2 in BPA; Figure 4). Generally, in all treatments,
G-enzyme L.marina
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
0
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
9E8
T0
T0: KW-H(6;28) = 18,6133; p = 0,0049
G-enzyme L.marina
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
2,8E9
T1
T1: KW-H(6;28) = 9,5616; p = 0,1444
G-enzyme L.marina
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
2,8E9
3E9
T2
T2: KW-H(6;28) = 13,7143; p = 0,0330
G-enzyme L.marina
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
-2E8
0
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
T3
T3: KW-H(6;28) = 15,7833; p = 0,0150
G-enzyme L.marina
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
0
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
T4
T4: KW-H(6;28) = 12,7167; p = 0,0478
Figure 3: The mean±SD of activity of
ß-D-glucosidase (nmol/g*h) of 4
replicates by treatment and sampling
time, in the experiment LM. The
capital letters display the significant
differences between treatments.
A B B
A B B C C
A B
A A C
A A B
A B
A A B B B
A B B
14
enzymatic activity increased until T2 and thereafter either remained more or less stable
(Ar, B, BP and BPA) or decreased (BE, BA and BPE).
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
0
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
lm-A
SW
r
lm-ASWr: KW-H(4;20) = 10,1; p = 0,0388
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
2,8E9
lm-B
ac
t
lm-Bact: KW-H(4;20) = 13,0429; p = 0,0111
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
-2E8
0
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
lm-B
+L
M
lm-B+LM: KW-H(4;20) = 13,5; p = 0,0091
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
2,8E9
3E9lm
-B+
En
op
loid
es
lm-B+Enoploides: KW-H(4;20) = 14,9286; p = 0,0049
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
2,8E9
lm-B
+A
do
nc
ho
laim
us
lm-B+Adoncholaimus: KW-H(4;20) = 12,2429; p = 0,0156
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
lm-B
+L
M+
En
op
lm-B+LM+Enop: KW-H(4;20) = 12,8714; p = 0,0119
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
lm-B
+L
M+
Ad
on
lm-B+LM+Adon: KW-H(4;20) = 10,3571; p = 0,0348
Figure 4: The mean±SD of activity of
ß-D-glucosidase (nmol/g*h) of 4
replicates by treatment and sampling
time, in the experiment LM. The
capital letters display the significant
differences between sampling times.
A B B B A B B B
A B C
A B B B
B B C
A B C A/B A
A A/C B C B A
A B B B
A B
15
Similarly for the activity of aminopeptidase, the interaction of the factors
Treatments and Time was also statistically significant (PERMANOVA TrxTi p<0.05;
Table 2).
Table 2: Results from the two –way crossed PERMANOVA analysis of activity
aminopeptidase in response to different treatments and sampling times in the
experiment LM.
Source df SS MS Pseudo-F P(perm)
Tr 6 7,9039E18 1,3173E18 13,451 0,0001
Ti 4 1,0186E19 2,5466E18 26,003 0,0001
TrxTi 24 6,9376E18 2,8907E17 2,9516 0,0002
Res 105 1,0283E19 9,7934E16
Total 139 3,5311E19
The pair-wise comparisons showed significant differences between treatments
within each level of the factor Time. In T0 the only significant differences were
displayed between control ASW and some treatments (PERMANOVA p<0.05, Ar < B,
BP, BPE, BPA; Figure 5). In T1 the treatment with prey presented an enzymatic activity
significantly higher than all the treatments with predator and the control ASW
(PERMANOVA p<0.05, BP > Ar, BE, BA; Figure 5). In T2 the control ASW still
exhibited significantly lower enzymatic activity than some treatments; the treatment
with only bacteria had a significantly higher enzymatic activity than those with only
prey nematodes and with prey nematodes + E. longispiculosus (PERMANOVA p<0.05,
Ar < B, BA, BPE, BPA and B > BP, BPE; Figure 5). In T3 the control ASW and the
treatment with A. fucus presented the lowest enzymatic activity with significant
differences with other treatments (PERMANOVA p<0.05, Ar < B, BP, BPE and BA <
B, BP, BPE, BPA; Figure 5). Finally in T4 there were many significant differences
between treatments (PERMANOVA p<0.05, Ar, BP, BE, BPE < B, BA and BPA > BP,
BE, BA, BPE; Figure 5).
16
The pair-wise comparisons between sampling times within each level of the
factor Treatment also were made. Unlike ß-D-glucosidase activity, there were no
significant differences in the aminopeptidase activity in the control with ASW through
time. Nevertheless, all treatments with nematodes exhibited a decreasing tendency of
the enzymatic activity at T3 and T4, which was reflected in the significant differences in
all the treatments (PERMANOVA p<0.05, T4 < T0, T1, T2, T3 in BP; T4 < T0, T1, T2
in BE; T2 > T1, T3, T4 in BA; T0 > T1, T3, T4 and T4 < T2, T3 in BPE; T2 > T1, T3,
T4 in BPA; Figure 6). Conversely, in the treatment with only bacteria, T1 showed the
lowest enzymatic activity differing significantly from other sampling times
(PERMANOVA p<0.05, T1< T0, T2 in B; Figure 6).
Box Plot (STATISTICA-L.marina 6v*28c)
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
2,8E9
T0
T0: KW-H(6;28) = 12,4729; p = 0,0522
Box Plot (STATISTICA-L.marina 6v*28c)
Mean
Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
T1
T1: KW-H(6;28) = 14,4754; p = 0,0248
Box Plot (STATISTICA-L.marina 6v*28c)
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
T2
T2: KW-H(6;28) = 16,33; p = 0,0121
Box Plot (STATISTICA-L.marina 6v*28c)
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
T3
T3: KW-H(6;28) = 18,1404; p = 0,0059
Box Plot (STATISTICA-L.marina 6v*28c)
Mean Mean±SD
AS
Wr
Bact
B+
LM
B+
Enoplo
ides
B+
Adonchola
imus
B+
LM
+E
nop
B+
LM
+A
don
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
T4
T4: KW-H(6;28) = 23,2611; p = 0,0007
Figure 5: The mean±SD of activity of
aminopeptidase (nmol/g*h) of 4 replicates
by treatment and sampling time, in the
experiment LM. The capital letters display
the significant differences between
treatments.
A B B B B A B A A
A B B/C C B/C
B D
A B A A C A
A A C A D
A B B A B
A C
17
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
4E8
5E8
6E8
7E8
8E8
9E8
1E9
1,1E9
lm-A
SW
r
lm-ASWr: KW-H(4;20) = 1,0143; p = 0,9076
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
lm-B
act
lm-Bact: KW-H(4;20) = 9,1857; p = 0,0566
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
lm-B
+LM
lm-B+LM: KW-H(4;20) = 11,5; p = 0,0215
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
lm-B
+E
no
plo
ide
s
lm-B+Enoploides: KW-H(4;20) = 13,0286; p = 0,0111
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
2,6E9
2,8E9
lm-B
+A
donchola
imus
lm-B+Adoncholaimus: KW-H(4;20) = 9,0286; p = 0,0604
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
2,4E9
lm-B
+L
M+
En
op
lm-B+LM+Enop: KW-H(4;20) = 15,5714; p = 0,0037
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1,8E9
2E9
2,2E9
lm-B
+L
M+
Ad
on
lm-B+LM+Adon: KW-H(4;20) = 11,5571; p = 0,0210
Figure 6: The mean±SD of activity of
aminopeptidase (nmol/g*h) of 4
replicates by treatment and sampling
time, in the experiment LM. The
capital letters display the significant
differences between sampling times.
A B A
A A A A B A A A B
A B A A
A B
A A/B B C
A B A A
18
Experiment with Diplolaimelloides sp. (1grain/ml) (D1):
The interaction between the effects of Treatments and Time on variability in the
activity of ß-D-glucosidase was statistically significant (PERMANOVA TrxTi p<0.05;
Table 3).
Table 3: Results from the two –way crossed PERMANOVA analysis of activity ß-D-
glucosidase in response to different treatments and sampling times in the experiment
D1.
Source df SS MS Pseudo-F P(perm)
Tr 8 2,8895E18 3,6118E17 125,67 0,0001
Ti 4 8,2537E17 2,0634E17 71,792 0,0001
TrxTi 32 1,8416E18 5,755E16 20,023 0,0001
Res 135 3,8801E17 2,8742E15
Total 179 5,9444E18
The pair-wise comparisons made between treatments within each level of the
factor Time displayed significant differences in every sampling time. Particularly, in T0
the controls with ASW and fresh sediment presented, respectively, significant
differences with almost all treatments (PERMANOVA p<0.05, FS > A, Ar, B, BP, BE,
BA, BPE, BPA; A < Ar, B, BP, BE, BA, BPE, BPA and Ar > BE, BPA; Figure 7). In
the T1 also the controls differed significantly from the rest of the treatments, besides
there were significant differences between these treatments (PERMANOVA p<0.05, FS
> A, Ar, B, BP, BE, BA, BPE, BPA; A < Ar, B, BP, BE, BA, BPE; Ar < B, BP, BE,
BA, BPE; BP < BE and BPA < B, BP, BE, BA; Figure 7). In T2 the significant
differences were between controls and treatments as well as between treatments with
predator and treatment with prey (PERMANOVA p<0.05, FS > A, BP, BPE, BPA; FS
< B, BE, BA; A < Ar, B, BP, BE, BA, BPA; Ar > BP, BPE, BPA; Ar < BA and B, BE,
BA > BP, BPE, BPA; Figure 7). Likewise in the T3 the controls differed significantly
from the rest of the treatments as well as these treatments displayed significant
differences (PERMANOVA p<0.05, FS >A, BP, BPE, BPA; FS < B, BA; A < B, BP,
BE, BA and B, BE, BA > BP; Figure 7). In the T4 the enzymatic activity in the control
FS was significantly higher than the rest of treatments and only some treatments
presented significant differences between them (PERMANOVA p<0.05, FS > A, Ar, B,
BP, BE, BA, BPE, BPA; BA > BP, BPE, BPA and Ar > BPE; Figure 7).
19
G-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E8
0
1E8
2E8
3E8
4E8
5E8
6E8
T0
T0: KW-H(8;36) = 28,1712; p = 0,0004
G-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E8
0
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
9E8
T1
T1: KW-H(8;36) = 31,1937; p = 0,0001
G-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E8
0
1E8
2E8
3E8
4E8
5E8
6E8
7E8
T2
T2: KW-H(8;36) = 31,4279; p = 0,0001
G-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E8
0
1E8
2E8
3E8
4E8
5E8
T3
T3: KW-H(8;36) = 29,2613; p = 0,0003
G-enzyme Diplo 1
Mean
Mean±SD
Fre
sh
se
d
AS
W
AS
W r
Ba
ct
B+
Dip
lo
B+
En
op
loid
es
B+
Ad
on
ch
ola
imu
s
B+
Dip
lo+
En
op
B+
Dip
lo+
Ad
on
Var1
-1E8
0
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
T4
T4: KW-H(8;36) = 23,1532; p = 0,0032
Figure 7: The mean±SD of activity
of ß-D-glucosidase (nmol/g*h) of 4
replicates by treatment and
sampling time, in the experiment
D1. The capital letters display the
significant differences between
treatments.
C D D
A B C C C C C C C
A D D D D F
D E
A B C D D D D D
B C B B
C B
A B B B B B B B B
A B C D A/C C B B
A C
A B A/C C D C C B/D D
20
The obtained results for the ß-D-glucosidase activity showed the same tendency
through time in all treatments as well as in the control Ar. The enzymatic activity
started low, then increased and finally decreased. The significant differences found
between treatments supported this hypothesis (PERMANOVA p<0.05, T2 > T0, T1, T3,
T4 and T1 < T4 in Ar; T2 > T3 > T1 > T0, T4 in B; T0, T4 < T1, T2, T3 in BP; T0 <
T1, T2, T3, T4; T4 < T1, T2, T3 and T1 < T2 in BE; T2 > T3 > T1 > T4 > T0 in BA;
T0, T4 < T1 in BPE; T2 > T0, T3, T4 in BPA; Figure 8). However, the control A
displayed an increasing ß-D-glucosidase activity through time (PERMANOVA p<0.05,
T0 < T3, T4; T1 < T0, T2, T3, T4 and T2 < T4; Figure 8). On the other hand, the
control with fresh sediment showed no a clear tendency and there were significant
differences between sampling times (PERMANOVA p<0.05, T3 < T0, T2 < T1, T4;
Figure 8).
21
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
9E8
Fre
sh s
ed
Fresh sed: KW-H(4;20) = 16,5286; p = 0,0024
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
-6E7
-4E7
-2E7
0
2E7
4E7
6E7
8E7
1E8
1,2E8
1,4E8
1,6E8
1,8E8
AS
W
ASW: KW-H(4;20) = 15,2; p = 0,0043
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
5E7
1E8
1,5E8
2E8
2,5E8
3E8
3,5E8
4E8
4,5E8
5E8
1-A
SW
r
1-ASW r: KW-H(4;20) = 12,7286; p = 0,0127
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
0
1E8
2E8
3E8
4E8
5E8
6E8
1-B
ac
t
1-Bact: KW-H(4;20) = 17,8857; p = 0,0013
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
4E7
6E7
8E7
1E8
1,2E8
1,4E8
1,6E8
1,8E8
2E8
2,2E8
2,4E8
1-B
+D
iplo
1-B+Diplo: KW-H(4;20) = 14,2571; p = 0,0065
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
0
1E8
2E8
3E8
4E8
5E8
6E8
7E8
1-B
+E
no
plo
ide
s
1-B+Enoploides: KW-H(4;20) = 17,0857; p = 0,0019
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
0
1E8
2E8
3E8
4E8
5E8
6E8
1-B
+A
donchola
imus
1-B+Adoncholaimus: KW-H(4;20) = 18,2857; p = 0,0011
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
-2E7
0
2E7
4E7
6E7
8E7
1E8
1,2E8
1,4E8
1,6E8
1,8E8
2E8
2,2E8
2,4E8
2,6E8
1-B
+D
iplo
+E
no
p
1-B+Diplo+Enop: KW-H(4;20) = 8,7429; p = 0,0679
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
-4E7
-2E7
0
2E7
4E7
6E7
8E7
1E8
1,2E8
1,4E8
1,6E8
1,8E8
2E8
2,2E8
1-B
+D
iplo
+A
don
1-B+Diplo+Adon: KW-H(4;20) = 10,3571; p = 0,0348
Figure 8: The mean±SD of
activity of ß-D-glucosidase
(nmol/g*h) of 4 replicates by
treatment and sampling time, in
the experiment D1. The capital
letters display the significant
differences between sampling
times.
A B A C B
A B C C
B A C
C D
A A B A A
A B C D A
A B B B A
A B C B/C D
A B C D E
A B A
A B A A
22
We also found statistically significant interaction in the effects of Treatments
and Time on variability in the activity of aminopeptidase (PERMANOVA TrxTi
p<0.05; Table 4).
Table 4: Results from the two –way crossed PERMANOVA analysis of activity
aminopeptidase in response to different treatments and sampling times in the
experiment D1.
Source df SS MS Pseudo-F P(perm)
Tr 8 2,6167E20 3,2709E19 111
0,0001
Ti 4 1,4925E20 3,7312E19 126,62
0,0001
TrxTi 32 1,132E20 3,5376E18
12,005 0,0001
Res 135 3,9782E19 2,9468E17
Total 79 5,639E20
As in the measurements of ß-D-glucosidase activity, the pair-wise comparisons
between treatments within each level of the factor Time concerning to aminopeptidase
activity displayed significant differences between the controls and the treatments with
bacteria in all the sampling times (PERMANOVA p<0.05, FS > B, BP, BE, BA, BPE,
BPA > Ar > A in T0; FS, B, BP, BE, BA, BPE, BPA > Ar > A and FS > BP, BE, BPE,
BPA in T1; B, BP, BE, BA, BPE, BPA > FS > Ar > A in T2; FS, B, BP, BE, BA, BPE,
BPA > Ar > A and FS < B. BP, BA, BPE in T3; FS > B, BP, BE, BA, BPE, BPA > Ar
and A < FS, Ar, BP, BPE in T4; Figure 9). Moreover in T1, T3, T4 there were
significant differences between treatments (PERMANOVA p<0.05, BPE > BPA in T1;
B >BE in T3; BE < B, BP, BA, BPE in T4; Figure 9).
23
Besides the aminopeptidase activity in this experiment exhibited the same
tendency through time as for ß-D-glucosidase, with an initial increase followed by a
decrease in activity after T2. All significant differences supported this tendency
(PERMANOVA p<0.05, T2, T3 > T0, T1 and T1 < T4 in Ar; T2, T3 > T1, T4 > T0 in
B, BP, BE, BA, BPE, BPA; Figure 10). However the control A presented no significant
differences, i.e. the activity did not change through time. On the other hand, the control
FS fluctuated a lot, without clear tendency, showing significant differences between all
sampling times (PERMANOVA p<0.05, T0, T1, T2, T3 < T4 and T2 < T1, T3; Figure
10).
P-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-5E8
0
5E8
1E9
1,5E9
2E9
2,5E9
3E9
3,5E9
T0
T0: KW-H(8;36) = 28,5225; p = 0,0004
P-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-5E8
0
5E8
1E9
1,5E9
2E9
2,5E9
3E9
3,5E9
4E9
T1
T1: KW-H(8;36) = 27,3288; p = 0,0006 P-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E9
0
1E9
2E9
3E9
4E9
5E9
6E9
7E9
8E9
T2
T2: KW-H(8;36) = 27,6216; p = 0,0006
P-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E9
0
1E9
2E9
3E9
4E9
5E9
6E9
T3
T3: KW-H(8;36) = 27,0721; p = 0,0007 P-enzyme Diplo 1
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-2E9
-1E9
0
1E9
2E9
3E9
4E9
5E9
6E9
7E9
T4
T4: KW-H(8;36) = 27,982; p = 0,0005
Figure 9: The mean±SD of activity of
aminopeptidase (nmol/g*h) of 4
replicates by treatment and sampling
time, in the experiment D1. The capital
letters display the significant differences
between treatments.
A B C D D D D D D
A B C D D D D D D
D E E E E
E A
A B C A A A A A A
D A
A B C D D D/A D D D/A
A B C C D C C C/D
B C C
24
T0 T1 T2 T3 T4
Var1
1E9
2E9
3E9
4E9
5E9
6E9
7E9
Fre
sh s
ed
Fresh sed: KW-H(4;20) = 16,1857; p = 0,0028
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
-1,5E9
-1E9
-5E8
0
5E8
1E9
1,5E9
2E9
2,5E9
AS
W
ASW: KW-H(4;20) = 1,0286; p = 0,9054
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
2E8
4E8
6E8
8E8
1E9
1,2E9
1,4E9
1,6E9
1-A
SW
r
1-ASW r: KW-H(4;20) = 15,8714; p = 0,0032
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
1E9
2E9
3E9
4E9
5E9
6E9
1-B
act
1-Bact: KW-H(4;20) = 16,8143; p = 0,0021 Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
1E9
1,5E9
2E9
2,5E9
3E9
3,5E9
4E9
4,5E9
5E9
5,5E9
1-B
+D
iplo
1-B+Diplo: KW-H(4;20) = 16,8286; p = 0,0021
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
0
1E9
2E9
3E9
4E9
5E9
6E9
7E91-B
+E
noplo
ides
1-B+Enoploides: KW-H(4;20) = 17,0286; p = 0,0019 Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
0
1E9
2E9
3E9
4E9
5E9
6E9
7E9
8E9
1-B
+A
donchola
imus
1-B+Adoncholaimus: KW-H(4;20) = 16,6143; p = 0,0023
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
1E9
2E9
3E9
4E9
5E9
6E9
7E9
1-B
+D
iplo
+E
nop
1-B+Diplo+Enop: KW-H(4;20) = 17,0286; p = 0,0019
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
T0 T1 T2 T3 T4
Var1
1E9
1,5E9
2E9
2,5E9
3E9
3,5E9
4E9
4,5E9
5E9
5,5E9
6E9
1-B
+D
iplo
+A
don
1-B+Diplo+Adon: KW-H(4;20) = 16,8143; p = 0,0021
Figure 10: The mean±SD of
activity of aminopeptidase
(nmol/g*h) of 4 replicates by
treatment and sampling time, in
the experiment D1. The capital
letters display the significant
differences between sampling
times.
C D C
A A A A B
C D
A A B B
A B C C B
A B C C B
A B C C B
A B C C B
A B C C B
A B C
A B C B
25
Experiment with Diplolaimelloides sp. (5grains/ml) (D2):
The interaction between the effects of Treatments and Time on variability in the
activity of ß-D-glucosidase was statistically significant (PERMANOVA TrxTi p<0.05;
Table 5).
Table 5: Results from the two –way crossed PERMANOVA analysis of activity ß-D-
glucosidase in response to different treatments and sampling times in the experiment
D2.
Source df SS MS Pseudo-F P(perm)
Tr 8 3,0696E18 3,837E17 40,92 0,0001
Ti 4 2,4667E18 6,1668E17 65,765 0,0001
TrxTi 32 2,3221E18 7,2566E16 7,7387 0,0001
Res 135 1,2659E18 9,377E15
Total 179 9,1244E18
The ß-D-glucosidase activity showed a similar behavior as in the previous
experiment D1. The control A presented the lowest enzymatic activity at all sampling
times, differing significantly from the rest of treatments, except in the T4
(PERMANOVA p<0.05, A< FS, Ar, B, BP, BE, BA, BPE, BPA in T0, T1, T3; A< FS,
Ar, B, BP, BE, BA, BPE in T2; A< FS in T4; Figure 11). In contrast, the enzymatic
activity in the controls Ar and FS fluctuted through time; there were few significant
differences between treatments at every sampling time (PERMANOVA p<0.05, FS >
Ar, B, BP, BE, BA, BPE, BPA; B, BPA < BA and BE > BPA in T0; Ar < FS, B, BP,
BE, BA, BPE, BPA; FS > BP and BE, BA > BPE, BPA in T1; FS, Ar < BE; B,BE, BA
> BPA and BP < B, BE in T2; FS > BPE; Ar > FS, BP, BA, BPE, BPA and BP, BPE <
BA in T3; Ar > BP in T4; Figure 11).
26
The obtained results for the ß-D-glucosidase activity showed the same tendency
through time than in the experiment D1. The enzymatic activity started low, then
increased and finally decreased, which was reflected in the registered significant
differences in all the treatments (PERMANOVA p<0.05, T0, T1, T4 < T2, T3 in Ar; T0,
T4 < T1, T2 and T2 >T3 in B; T0, T2, T3, T4 < T1 in BP; T0, T3, T4 < T1, T2 in BE;
T4 < T0, T3 < T2 < T1 in BA; T0, T3, T4 < T1, T2 and T0> T3 in BPE; T0, T3, T4 <
T1 in BPA; Figure 12). The control A and FS were shared between the experiment D1
G-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E8
0
1E8
2E8
3E8
4E8
5E8
6E8
T0
T0: KW-H(8;36) = 26,4414; p = 0,0009
G-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-2E8
0
2E8
4E8
6E8
8E8
1E9
T1
T1: KW-H(8;36) = 27,0856; p = 0,0007
G-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-2E8
0
2E8
4E8
6E8
8E8
1E9
T2
T2: KW-H(8;36) = 25,5495; p = 0,0013
G-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-5E7
0
5E7
1E8
1,5E8
2E8
2,5E8
3E8
3,5E8
4E8
4,5E8
T3
T3: KW-H(8;36) = 22,8604; p = 0,0035 G-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E8
0
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
T4
T4: KW-H(8;36) = 21,3739; p = 0,0062
Figure 11: The mean±SD of activity of
ß-D-glucosidase (nmol/g*h) of 4
replicates by treatment and sampling
time, in the experiment D2. The capital
letters display the significant
differences between treatments.
A B C C C C C C C
C D C
E C
A B C A A A A A A
D A A D D
A B A A D
B A A A
A C A A A A C
B C
A B B B B B B B B
A A/E D E
A B C A/C A A/C A A A
27
and D2, therefore the same significant differences between them were observed in both
experiments.
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
1,2E8
1,4E8
1,6E8
1,8E8
2E8
2,2E8
2,4E8
2,6E8
2,8E8
3E8
3,2E8
3,4E8
3,6E8
3,8E8
4E8
4,2E8
4,4E82
-AS
W r
2-ASW r: KW-H(4;20) = 14,5857; p = 0,0056
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
9E8
1E9
2-B
ac
t
2-Bact: KW-H(4;20) = 14,3857; p = 0,0062 Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
5E7
1E8
1,5E8
2E8
2,5E8
3E8
3,5E8
4E8
4,5E8
5E8
2-B
+D
iplo
2-B+Diplo: KW-H(4;20) = 11,2143; p = 0,0243
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
0
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
9E8
1E9
2-B
+E
no
plo
ide
s
2-B+Enoploides: KW-H(4;20) = 14,1714; p = 0,0068 Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
9E8
2-B
+A
donchola
imus
2-B+Adoncholaimus: KW-H(4;20) = 17,5; p = 0,0015
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
2-B
+D
iplo
+E
no
p
2-B+Diplo+Enop: KW-H(4;20) = 16,8286; p = 0,0021 Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
-1E8
0
1E8
2E8
3E8
4E8
5E8
6E8
7E8
8E8
2-B
+D
iplo
+A
don
2-B+Diplo+Adon: KW-H(4;20) = 10,3; p = 0,0357
Figure 12: The mean±SD of
activity of ß-D-glucosidase
(nmol/g*h) of 4 replicates by
treatment and sampling time, in
the experiment D2. The capital
letters display the significant
differences between sampling
times.
A A B B A
B A
A B B A
A B A A A
A B B A A
A B C A D
A B B C A/C
A B A A
28
Regarding to the activity of aminopeptidase we also found statistically
significant interaction in the effects of Treatments and Time (PERMANOVA TrxTi
p<0.05; Table 6).
Table 6: Results from the two –way crossed PERMANOVA analysis of activity
aminopeptidase in response to different treatments and sampling times in the
experiment D2.
Source df SS MS Pseudo-F P(perm)
Tr 8 8,0719E20 1,009E20 197,92 0,0001
Ti 4 1,4036E20 3,5091E19 68,834 0,0001
TrxTi 32 1,0633E20 3,3228E18 6,518 0,0001
Res 135 6,8822E19 5,0979E17
Total 179 1,1227E21
The aminopeptidase activity also displayed a similar behaviour as in the
experiment D1. The control ASW showed the lowest enzymatic activity, followed by
the control ASWr. Both controls exhibited significant differences compared to the other
treatments (PERMANOVA p<0.05, A < Ar < FS, B, BP, BE, BA, BPE, BPA in T0, T1,
T2, T4; A < Ar, FS < B, BP, BE, BA, BPE, BPA in T3; Figure 13). Significant
differences were also found between treatments and control FS, but only few
differences between treatments with bacteria were statistically significant
(PERMANOVA p<0.05, FS < B < BPA; FS < BPE and FS, B, BP, BPE < BA in T0;
FS, B, BP < BA, BPA and FS < B, BE, BPE in T1; FS < B, BP < BA, BPE, BPA and
FS, B < BE in T2; B, BP < BA, BPE, BPA in T3; FS, B, BP, BA < BPE and B < BA,
BPA in T4; Figure 13).
29
The same tendency was observed in the aminopeptidase activity as in the
experiment D1 for all the treatments. All significant differences supported this tendency
(PERMANOVA p<0.05, T0 < T2,T3, T4 in Ar; T0 < T1, T3, T4 < T2 in B; T0, T1 <
T2; T2 > T3 and T0 < T3, T4 in BP; T0 < T1, T2, T3, T4 in BE; T0 < T4 < T1, T2, T3
in BA; T0 < T1 < T2, T3, T4 in BPE; T0 < T1 < T2, T3 and T0 < T4 < T2 in BPA;
Figure 14). The controls A and FS were shared between the experiment D1 and D2,
therefore the same significant differences were observed in both experiments.
P-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E9
0
1E9
2E9
3E9
4E9
5E9
6E9
T0
T0: KW-H(8;36) = 29,6171; p = 0,0002
P-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E9
0
1E9
2E9
3E9
4E9
5E9
6E9
7E9
8E9
9E9
T1
T1: KW-H(8;36) = 31,6532; p = 0,0001 P-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-2E9
0
2E9
4E9
6E9
8E9
1E10
1,2E10
T2
T2: KW-H(8;36) = 30,8919; p = 0,0001
P-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-1E9
0
1E9
2E9
3E9
4E9
5E9
6E9
7E9
8E9
9E9
T3
T3: KW-H(8;36) = 30,6216; p = 0,0002 P-enzyme Diplo 2
Mean Mean±SD
Fre
sh s
ed
AS
W
AS
W r
Bact
B+
Dip
lo
B+
Enoplo
ides
B+
Adonchola
imus
B+
Dip
lo+
Enop
B+
Dip
lo+
Adon
Var1
-2E9
0
2E9
4E9
6E9
8E9
1E10
T4
T4: KW-H(8;36) = 27,5135; p = 0,0006
Figure 13: The mean±SD of activity
of aminopeptidase (nmol/g*h) of 4
replicates by treatment and sampling
time, in the experiment D2. The
capital letters display the significant
differences between treatments.
A B A C C C/D D D D F F F F E
A B C D D E E E
F F F F F E E
A B C D A/D D D
A G
F F F F E
A B C A A A/D D A
A D D
D D D D D D E
A B C A A A A A
30
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
1,8E9
2E9
2,2E9
2,4E9
2,6E9
2,8E9
3E9
3,2E9
3,4E9
2-A
SW
r
2-ASW r: KW-H(4;20) = 13,1; p = 0,0108
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
2,5E9
3E9
3,5E9
4E9
4,5E9
5E9
5,5E9
6E9
6,5E9
7E9
2-B
ac
t
2-Bact: KW-H(4;20) = 14,1143; p = 0,0069 Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
3E9
3,5E9
4E9
4,5E9
5E9
5,5E9
6E9
6,5E9
7E9
7,5E9
2-B
+D
iplo
2-B+Diplo: KW-H(4;20) = 16,1714; p = 0,0028
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
2E9
3E9
4E9
5E9
6E9
7E9
8E9
9E9
1E10
1,1E10
2-B
+E
no
plo
ide
s
2-B+Enoploides: KW-H(4;20) = 11,8857; p = 0,0182 Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean
Mean±SD T0 T1 T2 T3 T4
Var1
3E9
4E9
5E9
6E9
7E9
8E9
9E9
1E10
2-B
+A
do
nc
ho
laim
us
2-B+Adoncholaimus: KW-H(4;20) = 15,8143; p = 0,0033
Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
3E9
4E9
5E9
6E9
7E9
8E9
9E9
1E10
2-B
+D
iplo
+E
no
p
2-B+Diplo+Enop: KW-H(4;20) = 15,3286; p = 0,0041 Box Plot (Spreadsheet en Fini STATISTICA ANOVAS x time 3 exp 24v*20c)
Mean Mean±SD
T0 T1 T2 T3 T4
Var1
3E9
4E9
5E9
6E9
7E9
8E9
9E9
2-B
+D
iplo
+A
don
2-B+Diplo+Adon: KW-H(4;20) = 15,6143; p = 0,0036
Figure 14: The mean±SD of activity
of aminopeptidase (nmol/g*h) of 4
replicates by treatment and sampling
time, in the experiment D2. The
capital letters display the significant
differences between sampling times.
A B B B A B C B/C B
A B C B/C
A B A B B B B
A B B C
A B
A C B
A B C C
A B C C C
31
Nested analysis of experiments D1 and D2:
On the other hand, the results of the experiments D1 and D2 were examined with
a nested PERMANOVA design to better evaluate the effect of the different rice extracts
on the variability of the enzymatic activity. The Treatment factor was nested within
Rice extract level and in turn the Time factor was nested within Treatment (Figure 15).
The amount of variation that was attributable to the different levels in the model was
estimated by the components of variation for these levels (Table 7).
Figure 15: Schematic diagram of the sampling design for enzymatic activity.
Rice extracts 1 ml /grain
Treatments: Ar B BP BE BA BPE BPA
Time: T0 T1 T2 T3 T4
Etc.
n = 4 replicates
Rice extracts 0.2 ml /grain
Treatments: Ar B BP BE BA BPE BPA
Time: T0 T1 T2 T3 T4
Etc.
n = 4 replicates
32
Table 7: Results of nested PERMANOVA analysis of activity ß-D-glucosidase in
response to different rice extracts, treatments and sampling times in the experiments D1
and D2.
Source df SS MS Pseudo-F
P(perm)
Ri 1 1,4028E18 1,4028E18 11,195
0,0084
Tr(Ri) 12 1,5037E18 1,2531E17 1,1607
0,3259
Ti(Tr(Ri)) 56 6,0454E18 1,0795E17 15,671
0,0001
Res 210 1,4466E18 6,8886E15
Total 279 1,0398E19
Estimates of components of variation
Source Estimate Sq.root
S(Ri) 9,1251E15 9,5526E7
V(Tr(Ri)) 8,6766E14 2,9456E7
V(Ti(Tr(Ri))) 2,5266E16 1,5895E8
V(Res) 6,8886E15 8,2997E7
The obtained results for activity of ß-D-glucosidase indicated that the variability
derived from the Rice extract level and Time level significantly affected the enzymatic
activity (PERMANOVA p<0.05, Table 7). Particularly, the greatest component of
variation occurred at the Time scale, followed by Rice extract (Table 7). However the
Treatment level was the less important term in the model to explain overall variation
(Table 7).
33
Table 8: Results of nested PERMANOVA analysis of activity aminopeptidase in
response to different rice extracts, treatments and sampling times in the experiments D1
and D2.
Source df SS MS Pseudo-F P(perm)
Ri 1 4,7575E20 4,7575E20 15,777
0,0045
Tr(Ri) 12 3,6185E20 3,0154E19 3,7208
0,0005
Ti(Tr(Ri)) 56 4,5384E20 8,1043E18 19,296
0,0001
Res 210 8,8202E19 4,2001E17
Total 279 1,3796E21
Estimates of components of variation
Source Estimate Sq.root
S(Ri) 3,1828E18 1,7841E9
V(Tr(Ri)) 1,1025E18 1,05E9
V(Ti(Tr(Ri))) 1,9211E18 1,386E9
V(Res) 4,2001E17 6,4808E8
The activity of aminopeptidase showed that all levels in the design contributed
significantly to the variability of the enzymatic activity (PERMANOVA p<0.05, Table
8). The estimates of components of variation indicated that the highest variability
occurred at Rice extract level, followed by Time level and lastly the Treatment scale
(Table 8).
34
DISCUSSION:
We performed experiments to assess effects of nematodes representing two
different trophic levels on microbial activity. Based on previous experiments, we
expected that bacterivorous nematodes would measurably affect microbial activity, and
that predatory nematodes would impact prey nematode abundances. The latter aspect
could not be completed within the time frame of this study, but several earlier
microcosm studies with the predator-prey combinations used in our experiments with
Diplolaimelloides species (Moens et al., 1999; Moens et al., 2000; Gallucci et al., 2005)
allow us to confidently assume that at least in the treatments with E. longispiculosus,
monhysterid prey abundance should have been substantially affected. Hence, we
hypothesized that the presence of predators would at least partly counteract the expected
effect of bacterivorous nematodes on bacterial activity.
We measured the activity of two exo-enzymes as proxies of bacterial activity to
test these cascading effects. The quantitative estimates of bacterial extracellular enzyme
activity may be used as indications of the bacterial growth and the bacterial substrate
uptake, despite these measurements do not provide a complete picture of microbial
activity (Hoppe, 1993). The main advantages of this method compared to more
informative tests, such as the measurement of bacterial carbon production, are a
relatively easy protocol and short incubation periods (Hoppe, 1993). Moreover enzyme
activity in natural sediments might provide valuable information regarding the organic
matter flux in the microbial loop as well as of the quantity and quality of the available
organic matter to heterotrophs (Fabiano & Danovaro, 1998).
In our study both ß-D-glucosidase and aminopeptidase activity showed a similar
temporal pattern in all experiments, except for the aminopeptidase activity in the LM
experiment that exhibited a decreasing tendency of the enzymatic activity after 6-day. In
the rest of experiments, in all the treatments with bacteria a maximum enzymatic
activity was displayed in the middle of the experiments and the enzymatic activity
decreased significantly towards the end of the 8-day incubations. Temporal patterns
were not mainly driven by nematodes, as the patterns appeared in the treatments with
nematodes as well as in the treatments without them. In fact, the estimates of the
components of variation in the nested design for the experiments D1 and D2 supported
that the most (for ß-D-glucosidase) or second most (for aminopeptidase) important
35
source of variability of enzymatic activity was time. We suggest that the decline of
enzymatic activity towards the end of the experiment was due to the depletion of
available nutrients through time, as we started with a poor medium, despite the
provision of external organic matter. The lack of nutrients, as well as the surface area,
has been previously recorded as limiting factors for the growth of microbial populations
in aquatic sediments (Novitsky, 1987). Traunspurger et al. (1997) studied the effects of
bacterivorous nematodes on bacterial activity and abundance in freshwater sediment
running lab experiments during 5, 7 and 17 days. The authors observed that the
presence of nematodes stimulated the activity and abundance of bacteria, but with a
decreasing influence through time, i.e. the effect was strongest in the early phase of the
experiments. Therefore the Time factor might partly mask the direct and/or indirect
effects of the bacterivorous and predatory nematodes on the bacterial activity.
Nevertheless, the factor nematode treatment was also highly significant for
aminopeptidase activity, and in the two-way analysis of all experiments, Treatment
factor always exhibited significant effects. Often, these effects revealed differences
between treatments with nematodes and controls without nematodes. We therefore have
substantial evidence that nematodes in our experiments affected bacterial activity to
some degree. Such effects can be the result of direct or indirect processes.
The response of microbial populations to grazing (direct process) has varied
greatly in previous studies. Bacterivorous nematodes may increase or decrease the
bacterial growth or the abundance of bacteria may remain unaffected by their nematode
grazers (Ingham et al., 1985). In our study the measurements of ß-D-glucosidase
activity showed in all the experiments a similar pattern. The ß-D-glucosidase activity
was significantly higher in the treatment with only bacteria than in the treatments with
bacterial-feeding nematodes, either was not significantly different. Particularly, in the
LM experiment, the ß-D-glucosidase activity in the treatment B was higher than in the
treatment BPA in T2 and in the treatment BPE inT4, as well as the enzymatic activity in
T3 was significantly higher in treatment B than in the treatments BP, BPE and BPA. In
the experiment D1 in T1 the treatment B showed significantly higher activity than
treatment BPA, and in T2 and T3 the treatment B presented significantly higher
enzymatic activity than the treatments BP, BPE and BPA. Ultimately in the experiment
D2 the ß-D-glucosidase activity in the treatment B was significantly higher than in
treatment BP and BPA in T2. This pattern may be due to the direct consumption of
36
bacteria by nematodes, what can sometimes reduce bacteria abundance and activity (De
Mesel et al., 2003; De Mesel et al., 2004; De Mesel et al., 2006; Hubas et al., 2010).
However the measurements of aminopeptidase activity in the diverse
experiments showed no consistent differences between treatments. In the LM
experiment, significant differences were found after 4 and 8 days, with higher
enzymatic activity in the treatment with only bacteria than in the treatments BP and
BPE. In the experiment D1, there were no significant differences in aminopeptidase
activity between treatments with and without nematodes, while in the experiment D2
the treatment B had a lower aminopeptidase activity than the treatments BPE and BPA
at all sampling times. These last results in experiment D2 supported the presence of
stimulatory effects on sediment-living bacteria that have been widely reported not only
in presence of nematodes, but also of other benthic invertebrates (e.g. Fabiano &
Danovaro, 1998; Moriarty 1986; van de Bund et al. 1994). There are different possible
mechanisms for stimulation of bacterial population growth by bacterivorous nematodes,
as described in the introduction. However, our study does not allow elucidating the
mechanisms underlying the observed results. More importantly, nevertheless, it is that
stimulatory effects on microbial activity were exception rather than rule in our
experiments, as the microbial activity was in most cases amongst the highest in the
absence of nematodes.
Moreover, Ingham et al. (1985) suggested that the response of bacterial biomass
to nematode grazing differs with nematode species identity and density. The three
nematode species used in our experiment belong to the same functional group, the
deposit feeders (Moens & Vincx, 1998), however these species showed differential
responses to food availability. Rhabditid nematodes typically graze at higher rates than
monhysterids if sufficiently high bacterial densities are present. Therefore we might
expect a stronger effect of L. marina than of Diplolaimelloides species. Nevertheless the
results in the different experiments did not differ widely. Two possible reasons, non-
mutually exclusive, may explain this situation: 1) bacterial densities were not
sufficiently high for L. marina (dos Santos et al. 2009), and/or 2) bacteria may have be
less available to L. marina in sandy sediments, as their usual habitat is macroalgal
wrack (dos Santos et al. 2009).
37
Specifically, the effects of D. meyli and D. oschei, and other monhysterid
nematodes, on bacterial activity and detritus decomposition rates have been previously
documented (De Mesel et al., 2003; De Mesel et al., 2004; De Mesel et al., 2006;
Hubas et al., 2010). As in our experiments, significant declines of the microbial activity
were recorded even at relatively low nematode densities; in addition the differences
were found dependent on the identity of the nematode species present (De Mesel et al.,
2003; De Mesel et al., 2004; De Mesel et al., 2006). However Hubas et al. (2010)
detected no negative impact on bacterial proliferation in the laboratory experiment at
low nematode densities and suggested that it probably only occur at high abundances of
nematodes with high grazing rates. In our experiments was remarkable an expected
grazing effect for the nematodes densities used, which were comparable to the densities
found in the natural intertidal sediments (Heip et al. 1985), but it would need to test
different densities of bacterivorous nematodes to evaluate the effect of the density.
Besides the effect of bacterivorous nematodes, the effects of predatory
nematodes may be important too for the bacterial activity. In previous laboratory
experiments high predation rates of predacious nematodes have been reported for
marine nematode species such as E. longispiculosus and A. fuscus (Moens et al., 1999;
Moens et al., 2000; Hamels et al., 2001; Gallucci et al., 2005; Moens et al., 2005; dos
Santos & Moens, 2011). Therefore we expected that the predation had a significant
effect on prey.
Trophic-dynamic theories predict that the abundance of a particular trophic level
decrease by the addition of a new level situated an odd number from the given level
(Carpenter et al., 1985; Mikola & Setala, 1998; Laakso & Setala, 1999). Accordingly, in
our experiments we expected that the presence of a predator level would reduce the
biomass of their prey, and that this reduction in turn would partly release bacteria from
grazing. Therefore, the treatments with only prey nematodes and those treatments with
combined prey and predator species were compared.
Particularly in the experiment LM there were no significant differences in β-D-
glucosidase activity between treatments with L. marina and treatments with L. marina
plus predators. In the case of aminopeptidase activity the only significant difference was
found in T4 between treatment BP and treatment BPA, the latter displaying a higher
enzymatic activity. The absence of a cascading effect in the LM experiment could be
38
linked to a lower suitability of L. marina as prey to the predatory nematodes. However,
Moens et al. (2000) still found substantial predation of E. longispiculosus on L. marina,
albeit less than on the monhysterids. In experiment D1, all treatments containing
bacterivorous nematodes had substantially lower β-D-glucosidase activities than the B
treatment or the predator-only treatments. This demonstrates a significant effect of
Diplolaimelloides sp. on bacterial activity, but one that was not counteracted by the
addition of predators. Differences between treatment BP and treatments BPE and BPA
were found only in T1 and T3, the ß-D-glucosidase activity in treatment BP was higher
rather than lower than in the treatments with predators. In contrast, the experiment D2
presented no differences between these treatments. For aminopeptidase activity, in the
experiment D1 no significant differences were found between these treatments, while in
the experiment D2 the treatment BP displayed a lower aminopeptidase activity than
treatments BPE and BPA at all sampling times. The latter data are the only ones
indicating a cascading effect of predators over their prey to the activity of their prey’s
prey. However, given that in this dataset, treatment with bacterivorous nematodes did
not significantly differ from treatment with only bacteria, we cannot really assume that
the predators counteracted an effect of the bacterivorous nematodes. Therefore the
comparisons of treatments regarding both enzymes showed little evidence in support of
trophic cascades.
Our results add to the increasing evidence that propagation of top-down
influences attenuates in the lower levels both in terrestrial (Mikola & Setala, 1998;
Laakso & Setala, 1999) and in aquatic (McQueen et al., 1989) food webs. For instance,
top-down cascades were recorded between piscivores, zooplankton and planktivores in
lakes and there were no evidence for cascading effects between zooplanktivores and
phytoplankton (McQueen et al., 1989). McQueen et al. (1989) using data from
freshwater pelagic ecosystems predicted that the cascading effects in a top-down control
weaken from top towards the bottom. In soil food webs it was observed that the length
of the food chain affected to microbial population when it was measured the
phospholipids fatty acids, respiration and ammonium-N concentration (Mikola &
Setala, 1998). However, appreciable differences between tested chains of two and three
trophic levels were not found, i.e the abundance of bacteria did not respond to the
presence of predator species (Mikola & Setala, 1998), as it is predicted by the trophic-
dynamic theories (Carpenter et al., 1985). In posterior experiments Laakso & Setala
39
(1999) measured respiration rates and biomass of microbivore nematodes and microbes
in a detrital food web. The authors recorded reductions of more than 50% in nematode
respiration and biomass when the predators were present while the reduction in
microbial respiration was only 16%. This suggested that the top-down control is
unlikely to exert a relevant influence on other organisms than their prey.
Several reasons have been given to explain the rare occurrence of cascading
effects. The excretion of nutrients by microbial grazers causes an indirect grazing-
induced stimulation of microbial activity. This supply of nutrients compensates the
effects of direct consumption; therefore it is one of the main mechanisms preventing the
occurrence of trophic cascades until the level of microbes (Laakso & Setala, 1999).
Consequently it is thought that cascading effects may possibly only occur in systems
with plants, as the indirect stimulation of grazing on microbial activity may be damped
by the uptake of nutrients by plants (Laakso & Setala, 1999). In natural decomposer
food webs, Mikola & Setala (1998) suggested that the high turnover rate of fungi and
bacteria loops is one of the main reasons to make the occurrence of trophic cascades
rare and unpredictable. Furthermore the authors pointed the grazing may be focused on
the most palatable microbes releasing to the less palatable species, which could increase
their biomass, thus preventing the trophic cascade. On the other hand specific-species
responses of prey to predation may affect differently to the strength of trophic cascades
and may reduce the cascading effects (Polis & Strong, 1996). However, we do not
dispose of enough information to explain the mechanisms underlying the no occurrence
of trophic cascades in our experiments.
The variability between experiments D1 and D2 were probably derived from the
different rice extract used, as the rice extract factor was the main component of
variation in analyses of aminopeptidase activity and the second main component in
analyses of β-D-glucosidase activity after Time factor. The experiment D2, where we
used the most concentrated rice extract, presented higher β-D-glucosidase activity and
aminopeptidase activity than experiment D1. However the obtained results of β-D-
glucosidase activity from D1 were more concluding than for D2. In turn the results of
aminopeptidase activity from D2 were more meaningful than in D1. Therefore we
concluded that the top-down effects were more pronounced at relatively low β-D-
glucosidase activity while regarding the aminopeptidase activity the cascading effects
were more remarkable at higher microbial activity. Considering this strong substrate
40
effect, we suggested that, probably, bottom-up effects may have occurred in our
experiments, i.e. the nutrient availability influenced the higher trophic levels in the food
web. Resource availability may limit the population or biomass of higher trophic levels
determining community structure and function through both direct and indirect
processes (Posey et al., 1995; Posey et al., 2002). Nevertheless further rice extracts
should be tested using different prey nematode species to confirm this hypothesis.
Further research involving more diverse and complex assemblages are required
to elucidate the mechanisms driving the trophic cascades and to be able to extrapolate
the lab results to natural habitats. These studies should help to understand better the
oceanic top-down control and its interactions with bottom-up control. Ultimately
advances of this nature would facilitate the mitigation and the forecast of effects of
changing oceanic predator abundances, both of which are crucial for the successful
long-term management of oceanic resources (Baum & Worm, 2009).
ACKNOWLEDGEMENTS
My thanks to Professor Moens for putting at my disposal his long experience in
the research and for their assistance in the preparation of this thesis. Thanks also to my
supervisors Tania Bezerra and Xiuquin Wu for their support and guidance, and to all
laboratory technicians of the Department of Marine Biology, whose invaluable
assistance has made possible to conduct this thesis.
41
REFERENCES
ABRAMS, P. A. & MATSUDA, H. (1996). Positive indirect effects between prey species
that share predators. Ecology 77, 610-616.
BAUM, J. K. & WORM, B. (2009). Cascading top-down effects of changing oceanic
predator abundances. Journal of Animal Ecology 78, 699-714.
CARPENTER, S. R., KITCHELL, J. F. & HODGSON, J. R. (1985). Cascading trophic
interactions and lake productivity. Bioscience 35, 634-639.
COULL, B. C. (1986). Long-term variability of meiobenthos - value, synopsis,
hypothesis generation and predictive modeling. Hydrobiologia 142, 271-279.
COULL, B. C. (1999). Role of meiofauna in estuarine soft-bottom habitats. Australian
Journal of Ecology 24, 327-343.
DANOVARO, R., ARMENI, M., DELL'ANNO, A., FABIANO, M., MANINI, E., MARRALE, D.,
PUSCEDDU, A. & VANUCCI, S. (2001). Small-scale distribution of bacteria,
enzymatic activities, and organic matter in coastal sediments. Microbial Ecology
42, 177-185.
DE MEESTER, N., DERYCKE, S. & MOENS, T. (2012). Differences in Time until Dispersal
between Cryptic Species of a Marine Nematode Species Complex. Plos One 7.
DE MESEL, I., DERYCKE, S., MOENS, T., VAN DER GUCHT, K., VINCX, M. & SWINGS, J.
(2004). Top-down impact of bacterivorous nematodes on the bacterial
community structure: a microcosm study. Environmental Microbiology 6, 733-
744.
DE MESEL, I., DERYCKE, S., SWINGS, J., VINCX, M. & MOENS, T. (2003). Influence of
bacterivorous nematodes on the decomposition of cordgrass. Journal of
Experimental Marine Biology and Ecology 296, 227-242.
DE MESEL, I., DERYCKE, S., SWINGS, J., VINCX, M. & MOENS, T. (2006). Role of
nematodes in decomposition processes: Does within-trophic group diversity
matter? Marine Ecology Progress Series 321, 157-166.
DOS SANTOS, G., DERYCKE, S., FONSÊCA-GENEVOIS, V., COELHO, L., CORREIA, M. &
MOENS, T. (2008). Differential effects of food availability on population growth
and fitness of three species of estuarine, bacterial-feeding nematodes. Journal of
Experimental Marine Biology and Ecology 355, 27-40.
DOS SANTOS, G. A., DERYCKE, S., GENEVOIS, V. G., COELHO, L. C., CORREIA, M. T. &
MOENS, T. (2009). Interactions among bacterial-feeding nematode species at
different levels of food availability. Marine Biology 156, 629-640.
DOS SANTOS, G. A. P. & MOENS, T. (2011). Populations of two prey nematodes and their
interaction are controlled by a predatory nematode. Marine Ecology Progress
Series 427, 117-131.
FABIANO, M. & DANOVARO, R. (1998). Enzymatic activity, bacterial distribution, and
organic matter composition in sediments of the Ross Sea (Antarctica). Applied
and Environmental Microbiology 64, 3838-3845.
GALLUCCI, F., STEYAERT, M. & MOENS, T. (2005). Can field distributions of marine
predacious nematodes be explained by sediment constraints on their foraging
success? Marine Ecology Progress Series 304, 167-178.
GAMFELDT, L., HILLEBRAND, H. & JONSSON, P. R. (2005). Species richness changes
across two trophic levels simultaneously affect prey and consumer biomass.
Ecology Letters 8, 696-703.
GRUNER, D. S., SMITH, J. E., SEABLOOM, E. W., SANDIN, S. A., NGAI, J. T., HILLEBRAND,
H., HARPOLE, W. S., ELSER, J. J., CLELAND, E. E., BRACKEN, M. E. S., BORER, E.
42
T. & BOLKER, B. M. (2008). A cross-system synthesis of consumer and nutrient
resource control on producer biomass. Ecology Letters 11, 740-755.
HAMELS, I., MOENS, T., MUTYLAERT, K. & VYVERMAN, W. (2001). Trophic interactions
between ciliates and nematodes from an intertidal flat. Aquatic Microbial
Ecology 26, 61-72.
HEIP, C., VINCX, M. A. & VRANKEN, G. (1985). The ecology of marine nematodes.
HOPPE, H.-G. (1993). Use of fluorogenic model substrates for extracellular enzyme
activity (EEA) measurement of bacteria. Handbook of methods in aquatic
microbial ecology. Lewis Publishers, pp. 423-431.
HUBAS, C., SACHIDHANANDAM, C., RYBARCZYK, H., LUBARSKY, H. V., RIGAUX, A.,
MOENS, T. & PATERSON, D. M. (2010). Bacterivorous nematodes stimulate
microbial growth and exopolymer production in marine sediment microcosms.
Marine Ecology Progress Series 419, 85-94.
INGHAM, R. E., TROFYMOW, J. A., INGHAM, E. R. & COLEMAN, D. C. (1985). Interactions
of bacteria, fungi, and their nematode grazers - effects on nutrient cycling and
plant-growth. Ecological Monographs 55, 119-140.
IVES, A. R., CARDINALE, B. J. & SNYDER, W. E. (2005). A synthesis of subdisciplines:
predator-prey interactions, and biodiversity and ecosystem functioning. Ecology
Letters 8, 102-116.
KENNEDY, A. D. (1994). Predation within meiofaunal communities - description and
results of a rapid-freezing method of investigation. Marine Ecology Progress
Series 114, 71-79.
KRISTENSEN, E. (1988). Benthic fauna and biogeochemical processes in marine
sediments: microbial activities and fluxes. In: SORENSEN, T. H. B. A. J. (Ed.)
Nitrogen cycling in coastal marine environments. John Wiley & Sons Ltd, pp.
275-300.
LAAKSO, J. & SETALA, H. (1999). Population- and ecosystem-level effects of predation
on microbial-feeding nematodes. Oecologia 120, 279-286.
MCQUEEN, D. J., JOHANNES, M. R., POST, J. R., STEWART, T. J. & LEAN, D. R. (1989).
Bottom-up and top-down impacts on freshwater pelagic community structure.
Ecological Monographs, 289-309.
MIKOLA, J. & SETALA, H. (1998). No evidence of trophic cascades in an experimental
microbial-based soil food web. Ecology 79, 153-164.
MOENS, T., BOUILLON, S. & GALLUCCI, F. (2005). Dual stable isotope abundances
unravel trophic position of estuarine nematodes. Journal of the Marine
Biological Association of the United Kingdom 85, 1401-1407.
MOENS, T., BRAECKMAN, U., DERYCKE, S., FONSECA, G., GALLUCCI, F., INGELS, J.,
LEDUC, D., VANAVERBEKE, J., VAN COLEN, C. & VANREUSEL, A. (2013).
Ecology of free-living marine nematodes. In: SCHMIDT-RHAESA, A. (Ed.)
Handbook of zoology: Gastrotricha, Cycloneuralia and Gnathifera, vol. 2:
Nematoda. Berlin, Germany, De Gruyter, pp. 109-152.
MOENS, T., HERMAN, P., VERBEECK, L., STEYAERT, M. & VINCX, M. (2000). Predation
rates and prey selectivity in two predacious estuarine nematode species. Marine
Ecology Progress Series 205, 185-193.
MOENS, T., VERBEECK, L. & VINCX, M. (1999). Feeding biology of a predatory and a
facultatively predatory nematode (Enoploides longispiculosus and
Adoncholaimus fuscus). Marine Biology 134, 585-593.
MOENS, T. & VINCX, M. (1997). Observations on the feeding ecology of estuarine
nematodes. Journal of the Marine Biological Association of the United Kingdom
77, 211-227.
43
MOENS, T. & VINCX, M. (1998). On the cultivation of free-living marine and estuarine
nematodes. Helgolander Meeresuntersuchungen 52, 115-139.
MOENS, T. & VINCX, M. (2000). Temperature, salinity and food thresholds in two
brackish-water bacterivorous nematode species: assessing niches from food
absorption and respiration experiments. Journal of Experimental Marine Biology
and Ecology 243, 137-154.
MORIARTY, D. J. W. (1986). Measurement of Bacterial Growth Rates in Aquatic
Systems from Rates of Nucleic Acid Synthesis. In: MARSHALL, K. C. (Ed.)
Advances in Microbial Ecology. Springer US, pp. 245-292.
NAVARRETE, S. A. & CASTILLA, J. C. (2003). Experimental determination of predation
intensity in an intertidal predator guild: dominant versus subordinate prey. Oikos
100, 251-262.
O'GORMAN, E. J., ENRIGHT, R. A. & EMMERSON, M. C. (2008). Predator diversity
enhances secondary production and decreases the likelihood of trophic cascades.
Oecologia 158, 557-567.
OLAFSSON, E. (2003). Do macrofauna structure meiofauna assemblages in marine soft-
bottoms? A review of experimental studies. Vie Et Milieu-Life and Environment
53, 249-265.
POLIS, G. A. & STRONG, D. R. (1996). Food web complexity and community dynamics.
American Naturalist 147, 813-846.
POSEY, M., POWELL, C., CAHOON, L. & LINDQUIST, D. (1995). Top-down vs bottom up
control of benthic community composition on an intertidal tideflat. Journal of
Experimental Marine Biology and Ecology 185, 19-31.
POSEY, M. H., ALPHIN, T. D., CAHOON, L. B., LINDQUIST, D. G., MALLIN, M. A. &
NEVERS, M. B. (2002). Top-down versus bottom-up limitation in benthic
infaunal communities: Direct and indirect effects. Estuaries 25, 999-1014.
RIEMANN, F. & HELMKE, E. (2002). Symbiotic relations of sediment-agglutinating
nematodes and bacteria in detrital habitats: The enzyme-sharing concept. Marine
Ecology-Pubblicazioni Della Stazione Zoologica Di Napoli I 23, 93-113.
RIEMANN, F. & SCHRAGE, M. (1978). The mucus-trap hypothesis on feeding of aquatic
nematodes and implications for biodegradation and sediment texture. Oecologia
34, 75-88.
SHURIN, J. B., BORER, E. T., SEABLOOM, E. W., ANDERSON, K., BLANCHETTE, C. A.,
BROITMAN, B., COOPER, S. D. & HALPERN, B. S. (2002). A cross-ecosystem
comparison of the strength of trophic cascades. Ecology Letters 5, 785-791.
SOMERFIELD, P., WARWICK, R. & MOENS, M. (2005). Meiofauna techniques. Methods
for the study of marine benthos. pp. 229-272.
TRAUNSPURGER, W., BERGTOLD, M. & GOEDKOOP, W. (1997). The effects of nematodes
on bacterial activity and abundance in a freshwater sediment. Oecologia 112,
118-122.
VAN DER MEER, J., HEIP, C., HERMAN, P., MOENS, T. & VAN OEVELEN, D. (2005).
Measuring the flow of energy and matter in marine benthic animal populations.
Methods for the study of marine benthos. pp. 326-407.
VANDEBUND, W. J., GOEDKOOP, W. & JOHNSON, R. K. (1994). Effects of deposit-feeder
activity on bacterial production and abundance in profundal lake sediment.
Journal of the North American Benthological Society 13, 532-539.
WARWICK, R. (1987). Meiofauna: their role in marine detrital systems.
WORM, B., LOTZE, H. K., HILLEBRAND, H. & SOMMER, U. (2002). Consumer versus
resource control of species diversity and ecosystem functioning. Nature 417,
848-851.
44
YEATES, G. W., FERRIS, H. & MOENS, T. (2009). The Role of Nematodes in Ecosystems.
In: KAKOULI-DUARTE, M. J. W. A. T. (Ed.) Nematodes as environmental
indicators. Wallingford, UK, CAB International, pp. 1-44.