PDF (2750 KB)

Microbial predators promote their competitors: commensalism
within an intra-guild predation system in microzooplankton
MARTIN GU¨NTHER JOACHIM LO¨DER, MAARTEN BOERSMA, ALEXANDRA CLAUDIA KRABERG,
NICOLE ABERLE,
AND
KAREN HELEN WILTSHIRE
Alfred-Wegener-Institut, Helmholtz-Zentrum fu¨r Polar- und Meeresforschung, Biologische Anstalt Helgoland,
P.O. Box 180, 27483 Helgoland, Germany
Citation: Lo¨der, M. G. J., M. Boersma, A. C. Kraberg, N. Aberle, and K. H. Wiltshire. 2014. Microbial predators promote
their competitors: commensalism within an intra-guild predation system in microzooplankton. Ecosphere 5(10):128.
http://dx.doi.org/10.1890/ES14-00037.1
Abstract. This study elucidates the interspecific interactions between competing unicellular predators in
an intraguild predation system. The organisms studied were two microzooplankton (MZP) predators
competing for the phototrophic dinoflagellate prey Scrippsiella trochoidea. Since the smaller dinoflagellate
predator Gyrodinium dominans was also potential prey for the larger predator, the tintinnid ciliate Favella
ehrenbergii, the experimental system included the probability of intraguild predation (IGP). The
development of the three species was studied in set-ups containing either one of the two predators or
both together with their prey. The IG predator F. ehrenbergii grew at a mean rate of 0.77 d1 independent of
the presence of the IG prey G. dominans. High grazing of the IG predator on the smaller IG prey was
detected in treatments containing only the two predators. However, when all three species were present,
the IG prey displayed significantly higher growth rates (0.42 d1) compared to treatments containing only
the IG prey as predator (0.32 d1). The results of further experiments allowed the exclusion of mechanical
or chemical signals induced by the IG predator being responsible for the observed increase in growth rate
of IG prey. Live observations revealed that the IG predator rejected a significant proportion of its S.
trochoidea catch after initial uptake. This behavior led to an immobilization of around 26% of the caught
cells. We tested if this prey immobilization by the IG predator facilitated prey uptake by the IG prey and
thus could be potentially responsible for the higher growth rates of the IG prey. Indeed, the smaller
predator selected positively for immobilized prey and reacted with higher grazing and growth rates.
Consequently, the IG prey benefitted from this commensalism between IG predator and IG prey and the
strength of this pattern predominated IGP in our model system. As both predators co-occur in the same
environment their feeding relationship could increase exploitation efficiency of common mobile prey items.
Furthermore, such commensalism potentially opens a loophole for a stable coexistence of MZP predators
despite their competition.
Key words: ciliates; commensalistic interaction; competition; dinoflagellates; Helgoland Roads; intraguild predation.
Received 1 February 2014; revised 29 April 2014; accepted 21 May 2014; final version received 9 September 2014;
published 27 October 2014. Corresponding Editor: K. L. Cottingham.
Copyright: Ó 2014 Lo¨der et al. This is an open-access article distributed under the terms of the Creative Commons
Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the
original author and source are credited. http://creativecommons.org/licenses/by/3.0/
E-mail: [email protected].
INTRODUCTION
ogy (Begon et al. 2006). The most obvious
interaction between two predators preying on
the same limiting resource is competition. Gause’s
law states that in such a case a stable coexistence
Interspecific interactions such as predation and
competition are of fundamental interest in ecolv www.esajournals.org
1
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
of the three players is not possible, because the
competitors are negatively affected by one another and the more competitive predator outcompetes the less competitive one (Gause 1934). In
reality, many competing organisms coexist, and
thus, communities appear to violate Gause’s law.
Phytoplankton species, for example, share a
seemingly homogeneous environment and the
same limiting resources (e.g., CO2, light and
nutrients), and thus competition and exclusion
of certain weaker competitors seems inevitable.
However, a huge number of phytoplankton
species coexist which led Hutchinson to coin the
term the ‘‘paradox of the plankton’’ (Hutchinson
1961). As there are also a huge number of
heterotrophic organisms that coexist in the
plankton the paradox can be expanded to this
component as well.
Several mutually non-exclusive solutions have
been brought forward to explain the plankton
paradox. Externally imposed variability in the
surrounding environment resulting in a system
which is never really homogeneous, can create
microhabitats with different competitive arenas
and different outcomes (Scheffer et al. 2003).
Ratios of nutrients might be considered to be
more important than just single nutrients, thus
creating many more potential niches and, as a
result, a higher potential for coexistence (Loladze
et al. 2004). Limiting similarity and interspecific
trade-offs concerning competitive ability, colonization ability and longevity can also enable
coexistence between competing species (Tilman
1994). Another very important factor explaining
the paradox is trophic interactions between
plankton species, potentially resulting in oscillatory cycles (Scheffer et al. 2003) or even chaotic
population patterns (Beninca et al. 2008) with,
e.g., more predators present than resources.
Moreover, mixotrophy is a common strategy in
plankton communities which enables the switching to additional nutritional sources (Sherr and
Sherr 2002). In short, there are many explanations, none of which alone is enough to explain
the high diversity of plankton. Nevertheless, the
fact that this high diversity is present suggests a
possible coexistence in the face of strong competition (Fox et al. 2010).
Since Azam (Azam et al. 1983) described the
‘‘microbial loop’’, microzooplankton (MZP) has
been recognized as an important structural and
v www.esajournals.org
functional group in planktonic ecosystems
(Landry and Calbet 2004). MZP can channel
carbon up the food chain as a trophic link from
smaller planktonic organisms to higher trophic
levels (Sherr et al. 1986, Gifford and Dagg 1988,
Stoecker and Capuzzo 1990), but as a result of
multiple steps within the microbial loop grazing
by MZP can also cause a reduction of the energy
transfer efficiency to higher trophic levels (Pomeroy et al. 2007, Franze and Modigh 2013). MZP
facilitates a rapid nutrient recycling back to the
primary producers (Calbet and Saiz 2005, Irigoien et al. 2005) and, due to its capability of a
fast population response to increasing phytoplankton availability (Johansson et al. 2004,
Aberle et al. 2007), MZP is considered as one of
the most important group of grazers in many
oceans. MZP can graze up to 60–75% of the daily
phytoplankton production (Landry and Calbet
2004), thus often exerting a stronger grazing
pressure than metazoan grazers such as copepods (Sherr and Sherr 2007).
MZP-phytoplankton interactions in the plankton are well-studied, mostly focusing on the
ecological implications of these interactions, for
instance, community structure and succession
patterns (Riegman et al. 1993, De Laender et al.
2010), the efficiency of energy transfer (Azam
1998, Sherr and Sherr 2002, Calbet and Landry
2004, Calbet 2008, Franze and Modigh 2013) or
the controlling potential of grazers on blooms
(Sherr and Sherr 2009, Lo¨der et al. 2011).
However, almost nothing is known about the
consequences of interspecific interactions between competing predators within the MZP.
Although MZP comprises a large variety of
taxonomic groups, the most important ones in
terms of abundance, biomass and grazing impact
are heterotrophic dinoflagellates and ciliates
(Capriulo et al. 1991, Lo¨der et al. 2012). As a
result of their preferred prey size, small heterotrophic dinoflagellates potentially compete with
larger ciliates for prey (Jakobsen and Hansen
1997). However, these dinoflagellates do not only
compete for food with large ciliates but they are
also potential prey. Polis and Holt (1992) coined
the term intraguild predation (IGP), to describe
this mixture of competition and predation within
a feeding guild. Here, we also use the term
‘‘guild’’ in a broad sense, i.e., for all taxa in a
community that compete for similar resources
2
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
regardless of different tactics in resource exploitation (Polis et al. 1989). IGP, i.e., consumption of
species that use the same limiting food source, is
more complex than either competition or predation alone. It is different from ‘‘classic’’ competition as there is an immediate energetic benefit
for the omnivorous player, the IG predator, and,
at the same time, IGP reduces potential competitors, the IG prey (Polis et al. 1989). However,
predation on smaller predators which are competitors for the same resource, so-called ‘‘closed
loop omnivory’’ (Sprules and Bowerman 1988) is
ubiquitous in the world’s terrestrial and aquatic
food webs (Polis et al. 1989). Investigations of this
phenomenon for MZP are rare (Franze and
Modigh 2013), although IGP has far-reaching
implications for energy transfer, plankton dynamics and consequently for the stability and
coexistence of marine food webs as a whole.
Therefore, it helps to explain the ‘‘Paradox of the
Plankton’’ phenomenon.
We aim to fill part of the gap of knowledge on
IGP in MZP by experimentally studying a system
with one IG predator, one IG prey and one prey
item as resource (Fig. 1). The competing species
used in our investigations, the large tintinnid
ciliate Favella ehrenbergii and the smaller, naked
heterotrophic dinoflagellate Gyrodinium dominans
co-occur in the North Sea MZP community,
(Lo¨der et al. 2012). Both predators prey on the
same type of prey. Furthermore, Favella ehrenbergii (IG predator) is a potential predator on the
smaller Gyrodinium dominans (IG prey). In this
study, we set out to investigate the mechanisms
that explain their coexistence, using the small
autotrophic dinoflagellate Scrippsiella trochoidea
as shared resource. A theoretical analysis of the
consequences for coexistence is published elsewhere (Shchekinova et al. 2014a, b). Here, we
addressed the following questions: (1) is there
any kind of measureable interaction between the
predators, and if so, (2) what are the mechanisms
that drive the interaction and (3) how can this
interaction be categorized?
MATERIAL
AND
of the three species in different combinations,
and showed that—contrary to expectations—the
IG prey seemed to benefit from the presence of
the IG predator. In the following experiments we
investigated the mechanisms that are potentially
responsible for this beneficial interaction. First,
we investigated the grazing rates of the IG prey
in the presence and absence of the IG predator
(Exp. II, ‘‘Enhanced IG prey feeding’’ experiment)
showing that indeed the IG prey consumed more
prey in the presence of the predator. Subsequently, we investigated whether this response was
related to chemical interactions (Exp. III), or
physical interactions (Exp. IV).
Cultures
The experiments were conducted with cultures
of the heterotrophic dinoflagellate Gyrodinium
dominans (;30 lm length), the tintinnid ciliate
Favella ehrenbergii (;160 lm length) and their
phototrophic dinoflagellate resource Scrippsiella
trochoidea (;20 lm length). To facilitate reading
the three species are named according to their
IGP role in the following text: F. ehrenbergii ¼ IG
predator, G. dominans ¼ IG prey, S. trochoidea ¼
resource. The cultures were obtained from
isolates of 20 lm net samples from Helgoland
Roads, Germany (54811.3 0 N; 7854.0 0 E). The
resource organism was isolated in May 2007 and
grown in 73.5-mL tissue culture bottles containing F/2 medium (Guillard and Ryther 1962) at
14.58C and an illumination of ;40 lmol photosynthetically active radiation (PAR) at constant
light conditions. Cultures of IG prey (July 2007)
and IG predator (October 2007) were established
in 73.5-mL tissue culture bottles containing F/2
medium and the resource. Both predator species
were maintained on a plankton wheel (1.1 rpm)
at 14.58C and constant illumination of ;30 lmol
PAR. Cultures without contaminations were
established by repeated re-isolation. Culture
bottles, food and media were renewed weekly.
As the first IG predator culture died after the first
experiment a new culture was established as
described above in August 2009. This culture was
used in the later experiments. Cell volumes of the
three species were estimated from the linear
dimensions of acid Lugol’s solution fixed cells
(2% final conc., n ¼ 43–50) assuming geometric
shapes according to Hillebrand et al. (1999). Cell
volume was converted to wet weight assuming a
METHODS
To investigate the interactions between the
three species we carried out a suite of experiments all building onto each other. The first
experiment (Exp. I) tested the numerical response
v www.esajournals.org
3
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Fig. 1. Conceptual scheme of the IGP relationship that was investigated in this study with the three species
Favella ehrenbergii (IG predator), Gyrodinium dominans (IG prey) and Scrippsiella trochoidea (resource). Arrows
indicate the direction of predation. IG predator and IG prey feed on the resource and the IG predator additionally
feeds on the IG prey.
(131 cells mL1).
Predation of the IG predator on the IG prey
was investigated in an additional approach. Only
IG prey (840 cells mL1) was offered to the IG
predator, yielding the same biomass as the
resource in the other experiments. Treatments
comprising single species served as controls.
Cultures were maintained in 73.5-mL tissue
culture bottles containing F/2 medium under
incubation conditions as described for the predator cultures above. Predator cultures used for
the experiment were not starved for more than
one day prior to the experiment.
This first experiment was carried out over 72
hours as a batch culture experiment. To realize an
independent design with time as an independent
factor, four replicates per treatment and sampling time point were set up for incubation and
the total volume of each of the four replicate
incubation bottles was harvested and fixed with
acid Lugol’s solution (2% final conc.) for the
specific gravity of 1 and is hereafter referred to as
‘‘biomass’’.
Exp. I: general patterns of interaction
This first experiment aimed at investigating
general patterns of interaction within the experimental system IG predator–IG prey–resource.
We thus determined population growth and
grazing rates of both predators on the resource
in single predator treatments and in a combined
predator treatment. In the single predator treatments the same biomass of either the IG predator
(1.2 cells mL1) or the IG prey (270 cells mL1)
was added to a starting resource concentration of
;590 cells mL1 which is roughly equivalent to
bloom concentrations of the resource organism
(Qi et al. 2004). In the combined predator
treatments the overall predator biomass was the
same as in the single predator treatments
comprising of roughly equal biomass proportions of IG predator (0.6 cells mL1) and IG prey
v www.esajournals.org
4
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
determination of cell concentrations after 0, 24, 48
and 72 hours of incubation.
Samples were enumerated under a Zeiss
Axiovert 135 inverted microscope using the
Utermo¨hl method (Utermo¨hl 1958). The samples
were settled overnight and 4–6 transects of each
settling chamber were counted at 200- or 400-fold
magnification depending on cell concentrations
of resource and IG prey. For the larger IG
predator the whole chamber was enumerated.
This counting procedure was used for all other
experiments conducted in this study.
Growth rates k [d1] of IG predator, IG prey
and resource without grazers were calculated
assuming exponential growth (Eq. 1, g ¼ 0).
Grazing rates g [d1] (Eq. 1), filtration rates F [mL
(predator cell)1 d1] (Eq. 2) and ingestion rates I
[prey cells (predator cell)1 d1] (Eq. 3) of the
predators in the single and combined predator
treatments were computed for each sampling
interval with average prey concentrations [Cprey]
(during each time interval) (Eq. 4 ) according to
Frost (1972) with the modification of Heinbokel
(1978) for the growth of predators [Cpred] (Eq. 5).
(The equations below were also the basis for the
calculations of the parameters investigated in the
latter experiments.)
Ct24 ¼ Ct0 3 eðkgÞDt
ð1Þ
F ¼ g 3½Cpred 1
ð2Þ
I ¼ F 3½Cprey ð3Þ
½Cprey ¼
½Cpred ¼
Ct0 3ðeðkgÞDt 1Þ
ðk gÞDt
Cpred;t24 Cpred;t0
lnCpred;t24 lnCpred;t0
treatment as main factors at significance levels of
0.05. Grazing rates were log-transformed to reach
homogeneity of variance before statistical analysis. Differences between groups were analyzed
by Tukey’s HSD (honestly significant difference)
post hoc tests. Due to extremely high grazing in
the combined predator treatment no remaining
resource cells were detected in three of four
replicates after 72 hours of incubation and thus, it
was possible to calculate grazing rates for one
replicate only in this case. Nevertheless, we
included this single measurement into our
statistical analysis as grazing rates in the other
three replicates were not measureable but definitely higher.
We maintained start concentrations, incubation
conditions and replication of this first experiment
in the following experiments if not explicitly
stated otherwise to avoid density-dependent
differences and allow for comparability between
the different experiments.
Contrary to our expectation the smaller IG
prey showed higher growth rates in the presence
of the IG predator during the first experiment
(see results). Thus, we put the main focus of the
subsequent experiments on the investigation and
explanation of this observation.
Exp. II: ‘‘Enhanced IG prey feeding’’ experiment
The second experiment was designed to test
the robustness of the results of the first experiment. We investigated if the presence of a newly
isolated different IG predator culture again leads
to a significantly higher growth rate of the IG
prey in a combined predator treatment. We
incubated four replicates of IG predator, IG prey
and the resource together for 24 hours under
conditions as described above. Bottles containing
only the resource S. trochoidea together with the
IG prey served as controls. We determined
growth and grazing parameters as described
above. Due to the possibility of IGP (i.e., IG prey
could also be prey items for IG predator) it was
not possible to directly disentangle the individual
grazing rates of each predator species in our IGP
model system. We therefore focused on the food
vacuole content of the IG prey as a proxy for
ingested prey. For this purpose we recorded the
IG prey cells with at least one filled food vacuole
separately (indicated by the dark color of the
vacuoles).
ð4Þ
ð5Þ
where Ct0 is the concentration of cells at the
beginning of each sampling interval (0, 24, 48, 72
h), Ct24 after the following 24 hours of incubation,
k is the growth rate, g the grazing rate and Dt the
incubation time in days.
Statistical analyses were conducted with the
software ‘‘Statistica 7.1’’ (StatSoft). After testing
for normality (Shapiro-Wilk test) and homogeneity of variances (Levene’s test) data on growth
and grazing parameters of the predators were
analyzed using two-way ANOVAs with time and
v www.esajournals.org
5
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
growing IG predator culture (15 cells mL1) was
filtered over 0.2-lm nylon filters (Falcon). Ten
milliliters of filtrate, equalling a final IG predator
concentration of 2 cells mL1 in the treatment,
was added to four replicate incubation bottles
containing IG prey and resource. Controls
received 10 mL of F/2 filtrate. Growth and
grazing parameters were investigated as described above.
We also investigated the swimming behavior
and velocity of the resource organisms and the
IG prey in the presence and absence of compounds released by the IG predator in this
experiment. After exposure to the filtrate for 24
hours under conditions comparable to the first
two experiments, resource and IG prey cells were
filmed as described before. Samples for cell
concentrations were fixed at the start and
immediately after filming. Statistical analyses
were conducted in the same way as for the
‘‘enhanced IG prey feeding’’ experiment (compare Exp. II).
As the enhancement of swimming speed can
increase predator-prey encounter rates (Gerritsen
and Strickler 1977) and thereby food uptake, thus
potentially promoting higher grazing rates, we
also investigated the swimming behavior and
velocity of the resource organism S. trochoidea and
the IG prey in the presence and absence of the IG
predator. For this purpose, resource and IG prey
cells were filmed under a stereo microscope
(SZX16, Olympus) at 50-fold magnification for
10 seconds at a rate of 15 frames per second after
incubation for 24 hours. Samples for cell concentrations were fixed as described above at the start
and immediately after filming. The films for the
investigations on swimming velocity and behavior were converted into single frame pictures
using the freeware program ‘‘Avi4Bmp’’ (Bottomap Software). The timeframe most suitable for
our investigation was two seconds as most of the
organism tracks did not go beyond the frame and
tracks rarely crossed one another during this time.
Thirty pictures (equalling two seconds of each
original film) were stacked with the function
‘‘overlay frames’’ of the freeware program ‘‘Trace’’
(Heribert Cypionka, 2000–2010). The resulting
images showed the swimming paths of the cells
(Fischer and Cypionka 2006). The lengths of the
swimming tracks of 60 cells per treatment (four
replicates, 15 per replicate, randomly chosen) in
the stacked pictures were measured with the open
source software ‘‘ImageJ’’ (downloaded from
http://rsbweb.nih.gov/ij/index.html). Swimming
velocity [lm s1] of the cells was then calculated
by dividing path length by the time needed. Mean
values of the 15 measurements per replicate were
used for statistics. Changes in swimming behavior
were analyzed by visually comparing the patterns
of the swimming paths in the stacked pictures.
After testing for normality (Shapiro-Wilk test)
and homogeneity of variances (Brown-Forsythe
test) food vacuole content of IG prey and
swimming speed of IG prey and resource were
tested for significant differences between treatment and control with Statistica 7.1 using twotailed t-tests at significance levels of 0.05.
Exp. IV: Experiments on the precondition
of the resource organism S. trochoidea
By observation of a freshly fed IG predator
culture we found that not every resource cell
captured by the IG predator was actually
ingested. A considerable number of cells were
rejected resulting in immobilization after this
‘‘manipulation’’ by the IG predator (compare
video in the Appendix). This led to the hypothesis that the IG predator promotes the IG prey by
immobilizing the potential prey for the smaller
predator, thus making it easier to catch.
Prey immobilization rate of the IG predator Favella
ehrenbergii.—The rate of resource cells not eaten,
but immobilized by the IG predator, was
experimentally determined after 24 and 48 hours
of incubation. For each time step four replicate
bottles were incubated under conditions as
described above. Bottles containing only the
resource without the IG predator served as
controls. Differentiation between mobile and
immobile resource cells in fixed samples is not
possible, thus, immobile cells were allowed to
settle in the experimental bottles for 15 minutes
after incubation. This timeframe was sufficient
for a complete sedimentation of the immobile
fraction of resource cells. To discriminate between mobile and immobile cells, a grid was
Exp. III: Experiment on chemical stimulation
and swimming behavior
To test whether chemical compounds excreted
by the IG predator influenced the grazing and
growth of the IG prey, a well-fed, exponentially
v www.esajournals.org
6
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
defined and ten sites of the bottom surface of
each experimental bottle were filmed for 5
seconds under conditions as described above.
The movements of the resource cells were also
evaluated by filming as described for the
‘‘enhanced IG prey feeding’’ experiment. Cells
that did not change their position during the film
were counted as immobile. Samples for the
determination of total cell concentrations were
fixed immediately after filming. The concentration of immobile cells was calculated via the
extrapolation of the immobile cells in the filmed
area to the whole bottom surface and thereafter
to bottle volume. We calculated the immobilization rate of the IG predator per day [cells
immobilized predator1 day1] by dividing the
mean immobile S. trochoidea concentration by the
mean IG predator concentration during the time
of incubation (Frost 1972, Heinbokel 1978). The
percentage of immobilization of cells caught by
the tintinnid was also calculated. Differences in
the percentage of immobile cells between treatment and control replicates were tested for
significance applying two-way ANOVAs with
time and treatment as main factors at significance
levels of 0.05 as described above (compare Exp.
I).
Growth and grazing response of the IG prey
Gyrodinium dominans on immobilized prey.—To test
if the IG prey really benefits from immobile
resource cells we investigated its growth when
fed with the artificially immobilized resource
organism. Cells were immobilized by sonication
in an ultrasound bath in six cycles lasting three
minutes each. The ratio of immobile cells was
determined via films before and after incubation
for 24 hours as described above. We measured
growth and grazing (calculated as described
above) as well as selectivity for mobile or
immobile resource of the IG prey after incubation
of four replicate bottles under the same conditions as in the previous experiments for 24 hours
with sonicated resource cells. Incubation bottles
with the IG prey and untreated resource cultures
served as controls. Samples for cell counts were
fixed with acid Lugol’s solution immediately
before and after incubation.
The prey selectivity index ‘‘a’’ of the IG prey
preying on a population of S. trochoidea containing both mobile and immobile cells was calculated for each prey type according to Chesson
v www.esajournals.org
(1978, 1983). We chose Chesson’s case 1 equation
(prey population assumed to be constant; Eq. 6)
(Chesson 1983) because our values of ingestion
and percentage of prey in the environment were
obtained by averaged prey concentrations and a
strong depletion of food was not observed
during our selectivity experiment.
ri =ni
ai ¼ X
m
rj =nj
ð6Þ
j¼1
where ri is the frequency of prey i (mobile or
immobile resource) in the diet of the IG prey
and n i is the frequency of prey in the
environment, divided by the sum of the
relationships between the frequency of mobile
and immobile resource in the diet and in the
environment.
Values of a were then used to calculate the
electivity index E* (Eq. 7) according to Vanderploeg and Scavia (1979a, b).Values of E* range
from 1 to 1. E* values of 0 indicate nonselective
feeding, values . 0 indicate preference, values ,
0 indicate discrimination against a prey type.
E ¼
ai 1n
ai þ 1n
ð7Þ
where n is the total number of prey types.
Differences in growth, grazing parameters and
selectivity between treatment and control replicates were tested for significance applying t-tests
as described above (compare Exp. II).
RESULTS
Exp. I: General interaction patterns
within the IGP system
The first experiment was designed to investigate general growth and grazing patterns of both
predators in single predator treatments, the
grazing impact of the IG predator on the IG
prey and finally to investigate if there is any kind
of measurable interaction between IG predator
and IG prey that affects population dynamics in
the IGP system. Thus, we followed population
development of the IG predator, the IG prey and
their common resource in treatments with one
predator and the resource, IG predator and IG
prey only and treatments with both predator
species and the resource present.
7
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Fig. 2. General development of the IG prey Gyrodinium dominans and the IG predator Favella ehrenbergii as well
as their joint prey resource Scrippsiella trochoidea [n mL1] in single predator (a, b, c) and combined predator
treatments (d) during the time of incubation. Error bars correspond to one standard deviation, n ¼ 4.
of 0.32 d1 during the three days of incubation
and its growth rate continuously increased from
0.21 to 0.42 d1 (Fig. 3a). Its ingestion decreased
from 0.91 to 0.45 cells predator1 d1 (mean 0.65
Single predator treatments.—In the single predator treatments both predators displayed positive
growth rates throughout the experiment (Fig.
2a, b). The IG prey (Fig. 2a) grew at a mean rate
v www.esajournals.org
8
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Fig. 3. Growth rates [d1] (a: IG prey Gyrodinium dominans, statistics given in Tables 9 and 10; b: IG predator
Favella ehrenbergii, statistics given in Tables 7 and 8) and (c) grazing rates [d1] of both predators preying on the
resource Scrippsiella trochoidea in single predator and combined predator treatments after 24, 48 and 72 hours of
incubation; calculated for 24 hour intervals (statistics given in Tables 3, 4, 5 and 6). Error bars correspond to one
standard deviation, n ¼ 4. Significant differences between single predator treatments and combined predator
treatment at each sampling point are marked by ‘‘*a,’’ differences between single predator treatments by ‘‘*b’’
(two-way ANOVAs and Tukey HSD post hoc tests).
v www.esajournals.org
9
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Table 1. Comparison of the growth rates of the IG predator and the IG prey in the single predator treatments.
Results of the two-way ANOVA with time and treatment as main factors. Significant results given in boldface.
Effect
SS
df
MS
F
p
Constant
Treatment
Time
Treatment 3 Time
Error
6.9396
1.2156
0.4144
0.0552
0.2101
1
1
2
2
18
6.9396
1.2156
0.2072
0.0276
0.0117
594.4657
104.1351
17.7479
2.3635
0.0000
0.0000
0.0001
0.1226
Note: Abbreviations are: SS, sum of squares; df, degree of freedom; MS, mean squares; F, F value; p, p value.
cells predator1 d1) during the experiment. With
a mean grazing rate of 0.48 d1 (0.44–0.57 d1)
the consumption rate of the IG prey roughly
averaged the production rate of the resource
(mean growth rate 0.45 d1) and the predator
was not able to graze down its prey substantially
during the experiment.
At each sampling point the IG predator (Fig.
2b) had grown faster in the single predator
treatments when compared to the IG prey (Fig.
3a, b, Tables 1 and 2). Mean growth rates of the
tintinnid increased constantly from 0.54 to 0.98
d1 and were on average 0.77 d1. Ingestion rates
dropped from 165 to 86 cells predator1 d1 and
averaged 64 cells predator1 d1 during the
course of the experiment. Within the first two
days, similar to the IG prey, the grazing rates of
IG predator hardly exceeded the growth rates of
the resource (0.43 and 0.51 d1) and we found no
difference between the grazing rates in both
single predator treatments (Fig. 3c, Tables 3 and
4). During the last day of the experiment and as a
result of the fast growth of the predator
population, the grazing rate of IG predator
increased to 2.55 d1 and led to a sharp decline
in the prey population and a significantly higher
grazing rate compared to the IG prey (Tables 3
and 4).
Despite high initial ingestion rates the IG
predator was not able to increase its growth rate
in treatments with the IG prey as the sole prey for
the IG predator as it had increased with the
resource S. trochoidea (Fig. 2c). The IG predator
displayed an average growth rate of 0.32 d1
which declined from 0.50 to 0.37 to 0.08 d1
during the experiment. The IG predator initially
ingested 114 and 111 IG prey cells predator1 d1
on the second day but then ingestion dropped to
5 cells predator1 d1 (mean 69 cells predator1
d1). A mean grazing rate of 0.20 d1 of the IG
predator on the IG prey was measured during
the experiment. After a slight increase of the
grazing rate after the second day (0.21 to 0.34
d1), it decreased to 0.04 d1 on the last day of the
experiment although the IG prey was always
available at high concentrations.
Combined predator treatment.—While IG predator and IG prey in the single predator treatments
initially consumed roughly only the daily production of the resource, a completely different
picture was seen in the combined predator
treatments (Figs. 2d, 3c): Although containing
the same initial predator biomass, the resource
biomass already decreased after 24 hours and
was almost completely grazed down after 72
hours by the end of the experiment. Consequent-
Table 2. Comparison of the growth rates of the IG predator and the IG prey in the single predator treatments.
Results of the Tukey’s HSD (honestly significant difference) post hoc test (p-values). Significant results given in
boldface.
IG predator
Treatment
IG predator
IG prey
Time
24
48
72
24
48
72
h
h
h
h
h
h
IG prey
24 h
48 h
72 h
24 h
48 h
72 h
...
0.0840
0.0004
0.0049
0.0598
0.5933
...
0.0964
0.0002
0.0003
0.0031
...
0.0002
0.0002
0.0002
...
0.8237
0.1289
...
0.6910
...
v www.esajournals.org
10
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Table 3. Comparison of the grazing rates of the IG predator and the IG prey in the single predator treatments.
Results of the two-way ANOVA with time and treatment as main factors. Significant results given in boldface.
Effect
SS
df
MS
F
p
Constant
Treatment
Time
Treatment 3 Time
Error
0.9956
0.3220
0.6269
0.8275
0.0670
1
1
2
2
18
0.9956
0.3220
0.3134
0.4138
0.0037
267.6181
86.5425
84.2492
111.2169
0.0000
0.0000
0.0000
0.0000
Note: Abbreviations are: SS, sum of squares; df, degree of freedom; MS, mean squares; F, F value; p, p value.
Table 4. Comparison of the grazing rates of the IG predator and the IG prey in the single predator treatments.
Results of the Tukey’s HSD (honestly significant difference) post hoc test (p-values). Significant results given in
boldface.
IG predator
Treatment
Time
IG predator
24
48
72
24
48
72
IG prey
h
h
h
h
h
h
IG prey
24 h
48 h
72 h
24 h
48 h
72 h
...
0.5320
0.0002
0.9993
0.1013
1.0000
...
0.0002
0.7367
0.8823
0.6457
...
0.0002
0.0002
0.0002
...
0.1869
1.0000
...
0.1428
...
ly, significant differences due to the consistently
higher grazing rates in the combined predator
treatments were found compared to those measured in the single predator treatments at the
same time (Fig. 3c, Tables 5 and 6).
No difference in the growth rates was observed for the IG predator between the single
and combined predator treatments when looking
at individual sampling days (Fig. 3b, Tables 7
and 8). This was also reflected in the mean
growth rate of 0.77 d1 over 72 hours of
incubation in both cases. A completely different
picture was observed in case of the IG prey. It
always showed significantly different daily
growth rates (Fig. 3a) in the combined predator
treatments when compared to the single predator
treatments (Tables 9 and 10). Even if predation of
the IG predator on the IG prey was shown in
treatments containing only both predators (Fig.
2c), growth rates of the IG prey (mean 0.58 and
0.66 d1) were twice as high in the combined
predator treatments compared to the single
predator treatments during the first 48 hours.
During the last 24 hours of the experiment
growth of IG prey dropped to a mean value of
0.01 d1 along with the complete disappearance
of the resource (Fig. 3a).
The higher growth rates of the IG prey when
incubated in the IGP system with a shared
resource and an IG predator were a first sign
for an interaction between both predators. Thus,
the subsequent experiments focused on the
mechanisms behind these observed patterns.
Table 5. Comparison of the grazing rates of the IG predator and the IG prey in the single predator treatments and
the combined predator treatment. Results of the two-way ANOVA with time and treatment as main factors.
Significant results given in boldface.
Effect
SS
df
MS
F
p
Constant
Treatment
Time
Treatment 3 Time
Error
0.0578
1.3922
1.3307
1.1789
0.0809
1
2
2
4
24
0.0578
0.6961
0.6653
0.2947
0.0034
17.1431
206.3950
197.2713
87.3888
0.0004
0.0000
0.0000
0.0000
Note: Abbreviations are: SS, sum of squares; df, degree of freedom; MS, mean squares; F, F value; p, p value.
v www.esajournals.org
11
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Table 6. Comparison of the grazing rates of the IG predator and the IG prey in the single predator treatments and
the combined predator treatment. Results of the Tukey’s HSD (honestly significant difference) post hoc test (pvalues). Significant results given in boldface.
IG predator
Treatment
Time
IG predator
24
48
72
24
48
72
24
48
72
IG prey
Both predators
h
h
h
h
h
h
h
h
h
IG prey
Both predators
24 h
48 h
72 h
24 h
48 h
72 h
24 h
48 h
72 h
...
0.6726
0.0002
1.0000
0.1255
1.0000
0.0462
0.0002
0.0002
...
0.0002
0.8662
0.9622
0.7873
0.7845
0.0002
0.0002
...
0.0002
0.0002
0.0002
0.0002
0.2835
0.0008
...
0.2426
1.0000
0.0988
0.0002
0.0002
...
0.1821
0.9999
0.0002
0.0002
...
0.0706
0.0002
0.0002
...
0.0002
0.0002
...
0.0002
...
Table 7. Comparison of the growth rates of the IG predator in the single predator treatment and the combined
predator treatment. Results of the two-way ANOVA with time and treatment as main factors. Significant
results given in boldface.
Effect
SS
df
MS
F
p
Constant
Treatment
Time
Treatment 3 Time
Error
13.9752
0.0000
0.3793
0.0797
0.3728
1
1
2
2
18
13.9752
0.0000
0.1897
0.0399
0.0207
674.8546
0.0001
9.1582
1.9251
0.0000
0.9919
0.0018
0.1747
Note: Abbreviations are: SS, sum of squares; df, degree of freedom; MS, mean squares; F, F value; p, p value.
Table 8. Comparison of the growth rates of the IG predator in the single predator treatment and the combined
predator treatment. Results of the Tukey’s HSD (honestly significant difference) post hoc test (p-values).
Significant results given in boldface.
IG predator
Treatment
Time
IG predator
24
48
72
24
48
72
Both predators
h
h
h
h
h
h
Both predators
24 h
48 h
72 h
24 h
48 h
72 h
...
0.2916
0.0048
0.6236
0.5583
0.0427
...
0.3163
0.9890
0.9955
0.8857
...
0.1154
0.1402
0.8919
...
1.0000
0.5593
...
0.6245
...
Table 9. Comparison of the growth rates of the IG prey in the single predator treatment and the combined
predator treatment. Results of the two-way ANOVA with time and treatment as main factor. Significant results
given in boldface.
Effect
SS
df
MS
F
p
Constant
Treatment
Time
Treatment 3 Time
Error
3.2324
0.0708
0.3024
0.7874
0.2109
1
1
2
2
18
3.2324
0.0708
0.1512
0.3937
0.0117
275.9155
6.0463
12.9061
33.6077
0.0000
0.0243
0.0003
0.0000
Note: Abbreviations are: SS, sum of squares; df, degree of freedom; MS, mean squares; F, F value; p, p value.
v www.esajournals.org
12
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Table 10. Comparison of the growth rates of the IG prey in the single predator treatment and the combined
predator treatment. Results of the Tukey’s HSD (honestly significant difference) post hoc test (p-values).
Significant results given in boldface.
IG prey
Treatment
IG prey
Both predators
Time
24
48
72
24
48
72
h
h
h
h
h
h
Both predators
24 h
48 h
72 h
24 h
48 h
72 h
...
0.8247
0.1300
0.0016
0.0003
0.1479
...
0.6925
0.0198
0.0023
0.0137
...
0.2944
0.0460
0.0007
...
0.8970
0.0002
...
0.0002
...
Fig. 4. (a) Percentage of the IG prey Gyrodinium dominans population containing food vacuoles in presence or
absence (control) of the IG predator Favella ehrenbergii before and after incubation for 24 hours. (b) Percentage of
the resource Scrippsiella trochoidea population being immobile in presence or absence (control) of the IG predator
after incubation for 24 and 48 hours (statistics given in Tables 13 and 14). Error bars correspond to one standard
deviation, n ¼ 4. Significant differences between treatments and the controls are marked by an asterisk (a: twotailed t-tests; b: two-way ANOVA and Tukey HSD post hoc test).
v www.esajournals.org
13
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Fig. 5. Swimming speed [lm s1] of the IG prey Gyrodinium dominans and the resource Scrippsiella trochoidea in
the controls (both species present and IG predator absent) and in the treatments where the IG predator Favella
ehrenbergii was present. Significant differences between the species are marked by an asterisk (two-tailed t-tests).
significant effect on the ingestion by the IG prey
as determined by the percentage of individuals
containing food vacuoles after 24 hours of
incubation (t ¼ 6.42, df ¼ 6, p , 0.001, two-tailed
t-test; Fig. 4a). While the share of individuals
containing food vacuoles increased only slightly
from 12% 6 2% to 18% 6 1% in the absence of
the IG predator, this proportion increased from
13% 6 2% to 46% 6 9% when the IG predator
was present (ranges represent standard deviations). As we detected a mortality rate of
approximately 0.22 d1 in the presence of the
IG predator, indicating predation on the smaller
IG prey, higher shares of food vacuole-containing
cells could have resulted from selective predation
of the IG predator on IG prey without vacuole
contents. This would have artificially increased
the share of vacuole-containing IG prey cells in
the population. However, this hypothesis could
be rejected as there were significantly higher total
concentrations of vacuole-containing cells in
treatments where the IG predator was present
when compared to the controls (t ¼ 9.89, df ¼ 6, p
, 0.0001, two-tailed t-test).
Exp. II: ‘‘Enhanced IG prey feeding’’ experiment
Our hypothesis was that the mechanism
behind the previously observed higher growth
of IG prey was an enhanced food uptake by the
IG prey. Thus, we investigated if the presence of a
different isolate of the IG predator also produced
the observed positive effect in growth of IG prey
and, consequently we also investigated food
uptake. Due to the possibility of IGP (i.e., IG
prey could also be prey items for IG predator), it
was not possible to directly disentangle the
individual predation impact of each predator
species in our IGP model system. We therefore
focused on the food vacuole content of the IG
prey as a proxy for ingested prey by the predator
(Fig. 4a).
Furthermore, we accounted for swimming
behavior and velocity of the resource and the
IG prey when exposed directly to the IG
predator. Our results showed no differences
between the controls and treatments with the
IG predator when looking at the swimming
velocity of resource and IG prey (Fig. 5). There
were also no detectable changes in swimming
patterns. However, swimming velocity differed
significantly between species (means of all
treatments with and without IG predator: IG
prey: 167 lm s1, resource: 361 lm s1, t ¼ 30.59,
df ¼ 14, p , 0.0001, two-tailed t-test; Fig. 5).
The presence of the IG predator had a strongly
v www.esajournals.org
Exp. III: Experiment on chemical stimulation
and swimming behavior
A possible mechanism behind the observed
enhanced food uptake by the IG prey was
stimulation by chemical compounds released by
14
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Fig. 6. Growth rates [d1] and ingestion rates [prey cells predator1 d1] of the IG prey Gyrodinium dominans
preying on Scrippsiella trochoidea after exposure for 24 hours to a filtrate of the IG predator Favella ehrenbergii (a) or
to artificially immobilized prey (b). Error bars correspond to one standard deviation, n ¼ 4. Significant differences
between treatments and the controls are marked by an asterisk (two-tailed t-tests).
the IG predator. To test this, we exposed the IG
prey for 24 hours to a filtrate of culture medium
of exponentially growing IG predators in the
presence of the resource. Furthermore, swimming velocity and behavior were monitored.
The IG prey showed only a weak ingestion and
no difference in ingestion rates when comparing
the treatments that received filtrate (mean 0.06
prey cells predator1 d1) to the controls (mean
0.02 prey cells predator1 d1) (Fig. 6a) were
found. Growth rates of IG prey in treatments
with filtrate (mean: 0.04 d1) and in the controls
(mean: 0.1 d1) were lower than those observed
in the previous experiments without showing
v www.esajournals.org
any significant difference from each other (Fig.
6a).
We found no visible differences in the swimming patterns of both species, when examining
the paths of the cells after exposure to the filtrate
for 24 hours. In addition, swimming velocity did
not differ between treatments with and without
filtrate of the IG predator. However, it differed
again significantly between both species (means
of all treatments with and without filtrate: IG
prey: 177 lm s1, resource: 414 lm s1, t ¼ 30.58,
df ¼ 14, p , 0.0001, two-tailed t-test).
15
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Table 11. Comparison of the absolute amount of immobile resource cells in the treatment with the IG predator
and the control without any predator. Results of the two-way ANOVA with time and treatment as main factor.
Significant results given in boldface.
Effect
SS
df
MS
F
p
Constant
Treatment
Time
Treatment 3 Time
Error
108455.2699
29905.5670
2256.2632
2843.2232
2670.7311
1
1
1
1
12
108455.2699
29905.5670
2256.2632
2843.2232
222.5609
487.3060
134.3703
10.1377
12.7750
0.0000
0.0000
0.0079
0.0038
Note: Abbreviations are: SS, sum of squares; df, degree of freedom; MS, mean squares; F, F value; p, p value.
the total number of immobile cells and relative
shares of the total number of cells when
compared to the controls (Tables 11, 12, 13 and
14). We calculated a constant mean immobilization rate of 1.4 cells predator1 h1 for 24 as well
as 48 hours. Overall 26% 6 3% of the cells caught
by the tintinnid were rejected and immobilized
during the experiment. Although using a different culture, mean ingestion rates of the IG
predator were similar to the first experiment
(after 24 hours: 102 cells predator1 d1; after 48
hours: 90 cells predator1 d1).
Experiment on ingestion, growth and selective
response of Gyrodinium dominans to artificially
immobilized prey.—Here we tested the hypothesis
that immobilized resource cells are easier to catch
and preferred, thus leading to a higher ingestion
of food by the IG prey (ranges in this paragraph
represent standard deviations).
The ultrasound treatment resulted in an
immobilization of 57% 6 8% of the prey cells at
the beginning of the experiment. After 24 hours,
22% 6 2% were still immobile. The shares of
immobile prey in the controls were 7% 6 1% at
the start and 3% 6 1% after 24 hours, respectively.
We measured a mean ingestion rate of the IG
prey of only 0.02 prey cells predator1 d1 in the
controls and this differed significantly by a factor
Exp. IV: Experiments on the precondition
of the resource S. trochoidea
We recorded that the swimming speed of the
resource was more than twice as high as of the IG
prey. Furthermore, our observation showed the
immobilization of resource cells after catch,
‘‘manipulation’’ and rejection by the IG predator.
These two observations led to the hypothesis that
the IG predator promotes the IG prey by
providing prey, which is easier to catch for the
smaller predator. The experiments on the precondition of the resource focused on this hypothesis.
Determination of prey immobilization by Favella
ehrenbergii.—Here, we directly investigated the
prey immobilization process by the IG predator
and determined the number of cells immobilized
by the IG predator (ranges in this paragraph
represent standard deviations).
During the time of incubation the amount of
immobile cells in the control treatments was 3%
6 1% after 24 hours and 8% 6 2% after 48 hours,
i.e., always below 10% of the population.
However, in the presence of the IG predator this
proportion was 24% 6 1% and 26% 6 2%,
respectively (Fig. 4b). The amount of immobile
prey cells in the presence of the tintinnid was
always significantly higher both with regard to
Table 12. Comparison of the absolute amount of immobile resource cells in the treatment with the IG predator
and the control without any predator. Results of the Tukey’s HSD (honestly significant difference) post hoc test
(p-values). Significant results given in boldface.
þ IG predator
Control
Treatment
Control
þ IG predator
Time
24
48
24
48
h
h
h
h
v www.esajournals.org
24 h
48 h
24 h
48 h
...
0.0023
0.0002
0.0002
...
0.0005
0.0007
...
0.9923
...
16
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Table 13. Comparison of the relative amount of immobile resource cells in the treatment with the IG predator and
the control without any predator. Results of the two-way ANOVA with time and treatment as main factor.
Significant results given in boldface.
Effect
SS
df
MS
F
p
Constant
Treatment
Time
Treatment 3 Time
Error
3578.8336
1511.4716
52.8575
12.8038
35.1692
1
1
1
1
12
3578.8336
1511.4716
52.8575
12.8038
2.9308
1221.1264
515.7261
18.0354
4.3688
0.0000
0.0000
0.0011
0.0586
Note: Abbreviations are: SS, sum of squares; df, degree of freedom; MS, mean squares; F, F value; p, p value.
of 20 from the immobilized prey treatment (t ¼
5.85, df ¼ 6, p , 0.01, two-tailed t-tests), where
we found an ingestion rate of 0.40 prey cells
predator1 d1 (Fig. 6b). This was also reflected
in the comparison of the percentage of IG prey
containing food vacuoles (t ¼ 19.80, df ¼ 6, p ,
0.0001, two-tailed t-test). Whereas only 15% 6
2% of the predator population contained visible
food vacuoles in the controls, 42% 6 2% of the
predator cells contained vacuoles when offered
artificially immobilized prey. The growth rates of
IG prey reflected the same pattern. While we
measured low growth rates in the control
treatments (mean: 0.10 d1), we detected significantly higher growth rates of the IG prey
preying on immobilized prey after 24 hours
(mean: 0.26 d1, t ¼ 4.17, df ¼ 6, p , 0.01, twotailed t-test) (Fig. 6b). Overall the findings were
similar to the results found in the previous
experiments when the IG predator was directly
present.
The finding of higher ingestion and increased
growth of the IG prey when fed on immobile
prey was supported by prey selectivity. Selectivity patterns for mobile and immobile prey were
significantly different (t ¼ 6.33, df ¼ 6, p , 0.001,
two-tailed t-test). IG prey selected for immobile
resource cells (E*: 0.20 6 0.06) and strongly
avoided mobile prey cells (E*: 0.34 6 0.16).
DISCUSSION
Interactions between IG predator and IG prey
We investigated the interactions between two
naturally competing MZP predator species in a
natural three species microbial IGP system. We
were able to show that (1) the presence of the IG
predator enhanced ingestion and consequently
population growth in the smaller IG prey, (2)
conversely, there was no effect of the IG prey on
the IG predator, (3) neither chemical nor mechanical signaling by the IG predator had an
effect on the IG prey, (4) the enhancement of
growth was caused by pre-conditioning of the
food for the IG prey via immobilization by the IG
predator. We conclude that this interaction may
contribute to the likelihood of coexistence of
these three species. Admittedly, our experiments
contained only one shared prey resource, whereas in the field there may be more prey organisms
available which would enable the switching to
alternative prey, weaken the competition and
thus further strengthen the possibility of coexistence of both predators (Holt and Huxel 2007).
Furthermore, the duration of the experiments
was relatively short and the experiments were
conducted as small volume batch experiments.
Under these conditions it is not possible to
investigate longer-term coexistence behavior of
Table 14. Comparison of the relative amount of immobile resource cells in the treatment with the IG predator and
the control without any predator. Results of the Tukey’s HSD (honestly significant difference) post hoc test (pvalues). Significant results given in boldface.
þ IG predator
Control
Treatment
Control
þ IG predator
Time
24
48
24
48
h
h
h
h
v www.esajournals.org
24 h
48 h
24 h
48 h
...
0.0037
0.0002
0.0002
...
0.0002
0.0002
...
0.4538
...
17
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
the IGP model system. Theoretical considerations
linked to our experiments (Shchekinova et al.
2014a, b) indicated that the region of coexistence
is increased as a result of the observed IGP
interactions, something that should be tested in
long-term chemostat experiments (Hiltunen et al.
2013).
show between-species exchange of digestive
enzymes e.g., via feeding on fecal pellets of
competitors (Brendelberger 1997, Aberle et al.
2005), such benefits from ‘‘pre-conditioned prey’’
have been reported for dinoflagellates before
(Poulsen and Iversen 2008). The IG prey selected
strongly for immobilized dinoflagellates in our
experiments even if mobile prey was available in
the same concentration (Fig. 7). This is most
probably related to the feeding habit of the IG
prey, G. dominans. This species always shows a
distinct pre-capture swimming behavior surrounding the prey before capture and ingestion
(Hansen 1992). Keeping in mind that the prey
organism S. trochoidea is more than twice as fast
as the IG prey G. dominans such a behavior makes
it clear that immobile prey cells are more easily
captured by the IG prey despite the lower
encounter rates compared to swimming prey
(Gerritsen and Strickler 1977). This was also
confirmed by our observations where G. dominans IG prey cultures grew best on old S.
trochoidea resource batch cultures that started to
form immobile cysts.
The mechanism that drives the interaction
There are different possible explanations for
the observed immobilization of the resource S.
trochoidea by the tintinnid IG predator during
feeding. One explanation could be a continued
capturing of prey by the IG predator after food
satiation leading to a release of the captured and
immobilized prey. However, Stoecker et al.
(1995) reported that Favella ehrenbergii stops
feeding until ingested material is digested and/
or egested. A very likely explanation of the
rejection is related to the size of the prey.
Tintinnids are able to ingest particles up to 41–
45% of the oral diameter of their lorica (Spittler
1973, Heinbokel 1978). It is thus possible that the
rejected cells were generally too large to be
ingested and were rendered immobile after
damage of their flagellar apparatus. However, a
direct link between the relatively high immobilization rates by the tintinnid and extraordinarily
large cells of S. trochoidea is unlikely as such cells
occurred only sporadically in our experiments.
However, S. trochoidea has an ovoid shape which
implies that cells have a ‘‘long’’ side (longitudinal
axis) and a ‘‘short’’ side (transversal axis). The
longitudinal length of S. trochoidea was close to
the oral size restriction (Spittler 1973, Heinbokel
1978) of F. ehrenbergii in our experiments. If the
predator encountered its prey in longitudinal
orientation ingestion was not possible simply
because the cells were too big. And if the
predator was not able to turn the cell around
using its oral cilia the consequence had to be
rejection. The manipulation with the oral cilia
certainly also leads to the damage of the flagella
of the prey dinoflagellate. Our video observation
suggests such a behavior by the IG predator F.
ehrenbergii resulting in the immobilization of prey
after catch and release (compare video in the
Appendix).
Categorization of the interaction:
IGP relationship with a commensalistic element
Our findings add another element to other
studies which found no differences in growth
rates for a dinoflagellate and its potential ciliate
predator competing for the same prey when
combined and single predator treatments were
compared (Jakobsen and Hansen 1997) or where
IGP between IG predator and IG prey favored
their common resource (Stoecker and Evans
1985). In our experiments both predators also
competed for the same prey organism but the IG
prey was directly supported by the prey-immobilizing ‘feeding’ behavior of the IG predator.
Commensal interactions are characterized by
an immediate gain for one partner in the
commensalistic relationship whereas the other
partner is neither harmed nor benefited (Begon et
al. 2006). We showed that the immobilization of
autotrophic prey by the tintinnid IG predator
facilitates ingestion of the dinoflagellate IG prey
(Fig. 7). This happened without apparent positive or negative effects for the IG predator. Their
interaction is thus unidirectional facilitative and
can therefore be categorized as a competitive
predator relationship with a commensalistic
‘‘Immobile benefits’’ for the IG prey
Similar to benthic grazers which are known to
v www.esajournals.org
18
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Fig. 7. Conceptual scheme according to the results of this study showing the commensalistic IGP relationship
of the species Favella ehrenbergii (IG predator) and Gyrodinium dominans (IG prey) preying on Scrippsiella trochoidea
(resource). Black arrows indicate the direction of predation, the thickness of the arrows indicates a preference for
a prey type (IG prey). The red arrow represents the immobilization process of Scrippsiella trochoidea by Favella
ehrenbergii as basis for the commensalistic relationship of IG predator and IG prey.
element.
commensalism. This could result in a faster
growth of MZP predator biomass and thus an
increased grazing impact with potential consequences for phytoplankton standing stock. Thus,
increased predator biomass could strengthen the
microbial loop. A stronger microbial loop could
on the one hand reduce the efficiency of
channeling carbon to the classical food chain
(Pomeroy et al. 2007), however, on the other
hand it makes carbon available to higher trophic
levels that might have been lost otherwise
(Franze and Modigh 2013).
It is important to keep in mind, however, that
in the field there are many more prey organisms
available which would enable the switching to
alternative prey and thus potentially weaken the
effects of a commensalistic IGP relationship.
Hence, the importance of commensalistic patterns between MZP predators in the field and
especially during the wax and wane of blooms
Potential implications of the findings
MZP is known to play an important ecological
role regarding the decline and disappearance of
phytoplankton blooms (Nakamura et al. 1995a, b)
and the subsequent transport of energy to higher
trophic levels (Tang and Taal 2005, Gentsch et al.
2009, Montagnes et al. 2010). The MZP predators
we investigated co-occur in the field, especially
when autotrophic dinoflagellate prey is abundant (Buskey and Stoecker 1989, Nakamura et al.
1995b). This has also been observed at Helgoland
Roads during a late summer bloom of the
resource S. trochoidea M. G. J. Lo¨der, personal
observation). When extrapolating our results to
the field, the observed facilitative pattern of
immobilization of prey could support higher
ingestion rates in the IG prey and could also
potentially benefit other MZP predators via
v www.esajournals.org
19
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
needs to be investigated further as it adds a new
aspect to the intricate interactions between
phytoplankton and its predators.
the presence of the IG predator opened up an
opportunity for the IG prey despite IGP. The
promotion of the IG prey due to pre-conditioned
prey was of such magnitude that its apparent
growth rate was double the growth rate of the IG
prey in the experiments without the IG predator.
Thus, the presence of the IG predator allowed the
IG prey to do better, which would improve the
chances of coexistence of both predators in this
system. In fact, Shchekinova et al. (2014a, b)
modelled the interactions described above and
clearly showed the stabilizing effect of prey
immobilization for an IGP system. Permanent
coexistence in the IGP system becomes more
likely as a result of this commensalistic relationship (Shchekinova et al. 2014a, b). Clearly, the
next step investigating the stability in this IGP
system experimentally would be long-term chemostat experiments (Hiltunen et al. 2013), to
investigate whether long-term coexistence is
indeed possible.
Potential coexistence of IG predator and IG prey
Although previous research focused on the
destabilizing effects of IGP there is growing
evidence for mutualistic interactions in various
IGP systems, which help to explain seemingly
paradoxical coexistence patterns under strong
competitive conditions (Crowley and Cox 2011,
Assaneo et al. 2013). Thus, the question arises
whether these interactions affect the probabilities
of coexistence between competing MZP predator
species. Several assumptions need to be considered when discussing the coexistence of IG
predator and IG prey (Diehl and Feissel 2000)
with their communal prey in a three species
system. According to Diehl and Feissel (2001) the
most crucial one is that the IG prey must be the
superior competitor for the resource. This means
it should be able to persist at lower levels of
resource than the IG predator when grown alone,
because with increasing resource availability in
an IGP system the IG prey will be driven to
extinction by the boosted IG predator population. This requirement can be elucidated by the
comparison of a three species linear food chain
with a three species IGP system: while with
increasing resource availability in the linear food
chain relationship the top predator biomass
increases and the intermediate predator biomass
stays constant—in the IGP system, where IG prey
and IG predator compete, an increasing resource
availability boosts the omnivorous IG predator
and therefore also its predation pressure on the
IG prey. Thus, the predation pressure on the
intermediate predator is always higher in an IGP
system (Diehl and Feissel 2001; for more details
please refer to this study).
We have shown that it is unlikely that the IG
prey G. dominans is the superior competitor for
the resource when compared to the IG predator
F. ehrenbergii, especially in terms of its low
growth rates. The fact that the IG predator could
drive the resource to extinction, whereas the IG
prey could not (compare Fig. 2a, b) illustrates
this. We, thus, would not expect a long-term
stable coexistence to be possible from what we
observed in the single predator treatments.
Considering the combined predator treatments,
v www.esajournals.org
Conclusion and outlook
Our experiments showed that the IG predator
F. ehrenbergii can have a facilitative relationship
with its IG prey G. dominans which leads to
enhanced ingestion and growth in the IG prey.
The promotion of the IG prey was caused by
immobilization of prey by the IG predator and
the interaction between the IG predator and IG
prey can thus be categorized as a competitive
predator relationship with a commensalistic
element. To our knowledge such a relationship
has not been reported before within the MZP
community.
Our results showed that interactions within the
MZP community, and, especially in IGP systems,
are more complex than previously thought.
Moreover, they help to explain seemingly paradoxical coexistence patterns under strong competitive conditions. Although IGP is common in
nature and found in high frequencies (Arim and
Marquet 2004) such IGP patterns are poorly
investigated in aquatic systems so far. There is
growing evidence that IGP systems could harbor
mutualistic or facilitative relationships (Crowley
and Cox 2011) which can have the potential to
enhance coexistence (Assaneo et al. 2013). Further research on mutualistic and facilitative
relationships is necessary to elucidate the diverse
interactive patterns that can occur between
20
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
and photic stimuli. Journal of Experimental Marine
Biology and Ecology 132:1–16.
Calbet, A. 2008. The trophic roles of microzooplankton
in marine systems. ICES Journal of Marine Science
65:325–331.
Calbet, A., and M. R. Landry. 2004. Phytoplankton
growth, microzooplankton grazing, and carbon
cycling in marine systems. Limnology and Oceanography 49:51–57.
Calbet, A., and E. Saiz. 2005. The ciliate-copepod link
in marine ecosystems. Aquatic Microbial Ecology
38:157–167.
Capriulo, G. M., E. B. Sherr, and B. F. Sherr. 1991.
Trophic behaviour and related community feeding
activities of heterotrophic marine protists. Pages
219–265 in P. C. Reid, C. M. Turley, and P. H.
Burkill, editors. Protozoa and their role in marine
processes. Springer Verlag, Berlin, Germany.
Chesson, J. 1978. Measuring preference in selective
predation. Ecology 59:211–215.
Chesson, J. 1983. The estimation and analysis of
preference and its relationship to foraging models.
Ecology 64:1297–1304.
Crowley, P. H., and J. J. Cox. 2011. Intraguild
mutualism. Trends in Ecology and Evolution
26:627–633.
De Laender, F., D. Van Oevelen, K. Soetaert, and J. J.
Middelburg. 2010. Carbon transfer in herbivoreand microbial loop-dominated pelagic food webs
in the southern Barents Sea during spring and
summer. Marine Ecology Progress Series 398:93–
107.
Diehl, S., and M. Feissel. 2000. Effects of enrichment on
three-level food chains with omnivory. American
Naturalist 155:200–218.
Diehl, S., and M. Feissel. 2001. Intraguild prey suffer
from enrichment of their resources: A microcosm
experiment with ciliates. Ecology 82:2977–2983.
Fischer, J. P., and H. Cypionka. 2006. Analysis of
aerotactic band formation by Desulfovibrio desulfuricans in a stopped-flow diffusion chamber. FEMS
Microbiology Ecology 55:186–194.
Fox, J. W., W. A. Nelson, and E. McCauley. 2010.
Coexistence mechanisms and the paradox of the
plankton: quantifying selection from noisy data.
Ecology 91:1774–1786.
Franze, G., and M. Modigh. 2013. Experimental
evidence for internal predation in microzooplankton communities. Marine Biology 160:3103–3112.
Frost, B. W. 1972. Effects of size and concentration of
food particles on the feeding behavior of the
marine planktonic copepod Calanus pacificus. Limnology and Oceanography 17:805–815.
Gause, G. F. 1934. The struggle for existence. Williams
and Wilkins, Baltimore, Maryland, USA.
Gentsch, E., T. Kreibich, W. Hagen, and B. Niehoff.
2009. Dietary shifts in the copepod Temora long-
various members of the MZP at the base of
aquatic food webs and, consequently, to clarify
their implications for food web interactions.
ACKNOWLEDGMENTS
This study was part of a PhD thesis at the Alfred
Wegener Institute Helmholtz Centre for Polar and
Marine Research and we are grateful for the funding.
Special thanks to Heribert Cypionka for providing the
software ‘‘Trace’’ and his valuable advice. Many thanks
to the crews of the Research vessels ‘‘Aade’’ and
‘‘Diker’’ for providing the samples, to Uwe Nettelmann for help during filming, and last but not least the
whole team of the BAH Food Web Project for fruitful
discussions and assistance whenever necessary.
LITERATURE CITED
Aberle, N., H. Hillebrand, J. Grey, and K. H. Wiltshire.
2005. Selectivity and competitive interactions between two benthic invertebrate grazers (Asellus
aquaticus and Potamopyrgus antipodarum): an experimental study using C-13- and N-15-labelled
diatoms. Freshwater Biology 50:369–379.
Aberle, N., K. Lengfellner, and U. Sommer. 2007.
Spring bloom succession, grazing impact and
herbivore selectivity of ciliate communities in
response to winter warming. Oecologia 150:668–
681.
Arim, M., and P. A. Marquet. 2004. Intraguild
predation: a widespread interaction related to
species biology. Ecology Letters 7:557–564.
Assaneo, F., R. M. Coutinho, Y. C. Lin, C. Mantilla, and
F. Lutscher. 2013. Dynamics and coexistence in a
system with intraguild mutualism. Ecological
Complexity 14:64–74.
Azam, F. 1998. Microbial control of oceanic carbon
flux: The plot thickens. Science 280:694–696.
Azam, F., T. Fenchel, J. G. Field, J. S. Gray, L. A. MeyerReil, and F. Thingstad. 1983. The ecological role of
water-column microbes in the sea. Marine Ecology
Progress Series 10:257–263.
Begon, M., C. R. Townsend, and J. L. Harper. 2006.
Ecology: From individuals to ecosystems. Fourth
edition. Blackwell, London, UK.
Beninca, E., J. Huisman, R. Heerkloss, K. D. Johnk, P.
Branco, E. H. Van Nes, M. Scheffer, and S. P. Ellner.
2008. Chaos in a long-term experiment with a
plankton community. Nature 451:822–827.
Brendelberger, H. 1997. Coprophagy: a supplementary
food source for two freshwater gastropods? Freshwater Biology 38:145–157.
Buskey, E. J., and D. K. Stoecker. 1989. Behavioral
responses of the marine tintinnid Favella sp. to
phytoplankton: influence of chemical, mechanical
v www.esajournals.org
21
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
icornis during spring: evidence from stable isotope
signatures, fatty acid biomarkers and feeding
experiments. Journal of Plankton Research 31:45–
60.
Gerritsen, J., and J. R. Strickler. 1977. Encounter
probabilities and community structure in zooplankton: a mathematical model. Journal of the
Fisheries Research Board of Canada 34:73–82.
Gifford, D. J., and M. J. Dagg. 1988. Feeding of the
estuarine copepod Acartia tonsa Dana: Carnivory
vs. herbivory in natural microplankton assemblages. Bulletin of Marine Science 43:458–468.
Guillard, R. R. L., and J. H. Ryther. 1962. Studies of
marine planktonic diatoms. I. Cyclotella nana
Hustedt and Detonula confervacea (Cleve) Gran.
Canadian Journal of Microbiology 8:229–239.
Hansen, P. J. 1992. Prey size selection, feeding rates and
growth dynamics of heterotrophic dinoflagellates
with special emphasis on Gyrodinium spirale.
Marine Biology 114:327–334.
Heinbokel, J. F. 1978. Studies on the functional role of
tintinnids in the Southern California Bight. I.
Grazing and growth rates in laboratory cultures.
Marine Biology 47:177–189.
Hillebrand, H., C.-D. Du¨rselen, D. Kirschtel, U.
Pollingher, and T. Zohary. 1999. Biovolume calculation for pelagic and benthic microalgae. Journal
of Phycology 35:403–424.
Hiltunen, T., L. E. Jones, S. P. Ellner, and N. G.
Hairston. 2013. Temporal dynamics of a simple
community with intraguild predation: an experimental test. Ecology 94:773–779.
Holt, R. D., and G. R. Huxel. 2007. Alternative prey
and the dynamics of intraguild predation: Theoretical perspectives. Ecology 88:2706–2712.
Hutchinson, G. E. 1961. The paradox of the plankton.
American Naturalist 95:137.
Irigoien, X., K. J. Flynn, and R. P. Harris. 2005.
Phytoplankton blooms: a ‘loophole’ in microzooplankton grazing impact? Journal of Plankton
Research 27:313–321.
Jakobsen, H. H., and P. J. Hansen. 1997. Prey size
selection, grazing and growth response of the small
heterotrophic dinoflagellate Gymnodinium sp. and
the ciliate Balanion comatum: a comparative study.
Marine Ecology Progress Series 158:75–86.
Johansson, M., E. Gorokhova, and U. Larsson. 2004.
Annual variability in ciliate community structure,
potential prey and predators in the open northern
Baltic Sea proper. Journal of Plankton Research
26:67–80.
Landry, M. R., and A. Calbet. 2004. Microzooplankton
production in the oceans. ICES Journal of Marine
Science 61:501–507.
Lo¨der, M. G. J., A. C. Kraberg, N. Aberle, S. Peters, and
K. H. Wiltshire. 2012. Dinoflagellates and ciliates at
Helgoland Roads, North Sea. Helgoland Marine
v www.esajournals.org
Research 66:11–23.
Lo¨der, M. G. J., C. Meunier, K. Wiltshire, M. Boersma,
and N. Aberle. 2011. The role of ciliates, heterotrophic dinoflagellates and copepods in structuring
spring plankton communities at Helgoland Roads,
North Sea. Marine Biology 158:1551–1580.
Loladze, I., Y. Kuang, J. J. Elser, and W. F. Fagan. 2004.
Competition and stoichiometry: coexistence of two
predators on one prey. Theoretical Population
Biology 65:1–15.
Montagnes, D. J. S., J. F. Dower, and G. M. Figueiredo.
2010. The protozooplankton-ichthyoplankton trophic link: an overlooked aspect of aquatic food
webs. Journal of Eukaryotic Microbiology 57:223–
228.
Nakamura, Y., S. Y. Suzuki, and J. Hiromi. 1995a.
Growth and grazing of a naked heterotrophic
dinoflagellate, Gyrodinium dominans. Aquatic Microbial Ecology 9:157–164.
Nakamura, Y., S. Y. Suzuki, and J. Hiromi. 1995b.
Population dynamics of heterotrophic dinoflagellates during a Gymnodinium mikimotoi red tide in
the Seto Inland Sea. Marine Ecology Progress Series
125:269–277.
Polis, G. A., and R. D. Holt. 1992. Intraguild predation:
The dynamics of complex trophic interactions.
Trends in Ecology & Evolution 7:151–154.
Polis, G. A., C. A. Myers, and R. D. Holt. 1989. The
ecology and evolution of intraguild predation:
potential competitors that eat each other. Annual
Review of Ecology and Systematics 20:297–330.
Pomeroy, L. R., P. J. I. Williams, F. Azam, and J. E.
Hobbie. 2007. The microbial loop. Oceanography
20:28–33.
Poulsen, L. K., and M. H. Iversen. 2008. Degradation of
copepod fecal pellets: key role of protozooplankton. Marine Ecology-Progress Series 367:1–13.
Qi, Y. Z., J. F. Chen, Z. H. Wang, N. Xu, Y. Wang, P. P.
Shen, S. H. Lu, and I. J. Hodgkiss. 2004. Some
observations on harmful algal bloom (HAB) events
along the coast of Guangdong, southern China in
1998. Hydrobiologia 512:209–214.
Riegman, R., B. R. Kuipers, A. A. M. Noordeloos, and
H. J. Witte. 1993. Size-differential control of
phytoplankton and the structure of plankton
communities. Netherlands Journal of Sea Research
31:255–265.
Scheffer, M., S. Rinaldi, J. Huisman, and F. J. Weissing.
2003. Why plankton communities have no equilibrium: solutions to the paradox. Hydrobiologia
491:9–18.
Shchekinova, E., M. G. J. Lo¨der, M. Boersma, and K. H.
Wiltshire. 2014a. Facilitation of intraguild prey by
its intraguild predator in a three-species LotkaVolterra model. Journal of Theoretical Population
Biology 92:55–61.
Shchekinova, E., M. G. J. Lo¨der, M. Boersma, and K. H.
22
October 2014 v Volume 5(10) v Article 128
¨ DER ET AL.
LO
Wiltshire. 2014b. Stable coexistence in a LotkaVolterra model with heterogeneous resources and
intraguild predation. Physical Review E 88(6). doi:
http://dx.doi.org/10.1103/PhysRevE.88
Sherr, E. B., and B. F. Sherr. 2002. Significance of
predation by protists in aquatic microbial food
webs. Antonie Van Leeuwenhoek International
Journal of General and Molecular Microbiology
81:293–308.
Sherr, E. B., and B. F. Sherr. 2007. Heterotrophic
dinoflagellates: a significant component of microzooplankton biomass and major grazers of diatoms
in the sea. Marine Ecology Progress Series 352:187–
197.
Sherr, E. B., and B. F. Sherr. 2009. Capacity of
herbivorous protists to control initiation and
development of mass phytoplankton blooms.
Aquatic Microbial Ecology 57:253–262.
Sherr, E. B., B. F. Sherr, R. D. Fallon, and S. Y. Newell.
1986. Small, aloricate ciliates as a major component
of the marine heterotrophic nanoplankton. Limnology and Oceanography 31:177–183.
Spittler, P. 1973. Feeding experiments with tintinnids.
Oikos Supplement 15:128–132.
Sprules, W. G., and J. E. Bowerman. 1988. Omnivory
and food chain length in zooplankton food webs.
Ecology 69:418–426.
Stoecker, D. K., and J. M. Capuzzo. 1990. Predation on
protozoa: its importance to zooplankton. Journal of
Plankton Research 12:981–908.
Stoecker, D. K., and G. T. Evans. 1985. Effects of
protozoan herbivory and carnivory in a microplankton food web. Marine Ecology Progress Series
25:159–167.
Stoecker, D. K., S. M. Gallager, C. J. Langdon, and L. H.
Davis. 1995. Particle capture by Favella sp. (Ciliata,
Tintinnina). Journal of Plankton Research 17:1105–
1124.
Tang, K. W., and M. Taal. 2005. Trophic modification of
food quality by heterotrophic protists: speciesspecific effects on copepod egg production and
egg hatching. Journal of Experimental Marine
Biology and Ecology 318:85–98.
Tilman, D. 1994. Competition and biodiversity in
spatially structured habitats. Ecology 75:2–16.
Utermo¨hl, H. 1958. Zur Vervollkommnung der quantitativen Plankton-Methodik. Mitteilung der Internationalen Vereinigung fu¨ r theoretische und
angewandte Limnologie 9:1–38.
Vanderploeg, H., and D. Scavia. 1979a. Calculation and
use of selectivity coefficients of feeding: zooplankton grazing. Ecological Modelling 7:135–149.
Vanderploeg, H., and D. Scavia. 1979b. Two electivity
indices for feeding with special reference to
zooplankton grazing. Journal of the Fisheries
Research Board of Canada 36:362–365.
SUPPLEMENTAL MATERIAL
APPENDIX
Live observations revealed that the IG predator Favella ehrenbergii rejected a significant proportion
of its Scrippsiella trochoidea catch after initial uptake. This behavior led to an immobilization of
approximately 26% of the caught cells. This video shows the immobilization process of F. ehrenbergii
when feeding on S. trochoidea. It can be observed after ;11 seconds of the film in the middle of the
scene. http://dx.doi.org/10.1890/ES14-00037.2
v www.esajournals.org
23
October 2014 v Volume 5(10) v Article 128

Download Report