Download biotic interactions alter top-down pressure on a leaf

!
!
!
!
!
!
!
!
BIOTIC INTERACTIONS ALTER TOP-DOWN PRESSURE ON
A LEAF-SHREDDING CADDISFLY, PHYLLOICUS HANSONI
(TRICHOPTERA: CALAMOCERATIDAE), IN TRINIDADIAN
STREAMS
by
KELLY MACKENZIE MURRAY
A Thesis Submitted to the Honors Program of the University of Georgia in Fulfillment of
the Requirements for ENTO 4990H
©2014
Kelly Murray
All Rights Reserved
!
!
KELLY MACKENZIE MURRAY
Biotic interactions alter top-down pressure on a leaf-shredding caddisfly, Phylloicus
hansoni (Trichoptera: Calamoceratidae), in Trinidadian streams
(Under the direction of CATHERINE PRINGLE, Thesis Director, and TROY SIMON,
Thesis Reader)
ABSTRACT
The influence of predators can play a prominent role in shaping characteristics of
prey populations, but it is also important to study species interactions within the context
of an ecological community. Headwater streams in Trinidad’s Northern Range Mountains
are often characterized by high densities of killifish (Anablepsoides hartii) upstream of
barrier waterfalls, with downstream reaches containing killifish at lower densities as well
as guppies (Poecilia reticulata); when sympatric, intraguild predation between these two
species results in changes in killifish populations and life history. Killifish are predators
of the leaf-shredding caddisfly Phylloicus hansoni (Trichoptera: Calamoceratidae), and
previous research has suggested that there is increased predation pressure where guppies
are absent. Larval P. hansoni play a key role in controlling rates of leaf decomposition in
Trinidadian streams; therefore, predator-mediated effects on this tropical caddisfly are
important to understand. In this study, we examined the effects of fish assemblage on the
size structure of larval P. hansoni populations in streams (n=5) with distinct killifish-only
(KO) and killifish-guppy (KG) reaches. Analysis of size distributions of P. hansoni from
KO and KG reaches confirmed our prediction that killifish alter the size structure of
larvae differently depending on the presence or absence of guppies: larvae from KO
reaches exhibited a size frequency peak at a smaller size than in KG reaches. These
results indicate that the influence of a predatory fish on an important macroinvertebrate
species differs at small spatial scales with changes in the community assemblage.
2
!
!
INTRODUCTION
Predators can affect prey populations in many different ways beyond exerting topdown control of the numerical abundance (Peckarsky et al. 2008). Predators can induce
population cycling in prey species (Krebs et al. 2001) and subsequent adaptation
(Abrams and Matsuda 1997, Yoshida et al. 2003) in prey species, influence trophic
cascades (Carpenter and Kitchell 1988), and alter the size structure of prey populations
(Brooks and Dodson 1965). Theory predicts that populations of organisms experiencing
predation pressure will be driven to reproduce at an earlier age and reduced size (Abrams
and Rowe 1996), and empirical studies have shown that predators can prompt a reduction
in size at emergence of aquatic insect prey (Peckarsky et al. 2001) as well as reductions
in fecundity in both aquatic (Greig and McIntosh 2008) and terrestrial systems (Dixon
and Agarwala 1999). Often, to elucidate the specifics of such impacts on prey
populations, studies must focus comparing prey populations with the presence or absence
of a predator, but other biotic interactions occurring within the community are important
to consider; intraguild predation, for example, is one source of complexity (Davenport
and Chalcraft 2012, Anderson and Semlitsch 2014). Here we examine how the effects of
a predator can shift in the context of an ecological community by evaluating how the topdown pressure of a predatory fish on a functionally important leaf-shredding caddisfly is
mediated by differences in the total stream fish assemblage.
Our research focuses on larval populations of the caddisfly Phylloicus hansoni
(Trichoptera: Calamoceratidae) in Trinidadian headwater streams. P. hansoni (syn. P.
angustior) is native to the island of Trinidad, as well as Venezuela, and the genus is
found throughout Central and South America (Flint 1996, Prather 2003). The aquatic
3
!
!
larvae are leaf-shredders, which is a role often essential to the distribution of fine
particulate organic matter resources to filter-feeding invertebrates (Cummins et al. 1989,
Wallace and Webster 1996), which can control secondary production throughout the
stream community (Wallace et al. 1997). Using allochthonous leaf inputs for both
feeding purposes and the construction of protective cases, Phylloicus spp. break down
leaves containing a broad range of nutrient, polyphenol, and lignin contents (Rincón and
Martínez 2006). Phylloicus spp. are consistently reported as key decomposers of organic
matter in tropical headwater streams when present (Table 1), making this genus a notable
exception to the general scarcity of insect shredders in the tropics compared to temperate
zones (Wantzen and Wagner 2006, Boyero et al. 2011); P. hansoni is thought to be the
only macroinvertebrate shredder on the island of Trinidad (Botosaneanu and Sakal 1992).
Its important role in the facilitation of leaf breakdown motivated the study of larval P.
hansoni populations experiencing varying levels of predation from the killifish
Anablepsoides hartii (syn. Rivulus hartii) in Trinidadian streams.
Headwater streams located in Trinidad’s Northern Range Mountains provide an
ideal system for studying the interaction of both ecological and evolutionary processes
(Reznick and Endler 1982, Palkovacs et al. 2009, Bassar et al. 2010). Large barrier
waterfalls segregate fish communities between stream reaches by preventing certain
species from migrating upstream. As elevation increases, there are fewer fish species
present (Gilliam et al. 1993). Often, the killifish A. hartii is the only fish present in the
highest-elevation stream reaches (killifish-only, or “KO” reaches). Immediately
downstream from these KO reaches, guppies (Poecilia reticulata) and killifish are
sympatric (in killifish-guppy, or “KG” reaches) – see Figure 1. In the KG reaches of our
4
!
!
focal streams, guppies are thought to primarily consume algae, detritus, and small
invertebrates (Zandonà et al. 2011). Guppy prey can also include killifish juveniles, and
killifish growth rates are altered when coexisting with guppies (Walsh et al. 2011, Fraser
and Lamphere 2013). Killifish are primarily insectivorous (Gilliam et al. 1993, Fraser et
al. 1993), and are more abundant in streams where they are the sole fish species (Gilliam
et al. 1993, Walsh and Reznick 2008, 2009, 2010, 2011). Killifish also occasionally
consume guppies in KG reaches (K. Murray, personal observation from gut content
analysis). Here, we investigate how intraguild predation between of these two fish species
can influence the effects of killifish on populations of P. hansoni, which occur in both
reaches of these streams.
Previous experiments examined how the presence of guppies alters trophic
cascades within this system (Binderup 2011) using exclusion experiments (as in Pringle
and Blake 1994) to isolate the effects of macroconsumers on leaf decomposition.
Binderup (2011) reported that the presence of guppies released P. hansoni larvae from
the top-down control of killifish, which indirectly altered leaf decomposition: within a
KO reach, killifish predation reduces P. hansoni abundance to a greater degree, thereby
reducing leaf decay rates. Alternatively, within a KG reach, killifish did not reduce P.
hansoni abundance and leaf decay rates were not affected by top consumers. This
strongly suggests that leaf-shredding P. hansoni larvae experience reduced predation
pressure in the presence of guppies, and this decoupling of a trophic cascade could
change resource dynamics within the stream ecosystem (Binderup 2011). Our study
investigates whether the presence of guppies produces a consistent influence on killifish’s
effects on P. hansoni populations by using multiple streams with similarly partitioned
5
!
!
fish communities. We also focus on characteristics of killifish impacts in the context of
guppy presence beyond reductions in abundance of P. hansoni larvae by evaluating
differences in larval size structure.
Growth of aquatic macroinvertebrates is an important process to understand
within the framework of a stream ecosystem. Holometabolous aquatic insects, such as
caddisflies, undergo complex life cycles and must achieve a balance between growth and
predator avoidance under time constraints (Rowe and Ludwig 1991). Once adult insects
emerge from their aquatic environment, they can become important subsidies of prey and
nutrients to terrestrial ecosystems (Nakano and Murakami 2001, Hoekman et al. 2012).
Secondary production of aquatic macroinvertebrates is tied to growth rates and other life
history characteristics (Huryn and Wallace 2000), but predators can mediate these values
(Huryn 1998). Studying the size structures of P. hansoni experiencing the effects of
different fish assemblages may provide insight into how the biomass of aquatic insect
populations can be mediated by other biotic interactions within the community.
We used the unique microgeographic variation found in stream systems of
Trinidad’s Northern Range Mountains as a natural “experiment” to investigate whether
the ecological context (in this case, the presence or absence of guppies) in which a
predator is found can alter characteristics of its prey population across small spatial scales
(above and below waterfalls). We examined whether larval populations of P. hansoni
exhibit different size structures between areas where killifish and its intraguild predator,
guppies, are sympatric (reduced predation pressure on larvae) and where killifish are
found alone above barrier waterfalls (higher predation pressure on larvae). We predicted
that there would be a greater proportion of large individuals where P. hansoni individuals
6
!
!
coexist with killifish and guppies, compared to the frequency of large individuals in
killifish-only reaches. We also investigated a potential mechanism for any such change,
size-selective predation by killifish on P. hansoni larvae, by performing a gut content
analysis of killifish. We predicted that killifish would selectively consume the largest P.
hansoni individuals.
DESCRIPTION OF STUDY SITES
This study was conducted in streams on the southern slope of Trinidad’s Northern
Range Mountains, within the Caroni drainage. The Caroni drainage is located on the
northwestern side of the island and encompasses a total of ≈ 60,000ha. We conducted our
sampling efforts in mid-May 2013, a time corresponding to the end of the dry season in
Trinidad. The montane, spring-fed streams we studied (n=5: El Cedro “CED,” Endler
“END,” Guard Dog Creek “GDC,” Ramdeen “RDN,” and Trip Trace “TRT”) were
composed of reaches with distinct fish assemblages due to the presence of large barrier
waterfalls, which prevent guppies, but not killifish, from migrating upstream. We
sampled streams above waterfalls where killifish were found alone (“KO” reaches), and
directly below waterfalls, where killifish and guppies were sympatric (“KG” reaches).
METHODS
In order to better understand the consequences of killifish predation on P. hansoni
population dynamics, we evaluated the size structure of larval populations in relation to
fish assemblage by using five streams with distinct paired KO and KG reaches. To
determine whether physical and environmental differences between these upstream and
downstream areas could also influence P. hansoni habitats, we compared the stream
7
!
!
morphology and habitat quality between paired reaches.
Stream Characteristics
We measured four types of environmental characteristics along transects in each
reach: wetted width (cm), depth (cm), canopy cover (%), and leaf standing stocks (coarse
benthic organic matter, or CBOM, g cm-2). Three transects perpendicular to the stream
were chosen at random points along each 100m reach where sampling was conducted.
Average depth along transects was calculated from five readings at uniform distance
along the wetted width of the stream. We used a spherical densiometer to estimate
percent canopy cover over the stream by taking measurements at each edge and in the
middle of the stream. We collected leaf standing stocks from the stream benthos within
28cm of each transect line between the wetted width, then washed, dried for 36 hours,
and weighed the leaves.
We estimated killifish abundance in the study reaches to ensure that our streams
were consistent with previous findings that killifish are generally more abundant in
reaches where they are the only fish species present (Gilliam et al. 1993, Walsh and
Reznick 2008, 2009, 2010, 2011, Fraser and Lamphere 2013), thus potentially exerting
greater predation pressure on P. hansoni. We used baited minnow traps and measured the
catch per unit effort (CPUE). Three traps were placed in deep pools evenly spaced across
our 100m sampling reach. All killifish captured within 10 minutes of placing the trap
were counted and released. Each killifish was measured and categorized into one of the
following size classes based on body length: Juvenile (<15mm), Small (15-30mm),
Medium (30-45mm), Large (45-60mm), and Extra Large (>60mm).
8
!
!
To test whether the environmental parameters were consistently different between
KO and KG reaches, we calculated the mean and standard error of depth, wetted width,
percent canopy cover, and leaf standing stocks of the three transects per reach, and the
number of killifish caught per trap. We compared these measurements between KO and
KG reaches across all five streams with a paired t-test. For this and all other analyses,
significance was determined at p≤0.05.
Phylloicus hansoni Size Structure
For our evaluation of differences in larval size structures between the two
different types of fish community, we collected P. hansoni larvae in the KO and KG
reaches of the five focal streams. In each stream reach, we collected P. hansoni larvae in
pools with ample leaf litter availability with a standardized sampling effort: searching ~7
minutes per pool, or until at least 50 individuals were found. P. hansoni individuals were
preserved in 5% formalin, and larval measurements were taken under a dissecting
microscope to the nearest 0.1mm in the laboratory.
Insect measurements included body length, head capsule width, and leaf case
length and width at three points evenly spaced along the length of the case. Body length
was measured from the top of the head capsule to the caudal end. The average of the
three case width measurements and case length were multiplied to estimate leaf case
surface area (mm2), to be compared with larva size. We measured the case surface area so
as to compare whether P. hansoni build larger leaf cases when under greater predation
threat.
The distributions of larval body size violated the assumptions of parametric
9
!
!
statistical tests, even when standard data transformations were used. Therefore, we used
non-parametric statistics for hypothesis testing. We used a two-sample KolmogorovSmirnov (K-S) test to evaluate differences in the distribution of P. hansoni larvae body
length between KO and KG reaches. We also used the Wilcox rank sum test to compare
the mean body length and head capsule width of larvae from KO and KG reaches.
Killifish Gut Content Analysis
To determine whether killifish preferentially consume larvae of a specific size
class, and whether this differs between killifish from different stream reaches, we
analyzed the size of P. hansoni found in killifish gut contents. Killifish used for gut
content analysis were collected from KO and KG reaches of 4 streams (3 of which were
used in the larval size structure collections: Endler, Guard Dog Creek, and El Cedro) in
late April 2011, as part of a separate study. All killifish used in this study were female.
Fish were dissected and the digestive tract was preserved in 10% formalin. Gut contents
were then extracted from the stomach and searched to identify P. hansoni larvae.
Typically only the head capsule remained intact, so this was measured to compare the
average size of larvae consumed. We also calculated the percentage of killifish with P.
hansoni larvae in their guts from both reaches, an “occurrence method,” as described by
Hyslop (1980).
RESULTS
Environmental variables evaluated by our stream transects did not show any
marked differences between upstream KO reaches and downstream KG reaches. Mean
values of wetted width, depth, percent canopy cover, and leaf standing stocks did not
10
!
!
significantly differ between KO and KG reaches overall (p = 0.2, 0.4, 0.5, and 0.2
respectively, Table 2). However, in accordance with our predictions, killifish CPUE was
higher in KO reaches than KG reaches (p=0.014, Table 2). Our estimates of killifish body
length showed that size distributions are similar, though KO reaches had greater
proportions of the largest (>60mm) individuals (Figure 2).
KO and KG reaches exhibited different patterns in the size of P. hansoni larvae.
Plots of head capsule size and body length for larvae (Figure 3) enabled us to classify a
range of body sizes into classes roughly corresponding to instar. Generally, the largest
size classes were most abundant in the KG samples, except in the case of GDC larvae,
which showed a size frequency peak shifted towards a smaller size (Figure 4). The mean
body size of all larvae was higher in KG reaches (7.8 ± 0.13mm [mean ± SE]) than in KO
reaches (6.81 ± 0.19mm), p=0.032, and the mean head capsule width of KG larvae (0.78
± 0.015mm) was also higher than KO larvae (0.69 ± 0.088mm), p=0.003 (Figure 5).
Comparing reaches within individual streams with our Wilcox rank sum tests, TRT,
RDN, CED had significantly higher larval body lengths (p=0.036, <0.001, 0.008
respectively) and head capsule widths (p=0.02, <0.001, 0.01 respectively) in KG reaches.
END KO and KG body lengths (p=0.93) and head capsule widths (p=0.69) were not
significantly different. In GDC, KO larvae had larger body lengths (p<0.001) and head
capsule widths (p=0.002). The length and surface area of the leaf case increased with
increasing larva size, and there were no differences in these relationships between KO
and KG reaches (Table 3).
With the two-sample K-S test of larval body length, we concluded that the overall
KO and KG size distributions are different (p=0.046). The majority of KO larvae body
11
!
!
sizes are localized around a smaller size, ~4-6mm, while KG larvae exhibit a peak near
8mm (Figure 6). Because this analysis did not control for the larvae’s stream of origin,
we conducted a two-sample K-S test for each of the four streams separately. The KO and
KG larval size distributions of CED, TRT, GDC, and RDN were different (p=0.008,
0.036, <0.001, <0.001 respectively), yet those of END were not (p=0.93).
From the gut content analysis of killifish collected in 2011, the means of head
capsule widths of larvae consumed by killifish in both KO (0.095 ± 0.0109mm) and KG
(0.144 ± 0.0235mm) reaches were smaller than 0.2mm (Table 4). Killifish from KG
reaches had a higher percentage of P. hansoni head capsules in their guts than killifish
from KO reaches, and in KO reaches, the killifish with P. hansoni in their gut had a mean
body length (42.57 ± 3.29mm) smaller than the mean size of all KO killifish (47.83 ±
1.48mm) analyzed (Table 4).
DISCUSSION
Our results show that size distributions of P. hansoni larvae in killifish-only
reaches are different than in killifish-guppy reaches within the same stream, with a shift
toward larger individuals in the KG reach compared to the KO. Based on a reduction in
both the mean body size and head capsule width in KO larvae, and because head capsule
width is an indicator of age, we can infer that the individuals in KO reaches are, on
average, smaller and younger than their KG counterparts. This supports our prediction
that large individuals would be more frequent in reaches where killifish and guppies
coexist, and less frequent where killifish are the only fish species present. These data
provide only a snapshot of P. hansoni populations, but differences are consistent in three
out of five streams surveyed.
12
!
!
The disparity in top-down pressure exerted by killifish on P. hansoni between KO
and KG reaches, evidenced by the differences in larval size structures, are likely the
result of several factors stemming from guppy-killifish interactions in KG reaches.
Previous studies of these stream systems (Walsh and Reznick 2008, 2009, 2010, 2011,
Walsh et al. 2011), in addition to our CPUE abundance estimates, suggest that killifish
densities are lower in reaches where they coexist with guppies. Rates of killifish growth
increase in the presence of guppies, likely as a response to guppy predation on killifish
juveniles (Walsh et al. 2011, Fraser and Lamphere 2013). In accordance with these
findings, our surveys found greater incidences of the largest (>60mm) killifish in KO
reaches (Figure 2). Additionally, there may be differences in the foraging behavior of
killifish when coexisting with guppies; it has been shown in other systems that habitat
partitioning within the fish community can result from interference competition between
two species (Nakano and Furukawa-Tanaka 1994, Katano and Aonuma 2001). We have
shown using multiple streams that these shifts in killifish population dynamics due to the
presence of guppies provide enough change to influence the shape of the size structure of
P. hansoni larvae.
The reduced frequency of larger individuals of P. hansoni in KO reaches is
ecologically significant in the context of the life cycle of this tropical caddisfly, providing
evidence of increased predation pressure altering development. The exact characteristics
of the P. hansoni life cycles in Trinidad have not been documented, but we can make
predictions based on those of other organisms. Studies of larval development of
Trichopterans in Costa Rica that included three Phylloicus species (P. elegans, P.
ornatus, and P. undescribed sp. nr. ornatus) provided evidence for multivoltinism
13
!
!
(having multiple generations within one year) in this group (Jackson and Sweeney 1995).
While emergence of P. hansoni adults may occur throughout the year, it may exhibit a
peak in a particular season; this would be similar to patterns demonstrated by studies of
adult caddisfly abundance in Panama (McElravy et al. 1982), which included sampling of
Phylloicus spp. adults, and surveys of another leaf-shredding Calamoceratid,
Anisocentropus kirramus, in tropical Australia (Nolen and Pearson 1992). Both studies
found increases in the number of adults in the summer rainy season. Additionally, a
survey of Phylloicus sp. larvae in Brazilian streams documented an increase in the largest
individuals in the spring, and predicted that adult emergence would most likely start after
the beginning of the wet season (Huamantinco and Nessimian 2000). If populations of P.
hansoni in Trinidad also follow this seasonal pattern, we would expect that our samples
taken in May (corresponding to the end of the dry season in Trinidad) would contain the
largest larvae in greatest proportions; however, this is only the case with samples from
KG sections of streams, not in the KO reaches.
The differential effects of killifish predation in KO and KG reaches as given by
our size structure data may be influencing the timing of emergence, the size of adults, or
both. Life histories of stream insects are marked by trade-offs. Caddisflies such as P.
hansoni typically possess an ephemeral adult stage, so it is beneficial for these organisms
to synchronize emergence with an environmental cue, like increasing discharge (which
may explain the pattern of increased adult emergence in rainy seasons). It is also
advantageous for larvae to achieve a maximum size before emerging, because adult size
after metamorphosis is governed by larval growth, and is directly correlated to fecundity
and reproduction; however, the threat of predation may drive caddisflies to escape the
14
!
!
stream earlier than is ideal (Rowe and Ludwig 1991). More research is necessary to how
the effects of predation on the P. hansoni life cycle are potentially manifested, but we
have shown that another organism in the fish community can mediate these influences of
a predator.
Even though emergence may be more common in one season, growth and
development of aquatic larvae can be controlled by a variety of physical factors
(Sweeney 1984). We found no consistent differences in the physical habitat between
upstream and downstream reaches based on the environmental features measured, which
included an estimation of food and habitat resources for P. hansoni larvae by leaf
standing stock mass. Temperature is an important element controlling rates of larval
insect development (Sweeney 1984), but differences are likely minimal over the small
spatial scale separating KO and KG reaches of these spring-fed headwater streams with
similar canopy cover. Therefore, it is more likely that overall differences observed
between KO and KG P. hansoni populations are primarily driven by the differences in
fish community.
However, patterns were not consistent or significant throughout every stream we
surveyed. The shape of the larval size distributions from both the KO and KG reaches of
Guard Dog Creek (GDC) were shifted towards smaller sizes (Figure 4) and had smaller
mean larval body length and head capsule widths (Table 3) compared to other streams,
and differences between reaches in Endler (END) were not significant. Certain
environmental and geomorphological factors could influence whether a particular stream
reach is suitable for larvae growth and survivorship. Studies of mayfly oviposition and
recruitment have found that the availability of suitable rock substrate corresponds to
15
!
!
female ovipoisition (Encalada and Peckarsky 2006), and these streams providing ideal
hydrogeomorphic habitats host greater egg densities even though they also have trout
predators (Encalada and Peckarsky 2011). One factor such as this that may have
influenced our results is the stream depth. The KO and KG reaches in GDC were on
average ~1.8x deeper than the means for both reaches from all five streams (Table 2),
which indicates that the quality of physical habitat for P. hansoni larvae may not be as
suitable in the sampled portions of this particular stream; Phylloicus spp. larvae are
usually found in pools along the stream margin (Wantzen and Wagner 2006, Turner et al.
2008). So certain abiotic characteristics may overwhelm the effects of fish community,
but with consistent differences in the size structure P. hansoni larval populations between
KO and KG reaches in three streams, the influence of fish assemblage is nevertheless an
important factor.
There could be several explanations for a reduced prey size due to increased
predation pressure, as found in the KO reaches. Our gut content analysis showed that the
prevalence of smaller P. hansoni larvae in such reaches is probably not solely due to
positive size-selection from killifish, contrary to our prediction. Killifish are not expected
to be gape-limited (Binderup 2011), and large, late-instar P. hansoni individuals were
witnessed inside killifish guts (K. Murray, personal observation), but killifish may avoid
larger instars due to their unpalatable leaf case, which increases with body size. However,
the difference of two years in the gut content sampling (2011) and the P. hansoni
collections (2013), even in similar seasons, might not reflect an alignment in the
availability of large larvae as prey in these streams.
Other factors to consider include the potential suppression of prey activity levels
16
!
!
in order to avoid predators, or “adaptive foraging behavior,” a common occurrence
(Schmitz et al. 2008) that prompts a trade-off with growth (Werner and Anholt 1993).
Reductions in prey activity as a response to predator presence have been observed in
other aquatic, detrital-based systems (Boyero et al. 2008), but this is not always the case
(Greig and McIntosh 2006). Even when foraging is unchanged, prey populations may
exhibit altered growth due to reductions in nutrient assimilation when under the stress of
predation threat (Stoks et al. 2005). Predators can also impose selection on prey life
history evolution (Rowe and Ludwig 1991) or phenotypically plastic responses, which
may include emergence at smaller sizes (Peckarsky et al. 2001) and reductions in
fecundity (Greig and McIntosh 2008) in the case of aquatic insects.
We do not know the extent of genetic mixing between P. hansoni from KO and
KG reaches due to dispersal in the winged adult stage, and source-sink dynamics within
the metacommunity are important to consider (Leibold et al. 2004). The presence of
predators has been shown to elicit plastic responses in a variety of aquatic systems where
prey experience variable predation pressure (Van Buskirk and Schmidt 2000, Peckarsky
et al 2002, Torres-Dowdall et al. 2012), and the differential predation pressure on P.
hansoni larvae between nearby KO and KG reaches may provide selection for such
plasticity.
Further investigations are needed to provide insight into the mechanisms driving
the altered size structures of P. hansoni populations observed in our study between KO
and KG reaches. Recently, there has been emphasis on the need to evaluate differences in
ecological communities that may be causing adaptation at the “microgeographic” scale
(Richardson et al. 2014). Our study suggests that a shift in biotic interactions occurring
17
!
!
along a small spatial scale (within-stream distance <500m) can have consequences
throughout the community, in this case affecting a functionally important
macroinvertebrate species. Closer examinations of community-level differences within an
ecosystem may elucidate more complexity in ecological and evolutionary interactions
than previously estimated.
18
!
!
ACKNOWLEDGEMENTS
This project received support from the following programs at the University of
Georgia: the 2013 Honors International Scholars Program, the Center for Undergraduate
Research Opportunities 2013 Summer Fellowship, and the College of Agricultural and
Environmental Sciences 2013 Undergraduate Research Initiative. I would not have been
able to conduct this research or complete this thesis without the continuous support,
guidance, and dedication of Troy Simon, PhD Candidate in the Odum School of Ecology,
and Dr. Catherine Pringle, Distinguished Research Professor in the Odum School of
Ecology. I am also very grateful to Dr. Marianne Shockley of the UGA Department of
Entomology for her support in applying for the CAES Undergraduate Research Initiative
grant and in the application of this thesis to fulfill the requirements of ENTO 4990H.
Thank you to David Stoker and John Kronenberger for their assistance with 2013 data
collection in Trinidad, as well as Travis McDevitt-Galles, William Roberts, Michael
Rautenberg, Anika Bratt, Tierney Schipper, Emily Nash, and Joshua Soden for the 2011
collections and initial processing of the killifish used in this study. Thank you to the
graduate students in the Pringle lab group for their advice and helpful comments on this
manuscript: Tom Barnum, Jessica Chappell, Maura Dudley, Carissa Ganong, Jeremy
Sullivan, and Pedro Torres. And thank you to Dr. Darold Batzer, UGA Department of
Entomology, for advice on appropriate methodologies prior my trip to Trinidad. Finally,
thank you to my friends and family, who all provided invaluable support throughout my
undergraduate research career.
19
!
!
REFERENCES
Abrams, P.A. and H. Matsuda. 1997. “Prey adaptation as a cause of predator-prey
cycles.” Evolution 51(6): 1742-1750.
Abrams, P.A. and L. Rowe. 1996. “The effects of predation on the age and size of
maturity of prey.” Evolution 50(3): 1052-1061.
Anderson, T.L. and R.D. Semlitsch. 2014. “High intraguild predator density induces
thinning effects on and increases temporal overlap within prey populations.”
Population Ecology 56: 265-273.
Bassar, R.D., M.C. Marshall, A. López-Sepulcre, E. Zandonà, et al. 2010. “Local
adaptation in Trinidadian guppies alters ecosystem processes.” Proceedings of the
National Academy of Sciences 107(8): 3616-3621.
Benstead, J.P. 1996. “Macroinvertebrates and the processing of leaf litter in a tropical
stream.” Biotropica 28(3): 367-375.
Binderup, A.J. 2011. “Isolating top-down effects of aquatic macroconsumers on benthic
structure and function in a Neotropical stream.” Masters Thesis, University of
Georgia.
Botosaneanu, L. and D. Sakal. 1992. “Ecological observations on the caddisflies (Insecta:
Trichoptera) from Trinidad and Tobago (W. Indies).” Revue d’hydrobiologie
tropicale 25(3): 197-207.
Boyero, L., P.A. Rincon, and R.G. Pearson. 2008. “Effects of predatory fish on a tropical
detritus-based food web.” Ecological Research 23: 649-655.
Boyero, L., R.G. Pearson, D. Dudgeon, M.A.S. Graça, et al. 2011. “Global distribution of
a key trophic guild contrasts with common latitudinal diversity patterns.” Ecology
20
!
!
92(9): 1839-1848.
Brooks, J.L. and S.I. Dodson. 1965. “Predation, body size, and composition of plankton.”
Science 150: 28-35.
Carpenter, S.R. and J.F. Kitchell. 1988. “Consumer control of lake productivity.”
BioScience 38: 794-769.
de Carvalho, E.M. and V.S. Uieda. 2009. “Diet of invertebrates sampled in leaf-bags in a
tropical headwater stream.” Zoologia 26(4): 694-704.
Cummins, K.W., M. A. Wilzbach, D.M. Gates, J.B. Perry, and W.B. Taliaferro. 1989.
“Shredders and riparian vegetation.” BioScience 39(1): 24-30.
Davenport, J.M. and D.R. Chalcraft. 2012. “Evaluating the effects of trophic complexity
on a keystone predator by disassembling a partial intraguild predation food web.”
Journal of Animal Ecology 81: 242-250.
Dixon, A.F.G. and B.K. Agarwala. 1999. “Ladybird-induced life-history changes in
aphids.” Proceedings of the Royal Society B-Biological Sciences 266(1428):
1549-1553.
Encalada, A.C., J. Calles, V. Ferreira, C. Canhoto, and M.A.S. Graça. 2010. “Riparian
land use and the relationship between benthos and litter decomposition in tropical
montane streams.” Freshwater Biology 55: 1719-1733.
Encalada, A.C., and B.L. Peckarsky. 2006. “Selective oviposition of the mayfly Baetis
bicaudatus.” Oecologia 148(3): 526-537.
Encalada, A.C., and B.L. Peckarsky. 2011. “The influence of recruitment on withingeneration population dynamics of a mayfly.” Ecosphere 2(10): 107.
Flint, O.S. Jr. 1996. “Studies of Neotropical caddisflies LV: Trichoptera of Trinidad and
21
!
!
Tobago.” Transactions of the American Entomological Society (1890-) 122(2/3):
67-113.
Fraser, D.F. and B.A. Lamphere. 2013. “Experimental evaluation of predation as a
facilitator of invasion success in a stream fish.” Ecology 94(3): 640-649.
Greig, H.S. and A.R. McIntosh. 2006. “Indirect effects of predatory trout on organic
matter processing in detritus-based stream food webs.” Oikos 112: 31-40.
Greig, H.S. and A.R. McIntosh. 2008. “Density reductions by predatory trout increase
adult size and fecundity of surviving caddisfly larvae in a detritus-based stream
food web.” Freshwater Biology 53(8): 1579-1591.
Gilliam, J.F., D.F. Fraser, and M. Alkins-Koo. 1993. “Structure of a tropical stream fish
community: a role for biotic interactions.” Ecology 74(6): 1856-1870.
Hoekman, D., M. Bartrons, and C. Gratton. 2012. “Ecosystem linkages revealed by
experimental lake-derived isotope signal in heathland food webs.” Oecologia 170:
735-743.
Huamantinco, A.A. and J.L. Nessimian. 2000. “Variation and life strategies of the
Trichoptera (Insecta) larvae community in a first order tributary of the Paquequer
River, Southeastern Brazil.” Rev. Brasil. Biol. 60(1): 73-82.
Huryn, A.D. 1998. “Ecosystem-level evidence for top-down and bottom-up control of
production in a grassland stream system.” Oecologia 115: 173-183.
Huryn, A.D. and J.B. Wallace. 2000. “Life history and production of stream insects.”
Annual Review of Entomology 45: 83-110.
Hyslop, E.J. 1980. “Stomach contents analysis—a review of methods and their
application.” Journal of Fish Biology 17: 411-429.
22
!
!
Jackson, J.K. and B.W. Sweeney. 1995. “Egg and larval development times for 35
species of tropical stream insects in Costa Rica.” Journal of the North American
Benthological Society 14(1): 115-130.
Katano, O. and Y. Aonuma. 2001. “Negative effect of ayu on the growth of omnivorous
pale chub in experimental pools.” Journal of Fish Biology 58(5): 1371-1382.
Krebs, C., S. Boutin, R.Boonstra, and A. Sinclair. 2001. “What drives the 10-year cycle
of snowshoe hares?” BioScience 51: 25-35.
Landeiro, V.L., N. Hamada, B.S. Godoy, and A.S. Melo. 2010. “Effects of litter patch
area on macroinvertebrate assemblage structure and leaf breakdown in Central
Amazonian streams.” Hydrobiologia 649: 355-363.
Nakano, S. and T. Furukawa-Tanka. 1994. “Intra- and interspecific dominance
hierarchies and variation in foraging tactics of two species of stream-dwelling
chars.” Ecological Research 9(1): 9-20.
Nakano, S. and M. Murakami. 2001. “Reciprocal subsides: Dynamic interdependence
between terrestrial and aquatic food webs.” Proceedings of the National Academy
of Sciences of the United States of America 98(1): 166-170.
Nolen, J.A. and R.G. Pearson. 1992. “Life history studies of Anisocentropus kirramus
Neboiss (Trichoptera: Calamoceratidae) in a tropical Australian rainforest
stream.” Aquatic Insects 14(4): 213-221.
Palkovacs, E.P., M.C. Marshall, B.A. Lamphere, B.R. Lynch, D.J. Weese, D.F. Fraser,
D.N. Reznick, C.M. Pringle, and M.T. Kinnison. 2009. “Experimental evaluation
of evolution and coevolution as agents of ecosystem change in Trinidadian
streams.” Philosophical Transactions of the Royal Society B-Biological Sciences
23
!
!
364: 1617-1628.
Peckarsky, B.L., P.A. Abrams, D.I. Bolnick, L.M. Dill, J.H. Grabowski, B. Luttbeg, J.L.
Orrock, S.D. Peacor, E.L. Preisser, O.J. Schmitz, and G.C. Trussell. 2008.
“Revisiting the classics: Considering nonconsumptive effects in textbook
examples of predator-prey interactions.” Ecology 89(9): 2416-2425.
Peckarsky, B.L., A.R. McIntosh, B.W. Taylor, and J. Dahl. 2002. “Predator chemicals
induce changes in mayfly life history traits: A whole-stream manipulation.”
Ecology 83(3): 612-618.
Peckarsky, B.L., B.W. Taylor, A.R. McIntosh, M.A. McPeek, D.A. Lytle. 2001.
“Variation in mayfly size at metamorphosis as a developmental response to risk of
predation.” Ecology 82(3): 740-757.
Prather, A.L. 2003. “Revision of the Neotropical caddisfly genus Phylloicus (Trichoptera:
Calamoceratidae).” Zootaxa 275: 1-214.
Pringle, C.M. and G.A. Blake. 1994. “Quantitative effects of atyid shrimp (Decapoda:
Atyidae) on the depositional environment in a tropical stream: Use of electricity
for experimental exclusion.” Canadian Journal of Fisheries and Aquatic Sciences
51: 1443-1450.
Reznick, D. and J.A. Endler. 1982. “The impact of predation on life history evolution in
Trinidadian guppies (Poecilia reticulata).” Evolution 36(1): 160-177.
Richardson, J.L., M.C. Urban, D.I. Bolnick, and D.K. Skelly. 2014. “Microgeographic
adaptation and the spatial scale of evolution.” Trends in Ecology and Evolution
29(3): 165-176.
Rincón, J. and I. Martínez. 2006. “Food quality and feeding preferences of Phylloicus sp.
24
!
!
(Trichoptera: Calamoceratidae).” Journal of the North American Benthological
Society 25(1): 209-215
Rowe, L. and D. Ludwig. 1991. “Size and timing of metamorphosis in complex life
cycles: Time constraints and variation.” Ecology 72(2): 413-427.
Schmitz, O.J., J.H. Grabowski, B.L. Peckarsky, E.L. Preisser, G.C. Trussell, and J.R.
Vonesh. 2008. “From individuals to ecosystem function: Toward an integration of
evolutionary and ecosystem ecology.” Ecology 89(9): 2436-2445.
Stoks, R., M. De Block, M.A. McPeek. 2005. “Alternative growth and energy storage
responses to mortality threats in damselflies.” Ecology Letters 8: 1307-1316.
Sweeney, B.W. 1984. “Factors influencing life-history patterns of aquatic insects.” In
V.H. Resh and D.M. Rosenberg (Eds.), The Ecology of Aquatic Insects 56-100.
New York: Praeger.
Torres-Dowdall, J., C.A. Handelsman, E.W. Ruell, S.K. Auer, D.N. Reznick, and C.K.
Ghalambor. 2012. “Fine-scale local adaptation in life histories along a continuous
environmental gradient.” Functional Ecology 26(3): 616-627.
Turner, D., D.D. Williams, and M. Alkins-Koo. 2008. “Longitudinal changes in benthic
community composition in four neotropical streams.” Caribbean Journal of
Science 44(3): 380-394.
Van Buskirk, J. and B.R. Schmidt. 2000. “Predator-induced phenotypic plasticity in
larval newts: Trade-offs, selection and variation.” Ecology 81(11): 3009-3028.
Wallace, J.B., S.L. Eggert, J.L. Meyer, and J.R. Webster. 1997. “Multiple trophic levels
of a forest stream linked to terrestrial litter inputs.” Science 277: 102-104.
Wallace, J.B. and J.R. Webster. 1996. “The role of macroinvertebrates in stream
25
!
!
ecosystem function.” Annual Review of Entomology 41: 115-139.
Walsh, M.R., D.F. Fraser, R.D. Bassar, and D.N. Reznick. 2011. “The direct and indirect
effect of guppies: implications for life-history evolution in Rivulus hartii.”
Functional Ecology 25(1): 227-237.
Walsh, M.R. and D.N. Reznick. 2008. “Interactions between the direct and indirect
effects of predators determine life history evolution in a killifish.” Proceedings of
the National Academy of Sciences of the United States of America 105(2): 594595.
Walsh, M.R. and D.N. Reznick. 2009. “Phenotypic diversification across an
environmental gradient: A role for predators and resource availability on the
evolution of life histories.” Evolution 63(12): 3201-3213.
Walsh, M.R. and D.N. Reznick. 2010. “Influence of the indirect effects of guppies on
life-history evolution in Rivulus hartii.” Evolution 64(6): 1583-1593.
Walsh, M.R. and D.N. Reznick. 2011. “Experimentally induced life-history evolution in a
killifish in response to the introduction of guppies.” Evolution 65(4): 1021-1036.
Wantzen, K.M. and R.Wagner. 2006. “Detritus processing by invertebrate shredders: a
Neotropical-temperate comparison.” Journal of the North American
Benthological Society 25(1): 216-232.
Werner, E.E and B.R. Anholt. 1993. “Ecological consequences of the trade-off between
growth and mortality rates mediated by foraging activity.” The American
Naturalist 142(2): 242-272.
Yoshida, T., L.E. Jones, S.P. Ellner, G.F. Fussmann, and N.G. Hairston Jr. 2003. “Rapid
evolution drives ecological dynamics in a predator-prey system.” Nature 424:
26
!
!
303-306.
Zandonà, E., S.K. Auer, S.S. Kilham, J.L. Howard et al. 2011. “Diet quality and prey
selectivity correlate with life histories and predation regime in Trinidadian
guppies.” Functional Ecology 25: 964-973.
!
27
!
!
Table 1. Studies discussing the presence of Phylloicus spp. in Central and South America and their role as shredders in the stream
community.
Location
Species
Ecuador
00°06’N, 78°39’W
Phylloicus sp.
Information
Phylloicus abundance is a major driver of litter breakdown in tropical
montane streams, having a “disproportionate impact”
Brazil
02°25’S, 59°43’W
Phylloicus sp.
Suggests that leaf breakdown in headwater streams is controlled by
shredder abundance: Phylloicus and Triplectides caddisflies
São Paulo State, Brazil
Ribeirão da Quinta stream
23°06’47”S, 48°29’46”W
Phylloicus sp.
Brazil
15°54’S, 55°10’W
Phylloicus sp.
Guanapo watershed,
Northern Range, Trinidad
P. angustior Ulmer 1905
(as syn. P. hansoni
Denning 1983)
Heredia Province, Costa
Rica
10°17’N; 84°02’W
NW Venezuela
10°42’-11°08N, 72°42’72°22’W
Calamoceratidae sp.
Phylloicus sp.
Phylloicus was the only true shredder species found in leaf-pack
samples; abundance may be negatively impacted by riparian
deforestation
Phylloicus is one of the few leaf-feeders in a Neotropical system;
inhabiting “slowly flowing zones near the stream bank, especially in
flood pools” – “localized specialists,” possibly in response to the
effects of current on surface area of leaf case
Phylloicus is one of the largest caddisfly species and are possibly the
only true shredders, but “relatively important;” inhabits 2nd, 3rd and
4th order streams.
Leaf decomposition study: case-building by Calamoceratid larvae
“responsible for marked losses in leaf area” and made up 15.7% of
community in litter bags. One of two insect species found to be
classified as a shredder.
Laboratory feeding trials of different leaves – larvae preferred to feed
on leaves with high N and P content and low content of polyphenols
and lignin, though used leaves with high polyphenol and lignin
content for case construction.
28
!
Source
Encalada et al. 2010
Landeiro et al. 2010
de Carvalho & Uieda 2009
Wantzen & Wagner 2006
Botosaneanu & Sakal 1992
Benstead 1996
Rincón & Martínez 2006
!
Table 2. Five streams were used to collect Phylloicus hansoni larvae for size frequency analysis, each with a downstream KillifishGuppy, or KG, reach and an upstream Killifish-Only, or KO, reach. Values for each stream reach reflect means obtained for wetted
width, depth, percent area of canopy cover, and leaf standing stocks (or, coarse benthic organic matter, “CBOM” mass) along three
transects. Killifish Catch Per Unit Effort (CPUE) is the average of individuals caught from three baited minnow traps within each
reach. Numbers in parentheses represent standard error.
Stream
Reach
RDN
KG
Mean Depth (cm)
% Canopy Cover
100.3 (26.08)
4.0 (1.38)
95.7 (1.80)
0.149 (0.115)
3.0 (3.0)
(7.02)
5.5 (1.85)
94.6 (0.35)
0.413 (0.141)
13.7 (2.0)
KG
244.3 (52.30)
12.1 (3.85)
94.9 (0.46)
0.199 (0.152)
8.3 (1.2)
KO
244.3 (34.70)
12.4 (1.45)
89.7 (2.50)
0.065 (0.003)
10.7 (1.2)
KG
357.0 (44.19)
3.5 (0.05)
84.2 (3.06)
0.176 (0.022)
7.0 (2.0)
KO
195.0 (21.13)
7.8 (2.81)
90.1 (1.82)
0.138 (0.066)
13.7 (5.0)
KG
378.7 (66.80)
8.5 (0.80)
93.3 (0.61)
0.180 (0.037)
13.0 (1.5)
KO
221.0 (62.17)
6.2 (1.69)
94.2 (2.28)
0.245 (0.080)
19.7 (0.7)
KG
141.3 (12.87)
5.7 (2.03)
92.4 (1.50)
1.103 (0.840)
12.0 (5.0)
KO
149.0
3.4 (0.63)
92.4 (1.25)
0.133 (0.068)
19.0 (8.0)
KO
GDC
CED
END
TRT
86.0
(9.29)
29
!
CBOM (g m-2)
Wetted Width (cm)
Killifish CPUE
!
Table 3. Average size measurements of P. hansoni larvae and their leaf cases from each stream (± standard deviation) and mean of
KO and KG reaches (with standard error). Larvae and cases were measured on a dissecting microscope to the nearest tenth of a
millimeter. Case surface area was measured by multiplying the length by the average of three width measurements. Total number of
larvae sampled: KO = 337, KG = 730.
Stream
Reach
Body Length
(mm)
Head Capsule Width
(mm)
Case Length
(mm)
Case Area
(mm2)
RDN
KG
8.7 (3.06)
0.85 (0.23)
15.7 (6.94)
123.1 (108.56)
KO
7.0 (2.96)
0.69 (0.24)
13.0 (6.11)
81.6 (68.70)
KG
3.9 (2.30)
0.49 (0.21)
8.0 (4.53)
35.2 (43.54)
KO
5.4 (1.83)
0.57 (0.19)
9.9 (2.86)
47.3 (25.44)
KG
9.0 (2.79)
0.84 (0.21)
18.4 (7.33)
188.1!(156.97)!
KO
7.7 (2.94)
0.74!(0.27)!
14.0 (5.87)
103.6 (94.16)
KG
7.3 (2.86)
0.78!(0.23)!
15.9 (6.97)
130.0 (107.90)!
KO
7.5 (3.30)
0.77!(0.26)!
16.9 (8.69)
172.0 (183.14)
KG
6.3 (3.21)
0.66 (0.29)
12.6 (6.80)
87.9!!(94.52)!
KO
5.4 (2.97)
0.57 (0.25)
10.8 (6.50)
68.9 (95.17)
KG
7.1 (0.92)
0.72 (0.07)
14.1 (1.78)
112.9 (25.20)
KO
6.6 (0.50)
0.67 (0.04)
12.9 (1.24)
94.7 (21.37)
GDC
CED
END
TRT
Mean
30
!
!
Table 4. Measurements of killifish gut contents from the 2011 site survey, from 4 streams: KG (n=40), and KO (n=52) killifish.
Killifish length and P. hansoni head capsule width are of KG and KO samples; numbers in parentheses represent standard error.!
Stream
Reach
% Killifish
with P. hansoni in gut
[with P. hansoni]
Killifish length (mm)
[Total]
Killifish length (mm)
P. hansoni
head capsule width (mm)
KG
30.0%
39.80 (3.38)
41.82 (3.29)
0.144 (0.0235)
KO
9.6%
42.57 (3.29)
47.83 (1.48)
0.095 (0.0109)
31
!
!
KO reach
KG reach
Figure 1. Representation of fish assemblage of the streams we studied in Trinidad’s
Northern Range: Upstream from barrier waterfalls are reaches with killifish as the only
fish species (Killifish-Only or “KO” reaches), and downstream are reaches with both
killifish (at lower densities) and guppies (Killifish-Guppy or “KG” reaches). Symbols
courtesy of the Integration and Application Network, University of Maryland Center for
Environmental Science (ian.umces.edu/symbols/).
32
!
!
60
% sampled
50
40
30
KG
KO
20
10
0
J
S
M
L
XL
A. hartii size class
Figure 2. Average percent size class composition of killifish caught in minnow traps in
KG and KO reaches of the streams (n=5) sampled. The size classes, measuring body
length are: J – Juvenile: < 15mm, S – Small: 15-30mm, M – Medium: 30-45mm, L –
Large: 45-60mm, and XL – Extra Large: >60mm. Error bars represent standard error.
33
!
!
16
Body length (mm)
14
y = 10.757x - 0.5961
R² = 0.81518
KO
12
10
8
6
4
2
0
0.2
0.4
0.6
0.8
1
1.2
Head capsule width (mm)
16
Body length (mm)
14
KG
y = 11.188x - 0.8914
R² = 0.81104
12
10
8
6
4
2
0
0.2
0.4
0.6
0.8
1
1.2
Head capsule width (mm)
Figure 3. Scatter plot, showing measurements of body length and head capsule width
(mm) of P. hansoni larvae for KO and KG larvae (KO, n = 281; KG, n = 598).
34
!
!
60
60
50
KO
40
KG
30
20
Frequency
Frequency
50
40
30
20
10
10
0
0
2
4
6
8
2
10 12 14 16
RDN - Body length (mm)
50
30
40
25
Frequency
Frequency
4
6
8
10
12
14
GDC - Body length (mm)
30
20
10
20
15
10
5
0
0
2
4
6
8
10
12
14
2
CED - Body length (mm)
4
6
8
10
12
14
END - Body length (mm)
70
Frequency
60
50
40
30
20
10
0
2
4
6
8
10
12
14
TRT - Body length (mm)
Figure 4. Size frequency distributions of larvae from KO and KG reaches of five streams
individually. A two-sample K-S test was used to determine the difference between
distributions of RDN, CED, END, and TRT: p=0.0002, 0.008, 0.93, and 0.036
respectively.
35
!
!
0.85
A
KO
KG
8
7.5
7
6.5
6
Head capsule width (mm)
Body length (mm)
8.5
5.5
KO
B
KG
0.8
0.75
0.7
0.65
0.6
Figure 5. Mean body length (A) and head capsule width (B) of total P. hansoni larvae
collected in KO and KG reaches of four streams (END, RDN, TRT, CED). Error bars
represent standard error.
140
KO
120
KG
Frequency
100
80
60
40
20
0
2
4
6
8
10
12
14
16
P. hansoni body length (mm)
Figure 6. Size frequency distribution of the body length of P. hansoni larvae collected in
KO and KG reaches of four streams (END, RDN, TRT, CED). Individual body length
was measured to the nearest 0.1mm. Total KO = 281, total KG = 598.
36
!