Download Evolution in Response to Direct and Indirect Effects in Pitcher Plant

Florida State University Libraries
Electronic Theses, Treatises and Dissertations
The Graduate School
2010
Evolution in Response to Direct and
Indirect Effects in Pitcher Plant (Sarracenia
Purpurea) Inquiline Communities
Casey P. TerHorst
Follow this and additional works at the FSU Digital Library. For more information, please contact [email protected]
THE FLORIDA STATE UNIVERSITY
COLLEGE OF ARTS AND SCIENCES
EVOLUTION IN RESPONSE TO DIRECT AND INDIRECT EFFECTS
IN PITCHER PLANT (SARRACENIA PURPUREA) INQUILINE COMMUNITIES
By
CASEY P. TERHORST
A Dissertation submitted to the
Department of Biological Science
in partial fulfillment of the
requirements for the degree of
Doctor of Philosophy
Degree Awarded:
Summer Semester, 2010
The members of the committee approve the dissertation of Casey P. terHorst defended
on April 28, 2010.
__________________________________
Thomas E. Miller
Professor Co-directing Dissertation
__________________________________
Don R. Levitan
Professor Co-directing Dissertation
__________________________________
J. Anthony Stallins
University Representative
__________________________________
David Houle
Committee Member
__________________________________
Jeanette L. Wulff
Committee Member
Approved:
_____________________________________
P. Bryant Chase, Chair, Department of Biological Science
The Graduate School has verified and approved the above-named committee
members.
ii
I dedicate this dissertation to all my siblings,
Carole, Tiani, Eric, and Bonnie,
who always make me happy
whenever they are around
iii
ACKNOWLEDGEMENTS
I cannot adequately express my appreciation for my co-advisors, Tom Miller and
Don Levitan, who have given me all of the time I needed whenever I’ve needed
it. Their input, both on science and how to be a scientist, has shaped my career
and will continue to do so for decades.
In addition, the other members of my committee, David Houle, Janie Wulff, and
Tony Stallins, have provided excellent advice throughout the dissertation.
Measuring protozoan traits required a great deal of time in the laboratory and
could not have been accomplished without the assistance of a large number of
people, including Fani Gruber, David Honig, Amy Jenkins, Megan Lowenberg,
John Mola, Emma Moran, Casie Reed, Amber Roman, Christine Stokes, and
Heather Wells.
Emma Moran was particularly helpful in figuring out culturing as well as other
techniques in the lab and participated in many helpful discussions about pitcher
plant communities.
Megan Lowenberg provided a great deal of training on DNA extraction, PCR, and
sequencing in a short period of time. She provided further training by continually
answering my pestering questions over the course of two years. Her presence in
my life has helped to maintain a balance between science and the other things in
life that are crucially important.
Members of the Levitan lab (Mia Adreani, Katie Lotterhos, Nikki Fogarty, Megan
Lowenberg, Pablo Munguia, and Andres Plata-Stapper) have always provided a
fun and supportive environment, in addition to providing feedback on the various
manuscripts that comprise this dissertation.
Many of the ideas presented here are the result of fruitful conversations with
Amanda Buchanan, Elise Gornish, Brian Inouye, David McNutt, David Reznick,
Nora Underwood, and Alice Winn.
This work could not have been accomplished without the stress relief afforded by
those who have participated in intramural sports with me. Sadly, I will leave FSU
without a championship t-shirt, but not without fond memories and a strong
cardiovascular system.
iv
The graduate students and faculty in the Ecology & Evolution program at FSU
provide a nurturing and beneficial environment for graduate students. I have
benefited broadly from interactions with many, including EERDG, and more
covert organizations, such as SSRG and Southwest-Southeast.
This work was financially supported by a Florida State University Dissertation
Research Grant and the Edward and Marie Kingsbury Fellowship from the FSU
English Department. The work was also supported by funding from the National
Science Foundation to T.E. Miller (DEB 0519170 and DEB 0716891) and D.R.
Levitan (DEB 0822626).
v
TABLE OF CONTENTS
List of Tables…………………………………………………………………………...
vii
List of Figures………………………………………………………………………….
viii
Abstract…………………………………………………………………………………
ix
1. INTRODUCTION……………………………………………………………………
1.1 Direct and Indirect Effects………………………………………………..
1.2 Study System……………………………………………………………...
2. EVOLUTION DECREASES ECOLOGICAL EFFECTS OF PREDATORS…..
2.1 Methods…………………………………………………………………….
2.2 Results……………………………………………………………………...
2.3 Discussion………………………………………………………………….
1
2
5
8
9
13
16
3. EVOLUTION IN RESPONSE TO DIRECT EFFECTS OF COMPETITORS…
3.1 Methods…………………………………………………………………….
3.2 Results………………………………………………………………………
3.3 Discussion………………………………………………………………….
20
21
25
30
4. EVOLUTION IN RESPONSE TO INDIRECT ECOLOGICAL EFFECTS……..
4.1 Methods…………………………………………………………………….
4.2 Results………………………………………………………………………
4.3 Discussion………………………………………………………………….
35
37
40
44
5. CONCLUSION………………………………………………………………………
48
APPENDICES
A. SUPPLEMENTAL DATA FOR CHAPTER 2…………………………………….
50
B. GENBANK ACCESSION NUMBERS…………………………………………….
52
REFERENCES…………………………………………………………………………
53
BIOGRAPHICAL SKETCH……………………………………………………………
62
vi
LIST OF TABLES
1.1
Predicted evolutionary responses to direct and indirect effects………………….
4
3.1
Average correlation coefficients among six traits………………………………….
28
vii
LIST OF FIGURES
1.1
Consequences of increasing effect and response on an associate species……..
3
1.2
The abundance of predators, consumers and resources through succession…..
7
2.1
The effect of predators on prey abundance over the course of 12 days………….
14
2.2
Population growth rate after selection (measured in the presence of predators)… 14
2.3
Traits after selection in three environments………………………………………….. 15
2.4
Correlations between population growth rate and cell area………………………… 16
3.1
Averages of six traits from two different selection environments………………….. 26
3.2
Bootstrap consensus tree using parsimony………………………………………….
29
3.3
Correlations between population growth rate and cyst production………………..
32
4.1
Mean values of traits from four different selection environments………………….
41
4.2
Effect sizes of three selection regimes on the values of six traits…………………
42
4.3
Mean of the average Colpdium density in two selection treatments………………
43
4.4
Hypothesized direct and indirect effects of predators and competitors…………..
46
viii
ABSTRACT
The fields of community ecology and evolution are theoretically tightly linked, but in
general, community ecologists discount evolution in describing the dynamics of present-day
community patterns. Yet, evolution in response to strong selection pressure might affect
species interactions and alter ecological patterns on a relatively short time scale. Conversely,
evolutionary studies are typically limited to examining the evolution of traits in response to, at
most, one other species. In more complex communities, higher-order indirect effects emerge
that might have significant effects on how species evolve. Evolution in a community context
provides more insight into how species evole in natural communities. This dissertation focuses
on the evolution of a ciliated protozoan, Colpoda sp. that lives within the water-filled leaves of
the purple pitcher plant (Sarracenia purpurea). The evolution of several traits was measured in
response to the direct effects of predators (specialist mosquito larvae that live in the same
inquilines community), competitors (another co-occuring ciliated protozoan) and indirect effects
that emerge when both predators and competitors are present.
Two traits (cell size and population growth rate) evolved in response to predation,
resulting in predator avoidance and tolerance, respectively. The evolution of these traits
resulted in a significant decrease of the ecological effect of predators in only 12 days
(approximately 40 prey generations), but less than one predator generation. These same two
traits, as well as cyst production, evolved in response to competition. However, evolution in
response to the indirect effects that occurred when both predators and competitors were
present was stronger and in the opposite direction of direct effects. The result was little net
evolution in response to the sum of direct and indirect effects. The importance of these indirect
effects suggests that evolution in a community context may proceed much differently from that
predicted by the simple additive effects of multiple species. To fully understand how species
evolve in natural communities, complex interactions between multiple species must be fully
understood. Similarly, to explain ecological patterns, it is imperative to account for the constant
evolution of species in response to the suite of other species in their environment.
ix
CHAPTER 1
INTRODUCTION
The central goal of ecology is to explain the distribution and abundance of species.
Interactions between organisms and their environment are tightly linked to the evolutionary
history of the species, so it follows that the ecology of a community should be linked to the
evolutionary history of the species in the community. Yet few experimental studies have
incorporated evolutionary theory into explanations of present day community patterns. While
incorporating evolution may be useful to community ecologists, the converse may also be true:
considering evolution in response to multiple species at the same time may alter our view of
how species evolve.
In natural communities, species do not exist in isolation; instead, each species interacts
with many other species, often simultaneously. Several species may intensify, neutralize or
reverse the selection induced by one strongly-interacting species (Silander and Antonovics
1982, Dungan 1986). Conversely, selection due to interactions with many other species may be
so diffuse in multi-species assemblages that the effects of individual species are negligible
(MacArthur 1972, Pianka 1974, Stone and Roberts 1991). Thus, comparing how a species
evolves in isolation to how it evolves in a community context can alter the net rate or direction of
evolution.
A clarification is necessary here on the difference between evolution in a community
context and community evolution. Community evolution often refers to the evolution of
communities as a whole and considers the community as the unit of selection (Wade 1977,
Goodnight 1990, Swenson et al. 2000). In another sense, community evolution may refer to a
change in community structure over time, typically in the sense of ecological succession
(Cowles 1899, Clements 1936, Connell and Slatyer 1977). For example, over ecological time
periods, terrestrial plant communities can change from grasslands to climax communities
composed of hardwood trees (Clements 1936). Neither of these definitions of community
evolution should be confused with the goal of this dissertation to address evolution in a
1
community context. Here, I am concerned with selection on populations of individuals, and in
particular, how the population of one species responds evolutionarily to multiple species in the
community simultaneously. Evolution of two or more species in response to one another is
often referred to as ―coevolution‖ (Futuyma and Slatkin 1983), but that is inappropriate here; I
focus on the evolution of a single species and all other species need not evolve.
1.1 Direct and Indirect Effects
Much of ecological and evolutionary theory is focused on a species or trait response in a
population to either the abiotic environment or to direct interactions with other species. For
instance, competition between two species may result in character displacement over
evolutionary time (MacArthur and Levins 1967, Grant 1972, Dayan et al. 1989). Plants may
evolve increased competitive ability when grown in the presence of competitors for multiple
generations (Miller 1995). Predator-prey or parasite-host relationships can result in an
evolutionary arms race between species (van Valen 1973, Abrams 1990, Brodie and Brodie
1991).
Indirect ecological interactions can arise from chains of direct interactions and may differ
in sign and magnitude from the direct interactions (Strauss 1991, Schmitz et al. 2000, Shurin et
al. 2002). Consider a simple food chain, where carnivores negatively affect herbivores, which
negatively affect primary producers. From this chain of negative effects arise positive indirect
effects between carnivores and producers. As carnivores reduce the number of herbivores, the
impact of herbivores on producers is reduced, thereby increasing the number of producers.
Competition between species, although often considered as a direct effect, is typically an
indirect effect of competitors on one another through a shared resource, except in cases of
direct interference competition. Indirect effects may also occur between non-adjacent species
in competitive hierarchies or within competitive networks (Buss and Jackson 1979, Miller 1994).
As more species are added to a community, the potential for indirect effects increases
exponentially (Abrams 1992).
Indirect effects of one species on another can occur through two mechanisms. Consider
the case where species A affects species C indirectly through species B (Fig. 1.1). The total
effect of one species on another may be decomposed into the per-capita effect and the total
density of that species. Species A may affect the abundance of species B, and thus alter the
total effect of B on C. Such density-mediated effects (Miller and Kerfoot 1987, Strauss 1991)
can be predicted from the direct effects among species. However, higher order interactions
(Vandermeer 1969, Case and Bender 1981, Billick and Case 1994), or trait-mediated
2
interactions (Miller and Kerfoot 1987, Strauss 1991), occur when the presence of species A
affects the per-capita interaction between A and C or B and C. For instance, an herbivore might
hide when a carnivore is present, but forage more actively in the presence of a carnivore if
producer abundance is higher. Such trait-mediated interactions may influence the outcome of
competitive and mutualistic interactions (Trussell et al. 2002, Preisser et al. 2005, Mooney
2006).
Ecologists have long recognized indirect interactions as an important factor in structuring
communities (Vandermeer 1969, Neill 1974, Caswell 1978, Miller and Kerfoot 1987, Strauss
1991, Lawler 1993, Wootton 1994, Miller 1994, Menge 1995). Predator-mediated indirect
effects may allow for coexistence of prey species (Caswell 1978). Apparent competition
between prey species may be important in determining the structure of communities (Holt 1977,
Schmitt 1987, Menge 1995). Community assembly rules may be dictated by indirect effects
(Lawler 1993). Besides the mere existence of indirect effects, several studies have
demonstrated their importance relative to direct effects in determining community structure.
Indirect effects may ameliorate direct effects and prevent the local extinction of species that
would otherwise not persist (Lawler 1993, Miller 1994, van Veen et al. 2005).
(a)
A
(b)
Effect
A
Effect
B
B
Response
C
Response
C
Figure 1.1. The consequences of increasing either (a) response or (b)
effect on direct and indirect effects in a simple community. The indirect
effect is assumed to be positive in (a), but negative in (b). The size of the
arrows indicates the strength of the interaction and the size of the letters
indicates the abundance of each speces. (a) When C increase its
response to A, it reduces the net interaction between A and C, but has little
effect on the abundance of A. Thus the indirect effect is not altered. (b)
When C increases its effect on A, it also reduces the net interaction
between A and C, but also affects the abundance of A and reduces the
indirect effect as well. Modified from Miller and Travis (1996).
3
Since indirect effects play such an important role in ecological interactions, they should
also play a role in evolution (Wootton 1994, Miller and Travis 1996). Miller and Travis (1996)
considered indirect effects as evolutionary forces and concluded that the manner in which
indirect effects influence natural selection is likely to vary from species to species. Consider a
simple three species community with direct and indirect interactions (Fig. 1.1). The species may
be competitors, mutualists, or predators and prey, so long as their interactions can be described
as positive or negative. Over evolutionary time, one focal species (C) can alter its interaction
with an associate species (A) in one of two ways: by changing its response to A or by changing
its effect on A (Fig. 1.1). An increased response alters the interaction between A and C, but has
no demographic effect on A. In the case of two competitors, this typically involves a switch in
resource use, or niche partitioning. In a predator-prey scenario, traits of prey C (e.g. fecundity)
might evolve, such that a prey population can tolerate a particular level of predation by A (i.e.
predator tolerance), but have little effect on the predator. Since the abundance of the associate
species remains unaffected, the indirect effect of A on C (through B) remains unaffected as well
(Fig. 1.1a). Alternatively, C may increase its effect on A. An increased effect may include
increased resource acquisition rates by a competitor, or increased production of anti-predator
defense (e.g. toxins) by prey. Since this interaction alteration will lead to a change in the
abundance of A, it will not only change the direct interaction, but will also affect the indirect
effect of A on C through B (Fig. 1.1b). The nature of the direct and indirect effects allow for
predictions about the direction of evolution of a given focal species (Table 1.1, adapted from
Miller and Travis 1996).
Table 1.1. Predicted evolutionary responses of a focal species
based on the nature of direct and indirect interactions. Adapted
from Miller and Travis (1996).
Indirect Effect
Direct
Effect
Positive
Negative
Positive
Increased
Effect
Increased
Response
Negative
Decreased
Response
Increased
Effect
4
Evolution in a community context requires examining both the evolution of a single trait
in response to different species and the evolution of different traits in response to a single
species. The evolutionary rate or direction of a single trait may change when considered in
response to multiple species. Additionally, the evolutionary response to multiple species may
be constrained by trade-offs among traits (Yoshida et al. 2004). For example, the production of
spines as an evolutionary response to predation may come at a cost of decreased size, which
may affect competitive ability. Such trade-offs may make the traits that don’t evolve in a
community context just as interesting as those that do.
Natural microcosms are ideal for studying the evolution of traits in a community context
because the generation times of many species are very short relative to ecological processes
(Srivastava et al. 2004). Since microcosm communities are often short-lived, ecological
succession occurs relatively quickly and selection pressures change rapidly. Rates of evolution
often occur on ecologically-important time scales (Thompson 1998, Hairston et al. 2005, Losos
et al. 2006). Such evolutionary changes may have a large effect on community dynamics
(Thompson 1998, Strauss et al. 2008). The species found in natural microcosms provide a
unique opportunity to measure not just selection on traits, but the evolution of those traits on a
short time scale. Additionally, protozoans have historically been used to address ecological
questions, but such laboratory experiments often use species from different communities that do
not naturally co-occur. Natural microcosms provide an opportunity to examine evolution with
co-occurring species, making these results applicable to other communities.
The goals of this dissertation were to: (1) determine whether evolution occurs in
response to predation and whether this affects ecological interactions between predator
and prey; (2) determine whether evolution occurs in response to competition; and (3)
determine whether evolution occurs in response to indirect ecological effects and if the
direction of trait evolution matches theoretical predictions (i.e. Table 1.1).
1.2 Study System
The inquiline community found in the water-filled leaves of the purple pitcher plant
(Sarracenia purpurea) provides an excellent system in which to test questions about evolution in
a community context. The ecology of these microcosm communities is well-known (Miller and
Kneitel 2005). The host plant is widely-distributed, extending from northern Florida to the
Northwest Territories. Each plant consists of 1-12 cup-shaped leaves that fill with rainwater and
attract insect prey—primarily ants. The insects drown in the water, and their nutrients (nitrogen
and phosphorous) are presumably used by the plant to allow its survival in highly acidic, nutrient
5
poor environments of bogs and savannahs. In northern Florida, leaves are produced from
March through December at approximately one leaf per plant per month (Miller and terHorst,
unpublished data). Leaves can hold water for over one year, but can develop holes due to
herbivory by grasshoppers or larvae of a specialist lepidopteran herbivore. Newly opened
leaves attract prey using nectar, especially during the first few weeks of opening.
The leaves also attract a community of dipterans, rotifers, protists, and bacteria. At the
core of this community are bacterial decomposers that feed on captured insects, bacterivorous
protozoa and rotifers, and a specialist mosquito (Wyeomyia smithii) that consumes bacterivores.
Species interactions in this community are well-studied (Addicott 1974, Bradshaw and Creelman
1984, Heard 1994, Cochran-Stafira and von Ende 1998, Kneitel and Miller 2002, Miller et al.
2002). Mosquitoes have strong effects in controlling rotifer and protozoa abundances (Kneitel
and Miller 2002, Miller et al. 2002). Increased ant capture has been shown to increase both
bacterial and protozoan abundances and may have a non-linear effect on mosquito larval
abundance (Hoekman et al. 2007). In short, community dynamics are controlled both by topdown effects from mosquito larvae and bottom-up effects due to the number of prey captured by
the plant.
Leaves are colonized by virtually all members of this community within days, if not hours,
of the leaf opening. After colonization, succession in each new leaf appears to be driven by
prey availability and predator abundance, which are in turn largely determined by leaf age.
Leaves capture most of their prey within a few weeks of first opening (Fig. 1.2), with virtually no
prey captured by leaves older than 12 weeks. Most mosquito eggs are also laid in the first few
weeks, after which larval abundances slowly decline as mosquitoes emerge as adults. The
mosquitoes go through four instars and apparently can survive very long periods of low food
availability, in part by not advancing to the next stage.
Changes in dead prey and mosquito numbers through time are predicted to have
different effects on protozoa. In young leaves, resources for protozoa should be abundant, but
mosquito abundances are high (Fig. 1.2). This should lead to high rates of predation, relatively
low competition for the bacterivores. In older leaves, mosquito numbers decline, as do ant
capture rates and bacterial densities, suggesting lower predation but higher resource
competition (Fig. 1.2). The data suggest that protozoan densities are more responsive to
predation than to competition; densities of the three most abundant protozoans and the rotifer,
Habrotrocha rosa, all remain low when mosquito abundances (and food levels) are at their
highest (Fig. 1.2). The goal of this dissertation is to measure the evolutionary response of
6
protozoa traits to direct effects of predators and competitors separately, and then measure the
traits’ response to indirect effects in communities with both predators and competitors.
Figure 1.2. The abundance of predators, consumers, and
resources in pitcher plant communities over the age of the
community. The top graph shows that both predators
(mosquitoes) and resources (dead ants and bacteria)
decline over time. The bottom graph shows that consumer
(protozoa and rotifers) densities increase over time. Data
are from field census by T.E. Miller (unpublished data).
7
CHAPTER 2
EVOLUTION DECREASES ECOLOGICAL EFFECTS OF PREDATORS
Interactions between predators and prey are among the most studied species
interactions in ecology. Predator-prey interactions are important not only for understanding
species population dynamics, but can have cascading effects on other species at lower trophic
levels (e.g., Schmitz et al. 2000, Shurin et al. 2002). Meta-analyses have revealed hundreds of
ecological experiments that document negative effects of predators on prey survival and
abundance (e.g., Englund et al. 1999, Gurevitch et al. 2000) and theory predicts that
overexploitation can lead to instability (e.g. Case 2000). Despite this, predators and prey
commonly coexist in natural communities. Numerous hypotheses have been proposed to
explain predator-prey coexistence, but many of these mechanisms have proven insufficient for
maintaining the coexistence of predators and prey for any significant length of time (Huffaker
1958, Fujii 1999).
A separate body of research has focused on the evolution of traits in prey in response to
predation. Various traits may evolve in response to predator or herbivore selection, including
coloration (Alatalo and Mappes 1996), size (Nunezfarfan and Dirzo 1994), chemical defenses
(Mauricio and Rausher 1997, Shonle and Bergelson 2000), and structural defenses (Reimchen
and Nosil 2002, Seeley 1986). Coevolution between defensive traits of the prey and predator
traits that overcome these defenses (Futuyma and Slatkin 1983) results in an evolutionary arms
race, with predators continuing to have significant negative effects on prey populations (e.g.
Red Queen Hypothesis, van Valen 1973).
A growing body of work demonstrates adaptive evolution on ecological time scales
(Thompson 1998, Hairston et al. 2005, Johnson and Stinchcombe 2007, Strauss et al. 2008).
Slobodkin (1961) defined ecological time as the time period over which populations could
maintain a steady state, thought to be hundreds of generations. Yet Hairston et al. (2005)
defined rapid evolution as trait changes that have the potential to affect the outcome of
simultaneous ecological change, which could occur in relatively few generations. Examples of
8
rapid evolution are important because they suggest that ecologists need to consider ongoing
evolutionary change in order to explain the ecological patterns resulting from species
interactions. However, the time-scale for ecological and evolutionary processes will be different
for different species. For instance, prey species often have shorter generation times than
predators, potentially allowing for faster evolutionary rates in prey relative to predators. This
potential discrepancy in evolutionary rates between predators and prey could reduce ecological
effect sizes and explain coexistence in some natural communities (Strauss et al. 2008). An
increase in the effect of predation may occur in cases where predators evolve faster than their
prey (e.g., arthropods on plants). Despite some indirect evidence, few studies directly link the
magnitude of responses measured in ecological experiments to the evolution of species traits
(Strauss et al. 2008, but see Yoshida et al. 2003, 2007).
Here I examine if the evolution of protozoa found within the leaves of pitcher plants
alters the effect size of mosquito predation on those populations. I tracked the evolution of
several traits in the prey species in communities with and without predation over ―ecological
time‖ (defined here as within a predator’s lifespan) and determined which prey traits evolved
over this time period. I predicted that the rapid evolution of anti-predator traits in prey
populations over ecological time would decrease the effect of predation.
2.1 Methods
Study System
The inquiline community found in the water-filled leaves of Sarracenia purpurea (purple
pitcher plant) has been well-described elsewhere (reviewed in Miller and Kneitel 2005) and so
will be only briefly described here. The host plant is widely-distributed, extending from northern
Florida in the US to the Northwest Territories in Canada, and the community is representative of
other types of natural microcosm communities (Srivastava et al. 2004). Pitcher plants produce
cup-shaped leaves that fill with rainwater and attract insect prey—primarily ants. The energy in
the prey captured by the leaves serves as a basis for a community of bacterial decomposers,
bacterivorous protozoa and rotifers, and a specialist mosquito (Wyeomyia smithii) that
consumes bacterivores. Mosquitoes strongly affect rotifer and protozoa populations, reducing
abundance by 2-5 orders of magnitude (Kneitel and Miller 2002, Miller et al. 2002). Mosquito
larvae develop through four instar levels over the course of several weeks to several months,
depending on food availability (Bradshaw and Johnson 1995).
9
Over the course of 12 days, I measured the ecological effect of mosquito larvae on
replicate populations of a protozoan prey species (Colpoda sp.) in laboratory microcosms.
Colpoda sp. is a ciliated protozoan that is intermediate in size, competitive ability, and predator
tolerance relative to other bacterivores in the community (Leibold and Miller 2004). Colpoda
reproduces asexually every 4-8 hours (Lüftnegger et al. 1990), but some sexual reproduction
may occur (Dunthorn et al. 2008). The experiment encompassed 36-72 prey generations but
only a single larval stage of the predator, allowing for evolution of only the prey. To determine
whether the evolution of prey traits could contribute to a change in predator effect size, I
followed the evolution of several prey traits in a separate set of selection experiments.
Measuring Ecological Effects
Because of practical limitations on the number of replicates in a given time, the
experiment was conducted in two time blocks (October 2008 and January 2009), using different
lines of Colpoda isolated from the field prior to each block to measure the ecological effect of
predators. In each block, samples of water within pitcher plant leaves were collected from
haphazard leaves at each of two sites in the Apalachicola National Forest in northern Florida:
Crystal Bog and Naczi Borrow Pit. The evolutionary history of the collected Colpoda was
unknown, but Colpoda populations in most pitcher plant leaves have almost certainly
experienced some level of predation by mosquito larvae in their recent past (Miller and terHorst,
unpublished data), as both species are frequently found in the same pitchers. In the laboratory,
Colpoda were isolated from all possible samples, generating 3 lines in October and 5 lines in
January, each isolated from an independent leaf. Mixed cultures from individual lines were used
to produce ~300 individuals to inoculate laboratory microcosms (along with the associated
bacteria) that mimic pitcher plants (50 mL plastic macrocentrifuge tubes). Multiple individuals
were used from each line to increase genetic diversity and allow for potential evolution to occur.
Protozoa in these laboratory microcosms have similar population dynamics to those in pitcher
plant leaves in the field (Miller et al. 1994). Each microcosm contained 20mL sterile well water,
plastic beads (2-5 mm diameter), and 6 mg of Tetramin (Tetra Holding, Inc., Blacksburg, VA,
USA) as a food source for the bacteria. The plastic beads served as a refuge from predation for
Colpoda and mimic the detritus and frass found at the bottom of natural pitchers. Mosquitoes
foraged in the beads as well as in the water column, but their size restricted their movement in
the beads.
Replicate populations (n=4 per treatment per month) were maintained with and without
predation (two 3rd instar W. smithii larvae collected from the same sites) for 12 days. Although
10
all mosquitoes survived during the experiment, I attempted to reduce satiation of the predators
and minimize competition among Colpoda by reinitiating the experimental microcosms every 3
days with the same conditions as above, using new mosquito larvae and a random sample of
300 Colpoda from each replicate. On average, a sample of 300 individuals represents 2.3%
and 1.7% of the total population with and without predators, respectively. Every day, Colpoda
abundance in each population was estimated by counting a 0.1mL subsample on a Palmer
counting cell (Wildlife Supply Company, Buffalo, NY, USA) and the effect of predation was
determined using the mean within each time block as:
Predator effect = (abundancemonoculture – abundancepredation)/abundancemonoculture
There are minor costs and benefits to using either Tetramin or dead ants as a food
source, though both result in similar protozoa population dynamics. Tetramin can be added
more precisely, resulting in less variance among populations, but ants are a more natural food
source for bacteria in these communities. In this experiment I chose Tetramin, because I was
concerned with reducing variance in the food source over time in order to test changes in effect
size over time. In the second experiment (see Selection experiment below), I chose dead ants
to examine if traits would evolve differently across treatments given natural variation in food
quality.
Selection Experiments
In a separate set of experiments, I followed the evolution of traits in field-collected
Colpoda in laboratory microcosms. In these selection experiments, replicate populations of
Colpoda were exposed to predation by mosquito larvae for six days (18-36 prey generations) in
six time blocks: February, April, May, and July of 2006, and March and June of 2007. In each
time block, new independent lines of Colpoda were collected from the same field sites as above.
Each experimental microcosm contained 30mL of sterile DI water, 5 dead sterile fire ants
(Solenopsis invicta) as a food source for bacteria, 5mL of sterile plastic beads and was
inoculated with ~300 Colpoda composed of an equal mixture of all independent lines and the
associated bacteria. Two 3rd instar mosquito larvae were added to half of the tubes. Sample size
varied among time blocks (n=4 in February 2006, n=5 in April 2006, n=10 in all other months)
and was further reduced in some months because some cultures failed to grow sufficiently well
to measure traits (actual sample sizes given in figures).
For both experiments, cultures were maintained in a growth chamber with a diurnal cycle
(12/12) of light and temperature (day=30°C, night=20°C). At the end of the selection
experiment, microcosms were well-mixed and ~300 Colpoda were removed from replicates in
11
both predator and no-predator treatments and placed into identical no-predator microcosms
containing 30 mL of sterile DI water and 5 dead sterile ants. Colpoda were grown under these
common conditions for 24 hours to minimize the role of environmental plasticity and maternal
effects before measuring traits. After 24 hours (approximately 3 generations), two random
subsamples of 300 individuals were taken from each replicate. The first was used to determine
the degree to which each replicate was adapted to predation by measuring population growth
rate as a proxy measure of fitness in the presence of a predator (except in February 2006).
Replicates with higher growth rates in the presence of two 3rd instar mosquito larvae were
deemed to be more adapted to predation. The second subsample was used to measure traits in
a ―common garden‖ without predation. Any difference in traits between treatments was assumed
to be due to evolved genetic differences.
Population growth rates were measured by growing each subsample of 300 individuals
in microcosms as above. Population size was estimated by counting the number of individuals
in a 0.1 mL random subsample of each culture every six hours for 48 hours, or until populations
peaked in abundance. Populations displayed exponential growth up to a peak and then
abundance dropped off rapidly. Growth rates were determined by estimating r from the best-fit
exponential growth model (N0 = N0ert), using abundance data up to the peak. In June 2007,
three additional traits were measured in each replicate: cell size, swimming speed, and refuge
use. Digital videos and image analysis (NIH Image J, http://rsb.info.nih.gov/ij/) were used to
estimate cell area (mm2) and swimming speed (mm/s) of the first 30 individuals encountered in
the video. Refuge use was determined by removing two 0.1 mL samples from each microcosm,
one in the water column and one in the beads, and determining refuge use as: log(# in beads / #
in water). This measure of refuge use provides an unbounded index, with a value of 0 indicating
equal densities in the beads and water.
Predators may select for prey traits in two ways: directly, by selectively killing organisms
with particular traits, or indirectly, by reducing prey densities and intraspecific competition
(Schröder et al. 2009). In March and June of 2007, I used a third selection treatment to control
for prey densities and attempted to elucidate the mechanism of predator selection. In this
density-reduction treatment, densities were reduced every 24 hours to the median density in the
predator-treatment. An appropriate amount of liquid was removed from each tube, containing a
random sample of Colpoda, and replaced with the same amount of sterile water.
Statistical analyses
All data met the assumptions of parametric tests and statistical analyses were performed
in JMP version 4.0.4 (SAS Institute Inc., USA, 2001). Linear regression was used to determine
12
whether predator effect size changed over the course of the ecological effects experiment. Data
within time blocks were tested for autocorrelation in JMP to ensure that they were suitable for
linear regression. All ANOVAs described below were unrestricted two-way mixed model ANOVA
accounting for unbalanced design (Quinn and Keough 2003) comparing selection environment
(no predator vs. predator) as a fixed factor and time block as a random factor. Population
growth rates in predator environments were compared among predator and no-predator
selection lines using ANOVA to examine differences in adaptation to predation. A second
ANOVA compared population growth rate using all three selection treatments (no predator,
predator, and density reduction). To determine which traits evolved in different selection
environments, I used a series of ANOVAs (one per trait) to test for differences in trait values
(measured in the no-predator common environment) between populations evolved with and
without a predator. Tukey post-hoc tests were used to determine significant differences among
pair-wise treatment levels. An additional single factor ANOVA for each trait was used to
compare trait values in all three selection treatments.
Traits are often not independent of one another and trade-offs among traits lead to the
evolution of one trait in response to selection on another. To test for such effects, I used
Pearson product-moment correlations to determine if there were any significant relationships
between each pair-wise combination of traits in June 2007.
2.2 Results
Initially, predators had a negative ecological effect on prey abundance, reducing the
abundance by approximately 50% (Fig. 2.1). This effect decreased significantly over the course
of 12 days. By the end of the experiment, prey abundance in some replicates was higher in
populations with a predator relative to control populations, consistent with an evolved response
of Colpoda that decreased predator effect size. Effect size data were not autocorrelated with
respect to any time lags.
In the selection experiment, populations that evolved with a predator had higher
population growth rates in a predator environment than populations that evolved without a
predator (F=16.2, df=1,4, P=0.013, Fig. 2.2). There was no effect of time block on population
growth rate (F=3.03, df=4,4, P=0.15), nor was there an interaction between time block and
selection environment (F=1.69, df=4,74, P=0.16). In the density reduction treatment, Colpoda
grown with predators exhibited population growth rates intermediate to that in the predator and
no-predator treatments (Appendix A).
13
0.6
Predator Effect Size
0.4
0.2
0
-0.2
0
2
4
6
8
10
12
-0.4
-0.6
-0.8
R2=0.26
P=0.012
-1
Day
Population Growth Rate (r)
Figure 2.1. The effect of predators on prey abundance over the course of 12 days in two time
blocks: October (open diamonds) and January (filled diamonds). Effect size is the abundance in the
predator treatment relative to the monoculture treatment. Negative values indicate a negative
effect of predation on prey abundance.
Evolutionary History
0.3
No Predator
6
0.25
Predator
5
0.2
0.15
9
10
10
4
10
10
10
10
0.1
0.05
0
April '06
May '06
July '06
March '07
June '07
Figure 2.2. Population growth rate after selection (measured in the presence of predators) of
populations evolved without predation (white bars) and populations evolved with predation by
mosquito larvae (gray bars) in five independent experiments. Bars indicate means 1 s.e.
Numbers above bars indicate the sample size for each.
14
Trait evolution was measured in a common, no-predator environment. Population growth
rate (assumed to be correlated with cell division rate) was again significantly higher in
populations that evolved with a predator (F=9.76, df=1,5, P=0.025, Fig. 2.3A). There was no
significant difference in growth rate among time blocks (F=2.72, df=5,5, P=0.15). There was a
significant block*treatment interaction (F=9.83, df=5,73, P<0.001) because the magnitude of the
effect varied among time blocks. However, the mean growth rate in each block was always
higher in the populations that evolved with a predator (Appendix A). Cell area decreased
significantly in populations that evolved with a predator (Fig. 2.3B). In the density reduction
treatment, population growth rate and cell area were intermediate to the predator and nopredator treatments and not significantly different from either the predator or no-predator
treatment (Fig. 2.3). There was no difference in swimming speed (Fig. 2.3C) or refuge use (Fig.
2.3D) among treatments.
Population Growth Rate (r)
0.2
C
Swimming Speed (mm/s)
b
A 0.25
a,b
a
0.15
0.1
0.05
0
a
a
a
0.4
0.38
0.36
0.34
0.32
No Predator
B
0.42
Density Reduction
Predator
6
No Predator
D
a
1
Density Reduction
a
a
a
0.8
a,b
b
5
Refuge Use
5.5
Cell Area (mm2)
Predator
0.6
0.4
4.5
0.2
0
4
No Predator
Density Reduction
Predator
No Predator
Density Reduction
Predator
Figure 2.3. Traits after selection in three environments, measured in a common environment
without predators: a) population growth rate, b) cell area, c) swimming speed, and d) refuge use. A
value of zero for refuge use indicates no preference for the refuge. Positive values indicate
increased use of the refuge. Bars indicate means 1 s.e. Letters indicate significant pairwise
differences among treatment means (p<0.05, using Tukey post-hoc tests).
15
All pair-wise correlations between trait values were non-significant (P>0.05). Marginally
significant correlations were found between population growth rate and refuge use (r=0.34,
P=0.066) and cell size and swimming speed (r=0.32, P=0.086). The two traits that evolved in
response to predation (population growth rate and cell size) were the least correlated traits
among all comparisons (r=0.004, P=0.99, Fig. 2.4).
8
No Predator
Density Reduction
Cell Area (mm2)
7
Predator
6
5
4
3
0.1
0.15
0.2
0.25
Population Growth Rate (r)
Figure 2.4. Correlations between population growth rate and cell area in
populations from three selection environments. Overall, there is no correlation
between the two traits, nor is there any correlation within environments [no
predator (r=0.09, p=0.81), density reduction (r=-0.05, p=0.90), predator (r=0.48,
p=0.16)].
2.3 Discussion
Evolutionary processes have generally been ignored in ecological experiments.
However, I have demonstrated that the evolution of prey traits is associated with a qualitative
and significant decrease in the ecological effect of predators on prey populations over the
course of a relatively short experiment (Figs. 2.1, 2.2). While evolution on ecological time scales
16
has been recently demonstrated in other systems, I note that the discrepancies in generation
times between predator and prey, and perhaps between rates of evolution, may be particularly
important for explaining ecological patterns. At the beginning of the experiment, predators had a
large negative effect on protozoa abundance, reducing abundance by over 50% relative to
control populations. Yet a short time later, prey reproduction was similar to the consumption
rate of the predators.
Although there is an overall decrease in the ecological effect of predators over time, I
note a ―saw-tooth‖ pattern in the data that repeats every three days (Fig. 2.1), corresponding to
each time the tubes were reinitiated with Colpoda, mosquito larvae, and the associated bacteria.
Two factors might explain this result: bacterial evolution or predator satiation. Bacterial
evolution, although likely to occur, is unlikely to explain the observed pattern, which would
require an increase in bacteria beneficial to the Colpoda population. However, consumption by
protozoa should select against these bacteria. Predator satiation may explain the decrease in
effect over each three-day period, since new mosquito larvae were introduced every three days.
However, the most likely explanation for the overall decrease in effect over the span of the
experiment is evolution of the protozoa prey.
Evolution of specific traits reduced the ecological effect of predators by both increasing
intrinsic population growth rates and reducing consumption through a size refuge (see below).
These results are similar to those from previous studies in aquatic microcosms, which also
suggest that prey populations can evolve predator resistant traits that mask the rate of predation
(Meyer et al. 2006, Yoshida et al. 2007). Evolution of such traits may affect the population
dynamics of both predator and prey and influence coexistence (Shertzer et al. 2002, Yoshida et
al. 2003).
The evolution of cell size and population growth rate conferred adaptations to predation
by different mechanisms: predator resistance and predator tolerance. Predator-resistant traits
are those that decrease consumption and have negative effects on the predator, such as toxin
production, refuge use, or in our case, body size. Mosquito larvae in pitcher plants prefer larger
prey items, such as rotifers (Habrotrocha rosa), over smaller protozoa such as
Poteroochromonas or Bodo (Kneitel 2002, Leibold and Miller 2004). Selective predation by
mosquito larvae likely drove the evolution of smaller cell sizes in our experiment (Fig. 2.3B).
Conversely, predator-tolerant traits are those that do not reduce consumption and have little
effect on the predator, but rather allow the population to compensate for losses to predation. In
this study, selection for rapidly-dividing genotypes may have resulted in the evolution of higher
population growth that allowed the prey to keep pace with the rate of predation (Fig. 2.3A,
17
Appendix A). I do not know all the individual traits that determine the growth rate of the
population. Predator-resistant traits, for instance, may confer higher population growth rate in
the presence of a predator, but I also observed higher growth rates in the absence of predators.
Since population growth rate was uncorrelated with cell size (Fig. 2.4) or any other trait (P >
0.10), an increase in cell division rate seems the most parsimonious explanation for the
increase in population growth rate.
Predator (or herbivore) tolerance traits, such as compensatory growth, have been wellstudied in terrestrial plants and are thought to be constrained by trade-offs with predator
resistance traits, such as physical or chemical defense (Coley et al. 1985, Mole 1994, Zangerl
and Berenbaum 1997, Messina et al. 2002). Conversely, a common relationship between size
and age at maturity predicts that I might find a positive association between tolerance (high
growth rate) and resistance (small cell size) in this system—if cells divide before they reach their
maximum size, quickly dividing cells should be smaller than slowly dividing cells. However, I
found no support for either hypothesis in our Colpoda populations (Fig. 2.4). Note however that
a trade-off between traits may be concealed by the different levels at which I measured such
traits (growth rate at the population level vs. size at the individual level). If different individuals
solve a trade-off restriction differently (e.g. one clone grows fast, but another clone grows big),
this will not be evident at the population level.
The density-reduction treatment was intended to distinguish between two mechanisms
by which predators might alter the evolution of traits in prey populations—direct selection or
indirect effects on intraspecific competition (Schröder et al. 2009). However, the power of the
design was insufficient to distinguish among these alternate hypotheses. This ambiguous result
could have been influenced by both direct and indirect factors operating simultaneously or
because the frequency and intensity of the density reduction was less than the predator effect.
Ongoing work will attempt to tease apart these effects. Regardless of the mechanism,
predation significantly alters the evolution of prey.
Experiments in the laboratory often differ from natural conditions. Populations with
different evolutionary histories may respond differently when placed in a ―common garden‖ in
the laboratory and I cannot entirely dismiss this as a possibility in explaining our results. The
evolution of at least two traits in Colpoda (cell size and growth rate) conferred adaptation to
predation in the laboratory, but these same traits may not be adaptive in all environments.
Evolution of predator-resistant traits often comes at the cost of lower competitive ability (e.g.,
Kerfoot 1977, Wulff 2005). Previous work demonstrated a trade-off between predator tolerance
and competitive ability among bacterivorous species in pitcher plant inquiline communities
18
(Kneitel 2002, Leibold and Miller 2004). Species composition of these communities changes
dramatically with leaf age. Mosquito larvae are most abundant in young leaves (younger than 10
weeks; Miller and Kneitel 2005), but both larval and bacterial abundance decline with leaf age
as bacteriovore abundance increases, resulting in more competition for bacteria (Miller and
terHorst, in prep). It is unknown whether the evolution of predator tolerance and resistance traits
is maladaptive in competitive communities or whether Colpoda is able to adapt via evolution or
plasticity to these conditions.
Experimental manipulations often have unintended consequences on ecological
experiments. For example, recognition that phenotypic plasticity may alter outcomes ensures
that common gardens are regularly used in experiments. Likewise, strong selection imposed by
ecological experiments requires that ecologists incorporate controls for evolutionary effects in
ecological experiments (Strauss et al. 2008), but this is rarely done. These results demonstrate
that evolution can have large effects on the outcome of ecological experiments. These results
need not be limited to microcosm communities. Treatments in many ecological experiments
impose strong selection on the manipulated population, which may result in evolution even in
experiments over only a few generations (Strauss et al. 2008). Species with long generation
times may not be exempt from evolutionary effects. Although longer-lived predators were unable
to evolve over the course of our experiment, their population dynamics are likely to be affected
by evolution of their food source. Discrepancies in generation times are not unique to
microcosm communities. For instance, many plants have short generation times relative to
mammalian herbivores, but long generation times relative to arthropod herbivores.
Discrepancies in generation times increase when one considers other associate species such
as microbial symbionts, mycorrhizal fungi, or pathogenic bacteria or viruses. Evolution in any
associate species may affect ecological measurements on the focal species. I caution
experimental ecologists to begin recognizing evolution as a possible explanation of ecological
results and to incorporate appropriate controls into experimental protocols.
19
CHAPTER 3
EVOLUTION IN RESPONSE TO DIRECT EFFECTS OF COMPETITORS
Competition among species for resources is known to be important to the ecology and
evolution of species. Hundreds of ecological experiments have documented that interspecific
competition limits the distribution and abundance of species in natural communities (Connell
1983, Schoener 1983, Goldberg and Barton 1992, Gurevitch et al. 1992). There is considerably
less evidence of the evolutionary implications of such competition. Decades of research on
character displacement invoked competition as the likely mechanism driving the divergence of
species traits (Brown and Wilson 1956, Simberloff and Dayan 1991, Moen and Wiens 2009).
However, character displacement has often been difficult to confirm (e.g. Strong et al. 1979) and
character displacement alone does not implicate competition as the driving evolutionary force
(Connell 1980). Further, theory suggests that competition is just as likely to drive convergence
of traits as it is to cause divergence of traits (Aarssen 1983, Fox and Vasseur 2008, C. P.
terHorst and T. E. Miller, unpublished simulations).
Demonstration of the evolutionary effect of competition is best accomplished with
experiments that compare the evolution of several traits in competitive environments relative to
evolution in control environments. However, there are very few experimental studies of
evolution due to competition, so a general statement about the evolutionary effect of competition
is difficult. Studies of intraspecific competition have often found significant evolutionary changes
in traits such as size and growth rate (e.g. Santos et al. 1997, Sokolowski et al. 1997, Stanton et
al. 2004). There are fewer examples of trait evolution in response to interspecific competition
(but see for example, Schluter 1994, Smith and Rausher 2008, McGoey and Stinchcombe
2009) and these have focused largely on plant taxa. Others have documented significant
changes in competitive ability as a result of evolution, but have not found significant evolution of
traits that might cause such a change (e.g. Miller 1995, Goodnight and Craig 1996, Joshi and
Thompson 1996).
20
Historically, protozoa have proven to be useful model systems in ecological competition
experiments (e.g. Gause 1934, Vandermeer 1969, Cadotte et al. 2006), but evolution of traits in
these taxa has rarely been examined. Here, I used naturally co-occurring protozoa species
from the inquiline community of purple pitcher plant (Sarracenia purpuea) leaves to examine the
evolution of species traits in response to competition. The evolution of six traits in one species
were measured in response to both intra- and interspecific competition and compared to the
traits of the competitor to determine whether species traits converged or diverged.
3.1 Methods
Study System
The inquiline community found in the water-filled leaves of Sarracenia purpurea (purple
pitcher plant) has been well-described elsewhere (reviewed in Miller and Kneitel 2005) and so
will be only briefly described here. The host plant is widely-distributed, extending from northern
Florida in the US to the Northwest Territories in Canada, and the community is representative of
other types of natural microcosm communities (Srivastava et al. 2004). Pitcher plants produce
cup-shaped leaves that fill with rainwater and attract insect prey—primarily ants. Prey captured
by the leaves serves as a basis for a community of bacterial decomposers, bacterivorous
protozoa and rotifers, and a specialist mosquito (Wyeomyia smithii) that consumes bacterivores.
Previous work has demonstrated that evolution of a particular suite of protozoan traits evolves in
response to predation by mosquito larvae (terHorst et al. 2010). Competitors can have large
ecological effects on one another in this community, reducing bacterivore abundances by 30100% (Kneitel 2002), but their effects on trait evolution have never been demonstrated.
Selection experiment
The evolutionary effects of competition were measured in replicate populations of
Colpoda spp., a group of unicellular ciliated protozoa that are intermediate in size, competitive
ability, and predator tolerance relative to other bacterivores in the community (Kneitel 2002,
Leibold and Miller 2004). Because of practical limitations on the number of replicates, the
experiment was conducted in three time blocks (October 2008, April 2009, and June 2009),
using unique lines of Colpoda from the field in each block. Samples of water within pitcher plant
leaves were collected from haphazard leaves at each of two sites in the Apalachicola National
Forest in northern Florida: Crystal Bog and Naczi Borrow Pit. In each time block, Colpoda were
collected from 4 leaves to create 4 independent stocks, that were later mixed to create a pool of
21
genotypes in each replicate. Stocks from independent leaves were later genotyped to determine
species and clonal identity (see below).
Approximately 300 individuals from the mixed-genotype stock culture were used to
initiate a selection experiment by inoculating replicate laboratory microcosms (along with the
associated bacteria) that mimic pitcher plants (50 mL plastic macrocentrifuge tubes). Protozoa
in these laboratory microcosms have similar population dynamics to those in pitcher plant
leaves in the field (Miller et al. 1994). Each microcosm contained 20 mL sterile well water,
plastic beads (2-5 mm diameter), and 6 mg (± 0.15 mg) of Tetramin (Tetra Holding, Inc.,
Blacksburg, VA, USA) as a food source for the bacteria. The plastic beads mimicked the
detritus and frass found at the bottom of natural pitchers (terHorst et al. 2010).
Replicate populations (n=4 per treatment per block) were maintained in (1) monoculture
with only intraspecific competition and (2) two-species cultures with both intra- and interspecific
competition (hereafter referred to as Monoculture and Competition, respectively) for 20 days.
Asexual reproduction in Colpoda occurs every 4-8 hours (Lüftnegger et al. 1990), but some
sexual reproduction may occur infrequently (Dunthorn et al. 2008). Selection experiments were
run for 20 days, which encompassed 60-120 Colpoda generations. Another species of ciliated
protozoan, Colpidium sp., is known to be competitively dominant over Colpoda (Kneitel 2002).
In the Competition treatment, Colpidium was added at twice the density (~600 individuals) of
Colpoda. To avoid competitive exclusion of Colpoda by Colpidium and to prevent co-evolution
between these species, I used serial transfers to reestablish the experimental microcosms in
both treatments every 3 days with the same conditions as above, using a random sample of 300
Colpoda from each replicate and Colpidium from a laboratory stock culture.
All cultures were maintained in a growth chamber with a diurnal cycle (12/12) of light and
temperature (day=30°C, night=20°C). At the end of the selection experiment, microcosms were
well-mixed and ~300 Colpoda were isolated and removed from replicates in both Monoculture
and Competition treatments and placed into identical environments containing only 20 mL of
sterile well water and 6mg of Tetramin. Colpoda were grown under these common conditions
for 24 hours (3-6 generations) to diminish the role of phenotypic plasticity, due to environmental
or maternal effects, before measuring traits.
Measurement of Traits
After growth in a common environment for 24 hours, two random subsamples of 300
individuals were taken from each replicate in both Monoculture and Competition treatments and
used to measure traits in two common garden environments: monoculture and competition. I
will refer to the environment in which selection occurred for 20 days as the Selection
22
Environment and the environment in which traits were measured as the Measurement
Environment. In each replicate from both selection environments, six traits of Colpoda were
measured in both measurement environments: cell size, cell speed, cyst production, peak
population abundance, population growth rate, refuge use.
Digital videos and image analysis (NIH Image J, http://rsb.info.nih.gov/ij/) were used to
estimate cell area (mm2) and swimming speed (mm/s) of the first 30 individuals encountered in
the video. Refuge use was determined by removing two 0.1 mL samples from each microcosm,
one in the water column and one in the beads, and counting the number of individuals on a
Palmer counting cell. Refuge use was determined as: log(# in beads / # in water). This
measure of refuge use provides an unbounded index, with a value of 0 indicating equal
densities in the beads and water.
Population growth rates were measured by growing each subsample of 300 individuals
in microcosms with 20 mL sterile well water and 6 mg Tetramin. In the competition
measurement environment, 600 Colpidium were added to each microcosm. Population size
was estimated by counting the number of individuals in a 0.1 mL random subsample of each
culture on a Palmer counting cell every six hours, for 48 hours, or until populations peaked in
abundance. Populations displayed exponential growth up to a peak and then abundance
dropped off rapidly. Growth rates were determined by estimating r from the best-fit exponential
growth model (N0 = N0ert), using abundance data up to the peak. Colpoda regularly produces
visible resting cysts that are resistant to dessication, and presumably are immune to competition
and predation as well. Cyst production was estimated by counting the number of cysts in a 0.1
mL subsample from each microcosm after six days.
Additionally, traits of Colpidium were measured, but only in the measurement
environment with competition. Cell size, cell speed, population growth rate, and refuge use of
Colpidium populations were measured in the same way as Colpoda populations above.
DNA extraction, amplification, purification, and sequencing
To determine the genetic identity of Colpoda lines used in the experiment, 7 of the
original 13 stocks, along with two additional stocks collected from the same sites in August
2009, were recovered in February 2010 and used to extract DNA for sequencing. DNA was
extracted from 100 microliters (μL) of stock cultures varying in density, from 400 – 11,000
cells/mL, using SprintPrep® magnetic beads (Beckman Coulter, Brea, CA, USA). Each sample
was loaded into a 96-well round bottom plate, mixed with 80μL 100% isopropanol and 10μL of
SprintPrep beads, and then placed on a SPRIP 96R® ring magnetic plate (Beckman Coulter) for
5 minutes. After the supernatant was drained, samples remained on the magnetic plate through
23
5 rinses (30 seconds each) with cold 70% ethanol. Samples were allowed to dry for 30 minutes
to ensure evaporation of the ethanol. The round-bottom plate was removed from the magnetic
plate and 15μL of ddH20 were added to each sample and the plate was placed on a shaker for
one hour to unbind the DNA from the magnetic beads. Samples were again placed on the
magnetic plate for 5 minutes to separate the beads from the DNA in the supernatant. All
supernatant was removed and DNA concentration was measured using a Nanodrop
spectrophotomer (Thermo Scientific, Wilmington, DE, USA).
A portion of the 18S rRNA gene was obtained using polymerase chain reaction (PCR)
with protocols modified from Spears and Abele (2000). 100μL reactions were composed of:
53.5μL ddH20, 16μL dNTPs (10μM), 8μL MgSO4 (25mM), 10μL PlatinumTaq buffer, 0.5μL DNA
Polymerase High Fidelity PlatinumTaq (Invitrogen, Carlsbad, CA, USA), 5μL 329 forward primer
(20μM), 5μL HI reverse primer (20μM), and 2μL extracted DNA (5ng/μL). Primers were those
amongst those used by Spears et al. (1992): 329 forward primer: 5': TAA-TGA-TCC-TTC-CGCAGG-TT:3'; HI reverse primer: 5’:GTG-CAT-GGC-CGT-TCT-TAG-TTG:3’. PCR was carried out
on a Veriti 96-well Thermal Cycler (Applied Biosystems, Foster City, CA, USA) using the
following profile: 5 min @ 95°C; 35 cycles of 40s @ 95°C, 25s @ 52°C, 3 min @ 72°C, and a
final extension of 10 min @ 72°C. PCR product was purified using the same magnetic bead
protocol used for extraction. PCR products were sequenced at the DNA Sequencing Laboratory
in the Biological Science Department at Florida State University using a 3130x1 Genetic
Analyzer with Capillary Electrophoresis (Applied Biosystems).
DNA sequences were used to create a phylogenetic tree in PAUP (Version 4.0), using
reference sequences from Genbank at the National Center for Biotechnology Information (see
Appendix B for accession numbers) from similar species within the Colpodida (Ciliophora:
Colpodea), and three outgroups (Bryometopus pseudochilodon, Cyrtolophosis mucicola, and
Oxytricha nova, all obtained from the phylogeny of Dunthorn et al (2008). A bootstrap
consensus tree was created using parsimony, retaining up to 1000 trees in each of 1000
bootstrap replicates. All new sequences obtained in this study have been uploaded to Genbank
(see Appendix B for accession numbers).
Statistical analyses
All data met the assumptions of parametric tests or were transformed to meet these
assumptions and statistical analyses were performed in JMP version 4.0.4 (SAS Institute Inc.,
USA, 2001). Three-way mixed model ANOVAs, using selection environment and measurement
environment as fixed factors and time block as a random factor, were used to separately test for
differences in each of the six traits. Maximum likelihoods were calculated for separate linear
24
models with and without all combinations of interactions between main effects. Akaike’s
Information Criterion (AIC) was used to find the best-fit model, which was used for statistical
analysis of each trait.
Traits are often not independent of one another and trade-offs among traits can lead to
the evolution of one trait in response to selection on another. As only population averages were
measured for most traits, the distribution of traits within populations were not available, and
normal quantitative genetic approaches were inappropriate (e.g. Lande and Arnold 1983).
Instead, Pearson product-moment correlations were used to determine if there were any
significant relationships between each pair-wise combination of traits. To account for multiple
comparisons among traits, the data were bootstrapped 10,000 times in R (The R Development
Core Team), maintaining pairs of data during bootstrapping, to obtain mean correlation
coefficients and 95% confidence limits. Correlations between traits were considered significant
if the 95% confidence interval did not include zero.
3.2 Results
Three traits showed significant evolution due to competition: population growth rate, cell
size, and cyst production. Population growth rate of Colpoda was significantly higher in
populations that had evolved in the competition selection environment than those that evolved in
the monoculture environment (Fig. 3.1A, F1,41 = 10.2, P = 0.003). The environment in which
traits were measured had no significant effect on population growth rate (Fig. 3.1A, F1,2 = 0.029,
P = 0.88). There was also no effect of time block on population growth rate (F2,2 = 4.3, P =
0.19). The best-fit model included a significant interaction between time block and
measurement environment (F2,41 = 3.84, P = 0.03). Population growth rates of Colpidium were
more difficult to measure because populations had not peaked within 48 hours. Fitting an
exponential curve to the available data may overestimate population growth rate, but gave a
population growth rate of 0.086 (± 0.006 s.e.)—considerably lower than the population growth
rate of Colpoda (Fig. 3.1A).
Selection environment had a significant effect on cell size (F1,2 = 113, P = 0.009).
Individuals that evolved with competition were significantly smaller than those that evolved in
monoculture (Fig. 3.1B). Measurement environment had the opposite effect on cell size (F1,2 =
41.7, P = 0.023). Cells that were measured in competition were significantly larger than those
that were measured in monoculture (Fig. 3.1B). Time block also had a significant effect on cell
size (F2,2 = 29.5, P = 0.033). The smallest cells were found in the October 2008 experiments,
while the largest cells were found in June 2009, with intermediate sizes in April 2009. The best25
fit model included all interactions, but the two-way interactions were non-significant (P > 0.10).
There was a significant three-way interaction (F2,36 = 7.75, P = 0.002) indicating that the strength
of the Selection*Measurement Environment interaction varied among blocks. The three-way
interaction was due to a change in magnitude and not direction among blocks. Cells selected in
competition were always smaller than those in monoculture, and cells measured in competition
0.2
A
Refuge use
0.16
0.12
0.08
0.04
0
0.5
0
-0.5
Competition
BMeasurement Environment
2
1
Monoculture
Population Peak
Cell Area (x10-3 mm2)
3
D
-1
Monoculture
350
Competition
EMeasurement Environment
300
250
200
150
100
50
0
0
Monoculture
# cysts per 0.1mL
1
80
Monoculture
Competition
CMeasurement Environment
60
40
20
Cell Speed (mm/s)
Population Growth Rate
Selection Environment
Monoculture
Competition
0.6
Competition
FMeasurement Environment
0.4
0.2
0
0
Monoculture
Monoculture
Competition
Competition
Measurement Environment
Measurement Environment
Figure 3.1. Averages (± s.e.) of six traits of replicate Colpoda populations from two different selection
environments (monoculture: white bars; interspecific competition (gray bars), measured in two different
common garden environments (monoculture and competition). In D, positive values indicate greater density
in the refuge; negative values indicate greater density in the water column.
26
were always larger than those measured in monoculture, but the degree to which they were
different varied among time blocks. The average cell size of Colpidium (mean cell area = 9.5 x
10-4 mm ± 4.6 x 10-5) was smaller than Colpoda (Fig. 3.1B).
Selection environment has a significant effect on cyst production (F1,41 = 6.46, P =
0.015), as populations of Colpoda that evolved with competition produced significantly fewer
cysts than those that evolved in monoculture (Fig. 3.1C). Neither measurement environment
(Fig. 3.1C, F1,2 = 0.46, P = 0.57) nor time block (F2,2 = 10.0, P = 0.09) had a significant effect on
cyst production. Although the best-fit model included a block*measurement environment
interaction, it was non-significant (F2,41 = 1.47, P = 0.24).
Selection environment had no significant effect on Colpoda refuge use (F1,43 = 2.33, P =
0.13), nor did measurement environment (F1,43 = 3.05, P = 0.09). There was a trend for
populations of Colpoda that evolved in competition to use the refuge more heavily, but only in
the monoculture measurement environment (Fig. 3.1D). Time block had a significant effect on
refuge use (F2,43 = 7.32, P = 0.002), as Colpoda used the refuge more heavily in October and
June than in April. The best-fit model included only these main effects. Conversely, Colpidium
used the refuge very little and were found in higher densities in the water column relative to the
refuge (refuge use = -0.36 ± 0.12).
There was no effect of selection environment (Fig. 3.1E, F1,40 = 0.093, P = 0.76),
measurement environment (Fig. 3.1E, F1,2 = 0.52, P = 0.39), or time block (F2,2 = 1.91, P = 0.34)
on the peak population abundance. The best-fit model included two interactions
(measurement*selection environment and block*measurement environment), but neither were
significant (P > 0.55). Selection environment had no effect on the speed of Colpoda (Fig. 3.1F,
F1,2 = 0.23, P = 0.68), but there was a marginally significant effect of measurement environment
(Fig. 3.1F, F1,41 = 3.60, P = 0.064) and time block (F2,2 = 11.5, P = 0.08) on cell speed. The
best-fit model included a block*measurement environment interaction that was non-significant
(F2,41 = 0.65, P = 0.53). The speed of Colpidium (0.20 ± 0.007 mm/s) was much slower than the
average speed of Colpoda (Fig. 3.1F).
Most traits were significantly correlated with other traits, although the nature and
strength of the correlation was dependent on the specific traits under consideration (Table 3.1).
When data were subdivided into different measurement and selection environments,
correlations were qualitatively similar to one another, although not significant in some cases,
likely due to lower power with fewer data points. As such, the values in Table 3.1 reflect pooled
data with increased power. The one exception to this case was the correlation between cell
area and refuge use. The overall correlation with pooled data was significantly negative, but
27
this is largely due to a strong negative correlation of individuals from the monoculture selection
environment grown in the monoculture measurement environment (r = -0.45, P < 0.05). In all
other selection and measurement environments, the correlation between cell size and refuge
use was non-significant (range of r: -0.04 to 0.16, P > 0.05).
Table 3.1. Average correlation coefficients after selection among six traits after bootstrapping
data pairs. Asteriks indicate values for which the 95% confidence interval does not include zero.
Pop.
Peak
Cell
Swimming
Cyst
Refuge
Growth
Abundance
Area
Speed
Production
Use
0.36*
-0.56*
-0.32*
0.25*
0.29*
1.0
-0.28*
-0.37*
0.29*
0.15
1.0
0.36*
-0.17
-0.21*
1.0
-0.34*
-0.22*
1.0
0.18
Rate
Pop. Growth
Rate
Peak
Abundance
Cell
Area
1.0
Swimming
Speed
Cyst
Production
Refuge
1.0
Use
The phylogenetic tree generated (Fig. 3.2) indicates that most of the samples used in the
selection experiments were from a monophyletic group and likely belonged to the same species.
None of the reference sequences fall within the same group however, suggesting that the
species used has not previously been sequenced. The exceptions to this are two lines used in
the October 2008 experiments that fall outside of this monophyletic group. One of these lines
(Oct1) is likely Colpoda steinii, while the other (Oct2) falls into its own clade (Fig. 3.2). By the
April experiments of the following year, I recognized two distinct morphotypes among samples.
All species in the monophyletic group have similar morphologies and are conspicuously larger
than individuals in the Oct1 sample, with more obvious coloration and frequently visible
28
63
53
Bresslaua vorax
Colpoda henneguyi
Bresslauides discoideus
Pseudoplatyophyra nana
94
Mykophagophrys terricola
Colpoda inflata
Colpoda magna
99
Colpoda minima
Colpoda cucullus
Colpoda aspera
Ottowphyra dragescoi
Colpoda steinii
100
Oct1
82
Hausmanniella
Hausmannielladiscoidea
discoidea
Colpoda lucida
57
Nototoxoma parabryophryides
Bardeliella pulchra
Jun1
Jun3
92
Aug1
76
Jun2
Aug2
Apr1
Oct3
Oct2
Ilsiella palustris
Oxytricha nova
55
Bryometopus pseudochilodon
Cyrtolophosis mucicola
Figure 3.2. Bootstrap consensus tree using parsimony generated using nucleotide sequences from a region
of 18S rRNA. Samples in bold were sequenced in this study. Oct, Apr, and Jun samples represent a
portion of the lines from different time blocks; lines from Aug were collected in similar fashion, but not used
in any experiment. Reference sequences were obtained from Genbank (see Appendix B for accession
numbers).
29
vacuoles within the cell. Further, the gliding movement of the former morphotype is smoother
than the more jerky movement of the latter. The Oct2 sample is composed of both
morphotypes, which may explain why it does not group distinctly with either group. Although
several stock lines were unable to be recovered for sequencing, the unrecovered lines were
morphologically similar to that of the large monophyletic group.
3.3 Discussion
Protozoa have been used successfully as a model system of resource competition for
decades (e.g. Gause 1934, Vandermeer 1969, Cadotte et al. 2006), but this is the first study to
measure the evolution of protozoan traits in response to competition. Populations grown in
interspecific competition evolved faster population growth rates and smaller cell sizes and
produced fewer cysts relative to populations that evolved with only intraspecific competition (Fig.
3.1). Population growth rates and cell sizes measured before the start of the experiment,
suggest that these traits evolved in both selection treatments, but that populations in
interspecific competition evolved to a greater extent (unpublished data). However, cyst
production rates prior to selection are unknown. The difference between treatments could be
due to fundamental differences in the nature of selection in response to intra- and interspecific
competition or to differences in the strength of selection between these environments. Initial
densities of interspecific competitors were twice as high as intraspecific competitors, resulting in
triple the total density of organisms, which likely resulted in increased competition. Interspecific
competition could have resulted in a difference in both the type (different species) and strength
(different density) of selection. Regardless, competition has a significant effect on the evolution
of these traits.
At least two species were likely present in the October 2008 experiments (Fig. 3.2) and
selection at either the species or the population level may have produced the observed result.
All samples used in the April and June experiments though had similar morphotypes. The
phylogeny suggests that these genetic lines belong to a monophyletic group (Fig. 3.2). In these
latter experiments, evolution may have occurred via lineage sorting, in which particular
genotypes were more successful than others, or by a process of mutation and selection within
genotypes. These same processes may have occurred in the October experiments, but cannot
be distinguished from species-level selection. Development of microsatellite markers for these
species for use in future experiments will allow for distinguishing between different mechanisms
of evolution and whether different traits evolve via alternate mechanisms. Species-level
selection should have resulted in significant time block effects due to differences in trait
30
evolution in the October experiments (i.e. October different from April or June), but Tukey posthoc tests reveal that this is not the case. There were also no selection environment*time block
effects, suggesting that selection was similar in all time blocks. This suggests that if specieslevel selection did occur, it was unimportant relative to the overall pattern of selection.
Population growth rate is a function of both birth and death rates. Although nothing is
known about the death rates in these experiments, an evolved increase in cell division rate is
likely to have led to an increase in population growth rate. Although traits that reduce the death
rate in competitive environments could have led to this same increase, population growth rates
were also higher in the monoculture measurement environment, where death rates were
presumed to be lower. This suggests that birth rates of the populations increased. Both an
increase in cell division rate and cyst production may have allowed these populations to tolerate
competition from other species. Cell division rate has been suggested to increase tolerance of
predation in these communities (terHorst et al. 2010). Selection by competitors for rapidlydividing genotypes may have resulted in the evolution of increased population growth rate.
Likewise, the production of resting cysts may have allowed for tolerance of competition
by Colpoda populations. The production of cysts may be density-dependent and related to the
availability of bacteria in the environment (Yamaoka et al. 2004, Foissner et al. 2009), which
may be a reliable cue for the imminent demise of transient aquatic habitats (Ekelund et al.
2002). Populations with increased encystment rates may be more capable of rapid population
growth when transferred to a new environment. Tubes were reinitiated every three days with
equal densities of adults, but cysts that were transferred along with the adults could have
contributed to greater population growth rates. Such dynamics occur in several invertebrate
taxa, where resting cysts allow for tolerance of poor environmental conditions (e.g. Fenchel et
al. 1997, Caceres 1998, Dumont et al. 2002), followed by rapid population growth in favorable
environments.
Other traits may have contributed to reduced death rates in environments with
competitors. Although not significant, Colpoda populations that evolved with competitors tended
to use the refuge more heavily, possibly to avoid competition with Colpidium in the water column
(Fig. 3.1D). Use of the refuge by Colpoda was size-dependent, with smaller cells using the
refuge more heavily than larger cells that tend to remain in the water column (Table 3.1,
unpublished data). Since Colpidium used the water column the most, the larger Colpoda cells
most likely experienced more competition, resulting in selection for decreased cell size (Fig.
3.1B) as a function of refuge use. Cell size is positively correlated with competitive effect
among species of bacteriovores in the pitcher plant inquiline community (Kneitel 2002, Leibold
31
and Miller 2004), so the decrease in cell size presumably resulted in a decrease in competitive
effect. However, a niche shift in refuge use by Colpoda would constitute an increased
competitive response that may have reduced competition between Colpoda and Colpidium. It is
impossible to determine whether cell size evolved as a result of direct selection on cell size or
as a consequence of direct selection on refuge use, coupled with the correlation between refuge
use and cell size.
6
ln (cysts +1)
5
4
3
Mono
Selection Environment
Comp
Monoculture
Competition
2
1
0
-3
-2.5
-2
-1.5
-1
-0.5
0
ln population growth rate
Figure 3.3. Correlations between population growth rate and cyst production
from two selection environments. The two traits are positively correlated, but
selection is not constrained by the correlation. Evolution in competition
results in decreased cyst production and increased growth rate, but the
correlation is maintained in both treatments.
The possibility of indirect selection on cell size occurs when traits are correlated with one
another and selection on one trait results in the evolution of correlated traits. However,
correlations between traits do not necessarily constrain the response to selection, but rather
must be considered relative to the strength of selection on each trait. For instance, population
growth rate and cyst production are positively correlated with another (Table 3.1). Yet, evolution
in competition resulted in increased growth rates and decreased cyst production (Fig. 3.1). Had
32
selection on cyst production been weak, selection for increased growth rates probably would
have resulted in increased cyst production, owing to the positive correlation between these traits
(Fig. 3.3). The reduction in cyst production suggests that selection on this trait was stronger
than that on population growth rate and able to overcome the correlation between traits. The
effect size of selection environment on these traits suggests that indeed selection for cysts may
have been stronger than that on growth rate (Fig. 3.1).
In addition to evolved differences in three traits (population growth rate, cell size, cyst
production), there was also significant phenotypic plasticity between monoculture and
competition measurement environments in cell size and swimming speed. Cells tended to
move slightly faster in the competition measurement environment (Fig. 3.1F), but it is unclear
whether this 4% increase in speed is ecologically relevant. In the competition measurement
environment, Colpoda tended to be larger than in the monoculture environment, despite the fact
that the competition selection environment caused a decrease in the size of Colpoda. The
evolution of smaller cell size in response to competition may have been the result of indirect
selection on cell size (see above) and size itself may not be adaptive in environments with
competitors. Larger Colpoda cells may be better able to compete with smaller Colpidium cells
and exhibit some degree of plasticity in their size and competitive ability. However, the
evolution of this trait may be constrained by stronger selection on correlated traits
Traditional niche-based theory dictates that competitors’ traits should diverge through
evolutionary time (MacArthur and Levins 1967, May and MacArthur 1972, Chase and Leibold
2003) to minimize the effects of competition. Other theory suggests that either convergence or
divergence could occur among competitors (Aarssen 1983, Fox and Vasseur 2008), particularly
if rates of evolution are faster than the rate of competitive exclusion (C.P. terHorst and T.E.
Miller, unpublished simulations). While there are many documented cases of character
divergence (e.g. Brown and Wilson 1956, Simberloff and Dayan 1991, Schluter 1994), there are
considerably fewer, if any, cases of convergence in the literature. In this experiment, the
convergence or divergence of traits was trait-dependent. Population growth rates of Colpoda
diverged from Colpidium traits (Figs. 1, 4). However, cell size of Colpoda converged on that of
Colpidium. Note though that this is not the result of coevolution between these species, as
Colpidium was continually restocked from a stock culture and unable to evolve in response to
Colpoda. Most theory on character convergence and divergence examines a single trait (e.g.
resource use or competitive ability), but my results suggest that independently evolving traits
that each contribute to competitive ability may produce different outcomes. Convergence and
divergence in different traits may occur simultaneously. Further, evolution in this system
33
occurred rapidly relative to the lifespan of other species in this community (e.g. predatory
mosquito larvae) and the lifespan of the leaf itself (~40 weeks). Such rapid evolution may affect
the ecological outcome of species interactions (Carroll et al. 2007, Strauss et al. 2008, terHorst
et al. 2010). Ongoing work in this community is attempting to determine whether the
convergence or divergence of traits amongst a suite of competitors ultimately affects species
coexistence.
34
CHAPTER 4
EVOLUTION IN RESPONSE TO INDIRECT ECOLOGICAL EFFECTS
Although most ecological and evolutionary experiments examine the direct effects of
species on one another, species in natural biological systems exist within a network of species
interactions. This inevitably leads to both direct and indirect effects among many species
simultaneously. Indirect, or higher-order, ecological interactions occur when the direct effect of
one species on another is affected by the presence of a third species (Vandermeer 1969, Holt
1977, Strauss 1991, Wootton 1994). In more diverse and complex communities, indirect effects
are likely to increase in importance—as the number of species in a community increases, the
number of possible indirect pathways between species increases exponentially relative to the
number of direct effects (Abrams 1992, Menge 1995). Indeed, in some cases, multiple indirect
effects can swamp the influence of direct ecological effects (e.g. Stone and Roberts 1991,
Abrams 1992, Miller 1994). The effect of indirect interactions on ecological patterns has been
well-studied and can result in changes in species abundances, coexistence, and community
diversity (e.g. Vandermeer 1969, Strauss 1991, Lawler 1993, Miller 1994, Wootton 1994, Menge
1995).
Given their ecological importance, it seems likely that indirect effects would also be
critical to understanding how species evolve in a community context. Most experiments in
evolutionary biology remove species from a community context and measure selection or
evolution in response to abiotic factors or, at most, one other species. For instance,
experiments documenting the direct effect of predators on prey evolution are abundant.
Selection by predators or herbivores can result in the evolution of various traits, including
coloration (Alatalo and Mappes 1996), size (Nunezfarfan and Dirzo 1994, terHorst et al. 2010),
chemical defenses (Mauricio and Rausher 1997, Shonle and Bergelson 2000), and structural
defenses (Reimchen and Nosil 2002, Seeley 1986). The evolutionary consequences of
competitive interactions have been considered as well. Competition experiments have
demonstrated selection for traits such as growth rate (Schluter 1994, Santos et al. 1997,
35
McGoey and Stinchcombe 2009), body size (Santos et al. 1997, McGoey and Stinchcombe
2009, terHorst, in review), foraging behavior (Sokolowski et al. 1997), or floral traits (Smith and
Rausher 2008). The consequences of direct positive species interactions, such as facilitation
and mutualism, on trait evolution have been studied as well (Herre et al. 1999, Fenster et al.
2004). These evolutionary studies of direct species interactions have produced a century’s
worth of knowledge about how species evolve in controlled experiments with very simple
systems. However, they do not incorporate complex interactions, such as indirect effects, which
limits their applicability to evolutionary dynamics in nature.
Understanding evolution in a community context requires quantification of multiple direct
species interactions, as well as indirect interactions. The evolutionary consequences of indirect
ecological effects were considered by Miller and Travis (1996), whose heuristic model predicted
changes in species effect on, or response to, associate species, based on the type of direct and
indirect effects. This theoretical work was followed by a modest amount of empirical work
focused on selection in response to indirect ecological effects. Astles et al. (2005)
demonstrated significant genetic variation in a species response to indirect effects, indicating
that the raw material required for natural selection was present. Two other empirical studies
have examined indirect interactions among species of plants and insects and measured
significant selection that occurred in response to indirect ecological effects (Irwin 2006, Lau
2008). However, given variability in selection pressures across time and space, it is unclear
whether selection in response to indirect effects will ultimately result in the evolution of traits.
Selection experiments in controlled environments are required to quantify the effect of both
direct and indirect species interactions on the evolution of traits measured in common garden
environments to distinguish evolved genetic differences from effects of phenotypic plasticity.
Here I use the natural microcosm community that exists within the leaves of carnivorous pitcher
plants to measure the evolution of a suite of traits in a ciliated protozoan in response to both
direct and indirect ecological effects.
Ecological interactions in the inquiline community found in the water-filled leaves of
Sarracenia purpurea (purple pitcher plant) have been well-described (reviewed in Miller &
Kneitel 2005), but more recently this community has served as a model system to examine
evolution in response to direct species interactions (terHorst et al. 2010, terHorst, in review).
The host plant is widely-distributed, extending from northern Florida in the US to the Northwest
Territories in Canada, and the community is representative of other types of natural microcosm
communities (Srivastava et al. 2004). Purple pitcher plants produce cup-shaped leaves that fill
with rainwater and attract insect prey—mostly ants, and primarily to young leaves. The energy
36
in the prey captured by the leaves serves as a basis for a community of bacterial decomposers,
bacterivorous protozoa and rotifers, and a specialist mosquito (Wyeomyia smithii) that
consumes bacterivores. Species abundances change over weeks during succession in a
pitcher plant leaf. Predation by mosquito larvae on protozoan bacterivores is highest in young
leaves, but competition among these protozoans is strongest in older leaves where resources
become limiting (T.E. Miller and C.P. terHorst, unpublished data). Evolution of species traits
can occur in response to both predation and competition separately (terHorst et al. 2010,
terHorst, in review). Species in intermediately-aged leaves, however, experience an extended
period of time exposed to both predation and competition. The evolutionary consequences of
these multiple direct effects and the inherent indirect effects are the subject of this study.
4.1 Methods
The evolutionary response of traits to predation, competition and indirect effects were
measured in replicate populations of Colpoda sp., a unicellular ciliated protozoan that is
intermediate in size, competitive ability, and predator tolerance relative to other bacterivores in
the community (Kneitel 2002, Leibold & Miller 2004). Because of practical limitations on the
number of replicates, the experiment was conducted in three time blocks (October 2008, April
2009, and June 2009), using unique lines of Colpoda from the field in each time block. Samples
of water within pitcher plant leaves were collected from haphazard leaves at each of two sites in
the Apalachicola National Forest in northern Florida: Crystal Bog and Naczi Borrow Pit. In each
time block, Colpoda were collected from 4 leaves to create 4 independent stocks, that were later
mixed to create a pool of genotypes in each replicate. Stocks from independent leaves were
later genotyped to determine species and clonal identity. Sequences from 18s rRNA suggest
that individuals from the October 2008 time block belonged to at least two closely related
species (Colpoda steinii and a previously unsequenced species), but that all lines in the April
and June 2009 time blocks belonged to the same previously unsequenced species (terHorst, in
review). Despite this, time blocks effects were not significantly associated with differences in
species composition and all time blocks responded similarly to selection treatments (terHorst, in
review).
To initiate selection experiments, approximately 300 individuals from the mixedgenotype stock culture were used to inoculate replicate laboratory microcosms (along with the
associated bacteria) that mimic pitcher plant leaves (50 mL plastic macrocentrifuge tubes).
Protozoa in these laboratory microcosms have similar population dynamics to those in pitcher
37
plant leaves in the field (Miller et al. 1994). Each microcosm contained 20 mL sterile well water,
plastic beads (2-5 mm diameter), and 6 mg (± 0.15 mg) of Tetramin fish food (Tetra Holding,
Inc., Blacksburg, VA, USA) as a food source for the bacteria. The plastic beads served as a
potential refuge from predation and competition for Colpoda and mimicked the detritus and frass
found at the bottom of natural pitchers (terHorst et al. 2010).
Most reproduction in Colpoda is asexual and occurs every 4-8 hours (terHorst et al.
2010), although infrequent sexual reproduction may occur (Dunthorn et al. 2008). I used
selection experiments to examine the evolution of a suite of Colpoda traits after 20 days (60-120
generations). Replicate populations of Colpoda (n=4 per treatment per block, each containing a
mix of genotypes) were maintained in four selection regimes with different species interactions.
(1) In monocultures, individuals experienced only intraspecific competition. (2) The predation
treatment regimes used two predatory third-instar mosquito larvae (W. smithii) that were first
sterilized in H2O2 for 30 seconds (terHorst et al. 2010). (3) In the competition treatment, another
species of ciliated protozoan (Colpidium sp.) was added at twice the initial density of Colpoda
(terHorst, in review). Colpidium is known to be competitively dominant over Colpoda (Kneitel
2002). (4) The multispecies treatment contained the same density of Colpidium, as well as two
W. smithii larvae. This final treatment allowed for both direct and indirect species interactions to
occur. Hereafter, I refer to these four treatments as Monoculture, Predation, Competition, and
Multispecies, respectively. The abundance of both mosquito larvae and Colpidium were
monitored daily and all experimental microcosms were reestablished every three days to avoid
the exclusion of Colpoda by competiton or predation, coevolution between competitors, or
satiation of predators (terHorst et al. 2010, terHorst, in review). During each reestablishment,
300 Colpoda were transferred from the appropriate replicate and new mosquito larvae and
Colpidium were added from laboratory stock cultures.
All cultures were maintained in a growth chamber with a diurnal cycle (12/12) of light and
temperature (day=30°C, night=20°C). At the end of the selection experiment, microcosms were
well-mixed and ~300 Colpoda were isolated and removed from replicates in all treatments and
placed into identical environments containing only 20 mL of sterile well water and 6mg of
Tetramin. Colpoda were grown under these common conditions for 24 hours (3-4 generations)
to minimize the role of environmental plasticity and maternal effects before measuring traits.
Measurement of Traits
After growth in a common environment for 3-4 generations, two random subsamples of
~300 individuals from each replicate were used to measure traits in a monoculture ―common
38
garden‖ environment to discount effects of phenotypic plasticity. The generations in the
common garden insure that any differences in trait values between selection treatments should
be due to evolved genetic differences. In previous work, traits were measured in multiple
―common garden‖ environments (i.e. monoculture environments, environments with predators,
environments with competition), but analyses revealed no interactions in trait values between
measurement and selection environments (terHorst et al. 2010, terHorst, in review). For
simplicity, only traits measured in common monoculture environments are presented here. In
each replicate from each selection treatment, six traits of Colpoda were measured: cell size,
cell swimming speed, cyst production, peak population abundance, population growth rate, and
refuge use.
Digital videos and image analysis (NIH Image J, http://rsb.info.nih.gov/ij/) were used to
estimate cell area (mm2) and swimming speed (mm/s) of the first 30 individuals encountered in
the video. Refuge use was determined by removing two 0.1 mL samples from each microcosm,
one in the water column and one in the beads at the bottom of the microcosm, and counting the
number of individuals on a Palmer counting cell. Refuge use was determined as: log(# in beads
/ # in water). This measure of refuge use provides an unbounded index, with a value of 0
indicating equal densities in the beads and water.
Population growth rates were measured by growing each subsample of 300 individuals
in microcosms with 20 mL sterile well water and 6 mg Tetramin. Population size was estimated
by counting the number of individuals in a 0.1 mL random subsample of each culture on a
Palmer counting cell (Wildlife Supply Company, Buffalo, NY, USA) every six hours until
populations peaked in abundance. Populations displayed exponential growth up to a peak at
approximately 48 hours and then abundance dropped off rapidly. Growth rates were
determined by estimating r from the best-fit exponential growth model (N = N0ert), using
abundance data up to the peak. Protozoa produce cysts that allow populations to survive
desiccation (Ekelund et al. 2002). Cyst production was estimated by counting the number of
cysts in a 0.1 mL subsample from each microcosm after six days.
Statistical analyses
All data met the assumptions of parametric tests or were natural log-transformed to meet
these assumptions. Two-way mixed model ANOVAs (performed using restricted maximum
likelihood estimates in JMP version 4.0.4, SAS Institute Inc, USA, 2001), using selection
environment as a fixed factor and time block as a random factor, were used to separately test
for differences in each of the six traits. The effect of selection environment tested for evolved
39
differences in traits among treatments. Tukey post-hoc tests were used to test for differences
among selection treatment means.
The sizes of the direct effects of predation and competition on the evolution of each trait
were estimated as a percent change by subtracting the mean trait value in the Predation or
Competition selection treatment from the mean trait value in the Monoculture selection
treatment and then standardizing by the mean Monoculture trait value. The predicted additive
effects of predation and competition were calculated for each trait by adding the effect size of
predation and competition to the mean trait value in monoculture. In an example with a trait
value of 10 in monoculture, 15 in predation, and 17 in competition, the predicted additive effect
with both predation and competition present would be a trait value of 22.
Indirect effects were estimated by subtracting the mean trait value in the Multispecies
selection treatment from the predicted additive effect, such that the effect size on a given trait
was:
Effect sizeIndirect = Multispecies mean – (Monoculture mean + Pred. effect + Comp. effect)
In the above example, a trait value of 12 in the multispecies treatment, would yield an indirect
effect of -10. Direct and indirect effect sizes were bootstrapped 10,000 times using the R
statistical environment (R Core Development Team) and 95% confidence limits were estimated
to determine whether bootstrapped means significantly differed from zero.
4.2 Results
Population growth rates of Colpoda populations were significantly affected by the
selection treatments (F3,6 = 6.2, P = 0.029), as populations that evolved with predation and
competition grew faster than populations that evolved in the monoculture and multispecies
treatments (Fig. 4.1A). There was a significant effect of time block (F2,6 = 12.6, P = 0.007), but
there were no block*treatment interactions (F6,36 = 0.73, P = 0.63). The effect size of predation,
competition, and indirect interactions on population growth rate were significantly different from
zero (Fig. 2A). Indirect interactions were nearly as strong as the sum of direct interactions, but
had an effect of decreasing, rather than increasing, population growth rate (Fig. 4.2A).
Selection treatment also had a significant effect on Colpoda cell size (F3,6 = 10.8, P =
0.008). Cell size decreased in predation and competition treatments relative to cell sizes in the
monoculture and multispecies environments (Fig. 4.1B). There was an effect of time block (F2,6
= 8.9, P = 0.016) and the effect of treatment was dependent on time block (F6,35 = 13.6, P <
0.001). However, the interaction was one of magnitude and not direction—the rank-order of
selection treatments was the same in every time block. The effect size of predation,
40
competition, and indirect interactions were significantly different from zero (Fig. 4.2B). Although
indirect and direct effects were nearly equal in magnitude, they again had opposite effects on
0.22
b
0.2
b
0.18
0.16
a
a
0.14
0.12
0.5
0.4
0.3
D
120
a
b
1.6
c
1.2
0.8
0.4
B
0
1.2
0.8
0.4
0
-0.4
-0.8
-1.2
C
c
# Cysts per 0.1 mL
2
Cell Area (mm2 x 10-4)
0.6
0.1
A
Refuge Use
Cell Speed (mm/s)
0.7
100
80
60
40
20
0
E
Peak Density (# per 0.1mL)
Population Growth Rate (r)
cell size (Fig. 4.2B).
350
300
250
200
150
100
50
0
F
Figure 4.1. Mean values (+/- s.e.) of six traits from four different selection treatments, measured in
a monoculture common garden. The dashed bar represents the predicted trait value based on the
additive effects of predation and competition. Letters above bars represent significant differences
among means based on Tukey HSD in cases where the overall treatment effect was significant.
41
There was no overall effect of selection treatment on refuge use by Colpoda populations (F3,6 =
1.24, P = 0.38, Fig 4.1C). The effect size of predation was positive, but not significantly different
from zero (Fig. 4.2C). However, competitors had significantly positive
A
0.5
0.3
0.1
0.5
0.3
0.1
-0.1
-0.1
-0.3
-0.3
-0.5
-0.5
-0.7
D -0.7
Predator
Competitor
Speed
0.7
Population Growth Rate
Effect Size
Effect Size
0.7
Indirect
Predator
0.7
0.3
0.1
-0.1
-0.5
E
0.8
-0.4
-0.8
Predator
Competitor
Indirect
Predator
Competitor
Indirect
0.7
0.7
Peak
Refuge Use
Effect Size
Effect Size
1.2
0
-0.7
0.5
0.3
0.1
0.5
0.3
0.1
-0.1
-0.1
-0.3
-0.3
-0.5
-0.5
C
1.6
0.4
-0.3
B
Indirect
Cysts
2
Effect Size
Effect Size
Cell Area
0.5
Competitor
F
-0.7
Predator
Competitor
-0.7
Predator
Indirect
Competitor
Indirect
Figure 4.2. Effect sizes of three selection regimes on the values of six traits. Effects sizes were
calculated based on bootstrapped means. Error bars reflect bootstrapped 95% confidence
intervals. Intervals that do not include zero are significantly different from zero. In most cases,
indirect effects are stronger and opposed to the direct effects of predation and competition. Note
that the y-axis for E is on a larger scale.
42
effects on refuge use, while indirect effect sizes were significantly negative (Fig. 4.2C). There
was also a significant effect of time block (F2,6 = 6.2, P = 0.03), but the block*treatment
interaction was not significant (F6,36 = 1.6, P = 0.17).
There was a marginally significant effect of selection treatment on cell speed (F3,6 = 4.3,
P = 0.061, Fig. 4.1D), but the effect sizes of all treatments were not significantly different from
zero (Fig. 4.2D). There was a significant effect of time block (F2,6 = 19.5, P = 0.002), but the
block*treatment interaction was not significant (F6,35 = 0.64, P = 0.70).
Finally, there was no effect of selection environment on either cyst production (F3,6 =
1.06, P = 0.43, Fig. 4.2E) or population peak (F3,6 = 0.36, P = 0.79, Fig. 4.2F), nor were any
effect sizes significantly different from zero with respect to either trait (Fig. 4.2E,F). Time block
had an effect on both traits (Cysts: F2,6 = 41.5, P < 0.001; Peak: F2,6 = 5.1, P = 0.052). There
were no significant block*treatment interactions with respect to either trait (Cysts: F6,36 = 0.48, P
= 0.82; Peak: F6,36 = 0.56, P = 0.76).
Selection treatment had a significant effect on the density of competitors (F1,2 = 43.1, P =
0.02). In the Multispecies treatment, which contained both Colpidium and mosquito larvae,
Colpidium densities were reduced approximately five-fold (Fig. 4.3). There was no effect of time
block on Colpidium density (F2,2 = 0.95, P = 0.51), nor did treatments differ among time blocks
Average Colpidium density (per 0.1 mL)
(F2,18 = 2.35, P = 0.12).
350
300
250
200
150
100
50
0
Competition
Multispecies
Figure 4.3. Mean (+/- s.e.) of the average Colpidium density in
two selection treatments.
43
4.3 Discussion
The importance of indirect effects has long been appreciated in the ecological literature.
The results here demonstrate that indirect effects also play a large role in how species evolve in
a community context. These experiments confirm that evolution can occur in response to the
direct effects of predators and competitors on Colpoda populations (Fig. 4.1, terHorst et al.
2010, terHorst, in review). However, the novel result is that indirect effects also had significant
effects on the evolution of traits in Colpoda (Fig. 4.1, 4.2). In five of the six traits measured,
indirect effects had the opposite effect on trait values than the direct effect (Fig. 4.2). Further,
the strengths of indirect effects were often greater in magnitude than direct effects (Fig. 4.2) and
resulted in little net evolution of some traits in populations that experienced both direct and
indirect effects (Fig. 4.1).
Both population growth rate and cell size evolved in Colpoda populations in response to
the direct effects of predation and competition (Fig. 4.1A, B). Additionally, Colpoda populations
tended to use a refuge more heavily as a result of having evolved with either predators or
competitors (Fig. 4.1C, 4.2C). This confirms the results of previous work examining the
evolution of these traits in this species (terHorst et al. 2010, terHorst, in review). Although many
traits play a role in determining population growth rate, an increase in Colpoda growth rate is
likely the result of an increase in the rate of production of daughter cells (terHorst et al. 2010).
This trait serves as a tolerance strategy in the presence of predation, but may increase
competitive ability in the absence of predation. A genetic line may increase its competitive
ability by cell proliferation that leads to increased resource consumption, leading to an increase
in competitive ability at high resource levels. Among bacterivorous species in pitcher plant
communities, predator tolerance is negatively correlated with cell size, while competitive ability
is correlated with increasing cell size (Kneitel 2002, Leibold and Miller 2004), so one might have
expected divergent cell sizes in predation and competition treatments. However, cell size may
have evolved as an avoidance strategy in response to predation and competition. Smaller
individuals may be harder to capture by the filter-feeding mosquitoes and they tend to use the
refuge more frequently than larger individuals (terHorst, in review). By doing so, smaller
Colpoda cells experience less predation from mosquito larvae foraging in the water column
(terHorst et al. 2010) and less competition from Colpidium, which remain active largely in the
water column (terHorst, in review). Large cells that remain in the water column likely experience
the highest levels of both competition and predation resulting in relatively small survivors that
use the refuge more heavily.
44
Indirect effects can occur through density-mediated interactions, in which the abundance
of one species is reduced by the presence of another, or through trait-mediated effects, in which
traits of one species are altered in the presence of another (Miller and Kerfoot 1987, Trussell et
al. 2002, Pressier et al. 2005). In this experiment, both density- and trait-mediated indirect
effects are likely to have occurred. The indirect effects measured are a sum of both types of
indirect effect, but the independent effect of each is difficult to quantify. Trait-mediated effects
occurred via a habitat shift by mosquito larvae in the presence of Colpidium (Fig. 4.4).
Colpidium populations remained primarily in the water column and were infrequently found in
the refuge at the bottom of the microcosm (terHorst, in review). When Colpidium was present,
filter-feeding mosquito larvae were more likely to forage in the water column (terHorst,
unpublished data). In the absence of Colpidium, mosquito larvae were forced to forage for
Colpoda in the refuge (terHorst, unpublished data), where their feeding efficiency was likely
reduced due to limitations on their mobility within the refuge. Density-mediated indirect
interactions occurred when the pressure of competition on Colpoda populations was alleviated
by a reduction in Colpidium densities in the presence of mosquito larvae (Fig. 4.3). As a result
of these indirect effects, Colpoda experienced much less selection pressure from either
predators or competitors when both species were present (Fig. 4.4). Although the combination
of indirect effects on some traits is much higher than individual direct effects (Fig. 4.2), the sums
of indirect and direct effects on most traits are similar.
Miller and Travis (1996) made predictions about how focal species traits should evolve in
response to the direct and indirect effects of a focal species. In cases where direct effects are
negative and indirect effects are positive, as is the case with respect to both predator and
competitor effects on Colpoda, a decreased evolutionary response is predicted (Miller and
Travis 1996). A change in response is defined as a trait change that has little effect on the
associate species. A change in response is predicted because it reduces the direct negative
interaction with the associate species, but retains the positive indirect interaction. With respect
to predators as an associate species, the evolution of population growth rate, cell size, and
refuge use in Colpoda populations are consistent with this prediction. A change in these traits
likely has little effect on the population dynamics of mosquitoes in the multispecies treatment,
and thus maintains the positive indirect effects of mosquitoes (i.e. eating competitors) while
reducing interactions between Colpoda and mosquitoes. With respect to competitors, a match
to the prediction is less clear. The evolution of cell size and refuge use may reduce interactions
between Colpoda and Colpidium, but the evolution of population growth rate may increase
competitive interactions between these species.
45
Monoculture
Predation
Competition
Multispecies
Colpoda
Mosquito
Colpidium
Figure 4.4. The hypothesized direct and indirect effects of mosquito larval predators and Colpidium competitors on Colpoda populations
in experimental microcosms. The beads at the bottom of each microcosm serve as a refuge and are similar to refuges found in natural
pitchers. The size of the symbol reflects the abundance of each species and the location reflects use of different parts of a microcosm. In
the multispecies treatment, mosquito foraging behavior is altered because of the presence of Colpidium. This change in foraging
behavior reduces the abundance of Colpidium, resulting in both behavior- and density-mediated indirect effects on Colpoda populations.
46
Time block had a significant effect on the value of every trait, which may be a result of
slightly different laboratory conditions or culture techniques in each time block. Alternatively, the
time-block effect may be a result of properties of populations used in different time blocks. The
genetic variance of populations within and among leaves is unknown. One leaf may contain up
to hundreds of genotypes or, at the other extreme, one genotype may be pervasive in all leaves.
The four random populations chosen for each time block may simply have had different trait
values. Further, the genetic variance in the different four-population mixtures may have affected
the evolvability of populations and resulted in faster rates of evolution in some time blocks.
Finally, differences between time blocks may be due to selection at the species level in the
October 2008 experiment, but not in the other time blocks. Genetic analyses show that the
October experiments were conducted with two different, but closely related species from the
Colpodidae (Fig. 3.2). However, species-level selection is unlikely to have affected the general
conclusions of this study. If all differences in traits among time blocks were due to differences in
species-level selection, then one would expect consistent differences between October and
April/June time blocks. Post-hoc analyses reveal this to be the case for cyst production, where
significantly fewer cysts were produced in October, but not for the remaining five traits.
Considering evolution in a community context provides a more realistic depiction of how
species evolve in natural systems. An indirect ecological effect is an emergent property that
can only occur at the community level. Consideration of the importance of indirect effects
bridges the gap between evolutionary experiments with one or two species and evolutionary
dynamics in natural communities. Although complexity in the community used in this study was
greater than that of most evolutionary studies, this system was still very simple compared to
complex webs observed in natural communities. Certainly, more diverse communities, where
the number of indirect effects dwarfs the number of direct effects, may provide even more
complex dynamics Incorporating additional direct and indirect effects in future experiments will
better evaluate the evolutionary forces experienced by species in natural biological
communities.
47
CHAPTER 5
CONCLUSION
The integration of ideas from the fields of evolutionary biology and community ecology is
crucial for understanding both the evolution of species in a natural context and understanding
patterns of abundance and distribution in communities. Experiments in community ecology
have traditionally ignored the effects that recent trait evolution might have on species
abundance patterns. However, the strong selection imposed by most experiments can
frequently result in rapid evolution of at least some species (Strauss et al. 2008). The results of
this dissertation demonstrate that the evolution of traits can affect ecological interactions
between predator and prey. Previously, many would refer to such evolution over a 12 day time
span as ―rapid evolution‖ (e.g. Thompson 1998, Hairston et al. 2005), but this term is misleading
as it usually refers to an anthropogenic interpretation of what is rapid. What is rapid to one
species is not necessarily rapid to another. The evolution of traits in a population of elephants
over a few generations would be rapid relative to the ecological dynamics of elephant
populations, but incredibly long relative to the ecological dynamics of bacterial populations. In
this case, prey populations evolved over approximately 40 generations, but this evolution
occurred in less than one predator generation and within the timeframe of succession within
pitcher plant leaves.
This effect of evolution on ecological interactions need not be limited to microcosm
communities, like the one used here, but rather may apply more broadly to communities
composed of species with different generation times. Large differences in generation times are
common in many co-occuring species. For instance, annual plants have short generation times
relative to mammalian herbivores, but relatively long generation times when compared to
herbivorous arthropods. Discrepancies in generation times increase by orders of magnitudes
when one considers the population dynamics of megafauna and flora relative to pathogenic
bacteria, rhyzobia, mycorrhizal fungi, or algal endosymbionts. Such discrepancies in generation
times may lead to differences in rates of evolution and preclude coevolution that might
48
otherwise maintain ecological interactions (e.g. van Valen 1973). In order to understand how
species coexist within a community, it is imperative to understand how the traits of those
species are changing relative to one another over time.
Considering evolution in a community context is crucial to understanding how species
evolve in natural communities, where they can interact with many species simultaneously. Our
current understanding of evolution is largely based on two-species interactions. If the
evolutionary effects of multiple species are additive, then it is easy to predict evolution in a
community context based on the direct evolutionary effects determined from these relatively
simple experiments. At the community level, deviations from the predicted additive effect may
occur because of emergent indirect effects, as was the case in the experiments presented here.
Indirect effects proved to be stronger than individual direct effects in some cases. The indirect
effects here acted in the opposite direction of direct effects and resulted in reduced evolutionary
change as a result of the sum of indirect and direct effects. Given different species interactions
though, indirect effects might enhance direct effects and result in increased rates of evolution in
a community context.
To understand evolution in a community context, it is necessary to incorporate the
increased complexity that results from increasing species diversity. This dissertation took an
important first step in incorporating such complexity by incorporating multiple direct effects, as
well as indirect effects. However, indirect effects become both more numerous and more
complex as species diversity increases (Abrams 1992). Increasing diversity by only a few
species can result in dozens of new indirect effects via different pathways. In that sense, the
simple three species communities used in the experiments presented here are quite simple
compared to what is likely experienced by protozoan species growing in natural pitcher plant
inquiline communities. This begs the question, how much can we simplify our experimental
systems, yet still adequately understand how species evolve in a natural community context?
Most communities are composed of a few strongly interacting species, with many weakly
interacting species (Paine 1992, Polis 1994, Wootton 1994). This might indicate that only a
small subset of species are required to understand the evolutionary consequences of species
interactions. However, this may not be the case for two reasons. First, the sum of many weak
interactions may have large effects (Silander and Antonovics 1992, Dungan 1986). Second,
indirect effects via longer pathways are poorly understood; species that have weak direct effects
may have greater indirect effects via the sum of multiple indirect effects. This dissertation is a
small, but important, step down the road to understanding evolution in a community context.
49
APPENDIX A
Population Growth Rate
SUPPLEMENTAL DATA FOR CHAPTER 2
0.2
b
0.16
a
a,b
0.12
0.08
0.04
0
No Predator
Density Reduction
Predator
Appendix A.1: Adaptation to Predation in 3 treatments
Population growth rate after selection (measured in the presence of predators) of populations
evolved without predation, with predation, and under low density conditions. Bars indicate pooled cpt
means ( 1 s.e) from the March and June experiments. There was no significant difference among
months. Letters indicate significant pair-wise differences among treatment means (p<0.05, using
Tukey post-hoc tests).
50
Evolutionary History
Population Growth Rate
0.3
6
4
4
0.25
9
No Predator
Predator
10
4
0.2
10
4
0.15
10
10
7
7
0.1
0.05
0
Feb. '06
April '06
May '06
July '06
March '07
June '07
Appendix A.2: Population Growth Rate in All Time Blocks
Population growth rate after selection (measured in the absence of predators) of populations
evolved without predation (white bars) and populations evolved with predation by mosquito larvae
(gray bars) in six independent experiments. Bars indicate means 1 s.e. Numbers above bars
indicate the sample size for each.
51
APPENDIX B
GENBANK ACCESSION NUMBERS
Species Name
Sample Name
Bardeliella pulchra
Genbank
Accession Number
EU039884
Oct1
Genbank
Accession Number
HM215507
Bresslaua vorax
AF060453
Oct2
HM215506
Bresslauides discoideus
EU039885
Oct3
HM215505
Bryometopus pseudochilodon
EU039888
Apr1
HM215508
Colpoda aspera
EU039892
Jun1
HM215509
Colpoda cucullus
EU039893
Jun2
HM215510
Colpoda henneguyi
EU039894
Jun3
HM215511
Colpoda inflata
M97908
Aug1
HM215512
Colpoda lucida
EU039895
Aug2
HM215513
Colpoda magna
EU039896
Colpoda minima
EU039897
Colpoda steinii
DQ388599
Cyrtolophosis mucicola (Brazil)
EU039898
Hausmanniella discoidea
EU039900
Ilsiella palustris
EU039901
Mykophagophrys terricola
EU039902
Notoxoma parabryophryides
EU039903
Ottowphyra dragescoi
EU039904
Oxytricha nova
X03948
Pseudoplatyophrya nana
AF060452
52
REFERENCES
Aarssen, L. W. 1983. Ecological combining ability and competitive combining ability in plants:
toward a general evolutionary theory of coexistence in systems of competition. The
American Naturalist 122:707-731.
Abrams, P. A. 1990. The evolution of antipredator traits in prey in response to evolutionary
change in predators. Oikos 59:147-156.
Abrams, P. A. 1992. Predators that benefit prey and prey that harm predators: unusual effects
of interacting foraging adaptation. The American Naturalist 140:573-600.
Addicott, J. F. 1974. Predation and prey community structure: experimental study of effect of
mosquito larvae on protozoan communities of pitcher plants. Ecology 55:475-492.
Alatalo, R. V., and J. Mappes. 1996. Tracking the evolution of warning signals. Nature 382:708710.
Astles, P. A., A. J. Moore, and R. F. Preziosi. 2005. Genetic variation in response to an indirect
ecological effect. Proceedings of the Royal Society B-Biological Sciences 272:25772581.
Billick, I., and T. J. Case. 1994. Higher-order interactions in ecological communities: what are
they and how can they be detected. Ecology 75:1529-1543.
Bradshaw, W. E., and R. A. Creelman. 1984. Mutualism between the Carnivorous Purple
Pitcher Plant and Its Inhabitants. American Midland Naturalist 112:294-304.
Bradshaw, W. E., and K. Johnson. 1995. Initiation of metamorphosis in the pitcher-plant
mosquito: effects of larval growth history. Ecology 76:2055-2065.
Brodie, E. D., and E. D. Brodie. 1991. Evolutionary response of predators to dangerous prey:
reduction of toxicity of newts and resistance of garter snakes in island populations.
Evolution 45:221-224.
Brown, W. L., and E. O. Wilson. 1956. Character displacement. Systematic Zoology 5:49-65.
Buss, L. W., and J. B. C. Jackson. 1979. Competitive networks: non-transitive competitive
relationships in cryptic coral-reef environments. The American Naturalist 113:223-234.
Caceres, C.E. 1998. Interspecific variation in the abundance, production and emergence of
53
Daphnia diapausing eggs. Ecology 79:1699-1710.
Cadotte, M. W., D. V. Mai, S. Jantz, M. D. Collins, M. Keele, and J. A. Drake. 2006. On testing
the competition-colonization tradeoff in a multispecies assemblage. The American
Naturalist 168:704-709.
Carroll, S. P., A. P. Hendry, D. N. Reznick, and C. W. Fox. 2007. Evolution on ecological time
scales. Functional Ecology 21:387-393.
Case, T. J. 2000. An Illustrated Guide to Theoretical Ecology. Oxford University Press, New
York.
Case, T. J., and E. A. Bender. 1981. Testing for higher-order interactions. American Naturalist
118:920-929.
Caswell, H. 1978. Predator-mediated coexistence: a non-equilibrium model. The American
Naturalist 112:127-154.
Chase, J. M., and M. A. Leibold. 2003. Ecological niches: linking classical and contemporary
approaches. University of Chicago Press, Chicago, IL, USA.
Clements, F. E. 1936. Nature and structure of the climax. Journal of Ecology 24:252-284.
Cochran-Stafira, D. L., and C. N. von Ende. 1998. Integrating bacteria into food webs: studies
with Sarracenia purpurea inquilines. Ecology 79:880-898.
Coley, P. D., J. P. Bryant, and F. S. Chapin. 1985. Resource availability and plant antiherbivore
defense. Science 230:895-899.
Connell, J. H. 1980. Diversity and the coevolution of competitors, or the ghost of competition
past. Oikos 35:131-138.
Connell, J. H. 1983. On the prevalence and relative importance of interspecific competition:
evidence from field experiments. The American Naturalist: 5:661-696.
Connell, J. H., and R. O. Slatyer. 1977. Mechanisms of succession in natural communities and
their role in community stability and organization. The American Naturalist 111:11191144.
Cowles, H. C. 1899. The ecological relations of the vegatation on the sand dunes of Lake
Michigan. The Botanical Gazette 27:95-117,167-202,281-308,361-391.
Dayan, T., D. Simberloff, E. Tchernov, and Y. Yomtov. 1989. Interspecific and intraspecific
character displacement in Mustelids. Ecology 70:1526-1539.
Dumont, H. J., S. Nandini, and S. S. S. Sarma. 2002. Cyst ornamentation in aquatic
invertebrates: a defence against egg predation. Hydrobiologia 486:161-167.
Dungan, M. L. 1986. 3-way Interactions: barnacles, limpets, and algae in a Sonoran Desert
rocky intertidal zone. The American Naturalist 127:292-316.
54
Dunthorn, M., W. Foissner, and L. A. Katz. 2008. Molecular phylogenetic analysis of class
Colpodea (phylum Ciliophora) using broad taxonomic sampling. Molecular Phylogenetics
and Evolution 46:316-327.
Ekelund, F., H. B. Frederiksen, and R. Ronn. 2002. Population dynamics of active and total
ciliate populations in arable soil amended with wheat. Applied Environmental
Microbiology 68:1096-1101.
Englund, G., O. Sarnelle, and S. D. Cooper. 1999. The importance of data-selection criteria:
meta-analyses of stream predation experiments. Ecology 80:1132-1141.
Fenchel, T., G. F. Esteban, and B. J. Finlay. 1997. Local versus global diversity of
microorganisms: cryptic diversity of ciliated protozoa. Oikos 80:220-225.
Fenster, C. B., W. S. Armbruster, P. Wilson, M. R. Dudash, and J. D. Thomson. 2004.
Pollination syndromes and floral specialization. Annual Review of Ecology Evolution and
Systematics 220:375-403.
Foissner, W., B. Weissenbacher, W. D. Krautgartner, and U. Lutz-Meindl. 2009. A cover of
glass: first report of biomineralized silicon in a ciliate, Maryna umbrellata (Ciliophora:
Colpodea). Journal of Eukaryotic Microbiology 56:519-530.
Fox, J. W., and D. A. Vasseur. 2008. Character convergence under competition for nutritionally
essential resources. The American Naturalist 172:667-680.
Fujii, K. 1999. Overview of S. Utida’s research. Researches in Population Ecology 41:11-13.
Futuyma, D. J. and M. Slatkin. 1983. Coevolution. Sinauer Associates, Sunderland, MA.
Gause, G. F. 1934. Experimental studies on the struggle for existence. I. Mixed population of
two species of yeast. Journal of Experimental Biology 9:389-402.
Goldberg, D. E., and A. M. Barton. 1992. Patterns and consequences of interspecific
competition in natural communities: a review of field experiments with plants. The
American Naturalist 139:771-801.
Goodnight, C. J. 1990. Experimental studies of community evolution I: the response to selection
at the community level. Evolution 44:1614-1624.
Goodnight, C. J., and D. M. Craig. 1996. Effect of coexistence on competitive outcome in
Tribolium castaneum and Tribolium confusum. Evolution 50:1241-1250.
Grant, P. R. 1972. Bill dimensions of 3 Species of Zosterops on Norfolk Island. Systematic
Zoology 21:289-291.
Gurevitch, J., J. A. Morrison, and L. V. Hedges. 2000. The interaction between competition and
predation: a meta-analysis of field experiments. The American Naturalist 155:435-453.
Hairston Jr., N. G., S. P. Ellner, M. A. Geber, T. Yoshida, and J. A. Fox. 2005. Rapid evolution
55
and the convergence of ecological and evolutionary time. Trends in Ecology and
Evolution 8: 1114-1127.
Heard, S. B. 1994. Pitcher-plant midges and mosquitos—a processing chain commensalism.
Ecology 75:1647-1660.
Herre, E. A., N. Knowlton, U. G. Mueller, and S. A. Rehner. 1999. The evolution of mutualisms:
exploring the paths between conflict and cooperation. Trends in Ecology & Evolution
14:49-53.
Hoekman, D., C. terHorst, A. Bauer, S. Bruan, P. Gignac, R. Hopkins, S. Joshi, K. Laskis, N.
Sanscrainte, J. Travis, and T. E. Miller. 2007. Oviposition decreased in response to
enriched water: a field study of the pitcher plant mosquito, Wyeomyia smithii. Ecological
Entomology 32:92-96.
Holt, R. D. 1977. Predation, apparent competition, and structure of prey communities.
Theoretical Population Biology 12:197-229.
Huffaker, C. B. 1958. Experimental studies on predation: dispersion factors and predator-prey
oscillations. Hilgardia 27:343-383.
Irwin, R. E. 2006. The consequences of direct versus indirect species interactions to selection
on traits: pollination and nectar robbing in Ipomopsis aggregata. The American Naturalist
167:315-328.
Johnson, M. T. J., and J. R. Stinchcombe. 2007. An emerging synthesis between community
ecology and evolutionary biology. Trends in Ecology and Evolution 22:250-257.
Joshi, A., and J. N. Thompson. 1996. Evolution of broad and specific competitive ability in novel
versus competitive environments in Drosophila species. Evolution 50:188-194.
Kerfoot, W. C. 1977. Competition in Cladoceran communities: cost of evolving predation
against copepod predation. Ecology 58:303-313.
Kneitel, J. M., and T. E. Miller. 2002. Resource and top-predator regulation in the pitcher plant
(Sarracenia purpurea) inquiline community. Ecology 83:680-688.
Kneitel, J. M. 2002. Species diversity and trade-offs in pitcher plant (Sarracenia purpurea)
inquiline communities. Ph.D. Dissertation, Florida State University.
Lau, J. A. 2008. Beyond the ecological: biological invasions alter natural selection on a native
plant species. Ecology 89: 1023-1031.
Lande, R., and S. J. Arnold. 1983. The measurement of selection on correlated characters.
Evolution 37:1210-1226.
Lawler, S. P. 1993. Direct and indirect effects in microcosm communities of protists. Oecologia
93:184-190.
Leibold, M. A., and T. E. Miller. 2004. From metapopulations to metacommunities. Pages 133150 in I. Hanski and O. Gaggiotti, eds. Ecology, Genetics, and Evolution of
56
Metapopulations. Academic Press, San Diego, CA.
Losos, J. B., T. W. Schoener, R. B. Langerhans, and D. A. Spiller. 2006. Rapid temporal
reversal in predator-driven natural selection. Science 314:1111.
Lüftnegger, G., W. Foissner, and H. Adam. 1990. r- and K-selection in soil ciliates: a field and
experimental approach. Oecologia 66:574-579.
MacArthur, R. H. 1972. Geographical ecology: patterns in the distribution of species. Princeton
University Press, Princeton, New Jersey.
MacArthur, R. and R. Levins. 1967. The limiting similarity, convergence, and divergence of
coexisting species. The American Naturalist 101:377-385.
Mauricio, R., and M.D. Rausher. 1997. Experimental evidence of putative selective agents
provides evidence for the role of natural enemies in the evolution of plant defense.
Evolution 51:1435-1444.
May, R. M., and R. H. MacArthur. 1972. Niche overlap as a function of environmental variability.
Proceedings of the National Academy of Sciences of the United States of America
69:1109-1113.
McGoey, B. V., and J. R. Stinchcombe. 2009. Interspecific competition alters natural selection
on shade avoidance phenotypes in Impatiens capensis. New Phytologist 183:880-891.
Menge, B. A. 1995. Indirect effects in marine rocky intertidal interaction webs: patterns and
importance. Ecological Monographs 65:21-74.
Messina, F. J., S. L. Durham, J. H. Richards, and E. D. McArthur. 2002. Trade-off between
plant growth and defense? A comparison of sagebrush populations. Oecologia 131:4351.
Meyer, J. R., S. P. Ellner, N. G. Hairston, L. E. Jones, and T. Yoshida. 2006. Prey evolution on
the time-scale of predator-prey dynamics revealed by allele-specific quantitative PCR.
Proceedings of the National Academy of Science of the United States of America
103:10690-10695.
Miller, T. E. 1994. Direct and indirect species interactions in an early old-field plant community.
The American Naturalist 143:1007-1025.
Miller, T. E. 1995. Evolution of Brassica rapa L. (Cruciferae) populations in intra- and
interspecific competition. Evolution 49:1125-1133.
Miller, T., D. Cassill, C. Johnson, C. Kindell, J. Leips, D. McInnes, T. Bevis, D. Mehlman, D.,
and B. Richard, B. 1994. Intraspecific and interspecific competition of Wyeomyia smithii
(COQ) (Culcidae) in pitcher plant communities. American Midland Naturalist 131:136145.
Miller, T. E., and W. C. Kerfoot. 1987. Redefining indirect effects. Pages 33-37 in W. C. Kerfoot
and A. Sih, editors. Predation: direct and indirect impacts on aquatic communities. New
England University Press, Hanover, New Hampshire.
57
Miller, T. E., J. M. Kneitel, and J. H. Burns. 2002. Effect of community structure on invasion
success and rate. Ecology 83:898-905.
Miller, T. E., and J. M. Kneitel. 2005. Inquiline communities in pitcher plants as prototypical
metacommunities. Pages 122-145 in M. Holyoak, M. A. Leibold, and R. D. Holt, eds.
Metacommunities: Spatial Dynamics and Ecological Communities. University of Chicago
Press, Chicago, IL.
Miller, T. E. and J. Travis. 1996. The evolutionary role of indirect effects in communities.
Ecology 77:1329-1335.
Moen, D. S., and J. J. Wiens. 2009. Phylogenetic evidence for competitively driven divergence:
body size evolution in Caribbean treefrogs (Hylidae: Osteopilus). Evolution 63:195-214.
Mole, S. 1994. Trade-offs and constraints in plant-herbivore defense theory: a life history
perspective. Oikos 71:3-12.
Mooney, K. A. 2006. The disruption of an ant-aphid mutualism increases the effects of birds on
pine herbivores. Ecology 87:1805-1815.
Neill, W. E. 1974. Community matrix and interdependence of competition coefficients. The
American Naturalist 108:399-408.
Nowacki, M., V. Vijayan, Y. Zhou, K. Schotanus, and T. G. Doak. 2008. RNA-mediated
epigenetic programming of a genome-rearrangement pathway. Nature 451:153-158.
Nunezfarfan, J., and R. Dirzo. 1994. Evolutionary ecology of Datura stramonium L. in central
Mexico: natural selection for resistance to herbivorous insects. Evolution 48:423-436.
Paine, R. T. 1992. Food-web analysis through field measurement of per-capita interaction
strength. Nature 355:73-75.
Pianka, E. R. 1974. Niche overlap and diffuse competition. Proceedings of the National
Academy of Sciences of the United States of America 71:2141-2145.
Polis, G. A. 1994. Food webs, trophic cascades and community structure. Australian Journal of
Ecology 19:121-136.
Preisser, E. L., D. I. Bolnick, and M. F. Benard. 2005. Scared to death? The effects of
intimidation and consumption in predator-prey interactions. Ecology 86:501-509.
Quinn, G., and M. Keough. 2003. Experimental Design and Data Analysis for Biologists.
Cambridge University Press, Cambridge.
Reimchen, T. E., and P. Nosil. 2002. Temporal variation in divergent selection on spine number
in threespine stickleback. Evolution 56:2472-2483.
Santos, M., D. J. Borash, A. Joshi, N. Bounlutay, and L. D. Mueller. 1997. Density-dependent
58
natural selection in Drosophila: evolution of growth rate and body size. Evolution 51:420432.
Schluter, D. 1994. Experimental evidence that competition promotes divergence in adaptive
radiation. Science 266:798-801.
Schmitt, R. J. 1987. Indirect interactions between prey: apparent competition, predator
aggregation, and habitat segregation. Ecology 68:1887-1897.
Schoener, T. W. 1983. Field experiments on interspecific competition. The American Naturalist
122: 240-285.
Schröder, A., L. Persson, and A. M. de Roos. 2009. Culling experiments demonstrate size-class
specific biomass increases with mortality. Proceedings of the National Academy of
Science of the United States of America 106:2671-2676.
Schmitz, O. J., P. A. Hamback, A. P. Beckerman. 2000. Trophic cascades in terrestrial systems:
a review of the effect of carnivore removals on plants. American Naturalist 155:141-153.
Seeley, R. H. 1986. Intense natural selection caused by a rapid morphological transition in a
living marine snail. Proceedings of the National Academy of Science of the United States
of America 83:6897-6901.
Shertzer, K. W., S. P. Ellner, G. F. Fussmann, N. G. Hairston. 2002. Predator-prey cycles in an
aquatic microcosm: testing hypotheses of mechanism. Journal of Animal Ecology 71:
802-815.
Shonle, I., and J. Bergelson. 2000. Evolutionary ecology of the tropane alkaloids of Datura
stramonium L. (Solanaceae). Evolution 54:778-788.
Shurin, J. B., E. T. Borer, E. W. Seabloom, K. Anderson, C. A. Blanchette, B. Broitman, S. D.
Cooper, B. S. Halpern. 2002. A cross-ecosystem comparison of the strength of trophic
cascades. Ecology Letters 5: 785-791.
Slobodkin, L. B. 1961. Growth and Regulation of Animal Populations. Holt, Rinehart and
Winston, New York.
Silander, J. A., and J. Antonovics. 1982. Analysis of interspecific interactions in a coastal plant
community: a perturbation approach. Nature 298:557-560.
Simberloff, D., and T. Dayan. 1991. The guild concept and the structure of ecological
communities. Annual Review of Ecology Evolution and Systematics 22:115-143.
Smith, R. A. and M. D. Rausher. 2008. Selection for character displacement is constrained by
the genetic architecture of floral traits in the ivyleaf morning glory. Evolution 62: 28292841.
Sokolowski, M. D., H. S. Pereira, and K. Hughes. 1997. Evolution of foraging behavior in
Drosophila by density-dependent selection. Proceedings of the National Academy of
Science of the United States of America 94:7373-7377.
59
Spears, T., L. G. Abele, and W. Kim. 1992. The monophyly of Brachyuran crabs: a
phylogenetic study based on 18S rRNA. Syst Biol 41:446-461.
Srivastava, D. S., J. Kolasa, J. Bengtsson, A. Gonzalez, S. P. Lawler, T. E. Miller, P. Munguia,
T. Romanuk, D. C. Schneider, and M. K. Trzcinski. 2004. Are natural microcosms useful
model systems for ecology? Trends in Ecology and Evolution 19:379-384.
Stanton, M. L., D. A. Thiede, and B. A. Roy. 2004. Consequences of intraspecific competition
and environmental variation for selection in the mustard Sinapsis arvensis: contrasting
ecological and evolutionary perspectives. The American Naturalist 164:736-752.
Strauss, S. Y. 1991. Indirect effects in community ecology: their definition, study and
importance. Trends in Ecology & Evolution 6:206-210.
Strauss, S. Y., J. A. Lau, T. W. Schoener, P. Tiffin. 2008. Evolution in ecological field
experiments: implications for effect size. Ecology Letters 11:199-207.
Stone, L., and A. Roberts. 1991. Conditions for a species to gain advantage from the presence
of competitors. Ecology 72:1964-1972.
Strong, D. R., L. A. Szyska, and D. S. Simberloff. 1979. Tests of community-wide character
displacement against null hypotheses. Evolution 33:897-913.
Swenson, W., D. S. Wilson, and R. Elias. 2000. Artificial ecosystem selection. Proceedings of
the National Academy of Sciences of the United States of America 97:9110-9114.
terHorst, C.P. in review. Experimental evolution of protozoan traits in response to interspecific
competition. Evolution.
terHorst, C. P., T. E. Miller, and D. R. Levitan. 2010. Evolution of prey in ecological time
reduces the effect size of predators in experimental microcosms. Ecology 91:629-636.
Thompson, J. N. 1998. Rapid evolution as an ecological process. Trends in Ecology and
Evolution 13:329-332.
Trussell, G. C., P. J. Ewanchuk, and M. D. Bertness. 2002. Field evidence of trait-mediated
indirect interactions in a rocky intertidal food web. Ecology Letters 5:241-245.
Van Valen, L. 1973. A new evolutionary law. Evolutionary Theory 1:1-30.
van Veen, F. J. F., P. D. van Holland, and H. C. J. Godfray. 2005. Stable coexistence in insect
communities due to density- and trait-mediated indirect effects. Ecology 86:3182-3189.
Vandermeer, J. H. 1969. Competitive structure of communities: an experimental approach with
protozoa. Ecology 50:362-371.
Wade, M. J. 1977. Experimental study of group selection. Evolution 31:134-153.
Wootton, J. T. 1994. The nature and consequences of indirect effects in ecological
communities. Annual Review of Ecology and Systematics 25:443-466.
60
Wulff, J. L. 2005. Trade-offs in resistance to competitors and predators, and their effects on the
diversity of tropical marine sponges. Journal of Animal Ecology 74:313-321.
Yamaoka, M., T. Watoh, and T. Matsuoka. 2004. Effects of salt concentration and bacteria on
encystment induction in ciliated protozoan Colpoda sp. Acta Protozoologica 43:93-98.
Yoshida, T.,S. P. Ellner, L. E. Jones, B. J. M. Bohannan, R. E. Lenski, N. G. Hairston. 2007.
Cryptic population dynamics: rapid evolution masks trophic interactions. PLOS Biology
5:1868-1879.
Yoshida, T., N. G. Hairston, and S. P. Ellner. 2004. Evolutionary trade-off between defence
against grazing and competitive ability in a simple unicellular alga, Chlorelia vulgaris.
Proceedings of the Royal Society of London Series B-Biological Sciences 271:19471953.
Yoshida, T., L. E. Jones, S. P. Ellner, G. F. Fussmann, N. G. Hairston. 2003. Rapid evolution
drives ecological dynamics in a predator-prey system. Nature 424:303-306.
Zangerl, A. R., and M. R. Berenbaum. 1997. Cost of chemically defending seeds:
Furanocoumarins and Pastinaca sativa. The American Naturalist 150:491-504.
61
BIOGRAPHICAL SKETCH
Casey P. terHorst grew up in the San Gabriel Valley in Southern California and
graduated from the University of Southern California in 1998 with a Bachelor of Arts in History.
After taking classes in the biological sciences for two years, he enrolled in the Marine Biology
Semester at the Wrigley Marine Science Center on Catalina Island. From there, he enrolled in
the Department of Biology at California State University, Northridge, studying the effect of
disturbance in marine fouling communities. Under the guidance of Steve Dudgeon, he finished
his Masters of Science degree in 2004. He enrolled in the Ecology & Evolution doctoral
program in the Department of Biological Science at Florida State University in 2004, under the
guidance of Tom Miller and Don Levitan. He is moving on to a postdoctoral fellowship with Jen
Lau at the Kellogg Biological Station at Michigan State University.
Casey’s research focuses largely on how species evolve in a community context and
how the consequences of such evolution on species interactions. His other interests include
amphipod orientation behavior, ecological succession, how ecological equivalence among
suites of species evolves, and the evolution of life history traits.
Peer-reviewed publications
terHorst, C. P. Evolution in response to direct and indirect effects in pitcher plant inquline
communities. The American Naturalist (in review)
terHorst, C. P. Experimental evolution of protozoan traits in response to interspecific
competition. Evolution (in review)
terHorst, C.P., T.E. Miller and D.R. Levitan. 2010. Discrepancies in evolutionary rates change
ecological effect size of predators on prey. Ecology 91:629-636.
Miller, T. E., C.P. terHorst, and J. H. Burns. 2009. The ghost of competition present. The
American Naturalist 173: 347-353.
terHorst, C.P. and S.R. Dudgeon. 2009. Beyond the patch: disturbance affects species
abundances in the surrounding matrix community. Journal of Experimental Marine
Biology and Ecology 370: 120-126.
Burns, J.H., P. Munguia, B. Nomann, S. Braun, C.P. terHorst, and T.E. Miller. 2008. Vegetative
morphology and trait correlations in 54 species of Commelinaceae. Botanical Journal of
the Linnean Society 158: 257-268.
terHorst, C.P. and P. Munguia. 2008. Measuring ecosystem function: consequences arising
from variation in the biomass-productivity relationship. Community Ecology 9: 36-41.
Levitan, D.R., C.P. terHorst and N.D. Fogarty. 2007. The risk of polyspermy in three congeneric
sea urchins and its implications for gametic incompatibility and reproductive isolation.
Evolution 61: 2009-2016.
62
Hoekman, D., C.P. terHorst, A. Bauer, S. Braun, P. Gignac, R. Hopkins, S. Joshi, K. Laskis, N.
Sanscrainte, J. Travis, and T. E. Miller. 2007. Oviposition decreased in response to
enriched water: a field study of the pitcher-plant mosquito, Wyeomyia smithii. Ecological
Entymology 32: 92-96.
Fierst, J.L., C.P. terHorst, J.E. Kubler, and S. R. Dudgeon. 2005. Fertilization success can
drive patterns of phase dominance in complex life histories. Journal of Phycology
41:238-249.
Presentations
terHorst, C.P. Indirect ecological effects change an evolutionary trajectory. Paper presented at
the Western Society of Naturalists. Seaside, California. November 2009.
terHorst, C.P. Not dead yet: indirect effects make community ecology relevant for ecological
and evolutionary processes. Invited speaker at Florida State University Biology
Colloquium. October, 2009.
terHorst, C.P. and T.E. Miller. Indirect ecological effects alter evolutionary patterns. Paper
presented at the 94th Ecological Society of America annual meeting. Albuquerque, New
Mexico. August 2009.
terHorst, C.P. and T.E. Miller. The ghost of competition present. Paper presented at the
Benthic Ecology Meetings. Corpus Christi, Texas. March 2009.
terHorst, C.P. and T. E. Miller. Evolution in a Community Context. Organizer of Symposium at
the 93rd Ecological Society of America annual meeting. Milwaukee, Wisconsin. August
2008.
terHorst, C.P. and T. E. Miller. Rapid evolution of multiple species as the origin of functional
groups and ecological neutrality. Paper presented at the 93rd Ecological Society of
America annual meeting. Milwaukee, Wisconsin. August 2008.
terHorst, C.P. Context-dependent navigational cues of amphipods living on sandy shores.
Paper presented at the Benthic Ecology Meetings. Providence, Rhode Island. April
2008.
terHorst, C.P. and T.E. Miller. Prey rapidly adapt to predation in pitcher plant inquiline
communities. Paper presented at the Southeastern Ecology and Evolution Conference.
Tallahassee, Florida. March, 2008.
terHorst, C.P. and T.E. Miller. Rapid evolution in aquatic communities: the missing link between
community ecology and evolution. Invited seminar at the Smithsonian Environmental
Research Center. Edgewater, Maryland. March 2008.
terHorst, C.P. and T.E. Miller. Rapid evolution in a community context leads to ecological
neutrality. Paper presented at the Western Society of Naturalists Meeting. Ventura,
California. November 2007.
terHorst, C.P. and T. E. Miller. Convergence of ecological and evolutionary rates: predatordriven evolution of population growth rate. Paper presented at the 92nd Ecological
Society of America annual meeting, San Jose, California. August 2007.
terHorst, C.P. Evolutionary rates on ecological time scales. Presented at the Ecology &
Evolution Seminar. Florida State University. April 20, 2007.
terHorst, C. P. and T. E. Miller. Indirect ecological effects may alter evolution of competitors.
Paper presented at Western Society of Naturalists Meeting, Redmond, Washington.
November 2006.
terHorst, C. P. and T. E. Miller. Evolution of competitors in a community context. Paper
presented at the 91st Ecological Society of America annual meeting, Memphis,
Tennessee. August 2006.
terHorst, C. P. and P. Munguia. Variation in biomass-productivity relationships may result in
flawed measures of ecosystem function. Paper presented at Benthic Ecology Meetings,
Quebec City, Quebec. May 2006.
63
terHorst, C.P. and P. Munguia. Biomass and productivity relationships are like a box of
chocolates. Paper presented at the Western Society of Naturalists Meeting, Seaside,
California. November 2005.
terHorst, C.P. and S. Dudgeon. Differences in successional forces in a marine fouling
community are ameliorated over time. Paper presented at the 90th Ecological Society of
America annual meeting, Montreal, Canada. August 2005
terHorst, C.P. and S. Dudgeon. Community shifts in response to an invasive species. Paper
presented at Benthic Ecology Meetings, Mobile, Alabama. 2004.
terHorst, C.P. and S. Dudgeon. Alternate states and phase shifts: the role of a regionally rare,
non-native taxon in community development. Paper presented at Western Society of
Naturalists Meeting, Long Beach, California. 2003
terHorst, C.P. and S. Dudgeon. Alternative stable states and fouling community structure.
Poster presented at Benthic Ecology Meetings, Groton, Connecticut. 2003.
terHorst, C.P. and S. Dudgeon. Alternative stable states in marine fouling communities? Poster
presented at Western Society of Naturalists Meeting, Monterey, California. 2002.
terHorst, C.P., J.L. Fierst, J.E. Kubler, and S. Dudgeon. The effect of pre-settlement factors on
life history patterns in red algae (Rhodophyta). Paper presented at Benthic Ecology
Meetings, Orlando, Florida. 2002.
terHorst, C.P., J.L. Fierst, J.E. Kubler, and S. Dudgeon. Fertilization success affects the
dominant life history phase in red algae (Rhodophyta). Paper presented at Western
Society of Naturalists Meeting, Ventura, California. 2001
64