SEARCH

SEARCH BY CITATION

Keywords:

  • character displacement;
  • Colpoda;
  • convergence;
  • divergence;
  • 18S rRNA;
  • niche;
  • rapid evolution;
  • Sarracenia purpurea

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendices

Decades of experiments have demonstrated the ecological effect of competition, but experimental evidence for competitive effects on trait evolution is rare. I measured the evolution of six protozoan traits in response to competitors from the inquiline community of pitcher plants. Replicate populations of Colpoda, a ciliated protozoan, were allowed to evolve in response to intra- and interspecific competition for 20 days (approximately 100 generations), before traits were measured in two common garden environments. Populations that evolved with interspecific competition had smaller cell sizes, produced fewer cysts and had higher population growth rates relative to populations grown in monoculture. The presence of interspecific competitors led to differential lineage sorting, most likely by increasing the strength of selection. These results are the first to demonstrate protozoan evolution in response to competition and may have implications for species coexistence in this system.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendices

Competition among species for resources is known to be an important species interaction, with implications for the ecology of species and the evolution of species traits. Hundreds of ecological experiments have documented that interspecific competition limits the distribution and abundance of species in natural communities (Connell, 1983; Schoener, 1983; Goldberg & Barton, 1992; Gurevitch et al., 1992). There is considerably less evidence of the evolutionary implications of such competition. Decades of observations of character displacement invoked competition as the likely mechanism driving the divergence of species traits (Brown & Wilson, 1956; Schluter, 2000). However, character displacement has been often difficult to confirm (e.g. Strong et al., 1979), and character displacement alone does not implicate competition as the driving evolutionary force (Connell, 1980). 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.

There are few experimental studies of evolution owing to competition, so a general statement about the evolutionary effect of competition is difficult. Studies of intraspecific competition have 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). The few examples of trait evolution in response to interspecific competition (Taper, 1990; Schluter, 1994; Smith & Rausher, 2008; McGoey & Stinchcombe, 2009) have largely focused on plant taxa. Others have documented significant evolutionary changes in competitive ability, but have not examined the traits that might lead to such a change (e.g. Miller, 1995; Goodnight & Craig, 1996; Joshi & Thompson, 1996). More recently, microcosm studies have proven useful in demonstrating character displacement and evolutionary dynamics among species of bacteria (MacLean et al., 2005; Tyerman et al., 2008).

Historically, protozoa have proven to be useful model systems in ecological competition experiments (e.g. Gause, 1934; Vandermeer, 1969; Fox, 2002), but experimental evolution in these systems has rarely been explored (but see terHorst et al., 2010). 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. Evolutionary changes of six traits in one species were measured in response to both intra- and interspecific competitions.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendices

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 & Kneitel, 2005) and so will be only briefly described here. The host plant is widely distributed, extending from northern Florida in the United States 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. Previous work has demonstrated that the evolution of a particular suite of protozoan traits evolves in response to predation by mosquito larvae (terHorst et al., 2010).

However, competition also plays a large role in structuring these communities (Kneitel, 2002; Leibold & Miller, 2004). A suite of protozoa, along with rotifers, compete for bacteria, presumably via exploitative competition, and can reduce the densities of other bacterivores by 30–100% (Kneitel, 2002). In particular, Pseudocyrtolophosis alpestris (sometimes misidentified as Colpidium in this system) has strong competitive effects on Colpoda and most other bacterivores in the community (Kneitel, 2002).

Selection experiments on Colpoda were conducted in three time blocks. To confirm the species identity of Colpoda lines used in each experiment, 7 of the original 12 stocks (see Selection Experiments), along with two additional stocks collected from the same sites in August 2009, were used to extract DNA for sequencing of a portion of the 18S rRNA gene (Spears et al., 1992, see Appendix A for methods).

DNA sequences were used to create a phylogenetic tree in PAUP (Version 4.0; Sinauer Associates, Inc., Sutherland, MA, USA), 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 (Appendix B).

Selection experiments

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 & 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 four leaves to create four independent stocks that were later mixed to create a pool of genotypes in each replicate.

In each time block, approximately 300 individuals from a mixed-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). Each microcosm contained 20 mL of 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). Protozoa in these laboratory microcosms have similar population dynamics to those in pitcher plant leaves in the field (Miller et al., 1994).

Replicate populations (n = 4 per treatment per block) were maintained in (i) monoculture with only intraspecific competition and (ii) two-species cultures with both intra- and interspecific competition from P. alpestris (hereafter referred to as monoculture and competition, respectively) for 20 days, which encompassed 60–120 Colpoda generations. Asexual reproduction in Colpoda occurs every 4–8 h (Lüftnegger et al., 1990), but some sexual reproduction may occur infrequently (Dunthorn et al., 2008). The abundance of Colpoda in each treatment was measured to determine the ecological effect of P. alpestris. The ecological effect was determined after 24 h to minimize the evolutionary changes that can result in a change in ecological effect size (terHorst et al., 2010).

In the competition treatment, P. alpestris was added at twice the density (approximately 600 individuals) of Colpoda. Although natural densities of these species vary widely, these densities fall within the range found in natural pitchers. To avoid competitive exclusion of Colpoda by P. alpestris and to prevent co-evolution between these species, I used serial transfers to reestablish the experimental microcosms in both treatments every 3 days. A random sample of 300 Colpoda from each replicate was used to inoculate new microcosms with the same conditions as mentioned previously. New P. alpestris were added from laboratory stock cultures maintained in the same culture conditions as mentioned previously. Random samples were collected after inverting the microcosm twice and swirling the contents into a homogenous mixture.

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, approximately 300 Colpoda were isolated and removed from replicates in both monoculture and competition treatments and placed in identical environments containing 20 mL of sterile well water and 6 mg of tetramin. Colpoda were grown under these common conditions for 24 h (3–6 generations) to diminish the role of phenotypic plasticity, because of environmental or maternal effects, before measuring traits.

Measurement of traits

At the onset of the selection experiment, four traits of Colpoda were measured: cell size, cell speed, peak population density and population growth rate. Following 20 days in the selection environments and growth in a common environment for 24 h, 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 refer to the environment in which selection occurred for 20 days as the Selection Environment and the environment in which traits were measured as the Measurement Environment. In each replicate from both selection environments, the four traits that were measured initially were measured in both measurement environments, along with two additional traits: habitat use and cyst production. Traits of individuals are difficult to determine, although individual cell sizes and speeds can be determined. All traits were measured at the population level and reflect the average trait value of individuals within populations.

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. Habitat 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. Habitat use was determined as: log(# in beads/# in water). This dimensionless measure of habitat use provides an unbounded index, with a value of 0 indicating equal densities in the beads and water; a value of 1 indicates an order of magnitude higher density in the beads.

Population growth rates were measured by growing each subsample of 300 individuals in microcosms with 20 mL of sterile well water and 6 mg of tetramin. In the competition measurement environment, 600 P. alpestris 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 6 h, for at least 48 h, or until populations peaked in abundance. Populations displayed exponential growth up to a peak and then abundance dropped off rapidly. Per capita growth rates per hour were determined by estimating r from the best-fit exponential growth model (N0 = N0ert), using abundance data up to the peak. The decline in population abundance in Colpoda is followed shortly by the appearance of resting cysts, which are associated with a decline in food resources (Yamaoka et al., 2004; Foissner et al., 2009). These cysts are resistant to desiccation and may allow populations to survive transient aquatic environments (Ekelund et al., 2002). The peak in density (# per 0.1 mL) was used as a proxy measure of the efficiency of resource use. Cyst production was estimated by counting the number of cysts in a 0.1-mL subsample from each microcosm after 6 days.

Additionally, traits of P. alpestris were measured, but only in the measurement environment with competition. Cell size, cell speed, population growth rate and habitat use of P. alpestris populations were measured in the same way as Colpoda populations previously.

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., Cary, NC, USA). A two-way mixed model anova examined the ecological effect of competition from P. alpestris (fixed effect) and time block (random effect). manova examined the effect of selection environment and measurement environment, testing these mean squares over those of their respective interactions with time block. The effects on individual traits were analysed with a three-way mixed model anova for each of the six traits, using selection environment and measurement environment as fixed factors and time block as a random factor. Maximum likelihoods were calculated for separate linear models with and without all combinations of interactions between main effects. Akaike’s information criterion 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 was not available, and normal quantitative genetic approaches were inappropriate (e.g. Lande & Arnold, 1983). Instead, Pearson product-moment correlations were used to determine whether there were any significant relationships between each pairwise 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.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendices

Species identification

The primers used to sequence the 18S rRNA gene captured a region varying in length from 506–530 bp. These sequences resulted in a poorly resolved tree, but still distinguished two distinct species in the October time block. In the April and June time blocks, all of the stock lines belonged to the same monophyletic group and were considered the same species (Fig. 1). Lines within this clade shared identical sequences, except for an insert at the beginning and end of the sequence in some lines. None of the reference sequences fall within this same group, however, indicating that this portion of this gene in this species has not previously been sequenced. In October, at least one of the stock lines belonged to this same monophyletic group, but two other stock lines do not. One of these lines (Oct1) is likely Colpoda steinii (Fig. 1) and differs from the previously unsequenced species by 32 bp. The other October line (Oct2) falls into its own clade (Fig. 1) and differs from the previously unsequenced species by 16 bp.

image

Figure 1.  Bootstrap consensus tree using parsimony generated using nucleotide sequences from a region of the 18S rRNA gene. 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 (Appendix B).

Download figure to PowerPoint

Prior to the April and June experiments, I recognized two distinct morphotypes among samples and used only a single morphotype thereafter. All stock lines in the previously unsequenced species have similar morphologies and are conspicuously larger than Colpoda steinii, with more obvious colouration and frequently visible vacuoles within the cell. Further, the gliding movement of the unknown species is smoother than the more jerky movement of Colpoda steinii. The Oct2 line 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 those of the unsequenced species. To account for potential species-level selection in the October time block, I analysed the trait values with and without this block. Results were qualitatively similar, particularly with respect to the selection treatment, and for simplicity, I present the anova results with only a single morphotype (April and June time blocks). The analysis including the October time block is available in Appendix C.

Ecological effects of competition

The abundance of Colpoda was reduced significantly by P. alpestris after 24 h (F1,13 = 11.8, P = 0.004, Fig. 2). P. alpestris reduced abundances by 26%, on average. Time block had no significant effect on Colpoda abundance after 24 h (F1,13 = 0.46, P = 0.51), and the block × treatment interaction was not part of the best-fit model. In the October time block, P. alpestris had a similar effect and reduced Colpoda abundance by 24%.

image

Figure 2.  The significant ecological effect of competition from P. alpestris on the abundance of Colpoda after 24 h.

Download figure to PowerPoint

Evolutionary effects of competition

Selection environment had a significant effect on the evolution of Colpoda traits (manova: F6,6 = 52.1, P < 0.001) and affected the evolution of three of the six traits measured: population growth rate, cell size and cyst production. Measurement environment also had a significant effect on Colpoda traits measured (manovaF6,6 = 10.6, P = 0.006), largely because of cyst production and refuge use.

Population growth rate of Colpoda was significantly higher in populations that had evolved with competition than those that evolved in monoculture (F1,27 = 6.72, P = 0.015, Fig. 3a). Neither measurement environment (F1,1 = 0.36, P = 0.65, Fig. 3a) nor time block (F1,1 = 11.1, P = 0.19) had a significant effect on Colpoda population growth rate. The best-fit model included a nonsignificant interaction between time block and measurement environment (F1,27 = 2.49, P = 0.13). The initial population growth rate of Colpoda was lower than both selection environment treatments (r = 0.118 ± 0.009). Population growth rates of P. alpestris were more difficult to measure because populations had not peaked within 48 h. Fitting an exponential curve to the available data may overestimate population growth rate, but gave a population growth rate of 0.086 (± 0.006 SE) – considerably lower than the population growth rate of Colpoda (Fig. 3a).

image

Figure 3.  Averages (±SE) of six traits of replicate Colpoda populations from two different selection environments (monoculture: white bars; interspecific competition: grey bars) measured in two different common garden conditions. Hatched bars represent the ancestral state of population traits (when available) before the selection experiment. In (a), the units are per capita growth rates per hour. In (d), positive values indicate greater density in the refuge environment; negative values indicate greater density in the water column.

Download figure to PowerPoint

Selection environment had a marginally significant effect on cell size (F1,1 = 52.0, P = 0.088). When statistical power was increased by including the October time block, this effect was significant (P = 0.009, Appendix C). Individuals that evolved with competition were smaller than those that evolved in monoculture (Fig. 3b). Although measurement environment (F1,1 = 36.2, P = 0.105) and time block (F1,1 = 7.05, P = 0.229) had no significant effect on cell size, they did have a large effect when the October data were included in the analysis (Appendix C). The best-fit model included all two-way interactions. The interaction between measurement and selection environment was one of magnitude and not direction (F1,25 = 69.2, P < 0.001). Cells selected in competition were always smaller than those in monoculture, and cells measured in competition were always larger than those measured in monoculture (Fig. 3b). Similarly, the interactions between selection environment and time block (F1,25 = 15.4, P < 0.001) and measurement environment and time block (F1,25 = 3.81, P = 0.062) did not affect the interpretation of the main effects. The initial cell size of Colpoda grown in monoculture was larger than either selection environment treatment (1.75 ± 0.1 × 10−3 mm). The average cell size of P. alpestris (mean cell area = 9.5 × 10−4 mm ± 4.6 × 10−5) was smaller than Colpoda (Fig. 3b).

Selection environment has a significant effect on cyst production (F1,28 = 5.67, P = 0.024), as populations of Colpoda that evolved with competition produced significantly fewer cysts than those that evolved in monoculture (Fig. 3c). Measurement environment had a similar effect on cyst production in that populations measured in monoculture produced more cysts than those measured in competition (F1,28 = 6.14, P = 0.020, Fig. 3c). Time block also had a significant effect on cyst production, with more cysts produced in June (F1,28 = 7.89, P = 0.009). The best-fit model did not include any interactions.

Selection environment had no significant effect on Colpoda refuge use (F1,28 = 0.65, P = 0.426, Fig. 3d). Measurement environment had a marginally significant effect on refuge use (F1,28 = 3.73, P = 0.064). Colpoda measured in monoculture used the refuge more heavily than in competition (Fig. 3d). Time block had a significant effect on refuge use (F1,28 = 13.3, P = 0.001), as Colpoda used the refuge more heavily in June than in April. The best-fit model included only these main effects. Conversely, P. alpestris 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 (F1,26 = 0.002, P = 0.970, Fig. 3e), measurement environment (F1,1 = 0.002, P = 0.972, Fig. 3e) or time block (F1,1 = 0.005, P = 0.955) on the peak population density. The best-fit model included two interactions (measurement × selection environment and block × measurement environment), but neither were significant (P > 0.60). Selection environment had no effect on the speed of Colpoda (Fig. 3f, F1,27 = 0.024, P = 0.878). Neither measurement environment (F1,1 = 4.55, P = 0.279, Fig. 3f) nor time block (F1,1 = 19.9, P = 0.140) had a significant effect on cell speed. The best-fit model included a nonsignificant block × measurement environment interaction (F1,27 = 0.014, P = 0.906). The initial speed of Colpoda prior to selection was slightly slower than either selection environment treatment (0.508 ± 0.060 mm s−1). The speed of P. alpestris (0.20 ± 0.007 mm s−1) was much slower than the average speed of Colpoda (Fig. 3f).

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 1). When data were subdivided into different measurement and selection environments, correlations were qualitatively similar to one another. As such, the values in Table 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 this was largely because of 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 nonsignificant (range of r = −0.04–0.16, P > 0.05).

Table 1.   Average correlation coefficients among six traits after bootstrapping data pairs. Asterisks indicate values for which the 95% confidence interval does not include zero.
 Pop. growth ratePeak abundanceCell areaSwimming speedCyst productionRefuge use
Pop. growth rate1.00.36*−0.56*−0.32*0.25*0.29*
Peak abundance 1.0−0.28*−0.37*0.29*0.15
Cell area  1.00.36*−0.17−0.21*
Swimming speed   1.0−0.34*−0.22*
Cyst production    1.00.18
refuge use     1.0

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendices

Protozoa have been used successfully as a model system of resource competition for decades (e.g. Gause, 1934; Vandermeer, 1969; Fox, 2002), 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). Population growth rates and cell sizes measured before the start of the experiment suggest that these traits evolved to some extent in both selection treatments, but that populations in interspecific competition evolved to a greater extent (Fig. 3). The difference between treatments could be a result of increased competition for bacterial resources in the interspecific competition treatment, increasing the strength of selection. Initial densities of interspecific competitors were twice as high as intraspecific competitors, resulting in three times the total density of organisms. Regardless of mechanism, competition has a significant effect on the evolution of these traits.

Genetic variation in these experiments may have come from three sources of biological organization: among species (only in the October time block), among lineages within species and new variance within lineages created by mutations during the experiment. Assuming that reproduction was entirely asexual in Colpoda, lineage sorting at any of these levels is the most likely mechanism of evolution. However, sexual reproduction (Dunthorn et al., 2008) and non-Mendelian inheritance of traits, as observed in other protozoa taxa (Nowacki et al., 2008), may affect the heritability of traits within lineages and may play an important role in evolutionary dynamics beyond simple lineage sorting. However, given that the inclusion of the October time block in the anova analyses produced qualitatively similar results (Appendix C), lineage sorting within species, rather than species sorting, is the most likely mechanism of evolution in all three time blocks. Further, although significant time block effects in this latter analysis (Appendix C) could have been because of differences in species identity (i.e. October different from April or June), Tukey post hoc tests suggest that when time block effects did occur, they were largely because of differences in traits in the June time block. Time block effects may have occurred as a result of seasonal shifts in the bacterial community associated with pitcher plant inquiline communities through time. 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.

Both an increase in cell division rate and cyst production may have allowed these populations to tolerate intra- and interspecific competitions. Genetic lines that increased resource consumption via asexual cell proliferation would have been better equipped to monopolize resources and gain a numerical competitive advantage over other lineages and species. An increase in cell division rate would increase fitness during the 3-day bouts of competition within a microcosm. The faster dividing lineages also encysted more frequently (Table 1), allowing them to tolerate conditions when resources were scarce, but also increasing fitness in a new environment (i.e. a new microcosm or a newly opened leaf). The production of cysts is density dependent and thought to be 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). Such dynamics occur in several invertebrate taxa, where resting cysts allow for the tolerance of poor environmental conditions (e.g. Fenchel et al., 1997; Caceres, 1998; Dumont et al., 2002). Interestingly, cyst production decreased in Colpoda populations that evolved with competition, suggesting that encystment may be triggered by the abundance of conspecifics.

Other traits may have contributed to fitness by increasing survival 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 P. alpestris in the water column (Fig. 3d). Plate counts of bacteria from each habitat suggest that bacterial densities were two orders of magnitude higher in the refuge (9.7 ± 49 × 108 cells mL−1) than in the water column (8.6 ± 72 × 106 cells mL−1). 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 1, C.P. terHorst, unpublished data). Since P. alpestris used the water column the most, the larger Colpoda cells were most likely to experience competition, resulting in selection for decreased cell size (Fig. 3b) as a function of refuge use. Although cell size is positively correlated with competitive effect among species of bacteriovores in the pitcher plant inquiline community (Kneitel, 2002; Leibold & Miller, 2004), this is not true of all protozoan species (Fox, 2002). 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.

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 impose a constraint on 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 each other (Table 1). Yet, evolution in competition resulted in increased growth rates and decreased cyst production (Fig. 3). Had 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. 4). 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. 4).

image

Figure 4.  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.

Download figure to PowerPoint

Traditional niche-based theory dictates that competitors’ traits should diverge through evolutionary time (MacArthur & Levins, 1967; May & MacArthur, 1972; Chase & Leibold, 2003), although more recent theory predicts both convergence and divergence (Fox & Vasseur, 2008; terHorst & Miller, in review). Although there are many documented cases of character divergence (Schluter, 2000), there are considerably fewer, if any, cases of convergence in the literature. In this experiment, different traits converged or diverged from those of the competitor. Population growth rates of Colpoda diverged from P. alpestris traits, but cell sizes of these two competitors converged (Fig. 3). Most theory on character convergence and divergence examines a single trait (e.g. resource use or competitive ability), but these data highlight the need to examine independently evolving traits that may each evolve differently in response to competition.

Evolution is predicted to ultimately alter species interactions, particularly if evolution occurs on an ecological time scale (Carroll et al., 2007; Strauss et al., 2008; terHorst et al., 2010). Evolution in this system was rapid relative to the lifespan of other species in this community (e.g. predatory mosquito larvae) and the lifespan of the leaf itself (approximately 40 weeks). However, analysis of competitive response of Colpoda to P. alpestris over the course of this experiment reveals no significant change in ecological effect (C.P. terHorst, unpublished data). Coevolution between these competitors might maintain ecological effects, but is unlikely since P. alpestris was continually restocked from a stock culture. Colpoda traits that display significant plasticity, such as refuge use or cyst production, may also affect competitive interactions with P. alpestris and obscure changes in ecological interactions caused by evolutionary changes in traits. Ongoing work in this community is attempting to determine whether evolution of traits in natural pitchers affects competitive interactions among a suite of competitors and ultimately affects species coexistence.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendices

D. Levitan and T. Miller provided advice on the experiments and comments on previous versions of the manuscript, as did K. Lotterhos, N. Fogarty, M. Lowenberg, A. Plata-Stapper, J. Fox and one anonymous reviewer. Protozoa traits could not have been measured without assistance from E. Moran, F. Gruber, M. Lowenberg, J. Mola, C. Reed, A. Roman and H. Wells. M. Lowenberg assisted greatly with DNA extraction and PCR methodology. Primer sequences were supplied by T. Spears. This work was supported by a FSU Dissertation Research Grant and Kingsbury Fellowship (FSU English Dept.) to C.P.t and grants from the National Science Foundation to T.E.Miller (DEB 0519170 and DEB 0716891) and D.R. Levitan (DEB 0822626).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendices

Appendices

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendices

Appendix A: DNA extraction, amplification, purification, and sequencing

Genomic DNA was extracted from 100 microlitres (μL) of stock cultures varying in density, from 400 to 11 000 cells mL−1, using SprintPrep® magnetic beads (Beckman Coulter, Brea, CA, USA). Each sample was loaded into a 96-well round-bottom plate, mixed with 80 μL of 100% isopropanol and 10 μL of SprintPrep beads and then placed on a SPRIP 96R® ring magnetic plate (Beckman Coulter) for 5 min. After the supernatant was drained, samples remained on the magnetic plate through five rinses (30 s each) with cold 70% ethanol. Samples were allowed to dry for 30 min to ensure evaporation of the ethanol. The round-bottom plate was removed from the magnetic plate, and 15 μL of ddH20 was added to each sample, and the plate was placed on a shaker for 1 h to unbind the DNA from the magnetic beads. Samples were again placed on the magnetic plate for 5 min to separate the beads from the DNA in the supernatant. All the supernatants were removed, and DNA concentration was measured using a Nanodrop spectrophotometer (Thermo Scientific, Wilmington, DE, USA).

A portion of the 18S rRNA gene was obtained using polymerase chain reaction (PCR) with protocols modified from Spears et al. (1992). One hundred microlitre reactions were composed of 53.5 μL ddH20, 16 μL dNTPs (10 μm), 8 μL MgSO4 (25 mm), 10 μL PlatinumTaq buffer, 0.5 μL DNA Polymerase High Fidelity PlatinumTaq (Invitrogen, Carlsbad, CA, USA), 5 μL 329 forward primer (20 μm), 5 μL HI reverse primer (20 μm) and 2 μL extracted DNA (5 ng μL−1). Primers were amongst those previously used by Spears et al. (1992): 329 forward primer: 5′: TAA-TGA-TCC-TTC-CGC-AGG-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 40 s @ 95 °C, 25 s @ 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 3130 × 1 Genetic Analyzer with Capillary Electrophoresis (Applied Biosystems).

Appendix B: Genbank accession numbers for previous sequences used to create the phylogeny (left column) and new sequences from this study (right column)

Species nameGenbank accession numberSample nameGenbank accession number
Bardeliella pulchraEU039884Oct1HM215507
Bresslaua voraxAF060453Oct2HM215506
Bresslauides discoideusEU039885Oct3HM215505
Bryometopus pseudochilodon EU039888Apr1HM215508
Colpoda asperaEU039892Jun1HM215509
Colpoda cucullusEU039893Jun2HM215510
Colpoda henneguyiEU039894Jun3HM215511
Colpoda inflataM97908Aug1HM215512
Colpoda lucidaEU039895Aug2HM215513
Colpoda magnaEU039896  
Colpoda minimaEU039897  
Colpoda steiniiDQ388599  
Cyrtolophosis mucicola (Brazil)EU039898  
Hausmanniella discoideaEU039900  
Ilsiella palustrisEU039901  
Mykophagophrys terricolaEU039902  
Notoxoma parabryophryidesEU039903  
Ottowphyra dragescoi EU039904  
Oxytricha novaX03948  
Pseudoplatyophrya nanaAF060452  

Appendix C:

 Population growth rateCell sizeCyst production
d.f.FPd.f.FPd.f.FP
Selection Env.1,4110.20.0031,21130.0091,416.460.015
Measure Env.1,20.0290.881,241.70.0231,20.460.57
Time block2,24.30.192,229.50.0332,2100.09
Selection × Measure   1,28.10.104   
Selection × Time   2,20.480.678   
Measure × Time2,413.840.032,20.180.8492,411.470.24
Sel × Meas × Time   2,367.750.002   
 Refuge usePopulation peakCell speed
d.f.FPd.f.FPd.f.FP
  1. anova results using all three time blocks (October, April and June). For each trait, the best-fit model was determined by the lowest Akaike’s information criterion value. Empty cells indicate that the factor was not part of the best-fit model. Significant effects are indicated in bold. Results are qualitatively similar to those using only April and June time blocks, but because of increased power, there are more significant effects here.

Selection Env.1,432.330.131,400.0930.761,20.230.68
Measure Env.1,433.050.091,20.520.391,413.60.064
Time Block2,437.320.0022,21.910.342,211.50.08
Selection × Measure   1,40.250.616   
Selection × Time         
Measure × Time   2,400.580.5662,410.650.53
Sel × Meas × Time