Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Mar 6.
Published in final edited form as: Isr J Ecol Evol. 2011 Nov 21;56(3-4):10.1560/IJEE.55.3-4.325. doi: 10.1560/IJEE.55.3-4.325

COEXISTENCE, EXCLUSION, OR NEUTRALITY? A META-ANALYSIS OF COMPETITION BETWEEN AEDES ALBOPICTUS AND RESIDENT MOSQUITOES

Steven A Juliano 1
PMCID: PMC3589801  NIHMSID: NIHMS375737  PMID: 23482823

Abstract

Competition experiments estimating the relative effects of inter- and intraspecific competition can help to resolve whether interspecific competition results in coexistence or exclusion. For mosquitoes, most such experiments have focused on invasive Aedes albopictus and its interactions with resident Aedes. A meta-analysis of such experiments tested whether the effect of interspecific competition is greater than, less than, or equal to that of intraspecific competition, and whether competitive outcomes are dependent on food quality. For A. albopictus and A. aegypti, there was significant context dependence, with interspecific competitive advantage for A. albopictus with low food quality, and competitive equivalence with high food quality. Meta-analysis of survivorship yielded more significant effects than did estimated rate of increase. Competitive effects and competitive responses of each species yielded similar results. This meta-analysis suggests competitive exclusion of A. aegypti by A. albopictus, and is thus consistent with field sampling, qualitative reviews, and interpretations from individual publications. For A. albopictus and A. triseriatus, most results indicated competitive equivalence and no context dependence, and are thus contrary to previous qualitative reviews and to interpretations from individual publications. For both pairs of species, published results suitable for meta-analysis remain scarce, and better experimental designs and improved analysis and reporting of statistical results are needed. Greater emphasis needs to be placed on estimating species’ inter- and intra-specific competitive effects, rather than the more common, but theoretically less interesting, competitive responses. Experiments without low-density controls (i.e., replacement series) are inadequate for comparing competitive effects and responses.

Keywords: interspecific competition, Aedes albopictus, Aedes aegypti, Aedes triseriatus, container-dwelling mosquitoes, competitive effect, competitive response

INTRODUCTION

The archetypical question of community ecology is: Is local stable coexistence of competitors possible, or is competitive exclusion likely (Hutchinson, 1959; Chesson, 2000; Chase and Leibold, 2003)? Fundamental to answering this question is evaluation of the relative strengths of inter- and intraspecific competition (Chesson, 2000). If interspecific competition has a lower impact than intraspecific competition for both competitors, stable local coexistence is possible (Chesson, 2000). If interspecific competition has a greater impact than intraspecific competition for one or both competitors, local competitive exclusion is expected (Chesson, 2000). In contrast, when inter- and intraspecific competition have the same effects, competitors are equivalent, and neutral competitive dynamics may be the result, with community diversity determined by ecological drift (Hubbell, 2001; Adler et al., 2007). These outcomes are interesting to community ecologists posing questions about processes maintaining diversity in specific communities, and to the maintenance of diversity in general (Chesson, 2000; Shea and Chesson, 2002; Adler et al., 2007). When the competitors are vectors of human disease, and when vector species invade new areas, these questions about coexistence can take on practical importance (reviewed by Juliano and Lounibos, 2005; Juliano, 2009). If invaders exclude residents, local disease transmission may be altered by replacement of one vector by another. If invaders and residents coexist, local disease transmission may also change, if the two vectors have different characteristics (e.g., activity periods, vectorial capacity, host choice).

Introductions of nonnative mosquitoes have presented ecologists and medical entomologists with a challenge of understanding how communities of mosquitoes may be altered by those invasions and the possible health implications of those changes. One species, Aedes albopictus, has attracted the most recent attention as an introduced vector and potential competitor of resident mosquitoes (Hawley, 1988; Lounibos, 2002; Juliano and Lounibos, 2005; Juliano, 2009). Aedes albopictus is a widely introduced vector of arboviruses that competes with resident container-dwelling vectors, principally Aedes aegypti and Aedes triseriatus. There have been multiple investigations of competition between A. albopictus and resident species (reviewed by Juliano, 2009). These experiments often contain environmental manipulations in various ways, in an attempt to determine not only what the outcome of competition may be (exclusion, coexistence, neutral dynamics), but also, how those outcomes may vary with ecological context. Attempts to synthesize these results have led to several reviews of this literature (Juliano and Lounibos, 2005; Juliano, 2009) using simple enumeration (“vote counting”—Gurevitch and Hedges, 2001) to evaluate these experiments in aggregate. Formal meta-analysis is widely regarded as a preferable approach to synthesizing results from multiple investigations (Gurevitch et al., 1992, 2000; Gurevitch and Hedges, 1999, 2001), despite some potential limitations of how those meta analyses are implemented (Osenberg et al., 1997, 1999; Osenberg and St. Mary, 1998), and a meta-analysis of these investigations seems likely to provide new and better insight into the nature of competition among these Aedes species.

In this paper, I conduct a meta-analysis of published investigations of competition between larvae of A. albopictus and A. aegypti, testing specifically whether impact of interspecific competition on two population growth correlates, estimated rates of increase and survivorship, is greater or less than the impact of intraspecific competition. Thus, I evaluate whether coexistence or competitive exclusion is likely. A similar analysis is also done for a smaller set of published investigations of competition between larvae of A. albopictus and A. triseriatus. In both cases, I evaluate context dependence in the outcomes of competition for one variable, food quality, which has been postulated to alter the outcome of competition for these species (Murrell and Juliano, 2008; Juliano, 2009). In addition to evaluating expected outcomes for these species, I draw upon existing ecological theory (Chesson, 2000) to derive statistical approaches relevant to the methods used in most competition experiments. The statistical approach described can be useful not only for meta-analysis of competition experiments but also for evaluating the results of single experiments.

COMPETITION MODELS VS. EXPERIMENTS

Though the criterion for stable coexistence of competitors (interspecific competition has less impact than intraspecific competition) is well known, there is a mismatch between the way competition experiments are typically done and analyzed and the effects that must be compared in order to determine whether stable coexistence, exclusion, or neutrality are expected (Goldberg and Barton, 1992; Chesson, 2000). Chesson (2000) presents an alternative version of the Lotka–Volterra model that makes this mismatch clear. For two species, per capita rates of increase are given by:

dN1/N1dt=r1[1-α11N1-α12N2]dN2/N2dt=r2[1-α22N2-α21N1]

where Ns indicate abundance of the two competitors, rs are the maximal rate of increase, and αs are the absolute competition coefficient, quantifying the impact of density of each species (N1, N2) on rates of increase. These are related to the standard relative competition coefficients present in the standard version of the Lotka–Volterra competition model (Chesson, 2000) but differ primarily in that both inter- and intraspecific competition are quantified with coefficients that relate density (N) to decline in rate of increase (dN/Ndt) (Table 1). Graphically, the intraspecific absolute competition coefficients (α11, α22) and the interspecific absolute competition coefficients (α12, α21) define axis intercepts for zero growth isoclines (Fig. 1). Thus, we see from Fig. 1 that for, stable coexistence, 1/α11 < 1/α21 and 1/α22 < 1/α12. In words: the impact of interspecific competition from a species must be less than the impact of intraspecific competition from that species. This observation points out two meanings of the terms “impact of interspecific or intraspecific competition”. Coefficients may be described as competitive effects of a species (i.e., a species’ ability to suppress population growth; effects of species 1 are α11 and α21), or as the competitive responses of a species (i.e., a species’ ability to resist suppression of population growth; responses of species 1 are α11 and α12) (Goldberg and Fleetwood, 1987). Most competition experiments are well designed to compare inter- and intraspecific competitive responses of a species, but what we need in order to understand the potential for coexistence or exclusion is a comparison of inter- and intraspecific competitive effects of a species (Chesson, 2000).

Table 1.

Competitive effects from the standard (relative) two-species Lotka–Volterra model and in Chesson’s (2000) absolute two-species Lotka–Volterra model

Effect of species on species Relative Lotka–Volterra Absolute Lotka–Volterra
1 2 α1 α21 = α1/K1
1 1 1 α11 = 1/K1
2 2 1 α22 = 1/K2
2 1 α2 α12 = α2/K2
Open in a new tab
Fig. 1.

Fig. 1

Open in a new tab

Isoclines for Chesson’s (2000) absolute Lotka–Volterra competition model, showing isoclines for stable coexistence. Parameter definitions in Table 1.

A minimal competition experiment involving two species includes at least 5 treatments. Each species may be raised alone at some density N (treatments N1 and N2 in Table 2). Each species is also raised alone at some greater density (say 2N; treatments 2N1 and 2N2 in Table 2). Finally the two species are raised together at an overall density of 2N (treatment N1N2 in Table 2). More densities may be used, but this experiment represents the smallest response surface design (Goldberg and Scheiner, 2001), and combines the typical addition series and replacement series (Goldberg and Scheiner, 2001). For each replicate of each treatment, some response variable (Y), indicating the success or performance of each species, is quantified. Treatments with both species yield two responses (Y1 and Y2). Data analysis proceeds by analyzing competitive effects of N1 and N2 on means, either Ȳ1or Ȳ2. For example, analysis of Ȳ1 yields a comparison of means for treatments N1 and 2N1 (Y2N1 − YN1), quantifying α11, which is both the intra-specific competitive effect and response of species 1, and also comparison of means for treatments N1N2 and N1 (ȲN1N2 − ȲN1), quantifying α12, which is both the interspecific competitive effect of species 2 and response of species 1. Thus, data analysis is best designed for estimating and comparing inter- and intraspecific competitive responses of each species (Goldberg and Barton, 1992; Chesson, 2000; Goldberg and Scheiner, 2001). Formal statistical comparison of inter- and intraspecific competitive effects of each species would have to include both Y1 and Y2 which is typically not done, because analyses typically focus on each species’ responses individually as a target species (Goldberg and Scheiner, 2001).

Table 2.

Effects used in this meta-analysis. Ni refers to the density of individuals of species i in a treatment. YNi is the mean response variable (e.g., survivorship, rate of increase) from a treatment for species i. When two subscripts are present (e.g., ȲNiNj) the first subscript indicates the species for whom the response is measured. The subscript 2Ni indicates that species i is present at a density twice that indicated by the subscript Ni

1. Treatment means from experiment
Treatment N1 2N1 N1N2 2N2 N2
Mean for species 1 ȲN1 Ȳ2N1 ȲN1N2
Mean for species 2 ȲN2N1 Ȳ2N2 ȲN2
2. Effects calculated from treatment means
Description of effect Calculation (α parameters as in Table 1)
Interspecific vs. Intraspecific competitive effects of species 2 [ȲN1N2 − ȲN1] − [Ȳ2N2 − ȲN2] = α12 − α22 = Competitive Effect2
Interspecific vs. Intraspecific competitive effects of species 1 [ȲN1N2 − ȲN2] − [Ȳ2N1 − ȲN1] = α21 − α11 = Competitive Effect1
Interspecific vs. Intraspecific competitive response of species 1 [ȲN1N2 − ȲN1] − [Ȳ2N1 − ȲN1] = α12 − α11 = ȲN1N2 − Ȳ2N1 = Competitive Response1
Interspecific vs. Intraspecific competitive response of species 2 [ȲN1N2 − ȲN2 ] − [Ȳ2N2 − ȲN2 ] = α21 − α22 = ȲN2N1 − Ȳ2N2 = Competitive Response2
Open in a new tab

Although analyses of most experiments on competition do not include a direct comparison of inter- and intraspecific competitive effects or responses of a species, most experiments, like the minimal response surface described above, include the information necessary to estimate the relevant effect sizes. Thus, most competition experiments can allow a meta-analytic comparison of the inter- and intraspecific competitive effects and responses of a species, and my approach in this meta-analysis is to obtain and to analyze those effects wherever possible. The structure of the necessary comparisons is described in Table 2. For this meta-analysis of competition experiments, I test whether an effect estimating the difference in the interspecific competitive effects and responses and intra-specific competitive effects and responses of a species are significantly different from 0. For a simple competition experiment there would be four treatments involved (Table 2). As constructed in Table 2, if the mean of this effect size across a set of experiments is significantly greater than 0, the effect of interspecific competition is less than that of intraspecific competition. If, instead, the effect is significantly less than 0, the effect of interspecific competition is greater than that of intraspecific competition. I also do a meta-analysis testing whether responses to interspecific competition and intraspecific competition of a species (Table 2) are different, as this test may serve as a readily-available surrogate for the more theoretically relevant comparison of competitive effects of a species. From Table 2, it is evident that estimating the relative effects of inter- vs. intra-specific competition of a species requires treatments at both a low and a high density, and some competition experiments (replacement or substitution series) lack such treatments (Goldberg and Scheiner, 2001). Because replacement series designs are reasonably common in the literature on competition among Aedes (e.g., Novak et al., 1993; Daugherty et al., 2000), the approximate test based on inter- vs. intraspecific competitive responses of a species may be valuable since it can be estimated even when low-density, single-species treatments are absent (Table 2).

METHODS

META-ANALYSIS

The statistical tools of meta-analysis have been widely described in the ecological literature (e.g., Gurevitch et al., 1992, 2000; Hechtel and Juliano, 1997; Goldberg et al., 1999; Hedges et al., 1999; Gurevitch and Hedges, 1999, 2001; Borowicz, 2001; Morris et al., 2007; Vonesh and Blaustein, 2010), and will not be repeated here. All analyses were conducted using MetaWin 2.0 statistical software (Rosenberg et al., 2000). The most widely used meta-analytic measure of effect size is Hedges’ d (Hedges and Olkin, 1985), and this is the principal measure I use. I use Qtotal as my test for heterogeneity among effect sizes, and the random effects model for categorical effects (Gurevitch and Hedges, 1999, 2001; Rosenberg et al., 2000). Use of Hedges’ d has been criticized on several grounds (Osenberg et al., 1997, 1999, Osenberg and St. Mary, 1998) but several of these are circumvented by the approach taken in this analysis. Osenberg and St. Mary (1998) noted that Hedges’ d does not estimate per capita effects, which are often of primary importance for ecological questions such as those posed here. In the present meta-analysis, I compare results from density manipulations and pose a question explicitly about per capita effects: I compare results for manipulating densities of mosquitoes and choose treatments that used the same control and manipulated densities for both intra- and interspecific competition. Thus, although some of these experiments used different densities (see below), I always compare the effects of the same change in intra- and inter-specific densities within an experiment, and so contrast per capita inter- vs. intraspecific competitive impacts. Further, the preferred variable for this analysis was an estimate of per capita increase in cohort experiments with mosquitoes (see below). Osenberg et al. (1999) note that Hedges’ d, and meta-analysis in general, often overemphasize statistical tests of hypotheses even when the null hypothesis of an effect size of 0 is not biologically meaningful. The nature of my biological question renders this null hypothesis meaningful: theory outlined above indicates that testing whether the difference in impacts of inter- and intra-specific competition is 0 is precisely what is needed to assess the outcome of competitive interactions.

A further potential difficulty with Hedges’ d is its assumption that effects are additive (Osenberg and St. Mary, 1998; Osenberg et al., 1999; Morris et al., 2007) even though many biological effects may be multiplicative. Additive competitive effects may best be estimated as a difference between treatment means, but multiplicative competitive effects require an alternative metric involving a ratio of treatment means, like the log response ratio (Osenberg et al., 1999; Morris et al., 2007). In the Lotka–Volterra models that serve as the basis for my evaluation of competition, effects of density on per capita growth rates are assumed to be additive (Gotelli, 1995), suggesting that at least for effect sizes or per capita growth rates, Hedges’ d may indeed be appropriate. A further difficulty with ratios arises when some of the observed values for the meta-analysis take on a value of 0, or when numerator and denominator of the ratio have opposite signs, rendering use of ratios or logarithms impossible. Using log response ratio for meta-analysis of the data sets reported in this paper often yielded non-estimable log response ratios because of one or more of these problems, rendering analysis of log response ratios difficult to interpret. Thus, for this paper, I report only Hedges’ d.

In addition to testing overall effect sizes, I tested for differences between two a priori categories of experimental conditions: high food quality and low food quality (the nature of food quality in each experiment is explained below and in the Appendix). I used Qb (Rosenberg et al., 2000) as my test for differences between these groups. Because sample sizes were relatively small, I used bootstrapping and randomization tests (Manly, 1991) as provided by MetaWin 2.0 (Rosenberg et al., 2000) for statistical inference whenever possible.

DATA

The focus of my analysis is on competitive interactions of Aedes albopictus and its two most commonly investigated competitors, Aedes aegypti and Aedes triseriatus. Papers used in this analysis (see Appendix) were those cited in a recent review of experiments on competition among mosquito larvae (Juliano, 2009) plus publications that appeared shortly after that review was submitted (Leisnham et al., 2009). The minimum experimental design required for analysis comparing effects of inter- and intraspecific competitive effects of a species included treatments described in Table 2. Some experiments contained more treatment combinations (e.g., Braks et al., 2004) and these treatments were ignored for the purposes of meta-analysis. Some experiments lacked a treatment with two competing species at equal abundance (N1N2; Table 2) and instead had two species at an overall high density but in unequal abundances (e.g., Juliano, 1998). For these experiments, I used whichever treatments had species’ abundances that were the closest to equality, as long as there was an appropriate low-density control treatment for comparison. Further, for the data to be used in these analyses, authors had to provide (as summary values, graphs, or raw data) treatment means, some measure of variation (SE, SD, MSE), and sample size. Several potentially useful experimental studies had to be omitted because authors did not provide measures of variation or sample size (e.g., Black et al., 1989; Ho et al., 1989; Livdahl and Willey, 1991; Barrera, 1996). Other experiments on competition between these pairs of species had to be omitted because they did not include the relevant treatments (Juliano et al., 2004; Griswold and Lounibos, 2005; Costanzo et al., 2005a). In addition, several published studies made use of replacement series designs (Novak et al., 1993; Daugherty et al., 2000; Bevins, 2007) or other designs that lack a low-density control (Teng and Apperson, 2000). These investigations could not be used to compare inter- and intraspecific competitive effects of a species, but could contribute to comparisons of inter- and intraspecific competitive responses of a species (Table 2).

Many experiments on these competitors report results of competition under multiple environmental conditions, including different temperatures (e.g., Teng and Apperson, 2000; Lounibos et al., 2002), different food quality (e.g., Daugherty et al., 2000; Yee et al., 2007; Murrell and Juliano, 2008), quantity (e.g., Novak et al., 1993; Braks et al., 2004; Juliano, 2009), or renewal schedule (Bevins, 2007), different presence of pathogens (Aliabadi and Juliano, 2002), or different population origin (Leisnham et al., 2009). Because one of my questions concerns environmental effects on heterogeneity of the outcome of competition, I chose to count as a single experiment each set of replicated cohorts run under one set of conditions. This inevitably results in some dependence of results among linked experiments from the same study (Gurevitch and Hedges, 2001; Gurevitch et al., 1992, 2000). Such nonindependence seems to be present in most ecological meta-analyses (Gurevitch et al., 2000; Morris et al., 2007), and is an inevitable result of attempting to assess the context dependence of ecological effects across multiple studies.

Many experiments on these competitors measure multiple response variables (e.g., estimated rates of increase, survivorship, mean adult sizes, median development times). For the question of coexistence, the most relevant response variable is per capita rate of increase, hence the preferred variable is some estimate of rate of per capita increase. The estimate r′ (Livdahl and Sugihara, 1984) is a demographic estimate of instantaneous per capita rate of increase for a cohort based on survivorship, estimated female fecundity based on adult size, and time to maturity (described in detail by Livdahl and Sugihara, 1984; Livdahl and Willey, 1991). A transformed version of this index, λ′ = exp(r′) is also used commonly and provides an estimate of the finite rate of increase for a cohort (Juliano, 1998). One of these indices was reported by 7 published investigations of competition between A. albopictus and A. aegypti (yielding 23 experiments under distinct conditions), and studies using either index were used for a meta-analysis of effects of competition on rate of increase. All of these investigations also reported usable data on survivorship to adulthood, and meta-analysis of this variable was also done for comparison. One additional investigation (Daugherty et al., 2000) reporting results of a replacement series under 3 experimental conditions was added to the data set for analyses of competitive responses to inter- and intraspecific competition.

For experiments on A. albopictus and A. triseriatus, experimental designs and data reported were much more heterogeneous compared to those used for A. albopictus and A. aegypti. Relatively few investigations (2/5 published studies, yielding 8 experiments) reported estimates of rates of increase and had low density controls. An additional 3 investigations (Novak et al., 1993; Teng and Apperson, 2000; Bevins, 2007), reporting 7 separate experimental conditions, included sufficient data for analysis of survivorship, though all these experiments lacked low-density control treatments.

HYPOTHESES

I tested two hypotheses derived a priori from individual investigations (e.g., Murrell and Juliano, 2008) or from vote counting reviews (e.g., Juliano, 2009) of competition between these species. First, because A. albopictus has been deemed the superior competitor to both A. aegypti and A. triseriatus (e.g., Livdahl and Willey, 1991; Juliano, 1998; reviewed by Lounibos, 2002; Juliano and Lounibos, 2005; Juliano, 2009), I tested the hypothesis that stable coexistence of this invader with these competitors would not be possible. This hypothesis predicts that the interspecific competitive effect of A. albopictus will be significantly greater than its intraspecific competitive effect, and also that the interspecific competitive effect of the other species (A. aegypti or A. triseriatus) will be significantly less than its intraspecific competitive effect. This is a prediction of competitive asymmetry, consistent with A. albopictus competitively excluding either of the other species. Finding that either competitive effects of inter- and intraspecific competition are indistinguishable (i.e., neutrality) or that interspecific competitive effects are less than those of intraspecific competition for both species (i.e., stable coexistence expected) would fail to support this hypothesis.

Second, Juliano (2009) postulated that the outcome of interspecific competition between A. albopictus and A. aegypti or A. triseriatus was dependent on food quality, specifically on the rate of detritus decay or its nitrogen content (see also Livdahl and Willey, 1991; Daugherty et al., 2000; Murrell and Juliano, 2008; Yee et al., 2007). For the purposes of this meta-analysis, I deemed any detritus largely originating from deciduous or coniferous tree leaves to be low quality, and any detritus that included substantial animal material (dead insects, liver powder), yeast, or grass to be high quality (Appendix; see also discussion of the issue of detritus quality by Yee et al., 2007; Murrell and Juliano, 2008; Juliano, 2009). This hypothesis predicts that competitive asymmetry (see above) should be reduced or reversed as food quality increases. Because the same investigations also contributed to vote-counting reviews (e.g., Juliano, 2009) that generated these hypotheses, this meta-analysis cannot be viewed as an independent test of that hypothesis, but instead, serves as a test of whether a hypothesis generated by a qualitative analysis is supported by more rigorous statistical analysis.

RESULTS

As constructed in this paper, a mean effect size significantly >0 indicates that interspecific competition has a lesser effect than does intraspecific competition. A mean effect size significantly <0 indicates that interspecific competition has a greater effect than does intraspecific competition. A mean effect size not significantly different from 0 indicates statistical equivalence of inter- and intraspecific competition.

COMPETITION BETWEEN A. ALBOPICTUS AND A. AEGYPTI

COMPETITIVE EFFECT

For competitive effects of A. albopictus on estimated per capita rate of increase, there were no significant differences in results for high vs. low food quality, and none of the effect sizes were significantly different from 0, indicating that for per capita rate of increase, effects of inter- and intraspecific competition were indistinguishable for A. albopictus. Thus, for per capita rate of increase, meta-analysis is consistent with competitive equivalence of these species. In contrast, for competitive effects of A. albopictus on survivorship, there was a significant difference in the results for high- vs. low-food quality (Fig. 2). Low quality food yielded an effect size significantly <0, indicating that the interspecific effect of A. albopictus was significantly greater than its intraspecific effect. For high food quality, the effect size was positive, but not significantly >0, indicating equivalence of competitive effects with high-quality food.

Fig. 2.

Fig. 2

Open in a new tab

Effect sizes (± bootstrap 95% confidence intervals) for the difference in the inter- and intraspecific competitive effects of A. albopictus (top) and A. aegypti (bottom). Negative values indicate interspecific > intraspecific. Arrows indicate a significant (p < 0.05) difference between low quality food and high quality food. Effect sizes significantly different from 0 are highlighted with an *. Sample sizes: high quality food N = 4; low quality food N = 19.

For competitive effects of A. aegypti on estimated per capita rate of increase, there was also no significant difference between low and high food quality (Fig. 2), and all effect sizes were not significantly different from 0, consistent with equivalence of inter- and intraspecific effects (Fig. 2). For effects on survivorship, although there was no significant difference between low and high food quality, effect sizes for low quality food and for the combined data were significantly >0, indicating that the interspecific effect of A. aegypti on survivorship was significantly less than its intraspecific effect.

In aggregate, the results for competitive effects of these two species are partially consistent with competitive asymmetry, with A. albopictus having a competitive advantage over A. aegypti, particularly at low food quality. The results are consistent with high quality food producing equality of competitors (Fig. 2). The results for competitive effects of these species are not equivalent for estimated rate of increase and survivorship, with analysis of effects on survivorship showing more differences due to food quality, and more differences of inter- and intraspecific effects (Fig. 2).

COMPETITIVE RESPONSE

For competitive responses of A. albopictus estimated per capita rate of increase, patterns of effects were similar for estimated per capita rate of increase and for survivorship (Fig. 3). In both cases, there was a significant difference in effects with high vs. low food quality, with high food quality producing negative effects (significantly <0 for per capita rate of increase, but not for survivorship), and low food quality producing positive effects (significantly >0 for rate of increase, but not for survivorship). Thus, with low food quality, the trend was for A. albopictus to have a better competitive response to (i.e., be less affected by) interspecific competition than to intraspecific competition, whereas with high food quality, the trend was for A. albopictus to have a poorer competitive response to interspecific competition than to intraspecific competition.

Fig. 3.

Fig. 3

Open in a new tab

Effect sizes (± bootstrap 95% confidence intervals) for the difference in the inter- and intraspecific competitive responses of A. albopictus (top) and A. aegypti (bottom). Negative values indicate interspecific > intraspecific. Arrows indicate a significant (p < 0.05) difference between low quality food and high quality food. Effect sizes significantly different from 0 are highlighted with an *. Sample sizes: high quality food N = 6; low quality food N = 20.

For competitive response of A. aegypti estimated per capita rate of increase, differences between low and high food quality were never significant (Fig. 3). Effect sizes for both estimated per capita rate of increase and survivorship tended to be negative and were significantly <0 for the overall test in both cases (Fig. 3), indicating that A. aegypti had a poorer competitive response to interspecific competition than to intraspecific competition.

In aggregate, results for competitive responses of these species were similar to those for competitive effects of these species. Both competitive effects and competitive responses indicate competitive asymmetry favoring A. albopictus, particularly at low-food quality. Only A. albopictus yielded differences in competitive effects or responses with food quality.

Another way to examine effect sizes is to test for correlations across experiments in the effect sizes for competitive effects of each species. For both analyses of estimated rates of increase and of survivorship, these correlations were significantly <0 (Fig. 4), indicating that within one experiment, if one calculated effect size was negative (indicating that interspecific competition had a greater impact than intraspecific competition) the other effect size tended to be positive (indicating that interspecific competition had a lesser impact than intraspecific competition). This pattern strongly suggests competitive asymmetry, with competitive advantage changing across experimental conditions. Plots (Fig. 4) indicate that most experiments fell into quadrants of the graph with one effect >0 and the other <0, and very few effects were both positive (indicating interspecific competition low for both species and suggesting stable coexistence) or both negative (indicating interspecific competition high for both species and suggesting an unstable two-species equilibrium).

Fig. 4.

Fig. 4

Open in a new tab

Bivariate plots and correlation coefficients for the differences in effects of inter- and intraspecific competitive effects of A. albopictus and A. aegypti estimated from analysis of rate of increase (top) and survivorship (bottom). Expected outcomes of points falling into the four quadrants are indicated in boxes.

Effects sizes for competitive effects vs. competitive responses of the two species differed to some extent, and this may be interpreted in multiple ways. This disparity may simply be a result of random error. Alternatively, it may indicate that the two different ways of calculating effects (i.e., the theoretically important competitive effects of each species on the other and on itself, vs. the more traditional competitive responses to inter- and intraspecific competition) actually estimate fundamentally different aspects of competition. Finally, the disparity may derive from the fact that there were more experiments that could yield estimated competitive responses than competitive effect (see Introduction). Re-analysis of competitive responses of each species for only those investigations that also yielded estimates of competitive effects of each species (omitting 3 experiments by Daugherty et al., 2000) produced results virtually identical to those for all estimable competitive effects of each species (results not shown); hence, it seems that the last of these explanations is unlikely.

COMPETITION BETWEEN A. ALBOPICTUS AND A. TRISERIATUS

COMPETITIVE EFFECT

Fewer experiments were available for meta-analysis of competition between A. albopictus and A. triseriatus, and not surprisingly, meta-analysis yielded fewer significant effects and more equivocal conclusions. For competitive effects of A. albopictus, only one effect size, with high food quality and for effects on estimated per capita rate of increase, was significant, and was <0, indicating that the effect of interspecific competition was greater than that of intraspecific competition. Because there were no significant differences due to food quality (Fig. 5), it is probably best to assess overall effect sizes, which were not significantly different from 0 (Fig. 5), indicating equivalent effects of inter- and intraspecific competition. Similarly, for effects of A. triseriatus, there was no difference due to food quality and no significant difference from 0 for the effect sizes (Fig. 5). Thus, the data give no reason to reject the hypothesis of equivalent effects of inter- and intraspecific competition.

Fig. 5.

Fig. 5

Open in a new tab

Effect sizes (± bootstrap 95% confidence intervals) for the difference in the inter- and intraspecific competitive effects of A. albopictus (top) and A. triseriatus (bottom). Negative values indicate interspecific > intraspecific. Effect sizes significantly different from 0 are highlighted with an *. Sample sizes: high quality food N = 4; low quality food N = 4.

COMPETITIVE RESPONSE

Competitive responses of the two species show a somewhat different pattern. There were again no significant differences due to food quality for competitive response of A. albopictus or A. triseriatus for either estimated per capita rate of increase or for survivorship (Fig. 6). For competitive response of A. albopictus, effects with high and low food quality, and overall, were significantly <0, indicating that the response to interspecific competition was less than that to intraspecific competition (Fig. 6). All other effect sizes were not significantly different from 0 (Fig. 6).

Fig. 6.

Fig. 6

Open in a new tab

Effect sizes (± bootstrap 95% confidence intervals) for the difference in the inter- and intraspecific competitive responses of A. albopictus (top) and A. triseriatus (bottom). Negative values indicate interspecific > intraspecific. Effect sizes significantly different from 0 are highlighted with an *. Sample sizes: high quality food N = 4; low quality food N = 4.

Correlations of effect sizes across experiments again were negative, and the correlation was significant for survivorship, but not for estimated per capita rate of increase (Fig. 7). Thus, for this pair of species, correlations suggest competitive asymmetry, particularly survivorship, in contrast to the results estimating overall effects across all experiments. The much more limited data for competition experiments for A. albopictus and A. triseriatus render all conclusions for these species more tentative than those for A. albopictus and A. aegypti.

Fig. 7.

Fig. 7

Open in a new tab

Bivariate plots and correlation coefficients for the differences in inter- and intraspecific competitive effects of A. albopictus and A. triseriatus estimated from analysis of rate of increase (top) and survivorship (bottom). Expected outcomes of points falling into the four quadrants are indicated in boxes.

For A. albopictus and A. triseriatus, there were 3 investigations (Novak et al., 1993; Teng and Apperson, 2000; Bevins, 2007) that did not estimate rates of increase, and did not have low-density controls that would allow for estimation of effects of responses to inter- and intraspecific competition of the two species. These 3 investigations (a total of 6 experimental conditions) did allow for estimation of the competitive response of each of the species for survivorship, and it is interesting to ask what different conclusions may arise from this alternative view of effects of competition. Analysis of all of the experiments, totaling 15 separate experimental conditions, would have changed relatively few conclusions (data not shown). The only major change would have been that for competitive response of A. albopictus survivorship, this broader analysis would have yielded a significant difference in the responses observed with high vs. low food quality. The added experiments all used high food quality, and with these experiments added, with high food quality there was no significant difference in the effects of inter- and intraspecific competition on A. albopictus with high food quality (compare with Fig. 6). For ease of comparison to results for competitive effects of each species, and to results for per capita rate of increase, I report only the meta-analysis of the more limited set of experiments.

DISCUSSION

The central question addressed by this meta-analysis is whether competition between A. albopictus and resident species is likely to lead to: 1) a consistent competitive advantage for one species and resulting competitive exclusion; 2) low interspecific competition and resulting stable coexistence; or 3) competitive equivalence consistent with neutrality. For experiments on competition between A. albopictus and A. aegypti, the results are consistent in indicating that A. albopictus has a competitive advantage over A. aegypti under the conditions used in these laboratory and field experiments, and the meta-analysis suggests that competitive exclusion of A. aegypti by A. albopictus is expected. There was, however, some evidence for context dependence of competitive outcomes, with competitive asymmetry with low quality food resources (decaying deciduous and coniferous leaves), but competitive equality with high quality food (rapidly decaying detritus such as dead insects or grass clippings, yeast). The meta-analysis results for these two species were similar for evaluations of estimated per capita rate of increase and for survivorship, though patterns were clearer for survivorship. Meta-analysis results were also similar for analysis of competitive effects of each species, which is theoretically the most important result for inferring the outcome of competition, and for competitive responses of each species, which is typically what is analyzed in each individual study.

These results are broadly consistent with both patterns observed in much of southern North America (e.g., Hobbs et al., 1991; McHugh 1993; Hornby et al., 1994; Mekuria and Hyatt, 1995; O’Meara et al., 1995; Harrison et al., 1998; Britch et al., 2008; but see Juliano et al., 2004, for exceptions), and with the interpretation of these experiments by individual authors and vote counting reviews (e.g., Juliano and Lounibos, 2005; Juliano, 2009). The meta-analysis conclusion that there is context dependence, with high quality detritus leading to reduced competitive asymmetry, is consistent with a vote-counting review (Juliano, 2009), a laboratory experiment (Murrell and Juliano, 2008), and field distributions of A. aegypti and A. albopictus relative to detritus and nutrient composition (Murrell et al., in press). These combined results suggest an obvious hypothesis for apparent coexistence of these species at some locations in nature: if high-quality detritus is available, and competitive asymmetry is reduced, the time required for competitive exclusion may be long. Long-term persistence (if not stable coexistence) of the two species would then be observed. With competitive equivalence, fitness of two competitors is nearly equal, and even small stabilizing effects (e.g., small differences in resource use or habitat choice) may be sufficient to produce coexistence (Chesson, 2000; Adler et al., 2007). This hypothesis predicts systematic differences in the kinds of detritus (and microbial) resources available to larvae of these species in areas where they coexist vs. areas where A. albopictus has replaced A. aegypti, and this prediction should be testable in the field (Murrell et al., in press).

For competition between A. albopictus and A. triseriatus, results are less clear. Evidence for competitive asymmetry, or any deviation from competitive equivalence, is minimal, and evidence for context dependence lacking. The number of investigations of this pair of species is quite small. These meta-analysis results contrast sharply with conclusions of previous individual investigations (e.g., Livdahl and Willey, 1991; Novak et al., 1993; Teng and Apperson, 2000; Aliabadi and Juliano, 2002; Bevins, 2007; Yee et al., 2007) and previous reviews (e.g., Lounibos, 2002; Juliano and Lounibos, 2005; Juliano, 2009) that have suggested strong advantage for A. albopictus under most circumstances, but some instances of context-dependent reductions of that advantage (e.g., Livdahl and Willey, 1991; Bevins, 2007; Yee et al., 2007). Some of this disparity may arise because a substantial number of the investigations of competition between this pair of species could not be used in the meta-analysis, either because of absence of low-density control treatment (i.e., they were replacement series designs), or because of lack of reporting of measures of error variance. This problem was more frequent for experiments involving A. triseriatus than for experiments involving A. aegypti, both in absolute number of investigations and fraction of all published investigations that had to be excluded.

For both these species pairs, the actual number of experimental studies is still quite small, and the number of experiments done under different conditions is also small (compare to numbers of investigations reported by Gurevitch et al., 1992, 2000; Borowicz, 2001; Morris et al., 2007). The low number of investigations of these competitors is striking because A. albopictus and A. aegypti or A. triseriatus are the most commonly investigated competitors among mosquitoes (Juliano, 2009). Thus, even for these relatively well-investigated species, we lack a sufficient number of investigations to draw firm conclusions about competitive asymmetry, coexistence, and context dependence. Expanding the meta-analysis to include more competitors, such as Aedes japonicus (Armistead et al., 2008) or Culex pipiens (Carrieri et al., 2003; Costanzo et al., 2005b), would increase the number of investigations, but would also broaden the question from one of competitive outcomes between a pair of species to competitive outcomes among competing mosquitoes, in general. The question addressed by controlled (often laboratory) experiments, like most of those reviewed in this paper, cannot be about the existence of competition because the investigators created conditions likely to produce competitive interactions among the studied species. Rather, the questions addressed in these kinds of experiments are about which, if any, species has an advantage, and under what ecological context that advantage exists (Juliano, 2009). In this context, meta-analysis for particular pairs of competitors is the most relevant approach, akin to meta-analyses of effects of particular medical therapies (e.g., Markou et al., 2004) rather than a meta-analysis of detection of effects in field experiments across many taxa and ecosystems (e.g., Gurevitch et al., 1992, 2000; Morris et al., 2007). Clearly, both kinds of approaches would be valuable for investigations of interspecific competition amongst any group of organisms.

Virtually all of the publications used in this meta-analysis derive from two research groups (S.A. Juliano and L.P. Lounibos) that have been very active in investigating invasion biology of A. albopictus in North and South America. Further, many of the experiments included in this analysis are part of the same publications, representing different experimental conditions under which competitive interactions have been investigated. A number of important older studies did not report data in a form suitable for these analyses or did not include low-density control treatments needed for this analysis (e.g., Black et al., 1989; Ho et al., 1989; Livdahl and Willey, 1991; Novak et al., 1993; Barrera, 1996; Teng and Apperson, 2000; Bevins, 2007). As these investigations could be evaluated in vote-counting reviews (e.g., Juliano, 2009) they can contribute to our understanding of competition involving A. albopictus, but the greater rigor of meta-analysis compared to vote counting techniques places more stringent limits on which studies can be used. Nevertheless, the limited data set and lack of independence of most of the investigations raises questions about the generality of the conclusions reached in this review. The problem of independence is often present in meta-analysis (e.g., Gurevitch, 1992, 2000). Because some of the conclusions from the present meta-analysis are similar to those derived from earlier vote-counting reviews using more published data, some confidence in the conclusions may be justified, particularly for the inference that A. albopictus has a competitive advantage over A. aegypti under many conditions, but that high quality resources may reduce or eliminate that advantage. Conclusions about the interactions between A. albopictus and A. triseriatus are much more tentative because of the limited number of nonindependent data available and the discrepancy of meta-analytic and vote-counting results.

Finally, there is a need for new ways of designing and analyzing competition experiments. Typical analyses of individual competition experiments analyze each competitor separately and necessarily focus on competitive responses (in population growth or its correlates) of a species to inter- and intraspecific competition (Goldberg and Scheiner, 2001). However, it is clear from theory (Chesson, 2000) that what is important for determining coexistence or exclusion are effects of inter- and intraspecific competition from a species, which necessarily requires consideration of two response variables. The results of the meta-analysis, particularly for A. albopictus and A. aegypti, suggest that those two kinds of quantifications of competition may be similar (i.e., comparison of competitive responses of a species may be a reasonable approximation of competitive effects of a species). It would nonetheless be advantageous to devise experimental designs that are better for directly addressing competitive effects of a species via formal hypothesis tests. In this context, experiments without low-density controls (i.e., replacement series) are inadequate for evaluating potential for coexistence because they cannot be used to estimate the relative strengths of inter- and intraspecific competitive effects a species. Because there are other criticisms of replacement series designs (Gibson et al., 1999; Joliffe, 2000; Goldberg and Scheiner, 2001), it is advisable that investigators avoid using replacement series designs to investigate interspecific competition. Meta-analysis reduces statistical results to effect sizes, and may suggest a way to use data even from a single two-species experiment to compare effects of inter- and intraspecific effects of a species. By estimating effect sizes for inter- and intraspecific competition, an investigator may be more directly able to compare those effects, perhaps by randomization tests, even within a single experiment.

Acknowledgments

I thank L. Blaustein, R.D. Holt, and R. Ostfeld for the invitation to contribute to this compendium; my colleagues L.P. Lounibos, E.G. Murrell, P. Leisnham, M.P. Daugherty, D.A. Yee, M.A.H. Braks, and B.W. Aliabadi for providing raw data files that facilitated meta-analysis; V.A. Borowicz for sharing her ideas and expertise on meta-analysis; and V.A. Borowicz, L.P. Lounibos, J. Kneitel, and an anonymous referee for useful comments on the manuscript. While working on this paper, I was supported by a grant from NIAID (R01 AI044793, Illinois State University subcontract).

APPENDIX

Source Competitor Venue Food quality Food quantity Temp. (°C) Detritus or resource Other manipulated variables Density 2N Density N Estimated rate of increase Survivorship
Murrell and Juliano, 2008
Aedes aegypti Laboratory Low Low 28 Live oak leaves 40 20 X X
Aedes aegypti Laboratory High High 28 Grass clippings 40 20 X X
Aedes aegypti Laboratory Low Low 28 Slash pine needles 40 20 X X
Aedes aegypti Laboratory High High 28 Dead insects 40 20 X X
Juliano, 1998
Aedes aegypti Field Low Low Container water + live oak leaves 60 20 X X
Aedes aegypti Field Low High Container water 60 20 X X
Braks et al., 2004
Aedes aegypti Field Low Low Container water + low avocado leaves 60 30 X X
Aedes aegypti Field Low High Container water + high avocado leaves 60 30 X X
Daugherty et al., 2000
Aedes aegypti Laboratory Low Low 27 Live oak leaves 40 absent X X
Aedes aegypti Laboratory High High 27 Live oak leaves + low dead insects 40 absent X X
Aedes aegypti Laboratory High High 27 Live oak leaves + high dead insects
Lounibos et al., 2002
Aedes aegypti Laboratory Low Low 24 Container water 80 40 X X
Aedes aegypti Laboratory Low High 24 Container water + live oak leaves 80 40 X X
Aedes aegypti Laboratory Low Low 30 Container water 80 40 X X
Aedes aegypti Laboratory Low High 30 Container water + live oak leaves 80 40 X X
Alto et al., 2008
Aedes aegypti Laboratory High High 28 Live oak infusion + yeast:Lactalbumin 320 160 X X
Alto et al., 2005
Aedes aegypti Laboratory High High 28 Live oak infusion + yeast:Lactalbumin 320 160 X X
Leisnham et al., 2009
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus: E. St. Louis, IL 40 20 X X
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus: Bloomington, IN 40 20 X X
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus:Manassas, VA 40 20 X X
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus: Bartow, FL 40 20 X X
Leisnham et al., 2009
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus: Ft. Denaud, FL 40 20 X X
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus: Gainesville, FL 40 20 X X
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus: Ft. Meyers, FL 40 20 X X
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus: Tampa, FL 40 20 X X
Aedes aegypti Laboratory Low Low 28 Live oak leaves Geographic origin of A. albopictus: Daytona, FL 40 20 X X
Teng and Apperson, 2000
Aedes triseriatus Laboratory High Low+ high pooled 15 Rabbit chow + Yeast: Lact-albumin+ oak leaves 20;200;800 pooled absent X
Aedes triseriatus Laboratory High Low+ high pooled 23 Rabbitc how+Yeast:- Lact-albumin + Oak leaves 20;200;800 pooled absent X
Aedes triseriatus Laboratory High Low+ high pooled 31 Rabbit chow+ Yeast: Lact-albumin+ oak leaves 20;200;800 pooled absent X
Aliabadi and Juliano, 2002
Aedes triseriatus Laboratory Low Low 26 Red oak leaves Infection of A. albopictus: no Ascogregarina taiwanensis 45 15 X X
Aedes triseriatus Laboratory Low Low 26 Red oak leaves Infection of A. albopictus: with Ascogregarina taiwanensis 45 15 X X
Novak et al., 1993
Aedes triseriatus Laboratory High High 27 Liver powder 48 absent X
Aedes triseriatus Laboratory High High 27 Liver powder 48 absent X
Yee et al., 2007
Aedes triseriatus Laboratory High High 27 American elm leaves: dead insect 0:1g 12 6 X X
Aedes triseriatus Laboratory Low Low 27 American elm leaves: dead insect 1:0g 12 6 X X
Aedes triseriatus Laboratory Low Low 27 American elm leaves: dead insect 1:0.1g 12 6 X X
Aedes triseriatus Laboratory High High 27 American elm leaves: dead insect 1:1g 12 6 X X
Aedes triseriatus Laboratory High High 27 American elm leaves: dead insect 10:0.1g 12 6 X X
Aedes triseriatus Laboratory High High 27 American elm leaves: dead insect 10:1g 12 6 X X
Bevins, 2007
Aedes triseriatus Laboratory High High 24 Liver powder SCHEDULE: Detritus input schedule: Added in multiple pulses 48 absent X
Aedes triseriatus Laboratory High High 24 Liver powder Detritus input schedule: Added once 48 absent X
Open in a new tab

References

  1. Adler PB, HilleRisLambers J, Levine JM. A niche for neutrality. Ecol Lett. 2007;10:95–104. doi: 10.1111/j.1461-0248.2006.00996.x. [DOI] [PubMed] [Google Scholar]
  2. Aliabadi BK, Juliano SA. Escape from gregarine parasites affects the competitive impact of an invasive mosquito. Biol Invas. 2002;4:283–297. doi: 10.1023/A:1020933705556. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Alto BW, Lounibos LP, Higgs S, Juliano SA. Larval competition differentially affects arbovirus infection in Aedes mosquitoes. Ecology. 2005;86:3279–3288. doi: 10.1890/05-0209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Alto BW, Lounibos LP, Mores CN, Reiskind MH. Larval competition alters susceptibility of adult Aedes mosquitoes to dengue infection. Proc Royal Soc B. 2008;275:463–471. doi: 10.1098/rspb.2007.1497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Armistead JS, Arias JR, Nishimura N, Lounibos LP. Interspecific larval competition between Aedes albopictus and Aedes japonicus (Diptera: Culicidae) in northern Virginia. J Med Entomol. 2008;45:629–637. doi: 10.1603/0022-2585(2008)45[629:ilcbaa]2.0.co;2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Barrera R. Competition and resistance to starvation in larvae of container-inhabiting Aedes mosquitoes. Ecol Entomol. 1996;21:112–127. [Google Scholar]
  7. Bevins SN. Timing of resource input and larval competition between invasive and native container-inhabiting mosquitoes (Diptera: Culicidae) J Vector Ecol. 2007;32:252–262. doi: 10.3376/1081-1710(2007)32[252:torial]2.0.co;2. [DOI] [PubMed] [Google Scholar]
  8. Black WC, IV, Rai KS, Turco BJ, Arroyo DC. Laboratory study of competition between United States strains of Aedes albopictus and Aedes aegypti (Diptera: Culicidae) J Med Entomol. 1989;32:847–852. doi: 10.1093/jmedent/26.4.260. [DOI] [PubMed] [Google Scholar]
  9. Borowicz VA. Do arbuscular mycorrhizal fungi alter plant–pathogen relations? Ecology. 2001;82:3057–3068. [Google Scholar]
  10. Braks MAH, Honório NA, Lounibos LP, Oliveira RL, Juliano SA. Interspecific competition between two invasive species of container mosquitoes, Aedes aegypti and Aedes albopictus (Diptera: Culicidae), in Brazil. Ann Entomol Soc Am. 2004;97:130–139. [Google Scholar]
  11. Britch SC, Linthicum KJ, Anyamba A, Tucker CJ, Pak EW. Long-term surveillance data and patterns of invasion by Aedes albopictus in Florida. J Am Mosq Control Assoc. 2008;24:115–120. doi: 10.2987/5594.1. [DOI] [PubMed] [Google Scholar]
  12. Carrieri M, Bacchi M, Bellini R, Maini S. On the competition occurring between Aedes albopictus and Culex pipiens (Diptera: Culicidae) in Italy. Environ Entomol. 2003;32:1313–1321. [Google Scholar]
  13. Chase JM, Leibold MA. Ecological niches: linking classical and contemporary approaches. University of Chicago Press; Chicago, Illinois, USA: 2003. p. 221. [Google Scholar]
  14. Chesson P. Mechanisms of maintenance of species diversity. Annu Rev Ecol Syst. 2000;31:343–366. [Google Scholar]
  15. Costanzo KS, Kesavaraju B, Juliano SA. Condition-specific competition in container mosquitoes: the role of non-competing life-history stages. Ecology. 2005a;86:3289–3295. doi: 10.1890/05-0583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Costanzo KS, Mormann K, Juliano SA. Asymmetrical competition and patterns of abundance of Aedes albopictus and Culex pipiens (Diptera: Culicidae) J Med Entomol. 2005b;42:559–570. doi: 10.1093/jmedent/42.4.559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Daugherty MP, Alto BW, Juliano SA. Invertebrate carcasses as a resource for competing Aedes albopictus and Aedes aegypti (Diptera: Culicidae) J Med Entomol. 2000;37:364–372. doi: 10.1093/jmedent/37.3.364. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gibson DJ, Connolly J, Hartnett DC, Weidenhamer JD. Designs for greenhouse studies of interactions between plants. J Ecol. 1999;87:1–16. [Google Scholar]
  19. Goldberg DE, Barton AM. Patterns and consequences of interspecific competition in natural communities: a review of field experiments with plants. Am Nat. 1992;139:771–801. [Google Scholar]
  20. Goldberg DE, Fleetwood L. Competitive effect and response in four annual plants. J Ecology. 1987;75:1131–1143. [Google Scholar]
  21. Goldberg DE, Scheiner SM. ANOVA and ANCOVA, Field competition experiments. In: Scheiner SM, Gurevitch J, editors. Design and analysis of ecological experiments. 2. Oxford University Press; Oxford, UK: 2001. pp. 77–98. [Google Scholar]
  22. Goldberg DE, Rajaniemi T, Gurevitch J, Stewart-Oaten A. Empirical approaches to quantifying interaction intensity: competition and facilitation along productivity gradients. Ecology. 1999;80:1118–1131. [Google Scholar]
  23. Gotelli NJ. A primer of ecology. Sinauer Associates; Sunderland, MA: 1995. p. 206. [Google Scholar]
  24. Griswold MW, Lounibos LP. Competitive outcomes of aquatic container Diptera depend on predation and resource levels. Ann Entomol Soc Am. 2005;98:673–681. doi: 10.1603/0013-8746(2005)098[0673:COOACD]2.0.CO;2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gurevitch J, Hedges LV. Statistical issues in ecological meta-analyses. Ecology. 1999;80:1142–1149. [Google Scholar]
  26. Gurevitch J, Hedges LV. Meta-analysis: combining the results of independent experiments. In: Scheiner SM, Gurevitch J, editors. Design and analysis of ecological experiments. 2. Oxford University Press; Oxford, UK: 2001. pp. 347–369. [Google Scholar]
  27. Gurevitch J, Morrow LL, Wallace A, Walsh JS. A meta-analysis of competition in field experiments. Am Nat. 1992;140:539–572. [Google Scholar]
  28. Gurevitch J, Morrison JA, Hedges LV. The interaction between competition and predation: a meta-analysis of field experiments. Am Nat. 2000;155:435–453. doi: 10.1086/303337. [DOI] [PubMed] [Google Scholar]
  29. Harrison BA, Whitt PB, Powell EE, Hickman EY. North Carolina mosquito records. I Uncommon Aedes and Anopheles (Diptera: Culicidae) J Am Mosq Control Assoc. 1998;14:165–172. [PubMed] [Google Scholar]
  30. Hawley WA. The biology of Aedes albopictus. J Am Mosq Control Assoc. 1988;4 (Suppl):1–40. [PubMed] [Google Scholar]
  31. Hechtel LJ, Juliano SA. Effects of a predator on prey size at and time to metamorphosis: plastic response by prey or selective mortality? Ecology. 1997;78:838–851. [Google Scholar]
  32. Hedges LV, Olkin I. Statistical methods for metaanalysis. Academic Press; Boston, MA: 1985. p. 369. [Google Scholar]
  33. Hedges LV, Gurevitch J, Curtis PS. The metaanalysis of response ratios in experimental ecology. Ecology. 1999;80:1150–1156. [Google Scholar]
  34. Ho BC, Ewert A, Chew L. Interspecific competition among Aedes aegypti, Ae. albopictus, and Ae. triseriatus (Diptera: Culicidae): larval development in mixed cultures. J Med Entomol. 1989;26:615–623. doi: 10.1093/jmedent/26.6.615. [DOI] [PubMed] [Google Scholar]
  35. Hobbs JH, Hughes EA, Eichold BH., II Replacement of Aedes aegypti by Aedes albopictus in Mobile, Alabama. J Am Mosq Control Assoc. 1991;7:488–489. [PubMed] [Google Scholar]
  36. Hornby JA, Moore DE, Miller TW., Jr Aedes albopictus distribution, abundance, and colonization in Lee County, Florida and its effect on Aedes aegypti. J Am Mosq Control Assoc. 1994;10:397–402. [PubMed] [Google Scholar]
  37. Hubbell SP. The unified neutral theory of biodiversity and biogeography. Princeton University Press; Princeton, NJ: 2001. p. 375. [Google Scholar]
  38. Hutchinson GE. Homage to Santa Rosalina, or why are there so many kinds of animals. Am Nat. 1959;93:145–159. [Google Scholar]
  39. Jolliffe PA. The replacement series. J Ecol. 2000;88:371–385. [Google Scholar]
  40. Juliano SA. Species introduction and replacement among mosquitoes: interspecific resource competition or apparent competition? Ecology. 1998;79:255–268. [Google Scholar]
  41. Juliano SA. Species interactions among larval mosquitoes: context dependence across habitat gradients. Ann Rev Entomol. 2009;54:37–56. doi: 10.1146/annurev.ento.54.110807.090611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Juliano SA, Lounibos LP. Ecology of invasive mosquitoes: effects on resident species and on human health. Ecol Lett. 2005;8:558–574. doi: 10.1111/j.1461-0248.2005.00755. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Juliano SA, Lounibos LP, O’Meara GF. A field test for competitive effects of Aedes albopictus on Aedes aegypti in South Florida: Differences between sites of coexistence and exclusion? Oecologia. 2004;139:583–593. doi: 10.1007/s00442-004-1532-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Leisnham PT, Lounibos LP, O’Meara GF, Juliano SA. Geographic divergence in competitive effect and response of the mosquito Aedes albopictus and the potential role of a non-competing life-stage in its invasion success. Ecology. 2009;90:2405–2413. doi: 10.1890/08-1569.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Livdahl TP, Sugihara G. Non-linear interactions of populations and the importance of estimating per capita rates of change. J Anim Ecol. 1984;53:573–580. [Google Scholar]
  46. Livdahl TP, Willey MS. Prospects for invasion: competition between Aedes albopictus and native Aedes triseriatus. Science. 1991;253:189–191. doi: 10.1126/science.1853204. [DOI] [PubMed] [Google Scholar]
  47. Lounibos LP. Invasions by insect vectors of human diseases. Annu Rev Entomol. 2002;47:233–266. doi: 10.1146/annurev.ento.47.091201.145206. [DOI] [PubMed] [Google Scholar]
  48. Lounibos LP, Suárez S, Menéndez Z, Nishimura N, Escher RL, O’Connell SM, Rey JR. Does temperature affect the outcome of larval competition between Aedes aegypti and Aedes albopictus? J Vector Ecol. 2002;27:86–95. [PubMed] [Google Scholar]
  49. Manly BFJ. Randomization and Monte Carlo methods in biology. Chapman and Hall; New York: 1991. p. 281. [Google Scholar]
  50. Markou S, Perimenis P, Gyftopoulos K, Athanasopoulos A, Barbalias G. Vardenafil (Levitra) for erectile dysfunction: a systematic review and meta-analysis of clinical trial reports. Int J Imp Res. 2004;16:470–478. doi: 10.1038/sj.ijir.3901258. [DOI] [PubMed] [Google Scholar]
  51. McHugh CP. Distributional records for Aedes mosquitoes from the U.S. Air Force ovitrap-ping program 1992. J Am Mosq Control Assoc. 1993;9:352–355. [PubMed] [Google Scholar]
  52. Mekuria Y, Hyatt MG. Aedes albopictus in South Carolina. J Am Mosq Control Assoc. 1995;11:468–470. [PubMed] [Google Scholar]
  53. Morris WF, Hufbauer RA, Agrawal AA, Bever JD, Borowicz VA, Gilbert GS, Maron JL, Mitchell CE, Parker IM, Power AG, Torchin ME, Vazquez DP. Direct and interactive effects of enemies and mutualists on plant performance: a meta-analysis. Ecology. 2007;88:1021–1029. doi: 10.1890/06-0442. [DOI] [PubMed] [Google Scholar]
  54. Murrell EG, Damal K, Juliano SA. Distributions of competing aquatic insects depend on detritus types, nutrient ratios, and food availability. Ann Entomol Soc Am. 2011 doi: 10.1603/AN10158. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Murrell EG, Juliano SA. The role of detritus type in interspecific competition and population distributions of Aedes aegypti and Aedes albopictus. J Med Entomol. 2008;45:375–383. doi: 10.1603/0022-2585(2008)45[375:dtatoo]2.0.co;2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Novak MG, Higley LG, Christianssen CA, Rowley WA. Evaluating larval competition between Aedes albopictus and A. triseriatus (Diptera: Culicidae) through replacement series. Environ Entomol. 1993;22:311–318. [Google Scholar]
  57. O’Meara GF, Evans LF, Gettman AD, Cuda JP. Spread of Aedes albopictus and decline of Ae. aegypti (Diptera: Culicidae) in Florida. J Med Entomol. 1995;32:554–562. doi: 10.1093/jmedent/32.4.554. [DOI] [PubMed] [Google Scholar]
  58. Osenberg CW, St Mary CM. Meta-analysis: synthesis or statistical subjugation? Integ Biol: Issues, News and Views. 1998;1:43–48. [Google Scholar]
  59. Osenberg CW, Sarnelle O, Cooper SD. Effect sixe in ecological experiments: the application of biological models in meta-analysis. Am Nat. 1997;150:798–812. doi: 10.1086/286095. [DOI] [PubMed] [Google Scholar]
  60. Osenberg CW, Sarnelle O, Cooper SD, Holt RD. Resolving ecological questions through meta-analysis: goals, metrics, and models. Ecology. 1999;80:1105–1117. [Google Scholar]
  61. Rosenberg MS, Adams DC, Gurevitch J. MetaWin 2.0. Sinauer Associates; Sunderland MA: 2000. p. 129. [Google Scholar]
  62. Shea K, Chesson P. Community ecology theory as a framework for biological invasions. Trends Ecol Evol. 2002;17:170–176. [Google Scholar]
  63. Teng HJ, Apperson CS. Development and survival of immature Aedes albopictus and Aedes triseriatus (Diptera: Culicidae) in the laboratory: effects of density, food, and competition on response to temperatures. J Med Entomol. 2000;37:40–52. doi: 10.1603/0022-2585-37.1.40. [DOI] [PubMed] [Google Scholar]
  64. Vonesh JR, Blaustein L. Predator-induced shifts in mosquito oviposition site selection: a meta-analysis and implications for vector control. Israel J Ecol Evol. 2010;56:263–279. (this issue) [Google Scholar]
  65. Yee DA, Kaufman MG, Juliano SA. The significance of ratios of detritus types and microorganism productivity to competitive interactions between aquatic insect detritivores. J Anim Ecol. 2007;76:1105–1115. doi: 10.1111/j.1365-2656.2007.01297.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES