Interspecific Competition between Aedes albopictus and A. sierrensis: Potential for Competitive Displacement in the Western United States

The Asian tiger mosquito, Aedes albopictus, was first detected in North America twenty five years ago. It utilizes water-holding container habitats as immature development sites, and has rapidly spread throughout the eastern United States. Aedes albopictus has occasionally been detected in the western United States, but until recently no established populations of A. albopictus were reported. The western tree-hole mosquito, Aedes sierrensis, is the most common tree-hole mosquito throughout the western United States, and is expected to more frequently encounter A. albopictus. In this study, competition between A. albopictus from the eastern United States and A. sierrensis from the western United States was tested in order to better understand the potential for either competitive displacement of A. sierrensis by A. albopictus or competitive resistance of A. sierrensis to A. albopictus. Varying densities of each species were reared with limited resources in a response surface design. Consistent with a prior study, we found that A. albopictus was clearly a superior larval competitor than A. sierrensis. Aedes sierrensis λ′ (finite rate of increase) decreased with increasing A. albopictus density, but in contrast, A. albopictus λ′ actually increased with increasing A. sierrensis density; a result that was not reflected by individual fitness parameters. These results indicate that A. sierrensis will not be an effective barrier to A. albopictus invasion into tree-holes in the western United States.


Introduction
Interspecific competition is recognized as an important process in structuring many aquatic communities. Classic studies by Paine [1,2] showed that superior competitors could monopolize a community through competitive exclusion. Interspecific competition is often the dominant species interaction determining the success and outcome of biological invasions [3,4]. Many invasive species are superior competitors, which help them establish and spread throughout their introduced range [5,6]. Some invasive species become established and displace competitively inferior native species if there are no native predators or pathogens to control their population [7].
Tree-holes are natural depressions that are formed in trees which over time collect detritus and water. These natural container habitats are called phytotelmata and are utilized by the developmental stages of numerous dipteran insects. Among the most common taxa to colonize tree-holes worldwide are detritivorous mosquitoes (Diptera: Culicidae), whose communities are usually structured by competition for limited resources [8,9]. Some mosquito species that develop in tree-holes may also utilize artificial container habitats that similarly collect rain water and detritus, including tires, cemetery vases, and toys [9]. Tree-holes and artificial containers often occur in close proximity to each other within urban landscapes, allowing some container-utilizing mosquito species to move between the two habitat types.
The Asian tiger mosquito, Aedes albopictus (Skuse), is a containerutilizing invasive species that was first reported in North America in the mid-1980s through used tire trade shipments from Asia [10]. Aedes albopictus has since become widespread over the eastern United States, and is the dominant Aedes species in many urban areas [9]. Aedes albopictus was first reported from southern California in 1972, then again in 2001, after which it was claimed to have been eradicated [11,12]. However, in 2011, mosquito abatement districts in the Los Angeles metropolitan area, CA, discovered A. albopictus and since the initial discovery larvae and adults, have been commonly reported indicating a likely well established A. albopictus populations in the city [13]. Genetic studies comparing A. albopictus in Los Angeles from 2001 vs. 2011 have concluded that the A. albopictus in 2011 are similar to the A. albopictus from 2001, suggesting that that A. albopictus was either not eradicated in 2001and persisted undetected in presumably small densities, or that A. albopictus was reintroduced from the same origin population in Asia [13].
Aedes albopictus utilizes both shaded tree-holes and artificial containers as larval development habitats in the eastern United States [14], where it commonly co-occurs with resident mosquitoes, most notably the eastern tree-hole mosquito, Aedes triseriatus (Say) [9]. Numerous field and laboratory studies have shown that A. albopictus is a superior resource competitor over resident North American mosquito species [7], including A. triseriatus [15][16][17], and that larval mosquito competition is often the dominant species interaction dictating the distributions and abundances of species [18]. Given its successful invasion throughout the eastern United States, the ability to survive climate conditions in southern California, and ability to outcompete resident North American mosquitoes, A. albopictus presents a threat to spread beyond its present distribution in Los Angeles, CA, and increase its range throughout the western United States.
The western tree-hole mosquito, Aedes sierrensis (Ludlow), is the most common mosquito species found in tree-holes in the western United States [19]. Only one study has rigorously tested competition between A. albopictus and A. sierrensis [19], despite the possibility that A. albopictus may have already colonized treehole habitats in the western United States. This study showed that A. albopictus generally performed better in the presence of A. sierrensis than with conspecifics, and that A. sierrensis performed poorly under severe food limitation when larvae developed with A. albopictus. These findings suggest that A. sierrensis may not be a substantive barrier to A. albopictus invasion, and may be competitively excluded from tree-hole habitats. Washburn and Hartman [19] made an important first step in investigating competition between A. albopictus and A. sierrensis, but the study had key limitations that may distort the true outcome of interspecific competition and the invasion potential of A. albopictus.
The first limitation of Washburn and Hartman [19] was the use of a substitutive experiment (replacement series) to test competition, wherein total mosquito density was kept constant, and the densities of each species were varied. Substitutive designs are not recommended for experiments on natural communities [20], because they test only the relative intensity of interspecific and intraspecific competition, and not the occurrence or magnitude of competition [20]. Response surface designs address this limitation by manipulating the density of both focal and associate species [20]. If the competition experiment includes an invasive species, employing a response surface design becomes more important in understanding the success and impact of the invader. The outcome of invasion depends upon the degree of asymmetry between competitors, with competitive exclusion most likely to occur when interspecific competition is highly asymmetrical [4,21,22].
The second limitation of the Washburn and Hartman [19] study was that inferences on the competitive abilities of A. albopictus and A. sierrensis were based solely on individual parameters of fitness. Experimental comparisons of competitive abilities are ideally based on competitive effects on and responses of per capita rate of change [23]. In mosquito competition experiments, population performance can be estimated by calculating an estimate of the finite rate of population increase (l9), which is a composite index based on individual fitness parameters: survivorship, female development time, and female wing length (as a fecundity surrogate). Experimental methods that only consider individual fitness parameters yield limited inference of competitive abilities. For example, mosquito larvae under strong densitydependent competition often grow more slowly, and thus cohorts under strong competition may have the same or greater survivorship as larvae that do not compete, simply because larval development is delayed [24]. Further, l9 is a more biologically meaningful measure of population performance than considering individual fitness parameters, as it accounts for nonlinear interactions among these parameters [25]. Prior experiments using l9 have generated different conclusions for both l9 and survivorship of species, reiterating the importance of including an analysis of l9 in competition studies [24][25][26].
To measure the absolute magnitude of interspecific and intraspecific competition, of A. albopictus and A. sierrensis we employed a response surface design, design using A. albopictus from the eastern United States and A. sierrensis from the western United States, in which regression slopes of population performance vs. heterospecific and conspecific densities quantify per capita competitive effect and response to interspecific and intraspecific competition, respectively [23]. Based on prior experiments that have demonstrated the superior competitive capabilities of A. albopictus, we predict A. albopictus will have greater competitive effect, or better competitive response, than A. sierrensis.

Collection and maintenance of mosquitoes
Aedes albopictus larvae were collected from multiple populations in Maryland, Virginia, and New Jersey, United States (A. albopictus are not endangered species and permits are not required to collect them). Field collected larvae of A. albopictus were reared to adulthood at 26uC at 16:8 (L:D) h photoperiod and then released into 0.5-m 3 cages. Adults were kept at 26uC and 75% RH at 16:8 (L:D) h photoperiod. Adults had continuous access to 20% sugar solution. Females were regularly fed anesthetized mice (IACUC protocol: R-12-41, approved by the University of Maryland Institutional Animal Care and Use Committee), and laid eggs on seed paper in water-filled cups. Aedes sierrensis larvae were collected from tree-holes within Salt Lake City, Utah, USA (A. sierrensis are not endangered species and permits are not required to collect them) and were reared at 26uC at 16:8 (L:D) h photoperiod and then adults were released into a 0.5-m 3 cage. The adult females were fed horse blood with the Hemotek blood feeding system, and were allowed to lay eggs on paper napkins in black, water-filled cups. Field collected larvae for each species originated from urban and suburban landscapes, which are representative of where A. albopictus is known to have invaded in California, and also where further spreading is expected to occur. F 1 generation individuals from each colony were used in the experiment so that experimental populations would have similar competitive abilities as populations in the field where competition is important, and thus be able to better determine potential effects of competition on A. sierrensis persistence and A. albopictus expansion.

Competition
Both species were hatched synchronously in a solution of 0.30 g nutrient broth per 1 L distilled water. Within 24 h, larvae were rinsed and transferred into the experiment. The experiment consisted of the following initial combinations of larvae (A. albopictus: A. sierrensis): 10:0, 20:0, 40:0, 10:10, 20:20, 10:30, 30:10, 0:10, 0:20, and 0:40 to create an asymmetric response surface design [20]. These density combinations have been shown to reflect the field densities and have been used in experiments to evaluate competition between other container mosquitoes including A. albopictus [9]. Each combination was replicated five times yielding 50 experiment units. 400 ml cups were filled with 350 ml distilled water and provisioned with 0.70 g of dried senescent white oak (Quercus alba L.) leaves. Although Q. alba are not native to the southwestern United States, they are one of the most common trees in urban and suburban areas (in the region (including Los Angeles, CA), and the leaves are frequently found in containers that A. sierrensis inhabit [27]. Cups were set up four days prior to the addition of larvae and inoculated with tree-hole water (100 mmol) to allow microbial communities to establish. On days 14, 28, 42, and 56 after the start of each replicate, 0.70 g of additional dried live oak was added to each cup to avoid complete resource depletion and to mimic natural conditions. All cups were maintained at 350 ml by being topped up with distilled water daily to account for evaporative water loss.
The experiment was housed in an environmental chamber at 26uC and 14:10 (L:D) h photoperiod to approximate summer climate and photoperiod conditions in the southwestern United States. Treatments were randomly assigned cups and cup position was shuffled daily. Each day we collected pupae into individual vials and held them until adult emergence. Adults were killed by drying (24 h, 50uC) and females were weighed and their wing lengths measured. For each cup, the proportion of survivorship to adulthood (both sexes), mean female dry mass, and mean female wing length was recorded. Daily eclosion of females and their wing lengths were used to calculate l9, a composite index of population finite rate of increase based on r9, which estimates the realized per capita rate of population change (dN/N dt = r, the exponential growth rate) for each replicate cohort (Juliano 1998): where N 0 is the initial number of females (assumed to be 50% per microcosm), x is the mean time to eclosion (measured in days), A x is the mean number of females eclosing on day x, w x is the mean body size on day x, and f(w x ) is a function describing size dependent fecundity for each species, estimated from the mean wing length on day x, wx, of female mosquitoes [25,26]. D is the mean number of days it takes for an adult mosquito to mate, blood-feed, and oviposit. D is estimated at 14 days for A. albopictus [28]. There is no estimate for D with regards to A. sierrensis in the literature; therefore we use the estimate for the ecologically similar eastern tree-hole mosquito, A. triseriatus, which is 12 days.

Aedes sierrensis size-fecundity
We used a regression equation relating female wing length to fecundity for A. albopictus: f(w x ) = 2121.240+78.02w x , where w x is wing length (mm) [28]. We found no regression equation of A. sierrensis body size on fecundity in the literature. Thus, A. sierrensis larvae were reared to adulthood in the laboratory in order to examine the relationship between female wing length and numbers of eggs. Larvae were reared at low densities (20 larvae per 200 ml) in 250-ml cups provisioned with either 20 or 30 mg of bovine liver power with the goal of providing variable submaximal nutritional levels to produce a wide range of adult sizes. As adults eclosed, they were placed in 20-L plastic cages and within 5-10 days were fed to repletion from an anaesthetized mouse, then isolated in 600 ml containers with a 40 ml cup of water lined with seed paper for oviposition. For each female, oviposited eggs were counted. After oviposition, females were killed, dissected, and the number of mature eggs [stages 4 and 5, 1] in their ovaries counted. Fecundity was calculated by adding laid and unlaid mature eggs. Wings of all females were removed and measured. A total of 69 females entered the experiment. Killing and dissecting females after the first gonotrophic cycle is consistent with most prior studies that have examined the fecundity of A. sierrensis [2]. Data on the parity of wild A. sierrensis females suggests that the average female matures with one batch of eggs [3]. Linear regression of number of eggs vs. wing length and wing length vs. female dry mass were both highly significant (Fig. 1A,B). The regression of number of eggs vs. wing length was used to calculate A. sierrensis l9.

Statistical analyses
For each species, linear models (PROC GLM, SAS Institute 2004) with effects of densities of A. albopictus and A. sierrensis (continuous variables) were tested with l9 and its demographic fitness parameters (survivorship, mean female mass, mean male mass, mean female development time, and mean male development time) as dependent variables. To better approximate parametric assumptions of normality and homogeneity of variance, we log 10 +1 transformed A. albopictus l9 and arcsine-square-root +0.5 transformed A. albopictus survival from the mosquito competition. No transformations allowed A. sierrensis data to meet parametric assumption, hence tested for effects using randomization models [Randomization-wrapper for SAS PROCs ; 29]. Randomization models yielded the same significant effects as the parametric models; therefore we report only parametric results.

Results
Both A. sierrensis and A. albopictus l9 were affected by heterospecific densities but not conspecific densities (Table 1). Aedes sierrensis l9 decreased sharply with increasing A. albopictus density whereas A. albopictus l9 actually increased with increasing A. sierrensis density ( Fig. 2A). Aedes sierrensis survivorship was negatively affected by densities of both conspecifics and heterospecifics, whereas A. albopictus survivorship was not affected by either A. albopictus or A. sierrensis densities (Fig. 2B). Aedes albopictus female developmental time was negatively affected by conspecific density, but not affected by heterospecific density (Table 1; Fig. 2C). Female development time of A. sierrensis and female mass of both A. albopictus and A. sierrensis were not affected by either conspecific or heterospecific densities (Table 1; Fig. 2C, D).

Discussion
Experimental comparisons of competitive abilities are ideally based on competitive effects on and responses of l9. Using a response-surface design, we tested competition between the dominant tree-hole mosquito in the western United States, the native A. sierrensis, and the exotic congener A. albopictus, which has recently been collected in large numbers in the Los Angeles metropolitan area, CA. We found A. sierrensis l9 sharply decreased with increasing A. albopictus density. In contrast, A. albopictus l9 actually increased with increasing A. sierrensis density; a result that was not reflected by individual fitness parameters. These results are broadly consistent with the findings of the only prior experiment on competition between these two Aedes species, suggesting that A. albopictus is a superior competitor to A. sierrensis. Based on these results, we may expect that A. albopictus will not only competitively exclude A. sierrensis from tree-holes in the western United States, but that its range expansion in this part of the country may actually be facilitated by the presence of the competitively inferior native.
Our response-surface experiment showed that the superior competitive ability of A. albopictus over A. sierrensis consisted of both a strong competitive effect and competitive response of A. albopictus, and no competitive effect and poor competitive response of A. sierrensis. The negative effects of one species on another (competitive effect) can be influenced by ecological and physiological factors. Ecological factors include the ability to better harvest and deplete a scarce resource, such as food [22], and the physiological factors include the ability to manage metabolic demands depending on the availability of resources [23,30,31]. It has been shown that by manipulating resource levels resource depletion could be the principal factor involved in competition between Aedes [26,32]. However, both A. albopictus and A. sierrensis may also be affected by interference competition produced by water-borne substances [33][34][35]. While this study clearly demonstrated competitive superiority of A. albopictus over A. sierrensis, further investigation is needed to understand the specific mechanisms that make A. albopictus competitively dominant over A. sierrensis.
Of particular interest is our unexpected result that A. albopictus l9 increased with higher A. sierrensis density. The most plausible explanation for this result is that decaying A. sierrensis carcasses provided additional food resources for A. albopictus. Past studies have shown that larvae raised on insect detritus develop faster and attain larger body size than larvae raised on plant detritus [36][37][38][39], and that the negative effects of resource competition can be eliminated by supporting higher quantities (and possibly different species) of microorganisms [37][38][39]. If density dependent mortality of A. sierrensis occurred primarily at the first and second instars, then it is probable that the growth and development of surviving, predominantly A. albopictus, larvae would have utilized the pulse of animal detritus and associated microbial production to support their growth and development. This is especially true if A. albopictus larvae were better able to feed on the resultant microbial growth than surviving A. sierrensis. Classic competition theory predicts that, for coexisting species, intraspecific competition is greater than interspecific competition. The findings of this experiment are inconsistent with this prediction, and suggest that A. albopictus should competitively exclude A. sierrensis when they co-occur in the same larval habitat. Superiority in interspecific competition is often listed as a characteristic of non-native species that enhances the likelihood of becoming invasive [40]. However, despite being a similarly dominant competitor over almost all resident mosquitoes in the eastern United States [7], tree-hole based A. albopictus has failed to competitively exclude these species from many areas. Numerous hypotheses have been proposed for the coexistence of competitively inferior Aedes with A. albopictus, including condition specific competition [41], differential susceptibility to low temperatures [16], interspecific aggregation among individual containers [24], spatial partitioning among landscape variables [24], differential vulnerability to intraguild predation [42], and trade-offs between competitive ability and susceptibility to other predators or parasites [43]. These ecological processes may be important in the invasion success of A. albopictus in the western United States, and mediating its impact on A. sierrensis as well as other resident species.
This study employed a laboratory-based response surface design to test larval competition between A. albopictus and A. sierrensis consistent with other laboratory studies that have rigorously examined competition between A. albopictus and resident mosquito species in North America [7]. Laboratory-based competition studies are powerful at addressing biological details of competition [7], including the relative roles of competitive effect and response, which was a focus in this study. To answer questions about the impact of A. albopictus larval competition on A. sierrensis in nature requires field experiments to manipulate species densities under realistic conditions [20]. Although conditions in this study mimicked those in nature, only larval densities were manipulated. Future studies on larval competition between these Aedes species should manipulate other factors in addition to larval density to understand how larval competition may vary across habitat gradients specific to western United States. Important abiotic and biotic factors that have been shown to affect the outcomes of interspecific competition involving A. albopictus, including temperature [16], habitat permanence, resource amount and type [29], nutrient concentrations, and interactions with third species such as parasites or predators [44,45].
Among the most well documented and likely important ecological processes that promote coexistence of resident Aedes with A. albopictus is the higher susceptibility of A. albopictus to shared predators and parasites. In the eastern United States, the predators Corethrella appendiculata (Grabham) and Toxorhynchites rutilus (Coquillet) produce strong top-down pressure on tree-hole communities in forested areas dominated by A. triseriatus [44,46]. Aedes triseriatus larvae exhibit a higher frequency of low risk behaviors in the presence of predation risk cues from C. appendiculata and T. rutilus larvae when compared to A. albopictus larvae [47]. Field surveys have also shown that A. triseriatus populations are positively correlated with C. appendiculata and T. rutilus whereas A. albopictus are negatively correlated [48]. These predation studies have concluded that T. rutilus and C. appendiculata are keystone species which act as barriers to complete invasion by A. albopictus in the eastern regions of the United States. Tree-hole communities in the western United States do not appear to have predators that produce the same top-down pressure on larval mosquitoes, and may be vulnerable to the invasion of A. albopictus. Therefore, in the absence of predatory barriers, it is highly probable that A. sierrensis will be displaced by competitively superior A. albopictus from treeholes. On the other hand, evidence from experimentally placed oviposition traps suggests that the relative absence of A. albopictus from forested areas may also be due to an oviposition preference for open habitats [49].
Ascogregarina is a genus of protozoan parasites found in tree-holes that infect many mosquitoes in the eastern United States. Studies show that Ascogregarina has more impact on the invasive A. albopictus than the native eastern tree-hole mosquito, A. triseriatus [50]. Competition experiments between A. albopictus and A. triseriatus in the presence of Ascogregarina infections indicate that these parasites might be a disadvantage for range expansion of A. albopictus in the eastern United States [51]. The western tree-hole mosquito, A. sierrensis, is commonly infected by the endoparasite Lambornella clarki [45]. Lambornella clarki has significant deleterious effects on A. sierrensis, and has been proposed to be a good biological control agents [52]. However, studies have shown that A. albopictus is not a suitable host for L. clarki, and that the parasite will not impede A. albopictus [19]. Thus, it's likely that L. clarki will further facilitate invasion by preferentially infecting A. sierrensis, but not A. albopictus. In spite of any abiotic or biotic factors that may affect competition with A. albopictus at the larval stage, this study showed A. albopictus to be an overwhelmingly dominant competitor over A. sierrensis. Varying any of these factors is unlikely to reverse the outcome of larval competition in favor of A. sierrensis but rather the intensity of competition.
Although direct interactions among mosquito species are largely restricted to the larval life-stage, variation in the responses of species to environmental gradients at other life stages can affect the outcome of population-level competition [41]. In addition to larval competition for food resources, asymmetric reproductive competition might also be important in determining the population-level competition between A. albopictus and A. sierrensis. Upon mating, female mosquitoes generally become less interested in further mating due to proteins that were transferred by the male [53]. Recent findings also indicate that A. albopictus males mate with female A. aegypti more frequently than A. aegypti males with female A. albopictus thereby reducing their relative reproductive fitness as interspecific mating does not produce any offspring [53]. Sympatric A. aegypti females are more resistant to mating by A. albopictus males than allopatric females indicating a reproductive character displacement [54]. Similar asymmetric reproductive competition might be relevant for other mosquitoes that compete with A. albopictus including A. sierrensis.
Recent discovery of an established population of A. albopictus in the Los Angeles metropolitan area, CA is an important reason to revisit the effect of competition between A. albopictus and A. sierrensis. Consistent with the only prior study to examine competition between these species by Washburn and Hartman [19], we found strong evidence for competitive superiority of A. albopictus over A. sierrensis, supporting the exclusion of A. sierrensis from tree-holes where both species co-occur. Unlike in the eastern United States, there appear to be no predators and few parasitic barriers to A. albopictus invasion in tree-holes in the western United States. Despite the absence of these natural enemies in the western United States, A. albopictus has only emerged as an invasive threat in the southwestern region in the last two years, since its prior introduction and reported eradication in 2001. Genetic studies comparing A. albopictus individuals from 2001 and 2011 have concluded that the A. albopictus found in 2011 could be the descendants of the 2001 population, consistent with the idea that A. albopictus were never eradicated in 2001, or the alternative idea that a re-introduction of A. albopictus has occurred from the same region in Asia. The most common hypothesis for the lack of A. albopictus in Los Angeles, CA, either spreading from the eastern United States or from a re-introduction from Asia has been that the southwestern United States is too dry and hot. However, inconsistent with this hypothesis have been reports of A. albopictus from other states in the western United States [55]. In the past, when A. albopictus has been reported from western states their initial populations appear to never survive the winter or were reported to have been eradicated [11]. Aedes albopictus discovered in Los Angeles in 2011 appear to be persisting and possibly spreading [13]. These findings indicate the likely possibility that A. albopictus is now firmly established in the region. Dry and hot climates are likely to be still a barrier to A. albopictus spread, and in the past these conditions have likely prevented a rapid invasion like that seen in the eastern United States, but the established population in Los Angeles CA, now suggests that A. albopictus will be an immediate and persistent threat in the southwestern United States not previously experienced before.
The results of this study using an experimental laboratory approach that closely mimicked field conditions showed that A. albopictus was clearly the dominant competitor over A. sierrensis. Aedes albopictus in the experiment were F1 generation individuals of field collected larvae from multiple populations within urban and suburban landscapes in the eastern United States. There is a possibility that A. albopictus from the eastern United States may not exactly represent the competitive abilities of A. albopictus in LA County. Effects of A. albopictus origin on competitive ability are unclear. While Leisnham et al. [24] showed inherent interpopulation variation in competitive ability of A. albopictus, and other studies have shown interpopulation variation among other A. albopictus traits [56], only egg diapause has shown a clear trends between latitudes [57,58]. Therefore it is difficult to conclude that there are be systematic differences from eastern vs. western A. albopictus. Moreover, Leisnham et al [24] showed that populations within FL had different competitive abilities. Should A. albopictus spread throughout the southwestern United States, interpopulation variation among even geographically close populations of A. albopictus may evolve. However, by testing interspecific competition using A. sierrensis and A. albopictus strains that likely experience strong competition in the field using a response surface design, this study has provided a rigorous examination larval competition, and has likely represented the likely outcome of competition between these species in the field, especially given the overwhelming dominance of A. albopictus that was demonstrated. Nevertheless, additional experiments are needed to better understand factors that have prevented their successful invasion of A. albopictus in the past and those factors that have facilitated their establishment now. The results from this study here indicate that it is unlikely that A. sierrensis will present a substantive barrier to the colonization of A. albopictus in tree holes habitats in particular and the further spread of A. albopictus in general.