Copper pollution enhances the competitive advantage of invasive ascidians

Invasive species dominate and often outcompete native species in marine harbors. The abundance of introductions due to shipping activity combined with artificial substrates and poor water quality helps to facilitate invasion success in these areas. Colonial ascidians, in particular, have broad environmental tolerances that allow them to invade novel ecosystems including urban harbors. While the effects of invasive species and poor water quality on native species have been explored, the relative influence and how these pressures may interact to degrade biodiversity is not well characterized. The purpose of this study was therefore to understand the role of interspecific competition and pollution on the native Aplidium glabrum and the invasive Botrylloides violaceus. Colony surface area has been previously identified as a sensitive toxicity endpoint and was measured weekly to assess impacts both separately and together at 0, 1, 10, 20, and 40 μg/L copper concentrations. A. glabrum was significantly impacted at 1 μg/L in the presence of interspecific competition and at 10 μg/L with no competition; however, the copper EC50 revealed increased sensitivity in the presence of B. violaceus (11.3 μg/L versus 6.9 μg/L while in competition). Conversely, significant impacts to B. violaceus growth did not occur until much higher copper concentrations with an EC50 of 37.1 μg/L while alone and 31.7 μg/L when in competition. Interspecific competition was found to significantly impact A. glabrum surface area growth at 1, 10, and 40 μg/L, while there was no significant effect on B. violaceus at any of the concentrations tested. Finally, an interaction effect was found between copper pollution and competition status only for A. glabrum. The results of this study support the hypothesis that invasive species are more tolerant of pollution while also revealing the interactive effects of pollution exposure and interspecific competition.


Introduction
Invasive species pose significant threats to benthic marine communities in urban environments via competition, predation, and disease transmission (Wilcove et al. 1998;Parker et al. 1999;Lodge et al. 2006). Their presence in marine ecosystems is increasing with no sign of saturation (Seebens et al. 2017). Invasive species threaten vulnerable native ecosystem structure and function (Maxwell et al. 2016;Walsh et al. 2016;Tilman et al. 2017) The presence of invaders can lead to declining reproduction rates and survival, ultimately impacting population growth (Lockwood et al. 2007). Further, US economic losses associated with terrestrial, freshwater, and marine invasive species management are estimated at approximately $120 billion annually (Pimentel et al. 2005).
Frameworks have been proposed to identify critical steps in the invasion process for species to become successful in a novel environment. Blackburn et al. (2011) developed a unified framework from existing models to establish universal language describing the invasion process while including new information about invasion barriers and management strategies, all in the context of invasion stage. This model was then later applied to ascidians to identify the invasion barriers and factors which enable their success (Zhan et al. 2015). Anthropogenic impacts such as pollution and invasion have the potential to interact (Piola and Johnston 2006) and these disturbances can possibly result in decreased biodiversity (Ruiz et al. 1997). The interaction between anthropogenic stressors and the invaders themselves could significantly influence their colonization, distribution, and overall impacts in novel environments (Ruiz et al. 1999).
The Invader Tolerance Model (ITM) developed by Osborne et al. (2018) provides a framework to explain the reduction of biodiversity by invasive species and marine pollution. This framework proposes that invaders are more tolerant to pollution relative to native species, and that native species diversity will decrease following an exposure event. This decrease will further enable invasives to settle, thus decreasing overall native species diversity even more. However, within the ITM there are unanswered questions regarding the relative importance of pollution and invasive species on native species as well the combined effects.
Invasive ascidians have been identified in several anthropogenicallydisturbed environments along the Northwest Atlantic coast through rapid assessment surveys (e.g. McIntyre et al. 2013;Wells et al. 2014). The colonial ascidian species Botrylloides violaceus (Oka, 1927) has been found in the Gulf of Maine (GOM) since the 1980s (Dijkstra et al. 2007). B. violaceus was likely first introduced in the Damariscotta River in Maine via oyster aquaculture imports and has since spread rapidly both north and south along the coastline, demonstrating strong competitive abilities in the GOM (Dijkstra et al. 2007;Dijkstra and Harris 2009). Conversely, Aplidium glabrum (formerly "Amaroucium glabrum") (Verrill, 1871), is native to the North Atlantic Ocean (Van Name 1945) and is distributed along the North American coasts of the Bay of Fundy to Cape Hatteras (Van Guelpen et al. 2005). The distribution of A. glabrum is likely temperature-dependent, as this species tends to reside in cooler, often deeper waters (Van Name 1945). Both species are available for collection in New England (Wells et al. 2014), the larvae are easy to retrieve in the laboratory (pers. obs.), and rapidly reproduce. For these reasons, B. violaceus and A. glabrum serve as excellent model species to investigate the interacting impacts of competition and pollution exposure and can concurrently provide a native-invasive tolerance comparison.
Copper pollution is one of the primary active ingredients in many antifouling hull paints since the widespread ban of tributyltin (TBT) in the 1980s. It is commonly found in urban environments at concentrations of approximately 1 μg/L (Hall Jr. et al. 1998;Lawes et al. 2016) but can reach upwards of 80 μg/L in some areas including heavily trafficked locations of Chesapeake Bay (Hall Jr. et al. 1998). Chronic copper toxicity of the invasive B. violaceus and the native A. glabrum was previously assessed (Osborne et al. 2018) to identify sublethal concentrations impacting growth and morphology. The results of the study indicated that B. violaceus is more tolerant of copper pollution than A. glabrum in laboratory conditions. Both the number of zooids and colony surface area were significantly higher in the invasive species relative to the native at the conclusion of the 11-week experiment. Further, heart rates indicated greater physiological stress to the natives at lower concentrations relative to invasive species.
Biotic resistance, or the hypothesis that increased native community biodiversity can stave off invader establishment, is attributed to high species diversity and thus increased competitor presence (Elton 1958). Therefore, to understand how these sensitivity differences between invasive and native ascidians will play out within the biofouling community, interspecific competition must be directly explored. The purpose of this study was to investigate if and how competition and pollution exposure factors interact, focusing on the model ascidian species B. violaceus and A. glabrum. This approach tests important predictions of the ITM that the dual impacts of pollution exposure and competition will have a negative synergistic effect on the ability of A. glabrum to persist. Survival, growth, and reproduction of the invasive and native species were assessed both separately and together, and while exposed or not to elucidate if an interaction effect exists between these two factors.

Experimental Design
We conducted a 10-week exposure to 0, 1, 10, 20, and 40 μg/L Cu to test the combined and separate effects of copper pollution and interspecific competition on the native A. glabrum and the invasive B. violaceus. We found surface area to be a sensitive endpoint for negative impacts on colonial ascidians (Osborne et al. 2018), and therefore it was used as a primary endpoint in the present study. Table 1 describes the experimental treatment design used to test factors separately and together. Five replicates per treatment condition were included. Although it would have been worthwhile to also include a treatment of two larvae from the same species Table 1. Organization of experimental conditions to show how factors were tested either separately or together. Treatments are named by the following scheme: "A" = A. glabrum (native), "B" = B. violaceus (invasive), "CTRL" = control, "CU" = copper treatment; either 1, 10, 20, or 40 µg/L concentration.

Condition
Native to test for possible intraspecific competition, this would have almost doubled the number of treatments and controls and was not possible due to space and resource constraints. However, because space was not a limiting factor in these studies (percent cover of the disks never exceeded 80%) we believe that intraspecific competition would not be a major factor influencing colony growth.
To account for the more rapid growth and colony area of B. violaceus compared with A. glabrum and allow for direct comparison of interactive effects due to competition and copper exposure, the percent growth was calculated by dividing the colony growth in all treatments with their controls. Particular interest was given to the impacts on colony growth at concentrations within the range of previously identified EC 50 values for each species, 10 μg/L for A. glabrum and 40 μg/L for B. violaceus. P-values indicate differences in actual growth between competition and no competition treatments at these concentrations.

Larvae Collection and Initial Settlement
We collected larvae using a modified version of the Watanabe and Lambert (1973) larvae collector design (as described in Osborne et al. 2018). Briefly, adult colonies were retrieved from the field, immediately added to the larvae collector tanks, and allowed to passively release their larvae. Colonies were collected from the floating dock at Massachusetts Maritime Academy (MMA) in Buzzards Bay, Massachusetts (41°44′19.58″N; 70°37′36.44″W) and transported to the MMA aquaculture laboratory. Colonies were held in larvae collector tanks with a continuous low-flow of 100 μm filtered seawater until larvae release was observed; no additional food was added. Larvae release occurred within 1-2 hours of collection. All adult colonies were obtained simultaneously and therefore larvae were released at approximately the same time. Larvae were directed to a submerged Nitex mesh filter by a current generated at the gravity-fed carboy water input, where they were then collected. Only larvae which were free-swimming were used for settlement and unmoving larvae were discarded. We gently pipetted larvae by hand to ensure no damage was imposed once collected in the Nitex filter.
Larvae were settled on 20 mm diameter white plastic disks according to the methods in Osborne et al. (2018). Since treatments were designed to determine both individual effects and interactions of pollution exposure and interspecific competition, disks contained either 2 larvae (1 larvae of each species) settled 1 mm apart, or 1 larvae of a single species. Once larvae metamorphosed, disks were transferred to compartmentalized plastic containers and grown for five days until the exposure period began. Prior to exposure, all juveniles were visually inspected with a dissecting microscope to ensure a healthy appearance (strong heartbeat, plump and organized ampullae, siphons open and feeding, and a well-formed zooid) and probed to monitor stimuli reaction. Juveniles were fed 20 μL of Shellfish Diet 1800® (Reed Mariculture Inc., CA, USA) every alternating day for 5 days prior to exposures.

Experimental Units
Each disk with settled juveniles was suspended inside a 1 L plastic container with 750 ml of water. Copper concentrations of 0, 1, 10, 20, and 40 μg/L were used in this study and have been previously identified as sub-lethal concentrations for both species in chronic toxicity assays (Osborne et al. 2018). Stock solutions were prepared in deionized water using cupric chloride dihydrate (CuCl 2 *2H 2 O; MP Biomedicals, Santa Ana, CA). Experimental concentrations were prepared by adding the appropriate stock to 5μm filtered seawater at 32-34 psu salinity and ambient laboratory temperature (18-21 °C). We used a static renewal system for these exposures, changing water 3 times per week. Each container was fed 40 μL of Shellfish Diet 1800® (Reed Mariculture Inc., CA, USA) following water changes.

Toxicity Endpoints
Surface area was recorded once weekly from photographs taken with a USB microscope (USB-2-Micro-200X, Plugable Technologies) on both front and back surfaces of each disk. Individual zooids were counted during this time. Large-scale morphological observations were recorded on all water change days. Surface area was analyzed on photographs using ImageJ photo processing software (Schneider et al. 2012) and was calculated for all treatments. The number of zooids were counted using a dissecting microscope (EZ4, Leica) under low power (4x magnification) for each colony. Measurements for each pair of competition treatments at each pollution concentration were then compared with one another to determine at which copper concentrations competition significantly impacted surface area for each species separately. For example, the competition and no competition surface areas of A. glabrum were compared at 1 μg/L and were separately analyzed from those of B. violaceus.

Statistical Analyses
Statistical analyses were performed using R version 3.3.0. For all statistical tests, p-values < 0.05 were considered statistically significant and indicates those treatments where competition, pollution, or an interaction between the two factors had a significant impact on surface area growth.
The effect of copper between treatment concentrations was determined using ANOVA (normally distributed, homogeneous data) followed by Tukey's HSD or a Kruskal-Wallis (non-normally distributed, homogeneous data) followed by Dunn's test. Separate t-tests were used to determine differences between competition/no competition treatments at each concentration.
A factorial ANOVA was used to determine if an interaction effect exists between competition status and copper treatment. Competition status (competition, no competition) and copper pollution (0, 1, 10, 20, and 40 μg/L) were the independent variables and surface area was the dependent variable. Both variables will affect the outcome and the effect of changes in one variable will depend on the level of the other variable if an interaction exists; if an interaction does not exist, the effects of changes in one variable are the same at every fixed level of the other variable (Seltman 2015).
The Lowest Observable Effect Concentration (LOEC; lowest copper concentration which elicits a significant effect on surface area) was calculated using ANOVA. EC 50 concentrations (the Effective Concentration which negatively impacts 50% of the animals tested) were derived from dose-response curves constructed using a log-logistic regression model (Hasenbein et al. 2017).

Species Differences in Growth Rate
Surface area was assessed as a measure for colony growth throughout the 10-week exposure period (see Figure 1). In the absence of any copper treatment or competition, B. violaceus showed an accelerated growth rate compared with A. glabrum. Surface area growth generally followed a linear trend for both species, although final control surface area measurements for B. violaceus were an order of magnitude higher than those of A. glabrum. As shown in Figure 1, this trend began as early as four weeks postsettlement when the average colony size of B. violaceus was over 100 mm 2 , while A. glabrum average colony size was approximately 10 mm 2 . Zooid counts followed a similar trend, as observed in previous studies (Osborne et al. 2018). These results indicate that in the absence of any stressors, the invasive species has a significant growth advantage over the native species.

Pollution and Competition Effects
In all treatment types, copper exposure had negative impacts on colony growth of both native and invasive ascidians. Copper treatment significantly impacted Aplidium glabrum colony growth (p = < 0.00001) and post-hoc analysis indicated significant effects at 10 μg/L (p = 0.0001) and all other higher concentrations (p = < 0.00001). Botrylloides violaceus was also impacted by increasing copper concentrations, but only the 40 μg/L exposure resulted in a significant effect on colony growth (p = 0.0007). As shown in Figure 2, there was a dramatic decrease in A. glabrum growth between the 1 and 10 μg/L exposures, while B. violaceus did not appear to be similarly affected until 40 μg/L. Dose-response curves were generated to determine the EC 50 values (Table 2) for colony growth, or the concentration  resulting in a 50% reduction in surface area. Again, the native species shows increased sensitivity with an EC 50 value of less than a third of the value for B. violaceus.
In the presence of low concentrations of copper (1 μg/L), competition played a role in further suppressing native (p = 0.009), but not invasive species growth (Figure 2). Competition also had a significant effect on A. glabrum in the 10 and 40 μg/L copper exposures (Figure 2). The surface area of A. glabrum was 50% less compared to the control without competition at 10 μg/L, and nearly 65% less while in competition (p = 0.001). At 40 μg/L, surface area was 85% less without competition and 95% less in the presence of competition (p = 0.05). These findings suggest that as copper concentration increased, the effect of competition became stronger on A. glabrum. The presence of competition affected the growth of A. glabrum at all concentrations, further decreasing overall growth beyond the impacts of copper alone. For example, in the 10 μg/L exposure with the presence of competition, colony growth of A. glabrum is lower than the growth at 20 μg/L in the absence of competition (Table 3, Figure 1). Dose-response curves revealed that the EC 50 values were lower in the presence of competition. In addition, EC 50 values were nearly five times greater for B. violaceus than A. glabrum in treatments with competition, indicating a heightened sensitivity of the native species relative to the invader, especially when in competition with B. violaceus for resources.
A factorial ANOVA was used to compare the main effects of competition status and copper exposure (as reported in the sections above) and determine if there was an interaction effect on ascidian growth. The results reported in Table 4 indicated that a competition status-copper exposure interaction was significant for A. glabrum (p = 0.03) but not for B. violaceus at environmentally relevant concentrations.

Discussion
Understanding the mechanisms responsible for the success of marine invasive species is critical for their management. Under this context, the influences of copper pollution and interspecific competition were compared on a prolific marine invader, Botrylloides violaceus and a native species, Aplidium glabrum. Both species are abundant along the Northwest Atlantic coast where they compete for resources within the marine biofouling community. We argue that colony surface growth is a strong predictor of invasion success because space is a primary resource and under constant competition from the benthic community (e.g. Stachowicz et al. 2002). The major findings of this study indicate that 1) competition does not play a significant role in limiting A. glabrum surface area growth until copper is introduced and 2) B. violaceus is not significantly impacted by competition at any copper concentration tested. Additionally, similar EC 50 values to those previously established (Osborne et al. 2018) confirm relatively higher copper tolerance in invaders.
B. violaceus continues to be globally transported and field studies have confirmed the dominance of B. violaceus in the overlapping range of A. glabrum (Dijkstra and Harris 2009;Sorte and Stachowicz 2011;McIntyre et al. 2013;Wells et al. 2014;Dijkstra et al. 2017). Here, A. glabrum demonstrated a much slower proliferation overall than B. violaceus. Since this disparity in growth rate has been observed in the field (pers. obs.), this study provides a potential mechanism for this occurrence.
It was previously established that B. violaceus is more tolerant to copper pollution compared with native species, A. glabrum. The current study confirmed these findings where significant impairment of growth was observed in A. glabrum at 10 μg/L, while B. violaceus showed no sign of reduced growth until concentrations nearly four times higher. The physiological "tipping point" is therefore measurably lower in the native versus the invasive species. Copper toxicity to colonial ascidians has been previously investigated on a large-scale using community experiments Johnston 2008, 2009;Crooks et al. 2011) and also specific species (Bellas et al. 2004;Osborne et al. 2018). In general, native species displayed lower tolerance to copper. In the community experiments, the lower native tolerance resulted in decreased native biodiversity, which in turn, further enabled the growth and spread of invasive species. To conclude, Piola and Johnston (2008) emphasized the need to consider pollution effects not on diversity as a whole, but to distinguish native and non-native species as they were found to respond to exposure differently.
In addition to copper, the increased tolerance of invasive species to other environmental stressors may facilitate their spread. Studies have shown that B. violaceus can tolerate broad temperature fluctuations ranging 5-25 °C (McCarthy et al. 2007;Sorte et al. 2010) and wide salinity changes of 10-38 psu (Dijkstra et al. 2008;Epelbaum et al. 2009a). These observations are in direct contrast to native species sensitivity to environmental factors. In a temperature tolerance study by Sorte et al. (2010), the recovery period following a simulated heat wave was significantly longer for native than invasive ascidians, with natives demonstrating an overall lower recovery in general. Further, juvenile invasive ascidians were not significantly impacted by the heatwave, yet the abundance of the native colonial species decreased by 40% (Sorte et al. 2010). Lagos et al. (2017) investigated the relative tolerance of several native ascidians (i.e. Pyura dalbyi Rius and Teske, 2011, Pyura doppelgangera Rius and Teske, 2013, Herdmania grandis (Heller, 1878, and Botrylloides magnicoecum (Hartmeyer, 1912)) and invasive ascidians (i.e. Steyla plicata (Lesueur, 1823), Styela clava Herdman, 1881, Didemnum spp., and Diplosoma spp.) to hypoxia. Invasive species were found to be nearly twice as tolerant as natives to low oxygen conditions. Further, it was determined that manmade structures, including those found within marinas, lead to reductions in local oxygen levels, which likely impacts native species survival and growth (Lagos et al. 2017). The increased tolerance of invasive species to low oxygen, may help explain why invasive species occur in great abundance on artificial structures in urban harbors and marinas (e.g. Simkanin et al. 2012;Cordell et al. 2013). Taken together, these studies provide further evidence that environmental stressors can impact species biodiversity, by inhibiting the growth of native species, reducing biotic resistance, and providing habitat for invasive species to settle. However, the relative importance of competition versus environmental stress and whether these stressors interact to further inhibit native species growth has not been directly explored. Therefore, in the current study we went beyond measuring the difference in tolerances between A. glabrum and B. violaceus, to investigate if competition impacts ascidian growth both in the presence and absence of copper pollution. A. glabrum was not inhibited by invasive competitive pressure alone, but the combined effects of competition and pollution, even at low levels, were extremely detrimental to native growth. Concentrations of 1 μg/L caused a difference in surface area growth of A. glabrum in the presence of B. violaceus despite no observable impacts of competition in controls. Further, 1 μg/L is an environmentally concentration in urban marine settings (Stauber et al. 2000;Schiff et al. 2007), supporting the possibility that copper is negatively influencing native species. Previous studies have found that copper concentrations in urban environments can range from 0-80 μg/L (Manisseri and Menon 1995;Hall Jr. et al. 1998;Negri and Heyward 2001;Katranitsas et al. 2003;Bellas et al. 2004). According to these estimates, areas with low levels of copper pollution are at risk for invasion once their ability to compete is compromised. Although copper inhibited A. glabrum at relatively lower concentrations than B. violaceus, the crucial finding is that the pressure of competition overwhelmed the native species only once copper was introduced.
Previous studies have demonstrated that invasive species success significantly increases with the availability of free space (Stachowicz et al. 2002). Although we did not observe total resource exploitation during the 10-week duration of this experiment, percent cover has also been utilized as a proxy of competitive pressure in other studies (Stachowicz et al. 1999;Paetzold et al. 2012). The release of inhibitory chemical compounds by invaders (allelopathy) has been proposed as a potential important nonresource interaction with native counterparts (Hierro and Callaway 2003). Paetzold et al. (2012) determined that B. violaceus only inhibited its competitor Ciona intestinalis (Linnaeus, 1767) on pre-settled plates and not on bare plates which were available for settlement to both concurrently. The authors therefore suggest that allelopathic impacts may be derived from established B. violaceus colonies in pre-settled treatments only. Although possible, it is unlikely that competition for food played a role because effects of food type and concentration have been previously explored (Epelbaum et al. 2009b). Further, concentrations were doubled from recently published values (Osborne et al. 2018) to ensure abundant food availability.
It has been proposed that rapid environmental disturbances may limit native species with prior advantageous adaptations while in competition with non-native species (Byers 2002). Our findings provide key evidence for the Invader Tolerance Model (ITM) that native species are relatively more impacted than invaders by the dual pressures of copper exposure and competition, as described in Osborne et al. (2018). The ITM provides a framework to test the hypothesis that invasive species are more tolerant of marine pollution than natives, and high native species diversity is a crucial component of the community to resist invasion. Our study has shown that in the absence of an additional stressor (i.e. copper), A. glabrum growth is not hindered by the presence of invasive species. This implies that under pristine conditions, A. glabrum and likely other native ascidians, have the capacity to defend against invasive species through biotic resistance. Further studies with additional native ascidians exposed to other environmental stressors are necessary to confirm the generality of this finding. In addition, future studies should account for propagule pressure of invasive species, which was not possible with our current experimental set-up. However, this is the first known attempt to begin to untangle the effects of pollution, competition, and native/invasive status with a widespread marine invader. This important interaction effect sheds new light on community biotic resistance; tolerances to external pressures including marine pollution may play a much more significant role than originally proposed in forming a biotic defense against invaders.
The management of invasive colonial ascidians poses many challenges, including their ability to be transported on ship hulls, tolerate broad physical conditions, and utilize multiple reproductive strategies to proliferate rapidly (Carlton 1989;Carlton and Geller 1993;Collin et al. 2013). Biological (Carman et al. 2009;Switzer et al. 2011), chemical (Switzer et al. 2011, Carman et al. 2016, and physical (Switzer et al. 2011) removal treatments have been employed, but there has been little success in large-scale field applications. Therefore, management practices should continue to strive for avoiding initial introduction by identifying areas at risk, or hot spots of invasion. As seen here, invasive ascidians are stronger competitors for resources than native ascidians and copper pollution compounds the stress impacting native ascidians. Understanding that native species are threatened by both stressors, both independently and together, highlights the importance of multipronged approaches to invasive species management that includes reducing marine pollution.
We found that low and environmentally relevant concentrations of copper significantly impaired the native ascidians species in this study. This finding offers support for using copper regulation as a new management tool for combating biological invasions. Currently, the U.S. EPA marine water quality criteria (WQC) values are 2.0 μg/L for acute and 1.3 μg/L for chronic copper protection (U.S. EPA 2016). Considering the concentration yielding an interactive effect with competition in this study and measured field copper concentrations together highlights the importance of tightening pollution control to be a more effective prophylactic regulatory strategy. The U.S. National Invasive Species Council's (2016) current invasive species management strategies include introduction prevention, eradication, containment, and ecosystem restoration. This approach focuses on the invasive species themselves and does not consider the management of other potentially additive or synergistic stressors. A trans-disciplinary approach combining invasive species management with stricter copper and perhaps other heavy metal pollution regulations will produce more effective defense strategies for the communities both separately seek to protect.