Coral Disease and Ingestion: Investigating the Role of Heterotrophy in the Transmission of Pathogenic Vibrio spp. using a Sea Anemone (Exaiptasia pallida) Model System

ABSTRACT Understanding disease transmission in corals can be complicated given the intricacy of the holobiont and difficulties associated with ex situ coral cultivation. As a result, most of the established transmission pathways for coral disease are associated with perturbance (i.e., damage) rather than evasion of immune defenses. Here, we investigate ingestion as a potential pathway for the transmission of coral pathogens that evades the mucus membrane. Using sea anemones (Exaiptasia pallida) and brine shrimp (Artemia sp.) to model coral feeding, we tracked the acquisition of the putative pathogens, Vibrio alginolyticus, V. harveyi, and V. mediterranei using GFP-tagged strains. Vibrio sp. were provided to anemones using 3 experimental exposures (i) direct water exposure alone, (ii) water exposure in the presence of a food source (non-spiked Artemia), and (iii) through a “spiked” food source (Vibrio-colonized Artemia) created by exposing Artemia cultures to GFP-Vibrio via the ambient water overnight. Following a 3 h feeding/exposure duration, the level of acquired GFP-Vibrio was quantified from anemone tissue homogenate. Ingestion of spiked Artemia resulted in a significantly greater burden of GFP-Vibrio equating to an 830-fold, 3,108-fold, and 435-fold increase in CFU mL−1 when compared to water exposed trials and a 207-fold, 62-fold, and 27-fold increase in CFU mL−1 compared to water exposed with food trials for V. alginolyticus, V. harveyi, and V. mediterranei, respectively. These data suggest that ingestion can facilitate delivery of an elevated dose of pathogenic bacteria in cnidarians and may describe an important portal of entry for pathogens in the absence of perturbing conditions. IMPORTANCE The front line of pathogen defense in corals is the mucus membrane. This membrane coats the surface body wall creating a semi-impermeable layer that inhibits pathogen entry from the ambient water both physically and biologically through mutualistic antagonism from resident mucus microbes. To date, much of the coral disease transmission research has been focused on mechanisms associated with perturbance of this membrane such as direct contact, vector lesions (predation/biting), and waterborne exposure through preexisting lesions. The present research describes a potential transmission pathway that evades the defenses provided by this membrane allowing unencumbered entry of bacteria as in association with food. This pathway may explain an important portal of entry for emergence of idiopathic infections in otherwise healthy corals and can be used to improve management practices for coral conservation.

a model system was employed utilizing sea anemones (Exaiptasia pallida) and brine shrimp (Artemia sp.) to mimic natural coral feeding. Prior research has demonstrated the utility of sea anemones in the genus Exaiptasia (formally Aiptasia, see Grajales & Rodriguez, 2014 [41] for reclassification) as lab-friendly surrogates for coral experimentation (41)(42)(43)(44). Structurally, Exaiptasia spp. resemble large non-colonial coral polyps and feed both heterotrophically on zooplankton and autotrophically though the use of their algal symbionts (zooxanthellae) (45). Using this model system, we traced the acquisition of the putative coral pathogens V. alginolyticus, V. mediterranei, and V. harveyi (Table 2).

RESULTS
Artemia Colonization by Vibrio. Colonization experiments first assessed the ability of Vibrio spp. to attach to/associate with Artemia. Substantial colonization of Artemia gastrointestinal (GI) tracts was observed for all tested vibrios following overnight (18 h) exposure via ambient water at 28°C. Total colonization for each Vibrio spp. exposure (;250 Artemia) was 4.9 Â 10 6 , 1.5 Â 10 6 , and 7.6 Â 10 6 CFU per ; 250 individuals for V. alginolyticus, V. harveyi, and V. mediterranei, respectively. These levels equate to a mean acquisition of 4.3%, 2.1%, and 50.2% of the initial exposure dose for V. alginolyticus, V. harveyi, and V. mediterranei, respectively (Fig. S1). It should be noted that these counts are based on culturable GFP-Vibrio spp. and thus may be confounded by the presence of Vibrio spp. in a viable but non-culturable state (VBNC) (46). However, warm water conditions and short duration exposures minimized the likelihood of VBNC formation, which is typically associated with longer duration exposure to stressful conditions (46,47). Epifluorescence microscopy showed GFP-tagged cells were concentrated throughout the length of Artemia GI tracts in association with ingested material and  feces ( Fig. 1). GFP cells were also observed in association with Artemia feces following defecation. Low exoskeletal association was observed in all experimental trials, though minor attachment and/or entanglement was noted in association with Artemia appendages ( Fig. 1). GI association was consistent across naupliiar sizes excluding the smallest, most recently hatched individuals (Fig. S2), which showed little to no GFP-Vibrio accumulation. Visual patterns of GI association did not differ between Vibrio species. No distinctive behavioral changes or swimming impairment was observed in colonized Artemia throughout the duration of exposure (up to 24 h). Uptake of Vibrio by E. pallida. Anemone feeding studies evaluated the efficacy of an ingestion-based transmission pathway by confirming consumption of GFP-Vibriocolonized Artemia and quantification of the acquired GFP-Vibrio dose. Gross observations of feeding demonstrate that E. pallida readily ingested Vibrio-colonized Artemia, responding rapidly with predatory tentacle behavior when Artemia were introduced into the microcosm water (Fig. S3). No differences in anemone feeding behavior (i.e., tentacle response) were observed for exposures using spiked and non-spiked Artemia.
Assessment of the acquired dose compared 4 major feeding/exposure treatments: (i) spiked fed, where no GFP-Vibrio were inoculated into the microcosm water and anemones were fed with Vibrio-colonized Artemia, (ii) water exposed control fed, where GFP-Vibrio were inoculated into the microcosm water and anemones were fed with non-spiked Artemia, (iii) water exposed not fed, where GFP-Vibrio were inoculated into the microcosm water and no Artemia were added, and (iv) control, where no GFP-Vibrio were inoculated into the microcosm water and anemones were fed non-spiked Artemia (Fig. 2). Significantly greater GFP-Vibrio levels were observed in E. pallida individuals exposed via spiked Artemia (spiked fed) compared to individuals exposed through the ambient water, regardless of the presence of Artemia (i.e., all other experimental conditions). Anemone homogenate from spiked fed trials showed a mean GFP- Vibrio concentration of 6.9 Â 10 4 , 2.6 Â 10 5 , and 1.7 Â 10 5 CFU mL 21 for V. alginolyticus, V. harveyi, and V. mediterranei, respectively. Conversely, water exposed anemones showed a mean concentration of 3.3 Â 10 2 and 8.3 Â 10 1 CFU mL 21 for V. alginolyticus, 4.1 Â 10 3 and 8.3 Â 10 1 CFU mL 21 for V. harveyi, and 5.9 Â 10 3 and 3.8 Â 10 2 CFU mL 21 for V. mediterranei for water exposed control fed (non-spiked) and water exposed not fed (no Artemia) treatments, respectively (Table 3). These concentrations equate to a 207-fold, (P = 0.03), 62-fold (P = 0.013), and 27-fold (P = 0.013) increase in the GFP-Vibrio burden of spiked fed compared to water exposed control fed anemones and a 830-fold (P = 0.028), 3,108-fold (P = 0.026), and 435-fold (P = 0.030) increase in spiked fed compared to water exposed not fed anemones for V. alginolyticus, V. harveyi, and V. mediterranei, respectively (Fig. 3). Between the 2 water exposures, fed (non-spiked Artemia) anemones showed a significantly greater burden of GFP V. harveyi (P = 0.026) and V. mediterranei (P = 0.030) compared to non-fed anemones but did not differ significantly for V. alginolyticus (P = 0.51). No GFP-Vibrio were recovered from anemones in the control group (no exposure) or from anemone wash water (carry-over control).

DISCUSSION
The mucus membrane serves as the front line of defense against infection for coral species. This mucus coats the epithelia creating a semi-impermeable barrier between the coral tissue and ambient water (12,14,48,49). Within this mucus layer, a variety of mutualistic and commensal microorganisms are maintained. The totality of these microbes and the coral colony are collectively known as the holobiont (50). Research has suggested that the coral-associated microbial community can confer improved fitness to the holobiont through community shifts in response to environmental change (14,51), the production of antimicrobial compounds (52) and/or antagonistic competition with potential pathogens (50,53,54). Together, the physical mucus barrier combined with the biological protection of the microbial community poses a substantial challenge to the direct transmission of waterborne pathogens. To date, the majority of coral disease transmission research has focused on mechanisms of pathogen spread associated with perturbance of this mucus membrane, such as direct contact, vectormediated (i.e., biting), and indirect transmission via preexisting lesions (11). While these  Table 3. studies provide important insight into the ecology of coral diseases, these transmission mechanisms are dependent on opportunistic occurrences related to host proximity and preexisting or active damage and there is substantial need to investigate transmission mechanisms related to disease emergence in uninjured corals.
Despite the presence of zooxanthellae, heterotrophic feeding is a critical component of coral nutrition, accounting for up to 35% of the daily metabolic needs of some coral species (17,18). Corals preferentially feed on small zooplankton thus, we investi- FIG 3 Recovered GFP-Vibrio spp. concentrations from anemone homogenate following completion of the controlled feeding study. Spiked fed anemones demonstrated a significantly greater GFP-Vibrio spp. burden compared to water exposed individuals. Spiked fed versus water exposed and fed resulted in P-values of 0.03, 0.013, and 0.013 and spiked fed versus water exposed not fed resulted in P-values of 0.028, 0.026, and 0.03 for V. alginolyticus, V. harveyi, and V. mediterranei, respectively. N = 6 anemones for each exposure type and Vibrio spp. 2.0 Â 10 7 6 5.6 Â 10 5 NA 6.9 Â 10 4 6 7.5 Â 10 3 V. alginolyticus Water exposed control fed NA NA 4.5 Â 10 8 3.3 Â 10 2 6 1.0 Â 10 2 V. alginolyticus Water exposed not fed Spiked fed 2.8 Â 10 8 5.9 Â 10 6 6 2.6 Â 10 5 NA 2.6 Â 10 5 6 1.1 Â 10 5 V. harveyi Water exposed control fed NA NA 2.6 Â 10 8 4.1 Â 10 3 6 1.6 Â 10 3 V. harveyi Water exposed not fed 6.1 Â 10 7 3.0 Â 10 7 6 7.1 Â 10 5 NA 1.7 Â 10 5 6 5.9 Â 10 4 V. mediterranei Water exposed control fed NA NA 6.1 Â 10 7 5.9 Â 10 3 6 1.4 Â 10 3 V. mediterranei Water exposed not fed a For each experimental trial ;1000 Artemia (individuals) and 6 anemones (individuals) were exposed. b Total GFP-Vibrio exposure represents the CFU concentration introduced to Artemia to promote colonization this was calculated during Artemia dose assessment as the sum of 4 exposure trials (;250 Artemia each). Initial exposures were administered at 1.1 Â 10 8 , 6.9 Â 10 7 , and 1.5 Â 10 7 CFU/; 250 Artemia for V. alginolyticus, V. harveyi, and V. mediterranei, respectively. Water exposure trials were inoculated directly into the microcosm using the same concentration. c Total GFP-Vibrio carried by Artemia represents the CFU concentration acquired by Artemia following exposure. This was calculated during Artemia dose assessment as the sum of 4 exposure trials (;250 Artemia each). Exposures resulted in an Artemia-acquired dose of 4.9 Â 10 6 , 1.5 Â 10 6 , and 7.6 Â 10 6 CFU/; 250 Artemia for V. alginolyticus, V. harveyi, and V. mediterranei, respectively. gated the ability of pathogenic Vibrio spp. To be transmitted to a cnidarian host via ingestion following colonization of a zooplankton vector. Using sea anemones (E. pallida) and brine shrimp (Artemia spp.) to model coral feeding, we demonstrate that ingestion of Vibrio-spiked brine shrimp results in a significantly higher bacterial burden in recipient anemones compared to ambient water exposures, both with and without food sources (i.e., Artemia). These data suggest that ingestion could play a role in the transmission of certain coral pathogens. Furthermore, this mode of transmission bypasses the natural defense mechanisms of corals provided by their mucus membrane (50,52), which may describe an important portal of entry related to pathogenic infection of uninjured corals. Acting as our model zooplankton, Artemia were readily colonized by all tested Vibrio spp. following direct waterborne exposure, similar to previous studies in V. cholerae (34). However, the preferential colonization of the GI tract noted here differed from previously described observations where colonization was predominately observed on zooplankton exoskeletons (34)(35)(36). We hypothesize that this difference may be due to the fact that the present research was conducted ex situ where certain environmental determinants of zooplankton colonization (i.e., substrate limitation) may not be present and/or as impactful (21,31,55). While some minor exoskeletal association was observed on Artemia appendages, we suspect that this may be the result of incidental entanglement rather than purposeful attachment. Due to the lack of strong external association, we postulate that the colonization of Artemia GI tracts is the result of active ingestion of Vibrio spp. by nauplii occurring over prolonged interaction ($4 h of exposure). This hypothesis is further supported by the observation that the smallest most recently hatched Artemia (Fig. S2) showed minimal GI colonization. At this stage of life, nauplii are nutritionally maintained through residual yolk protein and do not actively feed until they are larger (56,57). The total Artemia-acquired dose remained relatively consistent for all three Vibrio spp. at ; 10 6 CFU per ; 250 individuals. These data suggest that Artemia have a threshold for the maximum concentration of Vibrio spp. they can harbor via GI colonization.
Feeding experiments demonstrate that spiked fed anemones acquire a significantly greater GFP-Vibrio burden compared to water exposed individuals regardless of the presence of food. This pattern was observed across all 3 Vibrio spp., suggesting that ingestion of Vibrio-colonized zooplankton can facilitate delivery of an elevated dose of these bacteria, broadly. The higher Vibrio levels are likely the result of bioaccumulation of these bacteria within Artemia facilitating acquisition of a highly concentrated dose through targeted feeding. This is consistent with prior observations of V. cholerae carriage by copepods where ingestion of a small number of individuals may facilitate receipt of a potentially pathogenic dose (#10 3 cells) (34,36). While low compared to spiked fed individuals, water exposed anemones did result in some uptake of GFP-Vibrio with higher levels acquired in the presence of food (non-spiked Artemia) than without. This observation is consistent with the findings of Gavish et al. (2021) (16) and suggests that even in the absence of Vibrio-colonization of food sources, active feeding and ingestion may contribute to the acquisition of Vibrio spp. cells from the surrounding water. It should be noted that corals are known to expel ingested pathogens as a mechanism of defense against infection (58,59). However, during the time period of the present experiment, we did not observe any expulsion from experimental anemones.
At ambient levels, Vibrio spp. typically range from 10 1 to 10 3 CFU mL 21 (60) with location-specific differences in community composition driven largely by temperature and salinity (21,25). However, Vibrio populations are known to be dynamic, fluctuating on a "boom-bust" cycle of growth and reduction in association with ephemeral pulses of limiting nutrients (61)(62)(63). During bloom events, total Vibrio can increase dramatically rising to levels 5 to 30 times greater than the typical background concentration of coastal waters (61). Prior research has shown that seasonal increases in Vibrio abundance facilitate increases in both free-living and zooplankton-associated abundance (64). Thus, bloom numbers could potentially promote zooplankton colonization and enhance the likelihood of transmission via ingestion during these events. While further studies on species-specific colonization rate, transmitted dose, and uptake in situ are needed to assess the potential importance in coral disease, we postulate that these mechanisms provide an ecological basis for foodborne transmission of certain coral pathogens.
While the scope of this research is targeted at understanding coral disease, the results of this study have broader implications for the spread of vibriosis. Vibrio spp. have been implicated as the causative or putative pathogens in numerous diseases of marine organisms, most notably important aquaculture species such as Pacific White Shrimp (Litopenaeus vannamei), Tiger Prawn (Penaeus monodon), Atlantic Salmon (Salmo salar), and Gilt-Head Sea Bream (Sparus aurata) (65)(66)(67)(68). Zooplankton serve as the base of the marine/estuarine food web, thus there is potential for ingestion to play a role in the acquisition of these and similar pathogens. This hypothesis is supported by the work of Goulden et al. (2012) (69) who utilized a similar GFP tracking system to demonstrate that Panulirus ornatus (ornate spiny lobster) mortality can be facilitated by ingestion of V. owensii-colonized Artemia in aquaculture settings. Furthermore, the non-discriminant acquisition of all 3 Vibrio spp. in the present study suggests that this pathway may be broadly viable within the Vibrionaceae and warrants continued investigation of the role of ingestion in the spread of other pathogenic vibrios.
Conclusion. Understanding coral disease transmission is critical to the conservation of reef habitats. The present study describes a mechanistic pathway for the acquisition of coral pathogens via zooplankton ingestion using a sea anemone (E. pallida) and brine shrimp (Artemia) model system to represent coral heterotrophy. The results of this research demonstrate that ingestion of Vibrio-colonized Artemia can facilitate receipt of a significantly elevated Vibrio dose when compared to exposure via the water column suggesting that heterotrophy may represent an important portal of entry for certain coral pathogens. Characterization of this pathway illustrates a means by which pathogenic bacteria may bypass the natural immune defenses of corals conferred by their mucus membranes allowing for unencumbered acquisition of a pathogenic dose. This mechanism may help to explain a potential source of idiopathic infections that arise in otherwise healthy unperturbed corals.

MATERIALS AND METHODS
Experimental Vibrio strains. Experimental Vibrio strains were obtained from our culture collection (E.K. Lipp, University of Georgia) and consisted of the known coral pathogens V. alginolyticus, V. mediterranei, and the putative coral pathogen V. harveyi ( Table 2). All strains were maintained at 280°C in a 1:1 mixture of 40% glycerol (20% final concentration) and LB (Sigma-Aldrich, Miller formulation) amended to 3% wt/vol NaCl (termed LBS 3%). To revive from storage, strains were inoculated into 4 mL LBS 3% and incubated at 30°C with 100 rpm shaking agitation (New Brunswick Scientific, C24 Incubator Shaker) for 18 to 24 h.
Brine shrimp cultures and maintenance. Artemia sp. were purchased as dehydrated cysts (Premium Grade Brine Shrimp Eggs: Brine Shrimp Direct Inc., Great Salt Lake Origin). Dehydrated cysts (0.3 g) were revived in 300 mL sterile artificial seawater (35 practical salinity units [PSU] Instant Ocean, termed ASW) incubated at room temperature under mild agitation from an aquarium bubbler (Whisper 20, Aquarium Air Pump). Cysts hatching occurred within 1 to 2 days of rehydration. Artemia were harvested at the nauplii stage, following 1 to 2 additional days of incubation, using a sterile serological pipette. Free swimming nauplii were collected from below the water surface to reduce collection of any discarded or unhatched cysts. Any Artemia cultures that appeared discolored (cloudy water), produced poorly swimming nauplii, or hatched insufficiently (,75% hatching, estimated visually) were discarded.
Anemone cultures and maintenance. E. pallida anemones were purchased live (Carolina Biological Supply, #162865) and maintained in laboratory holding tanks. Holding tanks were constructed using a 6 L glass aquarium equipped with a constant-flow water filter (Aqueon QuietFlow Aquarium Power Filter 10), an in-water aquarium heater (Aqueon Pro Heater 50W), and a 445 nm aquarium light (GloFish Blue, LED Aquarium Light). Holding aquaria were maintained under the conditions outlined in Tables S2 and S3. Prior to experimentation, all anemones were transferred to holding tanks and allowed to acclimate for a minimum of 2 weeks. Anemones were monitored daily, and any deceased individuals were removed. Long-term cultures (not used for experimentation) of E. pallida were kept with experimental anemones to stabilize holding tank water chemistry. While in the holding tank, anemones were fed twice per week with 50 mL (;2,000 individuals) of decapsulated Artemia. Water changes (50% of tank volume) were peformed every 2 weeks and replaced volumetrically with fresh ASW. Intermittent tank cleaning was performed as needed using a scrub brush and/or a serological pipette to remove anemone debris and algal build-up following feeding.
GFP tagging. All Vibrio spp. used in this experiment were tagged with GFP to enable localization and quantification of the bacterium. Tagging was accomplished using the methods outlined in Norfolk & Lipp, (2022) (70). In short, a tri-parental mating assay was used to transfer a gfp-containing plasmid to the target Vibrio sp. using bacterial conjugation. In this assay, 2 strains of Escherichia coli, the helper strain carrying the conjugative plasmid pEVS104 (tra trb Kn r ), and the donor strain carrying the gfp plasmid pVSV102 (gfp Kn r ), were combined in culture with the target Vibrio under mild kanamycin stress to promote transfer of the gfp plasmid. Mating cultures were then subsequently streaked onto thiosulfate bile salts sucrose agar (TCBS) agar to remove E. coli resulting in a purified GFP-tagged Vibrio strain. Purification was confirmed using subsequent growth on modified mTEC agar (Difco, Fischer Scientific), an E. coli specific medium. Fluorescence of all transconjugant (GFP-tagged) Vibrio spp. was confirmed using fluorescence microscopy (Olympus BX41 Fluorescence Microscope). Working stocks of transconjugant strains were maintained at room temperature in deep agar stabs containing LBS 3% amended with 300 mg mL 21 kanamycin to ensure retention of the plasmid. GFP strains were maintained at 280°C in a 1:1 mixture of 40% glycerol and LBS 3% broth amended with 300 mg mL 21 kanamycin for long-term storage.
Artemia colonization. GFP-Vibrio spp. were revived from 280°C storage in 4 mL of LBS 3% broth amended with 300 mg mL 21 kanamycin and incubated at 30°C with 100 rpm of shaking agitation for 18 to 24 h. Following incubation, 1 mL of the overnight culture was pelleted by centrifugation at ; 4,000 Â g, the supernatant was discarded, and replaced with 1 mL of sterile 1X phosphate-buffered saline (PBS). This process was repeated three times to ensure adequate removal of residual kanamycin from the culture. Concurrently, Artemia cultures were grown as described above in "Brine Shrimp Cultures and Maintenance" to produce free swimming nauplii. Six mL of decapsulated nauplii (; 250 individuals) were transferred to each well of a sterile 6-well tissue culture plate (Cellstar 6-Well Suspension Culture Plate). Each well of the culture plate was inoculated with 50 mL of washed GFP V. alginolyticus (; 1.1 Â 10 8 CFU), V. harveyi (; 6.9 Â 10 7 CFU), or V. mediterranei (; 1.5 Â 10 7 CFU). The Artemia-Vibrio mixture was covered and incubated at 28°C under 50 rpm of shaking agitation for 18 h. This exposure duration was selected to facilitate sufficient colonization of Artemia, which appeared too low for experimental needs after only 3 h of exposure (Fig. S5). Following incubation, the contents of each well was collected onto a 3.0 mm polycarbonate (PCTE) membrane (Sterlitech 47 mm and 3.0 mm PCTE membranes) using vacuum filtration to capture the suspended Artemia while allowing any non-associated Vibrio cells to be discarded as flow through. The Vibrio-colonized Artemia were resuspended from the membrane by vortexing for 30 s in 6 mL of sterile ASW. Colonization or apparent attachment of GFP-labeled cells to Artemia nauplii was confirmed using epifluorescence microscopy. Spiked Artemia were then homogenized (PRO Scientific, Series 250 Homogenizer) at max speed for 120 s, and homogenate was serial diluted (10-fold in 1X PBS and spread plated using glass rattler beads [Zymo Rattler Plating Beads, 4.5 mm]) onto TCBS agar amended with 300 mg mL 21 kanamycin in duplicate. The addition of kanamycin to the TCBS plates selected against any non-GFP-tagged Vibrio cells that may have been present. TCBS plates were incubated overnight at 30°C. The resulting plate counts were used to calculate the approximate level of acquired dose.
Uptake by E. pallida. To establish a connection between ingestion and Vibrio uptake, a controlled feeding study was conducted to measure the level of acquired GFP-tagged Vibrio spp. following exposure in a microcosm. Cultures of GFP-tagged Vibrio spp. and Artemia were prepared and combined as described above in "Artemia Colonization" to produce spiked Artemia. To increase the feeding opportunity, 4 Artemia spike exposures (; 250 individuals each) were combined for a total exposure of ; 1,000 individuals resulting in a maximum feeding dose (assuming ingestion of all Artemia) of ; 2.0 x10 7 CFU, ;5.8 Â 10 6 CFU, and ; 3.0 Â 10 7 CFU for V. alginolyticus, V. harveyi, and V. mediterranei trials, respectively ( Table 3). Colonization of the spiked Artemia was confirmed prior to anemone feeding using fluorescence microscopy. Control Artemia (non-spiked) were prepared in tandem using the protocol but were inoculated with 50 mL of sterile 1X PBS instead of GFP-Vibrio spp.
Experimental microcosms were constructed to house the anemones during exposure trials. Microcosms were created using 18 Â 12.5 Â 5 cm Pyrex dishes filled with 750 mL of sterile ASW. Each microcosm contained a submerged six well tissue culture plate to provide substrate for E. pallida (N = 6 per treatment). Prior to exposure, experimental E. pallida were transferred to the microcosm chambers and allowed to acclimate for 18 h. Anemones used in experiments were selected based on size and consisted of individuals ranging from 1.5 cm to 3 cm (at full extension) to reduce the influence of feeding bias by large or small individuals. No discolored or wilting anemones were selected (see Fig. S4 for an example of healthy E. pallida appearance). Care was taken during anemone detachment to ensure no damage to the tentacles or oral disk occurred. All anemones were checked visually for viability following acclimatization and replaced as needed. Experimental exposures were administered as detailed in Table 3 for a duration of 3 h. For trials where Artemia were fed to E. pallida, anemones were observed for the first 20 min following exposure to visually confirm ingestion. Anemones were rechecked every 30 min to ensure feeding behavior was continued and to stir microcosm water (to prevent Artemia from congregating out of anemone reach). Following exposure, anemones were collected from the chambers, transferred into individual 50 mL conical tubes containing 40 mL of sterile ASW, and vortexed for 30 s. This process was repeated twice to remove any non-ingested GFP-Vibrio cells. Washed anemones were then transferred into 10 mL of sterile ASW for homogenization. A total of 100 mL of ASW was removed prior to homogenization and spread plated with glass rattler beads (Zymo Rattler Plating Beads, 4.5 mm) onto TCBS agar amended with 300 mg mL 21 kanamycin to ensure no ambient GFP-Vibrio (non-ingested) remained in the wash water (carry-over control). All anemones were then homogenized (PRO Scientific, Series 250 Homogenizer) at max speed for 120 s. E. pallida homogenate was serial diluted (10-fold) in 1X PBS and spread plated with glass rattler beads onto TCBS agar amended with 300 mg mL 21 kanamycin, in duplicate. Plates were incubated at 30°C for 18 h. The resulting plate counts (CFU/mL) were used to calculate the uptake of GFP-Vibrio cells by the anemones under each experimental condition. Culture results were summarized and visualized in Rstudio using the packages 'tidyverse' and 'readxl.' Feeding exposures were compared using a pairwise Wilcoxon rank-sum test with a Bonferroni correction for significance.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.2 MB. SUPPLEMENTAL FILE 2, XLSX file, 1.4 MB.