Biol Invasions (2025) 27:44 https://doi.org/10.1007/s10530-024-03476-2 ORIGINAL PAPER Analysis of bulk stable isotopes and trophic positions of invasive lionfish (Pterois volitans) on deep versus shallow reefs at Curacao Megan M. Ewing · Rachel Welicky · Carole C. Baldwin · D. Ross Robertson · Katherine P. Maslenikov · Luke Tornabene Received: 11 May 2023 / Accepted: 25 October 2024 © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2024 Abstract In the decades since their introduction, due to challenges collecting lionfish below tradi- invasive Red lionfish (Pterois volitans) have spread tional SCUBA-diving depths. In the present study, throughout reefs across the entire tropical western lionfish were collected at Curaçao from deep reefs Atlantic Ocean (TWA). Adult lionfish are largely (60–174 m) using the Curasub submersible and were free from predation in the TWA, where they prey compared to specimens collected from adjacent shal- on native fish populations and compete with native low reefs (5–37 m). From each fish, scale, muscle and mesopredatory fishes. Recent explorations of deep- heart tissue were subjected to bulk stable carbon and reef habitats have found that lionfish are common nitrogen isotope analyses to examine trophic differ- on mesophotic and rariphotic reefs down to 304  m, ences over long, intermediate, and short-term tempo- bringing into question the extent of impact that inva- ral scales, respectively. We found that deep lionfish sive lionfish have on deep vs shallow reefs. A recent have a significantly higher (~ 0.3‰) trophic position Caribbean study found that lionfish from mesophotic than shallow lionfish. Our results also indicate that on habitats are older, bigger, more reproductively active, intermediate and longer temporal scales, δ13C values and have denser populations compared to those from of deep lionfish overlap with but are slightly broader shallow reefs. However, variation in trophic ecology than that of shallow lionfish. However, on shorter between these two populations is not well understood time scales, deep lionfish are depleted in δ13C. This may suggest lionfish migrations between deep and Supplementary Information The online version shallow reefs, or that the carbon signatures across contains supplementary material available at https:// doi. depth are more similar than previously considered. org/ 10.1 007/ s10530-0 24- 03476-2. M. M. Ewing · R. Welicky · K. P. Maslenikov · R. Welicky  L. Tornabene (*)  Unit for Environmental Sciences and Management, School of Aquatic and Fishery Sciences, University North-West University, North-West, Potchefstroom 2523, of Washington, Seattle, WA 98105, USA South Africa e-mail: Luke.Tornabene@gmail.com C. C. Baldwin  M. M. Ewing · K. P. Maslenikov · L. Tornabene  Department of Vertebrate Zoology, National Museum Burke Museum of Natural History and Culture, Seattle, of Natural History, Smithsonian Institution, Washington, WA 98105, USA D.C 20560, USA R. Welicky  D. R. Robertson  College of Arts and Sciences, Neumann University, Aston, Smithsonian Tropical Research Institute, Balboa, Panama PA 19014, USA Vol.: (0123456789) 44 Page 2 of 14 M. M. Ewing et al. These findings underscore the importance of under- history stages. This may be the case with lionfish, standing the impact invasive lionfish have on reef which occupy a variety of nearshore habitats during ecosystems and may have broad implications for the their ontogeny including euryhaline estuaries, sea- effectiveness of invasive-species management on grasses, mangroves, and coral reefs, where they can deep reefs. exploit different prey communities. Recently, lionfish were discovered on western Atlantic reefs at meso- Keywords Diet · Caribbean · Fishes · Mesophotic photic (40–150 m) and rariphotic (150–309 m) depths reefs · Rariphotic reefs · Non-native · Coral reefs to a maximum depth of 304  m (Lesser and Slattery 2011; Gress et al. 2017). Although comparatively less is known about fishes on mesophotic and rariphotic Introduction reefs compared to shallow reefs, recent surveys have shown that native fish communities on deep reefs Since the mid-1980s, invasive Red Lionfish (Pterois (to 300  m) in the Greater Caribbean are taxonomi- volitans) has spread throughout the entire tropical cally and ecologically distinct from those on shal- western Atlantic (TWA), outcompeting native preda- low reef habitats (< 40 m; Baldwin et al. 2018; Ste- tors, and posing a threat to native prey species by fanoudis et  al. 2019). Due to logistical constraints causing significant decreases in prey abundance and on conducting research on deep reefs, the biology reductions in recruitment in some regions (Whitfield of many deep-reef fishes is poorly understood; some et al. 2002; Semmens et al. 2004; Albins and Hixon species have only been described recently, and others 2008; Green et  al. 2012; Soares et  al. 2022). The remain undescribed (Asher et al. 2017; Baldwin et al. extent to which P. volitans has spread in the Atlantic 2018; Kahng 2010; Robertson et  al. 2022). We cur- is concerning because, while some instances of pre- rently lack a clear understanding of the threats posed dation on lionfish by native piscivorous fishes have to deep-reef fishes and whether lionfish pose a simi- been observed there, such predation is not frequent lar danger to populations on deep reefs as they do on or widespread enough to control lionfish population shallow reefs. growth (Maljkovic et  al. 2008; Munoz 2017). This, In the face of the potential threat posed to deep- in combination with lionfish’s high fecundity, has led reef organisms that may be consumed by lionfish, to an uncontrolled population increase in the Atlantic recent research efforts have sought to understand how (Gardner 2015). Lionfish in the TWA typically reach the biology of deep lionfish populations differs from maturity within 1 year of age and spawn as frequently that of lionfishes on shallow reefs. Green et al. (2021) as every 2–3 days once mature (Gardner et al. 2015). found that lionfish home ranges are larger than pre- Lionfish in their invasive range have been observed viously thought and that during long-distance travel to reach up to 10  years old, though the average age (> 1 km), some individuals move from shallow reefs of a lionfish population at a particular reef varies site to mesophotic reefs. This relocation potentially is uni- to site with some regions averaging from 2 to 4 years directional and, given that most lionfish management (Blakeway et  al. 2022; Edwards et  al. 2014). This efforts are through shallow-reef culling, it indicates combined with their high reproductive output has led that deep reefs may act as a refuge from culling and to the success of their invasion over the last several a source of new recruits to shallow reefs where cull- decades. Initially introduced off the coast of Florida ing occurs (Hinderstein et al. 2010; Green et al. 2021; in the 1980s, lionfish have since spread through- Airey et al. 2023). One possible cause of this move- out the western Atlantic including the east coast of ment to deep reefs could be increased prey availabil- the United States, Bermuda, the Gulf of Mexico, ity compared to the shallow reefs that have had many throughout the Caribbean, and, most recently, to Bra- resources exhausted by the lionfish invasion (Mal- zil (Soares et al. 2022). pica-Cruz et al. 2019). Andradi-Brown et al. (2017a) Success of invasive species may be linked to plas- found that, on reefs at Utila Island, Honduras, that ticity in their habitats and dietary niches (Davidson are subject to heavy shallow culling, lionfish from et al. 2011; Balzani et al. 2021; Glassic et al. 2023). upper-mesophotic habitat (30–85  m) tended to be This plasticity may coincide with species segregat- larger, occurred at higher densities, had a higher pro- ing spatially and temporally based on different life portion of spawning females, and had more mature Vol:. (1234567890) Analysis of bulk stable isotopes and trophic positions of invasive lionfish (Pterois volitans)… Page 3 of 14 44 individuals than did shallow-reef (< 30  m) popula- muscle, and scale tissues of fishes enable ecolo- tions. These characteristics point to deep lionfish gists to examine differences in fish trophic ecology having greater energetic demands, and they could be across short, intermediate, and longer time scales consuming more and/or larger prey. Consequently, in (i.e. ~ 1  week, ~ 2  months and ~ 4–5  months, respec- such situations deep lionfish may pose a greater threat tively; Vander Zanden et al. 2015; Busst and Britton to their prey communities than their shallow counter- 2018). Thus, by examining differences in 15N content parts. This is particularly problematic given observa- in lionfish tissues, we can infer the trophic ecology tions of deep lionfish preying upon undescribed spe- across time of our shallow and deep reef populations. cies (Tornabene and Baldwin 2017). While studies From stable isotope analyses we know that lion- on mesophotic lionfish are limited, recent studies in fish across the wider Caribbean are generalist preda- the Bahamas and Bermuda showed that lionfish in the tors and forage similarly to native meso- and apex upper mesophotic zone congregate in areas of high predators, including the coney (Cephalopholus fulva) prey-fish density (Goodbody-Gringley et  al. 2019), and Nassau groupers (Epinephelus striatus) (Eddy and that the invasion of upper mesophotic reefs by et al. 2020; Muñoz et al. 2011; O’Farrell et al. 2014; lionfish caused shifts in community structure of prey Aragonés Fred 2022). Lionfish niche breadth and species and declines in herbivore abundance (Lesser position are dependent on body size. In the inva- and Slattery 2011). Nonetheless, basic information sive range, the dietary niche of lionfish is broader regarding the diet, feeding behavior, and trophic posi- and they occupy a slightly higher trophic position tion of deep lionfish in comparison to those on shal- than in the native range (Malpica-Cruz et  al. 2019). low reefs currently is lacking. This shift coincides with a shift in prey community One useful method for analyzing the feeding ecol- structure (Malpica-Cruz et  al. 2019) resulting from ogy of an organism is via bulk stable carbon and small lionfish consuming abundant, small, solitary, nitrogen isotope analyses. The unique δ13C signatures shallow-bodied prey-fish before moving on to larger- from different reef environments allow us to evalu- bodied prey as they grow (Muñoz et al. 2011; Green ate the breadth of habitats from which lionfish’s prey and Côté 2014; Ritger et al. 2020). Specifically, Rit- originated. Specifically, δ13C values differ among ger et  al. (2020) found that invasive lionfish, espe- coastal habitats based on variation in the sources of cially larger individuals and those with increased primary production, because shallow inshore sys- condition (weight/length3), tend to prefer larger and tems driven primarily by benthic carbon sources (e.g. more frequently-encountered prey. However, those mangroves, seagrass beds, shallow reefs) will have studies focused mainly on lionfish on shallow reefs. greater δ13C values compared to deeper or offshore Two studies examined specimens collected from the ecosystems driven by planktonic carbon sources (Fry upper mesophotic zone at around 60  m (Eddy et  al. et al. 1982; France 1995; Hofmann and Heesch 2018; 2020; Aragonés Fred 2022) and two studies examined Lamb and Swart 2008). These habitat-specific dif- lionfish below 60 m (7 specimens collected between ferences in δ13C values are reflected in herbivorous 60–160  m, Sanjuan-Muñoz et  al. 2022; 155 speci- consumers and predatory fishes, including lionfish mens collected between 40–72  m; Andradi-Brown (France 1995; Pimiento et  al. 2015; Bradley et  al. et  al. 2017a). Thus, little is known at present about 2016; Cocheret de la Morinière et al. 2003). Moreo- the trophic ecology of lionfish on mesophotic and ver, fish communities on deep reefs are increasingly rariphotic reefs – which extend from 40 to ~ 300  m decoupled from benthic carbon sources due to a dra- and thus make up approximately three-quarters of the matic decrease in herbivory that reflects reductions in vertical extent of the entire reef slope. photosynthesis with depth (Slattery et al. 2011; Pin- One of the main challenges with studying deep- heiro et al. 2023). Thus, we expected to see a corre- reef lionfish diet is collecting specimens from below sponding depletion of δ13C signatures in lionfish liv- the traditional SCUBA-diving range. The Smith- ing on deep reefs compared to those found on shallow sonian Institution’s Deep Reef Observation Project reefs (Bradley et al. 2016). (DROP) has studied deep-reef fish communities in The accumulation of 15N in body tissues increases the Caribbean since 2010 using the human-occupied as an individual increases in trophic level, thus submersibles CURASUB and IDABEL, which have indicating its trophic position (Fry 1988). Heart, proven to be very effective tools for both surveying Vol.: (0123456789) 44 Page 4 of 14 M. M. Ewing et al. larger reef fishes and collecting smaller fishes (Rob- 2012 the government exempted the spearing of lion- ertson et  al. 2022). Submersibles, however, have fish from the spearfishing ban for registered fisher- rarely been used for capturing larger fishes such men hunting with pole-spears. Our study site is near as lionfish. In 2019, as part of the effort to assess popular dive sites and thus has experienced lionfish the impact of lionfish on deep reefs, DROP outfit- culling down to traditional SCUBA-diving depths (0 ted the CURASUB with multi-prong pole-spears to to ~ 30 m). collect lionfish. This ability to capture deep lionfish We targeted lionfish from shallow reefs (5–40 m) allowed us to use bulk stable carbon and nitrogen iso- via SCUBA and from deep reefs (60–175  m) using tope analysis to compare trophic position and dietary the submersible CURASUB (max depth rating of niche breadth between shallow and deep lionfish from 300 m) and technical divers operating down to 70 m. Curaçao in the Southern Caribbean. Because lionfish The CURASUB was outfitted with two pole spears on upper-mesophotic reefs have been shown to be that were attached to the front of the sub (Fig.  1A). more mature and larger than lionfish on shallow reefs One pole spear was attached to a robotic arm and was (Andradi-Brown et al. 2017b), deep lionfish may have reloadable via a motor. A second spear was attached increased energetic demands and the ability to eat to a fixed position on the front of the sub (Fig. 1B) and more and or larger prey. Thus, we hypothesized deep was reloaded manually using one of the sub’s robotic lionfish occupy a higher trophic position than shal- arms. Speared lionfish were sedated with a solu- low lionfish. We predicted that deep populations of tion of quinaldine in ethanol while still on the spear lionfish would have enriched δ15N values compared (Fig. 1C), and, once sedated, they were removed from to shallow populations, and, because δ13C typically the spear using one of the robotic arms and deposited becomes depleted with depth (e.g. Ceia et al. 2018), in a collection basket on the front of the submersible we predicted that deep populations of lionfish would (video available on Zenodo digital repository https:// have depleted δ13C values compared to members of doi.o rg/ 10.5 281/ zenodo. 14369 368). All collected fish shallow populations if deep lionfish restrict their were measured for Standard Length (SL) and frozen feeding to deep habitats. in a −20C freezer before being shipped on ice from Curaçao to the University of Washington Fish Collec- tion in Seattle, Washington, USA, where they were Methods stored at −80C. Site description and Lionfish Collection Stable isotope analysis Our study was conducted at a single site off the south- western coast of Curaçao (12.0831°N, −68.8991°W). Lionfish were thawed at −20C for 24  h and then at The area sampled in this study is < 1 km2. The reef room temperature to facilitate extracting tissue. We at our site matches the general characteristics of collected heart, muscle (taken from the back below mesophotic reefs of Curaçao and the nearby island of the spinous dorsal fin), and scale tissues following Bonaire (see detailed description in Frade et al. 2019; standard operating procedures for bulk stable carbon Baldwin et  al. 2018). Briefly, at our study site there and nitrogen isotope analysis (Søreide et  al. 2006). is moderate coral cover (16–40%) on the shallow We used these tissues to examine dietary changes (< 20 m) leeward side of the island (WAITT Institute, over short (~ 1  week; heart tissue), intermediate 2017). The gradually sloping shallow reef has more (~ 2 months; muscle tissue), and long temporal scales steep drops between 7–17 m depth, followed by sev- (~ 4–5  months; scale tissue) (Vander Zanden et  al. eral near vertical walls starting at 70–90  m depth, 2015; Busst and Britton 2018). Prepared samples depending on location. From 90 to 300 m depth, the were analyzed at the UC Davis Stable Isotope Facil- reef is characterized by moderately sloped to steep ity. The mean standard deviation for reference mate- cliffs separated by intermittent flat rubble and sand rial replicates of bovine liver were 0.06‰ for δ13C beaches. The government of Curaçao initiated a lion- and 0.05‰ for δ15N. The mean absolute accuracy for fish removal program starting in 2011 (de León et al. calibrated reference materials was 0.04‰ for δ13C 2013). While spearfishing is banned in Curaçao, in and 0.04‰ for δ15N. Vol:. (1234567890) Analysis of bulk stable isotopes and trophic positions of invasive lionfish (Pterois volitans)… Page 5 of 14 44 Fig. 1 CURASUB submersible. A Layout of specimen collection equipment. B Spear ready to be fired by depressing trigger with robotic arm. C Applying anesthetic to speared lionfish prior to moving specimen into the storage basket Statistical analysis predict  theoretical  residuals. Simulated residuals were used because of the small sample size of our Descriptive statistics were calculated using the sum- data (n = 76). All response variables were normally marize function in the tidyverse R package (Wickham distributed with the exception of δ13C of scale tis- et  al. 2019). To determine how the bulk stable car- sue, and we verified the normality of this model after bon and nitrogen isotope values of shallow and deep implementing the complete model. Thus, all models lionfish vary, we first tested normality by simulating were fitted with a Gaussian distribution. Because SL the residuals of generalized linear models where the and depth are highly correlated (see Results below), response variable was δ13C and δ15N of each tissue we included SL as a covariate in our linear models type, and the categorical independent variable was (ANCOVA). The complete models for each response depth. Because prey fish communities are highly variable were as follows: structured with depth into distinct zones (Pinheiro et  al. 2016; Baldwin et  al. 2018), we analyzed the Stable Isotopeij ∼ standard lengthij + depth categoryj effect of depth as a categorical variable (shallow vs. (1) deep), with the break between shallow and deep lion- where the response variable ij represents δ13C or δ15N fish being 40 m, which corresponds both to the dis- values of a particular tissue type from the ith fish col- tinction between the altiphotic and mesophotic fau- lected at jth depth. nal zones, as well as the point below which spearing To determine the trophic position of deep and shal- lionfish becomes logistically unfeasible for SCUBA low lionfish we calculated trophic position using the divers. formula, We implemented normality tests using the glm (( ) )15 15 15 function in the R package lme4 (Bates et  al. 2014) Trophic position =  Ni −  Ng ∕Δ N + 1 and predicted their residuals using the simulateR- (2) esiduals and testUniformity functions in the R pack- where δ15N values of muscle tissue are from ith age DHARMa (Hartig et  al. 2022). The assump- lionfish, the value of δ15Ng is 4‰, and the value tion of normality is tested based on theoretical of Δδ15N is 3.4‰. The δ15Ng value is that of the not observed residuals of a model. Therefore, we masked goby (Coryphopterus personatus) (Zhu leveraged the observed residuals of the glm to et  al. 2019), a common prey item of lionfish across Vol.: (0123456789) 44 Page 6 of 14 M. M. Ewing et al. the Caribbean. The given Δδ15N value is the stand- Results ard increase in 15N per trophic level for marine fishes (e.g., Post 2002). Although traditional bulk isotope We captured 31 lionfish from shallow reefs (5–37 m; reference and comparative material includes primary 178.2 ± 50.2  mm SL) and 45 lionfish from deep producers and a range of secondary consumers, such reefs (60–174  m; 246.0 ± 48.1  mm SL). The linear reference material was not collected as this was not model (ANCOVA) showed a significant relationship an original objective of the lionfish spearing pro- between SL and depth, with deeper lionfish being sig- gram. Therefore, we optimized our reference mate- nificantly larger than shallower lionfish (p < 0.0001; rial choice by selecting one of the most common prey Fig.  2). After controlling for the effect of SL in the items of lionfish that is ubiquitous in the Caribbean. ANCOVA analysis, values for δ13C were only sig- Because we recognize this substitution could con- nificantly different between deep and shallow lionfish found results for true values (Kjeldgarrd et al. 2021), for heart tissues, with deeper lionfish being depleted our objectives are to solely look at the relative differ- for δ13C ( Table  1; Fig.  3). Stable carbon isotope ences among the shallow and deep lionfish collected values for muscle and scale tissues showed no rela- for this study. tionship with depth, (Table  1; Fig.  3). When com- Because trophic level is a more commonly used paring only fish between 210–270 mm TL (Fig. 3B; metric than δ15N across ecological literature out- Table  S1) there was no difference in average δ13C side of stable isotope ecology and for food web values between shallow and deep lionfish as, regard- dynamics studies (e.g., Stuthmann and Castellanos- less of tissue type. While average δ13C values showed Galinda 2020; Quintana et al. 2023), we also tested no significant difference between shallow and deep whether trophic position was related to depth using reefs in muscle and scale tissues in our model, there the aforementioned models with trophic position as were differences in the overall range of δ13C values. the dependent variable. When looking at the entire dataset, deep lionfish have Finally, because SL and depth were correlated, a broader range of δ13C values than shallow lionfish to further try and isolate the effect of depth on our in scale and heart tissues (Fig. 3A). Similarly, for fish stable isotope values, we repeated the linear mod- between 210–270 mm TL, carbon values are broader els above for δ13C, δ15N and trophic position using in deep fish compared to shallow for all tissue types a subset of our data that included only lionfish (Fig. 3B). that were between 210–270  mm total length (TL) The ANCOVA analysis showed that δ15N values (160–212 mm SL). This size bin was selected based of all three tissue types were significantly enriched on studies from multiple locations in the Western in deep lionfish (Table  1; Fig.  3). There was also a Atlantic that show that lionfish at this size are at significant and positive relationship between SL and least one year old and both sexes are mature (John- δ15N values in all tissue types (Table 1; Fig. 3A). son and Swenarton 2016; Gardner et al. 2015; Fogg The trophic position of deep lionfish was sig- et  al. 2017, 2019). This also represents the area nificantly higher than shallow lionfish (deep, of our dataset with the most size overlap between 2.779 ± 0.213; shallow, 2.458 ± 0.141; Table  1; shallow and deep lionfish. This subset included 15 Fig. 3A). When analyzing the subset containing only shallow lionfish (mean depth 17  m) and 13 deep fish between 210–270  mm TL, there was no longer lionfish (mean depth 100  m). While the reduced a significant relationship between SL and depth, but sample size of this size-corrected subset may limit δ15N values and trophic position were all still signifi- the overall strength of this subset of our analysis cantly higher in deep lionfish (Fig. 3B, Table S1). and subsequent extrapolation of trends, it is the most straightforward way to remove the effect of size. Discussion Data and code are available at https://g ithub. com/ rlweli cky/E wing-e t-a l.- 2025 as well as on Our findings support the hypothesis that deep lion- Dryad (doi: https://d oi. org/ 10. 5061/ dryad.8 66t1 fish populations are enriched in δ15N, they occupy a g21f). higher trophic position than the shallow populations, and that this relationship is consistent across different Vol:. (1234567890) Analysis of bulk stable isotopes and trophic positions of invasive lionfish (Pterois volitans)… Page 7 of 14 44 Fig. 2 Relationship between standard length and depth of capture, p = 1.053e-06, R.2 = 0.526 time scales (i.e. tissue types). Our hypothesis that that increased concentrations of elements correlat- δ13C values are depleted in deep lionfish is supported ing with δ15N can be explained simply by the lion- for heart tissue, which provides short term informa- fish consuming increased quantities of prey. Indeed tion, but not in scale or muscle tissues (Table  1; on shallow reefs where invasive lionfish are present, Fig. 3). Interestingly, we also found the range of δ13C local prey fish abundance decreased by 45% or more values in deep lionfish broader than, and overlapping at some sites (Ballew et al. 2016; Green et al. 2012), with, that of shallow lionfish (Fig.  3). The possible and such changes might be reflected in the isotope mechanisms driving the patterns in δ15N and δ13C signatures of lionfish. Therefore, it is plausible that include prey quantity (e.g., Ritger et  al. 2018; Cure deep lionfish eating more prey could explain the δ15N et al. 2012), prey habitat (Eddy et al. 2020; Sanjuan- enrichment of their tissues compared to shallow lion- Muñoz et  al. 2022; Aragonés Fred 2022), lionfish fish (Fry et al. 1988; Fry 2006) and if this is the case, vertical movement along the reef slope, and age/size then deep-reef populations may have a more detri- differences between lionfish on deep versus shallow mental impact on deep native prey fish abundance reefs (e.g., Andradi-Brown et al. 2017b), and are dis- than previously observed. cussed below. An alternative or additional explanation is that deep lionfish may be eating larger prey, and that such Deep lionfish have a higher trophic position prey are generally enriched in δ15N compared to shal- low lionfish prey. Compared to their native range, Our results show that deep lionfish are feeding at a invasive lionfish tend to consume larger prey– nearly higher trophic position than shallow lionfish (Fig. 3, double the size of their prey in the Pacific (Cure et al. Table 1), and that this relationship is true even when 2012). The trophic position of deep lionfish may analyzing lionfish that are of a similar size range therefore reflect the proportional composition and (Fig.  3B; Table  S1). The higher trophic position abundance of deep-reef prey items. For example, deep of deep lionfish populations may be explained by lionfish may be feeding on a higher proportion of pis- the quantity of prey they are consuming and/or the civores than shallow lionfish. Bauer (2020) found that trophic level of their prey. Ritger et al. (2018) investi- shallow lionfish consumed more decapods, whereas gated invasive shallow lionfish in Curaçao and found lionfish from the upper mesophotic zone (30–57  m) Vol.: (0123456789) 44 Page 8 of 14 M. M. Ewing et al. seem to favor more δ15N enriched prey, including tel- eosts (Bauer 2020). Another possibility is that the proportions of prey types (e.g. invertebrates, and fishes that include benthic invertivores, planktivores and piscivores) between shallow and deep lionfish may be simi- lar, but deep-reef prey species may inherently have higher trophic positions, even within prey feeding guilds. For example, Bradley et  al. (2016) showed that benthic invertivores had higher trophic positions on mesophotic reefs versus shallow reefs. This differ- ence between the trophic position of deep and shal- low prey species may explain the subtle but signifi- cant increase (0.3‰) in trophic position we observed between deep and shallow lionfish. The masked goby we used as a standard was obtained from within our shallow reef depths (Zhu et al. 2019) as data on deep sea prey items are currently limited or unavailable, but the masked goby represents a commonly avail- able prey species in the Caribbean (Randall 1996). Interestingly, our study differs from previous studies on mesophotic lionfishes, where ‘deep’ lionfish had depleted δ15N values as compared to shallow lion- fish (Eddy et al. 2020; Aragones Fred 2022; Sanjuan- Muñoz et  al. 2022). These conflicting results may be explained by those studies not sampling lionfish below 60  m (which is near the shallow end of our ‘deep’ category), or their deeper sites being progres- sively further from shore and disconnected from land- based nitrogen sources (Sanuan-Muñoz et al. 2022). Unsurprisingly, there was a significant correlation between depth and lionfish size of our samples, with lionfish tending to increase in size with depth (Fig. 2). This is consistent with previous observations of lion- fish on mesophotic and shallow reefs in Honduras (Andradi-Brown et al. 2017a). Such a pattern may be an artifact of culling larger lionfish on shallow reefs, or an indication of an ontogenetic shift from shallow to deep reefs (Claydon et  al. 2012; Andradi-Brown et  al. 2017b); evidence for the latter explanation is conflicting in the literature (Andradi-Brown 2019; Airey et  al. 2023). However, even when compar- ing lionfish of identical sizes (Fig.  3B; Table  S1), thus controlling for the relationship between depth and size, deep lionfish were still enriched for δ15N. This suggests that if deep lionfish are indeed eating more prey, or higher-trophic-level prey, or both, it is not simply an artifact of deep lionfish being larger. Instead, it may have to do with differences in prey Vol:. (1234567890) Table 1 Results from linear models of full dataset testing a relationship between δ13C and δ15N and depth, with Standard Length (SL) as a covariate δ13C δ15N Sum Sq Shallow mean ± SD Deep mean ± SD F P Sum Sq Shallow mean ± SD Deep mean ± SD F P Heart Depth 6.800 − 17.1 ± 0.5 − 17.7 ± 1.0 9.839 0.002 14.134 8.5 ± 0.7 10.0 ± 0.8 28.771 0.000 SL 1.240 1.789 0.185 7.104 14.461 0.000 Muscle Depth 0.110 − 16.8 ± 0.8 − 17.1 ± 0.9 0.167 0.684 6.723 9.0 ± 0.5 10 ± 0.7 19.661 0.000 SL 2.450 3.585 0.062 5.366 15.692 0.000 Scale Depth 0.500 − 14.2 ± 0.5 − 14.3 ± 0.9 0.881 0.351 3.520 8.7 ± 0.3 9.2 ± 0.6 12.375 0.001 SL 0.200 0.355 0.553 0.009 0.031 0.860 Trophic Position Depth 0.582 2.5 ± 0.1 2.8 ± 0.2 19.661 0.000 SL 0.464 15.692 0.000 Bolded p-values indicate significance at the a = 0.05 level. All analyses have one degree of freedom Analysis of bulk stable isotopes and trophic positions of invasive lionfish (Pterois volitans)… Page 9 of 14 44 Fig. 3 Stable nitrogen and carbon isotope average values and shallow (light) lionfish, and ellipses showing a range of 95% ranges. Isoplots showing standard error bars for scale (square), of samples for each tissue type, for A All lionfish and B Using muscle (triangle), and heart (circle) tissues of deep (dark) and only 210–270 mm TL lionfish community composition, or alternatively, increased lionfish are abundant on deep reefs. Larger prey fishes overall feeding activity in deep vs shallow waters. can potentially be those that are more likely to be sex- Increased feeding on deep reefs could be due to the ually mature and more fecund, and losing such fishes perpetual low-light conditions that would be favora- could influence population dynamics. For example, ble for lionfish, who are otherwise known to be cre- Albins and Hixon (2008) found that in controlled puscular predators in the shallows (Green et al. 2021). field experiments on shallow reefs, invasive lionfish Regardless of the mechanism, lionfish presence on could strongly reduce native prey fish recruitment by deep reefs might indirectly affect prey fish recruit- an average of 79%. The increased trophic position of ment more than previously expected, especially if invasive lionfish on deep reefs could therefore make Vol.: (0123456789) 44 Page 10 of 14 M. M. Ewing et al. this impact on recruitment be felt to an even greater by crepuscular feeding lionfish; Green et  al. 2011), extent by the prey species on deep reefs, but recruit- versus ontogenetic and unidirectional. Nevertheless, ment dynamics of deep-reef fishes are poorly studied deep lionfish in Curaçao show evidence of exploiting when compared to fishes on shallow reefs. unique carbon sources at the timescale of several days prior to capture. The potential then to observe deep Resource overlap and partitioning between shallow lionfish at shallow reefs has important implications and deep lionfish for the management potential of this invasive species. It is imperative to note that the second explana- We expected to see a depletion of δ13C signatures in tion, that there is no difference in carbon signatures lionfish living on deep reefs compared to those found with depth at this site, cannot be ruled out. Currently on shallow reefs (Bradley et al. 2016). In both our full there is limited understanding of the carbon signa- dataset (Fig.  3A) and the dataset containing lionfish tures in Curaçao, and different regions are subject to from only a narrow size range (Fig. 3B), δ13C values having unique carbon signatures and trends (Hoffman of deep lionfish overlap with, but are slightly broader and Heesch 2018). To confirm whether or not our and more variable than those of shallow lionfish. This observed results indicate shared habitat and/or prey is true for nearly all tissue types, the only exception resources, a better understanding of carbon isotope being heart tissue in our full dataset, where δ13C is signatures of deep and shallow reefs in Curaçao is lower in deep lionfish (Fig.  3A). While little is cur- necessary. Further, to fully understand lionfish move- rently known about δ13C signatures of different Car- ment, a better proxy for movement is needed, poten- ibbean reef environments specifically, broader stable tially with future studies on gut content analysis. isotope literature suggests that δ13C values are dis- tinct source to source, including different reef envi- Management implications and limitations ronments, and consequently different depths (Böhm et al.1996; Davias et al. 2013; Fry et al. 1982; Lesser Lionfish removal efforts rely almost exclusively on et al. 2022). This leads to two possible explanations spear-fishing by divers, have not been scaled up to for our observed results: 1) that lionfish from the deep widely effective levels, and have been limited to shal- are migrating to the shallow reef environments on a low reefs as deep reefs are often beyond the limits of regular basis, or 2) that there is no difference in car- conventional SCUBA (de Leon et  al. 2013; Morris bon signatures between deep and shallow reefs. 2012). Short term local successes have been possible, Should the first explanation be true, the overlap- but without consistent removal efforts, the recruit- ping δ13C values of the shallow and deep lionfish may ment from unculled areas can reestablish lionfish on indicate some degree of shared habitat and/or prey previously culled reefs (Andradi-Brown et al. 2017b; resources. Under this assumption, deep lionfish show- de Leon et al. 2013, Harms-Tuohy et al. 2018). Trap- ing depleted δ13C values on short time scales (heart ping lionfish may be the only realistic alternative to tissue) but not intermediate and long-time scales spearing at depths beyond SCUBA limits. Several (muscle, scale tissue), suggests that over the course types of traps have been used for catching lionfish of weeks to months, habitat use of deep lionfish over- on mesophotic reefs, including lobster traps (Pitt and laps considerably with that of shallow lionfish. Yet, Trott 2013, 2014) and dome and purse traps baited on more immediate time scales, deep lionfish may be with fish aggregating devices (Gittings et  al. 2017). utilizing different carbon sources. Given that, when While these traps are currently less effective than controlled for size, the range of deep lionfish δ13C is spearing at removing lionfish from reefs, continued broader than shallow lionfish, we suspect this overlap refinement may improve capture rates and decrease is due to the deep lionfish moving across habitats and by-catch (Harris et al. 2020). feeding on shallow-reef prey. However, it is possible Overall, our data suggest that lionfish on deep that shallow lionfish are also undergoing migrations reefs are larger, at a higher trophic position, and to the deep. We currently lack the data to determine may have different carbon sources than those on the extent to which lionfish are moving vertically shallow reefs when considering short time scales. between reef zones, and whether this movement is On shallow reefs, lionfish are crepuscular feeders frequent and bidirectional (e. g. diurnal migrations that maximize the decreased light at dawn and dusk Vol:. (1234567890) Analysis of bulk stable isotopes and trophic positions of invasive lionfish (Pterois volitans)… Page 11 of 14 44 to optimize foraging success (Green et  al. 2021). References The perpetual low-light environment of mesophotic and rariphotic reefs may explain any prolonged resi- Airey M, Fogg AQ, Drew J (2023) Invasive lionfish dispersal dence on deep reefs, potentially offering lionfish an between shallow- and deep-water habitats within coastal Floridian waters. https:// doi. org/1 0. 21203/ rs.3. rs- 22280 extended period of crepuscular foraging, despite the 28/v 1 reduced prey availability of deep reefs (Pinheiro Albins MA, Hixon MA (2008) Invasive Indo-Pacific lionfish et  al. 2023). The ability of managers to effectively Pterois volitans reduce recruitment of Atlantic coral-reef manage deep lionfish populations will depend on fishes. Mar Ecol Prog Ser 367:233–238Andradi-Brown DA (2019) Invasive lionfish (Pterois volitans our ability to trap deep lionfish, as well as much and P. miles): Distribution, impact and management. In: better understanding of their movements between Loya Y, Puglise KA, Bridge TCL (ed) Mesophotic Coral depth zones. Ecosystems, Springer Cham, 931–941. https:// doi. org/ 10. The findings and conclusions of this study with 1007/9 78-3- 319- 92735-0_4 8Andradi-Brown DA, Gress E, Wright G, Exton DA, Rogers AD respect to the movement patterns of lionfish across (2016) Reef fish community biomass and trophic struc- depth are limited by the paucity of available 13C ture changes across shallow to upper-mesophotic reefs and gut content data for organisms inhabiting in the Mesoamerican Barrier Reef. Caribbean Plos ONE depths deeper than 60  m. Therefore our carbon 11(6):e0156641Andradi-Brown DA, Grey R, Hendrix A, Hitchner D, Hunt comparisons between deep and shallow popula- CL et al (2017a) Depth-dependent effects of culling—do tions are the comparison of the δ13C values, rela- mesophotic lionfish populations undermine current man- tive to each other, as these are the sole data avail- agement? R Soc Open Sci 4:170027 able at the time of this publication. We suggest that Andradi-Brown DA, Vermeij MJA, Slattery A, Lesser M, Beja-rano I et al (2017b) Large-scale invasion of western Atlan- future studies to identify the 13C signatures at this tic mesophotic reefs by lionfish potentially undermines or similar sites and on gut content analysis be con- culling-based management. Biol Invasions 19:939–954 ducted to confirm patterns in resource partition- Aragonés Fred, C (2022) Stable Isotope Analyses (δ13C and ing for lionfish at deep versus shallow reefs.. The δ15N) of the invasive lionfish, Pterois volitans, and native groupers, red hind, Epinephelus guttatus and coney, difference in size ranges of fish between deep and Cephalopholis fulva in Puerto Rico. MS Thesis. Univer- shallow reefs may have influenced these findings. sity of Puerto Rico – Mayagüez Campus, 52 pg. Although we did analyze a size-standardized sub- Asher J, Williams ID, Harvey ES (2017) Mesophotic depth set of fish (210–270 mm TL) to control for any size gradients impact reef fish assemblage composition and functional group partitioning in the main Hawaiian difference, it is unknown whether other factors (eg. Islands. Front Mar Sci 4:98 size at age) contributed to our findings. Future stud- Baldwin CC, Tornabene L, Robertson DR (2018) Below the ies may be done to incorporate age and reproductive mesophotic. Sci Rep 8:4920 stage across reef sites and fish size to confirm that Ballew NG, Bacheler NM, Kellison GT, Schueller AM (2016) Invasive lionfish reduce native fish abundance on a differences observed in feeding behavior or trophic regional scale. Sci Rep 6:32169 position were not due to differences in size at age Balzani P, Vizzini S, Frizzi F, Masoni A, Lessard J, Bernas- or reproductive stage. Further, since this study only coni C, Francoeur A, Ibarra-Isassi J, Brassard F, Cherix captured lionfish from one site, it is important to D, Santini G (2021) Plasticity in the trophic niche of an invasive ant explains establishment success and long-term further examine if patterns observed here are con- coexistence. Oikos 130:691–696 sistent across other regions. Bates DM, Mächler M, Bolker BM, Walker S (2014) Package Lme4: Linear mixed-effects models using Eigen and S4. J Acknowledgements We thank the crew of Substation Cura- Stat Softw 67:69 cao for their assistance in the field. We are especially grateful Bauer H (2020) Trophic interactions and habitat quality of to Tico Christiaan and Jordy Stolk for their technical-diving invasive lionfish in the Gulf of Mexico. Master’s Theses, efforts, and to Bruce Brandt for assistance with modifications Louisiana State University, Baton Rouge, LA to the Curasub. William Ludt, Maddie Willert and Sarah Yer- Blakeway RD, Fogg AQ, Johnston MA, Rooker JR, Jones GA race provided assistance in the field. Funding was provided by (2022) Key life history attributes and removal efforts of the William W. and Dorothy T. Gilbert Ichthyology Research invasive lionfish (Pterois volitans) in the Flower Garden Fund and the U.S. branch of the Prince Albert II of Monaco Banks National Marine Sanctuary, northwestern Gulf of Foundation gift funds. This study is Ocean Heritage Founda- Mexico. Front Mar Sci 9. tion/Curacao Sea Aquarium/Substation Curacao contribution Böhm F, Joachimski MM, Lehnert H, Morgenroth G, number OHF/CSA/SC#56. Kretschmer W, Vacelet J, Dullo W-C (1996) Carbon iso- tope records from extant Caribbean and South Pacific Vol.: (0123456789) 44 Page 12 of 14 M. M. Ewing et al. sponges: Evolution of δ13C in surface water DIC. Earth Fry B (1988) Food web structure on Georges Bank from sta- Planet Sci Lett 139:291–303 ble C, N, and S isotopic compositions. Limnol Oceanogr Bradley CJ, Longenecker K, Pyle RL, Popp BN (2016) Com- 33(5):1182–1190 pound-specific isotopic analysis of amino acids reveals Gardner PG, Frazer TK, Jacoby CA, Yanong RPE (2015) dietary changes in mesophotic coral-reef fish. Mar Ecol Reproductive biology of invasive lionfish (Pterois spp.). Prog Ser 588:65–79 Front Mar Sci 2:7 Busst GMA, Britton JR (2018) Tissue-specific turnover rates of Gittings SR, Fogg AQ, Frank S, Hart J V, Clark A, Clark the nitrogen stable isotope as functions of time and growth B, Noakes SE, Fortner RL (2017) Going deep for lion- in a cyprinid fish. Hydrobiologia 805:49–60 fish: designs for two new traps for capturing lionfish in Ceia FR, Cherel Y, Pavio VH, Ramos JA (2018) Stable isotope deep water. Marine Sanctuaries Conservation Series dynamics (δ13C and δ15N) in neritic and oceanic waters ONMS17–05. U.S. Department of Commerce, National of the North Atlantic inferred from GPS-tracked Cory’s Oceanic and Atmospheric Administration, Office of shearwaters. Front Mar Sci 5:377 National Marine Sanctuaries, Silver Spring, MD. 9 pp. Claydon JAB, Calosso MC, Traiger SB (2012) Progression of Glassic HC, Guy CS, Tronstad LM, Lujan DR, Briggs invasive lionfish in seagrass, mangrove and reef habitats. MA, Albertson LK, Koel TM (2023) Invasive preda- Mar Ecol Prog Ser 448:119–129. https:// doi. org/1 0. 3354/ tor diet plasticity has implications for native fish con- meps09 534 servation and invasive species suppression. PLoS ONE Cocheret de la Morinière E, Pollux BJA, Nagelkerken I, 18:e0279099 Hemminga MA, Huiskes AHL, van der Velde G (2003) Goodbody-Gringley G, Eddy C, Pitt JM, Chequer AD, Smith Ontogenetic dietary changes of coral reef fishes in the SR (2019) Ecological drivers of invasive lionfish (Pterois mangrove–seagrass–reef continuum: stable isotopes and volitans and Pterois miles) distribution across mesophotic gut-content analysis. Mar Ecol Prog Ser 246:279–289 reefs in Bermuda. Front Mar Sci 6:258 Cure K, Benkwitt CE, Kindinger TL (2012) Comparative Green SJ, Côté IM (2014) Trait-based diet selection: prey behavior of red lionfish Pterois volitans on native Pacific behaviour and morphology predict vulnerability to preda- versus invaded Atlantic coral reefs. Mar Ecol Prog Ser tion in reef fish communities. J Anim Ecol 83:1451–1460 467:181–192 Green SJ, Akins JL, Maljković A, Côté IM (2012) Invasive Davias LA, Kornis MS, Breitburg DL (2013) Environmental lionfish drive Atlantic coral reef fish declines. PLoS ONE factors influencing δ13C and δ15N in three Chesapeake 7(3):e32596 Bay fishes. ICES J Mar Sci 71:689–702 Green SJ, Matley JK, Smith DE, Castillo B II, Akins JL et al Davidson AM, Jennions M, Nicotra AB (2021) Do invasive (2021) Broad-scale acoustic telemetry reveals long- species show higher phenotypic plasticity than native spe- distance movements and large home ranges for inva- cies and if so, is it adaptive? A meta-analysis. Ecol Letters sive lionfish on Atlantic coral reefs. Mar Ecol Prog Ser 14:419–431 673:117–134 De León R, Vane K, Chamberland VF (2013) Effectiveness of Green SJ, Akins JL, Côté IM (2011) Foraging behaviour and lionfish removal efforts in the southern Caribbean. Endan- prey consumption in the Indo-Pacific lionfish on Baha- gered Species Research 22(2):175–182 mian coral reefs. Mar Ecol Prog Ser 433:159–167 Eddy C, Pitt JM, Larkum J, Altabet MA, Bernal D (2020) Sta- Gress E, Andradi-Brown D, Woodall L, Schofield PJ (2017) ble isotope ecology of invasive lionfish (Pterois volitans Lionfish (Pterois sp.) invade the upper bathyal zone in the and P. miles) in Bermuda. Front Mar Sci 7:435 western Atlantic. PeerJ 5:83683 Edwards MA, Frazer TK, Jacoby CA (2014) Age and growth of Harms-Tuohy CA, Appeldoorn RS, Craig MT (2018) The invasive lionfish (Pterois spp.) in the Caribbean Sea, with effectiveness of small-scale lionfish removals as a man- implications for management. Bull Mar Sci 90:953–966 agement strategy: effort, impacts and the response Fogg AQ, Brown-Peterson NJ, Peterson MS (2017) Reproduc- of native prey and piscivores. Manag Biol Invasion tive life history characteristics of invasive red lionfish 9(2):149–162 (Pterois volitans) in the northern Gulf of Mexico. Bull Harris EH, Fogg AQ, Gittings SR, Ahrens RNM, Allen MS, Mar Sci 93:791–813 Patterson WF III (2020) Testing the efficacy of lion- Fogg AQ, Evans JT, Peterson MS, Brown-Peterson NJ, Hoff- fish traps in the northern Gulf of Mexico. PLoS ONE mayer ER, Ingram GW Jr (2019) Comparison of age and 15(8):e0230985. https://d oi.o rg/1 0.1 371/ journ al.p one. growth parameters of invasive red lionfish (Pterois voli- 023098 5 tans) across the northern Gulf of Mexico. Fish Bull 117:1 Hartig, F (2022). DHARMa: Residual Diagnostics for Hierar- Frade PR, Bongaerts P, Baldwin CC, Trembanis AC, Bak RP, chical (Multi-Level / Mixed) Regression Models. R pack- Vermeij MJA (2019) Bonaire and Curacao. In: Puglise age version 0.4.5. https://C RAN.R- proje ct. org/ packa ge= KA, Bridge TCL (eds) Loya Y Mesophotic Coral Ecosys- DHARMa tems. Springer, Cham, pp 149–162 Hinderstein LM, Marr JCA, Martinez FA, Dowgiallo MJ, Pug- France RL (1995) Carbon-13 enrichment in benthic compared lise KA et al (2010) Theme section on “mesophotic coral to planktonic algae: foodweb implications. Mar Ecol Prog ecosystems: characterization, ecology, and management.” Ser 124:307–312 Coral Reefs 29:247–251 Fry B (2006) Stable isotope ecology. Springer, New York Hofmann LC, Heesch S (2018) Latitudinal trends in stable Fry B, Lutes R, Northam M, Parker PL (1982) A δ13C com- isotope signatures and carbon-concentrating mecha- parison of food webs in Caribbean seagrass meadows and nisms of northeast Atlantic rhodoliths. Biogeosciences coral reefs. Aquat Bot 14:389–398 15:6139–6149 Vol:. (1234567890) Analysis of bulk stable isotopes and trophic positions of invasive lionfish (Pterois volitans)… Page 13 of 14 44 Johnson EG, Swenarton MK (2016) Age, growth and popula- the 66th Gulf and Caribbean Fisheries Institute. Corpus tion structure of invasive lionfish (Pterois volitans/miles) Cristi, TX, pp. 187–190. in northeast Florida using a length-based, age-structured Pitt JM, Trott TM (2014) Trapping lionfish in Bermuda, Part II: population model. PeerJ 4:e2730 Lessons learned to date. In: Proceedings of the 67th Gulf Kahng SE, Garcia-Sais JR, Spalding HL, Brokovich E, Wagner and Caribbean Fisheries Institute, Christ Church, Barba- D et  al (2010) Community ecology of mesophotic coral dos, pp. 221–224. reef ecosystems. Coral Reefs 29:255–275 Post DM (2002) Using stable isotopes to estimate trophic Kjeldgaard MK, Hewlett JA, Eubanks MD (2021) Widespread position: models, methods, and assumptions. Ecology variation in stable isotope trophic position estimates: pat- 83:703–718 terns, causes, and potential consequences. Ecol Monogr Quintana Y, Keppeler FW, Winemiller KO (2023) Does inva- 91(3):e01451 sion by armored catfish shift trophic ecology of native Lamb K, Swart PK (2008) The carbon and nitrogen isotopic fishes? Evid Stable Isotope Anal Ecol 104(5):e4024 values of particulate organic material from the Florida Randall JE (1996) Caribbean reef fishes. Third Edition - Keys: a temporal and spatial study. Coral Reefs 27:351– revised and enlarged. T.F.H. Publications, Inc. Ltd., Hong 362. https://d oi. org/ 10. 1007/s 00338-0 07- 0336-5 Kong. 3nd ed. 368 p. Lesser MP, Slattery M (2011) Phase shift to algal dominated Ritger AL, Curtis AN, Chen CY (2018) Bioaccumulation of communities at mesophotic depths associated with lion- mercury and other metal contaminants in invasive lionfish fish (Pterois volitans) invasion on a Bahamian coral reef. (Pterois volitans/miles) from Curaçao. Mar Pollut Bull Biol Invasions 13:1855–1868 131(Pt A):38–44 Lesser MP, Slattery M, Macartney KJ (2022) Using stable iso- Ritger AL, Fountain CT, Bourne K, Martín-Fernández JA, tope analyses to assess the trophic ecology of Scleractin- Pierotti MER (2020) Diet choice in a generalist preda- ian Corals. J Geophys Res C Oceans 3:527–546 tor, the invasive lionfish (Pterois volitans/miles). Jour Exp Maljković A, Van Leeuwen TE, Cove SN (2008) Predation on Mar Biol Ecol 524:151311 the invasive red lionfish, Pterois volitans (Pisces: Scor- Robertson DR, Tornabene L, Lardizabal CC, Baldwin CC paenidae), by native groupers in the Bahamas. Coral Reefs (2022) Submersibles greatly enhance research on the 27(3):52 diversity of deep-reef fishes in the Greater Caribbean. Malpica-Cruz L, Green SJ, Côté IM (2019) Temporal and Front Mar Sci 8:800250 ontogenetic changes in the trophic signature of an invasive Ruiz-Navarro A, Jackson MC, Almeida D, Britton JR (2020) marine predator. Hydrobiologia 839:71–86 Influence of nutrient enrichment on the growth, recruit- Morris J (2012) Invasive Lionfish: A Guide to Control and ment and trophic ecology of a highly invasive freshwater Management. Gulf and Caribbean Fisheries Institute Spe- fish. Aquat Ecol 54:1029–1039 cial Publication, Marathon Sanjuan-Muñoz A, Bustons-Montes D, Polo-Silva CJ, Henao- Muñoz RC, Currin CA, Whitfield PE (2011) Diet of inva- Castro A, Marrugo M, Delgado-Huertas A, Vinyoles- sive lionfish on hard bottom reefs of the Southeast USA: Cartanya D, Acero AP (2022) Biología y ecología del Insights from stomach contents and stable isotopes. Mar pez león (Pterois volitans) en el Parque Nacional Natural Ecol Prog Ser 432:181–193 Corales de Profundidad, Caribe colombiano. Boletín de Muñoz RC (2017) Evidence of natural predation on invasive Investigaciones Marinas y Costeras 51(1): 75–98. https:// lionfish, Pterois spp. By the spotted moray eel, Gymno- doi.o rg/1 0. 25268/ bimc. invem ar. 2021. 51.1.1 087 thorax moringa. Bull Mar Sci 93(3):52 Semmens BX, Buhle ER, Salomon AK, Pattengill-Semmens O’Farrell S, Bearhop S, McGill RAR, Dahlgren CP (2014) CV (2004) A hotspot of non-native marine fishes: evi- Habitat and body size effects on the isotopic niche space dence for the aquarium trade as an invasion pathway. Mar of invasive lionfish and endangered Nassau grouper. Eco- Ecol Prog Ser 266:239–244 sphere 5(10):123 Slattery M, Lesser MP, Brazeau D, Stokes MD, Leichter JJ Pimiento C, Nifrong J, Hunter ME, Monaco E, Silliman BR (2011) Connectivity and stability of mesophotic coral (2015) Habitat use patterns of the invasive red lionfish reefs. J Exp Mar Biol Ecol 408:32–41. https://d oi. org/ 10. Pterois volitans: a comparison between mangrove and reef 1016/j. jembe.2 011. 07. 024 systems in San Salvador, Bahamas. Mar Ecol 36:28–37. Soares MO, Feitosa CV, Garcia TM, Cottens KF, Vinicius https:// doi.o rg/1 0. 1111/ maec.1 2114 B, Paiva SV, Duarte OdS, Gurjão LM, Silva GDV, Maia Pinheiro HT, Goodbody-Gringley G, Jessup ME, Shepher B, RC, Previatto DM, Carneiro PBM, Cunha E (2022) Lion- Chequer AD, Rocha LA (2016) Upper and lower meso- fish on the loose: Pterois invade shallow habitats in the photic coral reef fish communities evaluated by under- tropical southwestern Atlantic. Front Mar Sci 9:956848. water visual censuses in two Caribbean locations. Coral https:// doi.o rg/1 0.3 389/f mars. 2022.9 56848 Reefs 35:139–151 Søreide JE, Tamelander T, Hop H, Hobson KA (2006) Sam- Pinheiro HT, MacDonald C, Quimbayo JP, Shepherd B, Phelps ple preparation effects on stable C and N isotope values: a TA, Loss AC, Texeira JB, Rocha LA (2023) Assembly comparison of methods in Arctic marine food web studies. rules of coral reef fish communities along the depth gra- Mar Ecol Prog Ser 328:17–28 dient. Curr Biol 33:1421–1430. https:// doi.o rg/ 10. 1016/j. Stefanoudis PV, Gress E, Pitt JM, Smith SR, Kincaid T et  al cub. 2023. 02. 040 (2019) Depth-dependent structuring of reef fish assem- Pitt JM, Trott TM (2013) Efforts to develop a lionfish-spe- blages from the shallows to the rariphotic zone. Front Mar cific trap for use in Bermuda waters. In: Proceedings of Sci 6:307 Vol.: (0123456789) 44 Page 14 of 14 M. M. Ewing et al. Stuthmann LE, Castellanos-Galindo GA (2020) Trophic posi- Ooms J, Robinson D, Seidel DP, Spinu V, Takahashi K, tion and isotopic niche of mangrove fish assemblages Vaughan D, Wilke C, Woo K, Yutani H (2019) Welcome at both sides of the Isthmus of Panama. Bull Mar Sci to the tidyverse. J Open Source Softw 4(43):1686. https:// 96(3):449–468 doi. org/1 0.2 1105/j oss. 01686 Tornabene L, Baldwin CC (2017) A new mesophotic goby, Zhu Y, Newman P, Reid WD, Polunin NV (2019) Fish stable Palatogobius incendius (Teleostei: Gobiidae), and the first isotope community structure of a Bahamian coral reef. record of invasive lionfish preying on undescribed biodi- Mar Biol 166:1–14 versity. PLoS ONE 12(5):e0177179 Vander Zanden MJ, Clayton MK, Moody EK, Solomon CT, Publisher’s Note Springer Nature remains neutral with regard Weidel BC (2015) Stable isotope turnover and half- to jurisdictional claims in published maps and institutional life in animal tissues: a literature synthesis. PLoS ONE affiliations. 10(1):e0116182 Whitfield PE, Gardner T, Vives SP, Gilligan MR, Courtenay WR Jr (2002) Biological invasion of the Indo-Pacific Springer Nature or its licensor (e.g. a society or other partner) lionfish Pterois volitans along the Atlantic coast of North holds exclusive rights to this article under a publishing America. Mar Ecol Prog Ser 235:289–297 agreement with the author(s) or other rightsholder(s); author Wickham H, Averick M, Bryan J, Chang W, McGowan LD, self-archiving of the accepted manuscript version of this article François R, Grolemund G, Hayes A, Henry L, Hester J, is solely governed by the terms of such publishing agreement Kuhn M, Pedersen TL, Miller E, Bache SM, Müller K, and applicable law. Vol:. (1234567890)