What Doesn't Kill You Makes You Wary? Effect of Repeated Culling on the Behaviour of an Invasive Predator

As a result of being hunted, animals often alter their behaviour in ways that make future encounters with predators less likely. When hunting is carried out for conservation, for example to control invasive species, these behavioural changes can inadvertently impede the success of future efforts. We examined the effects of repeated culling by spearing on the behaviour of invasive predatory lionfish (Pterois volitans/miles) on Bahamian coral reef patches. We compared the extent of concealment and activity levels of lionfish at dawn and midday on 16 coral reef patches off Eleuthera, The Bahamas. Eight of the patches had been subjected to regular daytime removals of lionfish by spearing for two years. We also estimated the distance at which lionfish became alert to slowly approaching divers on culled and unculled reef patches. Lionfish on culled reefs were less active and hid deeper within the reef during the day than lionfish on patches where no culling had occurred. There were no differences at dawn when removals do not take place. Lionfish on culled reefs also adopted an alert posture at a greater distance from divers than lionfish on unculled reefs. More crepuscular activity likely leads to greater encounter rates by lionfish with more native fish species because the abundance of reef fish outside of shelters typically peaks at dawn and dusk. Hiding deeper within the reef could also make remaining lionfish less likely to be encountered and more difficult to catch by spearfishers during culling efforts. Shifts in the behaviour of hunted invasive animals might be common and they have implications both for the impact of invasive species and for the design and success of invasive control programs.     

Differential effects of native vs. invasive predators on a common Caribbean reef fish

Predators may have consumptive (lethal) and non-consumptive (sub-lethal) effects on prey. Non-consumptive effects include altered behavior and reduced growth and fecundity. Native prey may not recognize non-native predators as a threat, and therefore may suffer pronounced effects. Additionally, non-native predators may elicit different behavioral responses from prey compared to native predators. Theory predicts that consumptive effects should be greater for non-native predators (due to prey naiveté), and non-consumptive effects should be greater for native predators (due to predator recognition). To test these hypotheses, I monitored bicolor damselfish (Stegastes partitus) in the presence of invasive predatory Pacific lionfish (Pterois spp.), a native predator (graysby, Cephalopholis cruentata), and an egg predator (bluehead wrasse, Thalassoma bifasciatum). Body size and location of lionfish and graysby were monitored on reefs in the Bahamas. Bicolor fecundity was measured as the number and size of egg-masses that individual fish laid. Bicolor fecundity was negatively correlated with lionfish density but not graysby or bluehead density. Neither predator had a detectable effect on bicolor body size, but lionfish density was negatively correlated with the size of mature adult damselfish. I observed behavioral responses of bicolors to the two piscivores, to bluehead wrasse, and to two herbivorous fishes (Acanthurus coeruleus, Scarus spp.) as non-aggressive controls. Bicolors changed behavior (feeding and aggression) in the presence of all native fishes, but not in the presence of lionfish. Thus, differential effects exist between native and non-native predators, and invasive lionfish pose a non-consumptive threat to bicolor damselfish via reduced growth and fecundity.     

Effects of invasive Pacific red lionfish Pterois volitans versus a native predator on Bahamian coral-reef fish communities

The recent irruption of Pacific red lionfish (Pterois volitans) on Caribbean and Atlantic coral reefs could prove to be one of the most damaging marine invasions to date. Invasive lionfish are reaching densities much higher than those reported from their native range, and they have a strong negative effect on the recruitment and abundance of a broad diversity of native coral-reef fishes. Otherwise, little is known about how lionfish affect native coral-reef communities, especially compared to ecologically similar native predators. A controlled field experiment conducted on small patch-reefs in the Bahamas over an 8-week-period demonstrated that (1) lionfish caused a reduction in the abundance of small native coral-reef fishes that was 2.5?±?0.5 times (mean?±?SEM) greater than that caused by a similarly sized native piscivore, the coney grouper Cephalopholis fulva (93.7 vs. 36.3?% reduction); (2) lionfish caused a reduction in the species richness of small coral-reef fishes (loss of 4.6?±?1.6 species), whereas the native piscivore did not have a significant effect on prey richness; (3) the greatest effects on the reef-fish community, in terms of both abundance and richness, occurred when both native and invasive predators were present; and (4) lionfish grew significantly faster (>6 times) than the native predator under the same field conditions. These results suggest that invasive lionfish have stronger ecological effects than similarly sized native piscivores, and may pose a substantial threat to native coral-reef fish communities.     

Indo-Pacific lionfish are larger and more abundant on invaded reefs: a comparison of Kenyan and Bahamian lionfish populations

The invasion by Indo-Pacific lionfish (Pterois volitans and P. miles) of the western Atlantic, Caribbean and Gulf of Mexico is emerging as a major threat to coral reef communities across the region. Comparing native and introduced populations of invasive species can reveal shifts in ecology and behaviour that can accompany successful invasions. Using standardized field surveys replicated at multiple sites in Kenya and the Bahamas, we present the first direct comparisons of lionfish density, body size, biomass and behaviour between native and invaded coral reefs. We found that lionfish occur at higher densities with larger body sizes and total biomass on invaded Bahamian coral reefs than the ecologically equivalent species (P. miles) does on native Kenyan reefs. However, the combined average density of the five lionfish species (Pterois miles, P. antennata, P. radiata, Dendrochirus brachypterus and D. zebra) on Kenyan reefs was similar to the density of invasive lionfish in the Bahamas. Understanding the ecological processes that drive these differences can help inform the management and control of invasive lionfish.     

Consumptive and non-consumptive effects of an invasive marine predator on native coral-reef herbivores

Invasive predators typically have larger effects on native prey populations than native predators, yet the potential roles of their consumptive versus non-consumptive effects (CEs vs. NCEs) in structuring invaded systems remains unclear. Invasive lionfish (Pterois volitans) may have ecosystem-level effects by altering native fish grazing on benthic algae that could otherwise displace corals. Lionfish could reduce grazing by decreasing the abundance of herbivorous fishes (CEs), and/or the predation risk posed by lionfish could alter grazing behavior of fishes (NCEs). To test for these CEs, we manipulated lionfish densities on large reefs in The Bahamas and surveyed fish populations throughout June 2009–2011. In July 2011, NCEs of lionfish were measured by observing fish grazing behavior on algal-covered substrata placed in microhabitats varying in lionfish presence at different spatial scales, and quantifying any resulting algal loss. Lionfish reduced small herbivorous fish density by the end of the 2010 summer recruitment season. Grazing by small and large fishes was reduced on high-lionfish-density reefs, and small fish grazing further decreased when in the immediate presence of lionfish within-reefs. Lionfish had a negative indirect effect on algal loss, with 66–80 % less algae removed from substrata in high-lionfish-density reefs. Parrotfishes were likely driving the response of herbivorous fishes to both CEs and NCEs of lionfish. These results demonstrate the importance of considering NCEs in addition to CEs of invasive predators when assessing the effects of invasions.     

DNA barcoding significantly improves resolution of invasive lionfish diet in the Northern Gulf of Mexico

Invasive Indo-Pacific red lionfish (Pterois volitans) have become well-established residents within reef communities across the western Atlantic Ocean where they pose substantial threats to native fish communities and reef ecosystems. Species-specific identification of prey is necessary to elucidate predator–prey interactions, but can be challenging with traditional visual identification methods given prey are often highly digested, thus not identifiable visually. To supplement visual diet analysis of lionfish (n = 934) sampled in the northern Gulf of Mexico, we applied DNA barcoding to identify otherwise unidentifiable fish prey (n = 696) via amplification of the cytochrome c oxidase subunit I (COI) of the mitochondrial genome. Barcoding nearly doubled the number of identifiable fish prey, thereby greatly enhancing our ability to describe lionfish diet. Thirty-three fish prey species were identified via barcoding, twenty-four of which were not previously detected by traditional methods. Some exploited reef fishes were newly reported (e.g., red snapper, Lutjanus campechanus) or found to constitute higher proportions of lionfish diet than previously reported (e.g., vermilion snapper, Rhomboplites aurorubens). Barcoding added a significant amount of new dietary information, and we observed the highest prey diversity reported to date for invasive lionfish. Potential cannibalism on juveniles also was identified via DNA barcoding, with the highest incidence corresponding to high lionfish densities, thus suggesting density-dependent prey demand may have driven this response. Overall, DNA barcoding greatly enhanced our ability to describe invasive lionfish diet in this study, suggesting that even studies with relatively large diet sample sizes could benefit from barcoding analysis.     

Native Predators Do Not Influence Invasion Success of Pacific Lionfish on Caribbean Reefs

Biotic resistance, the process by which new colonists are excluded from a community by predation from and/or competition with resident species, can prevent or limit species invasions. We examined whether biotic resistance by native predators on Caribbean coral reefs has influenced the invasion success of red lionfishes (Pterois volitans and Pterois miles), piscivores from the Indo-Pacific. Specifically, we surveyed the abundance (density and biomass) of lionfish and native predatory fishes that could interact with lionfish (either through predation or competition) on 71 reefs in three biogeographic regions of the Caribbean. We recorded protection status of the reefs, and abiotic variables including depth, habitat type, and wind/wave exposure at each site. We found no relationship between the density or biomass of lionfish and that of native predators. However, lionfish densities were significantly lower on windward sites, potentially because of habitat preferences, and in marine protected areas, most likely because of ongoing removal efforts by reserve managers. Our results suggest that interactions with native predators do not influence the colonization or post-establishment population density of invasive lionfish on Caribbean reefs.     

Environmental and Biotic Correlates to Lionfish Invasion Success in Bahamian Coral Reefs

Lionfish (Pterois volitans), venomous predators from the Indo-Pacific, are recent invaders of the Caribbean Basin and southeastern coast of North America. Quantification of invasive lionfish abundances, along with potentially important physical and biological environmental characteristics, permitted inferences about the invasion process of reefs on the island of San Salvador in the Bahamas. Environmental wave-exposure had a large influence on lionfish abundance, which was more than 20 and 120 times greater for density and biomass respectively at sheltered sites as compared with wave-exposed environments. Our measurements of topographic complexity of the reefs revealed that lionfish abundance was not driven by habitat rugosity. Lionfish abundance was not negatively affected by the abundance of large native predators (or large native groupers) and was also unrelated to the abundance of medium prey fishes (total length of 5–10 cm). These relationships suggest that (1) higher-energy environments may impose intrinsic resistance against lionfish invasion, (2) habitat complexity may not facilitate the lionfish invasion process, (3) predation or competition by native fishes may not provide biotic resistance against lionfish invasion, and (4) abundant prey fish might not facilitate lionfish invasion success. The relatively low biomass of large grouper on this island could explain our failure to detect suppression of lionfish abundance and we encourage continuing the preservation and restoration of potential lionfish predators in the Caribbean. In addition, energetic environments might exert direct or indirect resistance to the lionfish proliferation, providing native fish populations with essential refuges.     

Parasite-mediated enemy release and low biotic resistance may facilitate invasion of Atlantic coral reefs by Pacific red lionfish (<Emphasis Type="Italic">Pterois volitans</Emphasis>)

Successful invasions are largely explained by some combination of enemy release, where the invader escapes its natural enemies from its native range, and low biotic resistance, where native species in the introduced range fail to control the invader. We examined the extent to which parasites may mediate both release and resistance in the introduction of Pacific red lionfish (Pterois volitans) to Atlantic coral reefs. We found that fewer lionfish were parasitized at two regions in their introduced Atlantic range (The Bahamas and the Cayman Islands) than at two regions in their native Pacific range (the Northern Marianas Islands and the Philippines). This pattern was largely driven by relatively high infection rates of lionfish by didymozoan fluke worms and parasitic copepods (which may be host-specific to Pterois lionfishes) in the Marianas and the Philippines, respectively. When compared with sympatric, native fishes in the Atlantic, invasive lionfish were at least 18 times less likely to host a parasite in The Bahamas and at least 40 times less likely to host a parasite in the Cayman Islands. We found no indication that lionfish introduced Pacific parasites into the Atlantic. In conjunction with demographic signs of enemy release such as increased density, fish size, and growth of invasive lionfish, it is possible that escape from parasites may have contributed to the success of lionfish. This is especially true if future studies reveal that such a loss of parasites has led to more energy available for lionfish growth, reproduction, and/or immunity.     

Genotyping confirms significant cannibalism in northern Gulf of Mexico invasive red lionfish, <Emphasis Type="Italic">Pterois volitans</Emphasis>

DNA barcoding is used in a variety of ecological applications to identify organisms, including partially digested prey items from diet samples. That particular application can enhance the ability to characterize diet and predator–prey dynamics but is problematic when genetic sequences of prey match those of consumer species (i.e., self-DNA). Such a result may indicate cannibalism, but false positives can result from contamination of degraded prey samples with consumer DNA. Here, nuclear-encoded microsatellite markers were used to genotype invasive lionfish, Pterois volitans, consumers and their prey (n?=?80 pairs) previously barcoded as lionfish. Cannibalism was confirmed when samples exhibited two or more different alleles between lionfish and prey DNA across multiple microsatellite loci. This occurred in 26.2% of all samples and in 42% of samples for which the data were considered conclusive. These estimates should be considered conservative given rigorous assignment criteria and low allelic diversity in invasive lionfish populations. The highest incidence of cannibalism corresponded to larger sized consumers from areas with high lionfish densities, suggesting cannibalism in northern Gulf of Mexico lionfish is size- and density-dependent. Cannibalism has the potential to influence population dynamics of lionfish which lack native western Atlantic predators. These results also have important implications for interpreting DNA barcoding analysis of diet in other predatory species where cannibalism may be underreported.     

Predicting impacts of lionfish (<Emphasis Type="Italic">Pterois volitans</Emphasis>) invasion in a coastal ecosystem of southern Brazil

The opportunistic feeder Pterois volitans is a voracious invader, causing large impacts in marine food-webs. We have used a Ecopath-with-Ecosim model to hypothesize an invasion by lionfish and to predict the likely impact of this potential generalist mesopredator in a subtropical food-web model. With thirty-three functional groups, the initial Ecopath model was balanced with a low biomass of lionfish (0.07 t/km²). In Ecosim, three scenarios of different vulnerability settings for the linkages between the introduced fish and its prey were tested, representing the default setting, a top-down control, and an extreme top-down control. The scenarios were tested using different assumptions on the ability of the invasive fish to change the proportions of prey consumed according to prey availability. Our model predicted that the hypothesized lionfish invasion would have a strong impact on this subtropical marine food web: (1) by reducing prey populations and, consequently, food for native predators, and; (2) by predating on key species, causing direct impacts and possibly cascading trophic effects. Reef fish were the most affected, including some groups ecologically and economically important, like lutjanids and groupers. However, some adaptations in the fishing strategy of fishermen are expected which may affect other fish groups. Stakeholders should be warned of the potential ecological and socio-economic impacts that may arise from a lionfish invasion and various strategies and policy options should be immediately developed and applied (1) to prevent the arrival and establishment of the lionfish, and; (2) to make the ecosystem more resilient to this and other possible exotic species.     

What are the characteristics of lionfish and other fishes that influence their association in diurnal refuges?

In Caribbean reefs, the lionfish Pterois volitans is an invasive species that causes severe negative ecological effects, especially as this crepuscular predator consumes very diverse prey. Lionfish are not active during the day and stay in their refuges, sharing these spaces with various other fishes. The aim of this study is to determine which fishes are associated with the lionfish in their shelters, and what characteristics of both the invasive and native species may influence and explain such coexistence between a predator and its potential prey. Through diving and snorkelling, we visited 141 lionfish refuges, mostly caves, where we observed 204 lionfish and 494 other fish from 16 native species. We recorded species and abundance, as well as lionfish size and abundance. Half of the lionfish were observed in groups and the majority were large-sized. The association with most fish species seems fortuitous, but three species, Gramma loreto, Chromis cyanea and Canthigaster rostrata, were frequently observed in association with lionfish. Numerous fish juveniles, most likely Scarus coeruleus, were also observed together with the invasive predator. The more commonly associated fishes, particularly G. loreto, are mostly associated with large-sized lionfish that were found in groups. The associated fishes are also generally found in groups. Gramma loreto is a potential cleaner of the lionfish; the reasons for the association between these fish species and the invasive lionfish may be more complex than a simple predator-prey relationship and are discussed based on their biological traits and previously reported lionfish trophic ecology and predation behaviour.     

Invasive lionfish drive Atlantic coral reef fish declines

Indo-Pacific lionfish (Pterois volitans and P. miles) have spread swiftly across the Western Atlantic, producing a marine predator invasion of unparalleled speed and magnitude. There is growing concern that lionfish will affect the structure and function of invaded marine ecosystems, however detrimental impacts on natural communities have yet to be measured. Here we document the response of native fish communities to predation by lionfish populations on nine coral reefs off New Providence Island, Bahamas. We assessed lionfish diet through stomach contents analysis, and quantified changes in fish biomass through visual surveys of lionfish and native fishes at the sites over time. Lionfish abundance increased rapidly between 2004 and 2010, by which time lionfish comprised nearly 40% of the total predator biomass in the system. The increase in lionfish abundance coincided with a 65% decline in the biomass of the lionfish's 42 Atlantic prey fishes in just two years. Without prompt action to control increasing lionfish populations, similar effects across the region may have long-term negative implications for the structure of Atlantic marine communities, as well as the societies and economies that depend on them.     

Grouper as a natural biocontrol of invasive lionfish

Lionfish (Pterois volitans/miles) have invaded the majority of the Caribbean region within five years. As voracious predators of native fishes with a broad habitat distribution, lionfish are poised to cause an unprecedented disruption to coral reef diversity and function. Controls of lionfish densities within its native range are poorly understood, but they have been recorded in the stomachs of large-bodied Caribbean groupers. Whether grouper predation of lionfish is sufficient to act as a biocontrol of the invasive species is unknown, but pest biocontrol by predatory fishes has been reported in other ecosystems. Groupers were surveyed along a chain of Bahamian reefs, including one of the region's most successful marine reserves which supports the top one percentile of Caribbean grouper biomass. Lionfish biomass exhibited a 7-fold and non-linear reduction in relation to the biomass of grouper. While Caribbean grouper appear to be a biocontrol of invasive lionfish, the overexploitation of their populations by fishers, means that their median biomass on Caribbean reefs is an order of magnitude less than in our study. Thus, chronic overfishing will probably prevent natural biocontrol of lionfishes in the Caribbean.     

Density-Dependent Growth in Invasive Lionfish (Pterois volitans)

Direct demographic density dependence is necessary for population regulation and is a central concept in ecology, yet has not been studied in many invasive species, including any invasive marine fish. The red lionfish (Pterois volitans) is an invasive predatory marine fish that is undergoing exponential population growth throughout the tropical western Atlantic. Invasive lionfish threaten coral-reef ecosystems, but there is currently no evidence of any natural population control. Therefore, a manipulative field experiment was conducted to test for density dependence in lionfish. Juvenile lionfish densities were adjusted on small reefs and several demographic rates (growth, recruitment, immigration, and loss) were measured throughout an 8-week period. Invasive lionfish exhibited direct density dependence in individual growth rates, as lionfish grew slower at higher densities throughout the study. Individual growth in length declined linearly with increasing lionfish density, while growth in mass declined exponentially with increasing density. There was no evidence, however, for density dependence in recruitment, immigration, or loss (mortality plus emigration) of invasive lionfish. The observed density-dependent growth rates may have implications for which native species are susceptible to lionfish predation, as the size and type of prey that lionfish consume is directly related to their body size. The absence of density-dependent loss, however, contrasts with many native coral-reef fish species and suggests that for the foreseeable future manual removals may be the only effective local control of this invasion.     

Herbivore functional traits and macroinvertebrate food webs have different responses to leaf chemical compounds of two macrophyte species in a tropical lake’s littoral zone

This research addressed the question of whether invertebrate food web structure varied between a native and an invasive macrophyte leaf species in the littoral zone of a tropical reservoir. We compared macroinvertebrate herbivore functional trait diversity composition with food web structure on the two macrophyte leaves, the invasive white ginger lily (Hedichium coronarium—Zingiberaceae) and the native pickerelweed (Pontederia cordata—Pontederiaceae). We predicted that the herbivore macroinvertebrate trait indices would decrease with macrophyte leaf species due to a lower resource quality with the flow-on effects in the food web structure. We calculated the number of functionally singular species (sing.sp) and herbivore functional trait richness (FRic) indices. For the macroinvertebrate food webs, we calculated the total number of trophic links (L), link density (L/S), connectance (C) and predator–prey ratios using a predator–prey matrix. We analysed the relationship between chemical traits of the macrophyte species’ leaves herbivore traits and food web indices using multivariate regression and Pearson’s correlation. Hedichium coronarium leaves had higher biomass and higher nitrogen content than the native P. cordata, which had higher phosphorus and carbohydrate content. Pontederia cordata leaves were associated with specialist macroinvertebrate species which primarily feed on biofilms (e.g. Ulmeritrus and Scirtidae) and plant leaves (e.g. Beardius). Food webs on P. cordata had lower numbers of trophic links (L), links per species (L/S) and predator–prey ratios. Connectance, which represents food web complexity, was similar between macroinvertebrate assemblages on the two leaf types. Our study suggests that chemical compounds of macrophyte leaves quality may have potential flow-on effects on food web structure.     

Shared behavioral responses and predation risk of anuran larvae and adults exposed to a novel predator

Invasive species are a regional and global threat to biological diversity. In order to evaluate an invasive predator species’ potential to harm populations of native prey species, it is critical to evaluate the behavioral responses of all life stages of the native prey species to the novel predator. The invasion of the African clawed frog (Xenopus laevis) into southern California provides an opportunity to evaluate the predation risk and behavioral responses of native amphibians. We performed predation trials and explored prey behavioral responses to determine how this invasive predator may impact native amphibian populations using Pacific chorus frogs (Pseudacris regilla) as a representative native California prey species. We found that X. laevis will readily prey upon larval and adult life stages of P. regilla. Behavior trials indicated that both larval and adult P. regilla exhibit prey response behaviors and will spatially avoid the novel invasive predator. The results suggest that native anurans may have a redundant predator response in both the larval and adult life stages, which could reduce the predatory impact of X. laevis but also drive emigration of native amphibians from invaded habitat.     

Ultimate Predators: Lionfish Have Evolved to Circumvent Prey Risk Assessment Abilities

Invasive species cause catastrophic alterations to communities worldwide by changing the trophic balance within ecosystems. Ever since their introduction in the mid 1980''s common red lionfish, Pterois volitans, are having dramatic impacts on the Caribbean ecosystem by displacing native species and disrupting food webs. Introduced lionfish capture prey at extraordinary rates, altering the composition of benthic communities. Here we demonstrate that the extraordinary success of the introduced lionfish lies in its capacity to circumvent prey risk assessment abilities as it is virtually undetectable by prey species in its native range. While experienced prey damselfish, Chromis viridis, respond with typical antipredator behaviours when exposed to a common predatory rock cod (Cephalopholis microprion) they fail to visibly react to either the scent or visual presentation of the red lionfish, and responded only to the scent (not the visual cue) of a lionfish of a different genus, Dendrochirus zebra. Experienced prey also had much higher survival when exposed to the two non-invasive predators compared to P. volitans. The cryptic nature of the red lionfish has enabled it to be destructive as a predator and a highly successful invasive species.     

Hot or not? Comparative behavioral thermoregulation,critical temperature regimes,and thermal tolerances of the invasive lionfish <Emphasis Type="Italic">Pterois</Emphasis> sp. versus native western North Atlantic reef fishes

Temperature influences the geographic range, physiology, and behavior of many ectothermic species, including the invasive lionfish Pterois sp. Thermal parameters were experimentally determined for wild-caught lionfish at different acclimation temperatures (13, 20, 25 and 32 °C). Preferences and avoidance were evaluated using a videographic shuttlebox system, while critical thermal methodology evaluated tolerance. The lionfish thermal niche was compared experimentally to two co-occurring reef fishes (graysby Cephalopholis cruentata and schoolmaster Lutjanus apodus) also acclimated to 25 °C. The physiologically optimal temperature for lionfish is likely 28.7 ± 1 °C. Lionfish behavioral thermoregulation was generally linked to acclimation history; tolerance and avoidance increased significantly at higher acclimation temperatures, but final preference did not. The tolerance polygon of lionfish shows a strong correlation between thermal limits and acclimation temperature, with the highest CT_max at 39.5 °C and the lowest CT_min at 9.5 °C. The tolerance range of invasive lionfish (24.61 °C) is narrower than those of native graysby (25.25 °C) and schoolmaster (26.87 °C), mostly because of lower thermal maxima in the former. Results show that lionfish display “acquired” thermal tolerance at higher and lower acclimation temperatures, but are no more eurythermal than other tropical fishes. Collectively, these results suggest that while lionfish range expansion in the western Atlantic is likely over the next century from rising winter sea temperatures due to climate change, the magnitude of poleward radiation of this invasive species is limited and will likely be equivalent to native tropical and subtropical fishes with similar thermal minima.