The authors have declared that no competing interests exist.
Conceived and designed the experiments: MWJ SJP. Performed the experiments: MWJ SJP. Analyzed the data: MWJ SJP. Contributed reagents/materials/analysis tools: MWJ SJP. Wrote the paper: MWJ SJP.
The Indo-pacific panther grouper
Marine invasive species are much less common than their freshwater counterparts; however, sightings of non-native species in Atlantic waters have been well documented by the United States Geological Survey Nonindigenous Species (USGS NAS) database
The panther grouper (
Records from the USGS NAS indicating locations of panther grouper captures or sightings.
The panther grouper is an Indo-pacific predatory fish species found in lagoons, hard bottom habitats, and seaward well-developed coral reefs, in depths up to 40 m
The panther grouper shows many potential invasive characteristics and shares ecomorphology and a breeding strategy similar to the efficacious Atlantic invasive species, the lionfish. The panther grouper and lionfish are also both Indo-pacific apex reef predators
Life-History Characteristic | Panther Grouper | Lionfish |
apex predator – teleosts, crustaceans | apex predator – teleosts, crustaceans | |
70 cm, 7.0 kg | 10 cm, 300–400 g |
|
19 years | up to 30 years in captivity |
|
coloration | coloration, venomous | |
likely moderate | high | |
16°C | 10°C | |
protogynous hermaprodite | monogametic | |
18 months | 12 months | |
floating, broadcast | floating, contained in a mucous sac | |
pelagic | pelagic | |
0.2 to 1.2 million | >2 million annually |
|
year round with a peak October – January | year round |
|
up to 4 | up to 7.5 |
|
40 days | 20 to 35 days |
Comparison of life history and reproductive traits of panther grouper and lionfish.
A significant dissimilarity between lionfish and the panther grouper are the morphological differences that lionfish exhibit from native Atlantic teleosts. Lionfish morphology is completely unique with expansive, venomous striped pectoral and dorsal fin rays – unlike any extant species in the Atlantic
This paper presents a suite of simulated scenarios that describe the potential spread of the panther grouper in the Atlantic, should a breeding population become established, based on the ISM previously utilized studying lionfish
Cellular automata (CA) models, such as the ISM, consist of four elements;
The initial version of the ISM excluded the temporal aspect of an invasion, instead focusing on the chronology of spatial occurrences. To include periodicity in the ISM, the model now integrates the timing of species life-history components, which are critical to predicting the progression of an invasion
Parameter Name | Value | Rationale | Source |
60 | |||
6 Arc Minutes | 10 fold increase in granularity from previous lionfish study | ||
16° C – 32.820°C | based on temperature extremes in their documented native range | ||
.02 | parameter does not largely influence initial distribution for a current-dispersed species | ||
.10 – 99.981 μgL−1 | chlorophyll concentrations on two sections of the Great Barrier Reef, a native habitat for PG, indicated a mean concentration of 0.2μgL−1 and 0.54 μgL−1 – lower limit of 0.10 μgL−1 based on comparative concentrations in its native Australia and similar concentrations in the Atlantic | ||
.02 | parameter does not largely influence initial distribution for a current-dispersed species | ||
1–40 M | known to inhabit lagoon type areas and shallow reefs to a depth of 40 meters; parameter does not largely influence initial distribution for a current-dispersed species | ||
.02 | parameter does not largely influence initial distribution for a current-dispersed species | ||
.90 | the most influential parameter to the spread of similar invasive lionfish | ||
40 days | durations documented by |
||
0.2 d −1 | In marine teleosts, larval |
||
18 months | documented in cultured conditions at approximately 18 months and 15.5 cm length | ||
.26 y −1 | based on two locations in Australia, the Great Barrier Reef and Torres Strait. | ||
15,000 | fertilization rates are estimated at 0–90% and hatching rates usually exceed 30% – estimated viable propagules per cycle (25%×200,000 (fertilization rate) ×30% (hatching rate)) based on natural reproduction, as opposed to controlled breeding situations in ideal circumstances | ||
20 | 20 – resulting in a larvae/kernel ratio of 0.0013 (approximately 750 larvae per kernel) | ||
January/December | natural reproduction has been documented year round | ||
30 days | breeding occurs on a monthly cycle around the full moon; conservatively, value has been set to one breeding session monthly | ||
January | arbitrary starting month |
Input values for all parameters considered in the ISM, including their source.
As documented in
The initial version of the ISM (using the default parameter set) examines a geographic area encompassing the western Atlantic Ocean, Caribbean Sea, and Gulf of Mexico from 45° to 5° N latitude and −100° to −50° W longitude, which corresponds to the approximate geographic extent of the lionfish invasion. In the enhanced ISM, the eastern Pacific is included for an area encompassing 50° N to 40° S latitude and −140° to −20° W longitude. For the purposes of this study, the panther grouper is presumed contained for the simulation duration to the Atlantic Ocean, Gulf of Mexico, and Caribbean.
The quantity and quality of eggs and larvae released are critical components when determining fecundity of a species
To identify potential settling locations (“hot spots”) regardless of origination in southern Florida, a composite simulation was created by selecting 1,411 points representing all grid locations (at a scale of 6 arc minutes or roughly 10 km) within 1° of USGS NAS panther grouper records (excluding the Gulf of Mexico record) and a water depth limit of 40 m. From these locations, a random number generator was used to select 1,000 points. One simulation was then created for each position, eliminating bias as to the exact introduction point. Two detailed test case scenarios were also chosen for closer examination to demonstrate differences between a south Florida (CSBC) and Florida Keys (CSFK) breeding population. A simulation duration of 78 months was deemed sufficient to illustrate the initial spread pattern and provide settling location guidance for all simulations.
In
Larvae survivorship in fish population models is inherently sensitive to small changes in the larval mortality rate, resulting in a pronounced effect on larval recruitment
Settling sites from the composite simulation study were summed per location and projected on a map (
Settlement rates of adult breeding populations for panther grouper on a ‘hot’ (red) to ‘cold’ (blue) scale using Jenks' natural breaks as class divisions (a method that reduces inter-class variance and maximizes variance between distinct classes) for CSFK (A), CSBC (B) and composite study (C) simulations for a duration of 78 months. Focus areas for early detection are indicated for the Jupiter Florida/Vero Beach (red), Cape Hatteras Tropical Limit/Myrtle Beach (orange), and Florida Keys/Ten Thousand Islands (green) locations.
CSFK assumes a breeding population of panther grouper in the Florida Keys. The USGS NAS records presently indicate a large specimen recently captured in the Florida Keys and the coordinates of 24.583° N and −81.217° W were chosen as an initial breeding population location. CSFK agrees with the composite simulation regarding settling points of larvae in the initial stages of an invasion. In this scenario, most larvae are transported east and north on the Gulf Stream current, eventually settling in two primary locations; 1) just south of the CHTL, and 2) north and west of Jupiter Florida near Vero Beach (
Due to the Florida Keys origination, CSFK also indicates a potential spread into the Ten Thousand Islands area off the tip of south Florida, where all parameters are within range for settling to occur. By year four, breeding populations exist at the CHTL, Jupiter Florida/Vero Beach, and Florida Keys/Ten Thousand Islands locations according to the model.
CSBC assumes a breeding population of panther grouper off the coast of south Florida in Broward County. The coordinates of 26.217° N and −80.083° W were chosen as the point of establishment for the initial breeding population as the USGS NAS indicates several records near this location. CSBC indicates that two of the same locations (south of the CHTL and Jupiter Florida/Vero Beach) have potential as settling points of larvae in the initial stages of an invasion (
The ISM indicated initial settling of larvae (non-breeding populations) 6–9 months after establishment of a breeding population in both CSFK and CSBC. The model predicts breeding populations of panther grouper would develop first in the northernmost CHTL settling point (month 20–22), followed secondly by Jupiter Florida/Vero Beach (month 28–31), and lastly, for CSFK, the Florida Keys/Ten Thousand Islands location (month 37) (
Map indicating the sequence and relative month [sequence(month)] for the first 10 steps for a Broward County origin (A) and a 12 steps for a Florida Keys origin (B).
It has been shown that coral reefs of the Florida Keys and south Florida show weaker connectivity to Bahamian reefs than would be expected based on distance alone, and are rather more closely associated with the upstream Mesoamerican Barrier Reef
To validate the ISM, 20 identical simulations were created for CSFK and CSBC using the parameter input values for this study (
SRCC calculation with a ρ = 0.80 for CSBC when comparing 20 individual model runs (y-axis) to the RM (A), and ρ = 0.49 when comparing the RM (y-axis) to Ho (B). SRCC with a ρ = 0.67 for CSFK when comparing 20 individual model runs (y-axis) to the RM (C), and ρ = 0.49 when comparing the RM (y-axis) to Ho (D). X-axis indicates the sequential order of establishment for the RM, and the y-axis indicates the order of establishment for each comparative simulation. Perfect correlation (SRCC of 1.0) is indicated by a point lying precisely on the diagonal from bottom-left to top-right.
When plotted on a map and summed by location, the results for each variation of
Settlement rates of adult breeding populations for panther grouper on a ‘hot’ (red) to ‘cold’ (blue) scale using Jenks' natural breaks as class divisions. CSFK with a larval mortality rate of 0.22 d −1 (A), 0.18 d −1 (B). CSBC with a larval mortality rate of 0.22 d −1 (C), 0.18 d −1 (D).
The ISM indicates several key locations which present a high likelihood for retention of larvae and the eventual development of breeding populations of panther grouper, given the constraints of the model. Common to all case studies, just south of the CHTL (a northernmost record of 34.817° N latitude was recorded in the model) near Myrtle Beach and north and west of Jupiter Florida centered close to Vero Beach, are high-risk areas. The Florida Keys/Ten Thousand Islands location is seen as lower risk with lower settlement rates. Based on the composite study, the highest likelihood of establishment of a breeding population of panther grouper, regardless of introduction point, is north of Jupiter Florida, centered near Vero Beach. Our modeling outputs suggest that Vero Beach is to be the highest priority for monitoring efforts, followed by the Myrtle Beach/CHTL and the Ten Thousand Islands area.
Based on the two individual case studies, a Florida Keys origin is most precocious as this would provide a conduit to the west coast of Florida and the Gulf of Mexico. If the invasion scenario follows the pattern documented by lionfish, this Florida Keys origin would short-circuit the progression stage sequence, allowing ingress of the panther grouper into the Gulf of Mexico much sooner than occurred with lionfish
As anticipated, and in accordance with studies by
Numerical models examining complex systems, such as the marine environment, suffer from uncertainty arising from the inevitable lack of a full understanding of the system. Approximation or underlying data errors or fundamental flaws in the model itself can introduce bias and undermine the model results. Acknowledging these limitations, this study aims to reduce inherent uncertainty within the model by eliminating bias when selecting origination locations and instead employs random locations within the study area. Additionally, the two case studies presented are validated against Ho and tested for precision using a standard metric, the SRCC. Sensitivity analysis is also performed to test model robustness to variances in larval mortality.
While the panther grouper has been found in the Florida Keys and Broward County, this does not confirm breeding populations. In both case studies, it is assumed that a breeding population persists at the origins and the lag period that is sometimes present with exotic invasions is ignored
This paper presents a rapid-response modelling study of the potential establishment and spread of the panther grouper in the western Atlantic in an effort to direct early detection and eradication efforts before the species has gained traction. This study identifies three areas of concern for potential establishment of the species in the western Atlantic, should a founder population occur in any location in the area examined (extending from approximately 29° N to 24° N on the Atlantic side of Florida in waters <40 m). These locations include; 1) just south of the Cape Hatteras Tropical Limit/Myrtle Beach, 2) north of Jupiter Florida/Vero Beach, and 3) the Florida Keys/Ten Thousand Islands location. As breeding populations are not yet thought to occur, it is suggested that these three locations should be high priority for monitoring and early detection efforts to prevent the proliferation of the panther grouper in the Atlantic. This study gives an early indication of potential hot spots of establishment to guide detection, containment, and perhaps eradication efforts.
We thank the USGS for compiling panther grouper records and making them publically available as well as support from the National Coral Reef Institute (NCRI) at Nova Southeastern University for this study. We also thank Dr. Patricia Illoldi-Rangel and the additional anonymous reviewers whose guidance was invaluable. This is NCRI contribution 153.