Low-temperature tolerance of African jewelfish, Hemichromis letourneuxi

One of the key environmental factors thought to restrict many tropical non-native fishes to southern Florida is their intolerance of cold water temperatures. The African jewelfish, H. letourneuxi, was introduced into the Miami area at least 30 years ago, and has recently expanded its range into the Everglades. At this point, it is not clear how far north in Florida the species may eventually spread. In this study, we determined the low-temperature tolerance of this species in a laboratory experiment and a complimentary field experiment. Because the African jewelfish has a broad salinity tolerance, we tested how the species low-temperature tolerance varied at three salinities: freshwater (0 ppt), brackish (10 ppt) and marine (35 ppt) in the laboratory. We then conducted a field experiment to examine the survivorship of individuals when caged in several common aquatic habitats during a cold snap.

Laboratory experiment

Fish were maintained in the laboratory at two temperatures (24 and 28°C) before the experiment. To determine the low-temperature tolerance of the species, we moved fish to individual containers in a large (3.3 m x 6.3 m) walk-in cooler. The temperature of the cooler was initially set at 28°C, at which time we introduced the fish that had been acclimated to that temperature. The cooler temperature was lowered to produce a 1°C decrease in water temperature each hour. At 24°C, the second batch of fish was added, and temperature continued to decrease by 1°C per hour until the cooler reached 8°C. Each hour, we checked the status of the fish and the temperature of the water.

Walk-in cooler used in temperature-tolerance trials (left). Close-up photo
of H. letourneuxi used in low-temperature tolerance experiments (right).

Fish were separated into six treatment combinations consisting of three salinities (0, 10 and 35 ppt) and two acclimation temperatures (24 and 28°C). Apart from these salinity/temperature combinations, we designated three additional treatments: control, loss of equilibrium, and death. Fish in the control treatment were housed in a nearby temperature-controlled lab (set at 25°C) throughout the experiment in containers similar to those housing the fish in the cooler. Loss of equilibrium (LOE) fish were chilled in the cooler until they exhibited an inability to right themselves. These fish were then removed from the cooler and placed in the adjacent lab at 25°C. Fish designated to the mortality treatment were left in the cooler after they lost equilibrium until they expired.

The low-temperature tolerance trial was run in December 2007. Mean loss of equilibrium temperature was 11.5°C (+ 0.9 SD), and mean expiration temperature was 9.6°C (+ 0.9 SD). Both of these endpoints were affected by acclimation temperature: fish acclimated at 24°C lost equilibrium and died at lower temperatures compared to the fish acclimated to 28°C.

Temperatures at LOE (left) and death (right) for H. letourneuxi. Thick line is the median; the box
bounds 50% of the data; whiskers denote data limits and the asterisk represents an outlier.

Salinity had little effect on low-temperature tolerance. Fish LOE was nearly equal at all three salinities; however, fish at 35 ppt died at a higher temperature than fish at 0 and 10 ppt.

Salinities at LOE (left) and death (right) for H. letourneuxi. Thick line is the median; the box
bounds 50% of the data; whiskers denote data limits and the asterisk represents an outlier.

No control fish died over the 14-day holding period at 25°C. Of the fish that lost equilibrium and were transferred to the warm (25°C) laboratory (n=37), 43% died less than 8 hours after transfer (before warming up to ambient temperature), about half (49%) lived for the following 14-day observation period and 8% (n=3) died of ich (Ichthyophthirius multifilis) infection during the 14-day observation period. Salinity had a strong effect on the survival of these fish-all fish at 35 ppt (n=12) died before warming to ambient temperature, while fish at 0 and 10 ppt were far more likely to survive the warming period (only 2 fish in each group died before warming).

Disposition of fish removed from the cooler at LOE. Percentage of fish shaded in blue are those that did not survive the first 8 hours after transfer from the cooler to the warm lab, and thus never warmed to ambient temperature. Percentage shaded in plum indicates fish that survived for several days after transfer to the cooler, but eventually succumbed to ich. Percentage in yellow indicates fish that survived 14 days after the transfer from the cooler.

Field experiment

In the low-temperature tolerance field experiment, we caged individual H. letourneuxi in metal mesh minnow traps (with the openings plugged) in three habitats: freshwater marsh, alligator ponds, and solution holes in Everglades National Park. We caged fish at three replicate sites within each habitat. At each site, we placed four cages (each with one fish), including one cage that housed a temperature data logger. In the deeper habitats (alligator ponds and solution holes), we caged fish at the top and bottom of the water column. A control group of 12 fish (three sets of four fish) were caged in insulated mesocosm tanks at Nova Southeastern University in Dania, Florida. Fish were placed in the field on January 1, 2008, the day before a cold front, and retrieved on January 3 after passage of the front.

Drs. Loftus and Schofield checking caged fish in the marsh (left). A dead fish in a marsh trap (right).

Dr. Loftus and J. Langston setting caged fish into a solution hole (left). Cages placed in solution hole (right).

A weak cold front entered southern Florida the night of January 1, 2008. A stronger reinforcing front passed through the next day, leading to the lowest temperatures in 5 years during the night of January 2. Our temperature logger data indicated that the three habitats responded differently to the cold snap

Temperature logger data for January 1-3, 2008. Each line represents a mean of three replicates. Temperature loggers collected data every 5 minutes.

No control fish died at the mesocosm facility at Nova Southeastern University throughout the study, where temperatures remained above 17.3°C. The first night of the study (January 2), when temperatures were cool but not lethal, fish at the bottom of the solution holes died. These fish likely died from low dissolved oxygen (mean = 0.43 mg/ L), because they were unable to reach the surface to perform aquatic surface respiration.

The morning of January 3, we observed many non-native fishes lying stunned or dead at the bottom of the alligator ponds, including Mayan cichlid (Cichlasoma urophthalmus), pike killifish (Belonesox belizanus), jaguar guapote (Cichlasoma managuense) and H. letourneuxi. All fish caged in marsh habitats, where temperatures dipped to 5°C, were dead on the morning of January 3 (n=12). Most fish caged at the top of the solution holes survived, and fewer fish from the pond habitats survived. We left a subsample of the LOE fish from pond habitats in place for an additional week to determine whether they would revive. However, those fish did not survive for the 7-day period following the cold front.

Survival of Hemichromis letourneuxi in Everglades National Park low-temperature tolerance study.

Acknowledgments

This study was funded by the South Florida Water Management District, the U.S. Fish and Wildlife Service and the U.S. Geological Survey, Invasive Species program. Jackie Langston and Denise Gregoire provided expert assistance in the field and laboratory. Shane Ruessler and Bob Lewis provided technical assistance in the laboratory. Jen Rehage and Ashley Porter kindly provided assistance with fish husbandry. This study was completed under IACUC approval USGS/FISC 2007-02 and Everglades National Park permit EVER-2007-SCI-0014. Buck Albert assisted with photography.

ABOUT FISC

OFFICE ADDRESSES