Salinity tolerance of goldfish: a widely-distributed
non-native cyprinid in the U.S.A.

P. J. Schofield, M. E. Brown and P. L. Fuller
U. S. Geological Survey, Gainesville, FL
 

Presented at the Southern Division, American Fisheries Society meeting, San Antonio, TX, February, 2006.

Introduction

       The goldfish, Carassius auratus, is native to East and Central Asia, but has been introduced into the U.S.A. since the early 1600s.  It is a hardy species that lives in a variety of habitats and has a wide diet.  Originally introduced as an ornamental, the species is now cultured for a number of purposes, including bait, forage and the aquarium trade. Escapes from aquaculture facilities and deliberate releases have resulted in the establishment of localized populations across much of the U.S.A. In fact, the goldfish is reported from every state except Alaska and maintains reproducing populations in at least 45 states (see Map). However, a more detailed distribution delineating specific ponds, lakes, streams and other water bodies inhabited by goldfish is difficult to ascertain as introductions continue intermittently throughout much of the country. To add to the confusion surrounding the goldfish's distribution, it is unclear whether the species is capable of surviving in brackish-water regions. The purpose of this study was to determine the limits to low levels of salinity for both standard and black salty goldfish under acute-transfer and chronic regimes.

Distribution of Carassius auratus in the U.S.A. Map courtesy of U.S.G.S., Nonindigenous Aquatic Species Program (http://nas.er.usgs.gov).  See below for explanation of data sources and definitions of "reproducing", "reported" and "states with records".

Explanation of data sources and coding for goldfish distribution map.

       Distribution information for goldfish is presented by drainage and by state, each coverage indicating somewhat different levels of information.

       Distribution by drainage --- U.S. Geological Survey Hydrologic Unit Codes (HUCs) were used as a base to build the map. This is a nationwide system that delineates watersheds based on surface hydrologic regions (for more information, see: http://nas.er.usgs.gov/hucs.asp). HUCs where the goldfish is known or believed to have naturally reproducing populations are coded as "reproducing". Usually, evidence for reproduction is based on reports that eggs, larvae or small juveniles have been taken from one or more sites within the HUC. HUCs are coded as "reported" if the species has been collected, stocked, or observed in one or more parts of a drainage, but no evidence exists that the species is successfully reproducing.

       Distribution by state --- States color-coded in green represent those from which the species has been recorded at least once from natural waters within state borders. In some instances, the record of occurrence may be non-specific, with no information about the precise drainage or location within the state where the species was reported, released, or captured. In cases without specific geographic data, an entire state is shaded green (but no drainage is shaded). In contrast to the HUC units described above, the state colors do not provide information concerning the population status of the species (i.e. reproducing versus non-reproducing).

       Water bodies included in the coverage -- Distributions delineated on the map pertain only to records from the wild, including both natural (e.g., rivers, lakes) and artificial (e.g., canals, farm ponds, reservoirs) habitats. Records based on captive indoor settings or aquaculture facilities (e.g., outdoor ponds and tanks) are not included.

       Source of map distribution data -- Data used to create the map were primarily derived from the Nonindigenous Aquatic Species (NAS) database at the U.S. Geological Survey and from the closely related publication on nonindigenous fishes by Fuller and co-workers (1999). The database is built from a variety of sources, including scientific literature, published and unpublished reports and museum specimens. The database also relies on information from personal communications with scientists and others, as these sources are often the most recent (or only) documentation of a species occurrence in a particular area.
 

Methods

       Fish were weighed to the nearest 0.1 g and placed in experimental aquaria (clear plastic Rubbermaid © bins; one fish per bin) with well water where they were held for about three weeks before the experiment began.  Each bin was plumbed with an airstone.

       Acute salinity tolerance: Individuals were transferred from holding conditions (well water) to one of six treatments (5, 10, 15, 20, 25, well-water control) where they were held for 72 hrs. Fish were checked at 2, 4, 6, 8, 12, 24, 30, 48 and 72 hrs after transfer. Replicates were as follows:  standard: 0.2, 15, 20, 25 ppt = 18 replicates each; 5, 10 ppt = 17 replicates each; black salty: 0.2, 5, 15, 20, 25 ppt = 6 replicates each; 10 ppt = 5 replicates. Standard goldfish ranged in mass from 8.8 to 55.0 g (mean = 22.7 + 12.9 SD); black saltys ranged in mass from 8.8 to 24.6 g (mean = 17.3 + 4.5 SD).  Both varieties were distributed uniformly among treatments, and variances did not differ significantly among treatments.

       Chronic salinity tolerance: Fish were exposed to progressively increasing salinities (a change of 2.0 ppt every 2-3 d) until each treatment had reached its terminal salinity (5, 10, 15 well-water control).  Goldfish reached their terminal salinities in a staggered (time-wise) fashion; however, each time we changed the salinity in one or more of the treatments, the water was changed for all fish to maintain similarity of handling across treatments.  Fish were checked once per day, four to five days per week. Fish were fed two days per week ad lib with floating pellet food, and we recorded whether each fish fed each week. After terminal salinities were reached for all fish, water was changed once per week.  Replicates were as follows: black salty: 0.2, 10 ppt = 8 replicates each; 5 ppt = 7 replicates; 15 ppt = 9 replicates – standard: 0.2, 5, 15 ppt = 12 replicates each; 10 ppt = 13 replicates.  Standard goldfish used in the chronic salinity tolerance study ranged from 23.2 to 64.1 g (mean = 36.2 + 9.0 SD); black saltys ranged in mass from 10.6 to 37.4 g (mean = 19.9 + 6.8 SD).  Both varieties were distributed uniformly among treatments, and variances did not differ significantly among treatments.
 

Results

Mean survival estimates of black salty and standard goldfish.  Survival was estimated with a Kaplan-Meier product limit estimator. The log-rank test was used for comparisons amongst survivorship curves generated by the Kaplan-Meier estimator. Comparisons (black salty versus standard goldfish) with values of P < 0.05 are marked with an asterisk. Error bars are + 95% confidence intervals.
 

Black salty goldfish
 

Results and Conclusions

• Both varieties of goldfish are able to persist in low-salinity environments (<10 ppt) for long periods of time and at higher salinities for short time periods.
   
• When acutely shifted from freshwater to low-salinity conditions (5-15 ppt) the species is capable of survival for at least 72 h. However, acute transfer to salinities of 20-25 ppt lead to 100% mortality by 8 h.
   
• Under chronic low-salinity conditions, the goldfish showed high levels of survival at salinities of 5 and 10 ppt, but significant mortality at salinities of 15 and 20 ppt.
   
• A newly developed variety of goldfish (black saltys) was basically equivalent to standard goldfish in its salinity tolerance.
   
• The ability of the goldfish to withstand low levels of salinity indicates that it may be able to invade estuarine areas periodically, especially during periods of high rainfall. Therefore, the species may be capable of using estuarine regions to gain access to new river systems and expand its distribution.
   
• Although the goldfish is capable of survival at low (<10 ppt) salinities for an extended period of time, these conditions may be stressful to the fish. Published reports indicate that holding goldfish at low salinities causes changes in the distribution and density of mitochondria-rich ("chloride") cells, evokes increases in blood cortisol, increases oxygen consumption by 54 to 64%, increases urea-N excretion, and decreases specific growth rate and feed-conversion ratio.  These sublethal effects may ultimately limit establishment of the species in brackish waters.
   

Acknowledgments

       Funding for this study was provided by the U.S. Fish and Wildlife Service (Region 4) and the U.S. Geological Survey. Amy Benson and Myriah Richerson lent assistance in the laboratory. Bob Howells (Texas Parks and Wildlife Division) and Howard Jelks provided many helpful suggestions. Photos courtesy of Buck Albert.

       This is a product of the Nonindigenous Aquatic Species Program, USGS, Florida Integrated Science Center and the US Fish and Wildlife Service.

Citation information for this website:

• Schofield, P.J., M.E. Brown and P. Fuller. 2006.  Salinity tolerance of goldfish Carassius auratus, a widely-distributed non-native cyprinid in the U.S.A. Poster presented at the Southern Division, American Fisheries Society meeting in San Antonio, February 2006. Available online at:
  http://fl.biology.usgs.gov/posters/Nonindigenous/Salinity_Tolerance_of_Goldfish/salinity_tolerance_of_goldfish.html
   
• Also see:
  Schofield, P.J., M.E. Brown & P. F. Fuller.  2006. Salinity tolerance of goldfish, Carassius auratus, a non-native fish in the United States. Florida Scientist 69(4): 258-268.