Atherina mordax (Mitchill, 1814), Osmerus spectrum (Cope, 1870), American Smelt, Atlantic Smelt, Freshwater Smelt, Leefish.

Spawning of Rainbow Smelt occurs after ice-out in early spring once water temperatures reach ~4.4°C with a peak at 10°C (Becker, 1983). In the Great Lakes, spawning occurs in tributary streams and occasionally on the shoreline. Eggs of the Rainbow Smelt are adhesive, negatively buoyant, and adhere to substrate until hatching 2 to 3 weeks later (Scott and Crossman 1973; Becker 1983). Age of maturation ranges between 1 to 4 years amongst populations (McKenzie 1958; Murawski and Cole 1978) and life expectancy can reach 8 years (Bailey 1964; Kirn and Labar 1996). They are sexually dimorphic, with females growing larger and living longer than males (McKenzie 1958; Bailey 1964).

Rainbow Smelt can survive in a broad range of environmental conditions. They inhabit temperatures from -1.8°C to around 20°C (Nellbring 1989; Raymond 1995), and can acclimate to extreme cold (< 5°C) by producing an antifreeze protein, glycerol, to facilitate faster-swimming speeds and increased muscle function (Coughlin et al. 2019). A laboratory study that assessed embryonic and larval Rainbow Smelt survival in response to various abiotic factors (dissolved oxygen, pH, salinity, nitrates, and phosphates) found that their survival was only significantly reduced when dissolved oxygen levels were less than 20% and the pH<5 (Fuda et al. 2007). However,a review of data from 8,842 Ontario lakes found that Rainbow Smelt were only caught in lakes with a pH>6 despite adequate opportunity to invade lakes with lower pH values (Evan and Loftus 1987).

Rainbow Smelt are zooplanktivorous at small sizes (<150 mm SL), consuming copepods and cladocerans (e.g., Daphnia and Bosmina); while at larger sizes (>150 mm SL) diet includes benthic crustaceans, and larval and small fishes  (Sheppard et al. 2012). They are an important forage fish for many predators, including Atlantic Salmon (Salmo salar), Walleye (Sander vitreus), and Lake Trout (Salvelinus namaycush) (Gorman 2019).

Two means have been proposed to explain the introduction of Rainbow Smelt into the Missouri and Mississippi rivers. It may have spread from Lake Michigan via the Chicago sanitary canal to the Illinois River and then to the Mississippi and Missouri rivers (Burr and Mayden 1980). Alternatively, the species may have gained access to these rivers as a result of a stocking at Lake Sakakawea, North Dakota, in 1971 (Bouc 1987; Mayden et al. 1987; Holton 1990). The second explanation seems more plausible because of a lack of records from the Illinois River. Records of first occurrences in other areas include the Mississippi River, Illinois, and Kentucky, 1978; Mississippi River, Louisiana, 1979; Mississippi River, Tennessee, and Arkansas, 1980; Missouri River, Missouri, 1980; Missouri River, Kansas, 1982 (Mayden et al. 1987). Mayden et al (1987) provided a map of the species' distribution, dates of first observation in new areas, and possible introduction pathways. The species was originally introduced into Lake Sakakawea, North Dakota, as forage for Salmonids (Mayden et al. 1987).

Great Lakes
Widespread invasive with high impact.  Reproducing and overwintering at self-sustaining levels have been recorded in all five Great Lakes.

Summary of species impacts derived from literature review. Click on an icon to find out more...

Environmental	Socioeconomic	Beneficial

Osmerus mordax has a high environmental impact in the Great Lakes.

Realized:
Since its introduction, Rainbow Smelt has become extremely integrated into the food webs of the Great Lakes. Its complex connections to various trophic levels have been documented by a wide range of studies. In different ecosystems, Rainbow Smelt may be an important prey item, predator, or competitor (Evans and Loftus 1987); in many cases, it may participate in multiple roles relative to a native species. Part of this complexity lies in the fact that Rainbow Smelt can have different impacts on native species during its different life stages (Hrabik et al. 2001). Impacts may also differ with the age class of the native species. For instance, Rainbow Smelt consumes larval fishes and is a documented predator of Lake Trout (Salvelinus namaycush) eggs (Evans and Loftus 1987); however, Rainbow Smelt is in many cases the primary food source of adult Lake Trout (Gorman 2007).

In the Great Lakes, Rainbow Smelt may compete with Cisco (Coregonus artedi) for food (Becker 1983). Christie (1974) supplied some evidence to support this, correlating Cisco decline with Rainbow Smelt increases in most of the lake. However, declines, local extirpations, and limitations to the recovery of Cisco populations have also been attributed to Rainbow Smelt predation on larval fish rather than competition (Hrabick et al. 1998, Stockwell et al. 2009). Both predation by and competition with Rainbow Smelt have been implicated in the declines of several endangered or special concern species in Canada, including Blackfin Cisco (C. nigripinnis) and Shortnose Cisco (C. reighardi), as well as Deepwater Sculpin (Myoxocephalus thompsonii) (COSEWIC 2005, 2006, 2007). Todd (1986) also reported that Rainbow Smelt may be partially responsible for the decline of whitefishes (Coregonus spp.) in the Great Lakes. The ultimate extent of Rainbow Smelt impact on native Cisco species in the Great Lakes remains uncertain (Dobiesz et al. 2005, Schmidt et al. 2009). Rainbow Smelt may have played a role in the decline of other threatened natives and has been suggested as a factor leading to the extinction of Blue Pike (Sander vitreus glaucus) (Dextrase and Mandrak 2006, Hartman 1973, USEPA 2008).

In a review of Rainbow Smelt introductions in inland Ontario lakes, Evans and Loftus (1978) found that 13 of 24 lakes with introduced Rainbow Smelt experienced a decline in Lake Whitefish (C. clupeaformis) recruitment while 5 of 19 reported declines in Cisco. Hrabik et al. (1998) found evidence of competition for food between introduced Rainbow Smelt and native yellow perch (Perca flavescens) in Wisconsin lake habitats. A study of Wisconsin inland lakes with and without introduced Rainbow Smelt from 1985-2004 found that young-of-the-year walleye (Sander vitreus) density was significantly lower in invaded lakes (Mercado-Silva et al. 2007). However, a separate study of two Wisconsin lakes found that fishing restrictions that allowed the recovery of adult walleye populations over several years were concurrent with a decline in Rainbow Smelt and recovery of native Cisco (C. artedi) (Krueger and Hrabik 2004).

A study on nighttime consumption by O. mordax in Lake Ontario revealed its primary food source to be Slimy Sculpin (Cottus cognatus) and opossum shrimp (Mysis relicta) (Brandt and Madon 1986). Juvenile Lake Trout relies heavily on C. cognatus, competing directly with O. mordax, while adult S. namaycush consumes O. mordax (Brandt and Madon 1986). The authors point out that S. namaycush may be a keystone predator in the relationship between O. mordax and C. cognatus. Rainbow Smelt is also known to feed on the young of Lake Whitefish, Lake Trout, and Burbot (Lota lota) (Evans and Loftus 1987). Stedman and Argyle (1985) found the diet of O. mordax to include young-of-the-year fishes, with consumption of particular prey species dependent upon prey availability (Stedman and Argyle 1985). Stedman and Argyle (1985) discovered that in Lake Michigan in late October of 1982, O. mordax consumed Bloater (C. hoyi) and Alewife (Alosa pseudoharengus). The authors also noted that although O. mordax did not appear to have had an impact on Bloater populations between 1980 and 1985, though future impacts are possible.

Rainbow Smelt can pose some health threats to its predators. It can contain high concentrations of thiaminase which when consumed in large quantities has been shown to decrease swimming performance and body condition, and decrease yellow pigmentation in Atlantic Salmon (Salmo salar) (Houde et al. 2015). Further, the consumption of thiaminase and subsequent deficiency in thiamine can cause early mortality syndrome (EMS) in Lake Trout (Honeyfield et al. 2005).  Blukacz-Richards et al. (2017) identified a strong trophic interaction between mercury concentrations in Rainbow Smelt and herring gull (Larus argentatus) eggs in Lake Superior and Lake Ontario. However, trophic elevation in forage fish is unlikely to result in harmful bioaccumulation (Swanson et al. 2003, 2006).

Current research on the socio-economic impact of Osmerus mordax in the Great Lakes is inadequate to support proper assessment.

Realized:
While many recreationally important species benefit from Rainbow Smelt as a prey item, in some instances, Rainbow Smelt has been associated with predation on and competition with these species (see above).

Osmerus mordax has a high beneficial effect in the Great Lakes.

Realized:
Havey (1973) reported increased growth of landlocked Atlantic Salmon following the introduction of Rainbow Smelt as a forage species in a lake in Maine. Osmerus mordax has also been used by USGS to monitor contaminant levels in the Great Lakes (Chernyak et al. 2005).

Rainbow Smelt can have a wide range of interactions with native species and sometimes demonstrates a neutral coexistence, such as with Smallmouth Bass (Micropterus dolomieu) and White Sucker (Catostomus commersonii) (Evans and Loftus 1987). Rainbow Smelt can also have a positive impact on natives because it provides a food source to many native and non-native piscivores in the Great Lakes, including native Burbot, Yellow Perch, and recreationally important introduced Salmonids. Species such as Walleye and Lake Trout may both benefit and suffer from introductions of Rainbow Smelt depending on the extent to which Rainbow Smelt acts as a prey item, predator, or competitor (Evans and Loftus 1987). Because so many species—including recreational and commercial species—depend on Rainbow Smelt as a food source, Rainbow Smelt are a vital member of the current food web and are considered by some to be an important species to manage and conserve (Schmidt et al. 2009).

Rainbow Smelt is not only valuable as a forage or bait fish for top predators; it is harvested commercially in both the U.S. and Canada, particularly in lakes Michigan and Erie (Dann and Schroeder 2003, Madenjian et al. 2002). It was estimated in 2003 that the commercial Rainbow Smelt harvest in the U.S. Great Lakes alone was worth over $750,000 yr-1—more than Lake Trout, Lake Herring, or Pacific Salmons (Dann and Schroeder 2003). Historically, the recreational harvest of Rainbow Smelt has also been popular (Scott and Crossman 1998); an annual harvest of over 150,000 Rainbow Smelt in the Great Lakes system was recently reported in a 2005 survey of anglers in Canada (Fisheries and Oceans Canada 2008). From 2010 to 2018, over 63 million pounds of Rainbow Smelt worth 14.6 million CAD was commercially harvested from Lake Erie and Lake St. Clair, primarily by the Canadian based Great Lakes Food Company Ltd. (OCFA 2018).

In the state of New York, it is unlawful to use Rainbow Smelt as bait except as provided in 6 NY CRR § 19.2. Furthermore, it is unlawful to take Rainbow Smelt for sale as bait or to sell as bait, except as otherwise provided as without a pursuant license as defined in NY ECL § 11-1315. In Pennsylvania, the use of commercial trap nets under license to capture Rainbow Smelt is regulated by Penn. Admin. Code § 69.33. In Ohio, Rainbow Smelt is defined as a commercial fish and an unrestricted species under Ohio Admin. Code 1501 § 31-1-02. Commercial fish are permitted to be taken, possessed, bought, or sold unless otherwise restricted in Ohio code. In Indiana, the Rainbow Smelt sport fishing season on Lake Michigan is defined as March 1-May 30, with capture allowed only by the use of dip nets, seines, or nets with limitations provided in 312 IAC § 9-7-2. There is otherwise no bag limit, possession limit, or size limit, as defined under 312 IAC § 9-7-14. In Illinois, the sport-fishing season of Rainbow Smelt is defined as March 1-April 30 under Illinois Admin. Code 17-1 § 810.10. In Wisconsin, Rainbow Smelt is defined as an established non-native fish species in Wis. Admin. Code § NR 40.02, and is restricted per the above definition by Wis. Admin Code § NR 40.05. In Minnesota, Rainbow Smelt is a regulated invasive species under Minn. Admin Rules § 6216.0260.

Note: Check federal, state/provincial, and local regulations for the most up-to-date information.

Control
Biological
Rainbow Smelt are heavily consumed by Atlantic Salmon (Salmo salar), Lake Trout (Salvelinus namaycush), Brook Trout (S. fontinalis), Coho Salmon (Oncorhynchus kisutch), Chinook Salmon (O. tshawytcha), Rainbow Trout (O. mykiss), Brown Trout (Salmo trutta), Splake (Brook Trout x Lake Trout), Burbot (Lota lota), Walleye (Sander vitreus), Northern Pike (Esox lucius), and many other freshwater piscivores (Brandt and Madon 1986; Crossman 1991; GLMRIS 2012; He and LaBar 1994; Kirn and LaBar 1996; Stewart et al. 1981). However, the significance of piscivore predation on Rainbow Smelt has only been studied for a few species. Observed Atlantic Salmon predation on smaller Rainbow Smelt, as well as bioenergetics modeling suggesting that by age 4, cumulative piscivory by Atlantic Salmon was nearly 10-fold greater than that of Lake Trout of the same age, implies its greater usefulness for management of Rainbow Smelt (Kirn and LaBar 1996). While Lake Trout consume large amounts of Rainbow Smelt, almost exclusively so in some studies, the species is believed to provide little potential for responsive management manipulation outside of stabilizing fluctuating prey populations, due to the long cycle of its predatory effect (peaking 3-5 years after stocking, lasting 7-8 years) (He and LaBar 1994; Kirn and LaBar 1996; Stewart et al. 1981). Chinook Salmon have been successfully used to eradicate Rainbow Smelt from small lakes in New Hampshire in 1936 (Stewart et al. 1981). Because the trade-off between fish species as agents of biological control is not directly correlated with consumption, management decisions involving shifts between species should not take consumption solely into account (Stewart et al. 1981).

Physical
The USACE Great Lakes and Mississippi River Interbasin Study notes the potential effectiveness of sensory deterrent systems in providing barriers to fish migration or eliciting fish movements (GLMRIS 2012). In situ testing of two models of strobe lights as a deterrent preventing entrainment of Rainbow Smelt through Oahe Dam, Lake Oahe, South Dakota demonstrated successful avoidance of 15-21 m horizontally and 6 m vertically by Rainbow Smelt (Hamel et al. 2008). Many large scale strobe systems consist of four individual lights that flash at a rate of 450 flashes/min., with an approximate intensity of 2634 lumens/flash (GLMRIS 2012).  Hamel et al. (2008) tested the AGL FH-901 flashhead, which consists of four horizontal lights positioned at 90-degree angles, flashing 450 times/min at 2,634 lumens/flash, and the newer AGL FH-920 flashhead, which consists of an omnidirectional vertical light tube, covering a full 360 degrees with 360 flashes/min at 6,585 lumens. When using physical deterrents as barriers, combining methods can increase effectiveness, as was the case for Patrick et al. (1985), who found that Rainbow Smelt and other pelagic fishes were successfully deterred by a barrier combining air bubbles and strobe lights.

Chemical
Of the four chemical piscicides registered for use in the United States, antimycin A and rotenone are considered general piscicides (GLMRIS 2012). Marking et al. (1983) found that the three most effective registered chemicals for potential use in the control of Rainbow Smelt eggs and larvae are rotenone, potassium permanganate, and chlorine, respectively. In exposures of 6-24 hours, all chemicals were effective at concentrations from 5 to >10 mg/L (Marking et al. 1983). Rotenone demonstrated a 96h LC50 of 0.015 mg/L for Rainbow Smelt eggs and 0.001 mg/L for larvae (derived calculating only the activity of rotenone in 5% Noxfish solution) (Marking et al. 1983). Potassium permanganate demonstrated 96h LC50s of 0.074 mg/L and 0.075 mg/L for eggs and larvae, respectively. Chlorine demonstrated 96h LC50s of 0.14 mg/L for eggs and 0.31 mg/L for larvae (Marking et al. 1983). Temperature, pH, and hardness of water all affected toxicity of rotenone and potassium permanganate, with higher temperatures, softer water, and higher pH increasing toxicity (Marking et al. 1983). It should be noted that tests were carried out in a laboratory, but natural waters usually contain oxidizable material, which produces a chlorine demand and reduces it to a less active form (Marking et al. 1983).

Increasing CO2 concentrations, either by bubbling pressurized gas directly into water or by the addition of sodium bicarbonate (NaHCO3), has been used to sedate fish with minimal residual toxicity, and is a potential method of harvesting fish for removal, though maintaining adequate CO2 concentrations may be difficult in large/natural water bodies (Clearwater et al. 2008). CO2 is approved only for use as an anesthetic for cold, cool, and warm water fishes in the US, not for use as euthanasia (Clearwater et al. 2008). Exposure to NaHCO3 concentration of 142-642 mg/L for 5 min. is sufficient to anesthetize most fish (Clearwater et al 2008).

It should be noted that chemical treatment will often lead to non-target kills, and so all options for management of a species should be adequately studied before a decision is made to use piscicides or other chemicals. Potential effects on non-target plants and organisms, including macroinvertebrates and other fishes, should always be deliberately evaluated and analyzed. The effects of combinations of management chemicals and other toxicants, whether intentional or unintentional, should be understood before chemical treatment.  Other non-selective alterations of water quality, such as reducing dissolved oxygen levels or altering pH, could also have a deleterious impact on native fish, invertebrates, and other fauna or flora, and their potentially harmful effects should therefore be evaluated thoroughly.

Note: Check state/provincial and local regulations for the most up-to-date information regarding permits for control methods. Follow all label instructions.