Conservation Hatchery

Managers may choose from two basic options for conservation or restoration of a population or community: (1) alter habitat (typically repair) or (2) manipulate the number of individuals of a species or population through harvest (removing individuals) and (or) stocking (adding individuals). Hatcheries are a critical tool for adding individuals to a population and have most often been focused on supplementing fish to directly support fisheries (Baillie and others, 2015; Trushenski and others, 2018). However, hatcheries are becoming vital tools for restoration, and research conservation hatcheries are key facilities for understanding the culture and release methods that are most effective for conservation and restoration of native fish populations (Lochmann, 2019). The U.S. Geological Survey (USGS) Great Lakes Science Center’s Tunison Laboratory of Aquatic Science (TLAS) is a research facility capable of rearing sufficiently large numbers of fish to support the field experimentation needed to determine the effectiveness of fish population manipulation efforts to restore extirpated and depleted native species. The TLAS research is the focus of this report but succeeds as a collaborative effort with several other facilities and agencies (Ontario Ministry of Natural Resources and Forestry [OMNRF] White Lake Fish Culture Station [FCS], U.S. Fish and Wildlife Service [FWS] Allegheny National Fish Hatchery [ANFH], FWS Northeast Fishery Center [NEFC], not shown in figure 1).

The Coregonine Research Program at TLAS includes a diverse array of research projects. An emphasis of the program was improved production and field assessments of juvenile bloater and cisco released experimentally to support restoration in Lake Ontario. Re-establishing and (or) rehabilitating coregonine populations as components of the Lake Ontario fish community is a high priority for United States and Canadian fisheries management agencies, and TLAS research is carried out in part to support fish community objectives of the Lake Ontario Committee (Stewart and others, 2017). This research is part of a broader effort throughout the Great Lakes to restore or conserve native coregonine fishes. The TLAS research program is conducted in collaboration with scientists from other laboratories and agencies, particularly the NEFC and FWS Lower Great Lakes Fish and Wildlife Resources Office; the USGS Lake Ontario Biological Station; the New York State Department of Environmental Conservation (NYSDEC) Lake Ontario Unit, regional offices, and hatchery system; and the OMNRF Glenora Station and White Lake FCS. Funding for this research is largely provided by the Great Lakes Restoration Initiative. Coregonine restoration research has been a major focus of the TLAS program for most of the last decade and is designed to determine the success of restoration methods. The feasibility of re-establishing bloater and cisco populations in Lake Ontario is being tested with notable successes discussed herein, including the first successful rearing of bloater; successful releases of these two native species into Lake Ontario; and collections of returning adults or surviving juveniles. We discuss culture and release methods, post release field evaluations of experimental fish, and additional research needs. A brief history of the program is provided, with an emphasis on our results for the last 2 years (2019–20).

Study Area and Conservation Hatchery

Lake Ontario is a deep Great Lake and losses of native species have left the deep waters depleted of fish life (Bronte and others, 2017; Stewart and others, 2017). Also, overharvest and other factors have left coastal Lake Ontario embayments that historically supported large spawning populations of cisco without native coregonines. Cisco collection sites and Lake Ontario cisco and bloater release sites are about 120 kilometers (km) from Cortland, New York, where the TLAS facility is located (fig. 1). The TLAS facility can rear tens of thousands of fish for experimental research and has an ultraviolet (UV) treatment facility constructed in 2012, which treats effluent and protects the local drainage basin from possible exposure to fish diseases. Most water in this flowthrough system is provided to a variety of tanks and raceways by wells and maintains a constant temperature of approximately (~) 9 °Celsius (°C) year-round. Oxygen and other dissolved gases are monitored, and additional oxygen can be added as needed.

Annual Cycle of Rearing, Release, and Evaluation—Bloater

Bloater restoration in Lake Ontario is a collaborative multiagency research project with close coordination among the NYSDEC, FWS, OMNRF, and USGS. The annual cycles of fish production, release, and evaluation include egg collection; care from spawn through development, hatch, and juvenile growth; experimental release; and field evaluation of surviving juveniles, returning adults, and wild production of eggs and larvae. There are two sources of bloater eggs for the Lake Ontario restoration project. The first is from wild bloater collected from Lake Michigan at spawning time by the FWS Jordan River National Fish Hatchery in Elmira, Michigan. The other (and presently the primary) egg source is from the bloater brood stock derived from those wild collections and maintained at the OMNRF White Lake FCS, Sharbot Lake, Ontario (and a few other Canadian hatcheries).

Fish were cared for using standard hatchery practices (Piper and others, 1982; Wedemeyer, 2002); a specific methods and procedures manual for coregonine culture was recently completed (Chalupnicki and others, 2024). Eggs were cared for at White Lake until eye up (that is, when the embryo has developed eyes that are visible though the chorion [eggshell]). Some eyed eggs were transferred to TLAS and some to other facilities (for example, NEFC or ANFH, Warren, Pennsylvania). Disease-free certification (provided by NEFC or NYSDEC) is required for care and release of bloater from wild stock or for bloater eggs transferred across the international border. Skilled fish culture staff at TLAS monitored bloater egg development weekly and standard fungal treatments were administered three times per week until hatch. Feed was provided in quantities that maximize growth for the water temperature and water volume and flow rate in each tank. Fish growth and survival were measured biweekly from hatch to stocking. A systemic mark was used to color the bones and identify the cohort source. Marking has involved single or double marks with oxytetracycline (OTC) or calcein administered either in feed (OTC) or via an immersion bath (calcein) at 2 months post-hatch (and 3 months for a double mark; table 1); a single mark was standard for bloater. At release, bloater fingerlings ranged from 75 millimeters (mm; 3 grams [g]) to 105 mm (8.5 g) and yearlings ranged from 97 mm (11.0 g) to 121 mm (13.0 g), depending on the temperature at which they were reared. Bloater have been consistently released offshore into deep waters of Lake Ontario (2012–20 and continuing) either from a haul tank on a research vessel (that is, USGS R/V Kaho) or from a stocking truck on a landing craft. Bloater fingerling stocking has usually been early in November and bloater yearling stocking is in mid-May. Recapture of released fish has been done using bottom trawls during annual USGS–NYSDEC–OMNRF fish surveys (Weidel and others, 2022).

Annual Cycle of Rearing, Release, and Evaluation—Cisco

Cisco restoration in Lake Ontario embayments that historically supported large spawning aggregations has been a collaborative multiagency research project with close coordination among field crews and spawning teams from the NYSDEC, FWS, and USGS. In late November, adult cisco were collected from Johnson, Middle, and Herrick Shoals in Chaumont Bay, Lake Ontario using Oneida-style trap nets (a box trap net with long wings projecting laterally from the mouth and a conical codend) soaked for 24 hours (hr). These adults were transported via an insulated and aerated haul tank on a truck to TLAS where eggs were spawned manually, usually within 24 hr. Tissue samples from spawned fish were sent to NEFC for disease screening. Eggs and fish were reared using standard hatchery practices like those applied to bloater culture. Eggs were cared for at TLAS until the eyed stage and disease-free certification was acquired. Some eyed eggs were then transferred to other facilities, such as NEFC or NYSDEC hatcheries. Cisco egg development was monitored weekly and standard fungal treatments were administered three times per week until hatch. Feed was provided in quantities that maximize growth for the water temperature and water volume and flow rate in each tank to produce fish most likely to survive well after release into the wild. Fish growth and survival were measured biweekly from hatch to stocking. Also, a systemic mark was used on cisco to color the bones and identify the cohort source and release site. Marking has involved single or double marks with OTC or calcein administered either in feed (OTC) or via an immersion bath (calcein) at 2 months post-hatch (and 3 months for a double mark; table 2). At release, cisco fingerlings range from 87 mm (4.4 g) to 133 mm (17.5 g) depending on the temperature at which they were reared.

Cisco were released into target embayments (specifically Irondequoit Bay or Sodus Bay) from 2011 to 2020 from either a stocking boat (operated by NYSDEC) or from shore directly out of the stocking truck (provided by either FWS or NYSDEC). Cisco fingerlings were stocked in late October or early November. However, because of limited capacity, summer releases have also happened in early July when fish were about 68 mm (2 g). Juvenile and adult ciscos were collected subsequent to experimental releases using trap nets or gillnets deployed in the experimental or adjacent embayments. Standard American Fisheries Society gillnets or trap nets (Bonar and others, 2009) were soaked for ~24 hrs for each set. As many as an additional five cisco annually have been caught and photographed by anglers in the past three years. Egg and larval collections were done annually within the experimental embayments (Sodus Bay and Irondequoit Bay) to detect any wild reproduction from 2015 to the present; collections were expanded to reference sites (Port Bay and Little Sodus Bay) in 2018 and continue to the present. Egg mats were deployed in December in randomly selected locations (3–9 per embayment, 2017–19; fig. 2) on presumed-appropriate habitat (based on the presence of relatively clean, gravel-cobble substrate). Ichthyoplankton tows were used in late April to sample larvae that may have been near the surface after the estimated hatch date (mid-April). A multi-agency effort sampled larval coregonines from numerous Lake Ontario embayments in spring of 2018, including Irondequoit Bay, Sodus Bay, Port Bay, Little Sodus Bay, Chaumont Bay, Henderson Bay (fig. 2), and Sandy Pond (Brown and others, 2022). The TLAS focused on Chaumont Bay (continuing a 17-year index sampling survey [2016 and 2017 missing]) with additional sampling in Little Sodus Bay, Port Bay, Sodus Bay, and Irondequoit Bay (the latter two being the experimental cisco release sites). Each year, larvae were sampled during a week in late April to early May. Seven to 20 sites were sampled within each bay using paired 1.0-meter (m) diameter x 3-m cone length ichthyoplankton nets with 500-micrometer (µm) mesh. Each net haul was at the surface for 5 minutes at 1,000 revolutions per minute (rpm) and distance travelled was recorded by global positioning system track (McKenna and Johnson, 2009; fig. 2). Contents from both nets were transferred to the same jar and preserved in 90 percent ethanol. Samples were then processed at TLAS.

Cultured Production and Experimental Releases—Bloater

Bloater egg production was from the wild Lake Michigan source from 2012 until 2018, producing as much as ~600,000 eggs. However, few wild eggs were collected in 2018 (~19,000; table 1). An additional collection happened in 2020 and consisted of ~72,000 eggs reared for Lake Ontario (fig. 3). Egg production from the OMNRF White Lake Hatchery brood stock began in 2018 and continues to the present; distribution to United States facilities (TLAS, NEFC, and FWS ANFH) has ranged from ~250,000 to nearly 700,000 eggs annually. Hatch rates for eggs from wild fish have been higher than that of brood stock eggs, but bloater egg quality from both sources was generally poor in 2020. Of the 171,738 wild green eggs received by TLAS (lot 1: 86,778; lot 2: 42,480) from the FWS Jordan River National Fish Hatchery in 2020, only 70 percent (120,217) survived to the eyed stage. The NEFC received 108,139 of these eyed eggs for experimental rearing. The remaining 12,078 were raised at TLAS. TLAS also received two shipments of eggs from the OMNRF brood stock cohort in 2020: 316,705 as green eggs (2013 year class (YC): 25,750; 2014 YC: 290,955) and 233,104 as eyed eggs (2013 YC: 74,740; 2014 YC: 158,364). After 431 temperature units (daily thermal units), ~174,187 and 128,207 of those eggs (~55 percent) hatched, respectively (table 1). A directed experiment was completed in 2023 to investigate a possible nutritional deficiency in the cultured brood stock that may be related to egg quality (Sweka and others, 2024).

At TLAS, fry and juvenile survival increase with time after hatch, and mortality is typically less than 0.1 percent/day/tank when the fish reach ~0.15–0.25 g at around 2–2.5 months post-hatch (depending on water temperature and density of fish per tank). Biweekly measurements of fish growth from hatch to release, displayed a typical power function pattern (Chalupnicki and others 2024). In 2020, calcein was unavailable, so bloater were marked with a manufactured feed containing OTC mixed in to provide identification of release year. They received a single mark at 10 months post-hatch with medicated feed manufactured by BioOregon.

Experimental releases of bloater fall fingerlings happened in 2012–18 and ranged from ~1,200 to nearly 150,000 (table 1; fig. 3). However, yearlings are suspected of having greater survival after release into the lake at a time of abundant food (spring) (Holeck and others, 2020). Therefore, yearling spring releases ranging from ~11,000 to 300,000 occurred 2018–present. In 2020, a total of 11,153 spring yearling bloater (2019 YC) were released into Lake Ontario off of Oswego, N.Y., in ~100 m of water depth using a NYSDEC-chartered landing craft. Since the beginning of restoration research, the total number of bloater released from U.S. facilities into Lake Ontario is 792,226. Approximately, 150,000 TLAS-produced bloater were released each spring of 2021 and 2022.

Cultured Production and Experimental Releases—Cisco

The cisco egg-take from Chaumont Bay spawners increased steadily from 2011 to a peak of nearly 1,000,000 eggs in 2017 and then was reduced to align better with experimental release targets (table 2; fig. 4). In November 2019, a total of 752,144 eggs were spawned from adult cisco collected from Johnson and Herrick Shoals in Chaumont Bay. At eye-up, 26,816 eggs were transferred to the NYSDEC Bath FCS, and 26,397 to the NYSDEC Oneida FCS. The remaining 641,506 eyed eggs developed at TLAS, and 545,280 hatched (85 percent) after 376 temperature units. The FWS NEFC also received 75,000 of these fry in mid-February. The cisco egg-take in 2020 was minimal and only intended to satisfy the most basic partner needs; a total of 38 adult cisco were collected from Chaumont Bay. The 15 females and 23 males spawned at TLAS produced 103,810 eggs. Like bloater, juvenile cisco survival at TLAS increased with time after hatch, and mortality was typically less than 0.1 percent/day/tank when the fish reach ~0.15–0.25 g at around 2–2.5 months post-hatch (depending on water temperature and density of fish per tank). Biweekly measurements of fish growth from hatch to release displayed a typical power function pattern (Chalupnicki and others, 2024). Because calcein was unavailable in 2020, cisco were marked with OTC to provide indications of release year; they received OTC marks at 2 and 3 months post-hatch via medicated feed manufactured by BioOregon.

The NYSDEC, FWS, and TLAS collaborated to accomplish experimental releases of cisco into Irondequoit Bay 2012–15 and in 2018, ranging from ~9,000 to 240,000 fall fingerlings (and one 2018 August release). Experimental releases happened in Sodus Bay 2016–20 and ranged from ~22,000 to 400,000 fall fingerlings (most 2017 fish were released as summer fingerlings). In 2020, 178,744 fall fingerling cisco were shore stocked into Sodus Bay by TLAS and 60,500 fall fingerling cisco were shore stocked into Sodus Bay by NEFC. The total release of cisco into Irondequoit Bay to date is 497,295 and into Sodus Bay is 1,197,858.

Recapture of Released Fish and Larval Production

Bloater were recaptured from Lake Ontario each year from 2015 to 2019, except for 2016, for a total of nine fish (Weidel and others, 2022). Collection locations were widely distributed throughout Lake Ontario. Because of the COVID–19 pandemic, open lake surveys capable of collecting bloater were limited and assessments of adult cisco returning to the experimental embayments in 2020 were not possible. Adult and juvenile cisco have been captured from experimental release sites or adjacent embayments each year during 2015–19 (except for 2017; table 3), for a total of 16; two of those ciscos had verifiable calcein marks (table 2). There were also annual reports from ice fishermen of adult coregonines collected (as many as 30 in a season) from Port Bay, Irondequoit Creek, and other embayments. However, whether they were experimental fish released into Sodus or Irondequoit Bays has not been determined. Naturally produced larval coregonines have been consistently found in Chaumont Bay each spring since 2004 (McKenna and Johnson, 2009). Since 2018, coregonine larvae have also been collected from six other Lake Ontario embayments, including greater than 300 from Sodus Bay (experimental cisco release site) and Henderson Bay (adjacent to Chaumont Bay; table 4) (Brown and others, 2022). Some of the Chaumont Bay larvae have been genetically identified as cisco and others as lake whitefish (McKenna and others, 2020; Brown and others, 2023). Some of those from other embayments have also been genetically identified as cisco (Brown and others, 2023). Egg mats have been deployed in six experimental embayments and the known spawning area in Chaumont Bay (fig. 2). Thousands of eggs have been collected in Chaumont Bay, but only two have been collected from a different embayment (that is, Little Sodus Bay).

Distribution, Survival, and Recapture After Release

Recapture numbers have been low but are encouraging given the number released and the low probability of detection in the main lake, as well as the fact that most of the year classes have yet to mature and return to spawning sites. To date, nearly 500,000 bloater have been released into Lake Ontario by USGS and FWS, and a similar number (555,000) have been released by OMNRF. Nearly 500,000 cisco have been released into Irondequoit Bay and nearly 1.2 million into Sodus Bay. These may seem like large numbers of fish, but Lake Ontario is an enormous water body (the 13th largest in the world), with an area of nearly 19,000 square kilometers (km²) and a water volume of greater than 1.6x10¹² cubic meters (m³; U.S. Environmental Protection Agency, 2021). If all the bloater released to date were spread uniformly throughout the entire volume of Lake Ontario, there would be ~1 bloater for every 1.6 million m³. It would require ~110 standard trawls to collect a single bloater (assuming a volume of ~14,500 m³ swept by each trawl haul; Weidel and Walsh, 2013). This is similar to Weidel and others' (2022) estimate. However, not all these fish have been released into Lake Ontario at the same time and the number surviving in the lake is likely to be much smaller in any one year. Greater than 1,000 trawls would have been required in the year with the first bloater recapture (2015). Despite the low probability of collection, the trawl survey has recaptured nine bloater and as many as four in a single year. This tells us that not only are the bloater surviving in Lake Ontario for at least months after release, but also that they are probably not uniformly distributed. They have been recaptured at widely different locations throughout the lake. Similarly, for cisco, which are generally a shallow water species, the likelihood of recapture is low. If all the cisco released into the experimental embayments were spread out uniformly within the waters of Lake Ontario shallower than 30 m deep (a total volume of 6.7x10¹⁰ m³), we would expect one cisco for approximately every 32,000 m³. Also, cisco are collected with gillnets and trap nets, which are passive gear occupying small areas of the embayments, and yet 16 cisco have been captured from Irondequoit Bay, Sodus Bay, and Port Bay, two of which had calcein marks, confirming those fish were from experimental releases. Additionally, several anecdotal reports of cisco catches from ice fishermen (as many as 10 at a time with about a dozen fishermen reporting) suggest that cisco may be more prevalent in the experimental embayments than indicated by the trap net and gillnet catches.

Tunison Laboratory Research: Present and Future

Research at TLAS supports the Lake Ontario Committee objectives and partner needs in a variety of ways. In addition to the broader field experiments and associated fish rearing done at TLAS, numerous directed TLAS and partner studies were required to successfully produce a sufficient number of high-quality fish for the field experiments and to effectively release and evaluate the re-introduction of coregonine fishes into Lake Ontario. These include determining the proper feed (size and composition) for each life stage of bloater and cisco in the laboratory, determining optimal water flow and water quality (for example, dissolved gases and temperature) for best growth and survival, developing effective marking techniques that provide identification of the producing facility and year class of recovered fish; evaluating the best release locations and conditions; assessing postrelease survival, quantifying returning adults, and detecting wild egg and larval production.

Rearing

Marking

The ability to identify cultured fish once they have been released into the wild is critical for field experimentation. Unique marks distinguish cultured fish from any wild fish that may already exist in the system and can also identify particular source facilities of cultured fish, the year class of the fish, and other factors, such as what rearing or release techniques were used. Marking is vital for the Lake Ontario coregonine research, particularly for cisco, where wild populations still exist in the Eastern Basin of Lake Ontario. TLAS and NEFC developed effective color marking of bones that persists for years and will identify experimental fish for field evaluations of survival, habitat use, and other aspects of cisco and bloater ecology (Chalupnicki and others, 2016). Initial marking of cisco was done with OTC applied to the fish in an immersion bath but a successful mark could not be verified. Calcein, which impregnates bones with a dye that fluoresces green, was determined to be durable and was used in all subsequent years, except 2020 when the chemical was unavailable. OTC mixed in feed was successfully used in 2020 to mark bloater and cisco. There can be many combinations of treatments and sources of experimental fish, and additional marking options would allow for more complex experimental treatments. Dyes fluorescing other colors than calcein green have also been tested but exhibited lesser degrees of effectiveness (M. Chalupnicki, TLAS, written commun., 2014); more research is needed. Genetic samples have been collected from all cisco and bloater used to produce fish for experimental releases, and parentage analysis may be used to identify the progeny of any of these fish that are recaptured from Lake Ontario.

Other Coregonine Species

At least one other coregonine species, kiyi, has been extirpated from Lake Ontario but exists within the Great Lakes Region. Coregonus reighardi (shortnose cisco) was considered extinct until 2023 (Eshenroder and others, 2016; Ackiss, U.S. Geological Survey, oral commun., 2023). In 2021, a collaboration among USGS, the Little Traverse Bay Bands of Odawa Indians and Sault Ste. Marie Tribe of Chippewa Indians, FWS, and others resulted in successful spawning of kiyi collected from Lake Superior, but only hybrid crosses with bloater were successful. Some of those hybrids are being reared at TLAS for research on their genetics and fertility. Efforts continue to acquire kiyi and other Great Lakes coregonine species that might benefit from research done at TLAS and other facilities of the broader collaboration.