Success of two methods for long distance transport and fertilization of Bull Trout

ABSTRACT

Objective

The use of captive breeding programs for the conservation of freshwater fishes, be it for research or reintroduction, is becoming more common. Interspecific differences in rearing ecology make applying common rearing techniques difficult for threatened species. This becomes increasingly difficult for species in remote areas, as accessing fish for hatchery propagation can take longer. Bull Trout Salvelinus confluentus has recently been listed as threatened in Alberta and is garnering greater consideration for artificial propagation. Currently, there are no active hatchery programs for Bull Trout, and historically, very few attempts have occurred with limited success. The goal of this research was to establish a laboratory population of Bull Trout and compare the effectiveness of delayed and same day fertilization when transporting gametes or embryos over large distances.

Methods

Here, eggs were either fertilized in the field prior to transport (field) or 1 d after transport (green). Both treatments were collected from Smith-Dorrien Creek, Alberta, and transported to Winnipeg, Manitoba, by airplane (approximately 1,300 km). Differences between treatments were tested using a binomial generalized linear model.

Results

Fertilization rates were similar and high (>99%) for both treatments, and our model did not detect a significant treatment effect. Although a significant treatment effect was not observed, the green treatment had 10% lower survival over the duration of the experiment. The greatest mortality rates were observed at the swim-up stage in both treatments.

Conclusions

Both methods had similar fertilization rates and high survival to the fry stage (green = 36% and field = 46%). The results suggest that both methods are viable options when transporting Bull Trout gametes or embryos long distances for species recovery and research purposes.

INTRODUCTION

There is a long history of fish propagation, initially for consumptive purposes and recreation and more recently for the conservation of fish species and experimentation (Greenberg, 2010). Research studying the rearing of at-risk fish species for conservation purposes has become more common, particularly to recover and reestablish native populations (i.e., fish that were not introduced by humans or allowed to invade due to anthropogenic activities) that have declined due to various anthropogenic effects on their habitats (Halverson, 2008; Jackson et al., 2004). The creation and use of laboratory populations of fish that are propagated from native sources rather than domestic hatchery strains gives researchers the opportunity to better understand the ecology and physiology of the species for conservation purposes, as differences are frequently observed between domestic and wild populations (Scott et al., 2015). When selecting source populations for establishing native broodstocks in hatcheries for reintroduction, it is important to select sources that are similar (genetics and life history) to past populations (Dunham et al., 2011). In some circumstances, source populations for experimentation or restocking may be located in areas that are logistically challenging to access and/or have highly variable spawning times that do not allow gametes to be crossed among populations on the same date.

Salmonidae is a family of fishes that has a long history of propagation for food, recreation, and conservation purposes (Greenberg, 2010; Halverson, 2010). Much of this work has been focused on a few commercially and recreationally important species: most notably Atlantic Salmon Salmo salar and Rainbow Trout Oncorhynchus mykiss, resulting in their establishment in various coldwater habitats throughout the world (Greenberg, 2010; Halverson, 2010). An assessment of stocking rates suggested that Rainbow Trout represent the most dominant fish species among those that have been stocked by biomass in the United States between 1931 and 2004 (Halverson, 2008). Unlike Rainbow Trout, Bull Trout Salvelinus confluentus, which is a culturally and ecologically important salmonid, has received limited attention for artificial rearing (not present in a review of U.S. stocking from 1931 to 2004; Halverson, 2008). To date, few documented attempts have occurred to rear Bull Trout in captivity (Hayes & Banish, 2017; Kissinger et al., 2024) and even fewer studies have been published on the practices of egg-take and early rearing of Bull Trout (Fredenberg et al., 1995). Bull Trout are not typically considered for captive breeding for recreational or commercial purposes (Hayes & Banish, 2017; Kissinger et al., 2024), probably because they spawn in the fall in cold headwater streams that can be difficult to access and require a longer incubation period than spring spawners (Nelson & Partz, 1992). Bull Trout exhibit slow growth because they occupy cold low-productivity habitats (Selong et al., 2001), they are more vulnerable to stress from high temperatures (Selong et al., 2001) and lower oxygen than other salmonids, and larger juveniles and adults are territorial and piscivorous (Nelson & Partz, 1992), resulting in potential harassment and cannibalism when they are held in captivity. Additionally, within parts of their natural distribution, Bull Trout have historically been considered a less-desirable fish and in some cases have suffered intentional eradication measures (Alberta Sustainable Resource Development, 2012), which may further explain their absence as a commonly propagated fish species.

Although Bull Trout are not an ideal candidate for propagation in commercial or recreational hatchery facilities, their listing as a species at risk (Canada Gazette, 2019; U.S. Fish & Wildlife Service, 1999) in parts of their southern distribution in the USA and Canada has heightened the need for laboratory research and the potential for captive breeding (Alberta Sustainable Resource Development, 2012; Dunham et al., 2011). Obtaining gametes from Bull Trout often requires transport over large distances and long periods. Typical fish husbandry practices encourage the fertilization of eggs immediately or soon after removal, but in cases where access to hatchery facilities is limited, fertilized eggs or gametes can be transported to rearing facilities (Bobe & Labbé, 2009). Although the success of fertilization declines the longer gametes are held, no change in fertilization success was observed between 0 and 7 d after the removal of eggs from Rainbow Trout in a study by Carpentier and Billard (1978). Successful fertilization has been documented in Rainbow Trout up to 18 d following the removal of eggs, but fertilization success was very low (near 0, Babiak & Dabrowski, 2003). Although success was observed following long periods of holding at controlled temperatures for Rainbow Trout, differences in fertilization success have been observed among species (Carpentier & Billard, 1978; Jeuthe et al., 2016) and the gametes in most studies that have assessed long-term holding were kept in controlled environments that had minimal exposure to a range of possible transport-related effects (e.g., jostling, low dissolved oxygen, pressure changes, and temperature change; Bobe & Labbé, 2009).

To better understand the physiology of threatened Bull Trout, we set out to establish a laboratory population at the Department of Fisheries and Oceans in Winnipeg, Manitoba, by retrieving gametes from Smith-Dorrien Creek, Alberta (approximately 1,300-km straight line and approximately 20-h transport time). We also sought to determine whether there is a significant difference in survival rate between two propagation methods: (1) transporting field-fertilized hardened eggs (field) and (2) transporting unfertilized eggs and milt separately (green) over large periods and distance followed by fertilization in the laboratory. We assessed survival rate to multiple critical developmental stages, including fertilization (1 d postfertilization [dpf]), 50% eyed, 50% hatch, 50% swim-up, and fry at 342 dpf. We also compared accumulated thermal units (ATU) among treatments to examine potential differences in developmental rates at these developmental stages. Finally, we documented the husbandry practices and timing of the critical developmental stages for both treatments because this information is limited in the literature for Bull Trout. Understanding the process and potential differences in the effectiveness of these two propagation methods will provide staff with a greater suite of options when collecting gametes should additional time be needed for transport or fish from one location become ripe prior to another, where crosses would be beneficial (i.e., broodstock development).

METHODS

Experimental design

All Bull Trout were captured on September 15, 2021, and eggs and milt were harvested from six females and six males from Smith-Dorrien Creek. The gametes were fertilized via two methods (field and green, Figure 1).

For the green treatment, gametes were taken from three females (4,048 eggs) and three males on September 15. Unfertilized eggs were placed in 3.78-L Ziploc bags and approximately 3–5 mL of milt was collected from each male and shipped in 15-mL falcon tubes. The gamete containers were filled with pure O₂ and placed in a cooler with ice packs. The gametes were then shipped by airplane in a pressurized compartment (flight time approximately 2 h) and received at Fisheries and Oceans Canada Freshwater Institute (Winnipeg, Manitoba) on September 16 and fertilized that day, resulting in an 18–22-h delay from egg-take to fertilization. Temperature was monitored inside the cooler using hobo temperature loggers (Onset HOBO TidbiT v2 Water Temperature Data Logger, 124 UTBI-001) and gradually increased during shipment from 1.9°C to 6.0°C. Due to concerns with milt from one of the males (i.e., water in the vial), milt from only two males was used to fertilize the eggs from the three females. For both treatments, milt was combined prior to fertilization to increase the chances of fertilization and to randomize the within-treatment influence of parental effects. The eggs and milt were measured and fertilized using a ratio of 200 mL of eggs to 100 µL of milt cocktail.

For the field treatment, gametes were taken from three females (6,111 eggs) and three males on September 16. By taking and fertilizing the eggs on September 16, both treatments (field and green) were fertilized on the same day. The eggs were then water-hardened in water from Smith-Dorrien Creek that had been boiled and cooled to eliminate the chance of disease transfer through contaminated water (Wagner, 2002). The water was saturated to 100% dissolved oxygen before the eggs were water hardened by bubbling them in water from a cylinder of compressed oxygen. The percentage of saturation was monitored using a handheld multiparameter water quality instrument (YSI ProDigital ODO/CT, SKU 627150-1). The fertilized eggs were water-hardened for 2 h prior to movement and were placed in a 22.7-L collapsible container and then in a cooler with ice packs. The cooler was then shipped to Fisheries and Oceans Canada Freshwater Institute by air and received September 16, a duration of 10–11 h after the gametes were collected and fertilized in the field. Temperature was monitored inside the cooler using hobo temperature loggers (Onset HOBO TidbiT v2 Water Temperature Data Logger, UTBI-001) and gradually increased during shipment from 2.9°C to 3.5°C upon arrival at the Freshwater Institute.

Once in the lab, the eggs from both treatments were placed in a Marisource Heath Tray flow-through stacked incubation system with five trays dedicated to eggs (green: trays 1, 2, and 3; field: trays 4 and 5). For the green treatment, a single tray was used for each female to keep the lineages separate, and for the field, two trays were used to hold the embryos from the fertilized treatment, as they were mixed during shipment. Each tray was 82 × 60× 64 cm, and the eggs were placed in the trays between 1 and 2 eggs deep on a trout screen (mesh size 1.4 × 1.8 mm). Fertilized eggs were approximately 5.5 mm in diameter after water hardening. The system was flow through (freshwater inflow rate 5 L/min), but some water (flow rate 3 L/min) was recirculated to increase flow and act as a redundancy should power be lost to the main water source. Initial fertilization was not enumerated until 24 h postfertilization, where staff considered successful fertilization as eggs with the presence of embryo development. Mortalities from each tray were enumerated daily and were identified as opaque eggs. To assess mortalities once hatched, alevins were gently probed to assess their status if they were not actively moving. The timing of different developmental stages (e.g., fertilization, 50% eyed, 50% hatched, and 50% swim-up) were documented. Alevins from each treatment were kept separate, but trays within treatments were pooled together and transferred to larger flow-through tanks (volume 188 L/tank, green: tanks A and B; field: tanks C, D, and E; Figure 1) on March 15, 2022. An extra tank was used for the field treatment to accommodate the higher number of fish, and each tank was held at approximately 1,900 fish/tank. Mortalities from each treatment were monitored until August 23, 2022, when the tanks were combined and treatments were no longer discernible.

Accumulated thermal units

The temperatures for all treatments were increased to mimic natural changes that are observed in the wild. In general, the eggs were fertilized around 4°C and held at this temperature until 121 dpf, when both treatments were 50% hatched, at which time temperature was increased gradually to approximately 10°C over the next 70 d (Table 1; Supplement 1 [see online Supplementary Material]).

Temperature was monitored in the heath trays and tanks using hobo temperature loggers that were set for 10-min measurements throughout the duration of the experiment. The temperature measures were averaged to obtain the daily temperature of each tray. The sum of average daily temperatures from each tray and tank was used to calculate ATU for survival to different developmental stages for each treatment. The number of surviving individuals out of the initial number of inseminated oocytes and ATU were calculated to 342 dpf and compared among treatments.

Statistical analysis

The proportion of individuals surviving at the end of each developmental stage allowed us to explicitly model survival within each treatment group. Due to the small sample size, we were unable to employ more complex modeling approaches. We assumed independence of observations, although we recognized the presence of unaccounted batch effects between developmental stages.

To model the probability, π_i, of the proportion of surviving individuals at the end of each discrete developmental stage as a function of the covariates, we applied a binomial generalized linear model with a logit link function:

where “survived” was the number of inseminated oocytes that survived at the end of a developmental stage out of the initial number, n_i, at the beginning of the stage. The covariates “treatment” (categorical with two levels: field and green) and developmental “stage” (categorical with five levels [1 dpf, 50% eyed, 50% hatched, 50% swim-up, and fry]) were included as fixed effects. Interactions between these factors were included to obtain treatment- and stage-specific survival estimates. The 95% point-wise confidence intervals were estimated on the linear predictor scale and transformed back to the response scale using the inverse of the link function. The post hoc analysis included pairwise comparisons between treatment groups for each stage using the “emmeans” R package (Lenth, 2024). All statistics were completed in R Statistics (R Core Team, 2023).

RESULTS

Development

The fertilized eggs from the field treatment reached each respective developmental stage slightly earlier (1–3 d) than the eggs from the green treatment (Table 1). Eyed development was first observed at 60 dpf, and less than 3 weeks later, all eggs were eyed. Hatching was first observed in the field treatment 106 dpf, and by 120 dpf, both treatments had achieved 50% hatching success (Table 1). Hatching was completed by 124 dpf in both treatments, and 50% swim-up was achieved by 212 dpf and completed after 234 dpf (Table 1).

Accumulated thermal units

Temperature ranged from 3.5°C to 13.2°C, with an average of 7.6°C throughout the duration of the experiment (Supplement 1). At all developmental stages, the field treatment had higher average ATU (Table 1). However, we observed no statistical differences between the field treatment and green treatment for ATU with 50% eyed (p = 0.098), 50% hatched (p = 0.427), and 50% swim-up (p = 0.321). At the time of 50% swim-up, values for ATU were slightly greater in the field treatment (1,082 ATU green and 1,093 ATU field).

Survival

Both treatments successfully reared fish to the fry stage (Figure 2), but linear trends in mortality significantly differed between each developmental stage, suggesting a cumulative 10% difference among treatments by the end of this experiment (Table 2; Figure 2B). Despite a 10% difference at the end of the experiment, pairwise comparisons of fertilization rates were similar and high in both treatments (>99%; Table 3) and no significant linear treatment effect was observed (p = 0.074; Table 2), suggesting that the transportation and timing of fertilization method did not significantly influence initial fertilization and overall survival (Figure 2). The greatest decrease in survival was observed at the 50% swim-up stage, and the pairwise comparisons indicate that survival of the green treatment was significantly lower at all developmental stages except fertilization and 50% hatched (Figure 2A; Table 3).

DISCUSSION

As far as we know, this is the first study to show that Bull Trout gametes and embryos can be successfully raised to hatch following long-distance transportation by air and up to 22 h of holding. The data suggest that similar amounts of ATU are required for each developmental stage among treatments but also show that mortality was higher in the green treatment with 10% less survival by the end of the experiment than in the field treatment. Although higher mortality was observed, both treatments successfully reared large numbers of Bull Trout to the fry stage and our model did not detect a significant treatment effect. The high and similar fertilization rates between treatments suggest that both techniques are equally effective.

Development

To date, little is known about the development of Bull Trout in hatchery systems, as there are only a few examples of hatchery and laboratory rearing (Gould, 1987; Hayes & Banish, 2017; Kissinger et al., 2024). The extant information about Bull Trout development suggests that average ATU in our study, for both treatments from fertilization to 50% emergence, was approximately 300 ATU higher than has been described previously (Bowerman et al., 2014; Gould, 1987). Variability in ATU within a species can be large (e.g., 650–1,440 ATU, for Chum Salmon Oncorhynchus keta; Merritt & Raymond, 1983), indicating that some level of plasticity or local adaptation is present within a species. The finding that Bull Trout from Smith-Dorrien Creek require higher ATU than populations that are found at more southern latitudes may be driven by regional differences in climate, where winters are longer and colder relative to southern range extents (Bowerman et al., 2014; Gould, 1987). Spawning and swim-up phenology can vary among populations (Austin et al., 2019; Bowerman et al., 2014), which is thought to be governed by differences in water temperature, and subsequently, ATU, where fish occupying colder streams typically spawn earlier (Austin et al., 2019). When four genetically differentiated strains of Brook Trout were brought into a laboratory and reared at the same temperatures, significant differences in degree-days to hatch were observed among strains, which was believed to be linked to local adaptation (Baird et al., 2002). Based on the geographic differences observed between the Bull Trout population in Smith-Dorrien Creek and other Bull Trout populations referenced prior, it is likely that there are genetic differences, as population differentiation within Bull Trout has been observed at much smaller scales (Warnock et al., 2010). Although they were not tested in this study, these genetic differences may be linked to observed differences in ATU among populations and warrant further investigation.

The presence of high mortality rates in both treatments during swim-up was not surprising because transitioning from endogenous to exogenous nutrient sources can be challenging for fish, especially when they are artificially reared (Jeuthe et al., 2015). The survival to swim-up that we observed in our experiment was lower than survival rates in hatchery systems (50–80%, Essington et al., 2000; Hilborn, 1992). However, this is not surprising, as Arctic Char Salvelinus alpinus, which are close relatives to Bull Trout, have similarly low survival in hatchery systems (Jeuthe et al., 2016), which could be linked to a preference for development at colder temperatures or difficulties transitioning to exogenous food sources (Jeuthe et al., 2015). Although survival to the fry stage in this Bull Trout population was lower than that of many hatchery-strain salmonids, an average survival of 41% for these two methods to 342 dpf suggests that both methods that were employed here represent a viable option for rearing Bull Trout in captivity. Should this become a common approach for species recovery, improvements to the swim-up and fry stage would have the greatest benefits to overall survival.

Managing at-risk species

Demonstrating that Bull Trout populations can be reared in captivity following long-distance transport opens the door for not only laboratory experiments, like those described here, but also the possibility of creating broodstocks for restoration purposes. Transport and holding of gametes for up to 22 h with nearly 100% fertilization success for the green treatment is highly encouraging because it allows users to keep milt and eggs separate if in situ fertilization is not possible. The ability to keep gametes separate until arrival in a laboratory facility and hold gametes for extended periods offers users the flexibility to conduct various parental crosses, even in situations where local populations spawn at different times (e.g., Austin et al., 2019). This can be especially important when attempting to incorporate as much genetic diversity as possible into a recovery composite broodstock from various sources (Whiteley et al., 2015). In addition, as many Bull Trout populations receive higher levels of disruption in areas that are near population centers, many of the high-density stocks that would be suitable donors are found in remote locations that are challenging to access (requiring long hikes and/or helicopter access).

In Alberta, there are four provincial hatcheries and three are within the historic distribution of Bull Trout. All three hatcheries within the historic distribution of Bull Trout are in the southwestern corner of the province and would be approximately 550 linear km from the most northern extent of Bull Trout within the province of Alberta, creating challenges for potential collection and transport of gametes or embryos. Alternatively, Alberta Provincial Fisheries biologists are testing different methods for rearing native salmonids at sites (i.e., Westslope Cutthroat Trout Oncorhynchus lewisi), including remote streamside incubation chambers and portable incubation trailers (B. Meagher, Alberta Environment and Protected Areas, personal communication). However, with fall-spawning species like Bull Trout, winter freezing presents challenges for remote streamside incubation chambers (as are used for Westslope Cutthroat Trout) and portable incubation trailers. Therefore, Alberta Environment and Protected Area staff are currently supporting a research initiative at the University of Calgary to assess the feasibility of egg incubation chambers for Bull Trout that are buried in the sediment (methods similar to those of Baxter & McPhail, 1999). The results from this research indicate high levels of survival to hatch (71%) and that this method could be a viable alternative to hatchery propagation for fall-spawning species in cold climates (Lepine, 2024).

Another risk when transporting fish is disease transfer through transporting live fish and/or water. This is of particular concern with the recent documentation of whirling disease in Alberta (Canadian Food Inspection Agency, 2019). Consequently, the water for the field treatment from this study had to be decontaminated prior to use and transport (Wagner, 2002). Transporting gametes without the need for water decreases the probability of spreading whirling disease because gametes are not susceptible to infection (Markiw, 1991). Showing that both methods had similar fertilization rates suggests that the green method can be used, which reduces the risk of transferring aquatic diseases to fish and between locations. Although risks of disease transfer are reduced using the green method, other factors must be considered prior to moving fish or gametes and practitioners should consider existing guidance when undertaking relocations (e.g., Dunham et al., 2011).

Conclusion

We show that the green and field methods for transporting Bull Trout embryos and gametes over long distances were equally effective at initial fertilization, no overall treatment effect was observed, and both approaches successfully reared large numbers of fish to the fry stage. Improvements to rearing techniques after the swim-up stage would increase overall survival rates, and we suggest that further research and considerations focus on these stages of development. We show that both approaches provide viable options for research and species recovery actions for at-risk Bull Trout populations.

SUPPLEMENTARY MATERIAL

Supplementary material is available at North American Journal of Aquaculture online.

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding author upon reasonable request.

ETHICS STATEMENT

All fish were treated in accordance with the guidelines of the following permits as required by the Government of Canada to transport and conduct research on aquatic species at risk. Alberta Environment and Parks (Research License #21-0217RL; Research and Collection Permit #21-384), DFO Freshwater Institute Animal Care Committee (Animal Use Protocol #FWI-ACC-2021-66), Canadian Food Inspection Agency (Domestic Movement Permit #D-2021-119874), Government of Manitoba (Live Fish Handling Permit #34097479 LFH-ITP 25-2021), and Species at Risk program (Species at Risk Act Section 73 Permit #21-PCAA-00048).

FUNDING

Funding was provided by Fisheries and Oceans Canada (DFO) through the Program and the Forestry Resource Improvement Association of Alberta (FOOMOD-01-030).

ACKNOWLEDGMENTS

The authors thank T. Rudolfsen, M. Teillet (DFO Edmonton), G. Eisler, A. Robinson, and H. Keyes (AJM Environmental Inc., Calgary) for assistance with planning logistics, fieldwork, and fish capture. John Enns (Calgary) provided valuable knowledge and experience collecting and shipping gametes/fertilized eggs. A. Spiteri, C. Scotland, and J. Symes (Alberta Environment and Parks, Kananaskis District) assisted with field navigation, planning logistics, and safety. L. Peterson and multiple volunteers from Trout Unlimited Canada (Calgary) provided field assistance with redd surveys. K. Wautier and T. Durhack (DFO Winnipeg) received and fertilized gametes in Winnipeg and provided support with the design and preparation of incubation units. M. Panciera and B. Romaniuk (DFO Winnipeg) helped care and raise fish after hatching. Thanks to T. Winter (Alberta Environment and Protected Areas) and R. Bennett (DFO Winnipeg) for help with review.

REFERENCES

Author notes