Research Paper

INTRODUCTION

Strategies used for evaluating the reintroduction of an extirpated species include tracking change in abundance and distribution (Dziminski et al. 2021). These metrics are generally the easiest to collect (compared to demographic rates), may involve active monitoring programs (Roque et al. 2021) or passive or collaborative monitoring programs (e.g., Sullivan et al. 2009), and they provide a good method for measuring population health. An improved strategy for assessing reintroduction includes identifying and measuring factors affecting population dynamics (Hunter-Ayad et al. 2020). Understanding how population demographic rates (e.g., reproduction and survival) are impacted by drivers of population growth rate is also important for assessing population health (Larson et al. 2004, Hunter-Ayad et al. 2020). Analysis of demographic rates helps to identify causes of change in distribution and population growth rate and can signal if the population is reaching habitat carrying capacity. Understanding demographic rates may also indicate if other conditions (e.g., inter-species interactions, dispersal issues, changes in habitat suitability, disease, contaminants, etc.) need to be investigated to assess potential barriers to a successful reintroduction.

The Trumpeter Swan (Cygnus buccinator) is the largest of North America’s waterfowl (Anatidae), and the species was close to extinction in the early 1900s (Banko 1960). A remnant population in western North America (Palmer 1976) was the source for reintroduction in the states and provinces of interior North America (Moser 2006). The continental reintroduction of the Trumpeter Swan, with translocations starting in the late 1930s (Banko 1960), is generally considered a success (Groves 2012, Handrigan et al. 2016) with the establishment of self-sustaining breeding populations that increased (Groves 2012, Badzinski and Earsom 2015) and spread throughout parts of what is believed to be their former breeding range (Banko 1960, Lumsden 1984, Gale et al. 1987, Handrigan et al. 2016). The release of these swans in Ontario, Canada started in 1988 and ceased in 2006 as the population was deemed to be self-sustaining through wild reproduction (Lumsden 2008, Intini 2009). Abundance has increased in south-central Ontario with continuing reproduction of descendants of the released captive-bred birds and in northern and western Ontario via immigration from restored populations in the northern United States (Minnesota, Wisconsin, Michigan; Lumsden 2008, Badzinski and Earsom 2015, Handrigan et al. 2016). A large proportion of south-central Ontario Trumpeter Swans do not migrate long distances but instead make relatively short distance migration movements, mostly within Ontario (Handrigan et al. 2016).

Among the North American Anatidae (Bellrose 1980, Baldassarre 2014, Koons et al. 2014), Trumpeter Swans take the longest to sexually mature, forming pair bonds after they are least 20 months old (Mitchell and Eichholz 2020) and breed for the first time when they are 4 to 7 years old (Banko 1960, Wilmore 1979). This species is therefore relatively slow to reach the carrying capacity of their habitat and populations may not exhibit as rapid growth as other Anatidae. Nevertheless, a species observed to increase in abundance and one that is expanding spatially into its former breeding and wintering range should eventually reach carrying capacity.

As populations approach carrying capacity, recruitment and productivity often decline. For relatively slow growing species like Trumpeter Swans, reduced productivity and recruitment may occur through several potential processes. A decline in nest survival, egg hatching success, or number of eggs laid may occur (Mitchell and Eichholz 2020) if swans nest in sub-optimal habitats. Density dependence in winter (Shea et al. 2002) might also affect productivity (Sedinger and Alisauskas 2014) where sub-optimal nutrition for breeding adults affects reproduction the following spring through a carry-over effect (e.g., Fowler et al. 2020). Reduced cygnet survival may also occur if sub-optimal habitats limit the quantity and quality of food that cygnets need for growth (Knudsen et al. 2002), or if predation of cygnets is higher in such habitats. Although not an exhaustive list, these are mechanisms observed for other Anatidae that were near their habitat’s carrying capacity (e.g., Canada Geese [Branta canadensis]: Brook et al. 2015, Snow Goose [Anser caerulescens]: Alisauskas et al. 2022, and Mallard [Anas platyrhynchos]: Kaminski and Gluesing 1987).

A reduction in recruitment will likely occur when available preferred breeding territories are filled. Territory size depends on habitat characteristics (Mitchel and Eichholz 2020), and territories are rigorously defended by breeding pairs (Banko 1960, Hensen and Cooper 1992). Trumpeter Swan pairs may attempt to breed in sub-optimal habitats as preferred breeding sites become filled by other pairs or other species (e.g., Mute Swan [Cygnus olor]: Lumsden 2016 or Canada Geese: Henson and Cooper 1992). Additionally, because of the relatively slow life-history characteristics of Trumpeter Swans, changes in the stable age distribution (i.e., a reduction in the proportion of breeding aged individuals) may cause a reduced population growth rate, more so than for other Anatidae (Cooch et al. 2014, Koons et al. 2014).

Our objectives were to assess whether the south-central Ontario population of Trumpeter Swan is at or near carrying capacity, and to determine the effects of endogenous factors on population growth rate and juvenile survival. This information is important for assessing if further conservation measures are needed. We hypothesize that the south-central Ontario Trumpeter Swan population will exhibit a slowing growth rate as it reaches carrying capacity. If the population has reached or exceeded carrying capacity, we expect that growth rate would decline. We predict that density dependence will manifest as a decline in recruitment of juveniles into the breeding population as it has with other large Anatidae (e.g., Williams et al. 1993, Nummi and Saari 2003, Aubry et al. 2013, Wood et al. 2016). We estimated demographic rates using an Integrated Population Model (IPM) that combined relative abundance, productivity, and mark-resighting data (Arnold et al. 2018, Plard et al. 2019, Nuijten et al. 2020). Our analysis has the added advantage of allowing posterior estimation of abundance trends by age or breeding class as latent parameters (not directly observable) to provide further insight into Trumpeter Swan population dynamics. We aimed to capture the age structure complexity inherent in a species with a longer life span and generation time (Koons et al. 2014). We used the integrated model to forecast population size for an additional 15 years for the south-central Ontario Trumpeter Swan population.

METHODS

Data sources

Ontario Trumpeter Swan reintroduction (1982 to 2006) was led by the Ontario Trumpeter Swan Restoration Group (OTSRG), a non-government organization of mostly volunteer members (Handrigan et al. 2016) now named Trumpeter Swan Conservation Ontario (TSCO, since 2022). In each year of the restoration program and continuing since, a portion of Trumpeter Swans in south-central Ontario were marked and banded (Lumsden and Drever 2002). Swans were banded using U.S. Fish and Wildlife Service standard aluminum butt-end leg bands or lock-on type legs bands. Each banded bird was also equipped with a yellow patagial tag (with unique alpha numeric code; Lumsden 2008) attached prior to release. Though attachment of a patagial tag is considered moderate-high risk for some species (Trefry et al. 2013, VKM et al. 2024), we did not find, through our own experience, nor did we find evidence in the peer reviewed literature that patagial tags cause a bias in Trumpeter Swan demographic rates. TSCO tested patagial tags on captive reared Trumpeter Swans prior to use on wild swans and swans did not appear to be hampered or harmed by them. Patagial marked Trumpeter Swans have been regularly observed breeding and rearing cygnets with no observed issues for the last 15 years of the project; however, discarded fishing gear or other material entangled on a patagial tag may cause negative effects for survival or production, though we feel those effects are minimal. Wild-hatched swans were captured, usually in winter, at feeding sites where individuals were lured with food and hand captured for marking and banding (Lumsden et al. 2012). Observations of marked swans were made by a network of volunteers and the public. Observations were also compiled and added to the database from those reported at online internet sites (iNaturalist, The Trumpeter Swan Society, TSCO, Wye Marsh), on eBird (eBird 2021), and on social media sites including Facebook and Instagram (G. Lane, personal communication). Repeat observations from the winter (November to March) were reduced to one per bird annually for model estimates of annual survival. All observations of marked birds were compiled into a database and linked to associated capture information for each bird. We limited our analysis to data from 2006 to 2019 because of insufficient banding and resightings prior to 2006 and in 2021 and 2022. This ensured that there were sufficient annual data for estimating parameters within a survival analysis.

We used Christmas Bird Count (CBC; National Audubon Society 2021) data for Ontario, corrected for effort (Soykan et al. 2016), which can be interpreted as relative abundance (i.e., annual estimates relative to one another but not true abundance). CBC relative abundance measures were used to estimate early winter abundance of south-central Ontario Trumpeter Swan data for the period of interest. In brief, relative abundance data are calculated as the log of expected counts as a function of the count circle (i.e., the 24 km area of each count), year, the total count hours and an overdispersion parameter (see Link et al. 2006, Sauer and Link 2011, and Soykan et al. 2016 for details).

We estimated annual productivity of the Trumpeter Swan population using data recorded by the OTSRG each year, including the number of eggs laid, cygnets hatched, and fledged (observed flying) for a sample of nest sites. In our models, we used the mean annual number of cygnets fledged per breeding pair as an estimate of productivity. These data account for breeding pairs that lost their entire clutch (i.e., failed breeding).

Integrated model

We used annual relative abundance data, mark-resight data, and a productivity index in an integrated population model to estimate demographic traits of the population. The integrated population model links the three datasets allowing improved inference of demographic parameters as opposed to modeling each alone. First, we developed an age/breeding-stage structured, female-only model describing the annual life cycle of Trumpeter Swans (Fig. 1). We built the model with pre-breeding relative abundance data (CBC conducted three months prior to nesting), tracking each age and breeding-stage in an annual time step. We assumed a female-only model because there is no evidence that males are limiting in the population. We also assumed that females do not attempt breeding before 3 years of age (Mitchell and Eichholz 2020). We assumed that cygnets survived at a lower annual rate than older birds based on a priori model fitting of candidate models that included different age-class structures and from previous swan research (Watola et al. 2003, Wood et al. 2018). Each two-year-old (2-yo) Trumpeter Swan has a probability of either becoming a breeding adult (ψ_3yo-br) or non-breeding 3-yo (1 - ψ_3yo-br) at each time step. Similarly, a non-breeding adult (3-yo+) has a probability of either becoming a breeding adult (ψ_nb-br) or staying a non-breeding adult (1- ψ_nb-br).

The relative abundance data do not provide any information about age, sex, or breeding-stage structure of the population. Therefore, estimates of abundance for these latent (unobservable) parameters (i.e., 1-yo, 2-yo, and breeding and non-breeding adults) were made by fitting data to a structured population model (Fig. 1) through a joint likelihood function. Using a hierarchical approach, we linked models describing the observations with the state (biological) process (Fig. 2). To model the temporal dynamics of abundance and productivity indices, we used recovery histories of the mark-resight data in a Cormack-Jolly-Seber (CJS) model (Cormack 1964, Jolly 1965, Seber 1965) integrated within a state-space framework.

Banding and resighting history summaries were arranged into an m-array (matrix) for analysis (Kéry and Schaub 2012). Prior to including a CJS model in our integrated model, we assessed CJS candidate sub-models using program MARK (Cooch and White 2019) to determine the most parsimonious model. We used data from both sexes together because we did not find evidence of sex related differences in survival or resighting probability (see also Lumsden and Drever 2002). The addition of male data in the CJS analysis increased sample size and parameter precision. We parameterized the survival and resighting probabilities in a two-age class models (juveniles and adults), and we did not find parsimonious evidence for additional age-class (2+ age classes) structure (see also Lumsden and Drever 2002). However, a small bias may exist in apparent female survival if there was a difference between sexes or if there was additional age structure that we could not account for. We did not model survival (or resighting probability) differently for captive-reared and wild-hatched Trumpeter Swans in Ontario as Lumsden and Drever (2002) found no difference between them.

Demographic estimates

We defined a state-process (Brooks et al. 2004) to describe the population trajectory with latent parameters for 1-yo, 2-yo, breeding-adult, and non-breeding-adult abundance (N₁, N₂, N_br, N_nb, respectively) for each year (t). The number of cygnets fledged per breeding female (ρ) was multiplied by 0.5 to reduce to females only, assuming an even sex ratio (we found the sex ratio was not significantly different from even in this study: 1.01 M:F for cygnets, binomial P = 0.027). The annual total number of cygnets fledged (TF_t) was estimated from total broods (TB_t) and the average number of female cygnets per brood (ρ_t) using a Poisson distribution (Eq. 1).

(1)

Using the m-array (m) format, we employed a multinomial likelihood to estimate specific apparent survival (ϕ) and resighting probability (p) for two age classes: juvenile (ϕ_juv, p_juv) and adults (ϕ_ad, p_ad). We estimated annual apparent survival probability (ϕ_t) for juveniles and adults using a random effect (Barry et al. 2003) and the CJS parameterization. We used a time-dependent, age-class, multistate model with a state-space formulation made up of state and observation processes.

We modeled N₁ (Eq. 2) and N₂ (Eq. 3) abundance with a Poisson distribution allowing for unbounded integer values. We included adult survival (adjust for three months: ϕ_{ad, t}^3/12) to bridge the survival of adult breeding Trumpeter Swans from the survey (end of December) to the start of breeding in April:

(2)

(3)

To estimate abundance of breeding- and non-breeding-adults (Eqs. 4–7) we used a binomial distribution with a probability and size n, with values between zero (if no individuals survived) and the maximum number of individuals the year before (if all individuals survived).

(4)

(5)

(6)

(7)

The relative abundance (Y_t, from the CBC estimate rescaled to help improve convergence) did not have to be reduced to females-only. Relative abundance was modeled using a normal distribution with error (σ²), where σ² includes the residual error and is made up of both lack of fit of the process model and observation error (Eq. 8 and 9).

(8)

(9)

The joint likelihood (L_IPM) of the complete model is the product of the survival (L_cjs), state-space (L_ss), and productivity (L_pr) likelihoods (Eq. 10):

(10)

Implementation of the IPM

We analyzed the data under a hierarchical Bayesian framework using Markov Chain Monte Carlo (MCMC) methods with JAGS 4.3.0 (Plummer 2003), in the R computing environment (R Core Team 2021) with the JagsUI package (Kellner 2019). We used three chains with a burn-in of 5,000 iterations, thinning 100,000 iterations by retaining every fifth iteration for posterior estimates. We assessed parameter convergence, where Ȓ ≤ 1.1 (potential scale reduction factor) indicated convergence (Gelman and Hill 2007). Unless stated otherwise, we report means and 95% credible intervals of posterior estimates.

We used uniform non-informative priors for each of the survival and recovery probabilities (logit linked) and their errors. We log transformed and bound productivity from a uniform prior distribution (0 to 7). The process error (σ²) prior was estimated from a uniform distribution (0 to 30).

Using methods from Koons et al. (2017), we calculated elasticities for demographic rates, i.e., the sensitivity of changes in growth rate to proportional changes in a demographic parameter, as an extension of the absolute change: θ_i,t/λ_t × მλ_t/მθ_i,t. This provided a method for comparing sensitivities for parameters at different scales. We also performed a transient life table response experiment (LTRE; Koons et al. 2016, 2017) to determine how changing vital rates and population structure contributed to variation in the realized population growth rate. Elasticities and LTREs account for the feedback between population structure, demographic rates (Tuljapurkar 1990), and the short-term transient dynamics (Koons et al. 2016). We rescaled LTRE contribution as a proportion of the total.

We tested IPM Goodness of Fit (GoF) following recommendations of Schaub and Kéry (2022) for testing GoF for each sub-model individually. We used a Freeman-Tukey statistic (Brooks et al. 2002) to test GoF of the CJS sub-model. For the productivity sub-model, we used a dispersion test (Schaub and Kéry 2022) to determine if the data were over-dispersed Poisson by comparing the ratios of variance to the mean between the observed and simulated (expected) data. To test GoF for the state-space sub-model, we compared mean absolute percent error between observed and expected abundance.

To forecast total Trumpeter Swan abundances for the south-central Ontario population, we ran models for each of 15 subsequent years by substituting data from the end of the data informative period (i.e., 2019) forward to 2034.

RESULTS

From 2006 to 2019, 714 adult (mean = 51 per yr, SD = 14.6) and 451 juvenile (mean = 32 per yr, SD = 7.9) Trumpeter Swans were banded and marked with patagial tags in south-central Ontario. The sex ratio (M:F) was 0.96 for adults and 1.01 for juveniles. During the same period there were 13,715 (mean = 1,055 per yr, SD = 571.7) and 20,008 (mean = 1,539 per yr, SD = 789.9) resightings of those birds, for each age class, respectively.

The IPM model converged with a potential scale reduction factor < 1.1 (Ȓ ≤ 1.002). Results for GoF tests were variable. The count data fit the model well (P = 0.52) as did the productivity data (P = 0.64). Mark-recapture data did not adequately fit the CJS model (P < 0.01) though over-dispersion was deemed acceptable (median ĉ = 1.33; Lebreton et al. 1992) when candidate models were compared. The global model (ϕ_{t age}, p_{t age}), where apparent survival (ϕ) and resighting rates (p) were modeled with full age-class (age) and time dependence (t), converged, providing estimates for all parameters expected. Additional parameters for estimating more age class structure did not improve model GoF, so we used the global model and included random effects (Burnham and White 2002) in the IPM to ameliorate potential unmodeled overdispersion.

Average posterior adult apparent survival (0.84, 95% CL = 0.83 to 0.86) was larger and less variable than juvenile apparent survival (0.80, 95% CL = 0.75 to 0.85, Fig. 3). For productivity data, the average number of breeding pairs monitored per year was 42.6 (SD = 21.2) and the mean number of cygnets fledged per breeding pair observed was 2.85 (95% CL = 2.70 to 3.01).

Posterior estimates of abundance for female age and breeding classes all showed an increasing trend, and the estimated total number of female swans in the model generally followed the same increasing trend as the CBC data, but with a smoother trajectory (Fig. 4). The estimated abundance of juveniles and nonbreeding adults outnumbered the abundance of breeding adults by an average ratio of 2.6:1 (SD = 1.28). The estimated population growth rate, though annually variable, was positive in all but one year (mean lambda = 1.16, 95% CL = 1.13 to 1.20, Fig. 5).

The posterior probability of a 3-yo pre-breeding female becoming a breeding adult (mean ψ_3yo-br = 0.07, 95% CL = 0.01 to 0.15) was considerably lower than the probability of a 4-yo or older female transitioning from non-breeding adult to breeding adult (mean ψ_nb-br = 0.53, 95% CL = 0.26 to 0.92).

Population growth rate (λ) was most strongly affected by changes in adult survival (elasticity = 0.80, 95% CL: 0.78 to 0.32, Table 1) followed by changes in breeding adult abundance (elasticity = 0.35, 95% credible limits: 0.33 to 0.37). The LTRE suggested that adult abundance and survival both had unambiguous impacts on population growth (i.e., adult breeder abundance and survival did not include zero in the credible limits; contribution = 0.857, 95% credible limits: 0.411 to 0.866, and 0.437, 95% credible limits: 0.254 to 0.511, respectively, Table 1). All other vital rates and life stage abundances had a considerably lower impact on growth rate and were ambiguous.

To assess a possible density dependent effect, we compared the posterior mean juvenile apparent survival with both the posterior mean number of female breeding adult swans (effect = -0.0001, SE = 0.0002, P = 0.80) and the posterior mean total female swans (effect < 0.000, SE < 0.000, P = 0.82, Figs. 6 and 7, respectively). Although the linear relationship was positive in both cases, the effect size was small and ambiguous. We also found no apparent trend in annual growth rate (year effect = 0.014, SE = 0.006, P = 0.05, Fig. 5).

We forecast total relative abundance of Trumpeter Swan females in south-central Ontario without applying any density related growth restriction. Relative abundance is forecast to rapidly increase from a modeled estimate of 317 (95% CL = 286 to 347) female swans in 2019 to a modeled estimate of 3487 (95% CL = 958 to 7309) in 2034; an increase of 11 times (Fig. 8).

DISCUSSION

The Trumpeter Swan population in south-central Ontario is increasing in abundance with evidence that its distribution is also expanding (Handrigan et al. 2016; G. Lane, personal communication). We used relatively broad measures to assess potential density-dependent effects, and we assessed hypotheses about potential feedbacks acting through recruitment. We found no trend in annual growth rates and did not find any unambiguous negative feedback from measures of abundance affecting survival of juveniles. Modeling density-dependent limitations on growth rate of the underlying mechanisms could help improve precision on our forecasted growth trajectory; however, there was no indication from our analysis that limitation is occurring at the population level during the period assessed. Continued expansion and dispersal into unoccupied habitat may mitigate density-dependent effects acting through habitat limitations on the current breeding range. Until breeding habitat (or winter habitat) becomes limiting, feedback of density-dependence may not be detectable at the population level using the metrics or methods that we used. The population will not continue to grow indefinitely, of course, but it appears as though population growth will likely continue provided vital rates remain in the ranges estimated and habitat for population growth and range expansion are available. The forecast of population size beyond the period of empirical data suggests increasing abundance for the next 15 years, though the projected rate of increase is increasingly uncertain toward the end of that period.

A potential factor that may increase the rapidity at which density-dependence is reached could be simultaneous increases in abundances of allospecifics with overlapping breeding ranges that may compete for territory and resources on breeding and wintering areas. Mute Swans appear to be spreading northward from breeding areas in coastal marshes along the north shores of Lake Ontario and Lake Erie (Petrie and Francis 2003, Badzinski 2007, Badzinski and Wood 2025) with increasing potential for breeding range overlap with Trumpeter Swans. However, Lumsden (2016) suggested that Trumpeter Swans may be behaviorally dominant to Mute Swans based on observation of winter interactions and anecdotal and opportunistic observations of breeding site competition. We suggest that as quality breeding locations become limited with increasing abundance of both species, it is unclear what factors (e.g., age and breeding experience of competitors, nest initiation timing and first establishment of breeding territory, experience at a particular site, food availability, etc.) will determine the victor in the competition for breeding sites and how those will contribute to the future growth and distribution of each species. Similarly, the increasing abundance of temperate breeding Canada Geese (Branta canadensis maxima) in south-central Ontario (Iverson et al. 2014), a species that was once extirpated and since has been reintroduced to the region (Dennis et al. 2000), may increase competition for limited breeding territories in the region. The interaction between these species with Trumpeter Swan breeding pairs and the limited available swan breeding habitats in south-central Ontario may expedite the effects of density-dependence for all three species and should be monitored.

We modeled this Trumpeter Swan population (i.e., age structure, recruitment, survival, and transition rates) to reflect the characteristic life history that swans typically exhibit rather than simplifying the model structure. Despite the relatively long period for adults to reach sexual maturity, and with most of the population being in the non-breeding class at any one time, the population had an average annual growth rate of ~16% (with a doubling time of about four or five-years). This growth rate is comparable to the overall growth rate (14.4%) reported for the interior Trumpeter Swan population (Groves 2017, Mitchell and Eichholz 2020) and the south-central Ontario Trumpeter Swan population is estimated to be growing slightly faster than Mute Swans breeding in the same region (10% with a doubling time of seven to eight years; Petrie and Francis 2003).

The Pacific Coast population of Trumpeter Swans appears to be growing more slowly (5.5%; Groves 2017). The Pacific Coast population breeds predominately in forested wetlands of Alaska, and they may be approaching their breeding ground carrying capacity (Groves 2017). However, Schmidt et al. (2009) suggested that the Alaskan breeding population may not be at carrying capacity and that their estimate of growth rate (5.9% for 1968 to 2005) was relatively high for a long-lived species like Trumpeter Swan. The Rocky Mountain population of Trumpeter Swan growth rate averaged 6.5% (1968 to 2015; Groves 2017, Mitchell and Eichholz 2020) but appears to vary between flocks. This difference may, in part, be because of density dependent factors. Most of this population relies on natural rather than agricultural foods in winter so may be more susceptible to density dependent effects during this period where winter habitat quality and availability is a conservation concern (Pacific Flyway Council 2017).

As is typical of other species with similar life-history traits (Koons et al. 2014), growth rate for Trumpeter Swans is most sensitive to changes in adult female (breeding) survival. For this population, we found that apparent adult female survival estimates appear relatively high and stable. However, mechanisms that have the potential to increase adult mortality (e.g., Highly Pathogenic Avian Influenza [HPAI], lead poisoning, injury due to ingestion of discarded or lost fishing tackle, etc.) need to be monitored as they may cause a decline in population growth if they become increasingly prevalent. Supplemental winter feeding of Trumpeter Swans by private individuals occurs at certain locations in south-central Ontario, also a relatively common management practice for reintroduced swan populations (Shea et al. 2002, Slater 2006). The influence that winter feeding has on apparent survival of the south-central Ontario population or its role in mitigating potential density dependent effects is not known. Winter feeding of swans may reduce within-winter movements and may therefore increase survival rates of swans (Gillette 2005) by limiting exposure to migration hazards. Handrigan et al. (2016) more fully explored the potential impact of winter feeding for reducing migratory distance, concluding that it is not clear what the impact of discontinuing winter feeding would be on swan demographic rates. Varner and Eichholz (2012) suggested that winter feeding is no longer necessary for conservation of the Trumpeter Swan population they studied in north-central United States. Winter mortality risk for Trumpeter Swans because of severe weather events (Shea et al. 2002), or disease from overcrowding during winter (Slater 2006), may have population level effects if the additional mortality is sufficient to influence overall adult survival. Avoiding artificially concentrating swans and encouraging a migratory tradition may be important for conservation of the Interior Trumpeter Swan population (Slater 2006, see also Handrigan et al. 2016), especially with the prevalence of diseases like HPAI in North America (Ramey et al. 2022, Giacinti et al. 2023). However, inducing a migration tradition to new wintering areas imparts its own additional mortality risks including the potential for increased lead poisoning, disease, incidental kill by hunting, and power line collisions (Slater 2006).

Estimating seasonal survival probabilities could help determine population growth limitations. Winter carry-over effects on productivity and winter survival might limit this population’s growth before the availability of suitable breeding territories become limiting. If the current growth rate persists, the south-central Ontario population may require increasing management to avoid negative human and property interactions. Continued research on the drivers of vital rates, migration and movements, increasing or decreasing population densities, and geographic distribution changes will help direct future management action and ensure that the Ontario Trumpeter Swan population remains healthy.

LITERATURE CITED

Alisauskas, R. T., A. M. Calvert, J. O. Leafloor, R. F. Rockwell, K. L. Drake, D. K. Kellett, R. W. Brook, and K. F. Abraham. 2022. Subpopulation contributions to a breeding metapopulation of migratory arctic herbivores: survival, fecundity and asymmetric dispersal. Ecography 7:e05653. https://doi.org/10.1111/ecog.05653

Arnold, T. W., R. G. Clark, D. N. Koons, and M. Schaub. 2018. Integrated population models facilitate ecological understanding and improved management decisions. Journal of Wildlife Management 82:266-274. https://doi.org/10.1002/jwmg.21404

Aubry, L. M., R. F. Rockwell, E. G. Cooch, R. W. Brook, C. P. Mulder, and D. N. Koons. 2013. Climate change, phenology, and habitat degradation: drivers of gosling body condition and juvenile survival in Lesser Snow Geese. Global Change Biology 19:149-160. https://doi.org/10.1111/gcb.12013

Badzinski, S. S. 2007. Mute Swan. Pages 64-65 in M. D. Cadman, D. A. Sutherland, G. G. Beck, D. Lepage, and A. R. Couturier, editors. Atlas of the breeding birds of Ontario, 2001–2005. Bird Studies Canada, Environment Canada, Ontario Field Ornithologists, Ontario Ministry of Natural Resources and Ontario Nature, Toronto, Ontario, Canada.

Badzinski, S. S., and S. D. Earsom. 2015. The status of Trumpeter Swans in Ontario, 2015: results of the fifth North American Trumpeter Swan survey. Canadian Wildlife Service and Ottawa, Canada and U.S. Fish and Wildlife Service, Laurel, Maryland, USA.

Badzinski, S., and R. Wood. 2025. The status of Mute Swans in Ontario, 2024. Environment and Climate Change Canada, Ottawa, Ontario, Canada.

Baldassarre, G. A. 2014. Ducks, geese and swans of North America, Volume 1. Johns Hopkins University Press. Baltimore, Maryland, USA. https://doi.org/10.56021/9781421407517

Banko, W. E. 1960. The Trumpeter Swan: its history, habitats, and population in the United States. North America Fauna no. 63. U.S. Fish and Wildlife Service, Washington, D.C., USA. https://doi.org/10.5962/bhl.title.86963

Barry, S. C., S. P. Brooks, E. A. Catchpole, and B. J. T. Morgan. 2003. The analysis of ring-recovery data using random effects. Biometrika 59:54-65. https://doi.org/10.1111/1541-0420.00007

Bellrose, F. C. 1980. Ducks, geese and swans of North America. Stackpoll Books, Harrisburg, Pennsylvania, USA.

Brook, R. W., J. O. Leafloor, K. F. Abraham, and D. C. Douglas. 2015. Density dependence and phenological mismatch: consequences for growth and survival of sub-arctic nesting Canada Geese. Avian Conservation and Ecology 10(1):1. https://doi.org/10.5751/ACE-00708-100101

Brooks, S. P., E. A. Catchpole, B. J. T. Morgan, and M. P. Harris. 2002. Bayesian methods for analysing ringing data. Journal of Applied Statistics 29:187-206. https://doi.org/10.1080/02664760120108683

Brooks, S. P., R. King, and B. J. T. Morgan. 2004. Bayesian approach to combining animal abundance and demographic data. Animal Biodiversity and Conservation 27:515-529. https://doi.org/10.32800/abc.2004.27.0515

Burnham, K. P., and G. C. White. 2002. Evaluation of some random effects methodology applicable to bird ringing data. Journal of Applied Statistics 29:245-264. https://doi.org/10.1080/02664760120108755

Cooch, E. G., M. Guillemain, G. S. Boomer, J.-D. Lebreton, and J. D. Nichols 2014. The effects of harvest on waterfowl populations. Wildfowl Special Issue 4:220-276.

Cooch, E. G., and G. C. White. 2019. Program MARK: a gentle introduction. http://www.phidot.org/software/mark/docs/book/

Cormack, R. M. 1964. Estimates of survival from the sighting of marked animals. Biometrika 51:429-438. https://doi.org/10.1093/biomet/51.3-4.429

Dennis, D. G., N. R. North, and H. G. Lumsden. 2000. Range expansion and population growth of Giant Canada Geese in Southern Ontario: benefits, drawbacks, and management techniques. Pages 159-165 in K. Dickson, editor. Towards conservation of the diversity of Canada Geese (Branta canadensis). Canadian Wildlife Service Progress Note 103. Environment Canada, Ottawa, Ontario, Canada.

Dziminski, M. A., F. M. Carpenter, and F. Morris. 2021. Monitoring the abundance of wild and reintroduced bilby populations. Journal of Wildlife Management 85:240-253. https://doi.org/10.1002/jwmg.21981

eBird. 2021. eBird: An online database of bird distribution and abundance. eBird, Cornell Lab of Ornithology, Ithaca, New York, USA. http://www.ebird.org

Fowler, D. N., E. B. Webb, M. P. Vrtiska, and K. A. Hobson. 2020. Winter carry-over effects on spring body condition driven by agricultural subsidies to Lesser Snow Geese (Anser caerulescens caerulescens). Avian Conservation and Ecology 15(2):21. https://doi.org/10.5751/ACE-01743-150221

Gale, R. S., E. O. Garton, and I. J. Ball. 1987. The history, ecology, and management of the Rocky Mountain population of Trumpeter Swans. U.S. Fish and Wildlife Service, Montana Cooperative Wildlife Research Unit, Missoula, Montana, USA.

Gelman, A., and J. Hill. 2007. Data analysis using regression and multilevel/hierarchical models. Cambridge University Press, New York, New York, USA. https://doi.org/10.1017/CBO9780511790942

Giacinti, J. A., A. V. Signore, M. E. B. Jones, L. Bourque, S. Lair, C. Jardine, B. Stevens, T. Bollinger, D. Goldsmith, British Columbia Wildlife AIV Surveillance Program (BC WASP), M. Pybus, et al. 2023. Avian influenza viruses in wild birds in Canada following incursions of highly pathogenic H5N1 virus from Eurasia in 2021–2022. mBio 15:e03203-23. https://doi.org/10.1128/mbio.03203-23

Gillette, L. N. 2005. Is migration necessary for restoration of Trumpeter Swan in the midwest? Pages 55-57 in M. H. Linck, and R. E. Shea editors. Selected Papers of The Twentieth Trumpeter Swan Society Conference. The Trumpeter Swan Society, Council Bluffs, Iowa, USA.

Groves, D. J. 2012. The 2010 North American Trumpeter Swan Survey. U.S. Fish and Wildlife Service, Division of Migratory Bird Management, Juneau, Alaska, USA.

Groves, D. J. 2017. The 2015 North American swan survey. A cooperative North American survey. U.S. Fish and Wildlife Service, Office of Migratory Bird Management, Juneau, Alaska, USA.

Handrigan, S., M. L. Schummer, S. A. Petrie, and D. R. Norris. 2016. Swans Cygnus buccinator re-introduced in southwest and central Ontario. Wildfowl 66 60-74.

Henson, P., and J. A. Cooper. 1992. Division of labour in breeding Trumpeter Swans Cygnus buccinator. Wildfowl 43:40-48.

Hunter-Ayad, J., R. Ohlemüller, M. R. Recio, and P. J. Seddon. 2020. Reintroduction modelling: a guide to choosing and combining models for species reintroductions. Journal of Applied Ecology 57:1233-1243. https://doi.org/10.1111/1365-2664.13629

Intini, K. 2009. TRUMPS: A GIS database of reintroduced nesting Trumpeter Swans (Cygnus buccinator) in Ontario. Bachelor of Science. McMaster University, Hamilton, Ontario, Canada.

Iverson, S. A., E. T. Reed, R. J. Hughes, and M. R. Forbes. 2014. Age and breeding stage-related variation in the survival and harvest of temperate-breeding Canada Geese in Ontario. Journal of Wildlife Management 78:24-34. https://doi.org/10.1002/jwmg.636

Jolly, G. M. 1965. Explicit estimates from capture-recapture data with both death and immigration-stochastic model. Biometrika 52:225-248. https://doi.org/10.1093/biomet/52.1-2.225

Kaminski, R. M., and E. A. Gluesing. 1987. Density and habitat-related recruitment in Mallards. Journal of Wildlife Management 51:141-148. https://doi.org/10.2307/3801645

Kellner, K. 2019. jagsUI: a wrapper around ‘rjags’ to streamline ‘JAGS’ analysis. Version 1.5.1. https://cran.r-project.org/web/packages/jagsUI/ https://doi.org/10.32614/CRAN.package.jagsUI

Kéry, M., and M. Schaub. 2012. Bayesian population analysis using WinBUGS - a hierarchical perspective. Academic Press, Boston, Massachusetts, USA.

Knudsen, H. L., B. Laubek, and A. Ohtonen. 2002. Growth and survival of Whooper Swan cygnets reared in different habitats in Finland. Waterbirds 25:211-220.

Koons, D. N., T. W. Arnold, and M. Schaub. 2017. Understanding the demographic drivers of realized population growth rates. Ecological Applications 27:2102-2115. https://doi.org/10.1002/eap.1594

Koons, D. N., G. Gunnarsson, J. A. Schmutz, and J. J. Rotella. 2014. Drivers of waterfowl population dynamics: from teal to swans. Wildfowl Special Issue 4:169-191.

Koons, D. N., D. T. Iles, M. Schaub, and H. Caswell. 2016. A life history perspective on the demographic drivers of structured population dynamics in changing environments. Ecology Letters 19:1023-1031. https://doi.org/10.1111/ele.12628

Larson, M. A., F. R. Thompson III, J. J. Millspaugha, W. D. Dijak, and S. R. Shifley. 2004. Linking population viability, habitat suitability, and landscape simulation models for conservation planning. Ecological Modelling 180:103-118. https://doi.org/10.1016/j.ecolmodel.2003.12.054

Lebreton, J. D., K. P. Burnham, J. Clobert, and D. R. Anderson. 1992. Modeling survival and testing biological hypotheses using marked animals: a unified approach with case studies. Ecological Monograph 62:67-118. https://doi.org/10.2307/2937171

Link, W. A., J. R. Sauer, and D. K. Niven. 2006. A hierarchical model for regional analysis of population change using Christmas Bird Count data, with application to the American Black Duck. Condor 108:13-24. https://doi.org/10.1093/condor/108.1.13

Lumsden, H. G. 1984. The pre-settlement breeding distribution of trumpeter, Cygnus buccinator, and Tundra Swans, C. columbianus, in eastern Canada. Canadian Field Naturalist 98:415-424. https://doi.org/10.5962/p.355185

Lumsden, H. G. 2008. Trumpeter Swans in Ontario 1982-2006. Toronto Birds 2:51-60.

Lumsden, H. G. 2016. Trumpeter Swans and Mute Swans compete for space in Ontario. Ontario Birds 34:14-23.

Lumsden, H. G., and M. C. Drever. 2002. Overview of the trumpeter swan reintroduction program in Ontario, 1982-2000. Waterbirds 25:301-312.

Lumsden, H. G., R. Kingdon, B. Kingdon, K. Intini, and J. Kee. 2012. The recent history of trumpeter swans in Ontario and Quebec and their status in 2010-2011. Ontario Birds 30:109-119.

Mitchell, C. D., and M. W. Eichholz. 2020. Trumpeter Swan (Cygnus buccinator), version 1.0. In P. G. Rodewald, editor. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.truswa.01

Moser, T. J. 2006. The 2005 North American Trumpeter Swan survey. Division of Migratory Bird Management, U.S. Fish and Wildlife Service, Denver, Colorado, USA.

National Audubon Society. 2021. Community science: Christmas Bird Count: Where have all the birds gone? https://www.audubon.org/community-science/christmas-bird-count/where-have-all-birds-gone

Nuijten, R. J., S. J. G. Vriend, K. A. Wood, T. Haitjema, E. C. Rees, E. Jongejans, and B. A. Nolet. 2020. Apparent breeding success drives long‐term population dynamics of a migratory swan. Journal of Avian Biology 51(11):e02574. https://doi.org/10.1111/jav.02574

Nummi, P., and L. Saari. 2003. Density-dependent decline of breeding success in an introduced, increasing Mute Swan Cygus olor population. Journal of Avian Biology 34:105-111. https://doi.org/10.1034/j.1600-048X.2003.02801.x

Pacific Flyway Council. 2017. Pacific Flyway management plan for the Rocky Mountain Population of Trumpeter Swans. Pacific Flyway Council, care of U.S. Fish and Wildlife Service, Division of Migratory Bird Management, Vancouver, Washington, USA.

Palmer, R. S. 1976. Handbook of North American birds, Volume 2. Yale University Press, New Haven, Connecticut, USA.

Petrie, S. A., and C. M. Francis. 2003. Rapid increase in the lower Great Lakes population of feral mute swans: a review and a recommendation. Wildlife Society Bulletin 31:407-416.

Plard, F., R. Fay, M. Kéfy, A. Cohas, and M. Schaub. 2019. Integrated population models: powerful methods to embed individual processes in population dynamics models. Ecology 100:e02715. https://doi.org/10.1002/ecy.2715

Plummer, M. 2003. JAGS: a program for analysis of Bayesian graphical models using Gibbs sampling. In K. Hornik, F. Leisch, and A. Zeileis, editors. Proceedings of the 3rd International Workshop on Distributed Statistical Computing (DSC 2003). Vienna, Austria.

R Core Team. 2021. R: a language and environment for statistical computing. Version 3.4.2. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/

Ramey, A. M., N. J. Hill, T. J. DeLiberto, S. E. J. Gibbs, M. C. Hopkins, A. S. Lang, R. L. Poulson, D. J. Prosser, J. M. Seeman, D. E. Stallknecht, and X.-F. Wan. 2022. Highly pathogenic avian influenza is an emerging disease threat to wild birds in North America. Journal of Wildlife Management 86:e22171. https://doi.org/10.1002/jwmg.22171

Roque, D. V., T. Gӧttert, V. A. Macandza, and U. Zeller. 2021. Assessing distribution patterns and the relative abundance of reintroduced large herbivores in the Limpopo National Park, Mozambique. Diversity 13:456. https://doi.org/10.3390/d13100456

Sauer, J. R., and W. A. Link. 2011. Analysis of the North American Breeding Bird Survey using hierarchical models. Auk 128:87-98. https://doi.org/10.1525/auk.2010.09220

Schaub, M. and M. Kéry. 2022. Integrated population models: theory and ecological applications with R and JAGS. Academic, Cambridge, Massachusetts, USA.

Schmidt, J. H., M. S. Lindberg, D. S. Johnson, B. Conant, and J. King. 2009. Evidence of Alaskan Trumpeter Swan population growth using Bayesian hierarchical models. Journal of Wildlife Management 73:720-727. https://doi.org/10.2193/2008-262

Seber, G. A. F. 1965. A note on the multiple recapture census. Biometrika 52:249-260. https://doi.org/10.1093/biomet/52.1-2.249

Sedinger, J. S., and R. T. Alisauskas. 2014. Cross-seasonal effects and the dynamics of waterfowl populations. Wildfowl 4:277-304.

Shea, R. E., H. K. Nelson, L. N. Gillette, J. G. King, and D. K. Weaver. 2002. Restoration of Trumpeter Swans in North America: a century of progress and challenges. Waterbirds 25:296-300.

Slater, G. L. 2006. Trumpeter Swan (Cygnus buccinator): a technical conservation assessment. United States, Forest Service, Rocky Mountain Region, Fort Collins, Colorado, USA.

Soykan, C. U., J. Sauer, J. G. Schuetz, G. S. LeBaron, K. Dale, and G. M. Langham. 2016. Population trends for North American winter birds based on hierarchical models. Ecosphere 7(5):e01351. https://doi.org/10.1002/ecs2.1351

Sullivan, B. L., C. L. Wood, M. J. Iliff, R. E. Bonney, D. Fink, and S. Kelling. 2009. eBird: a citizen-based bird observation network in the biological sciences. Biological Conservation 142:2282-2292. https://doi.org/10.1016/j.biocon.2009.05.006

Trefry, S. A., A. W. Diamond, and L. K. Jesson. 2013. Wing marker woes: a case study and meta-analysis of the impacts of wing and patagial tags. Journal of Ornithology 154:1-11. https://doi.org/10.1007/s10336-012-0862-y

Tuljapurkar, S. 1990. Population dynamics in variable environments. Springer-Verlag, Berlin, Germany. https://doi.org/10.1007/978-3-642-51652-8

Varner, D. M., and M. W. Eichholz. 2012. Annual and seasonal survival of Trumpeter Swans in the Upper Midwest. Journal of Wildlife Management 76:129-135. https://doi.org/10.1002/jwmg.280

VKM, K. Eldegard, M. W. Furnes, M. J. Grainger, B. Moe, B. K. Sandercock, G. A. Sonerud, B. Ytrehus, E. Rueness, A. Sayyari, L. Kirkendal, E. Granquist, K. Kausrud. 2024. Effects of capture, marking, and tracking on the welfare of wild birds. Scientific Opinion of the Norwegian Scientific Committee for Food and Environment. VKM Report 2024:03, ISBN: 978-82-8259-439-4, ISSN: 2535-4019. Norwegian Scientific Committee for Food and Environment (VKM), Oslo, Norway.

Watola, G. V., D. A. Stone, G. C. Smith, G. J. Forrester, A. E. Coleman, A. E., J. T. Coleman, M. J. Goulding, K. A. Robinson, and T. P. Milsom. 2003. Analyses of two Mute Swan populations and the effects of clutch reduction: implications for population management. Journal of Applied Ecology 40(3):565-579. https://doi.org/10.1046/j.1365-2664.2003.00811.x

Williams, T. D., E. G. Cooch, R. L. Jefferies, and F. Cooke. 1993. Environmental degradation, food limitation and reproductive output: juvenile survival in Lesser Snow Geese. Journal of Animal Ecology 62:766-777. https://doi.org/10.2307/5395

Wilmore, S. B. 1979. Swans of the World. Taplinger, New York, New York, USA.

Wood, K. A., J. L. Newth, G. M. Hilton, B. A. Nolet, and E. C. Rees. 2016. Inter-annual variability and long-term trends in breeding success in a declining population of migratory swans. Journal of Avian Biology 47(5):597-609. https://doi.org/10.1111/jav.00819

Wood, K. A., R. J. Nuijten, J. L. Newth, T. Haitjema, D. Vangeluwe, P. Ioannidis, A. L. Harrison, C. Mackenzie, G. M. Hilton, B. A. Nolet, E. C. Rees. 2018. Apparent survival of an Arctic-breeding migratory bird over 44 years of fluctuating population size. Ibis 160(2):413-430. https://doi.org/10.1111/ibi.12521