INTRODUCTION

The rapid extinction rate we see today is probably one of the highest in human history (Ceballos et al. 2015). Notably, around 55% of bird extinctions were estimated as undiscovered, with numerous species teetering on the brink of extinction, their critical status largely unrecognized on a global front (Cooke et al. 2023). Conservation efforts in developing countries are particularly challenged and hampered by a lack of comprehensive monitoring and thorough assessment of species conservation status despite these nations being pivotal to global biodiversity (Danielsen et al. 2005). Additionally, conservation measures for rare species in these regions often hinge on international aid, compelling the adoption of broad, appealing conservation strategies. This approach, however, is not consistently underpinned by scientific rigor. To bolster the effectiveness of rare species conservation, it is crucial to integrate fragmented information and utilize models to assess and refine conservation strategies, thereby facilitating more scientifically informed conservation decisions.

Although a few large flagship species are monitored and conserved, numerous species with a lower public profile can disappear unnoticed (Amat and Green 2010, Burger 2006). White-bellied Heron (Ardea insignis) is one such species that is on the brink of extinction. It is the second largest heron species, and its population is rapidly disappearing because of habitat loss and degradation (BirdLife International 2023). The heron, once distributed across South Asia, is now extinct in Nepal and likely in Bangladesh (BirdLife International 2023). Recent records confirm its existence only in Bhutan, India, Myanmar, and China (BirdLife International 2023). The estimated global population range is 49–249 adults with a geographic range spanning less than 165,000 km² of fast-flowing freshwater ecosystems in the Himalayan foothills (BirdLife International 2023). The known estimated global population size is just 60 individuals worldwide, with fewer than five known breeding pairs. Nearly half of the estimated population occurs in Bhutan, including four of five known breeding pairs (Acharja 2020, Maheswaran 2014). The heron is sensitive to human disturbance and specializes in feeding on fish in fast-flowing rivers. Its large size requires a long nesting season and significant fish resources; its riverine habitats are also human population centers and face numerous threats, including hydropower development, sand dredging, and stone quarries (BirdLife International 2001, Royal Society for Protection of Nature 2011).

Information regarding the breeding population is exceedingly rare. Prior to 1930, documentation only existed for two nests presumed to belong to this species (BirdLife International 2001). For over seven decades, no active nests were documented until the discovery of an active breeding pair in Bhutan in 2003 (Royal Society for Protection of Nature 2011).

In Bhutan, the Royal Society for Protecting Nature (RSPN) initiated a full-scale research and conservation program starting in 2003, when the first nest was discovered. The program included mapping its distribution, assessing population size, reproductive survey, breeding activity, and examining its feeding habits (Pradhan et al. 2007, unpublished manuscript). Additionally, the program involved enhancing public awareness, fostering community support and engagement, and conducting preliminary research and information development within Bhutan (Royal Society for Protection of Nature 2011; Pradhan et al. 2007, unpublished manuscript). A series of conservation actions were also implemented, including protection of critical feeding and nesting habitats as “heron zones,” restoration of degraded habitats, construction of artificial feeding ponds, enforcement of illegal fishing and providing communities with fisheries and alternate sources of income, hoping to reduce human dependency on natural resources that compete with herons (Royal Society for Protection of Nature 2011, unpublished report). However, despite concerted conservation efforts, the wild population continued to decline (BirdLife International 2023).

Recognizing its critical situation and risk of extinction of the wild population, a trial artificial incubation and rearing, with eggs collected from a wild nest was conducted and the bird was successfully raised and released back to the wild in 2011 (Royal Society for Protection of Nature 2011). It was piloted to explore possibilities of restoring the population through captive breeding and reintroduction (Royal Society for Protection of Nature 2011). Following that success, there was a gain in momentum in conservation efforts across the range, which resulted in two international conferences and the formulation of an international conservation strategy in 2015 led by the International Union for Conservation of Nature Species Survival Commission (IUCN SSC) Heron Specialist Group in collaboration with regional experts.

Based on the White-bellied Heron international strategy, a captive breeding program was also implemented to preserve the ex-situ gene pool, rescue and rehabilitate sick or injured herons, and breed to supplement the wild population by releasing them into suitable wild habitats (Royal Society for Protection of Nature 2021a). The establishment of captive populations and reinforcement are planned to support and augment wild populations.

The systematic population monitoring across Bhutan over the last two decades has found 20–30 adult birds, including 2–5 breeding pairs, producing 2–8 fledglings annually (Acharja 2019, Royal Society for Protection of Nature 2023). Despite the addition of a significant number of fledglings to the existing population, the total population size has remained critically low and is declining (Tobgay et al. 2022, BirdLife International 2023). This downward trend in population, in the face of concerted conservation efforts, is alarming and calls into question the effectiveness of the current conservation strategies. A more detailed assessment of the population’s status, including demographic parameters, is therefore imperative. Additionally, considering the significant economic costs associated with captive breeding and reinforcement, it is crucial to understand the effectiveness of these approaches.

The integrated population model and population viability analysis framework (IPM-PVA) simultaneously estimates population size, demographic parameters, and their contribution to population trends, and projection of future population dynamics under a range of scenarios, accounting for the uncertainty associated with all demographic parameters (Brooks et al. 2004, Zipkin and Saunders 2018, Schaub and Kéry 2021). A benefit of IPM-PVA is its ability to estimate demographic parameters in cases where there is limited information available (Millon et al. 2019, Plard et al. 2019), such as in the White-bellied Heron. Thus, we used IPM-PVA with data from the population count, carcass recovery, and breeding surveys to estimate population size, survival rate, and fecundity. Furthermore, PVA was used to simulate what conservation action (reinforcement, increasing reproductive success, increasing survival rate) best resulted in positive population growth.

Our objective is to evaluate demographic parameters and viability using the limited information available, such as population count, breeding records, and mortality, and promote strategic conservation actions that can enhance the likelihood of population recovery. This study directly informs White-bellied Heron conservation planning and provides a framework to assess the effectiveness of rare species conservation efforts using fragmented information where data are deficient and comprehensive monitoring presents challenges.

METHODS

Field survey

The data for this research was collected from across the White-bellied Heron range in Bhutan by RSPN in collaboration with Local Conservation Support Groups (LCSGs) established with members from local communities, foresters, students, birdwatchers, nature guides, and volunteers. Since 2003, the team at RSPN systematically maintained records of the White-bellied Heron population and nest counts, fecundity, and mortality, focusing on all major river systems in Bhutan (Fig. 1). An annual population and nest count program was established to monitor population size, which was conducted over five consecutive days at the start of the breeding season (typically the last week of February to the first week of March; White-bellied Heron annual population survey reports [unpublished] by RSPN). This period is most suitable for the survey because of the species’ increased visibility during the courtship period, when 2 to 4 birds may be found together and are most active during this season (Acharja 2019). For the annual population count, the entire stretch of the rivers where herons have been sighted in the past or even potential habitats were divided into transects of 5–10 kilometers, and on average, 70 surveyors were engaged to count herons. Depending on the site’s accessibility, one to four surveyors were deployed at each transect to search for herons from 6:00 AM to 6:00 PM for five consecutive days. For each sighting, the surveyors record the date, time, activity, and other associated information while also communicating with surveyors at adjacent transects to avoid double counting individuals. After the end of the survey, the population size was determined by analyzing the gathered data during that survey period. Next, based on the occurrence records from the population survey, nest surveys were conducted throughout the breeding season to find all active nests in the country. Once the active nests were located, the team regularly monitored and collected data on the reproductive success (success or failure) and the number of fledglings from the nest; this approach has been used consistently since 2003. However, with additional information on the occurrence of herons in other river basins, the survey area expanded significantly over the years. Consequently, the transects for the survey were determined based on the specific area for that particular year. Thus, different numbers of transects were laid at different years. The mortality in the wild was also monitored throughout the year through the recovery and preservation of any carcasses or remains located. This paper is the result of two decades of data collected by the team at RSPN as part of conservation and population recovery projects for the White-bellied Heron.

Integrated population model

The IPM integrated three separate data sources to jointly estimate population size, survival rate, and reproductive success: (1) carcass recovery, (2) reproductive data from nest surveys (reproductive success and number of fledglings), and (3) total population size from annual population count. Our model is an extension of the method developed for Eurasian Hoopoe (Upupa epops; Kéry and Schaub 2011) and Crested Ibis (Nipponia nippon; Okahisa and Nagata 2022). Three integrated population models were developed using different assumed values for breeding probability. We created a population projection matrix model. We specified “N” as the estimated abundance from the model, “C” as counted number of individuals, and “t” as the year. There is no information on the actual age of sexual maturity of White-bellied Heron. Thus, we treated individuals as sexually mature at two years of age, same as its congeners, the Great Blue Heron (Ardea Herodias; Vennesland and Butler 2020) and Gray Heron (Ardea cinerea; Martínez-Vilalta et al. 2020). Sensitivity analysis on the age of sexual maturity was also modeled, assuming sexual maturity at the age of three. However, the results remained the same. Three age-specific numbers were described as N_j,t for juveniles, N_y,t for yearlings, and N_ad,t for adults. Total number of the population was described as N_tot,t.

Two age-specific annual survival rates were estimated. φ_j,t denoted survival for juvenile and φ_ad,t denoted survival for individuals more than one year old. Survival rates were modeled with the mean coefficient μ_φ, coefficients of age α₁, and time random effect ε_φ as:

(1)

Population change due to mortality was modeled according to binomial distributions as:

(2)

To calculate the fecundity (F_t) of the heron, we used two reproductive criteria: reproductive success rate (r_st), and the number of fledglings in successful nests (f_t) following Okahisa and Nagata (2022). This is because the overall number of nests with data was 55, of which 11 nests had successful reproduction but the number of fledglings was not recorded. The reproductive success rate was modeled with the mean coefficient (μ_rs) and yearly random effect (ε_rs,t). The maximum observed number of fledglings was three. The number of fledglings using moment-matching of the binomial distribution (Kéry and Royle 2016) with the mean coefficient (μ_f) and yearly random effect (ε_f,t) as follows:

(3)

White-bellied Heron is sexually monomorphic, but there is no information on sex ratios. Thus, a sex ratio of 0.5 was assumed. Breeding probability cannot be accurately estimated in our case because there were no ringed individuals. Therefore, three different models were developed using three assumed values for breeding probability (Bp_t; model I: observed value, model II: 52%, and model III: 100%). In the model I, we assumed that all nests were found during breeding survey. Thus, the breeding probability was modeled as the number of found nests divided by number of females (0.5*estimated number of adults). Because nest detection rates were unlikely to be 100%, it is expected that the breeding probability in model I will be an underestimate. Thus, we applied the maximum breeding probability estimated in model I (52%) in model II. In model III, it was assumed that all surviving adults would participate in breeding. The total number of juveniles (N_j,t) was calculated by multiplying these parameters and number of adult females.

(4)

Binomial dead-recovery models were used to estimate survival rates, in which individuals were found based on the discovery rate q_t and mortality rate (1-φ_i,t) from the true population size. Number of juvenile and adults carcasses were described as C_deadj,t and C_deadyad,t. Discovery rate was modeled with the mean coefficient (μ_q) and yearly random effect (ε_q,t). Juvenile and adult birds were modeled separately because they have different survival rates, and their ages can be determined at the time of carcass recovery.

(5)

The number obtained from the population survey was modeled as normal distributions averaged over the true population size and observation error. The reproductive success was assumed to be observed by a Bernoulli distribution according to the reproductive success rate. The number of fledglings at each nest was counted according to a zero-truncated binomial distribution. The sex of each fledgling was assumed to be observed by a Bernoulli distribution. We used the goodness of fit with the chi-square discrepancy measure as in previous studies (Kéry and Schaub 2011, Besbeas and Morgan 2014, Margalida et al. 2020).

For population vitality analysis, we simulated population dynamics for 30 years with 17 scenarios in each model: current demographic parameters (5-year average), improved juvenile survival (10%, 20%, 30%, 40%), improved reproductive success (10%, 20%, 30%, 40%), the release of captive-reared individuals (5, 10, 15, 20 per year), improved juvenile survival (10%, 20%, 30%, 40%) with the release of 5 captive-reared individuals. Because of the already high estimated adult survival rate, we did not include a scenario focusing on improving adult survival. There is no information on survival rates for captive-reared individuals, thus first-year survival rate was assumed to be equal to wild-born juvenile survival rates. The survival rate of the released individual was assumed to be the same as wild-born adults from the second year after release. The extinction probability for the PVA was defined as the probability that the number of adults would be less than two in the final year of the simulation. In addition, the probability of an increase in the population in the final year of the simulation compared to 2022 was calculated.

The runjags package (Denwood 2016) in R 4.2.2 (R CoreTeam 2020) and JAGS 4.3.0 (Plummer 2003, 2017) were used to run the Markov chain Monte Carlo (MCMC) method. We ran four chains with different initial values for 500,000 iterations, with the first 100,000 iterations as burn-in, and thinned each chain by 20 to generate 20,000 posterior samples. The convergence diagnosis was determined because the Gelman-Rubin statistic was less than 1.1 for all parameters (Gelman et al. 2013) and each Markov chain converged to a steady-state by a visual judgement of the time series plot (Kéry 2010). The posterior mean value and 95% Bayesian credible intervals (CRIs) of the obtained samples were used as representative values for parameter estimation in IPM, while the median values were used as representative values for PVA. We specified uninformative priors with a uniform distribution for the sigma values for random effects. Uninformative priors with normal distributions were used for the model coefficients.

RESULTS

Demographic parameters

Goodness-of-fit tests indicated that the IPM models did not show any significant deviations from the expected patterns. This was observed across all three IPMs in the dead recovery model for juveniles (model I, P = 0.490; model II, P = 0.539; model III, P = 0.557) and adults (model I, P = 0.311; model II, P = 0.232; model III, P = 0.209), in reproductive success (model I, P = 0.549; model II, P = 0.555; model III, P = 0.559), number of fledglings (model I, P = 0.871; model II, P = 0.849; model III, P = 0.852), and population survey (model I, P = 0.533; model II, P = 0.532; model III, P = 0.529).

For breeding probability, a fixed value of 52% was used for model II and 100% for model III. The estimated breeding probability in model I varied among years, and it increased from 2010 to 2022 (Fig. 2a; range 0.039 to 0.261).

There were no significant differences in estimated fecundities among the three models (posterior mean and CRI of average over the study period: model I 1.693, 1.222–2.201; model II 1.659, 1.179–2.171; model III 1.654, 1.19–2.185). Fecundities were stable until 2011, but they became unstable in 2012 and declined from 2017 onward (Fig. 2b).

The estimated apparent annual survival rates of juveniles were extremely low in all models (Fig. 2c; posterior mean and CRI of average over the study period: model I 0.123, 0.000–0.289; model II 0.094, 0.000–0.257; model III 0.078, 0.000–0.234). In contrast, the adult survival rate was much higher than that of juveniles (Fig. 2d; posterior mean and CRI of average over the study period: model I 0.964, 0.925–0.996; model II 0.953, 0.890–0.994; model III 0.935, 0.834–0.995); however, the adult survival rates declined in 2019 and 2022.

Estimated numbers of juveniles varied among models, with higher breeding probability resulting in a higher number of juveniles (Fig. 3a; posterior mean and CRI of average over the study period: model I 6.53, 3.90–11.35; model II 10.65, 4.30–18.60; model III 18.75, 6.35–31.30). In all models, it was estimated that the number of juveniles peaked in 2014 and slightly declined in recent years. It was also estimated that most juvenile birds died, and few yearlings survived (Fig. 3b; posterior mean and CRI of average over the study period: model I 0.744, 0.05–3.10; model II 0.968, 0.05–3.90; model III 1.371, 0.05–5.00). Number of adults was consistent across all models (Fig. 3c; posterior mean and CRI of average over the study period: model I 23.39, 18.95–27.10; model II 23.09, 18.65–26.95; model III 22.62, 17.55–26.60). The number of adults has been declining since 2017.

The probability of extinction and the probability of population increase over the next 30 years estimated in the viability analysis differed among conservation measure scenarios and models. All models predicted with high certainty that this population would continue to decline without new conservation measures (Fig. 3d). Without any additional conservation measures, probabilities of extinction reached 10.4–26.6% (Fig. 4a, 4b, 4c, 4d; model I 26.6%; model II 25.2%; model III 10.4%), and probability of population increase were only 6.2–14.0% (Fig. 4e, 4f, 4g, 4h; model I 6.2%; model II 8.2%; model III 14%).

Improving juvenile bird survival significantly reduced the extinction probability, but the effect varied among models (Fig 4a). In model I, there was a 5.6% probability of extinction, even with a 40% improvement in juvenile survival. In contrast, the extinction probability dropped to 2.4% with 30% increased juvenile survival in model II, and 1.2% with 10% increased juvenile survival in model III.

Higher survival rates were required to increase population growth and reduce risk of extinction (Fig 4e). In model I, the probability of population increase was 67.2% with 40% improvement in juvenile survival. In model II the probability of population increase reached 94.6% with 40% improvement in juvenile survival. In model III the probability of population increase reached 98.6% with 20% improvement in juvenile survival.

There were no scenarios in any model where improved breeding performance would result in an extinction probability of less than 5% or a population increase probability of more than 90% (Fig 4b, 4f).

Releasing captive-bred individuals reduced the probability of extinction, but the effect varied between models (Fig 4c, 4g). To reduce the extinction probability below 5%, releasing more than 20 individuals was required for Model I and Model II and more than 5 individuals for Model III. However, no scenarios in either model with captive release resulted in a population increase probability of more than 90%. The probability of population increases when 20 birds were released each year was 53.6% in Model I, 49.0% in Model II, and 59.8% in Model III.

Releasing five birds annually and improving juvenile survival could avoid extinction in all scenarios (Fig 4d, 4h); the probability of extinction was below 5% with a 10% increase in juvenile survival, and the probability of population increase reached 99% with a 20% increase in juvenile survival.

DISCUSSION

Our IPM-PVA framework outlines the current population dynamics and demographic parameters of the White-bellied Heron in Bhutan. We assessed the effectiveness of various conservation measures and identified that the primary threat to the population at present is the extremely low survival rate of juveniles. Our model indicates that improving juvenile survival is essential for accelerating population growth. At the same time, it is crucial to protect the existing breeding pairs and their nesting sites to ensure continued reproduction of the wild population. Additionally, it is urgent to gain a deeper understanding of nest success rates, post-fledging behavior, dispersal patterns, and actual survival rates of juveniles in the wild, as well as to identify and address the key threats they face.

In the current situation, the model indicates that the impact of improved fecundity on the population growth rate and extinction risk is relatively small. This is primarily due to the already high levels of fecundity, which have been unstable since 2018, resulting in several years of low fecundity. Recent observations show that more herons are now breeding in dense, broadleaved forests compared to the open Chirpine forests they preferred before 2017 (Acharja 2020). However, there has also been an increase in nest failures attributed to higher rates of egg predation, sexual conflict, parental infanticide, non-hatching eggs, and nest abandonment (Acharja et al. 2021). The shift in breeding grounds from wide open valleys to narrower gorges, characterized by thick vegetation cover and limited open riverbanks and sandbars, is likely a result of habitat loss, degradation, and disturbances, which may be contributing to the declining fecundity (Tobgay et al. 2022).

Our model indicates that although releasing captive-bred birds can help reduce the risk of extinction, shifting the population toward a growth trajectory is challenging without improving juvenile survival rates. However, improving the juvenile survival rate by 10% and releasing five birds yearly could steadily increase the population. Therefore, the primary goal should be to improve survival rates in the wild over the long term while offering short-term support through headstarting and population reinforcement.

One effective short-term strategy for improving juvenile survival would be to collect wild chicks and raise them at the Conservation Center. This approach would also help maintain a healthy breeding stock and ensure genetic diversity. By raising all chicks in the Center for their first year and then soft-releasing them the following winter, when river volumes are lower, and foraging sites are more accessible, they will be more mature and experienced in foraging. Data indicates that adult survival rates are significantly high, so if chicks can survive their first year after being raised safely, their chances of long-term survival increase considerably. Although such practices are not common, the headstarting program for Spoon-billed Sandpiper (Calidris pygmaea) increased fledgling survival by 20 percent through egg collection, hatching, hand-rearing, and wild release (Loktionov et al. 2023). Headstarting is effective in protecting eggs and chicks from various threats that can limit breeding success, particularly predation (Laidlaw et al. 2021), flooding, and drought (Donaldson et al. 2024). Furthermore, it has effectively led to high hatch-to-fledging rates for various wader species, such as the Black-tailed Godwit (Limosa limosa; Donaldson et al. 2024) and the Piping Plover (Charadrius melodus; Roche et al. 2008).

Collecting eggs or young chicks, preferably before they reach 12 days old, would also allow breeding pairs to re-clutch and potentially produce a second brood within the same year (observation from failed nests in Bhutan since 2018). Observations by the Royal Society for the Protection of Nature (2021) have shown that White-bellied Herons renest up to three times in a single season if their initial breeding attempts are unsuccessful. However, the comparative success rate of subsequent clutches has not been thoroughly studied, making it crucial to evaluate this before considering the continuous collection of eggs or chicks beyond establishing a breeding stock. Moreover, for long-term population recovery, it is essential to understand and address the factors limiting breeding success and juvenile survival in the wild.

Key conservation challenges

The White-bellied Heron has historically maintained a low population despite its wide distribution across South and Southeast Asia (Stanford and Ticehurst 1939, Smythies 1953, Walters 1976, Ali and Ripley 1987, King et al. 2001). The reasons for this persistently low population, despite the availability of abundant suitable habitats, remain unclear. A recent ensemble distribution probability model postulates that habitat loss and prolonged population isolation are significant driving factors (Maheswaran et al. 2021). The study found a drastic habitat loss of over 59% from the past (pre-2000) to the present (post-2000; Maheswaran et al. 2021). Additionally, the heron’s high sensitivity to human disturbance, solitary nature, and specific habitat preferences have pushed remaining populations into a few undisturbed river basins and protected areas along the Himalayan foothills.

Earlier population estimates have been uncertain because of limited sightings. Rose and Scott (1997) estimated the population to be between 20 and 200 individuals, while Kushlan and Hafner (2000) estimated 250 to 1000 mature individuals, taking into account the species’ remote range and the rarity of reported sightings (BirdLife International 2001). However, at the White-bellied Heron International Meeting in 2015, it was realized, based on reports from range countries, that fewer than 60 individuals may be surviving across its entire range. There are no substantial historical records on breeding population size or reproduction. Since the Royal Society for the Protection of Nature began monitoring the population and documenting breeding success, at least 78 chicks have successfully fledged from wild nests in Bhutan between 2003 and 2022. This is a significant contribution to the population, especially given the extremely small breeding population size.

Despite reasonable fecundity, the population size did not increase during these periods; our model estimated almost no net recruitment to the population (Acharja 2020). Unfortunately, it remains unclear why juvenile survival is so low. Past studies on post-fledging behavior, dispersal, and juvenile survival were largely unsuccessful. The last three attempts (in 2011, 2016, and 2019) to study chick dispersal and life history using satellite trackers failed to yield substantive data because of technical issues and mortality of the tracked individuals (Royal Society for Protection of Nature, unpublished data). Initial assessments suggest that low population performance is due to various factors, including mortality from collisions with power lines, natural predation, habitat alteration, limited food availability during the fledging period, and challenges exacerbated by climate change (Tobgay et al. 2022, Birdlife International 2023).

White-bellied Heron chicks typically fledge in June and July, coinciding with the peak monsoon season in Bhutan, which leads to high river volumes and flooding of most accessible feeding grounds (Khandu et al. 2021). Over the last decade, riverine habitats where most herons are found have increasingly fragmented because of urbanization, infrastructure development, mining, quarries, and hydropower projects (Tobgay et al. 2022). When smaller feeding grounds are flooded and fragmented, adult birds can fly longer distances to find other feeding sites (Royal Society for Protection of Nature 2011), whereas fledglings may become stranded and potentially starve because of the unpredictable and prolonged inundation of feeding sites.

Mesocarnivores, such as monkeys (Macaca mulatta) and the Masked Palm Civet (Paguma larvata), have been observed preying on heron eggs (Khandu 2022) and may also target fledglings. However, adult birds, because of their large size, have few natural predators. Because the species is protected under the Forest and Nature Conservation Act and supported by extensive public awareness and education campaigns, there are no recorded incidents of killings, which contributes to a higher survival rate.

Future conservation and research

Bhutan’s complementary approach, which focuses on preserving the wild population and their habitats while reinforcing them through captive breeding and release through the Conservation Center is promising. However, the Center is relatively new (founded in 2021), and only the first batch of five birds are being raised. The Center plans to continue harvesting selective eggs and chicks and rescuing individuals from the wild until the founder population’s diversity represents that of wild relatives. The Center may take several years to start breeding and releasing individuals into the wild (Royal Society for Protection of Nature 2021b). Alongside establishing a foundational population and headstarting by strategically collecting chicks and eggs from wild nests to improve the current juvenile survival rate, the combined approach could serve as an important complementary conservation action for population reinforcement.

Captive breeding and reinforcement have proven successful in recovering many critically endangered avian species, such as the Crested Ibis (Zang et al. 2021, Okahisa and Nagata 2022) and Black Stilt (Himantopus novaezelandiae; van Heezik et al. 2005), from a handful of individuals. However, the long-term success of this conservation strategy will require an integrated approach that extends beyond captive breeding and reinforcement. It is essential to continue ongoing efforts such as awareness campaigns, community engagement, monitoring, and community-based conservation actions. Additionally, protecting critical habitats, such as foraging, roosting, and nesting sites, is crucial for the survival of this small population. Furthermore, it is necessary to actively engage governmental bodies and initiate policy changes to address the impacts of infrastructure projects, such as dam construction and land use changes, on the species and its habitat. Scaling up these measures to the landscape level and across the species’ global range will be vital, particularly if short-term actions can enhance the overall performance of the sub-population.

Given the precarious status of this species, every individual bird is critical for its future. Cooperation among the range countries, along with support from international organizations like BirdLife International and the IUCN Species Survival Commission (IUCN-SSC), will be vital. Additionally, technical support from regional and international experts, as well as knowledge sharing among the partners, will play a key role. With the current global extent of occurrence concentrated in fragile Eastern Himalayas and Indo-Burma biodiversity hotspots in less than 165,000 square kilometers (Birdlife International 2023), coordinated research, population monitoring, and standardized conservation actions could be applied at the metapopulation level.

A preliminary genetic assessment of samples collected from Bhutan has revealed low genetic diversity (Royal Society for the Protection of Nature and Aaranyak Wildlife Genetics Laboratory, unpublished report), suggesting potential inbreeding within Bhutan’s population, which may also be common among other sub-populations. In the long term, countries within the species’ range could explore possibilities for sharing birds for breeding stock and for the release of captive-bred birds across their range, but such actions should be approached cautiously and supported by thorough research.

Integrated population models (IPMs) are highly flexible and offer the advantage of estimating demographic parameters even when data gaps exist (Millon et al. 2019, Riecke et al. 2019, Frost et al. 2023). IPMs have been effectively utilized as comprehensive decision-making tools in the conservation and management of many species, including the Bald Eagle (Haliaeetus leucocephalus; Cappello et al. 2024), the Piping Plover (Saunders et al. 2018), and the Northern Bobwhite (Colinus virginianus; Rosenblatt et al. 2021). However, our model faces several inherent biases due to limitations in sample size, inconsistent spatial coverage, and variability in data collection methods. The lack of individual marking and territory mapping required the use of a dead-recovery model to estimate survival rates, accounting for estimated abundance. The complex topography of heron habitats resulted in uneven survey efforts, likely leading to underestimation of mortality rates. In the absence of empirical data on breeding probabilities, we used assumed values, which introduced variability in juvenile and total population estimates across models. To account for this, we included minimum (Model I) and maximum (Model III) breeding probabilities, with the true value expected to lie between them. Although there are uncertainties, this model offers essential evidence for informing conservation strategies in data-poor environments. For the White-bellied Heron, ongoing monitoring and additional research into breeding biology, dispersal, and survival rates will allow refinement of the model and improve its accuracy over time. This model, therefore, provides a scalable solution that can be applied to various species and environments, facilitating strategic conservation efforts.

Our model assumes that the population is restricted to Bhutan, but more than half of the total population is likely to be found in India and Myanmar. The mountainous Medog region, located at the confluence of Tibet and Yunnan in China, has landscapes similar to those found in Bhutan, Northeast India, and Northern Myanmar, and likely supports a small population. Limited information is available on the breeding population in India, and almost none exists for the breeding population in Myanmar and China. There are also possibilities that the populations along the southern border, such as Phibsoo Wildlife Sanctuary and Royal Manas National, could be living on both Bhutan and Indian sides, and survival rates may be estimated to be low because of the dispersal of juveniles to some undiscovered areas or across the border to India. Therefore, we expect to refine the model for the entire metapopulation in the future as new research and field monitoring provide additional information for other subpopulations. Nonetheless, this model and its future revisions will continue to provide strategic directions for the effective conservation of this critically endangered species.

LITERATURE CITED

Acharja, I. P. 2019. The current population, distribution, and conservation status of the critically endangered White-bellied Heron (Ardea insignis) in Bhutan. Bulletin of the Yale Tropical Resources Institute 38:1-10.

Acharja, I. P. 2020. Evaluation of nest habitat, site preferences and architecture of the critically endangered White-bellied Heron Ardea insignis in Bhutan. Bird Conservation International 30(4):599-617. https://doi.org/10.1017/S095927091900042X

Acharja, I. P., S. Tshering, T. Lhendup, and T. Phuntsho. 2021. First observation of sexual conflict and parental infanticide in the Critically Endangered White-bellied Heron Ardea insignis. BirdingASIA 35:38-42.

Ali, S., and S. D. Ripley. 1987. Compact handbook of the birds of India and Pakistan. Oxford University Press, New Delhi, India.

Amat, J. A., and A. J. Green. 2010. Waterbirds as bioindicators of environmental conditions. Pages 45-52 in C. Hurford, P. Schneider, and I. G. Cowx, editors. Conservation monitoring in freshwater habitats: a practical guide and case studies. Springer, Dordrecht, The Netherlands. https://doi.org/10.1007/978-1-4020-9278-7_5

Besbeas, P., and B. J. T. Morgan. 2014. Goodness‐of‐fit of integrated population models using calibrated simulation. Methods in Ecology and Evolution 5(12):1373-1382. https://doi.org/10.1111/2041-210X.12279

BirdLife International. 2001. Threatened birds of Asia: The BirdLife International Red Data Book. BirdLife International, Cambridge, UK.

BirdLife International. 2023. Species factsheet: Ardea insignis. BirdLife International, Cambridge, UK. http://datazone.birdlife.org/species/factsheet/white-bellied-heron-ardea-insignis

Brooks, S. P., R. King, and B. J. T. Morgan. 2004. A Bayesian approach to combining animal abundance and demographic data. Animal Biodiversity and Conservation 27(1):515-529. https://doi.org/10.32800/abc.2004.27.0515

Burger, J. 2006. Bioindicators: types, development, and use in ecological assessment and research. Environmental Bioindicators 1(1):22-39. https://doi.org/10.1080/15555270590966483

Cappello, C. D., K. V. Jacobson, J. T. Driscoll, K. M. McCarty, and J. M. Bauder. 2024. Evaluating the effects of nest management on a recovering raptor using integrated population modeling. Ecosphere 15(10):e4943. https://doi.org/10.1002/ecs2.4943

Ceballos, G., P. R. Ehrlich, A. D. Barnosky, A. García, R. M. Pringle, and T. M. Palmer. 2015. Accelerated modern human-induced species losses: entering the sixth mass extinction. Science Advances 1(5):e1400253. https://doi.org/10.1126/sciadv.1400253

Cooke, R., F. Sayol, T. Andermann, T. M. Blackburn, M. J. Steinbauer, A. Antonelli, and S. Faurby. 2023. Undiscovered bird extinctions obscure the true magnitude of human-driven extinction waves. Nature Communications 14:8116. https://doi.org/10.1038/s41467-023-43445-2

Danielsen, F., N. D. Burgess, and A. Balmford. 2005. Monitoring matters: examining the potential of locally-based approaches. Biodiversity and Conservation 14(11):2507-2542. https://doi.org/10.1007/s10531-005-8375-0

Denwood, M. J. 2016. runjags: an R package providing interface utilities, model templates, parallel computing methods and additional distributions for MCMC models in JAGS. Journal of Statistical Software 71(9):1-25. https://doi.org/10.18637/jss.v071.i09

Donaldson, L., R. Hughes, J. Smart, N. S. Jarrett, M. D. Burgess, C. Batey, N. Dessi, and G. M. Hilton. 2024. Headstarting boosts population of a threatened wader, the Black-tailed Godwit. Animal Conservation. https://doi.org/10.1111/acv.12984

Frost, F., R. McCrea, R. King, O. Gimenez, and E. F. Zipkin. 2023. Integrated population models: achieving their potential. Journal of Statistical Theory and Practice 17:6. https://doi.org/10.1007/s42519-022-00302-7

Gelman, A., J. B. Carlin, H. S. Stern, D. B. Dunson, A. Vehtari, and D. B. Rubin. 2013. Bayesian data analysis. Third edition. CRC, Boca Raton, Florida, USA. https://doi.org/10.1201/b16018

Kéry, M. 2010. Introduction to WinBUGS for ecologists: a Bayesian approach to regression, ANOVA, mixed models, and related analyses. Academic, Cambridge, Massachusetts, USA. https://doi.org/10.1016/B978-0-12-378605-0.00003-X

Kéry, M., and J. A. Royle. 2016. Applied hierarchical modeling in ecology: analysis of distribution, abundance and species richness in R and BUGS. Elsevier, London, UK. https://doi.org/10.1016/B978-0-12-801378-6.00001-1

Kéry, M., and M. Schaub. 2011. Bayesian population analysis using WinBUGS: a hierarchical perspective. Academic, Cambridge, Massachusetts, USA.

Khandu, P. 2022. Nest predation of the critically endangered White-bellied Heron Ardea insignis by masked palm civet Paguma larvata in Burichhu, Bhutan. Journal of Heron Biology and Conservation 7:1-11.

Khandu, P., G. A. Gale, and S. Bumrungsri. 2021. Ecological and environmental factors affecting the foraging activity of the White-bellied Heron Ardea insignis (Hume, 1878) in Bhutan. Bird Conservation International 31(3):420-437. https://doi.org/10.1017/S0959270920000684

King, B., H. Buck, R. Ferguson, T. Fisher, C. Goblet, H. Nickel, and W. Suter. 2001. Birds recorded during two expeditions to North Myanmar (Burma). Forktail 17:29-40.

Kushlan, J. A. and H. Hafner. 2000. Heron conservation. Academic, London, UK.

Laidlaw, R., J. Smart, H. Ewing, S. Franks, H. Belting, L. Donaldson, G. Hilton, N. Hiscock, A. Hoodless, B. Hughes, et al. 2021. Predator management for breeding waders: a review of current evidence and priority knowledge gaps. Wader Study 128(1):44-55. https://doi.org/10.18194/ws.00220

Loktionov, E. Y., R. A. Digby, N. N. Yakushev, I. A. Shepelev, J. P. Clements, P. S. Tomkovich, N. S. Jarrett, N. A. Clark, R. E. Green, E. G. Lappo, and E. E. Syroechkovskiy. 2023. Evaluating the impact of headstarting on the critically endangered Spoon-Billed Sandpiper Calidris pygmaea. Diversity 15(4):584. https://doi.org/10.3390/d15040584

Margalida, A., J. Jiménez, J. M. Martínez, J. A. Sesé, D. García-Ferré, A. Llamas, M. Razin, M. À. Colomer, and B. Arroyo. 2020. An assessment of population size and demographic drivers of the Bearded Vulture using integrated population models. Ecological Monographs 90(3):e01414. https://doi.org/10.1002/ecm.1414

Maheswaran, G. 2014. Update on the nesting White-bellied Heron Ardea insignis in Namdapha Tiger Reserve, Arunachal Pradesh, India. BirdingASIA 22:8-8.

Maheswaran, G., L. K. Sharma, H. S. Mondal, and T. Mukherjee. 2021. White-bellied Heron a species on the verge of extinction: ensemble model reveals loss of habitats and resultant prolonged isolation driving the species to extinction. Ecological Informatics 64:101383. https://doi.org/10.1016/j.ecoinf.2021.101383

Martínez-Vilalta, A., A. Motis, and G. M. Kirwan. 2020. Gray Heron (Ardea cinerea), version 1.0. in J. del Hoyo, A. Elliott, J. Sargatal, D. A. Christie, and E. de Juana, editors. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.graher1.01

Millon, A., X. Lambin, S. Devillard, and M. Schaub. 2019. Quantifying the contribution of immigration to population dynamics: a review of methods, evidence and perspectives in birds and mammals. Biological Reviews 94(6):2049-2067. https://doi.org/10.1111/brv.12549

Okahisa, Y., and H. Nagata. 2022. Evaluation of ongoing Crested Ibis (Nipponia nippon) reintroduction using an integrated population model and Bayesian population viability analysis. Ibis 164(4):1104-1122. https://doi.org/10.1111/ibi.13071

Plard, F., R. Fay, M. Kéry, A. Cohas, and M. Schaub. 2019. Integrated population models: powerful methods to embed individual processes in population dynamics models. Ecology 100(6):e02715. https://doi.org/10.1002/ecy.2715

Plummer, M. 2003. JAGS: a program for analysis of Bayesian graphical models using Gibbs sampling. Proceedings of the 3rd International Workshop on Distributed Statistical Computing (DSC 2003) 124:1-10. Vienna, Austria.

Plummer, M. 2017. JAGS version 4.3.0 user manual (Issue August). Pages 0-41. https://sourceforge.net/projects/mcmc-jags/files/Manuals/4.x/jags_user_manual.pdf/download

R Core Team. 2020. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.r-project.org

Riecke, T. V., P. J. Williams, T. L. Behnke, D. Gibson, A. G. Leach, B. S. Sedinger, P. A. Street, and J. S. Sedinger. 2019. Integrated population models: model assumptions and inference. Methods in Ecology and Evolution 10(7):1072-1082. https://doi.org/10.1111/2041-210X.13195

Roche, E. A., F. J. Cuthbert, and T. W. Arnold. 2008. Relative fitness of wild and captive-reared Piping Plovers: does egg salvage contribute to recovery of the endangered Great Lakes population? Biological Conservation 141(12):3079-3088. https://doi.org/10.1016/j.biocon.2008.09.014

Rose, P. M., and D. A. Scott, editors. 1997. Waterfowl population estimates (Vol. 44). Wetlands International, Wageningen, The Netherlands.

Rosenblatt, C. J., R. J. Gates, S. N. Matthews, W. E. Peterman, and N. J. Stricker. 2021. An integrated population model to project viability of a Northern Bobwhite population in Ohio. Ecosphere 12(9):e03731. https://doi.org/10.1002/ecs2.3731

Royal Society for Protection of Nature (RSPN). 2011. The critically endangered White-bellied Heron. R. Pradhan and P. Frederick, editors. Royal Society for Protection of Nature, Thimphu, Bhutan.

Royal Society for Protection of Nature (RSPN). 2021a. The White-bellied Heron Conservation Center: the center for White-bellied Heron conservation breeding, research and information. Royal Society for Protection of Nature, Thimphu, Bhutan. http://www.rspnbhutan.org/downloads/2621/?tmstv=1708464592

Royal Society for Protection of Nature (RSPN). 2021b. White-bellied Heron Conservation Center establishment and management plan. Royal Society for Protection of Nature, Thimphu, Bhutan. https://rspnbhutan.org/?downloads=2578&tmstv=1725221091

Royal Society for Protection of Nature (RSPN). 2023. The White-bellied Heron Ardea insignis annual population survey 2023. Royal Society for Protection of Nature, Thimphu, Bhutan. http://www.rspnbhutan.org/downloads/1902/?tmstv=1709130973

Saunders, S. P., F. J. Cuthbert, and E. F. Zipkin. 2018. Evaluating population viability and efficacy of conservation management using integrated population models. Journal of Applied Ecology 55(3):1380-1392. https://doi.org/10.1111/1365-2664.13080

Schaub, M., and M. Kéry. 2021. Integrated population models: theory and ecological applications with R and JAGS. Academic, Cambridge, Massachusetts, USA.

Smythies, B. E. 1953. Birds of Burma. Second edition. Oliver and Boyd, London, UK.

Stanford, J. K., and C. B. Ticehurst. 1939. On the birds of Northern Burma.—Part V. Ibis 81(1):1-45. https://doi.org/10.1111/j.1474-919X.1939.tb03959.x

Tobgay, T., I. P. Acharja, J. Tshering, and N. Wangdi. 2022. White-bellied Heron conservation action plan (2022-2031). Royal Society for Protection of Nature and Department of Forests and Park Services, Thimphu, Bhutan.

van Heezik, Y., P. Lei, R. Maloney, and E. Sancha. 2005. Captive breeding for reintroduction: influence of management practices and biological factors on survival of captive kaki (black stilt). Zoo Biology 24(5):459-474. https://doi.org/10.1002/zoo.20065

Vennesland, R. G., and R. W. Butler. 2020. Great Blue Heron (Ardea herodias), version 1.0. in A. F. Poole, editor. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.grbher3.01

Walters, M. 1976. Some observations on the eggs of the great White-bellied Heron Ardea insignis. Journal of the Bombay Natural History Society 73:213-214.

Zhang, Y., F. Wang, Z. Cui, M. Li, X. Li, X. Ye, and X. Yu. 2021. Can we reestablish a self-sustaining population? A case study on reintroduced Crested Ibis with population viability analysis. Avian Research 12:14. https://doi.org/10.1186/s40657-021-00250-z

Zipkin, E. F., and S. P. Saunders. 2018. Synthesizing multiple data types for biological conservation using integrated population models. Biological Conservation 217:240-250. https://doi.org/10.1016/j.biocon.2017.10.017