Reintroduction programs are crucial for conserving species at risk of extinction, but their success depends on several factors, with genetic concerns playing a central role. A major challenge in reintroductions is the potential loss of genetic diversity, often exacerbated by the small number of individuals typically involved. This can lead to inbreeding and the founder effect. Inbreeding may lead to inbreeding depression and, along with genetic drift, results in the loss of genetic variability, negatively affecting fertility, survival, and overall population viability (Falconer, 1989). The founder effect occurs when a small subset of individuals is separated from the larger population, carrying only a fraction of the genetic diversity, which over time leads to allele loss, reduced heterozygosity, and decreased population fitness (Shorrock, 1980). Genetic and demographic founder effects can have long-term consequences for the colonizing populations, compounding the risks of inbreeding depression and hindering the population’s recovery (Szűcs et al., 2017). These genetic challenges are particularly concerning for species like the marsh deer (Blastocerus dichotomus), given its restricted distribution and the risk of future genetic bottlenecks, that threaten its survival.

The marsh deer is the largest deer species in South America, reaching up to 100 kg for females and 130 kg for males (Piovezan et al., 2010). It inhabits floodplains and wetland areas, historically ranging from the southern Amazon Rainforest to northeastern Argentina (Piovezan et al., 2010). Several populations are currently at risk of extinction due to poaching, livestock disease exposure, floodplain drainage for agricultural expansion, and the construction of hydroelectric dams; therefore, the species is categorized as vulnerable by the IUCN (Szabó et al., 2003, 2007; Piovezan et al., 2010; Duarte et al., 2016). These factors may cause severe population declines and reduced genetic variability due to inbreeding and genetic drift, impairing reproductive performance and limiting adaptive potential (Roldan et al., 2006; Duarte et al., 2012).

In the early 1990s, following the construction of the Três Irmãos Hydroelectric Power Plant on the Tietê River by Companhia Energética de São Paulo (CESP), 45 marsh deer were rescued and placed in captivity. This led to the creation of the Marsh Deer Captive Management Program, currently managed by Tijoá Energia. The program was operated at the Marsh Deer Conservation Center (CCCP) in Promissão, São Paulo, and the Ilha Solteira Zoo in Ilha Solteira, São Paulo. Then within the same decade, the construction of the Porto Primavera Hydroelectric Power Plant (UHSM) on the Paraná River in 1998 brought the last significant population of marsh deer in São Paulo to the brink of extinction. In response, the Porto Primavera Marsh Deer Project was established to study and conserve the affected population through ex situ conservation efforts, including the reintroduction of animals into remnant floodplain areas in São Paulo (Zanetti and Duarte, 2008).

In late 1998, the Marsh Deer Conservation Program was initiated by researchers from the Deer Research and Conservation Center (NUPECCE) at São Paulo State University (UNESP-Jaboticabal) to study the impact of the UHSM on the local marsh deer population. The project aims to conserve the species through ex situ programs to preserve genetic diversity and provide animals for reintroduction efforts, such as those performed at the Jataí Ecological Station (EEJ) in São Paulo (Figueira et al., 2005; Zanetti and Duarte, 2008). Despite initial challenges, a reintroduced population at the EEJ was established from eight individuals (three males and five females) between 1998 and 2001. Then, between July 2004 and June 2007, individuals from this population were captured for transmitter ear-tag placement, biological material collection, and monitoring, estimating the population size between 20 and 25 individuals (Ferreira, 2011). Although the loss of genetic diversity and inbreeding were not yet concerns, simulations indicated that long-term genetic diversity could only be maintained by introducing new individuals every four generations (Ferreira, 2011).

Given the challenges of translocation efforts, including high mortality rates and diseases risks (Warne and Chaber, 2023), embryo translocation (ET) offers a promising alternative for genetic management and disease risk reduction in reintroduced populations (Rola et al., 2021, 2023). However, to fully utilize ET for species conservation, it is crucial to develop effective superovulation protocols for generating embryos from captive animals. Reproductive studies on marsh deer provide important insights into their breeding biology, which can help to develop and adapt these protocols. Female marsh deer reach sexual maturity at around two years of age, with an average estrous cycle of 21.3 ± 1.3 days. They are polystric and uniparous, with an average pregnancy length of 253 ± 4 days and no reproductive seasonality, as evidenced by postpartum estrus (Polegato et al., 2018). Males display individual variation in antler cycles, with minimal synchronization and no apparent photoperiod influence, indicating year-round fertility (Ramos, 2004; Pereira, 2010). These basic reproductive characteristics are crucial for optimizing ET techniques and enhancing genetic management in conservation programs (Comizzoli, 2020).

Despite progress, the program faces challenges in maintaining optimal population sizes and genetic diversity, particularly concerning male management and institution capacity (Ferreira, 2011). The original population of 82 animals distributed in 18 Brazilian institutions had decreased to 67 individuals distributed in 12 institutions by 2020 (considering only the UHSM population). These challenges highlight the need for a genomic embryo bank to enhance genetic management and ensure the long-term success of the conservation program (Comizzoli, 2020; Rola et al., 2021).

This study aimed to bridge captive populations with wild counterparts by collecting embryos from the Brazilian captive population to establish a cryogenic bank for the species and developing an embryo reintroduction technique to assist the wild EEJ population.

The establishment of the Marsh Deer Conservation Program represents a significant milestone in Brazil and South America’s conservation efforts, being one of the first fully local conservation programs. Over the course of two decades, the program has navigated challenges, setbacks, and successes, underscoring the complexities of implementing such a comprehensive initiative with the collaboration of all stakeholders. Despite notable advancements, the marsh deer remains at risk across its range, highlighting the critical importance of integrating reproductive management strategies and biotechnologies to support both at-risk wild and captive populations. While considerable progress has been made in animal reproductive sciences, continued research is essential to furthering our understanding of the reproductive physiology of Neotropical deer species. Based on previous studies of superovulation (SOV) in species such as the brown brocket deer (Subulo gouazoubira) (Duarte and Garcia, 1995; Zanetti and Duarte, 2012; Zanetti et al., 2014) and the red brocket deer (Mazama americana) (Cursino et al., 2014), this study aims to evaluate the application of SOV and embryo transfer techniques as potential tools for advancing the conservation of the marsh deer.

Due to the limited number of captive marsh deer, we could not use a standard experimental design as used with non-endangered species. Instead, we developed an adaptive system with sequential experiments, making small adjustments between trials based on earlier outcomes. In Treatment A, the inadequate superovulatory response may have resulted from several factors, including handling-induced stress, which affects oocyte maturation, suppress the preovulatory LH surge, and may influence adrenal progesterone production (Monfort et al., 1990; Brann and Mahesh, 1991; González et al., 2008; Zanetti et al., 2014). The absence of an ovulation inducer may have contributed to the lack of ovulation, as the proper timing and type of inducer are critical for follicle maturation and ovulation (Xu et al., 1995; D’Occhio et al., 1997; Liu and Sirois, 1998; D’Occhio et al., 1999; Kaim et al., 2003). Insufficient LH receptor development in ovarian follicles can lead to follicular luteinization without ovulation (Brann and Mahesh, 1991). Common inducers include LH, GnRH and human chorionic gonadotrophin (hCG). Zanetti et al. (2014) observed multiple anovulatory follicles following SOV protocols based on eCG and FSH in brown brocket deer (4.5 ± 1.7 and 7.6 ± 0.9, respectively). Elevated fecal glucocorticoid levels during the SOV period further supported this hypothesis. To address these issues in Treatment B, we added an additional day of exogenous progesterone, adjusted the timing of eCG administration, and included LH as an ovulation inducer after estrus detection.

Modifications made in Treatment B did not effectively induce a superovulatory response in marsh deer females, and female F210 showed no response at all. We suggest that F210’s lack of response may be related to stress from the management procedures during treatment, which could have disrupted her reproductive physiology (Monfort et al., 1990). However, glucocorticoid hormone levels were not measured to confirm this hypothesis. For the other females, we hypothesize that a delayed eCG application may have interfered with the timing of follicular development. By the time eCG was administered, a dominant follicle had already begun to suppress the development of subordinate follicles, which are still FSH-dependent at the recruitment stage (Castilho and Garcia, 2005). Although female F261 developed two ovulatory follicles, there was a lack of synchronicity between follicular maturation and the ovulation inducer. In contrast, female F344 exhibited two CL, which might be considered a superovulatory response, but the overall follicular response was suboptimal due to the limited number of follicles ( Table 2 ).

After the unsatisfactory response to Treatment B, further adjustments were made to the SOV protocol in Treatment C, including lowering the EB dose to 0.5 mg, adding GnRH (0.11 mg), administering eCG on day 5, and delaying LH application until 12-18 hours after estrus detection. These modifications successfully induced an adequate superovulatory response in marsh deer females. Previous studies support the effectiveness of a lower EB dose during progesterone implant insertion (Kumar et al., 2003; Zanetti and Duarte, 2012). Additionally, GnRH improved dominant follicle ovulation, and combining EB with CIDR^® promoted follicular atresia in subordinate follicles, aiding estrus synchronization (Bo et al., 1995; Diskin et al., 2002; McCorkell et al., 2008; Zanetti et al., 2014). The delayed eCG application was aimed at optimizing follicular growth, with studies indicating follicular wave emergence occurs 4-5 days after estradiol-17β and progesterone treatment (Burke et al., 2000; McCorkell et al., 2008), and selection occurs 3-4 days after GnRH (Twagiramungu et al., 1995). The delayed LH application allowed for proper follicular maturation, making them more responsive to the LH peak stimulus (D’Occhio et al., 1997, 1999). Despite being 15 years old, female F61 responded to Treatment C, showing 3 CL and 1 anovulatory follicle. Similar to Zebu cattle, older ruminants typically have a reduced ovarian follicle reserve (Peixoto et al., 2006), with a decrease in CL and recovered structures. This suggests that the treatment may be more effective in younger females, as seen with F315 (6 years) and F366 (3 years) ( Table 2 ). The superior response in Treatment C was characterized by higher CL counts and an improved total follicular stimulus (mean CL = 8, TFS = 10) compared to Treatments A and B (mean CL = 0.67 and 1, TFS = 5.33 and 2, respectively). These results suggest that the protocol used in Treatment C significantly enhances follicular development and superovulation outcomes compared to the less effective protocols employed in Treatments A and B.

Embryo recovery was successfully performed in all treatments, with Treatment C yielding the highest recovery rate. The embryo recovery rate in Treatment C was 37.5% (9/24), which is consistent with rates reported in other deer species, such as Elaphurus davidianus (29.3%) (Argo et al., 1994); Cervus elaphus (43.1%) (Argo et al., 1994), (32.7%) (Asher et al., 1995); and Dama dama (30.6%) (Asher et al., 1995). Of the nine embryos recovered in Treatment C, four were degenerated, and two showed developmental arrest. Only two of the remaining embryos were of sufficient quality for transfer and were subsequently used in this study ( Figure 5 ). Both viable embryos came from female F366, who exhibited the best response in terms of follicular development within Treatment C. This suggests that females of optimal reproductive age may respond more favorably to the SOV protocol used in Treatment C, further supporting the importance of age in reproductive success. The results presented here represent an opportunity to adjust the protocol for the implementation of a genetic resource bank for the species. This will be of vital importance for supporting both the captive population and the free-ranging population in the Marsh Deer Conservation Program.

The decision to transfer the only two viable embryos into the same female was a difficult one, made with the intention of maximizing the chances that at least one of them would implant. Although this species is considered uniparous and twin or multiple births have not been observed (Piovezan et al., 2010; Polegato et al., 2018), unfortunately, nine months later, during recapture, it was discovered that the female had lost both embryos. Suggesting that approximately three months after her reintroduction into the wild, it displayed estrus and had the opportunity to mate, eventually showing signs of pregnancy with approximately six months of gestation at the time of the ultrasound evaluation. While this was disappointing in terms of potential genetic diversity enhancement for the EEJ population, it nonetheless provided hope regarding the entire procedure conducted since the female’s capture from the wild. It indicates that the surgical procedures for embryo transfer did not negatively affect its physiological reproductive capacity, which is crucial for the viability of this procedure in other wild females.

This attempt demonstrated the feasibility of holding a free-ranging animal in captivity for four months and successfully reintroducing it to the wild, offering valuable insights into the potential of ex situ conservation techniques. It is important to note, however, that the female was reintroduced to an area very close (100 meters) to her original capture site, minimizing environmental variations and ensuring she was familiar with the surroundings. The successful reintroduction of this female marsh deer supports findings on the positive effects of longer acclimation periods on site fidelity (Mertes et al., 2019). Consistent with studies on elk (Ryckman et al., 2010; Bleisch et al., 2017), the marsh deer remained near the release site without signs of stress or displacement, suggesting that the extended captivity period facilitated proper acclimation. In contrast, shorter captive holding periods (4–11 days), as seen in elk, led to greater dispersal from the release site (22.6–26 km) (Ryckman et al., 2010). This case highlights the importance of controlled acclimation in familiar environments to improve reintroduction success and reduce risks such as post-release dispersal (Parker et al., 2008; Tuberville et al., 2008). Besides, it is important to consider that deer are prey species in the wild and are highly susceptible to stress stimuli. For this reason, their release must be carefully planned to ensure that the animal can adequately adapt and return to its natural habitat. Proper planning of reintroduction strategies, including minimizing handling and environmental stressors, is essential for the success of wildlife restoration efforts (Fischer and Lindenmayer, 2000; Seddon et al., 2007). Moreover, stress during reintroduction can negatively impact animal health and survival rates, further highlighting the need for a thorough, well-managed release process (Teixeira et al., 2007).

The use of GPS-GSM technology to monitor the reintroduced female marsh deer provided invaluable insights into her post-release behavior, survival, and adaptation to the wild. Successful reintroduction programs, as highlighted by previous studies, depend on tracking released animals to assess their integration into the environment (Parker et al., 2008; Mertes et al., 2019). In this study, the GPS collar with GSM data transmission was crucial for understanding the female’s movements and habitat use post-release. It enabled two types of data sampling: continuous hourly location tracking for broader area uses and survival, and intensive 15-minute intervals to monitor fine-scale movements during the immediate post-release period and some days of gestation. This dual approach provided comprehensive data on both daily behavior and long-term adaptation, aligning with findings from other reintroduction studies where detailed movement data is essential for evaluating habitat selection (Ryckman et al., 2010; Bleisch et al., 2017). Initially, the female showed higher mobility, moving an average of 124 meters per hour (3 km/day) during the first three days, typical of post-release behavior as animals re-establish home ranges (Mertes et al., 2019). Afterward, the female’s movement slowed to 50 meters per hour (1.2 km/day), indicating stabilization and adaptation. Therefore, GPS tracking proves a valuable tool for wildlife managers, supporting informed decisions on habitat suitability, release strategies, and the effectiveness of ex situ conservation efforts.

This study demonstrates the potential of embryo transfer as a viable strategy for enhancing genetic diversity in natural marsh deer populations, particularly those threatened by low genetic variability (Rola et al., 2021, 2023). Traditional reintroduction programs often face challenges related to inbreeding depression, which can lead to reduced fitness, lower reproductive success, and heightened disease susceptibility (Seddon et al., 2007). By introducing genetically diverse embryos sourced from ex situ conservation programs, this technique offers a controlled and scientifically grounded approach to mitigating these issues. The findings highlight key advancements, including improvements in hormonal protocols, optimized embryo collection techniques from the captive population, and the refinement of transfer methods. The success achieved in obtaining viable embryos and performing transfers indicates that these methods can significantly contribute to conservation programs to sustain healthy and diverse populations. Continued research and technological refinement are necessary to enhance the effectiveness of this approach and adapt it to the specific reproductive physiology of marsh deer and other threatened Neotropical deer species. By integrating embryo transfer into broader conservation strategies, this method holds promise as a valuable tool for addressing genetic bottlenecks and improving the long-term viability of wildlife populations facing environmental and genetic challenges.