Ornithological Methods

INTRODUCTION

Understanding the demographic characteristics of animal populations, including details such as individual age, sex, and population structure, is fundamental to effective wildlife management and conservation efforts (Beissinger and Westphal 1998, Kilpatrick et al. 2020). These demographic factors inform key aspects of population dynamics, such as reproductive success and survival rates, allowing wildlife managers and researchers to better model population dynamics to the near future and thus, predict potential susceptibility to environmental changes (Kilpatrick et al. 2020). Birds play a vital role in various ecosystems, contributing to seed dispersal, pollination, and pest control, as well as serving as key indicators of environmental health (García-Moreno et al. 2007). For example, shifts in the age structure of a bird population can signal environmental changes, such as habitat degradation or food scarcity that may otherwise go undetected (Martins et al. 2024). Additionally, knowing the sex of individual birds within a population is crucial for assessing reproductive viability and predicting future population growth (Elphick et al. 2007). Understanding and knowing these demographic factors allows conservationists and wildlife managers to develop targeted, species-specific strategies that support sustainable bird populations and preserve biodiversity in increasingly vulnerable ecosystems (Kilpatrick et al. 2020).

Despite its importance, accurately identifying demographic information like age and sex within wild populations presents numerous challenges, especially for bird species inhabiting sensitive or remote wetlands (Griffiths 2000). Many wetland birds, such as herons, bitterns, and certain migratory waterfowl, are particularly sensitive to human presence, making frequent data collection efforts difficult and risking behavioral changes or nest abandonment (Gray et al. 2013). Traditional methods, such as physical tagging, capture, and genetic sampling, can be invasive, time-consuming, and stressful for these birds, potentially disrupting breeding behaviors and exposing them to greater predation risk, particularly in endangered species in which minimizing human impact is crucial (Duarte 2013, Casas et al. 2014). The complexity and cost of these methods often necessitate specialized equipment and trained personnel, further limiting the feasibility of large-scale or long-term studies across vast and diverse wetland regions (Lieury et al. 2017). Furthermore, the remote and often inaccessible nature of wetland habitats adds logistical challenges to continuous bird monitoring (Gallant 2015).

Recent technological advancements offer promising solutions to overcome these barriers by providing non-invasive ways to assess age and sex and gather demographic data from a distance (Zemanova 2020). For instance, the use of drones equipped with high-resolution cameras has shown promise in monitoring bird colonies remotely, allowing researchers to capture images without disturbing the birds or their habitat (Francis et al. 2022, Colyn et al. 2024). Other methods such as photogrammetry, for example, use photography to capture precise measurements of animals, which allows researchers to analyze morphology in greater detail and non-invasively (Napoli et al. 2024). This technique has been applied successfully in identifying age and sex markers in various species, with potential for broader application as technology becomes more accessible and sophisticated (Fearnbach et al. 2019, Napoli et al. 2024). By minimizing disturbance and providing rapid, high-accuracy data, photogrammetry could enhance the efficiency and scalability of demographic assessments (Smith et al. 2020).

Flamingos (Phoenicopteriformes) are examples of wetland species for which demographic information of wild birds, such as age and sex, is essential yet challenging to obtain (Yang et al. 2024). These birds, known for their large, dense colonies and vibrant plumage, inhabit remote saline and alkaline lakes and coastal wetlands, which can be difficult to access regularly (Winkler et al. 2020). Furthermore, traditional data collection methods, such as tagging or genetic sampling, are often impractical with flamingos due to their sensitivity to human disturbance and the inaccessibility of their nesting grounds (Yosef 2000, Price 2008). Flamingo colonies may abandon their nests if they perceive a threat, leading to substantial breeding losses (Delfino and Carlos 2024). Consequently, efforts to gather demographic data on flamingos have been limited and sporadic, often preventing researchers from obtaining the data to support informed conservation decisions. However, past research has shown that flamingos exhibit subtle sexual dimorphism, with males generally being larger than females, as well as age-related differences, particularly visible in feather coloration (Montalti et al. 2012, Chiale et al. 2018, Boucheker et al. 2020). Despite these distinguishing features, collecting such data from free-ranging flamingos has proven highly impractical.

Building on our understanding of morphological variation and dimorphism in flamingos, our primary aim is to test modern photogrammetric methodologies for collecting key morphological data from wild flamingos. By employing differential equations and modeling techniques, we aim to develop a reliable approach for sex identification that can guide conservation and management decisions for these populations. To evaluate this approach, we photographed captive Chilean Flamingos (Phoenicopterus chilensis) previously sexed through molecular analyses and applied a measurement extraction method to assess the accuracy of these measurements in sex determination of the individuals. Additionally, we expanded and applied the same methods in a wild free-ranging population of the same species, investigating potential challenges in method application and accuracy under different context of distances and image conditions. Through this research, we aim to propose a field protocol for collecting digital measurements, providing wildlife experts and flamingo specialists with tools for improved data collection and analysis of wild populations.

METHODS

Captive population studied

To evaluate whether photogrammetry is a reliable and effective method for detecting sexual dimorphism in Chilean Flamingo populations, we conducted a study using a captive, sexed, and well-documented group at Fundación Temaikèn, located in Argentina. Fundación Temaikèn is a private zoological institution that plays a pivotal role in environmental conservation through education, awareness campaigns, and animal rehabilitation (https://temaiken.org.ar/). The institution also actively collaborates with scientific research in diverse fields, such as animal behavior, ecology, and welfare. Its facilities host an extensive variety of animal species, primarily from South America, but also from other parts of the world. Among the species housed are two flamingo species: the Chilean Flamingo and the Caribbean Flamingo (P. ruber). Although both species are maintained at the zoo, only the Chilean Flamingo population had complete and reliable sexing data. This population had been sexed in 2017 using molecular tools, as reported by González (2017). The group consisted of 131 individuals, including 63 males, 53 females, and 18 individuals whose sex remained undetermined (González 2017). In the captive flock, most individuals were originally born in the wild and later permanently housed in the zoo, whereas a minority (the 18 individuals that were not sexed) hatched in the zoo (A. de las Colinas, personal communication). The birds ranged in age from 10 to 20 years, ensuring that all individuals were sexually mature and fully developed, which is crucial for studies on dimorphism. The flock was notably stable in its composition and care conditions, reflecting consistent husbandry practices and creating an ideal context for the study (A. de las Colinas, personal communication).

The Chilean Flamingo enclosure is an open habitat designed to mimic natural conditions. It features a medium-sized artificial lake with a solid island at its center, providing space for the flamingos to move freely without fences or barriers restricting their movement. The lake offers access to water and designated feeding areas, and the flamingos spent most of their time engaging in natural behaviors such as resting, foraging, and socializing in and around the water. The enclosure was shared with other bird species, primarily waterfowl (Anatidae) and various wader birds, creating a dynamic and biodiverse environment that encouraged interspecies interactions. Despite the open design, the flamingo flock’s ability to move beyond the enclosure was controlled through wing clipping (A. de las Colinas, personal communication). This procedure involves trimming the primary flight feathers on one wing, effectively preventing the birds from achieving the lift or balance needed for flight. At the time of this research, the Chilean Flamingo habitat did not include any confined or covered areas. Tourists and visitors to the park could observe the flamingos from designated viewing areas, maintaining a safe distance of at least 50 meters from the lake’s borders.

Photography and morphometric measurements

In December 2022, we conducted a photographic survey of all the animals in the Chilean Flamingo flock using a Canon Rebel T7 camera with EF 75 300 mm f/4-5.6 III lens, set to standard photographic configurations. The photographs primarily focused on capturing detailed images of the beak and head of each animal, with particular emphasis on obtaining profile views (Medina et al. 2020). Additionally, the images consistently included the tarsal portion of the legs, ensuring that any identification rings were visible. This approach facilitated the precise identification of individuals within the flock for future reference. The photographs were taken from fixed observation points strategically located around the flamingo enclosure. The camera was also mounted on a tripod at a fixed height of 1.50 meters at each observation point. This setup ensured consistent camera elevation across all photographs, attempting to maintain uniformity in the perspective and alignment of the images (Dai and Lu 2010).

To ensure accurate documentation, the distance between the photographer and the animals was measured using a telemeter (Rangefinder; Humphries et al. 2023). The telemeter was mounted on top of the camera and used to measure the distance between the photographer and the animal prior to each photograph. The device operates by emitting a laser beam that, upon striking the target and being reflected back, provides the distance to the observer. The laser was preferentially aimed at the center of the animal’s body to ensure standardization of the measurements (Heal et al. 2021). All images were captured using the same camera and by a single observer, thereby minimizing potential intermeasurement bias and ensuring consistency in the data collection process (Schüßler et al. 2024). For calibration and reference purposes, photographs were also taken of a standard 30 cm ruler positioned at three predetermined distances: 60 meters, 30 meters, and 10 meters. These reference images provided a reliable scale, enabling accurate comparisons between the known ruler measurements and the measurements observed in the flamingo photographs during subsequent analyses.

After the photographs were taken, we processed them using ImageJ software (Schneider et al. 2012), extracting metadata from each image and utilizing the known measurements of the ruler to construct a reliable scale. This scale enabled accurate conversion and collection of measurements from the flamingos. For each image, we aimed to detect significant differences between sexes by collecting a set of 10 morphological measurements from each individual. All measurements were taken along straight lines to avoid biased variation introduced by curved lines in measurement software (Lürig et al. 2021). The 10 measurements were selected based on previous literature describing sexual dimorphism in flamingos, prioritizing morphological traits less likely to be influenced by environmental or physiological variables such as weight, age, or the health status of individuals (Childress et al. 2005, Montalti et al. 2012). The measurements included two head dimensions (head length and head height), five beak measurements (beak length base-to-tip, beak length base-to-bend, beak length bend-to-tip, beak depth at base, and beak depth at bend), two neck measurements (neck depth upper and neck depth at base), and one lower limb measurement (lower limb length; Table 1, Fig. 1). The researcher collecting morphological measurements and photographs of the captive birds was blind to the sex of the individuals, and sex was only verified after the analysis, avoiding the inclusion of bias in the measurements.

For each photograph, we determined the pixel dimensions by using the metadata of the images captured by the camera (Torres and Bierlich 2020). The pixel dimension was calculated by dividing the sensor width of the camera (in millimeters) by the image width (in pixels). This provided the physical size represented by a single pixel on the sensor. Next, we incorporated the distances between the photographer and the individual flamingos, the focal length of the camera (in millimeters), and the calculated pixel dimensions to determine the sampling distance (SD; Torres and Bierlich 2020; Fig. 1). The sampling distance represents the real-world size that each pixel corresponds to in the photograph. Finally, to scale the pixel-based measurements to real-world values, we multiplied the sampling distance by the lengths measured in pixels for each morphological trait (Torres and Bierlich 2020). This process allowed us to convert all pixel-based measurements into accurate real-world dimensions, expressed in centimeters. The scale was adjusted and calibrated using the actual scale provided by the ruler photographed on-site.

Data analysis

After processing the images and measuring the variables, we obtained different sets of variables for each individual from images taken at various distances. To streamline subsequent analyses, we calculated the mean value of each variable for each individual. All further analyses were conducted using these mean values, providing a standardized dataset for each Chilean Flamingo. Then, we calculated the mean value of each measurement for male and female Chilean Flamingos separately and performed a t-test to evaluate whether significant differences existed between the sexes (Dechaume-Moncharmont et al. 2011). Additionally, we calculated the mean difference index (MDI) for each measurement, following the methodology outlined by Delestrade (2001) and Helfenstein et al. (2004). The MDI is computed as MDI = 100× (mean female/mean male), providing a standardized measure of sexual dimorphism for each variable. We also assessed the reliability of each morphological measurement by calculating the mean error, calculated by the sum of variances divided by the sample size, which represents the average discrepancy between repeated measurements of the variables (Stevens 2002). This step ensured that the measurements were consistent and reproducible.

To determine whether the morphological measurements were effective in distinguishing between male and female Chilean Flamingos, we conducted a discriminant function analysis (DFA). The DFA is a statistical technique used to classify observations into predefined groups based on a set of predictor variables (Stevens 2002). It works by finding a combination of predictor variables that best separates the groups, creating one or more discriminant functions. These functions are linear combinations of predictors and aim to maximize the ratio of between-group variance to within-group variance (Stevens 2002). Initially, all 10 measured variables were included in the analysis. Subsequently, we refined the analysis by using only variables with an MDI greater than 95% because these were considered to exhibit a strong degree of sexual dimorphism (Lovich and Gibbons 1992). The analysis was performed using the lda function from the MASS package (Venables and Ripley 2002) in R, version 2.10.1 (R Core Team 2023), using 30% of the dataset as training set and 70% of testing group. We compared the two DFA methods by evaluating their discrimination rates and success at detecting correctly the sex of individuals at the testing set.

Finally, to assess whether variations in the distance between the photographer and the animal influenced the accuracy of the measurements, we first calculated the difference between each measurement and the mean value of the corresponding variable, ensuring that comparisons were always made within the same individual (Pohl and Farr 2010). Using these differences, we performed a correlation test to determine whether increasing the distance between the photographer and the animal affected the accuracy of the measurements (Stevens 2002). These analyses allowed us to assess the extent to which measurement accuracy and classification performance were affected by variations in the distance from which the photographs were taken. By identifying potential biases or limitations associated with distance, we aimed to ensure the robustness and reliability of our morphological measurements and the resulting discriminant function (Stevens 2002).

Replication with wild population

To assess whether the methods could potentially be applied to a wild population, we captured photographs of a flock of Chilean Flamingos at the Lagoa do Peixe National Park, between June and September 2022. This park is the only area in Brazil where Chilean Flamingos are regularly found, although no sexing or demographic studies have been conducted there (Delfino et al. 2023). The photos were taken using the same camera and setup as those used for the captive population described earlier. The distance between the photographer and the animals was also measured using the same telemeter as in the previous study. Because the animals in the park were not ringed or individually identifiable, we could not compare the measured morphology with known demographic data. Nevertheless, we assessed the reliability of the methods by comparing the sex ratio within the flock across multiple photographs taken on the same day (Mokkink et al. 2023). Measurements were taken from the same image for all animals within the flock to avoid potential bias caused by changes in the ecological context. Photographs of field birds were taken from 8:00 to 18:00 on each sampling day, during which individuals could move within the area or temporarily leave and rejoin the flock; therefore, sex ratios rather than raw counts were used to better account for these fluctuations.

We then applied both DFA methods described earlier and, for each photograph of the flock, counted the number of adult individuals classified as males and females. The sex ratio for each photo was calculated by dividing the number of females by the number of males (Ancona et al. 2017). The method was considered successful if the sex ratio for a single day’s flock did not present significant differences when tested with a chi-squared test, in which the dependent variable was the number of individuals classified as male or female, and the independent variable was the photograph. Although this approach does not provide definitive sex identification for unmarked wild individuals, it allows a practical evaluation of method reliability under natural field conditions. Non-significant variation in sex ratios indicates that the method can be applied in the field and provides an indicative measure of sex identification in limited context.

RESULTS

We captured 4594 photographs of the captive Chilean Flamingo population at the Temaikén Foundation in Argentina, from distances ranging from 9.2 m to 68.4 m. From these images, we successfully collected at least 15 complete sets of variables for each individual at various distances, totaling 19,650 measurements. Across all measured conditions, males consistently exhibited higher values than females, with the differences always being statistically significant, despite varying among the different variables (Table 2). Among the measured variables, upper and lower neck depth showed the greatest variation in both males and females (Fig. 2). In contrast, head length and head width exhibited the lowest variation in females, whereas head length and beak depth at base showed the lowest variation in males (Fig. 2). Overall, the males presented a higher intra-measurements variation then females, excepting the neck variables, which presented higher variances in the latter group. To compare the sexes, the most effective variables were beak length (base-to-tip), beak length (base-to-bend) and beak length (bend-to-tip), because they demonstrated the highest MDI values.

Regarding the DFA, both variable sets demonstrated excellent performance in distinguishing the sexes within the flamingo flock (Fig. 3). This was particularly evident in the DFA using the subset of variables with an MDI greater than 95%, which primarily comprised beak- and head-related measurements. The DFA incorporating all variables achieved discrimination values of 0.787 for LD1 and 0.012 for LD2, resulting in an accuracy of 67.02% in correctly identifying the sexes of individuals in the test sample. In comparison, the DFA based on the most distinct variables yielded discrimination values of 0.985 for LD1 and 0.014 for LD2, accurately classifying 86.25% of the sexes in the test population, a significant difference between the two sets. These results suggest that the reduced set of highly discrepant variables is sufficient to reliably determine the sexes of individuals within a high relative accuracy (Fig. 3).

Regarding measurement accuracy, the mean errors across variables showed minimal differences, remaining within an acceptable interval of less than 15 points of mean error. However, tarsal length and lower neck depth exhibited the highest mean errors among the variables measured (Table 2). This was likely due to the individuals being partially submerged in water, which complicated measurements at greater distances. When analyzing the influence of the distance between the photograph and the animal on the variance of the measurements, no correlation was found for any of the measured variables (Fig. 4). This indicates that distance had a minimal impact on the variation and standard deviation of measurements for the same individuals. However, when examining the graph relating distance to the standard deviation of measurements, a slight increase in variation with distance was observed for beak depth at bend, upper neck depth, and tarsal length (Fig. 4). Although this increase had a negligible impact on the overall accuracy of these measurements, it could affect measurements taken at greater distances.

From June to September 2022, we captured 132 photographs of flamingo flocks at Lagoa do Peixe National Park, with a minimum of 30 photos taken per month (all captured on the same day each month). The distance between the photographer and the animals ranged from 30 to 200 m. For each sampled month, photographs taken at varying distances were included in the analysis to minimize potential distance-related bias in the samples. The same variables were extracted for each individual in these photographs. The DFA using variables with an MDI greater than 95% showed relatively higher accuracy, thus, we chose to apply only this model to the photographs, thereby streamlining the analysis process (Fig. 5).

When we performed DFA on the variable sets for each flock in the photographs and repeated the process for all photos within a given month, we observed mean sex ratios in the flocks ranging from 1.27 to 2.46 females per male (Fig. 5). The lowest ratio was recorded in September, while the highest occurred in July. Comparing sex ratios within the same month, we found no significant differences across photos, demonstrating the reliability of the DFA method in determining flock sex ratios (June χ² = 1.18; July χ² = 0.45; August χ² = 0.68; September χ² = 0.15; all p-values were higher than 0.05). The month with the highest variance, though still statistically non-significant, was also the month with the largest flock size (Fig. 5). This suggests that higher variance might be attributed to observational biases during photo capture, despite efforts to minimize such errors. Nonetheless, the analysis proved to be a reliable approach for estimating sex ratios in wild, non-captive flocks.

DISCUSSION

Chilean Flamingos, like other flamingo species, exhibit discrete sexual dimorphism, i.e., differences between males and females that are subtle and not immediately apparent (Owens and Hartley 1998, Montalti et al. 2012, Boucheker et al. 2020). Unlike species with more obvious dimorphism, identifying sex differences in flamingos requires careful observation and precise measurement (Studer-Thiersch 2007, Montalti et al. 2012). The main distinction lies in body size and proportions, with males generally being larger. In these social birds, which live in large flocks, larger males often achieve higher dominance, greater access to resources, and increased reproductive success (Bildstein et al. 1991, Schmitz and Baldassarre 1992, Royer and Anderson 2014, Rose and Soole 2020). These factors can lead to frequent agonistic interactions among adult males (Schmitz and Baldassarre 1992), while for females, such physical traits are typically less important for competing over space, mates, or resources (Schmitz and Baldassarre 1992, Perdue et al. 2011). Our findings confirmed that males tend to have higher values in various head and beak measurements. Although not all variables showed statistically significant differences individually, the combined set proved effective for distinguishing males from females within the same flock.

Previous studies on Chilean Flamingos and related species have shown that morphological measurements can effectively distinguish between sexes (Richter and Bourne 1990, Richter et al. 1991, Childress et al. 2005, Montalti et al. 2012, Boucheker et al. 2020). However, these methods face technical and analytical limitations, including the need to capture and handle birds, account for age-related variation, and select appropriate variables for reliable sexing (Montalti et al. 2012, Watson et al. 2016, Boucheker et al. 2020). These challenges highlight the need for less invasive and more efficient alternatives. In this study, we introduced photogrammetry as a novel, non-invasive method for sex discrimination in adult Chilean Flamingos within a flock. Our results showed that, when proper protocols are followed, it is possible to distinguish males from females using photographs with high accuracy. This method minimizes physical handling, lowers costs, and is particularly suited for large populations in field settings (Kemper et al. 2016, Humphries et al. 2023). Its success depends on following several key steps during the photographic and measurement process.

The first key consideration in applying photogrammetry is the proper calibration of both the camera and the telemeter used to measure the distance between the camera and the bird. Accurate calibration is essential because image quality, embedded metadata, and measured distance directly affect the precision of morphological measurements (Fraser 2001, Luhman et al. 2016). Calibration refers both to standardizing the camera settings (focal length and resolution) across all photographs and to verifying the telemeter’s accuracy by cross-checking its distance readings against known reference distances prior to data collection (Luhman et al. 2016). These steps are critical to ensure photogrammetry functions as a reliable, non-invasive tool for studying sexual dimorphism in flamingos and other avian species. A second crucial aspect involves selecting appropriate variables and digital analysis software (Gaudioso et al. 2014, Leorna et al. 2022). In our study, we measured 10 variables from different body regions, focusing on the beak and head, areas previously shown to exhibit the greatest sex-related differences (Richter et al. 1991, Childress et al. 2005, Montalti et al. 2012). These traits also showed low measurement error across distances and were less affected by age, making them reliable indicators in adult individuals (Rowley et al. 2020), unlike in juveniles, whose features are not fully developed.

Variable selection was guided by previous research and by the stability of traits under environmental and physiological variation. For example, although tarsal measurements are common in morphometric studies (Studer-Tiersch 2007), they exhibited high variability in our dataset, likely due to water depth and uneven terrain distorting tarsus proportions despite careful imaging (Gilmore et al. 2013). Neck measurements, on the other hand, showed minimal sexual dimorphism and were likely influenced by factors such as health, body weight, and feather condition (Labocha and Hayes 2012). Relying solely on these traits reduced the method’s accuracy, as reflected in our results. Another important consideration is that, although absolute morphological measures may vary slightly between captive and wild individuals due to differences in diet, activity, or environment, previous studies indicated that the relative sexual size dimorphism in flamingos remains consistent across both settings, thereby supporting the use of captive birds as a reliable reference for comparative analyses with wild populations (Montalti et al. 2012). This highlights the need for a multivariate approach. By applying discriminant analysis, we showed that a carefully selected subset of variables can effectively distinguish sexes while reducing reliance on less informative traits (Dechaume-Moncharmont et al. 2011).

The third critical consideration is ensuring accurate measurement of variables in photographs using the selected software. We chose to perform all measurements using straight lines because this approach is more easily processed by software algorithms and minimizes potential biases introduced by curved or non-linear measurements, which often require more complex analytical methods (Jiang et al. 2008, Sapirstein 2016). Although straight-line measurements may not fully capture certain morphological features, they provide a practical and efficient solution for consistent and reliable results. The accuracy of this approach depends on precise identification and consistent use of anatomical landmarks across all images and individuals, facilitating streamlined data collection and enabling meaningful comparisons (Webster and Sheets 2010, Wärmländer et al. 2019). This uniformity is particularly advantageous in large-scale studies in which measurement consistency is critical. Although simplifying some aspects of morphology, the method produces high-quality, reproducible data essential for accurate sex differentiation in species with subtle sexual dimorphism, such as Chilean Flamingos.

Another important factor is the distance between the photographer and the photographed bird. Our results showed no significant effect of distance on measurement accuracy, nor a clear non-linear relationship. Nevertheless, maintaining an appropriate distance remains crucial (Dai et al. 2014). Distances greater than 200 m can complicate pixel-to-real-world conversions and reduce precision (Aksu et al. 2010), while distances below 10 m, though optimal for accuracy, are often impractical in the field due to logistics or the need to avoid disturbing the animals (Piratelli et al. 2015, Tablado et al. 2021). When zoom lenses are used, careful attention must be paid to recording the zoom level in the photograph’s metadata and properly accounting for it in measurements because uncorrected variations can introduce bias (Dai et al. 2014).

By following the established protocol, it is likely to obtain a reliable set of measurements, enabling effective use of DFA to differentiate sexes in identifiable individuals or estimate sex ratios in unmarked flocks (Dechaume-Moncharmont et al. 2011). Applying this method to the Lagoa do Peixe wild flock of Chilean Flamingos demonstrated its effectiveness in ecological contexts. The results revealed a more balanced sex ratio than previously reported, aligning with behavioral studies on the same population (Delfino and Carlos 2021, 2022, Delfino et al. 2023). Earlier studies from the 1980s had suggested a male- and juvenile-dominated population (Belton 1985, Rezende and Leeuwenberg 1987), but recent behavioral observations, such as agonistic interactions and synchronized displays, challenge this assumption (Delfino and Carlos 2021, 2022). Our photogrammetric analysis confirmed these findings, providing quantitative support for a balanced sex ratio and highlighting the value of combining behavioral and morphometric data in conservation research.

Although our results suggest good consistency in sex ratio estimates across photos, the lack of individual identification remains a key limitation. Without marked animals, we cannot confirm whether the same individuals were measured correctly, limiting the accuracy of sex classification. Future applications should consider combining this method with individual recognition to improve reliability. Although our method offers a non-invasive alternative for estimating sex ratios in wild flamingo populations, it is important to recognize the limitations associated with classification accuracy. In the captive validation, approximately 86% of individuals were correctly classified, indicating moderate reliability. This level of accuracy may be sufficient for broad-scale ecological inferences, but it does introduce a margin of error that must be considered when interpreting sex ratio estimates and their fluctuations. These fluctuations likely reflect both the natural challenges of working with unmarked, mobile individuals and the cumulative effect of classification uncertainty. Therefore, we caution against interpreting these estimates as exact counts. Rather, they should be viewed as approximations that can support broader population-level analyses when used with appropriate statistical and conceptual care. Future applications of this method should carefully consider the research context and integrate measures of uncertainty into any downstream analyses or management decisions.

Furthermore, other potential limitations should also be acknowledged. First, measurement accuracy depends heavily on proper calibration of equipment, consistent identification of anatomical landmarks, and careful control of photographic conditions, which may be challenging in field settings with large or mobile flocks (Fraser 2001, Webster and Sheets 2010, Luhman et al. 2016). Second, the method is most reliable for adult individuals because juveniles have not yet fully developed sexually dimorphic traits, limiting its applicability to mixed-age populations (Rowley et al. 2020). Third, environmental factors, such as water depth, uneven terrain, lighting, and feather condition, can introduce variation in measurements, particularly for traits like tarsus or neck length (Labocha and Hayes 2012, Gilmore et al. 2013). Finally, photogrammetry requires precise metadata and careful handling of zoom or distance adjustments, and small errors in these parameters can propagate into the morphological estimates (Aksu et al. 2010, Dai et al. 2014). Despite these challenges, when properly applied, photogrammetry remains a robust tool, though its limitations should be considered when interpreting results and designing studies.

Beyond their application to Chilean Flamingos, these methods hold significant potential for broader use across other flamingo species. Because all flamingo species exhibit distinct sexual dimorphism, with males generally being taller and larger than females, the fundamental principles of this approach may prove widely applicable (Ritcher and Bourne 1990, Ritcher et al. 1991, Childress et al. 2007, Boucheker et al. 2020). The shared morphological traits among flamingo species suggest that similar techniques could provide accurate and reliable results in diverse ecological and geographical contexts. However, to ensure the robustness of these methods across species, further research is essential. Such studies should aim to assess whether additional variables are needed to account for subtle interspecies differences in morphology or whether alternative variables might improve accuracy. Furthermore, it is crucial to investigate potential species-specific adaptations in methodology to accommodate unique behavioral or ecological characteristics. For example, variations in flock structure, habitat use, or growth rates could influence the efficacy of the measurements and require tailored adjustments.

Our study demonstrates that photogrammetry is a practical and minimally invasive method for determining the sex of flamingos in wild populations. This approach reduces stress from handling, provides cost-effective data collection, and enables accurate assessment of key demographic parameters, such as sex ratios, which are critical for management and conservation strategies (Colomina and Molina 2014, Mendes et al. 2016, Volis and Deng 2020). For example, knowing whether groups have balanced sex ratios allows assessment of their reproductive potential and long-term viability in the wild (Donald 2007, Wedekind 2012). This information is also essential for planning reinforcement or translocation programs in which ensuring the inclusion of both sexes is critical to the success of new or supplemented populations (Lambertucci et al. 2013). Furthermore, monitoring sex composition over time can serve as an indicator of demographic health in both wild and managed populations, helping to identify potential biases caused by environmental pressures or management practices (Wedekind 2012). By linking sex determination directly to these applications, our approach provides a practical tool for informing conservation decisions and sustaining viable flamingo populations. Another promising future direction for this approach is the use of photographs from citizen science platforms such as eBird or iNaturalist, allowing researchers to estimate sex ratios and group composition across broader spatial and temporal scales, complementing traditional field-based monitoring.

Our results confirm that careful selection of morphological traits and adherence to standardized photographic procedures allow reliable sex differentiation, particularly in adult birds. Although limitations exist, such as sensitivity to environmental conditions, photographic calibration, and the less distinct traits of juveniles (Fraser 2001, Webster and Sheets 2010, Dai et al. 2014, Rowley et al. 2020), photogrammetry remains a robust tool that can enhance our understanding of population dynamics and support the sustainable management of flamingo habitats (Aksu et al. 2010, Gilmore et al. 2013, Conde et al. 2019). Overall, our findings highlight the potential of combining advanced technological methods with conservation practice to improve monitoring and protection of these iconic birds and their ecosystems.

LITERATURE CITED

Aksu, M., D. Kaya, and I. Kocadereli. 2010. Reliability of reference distances used in photogrammetry. Angle Orthodontist 80:670–677. https://doi.org/10.2319/070309-372.1

Ancona, S., F. V. Dénes, O. Krüger, T. Székely, and S. R. Beissinger. 2017. Estimating adult sex ratios in nature. Philosophical Transactions of the Royal Society B: Biological Sciences 372:20160313. https://doi.org/10.1098/rstb.2016.0313

Beissinger, S. R., and M. I. Westphal. 1998. On the Use of demographic models of population viability in endangered species management. Journal of Wildlife Management 62:821–841. https://doi.org/10.2307/3802534

Belton, W. 1984. Birds of Rio Grande do Sul, Brazil. Part 1, Rheidae through Furnariidae. Bulletin of the American Museum of Natural History 178:371–631. https://digitallibrary.amnh.org/items/5046400f-066b-4dcf-874c-7046b8456ea2

Bildstein, K. L., P. C. Frederick, and M. G. Spalding. 1991. Feeding patterns and aggressive behavior in juvenile and adult American Flamingos. Condor 93:916–925. https://doi.org/10.2307/3247726

Boucheker, A., R. Nedjah, R. Prodon, M. Gillingham, F.-X. Dechaume-Moncharmont, A. Béchet, and B. Samraoui. 2020. Cohort effect on discriminant rate: the case of Greater Flamingo (Phœnicopterus roseus) chicks sexed with morphological characters. Web Ecology 20:153–159. https://doi.org/10.5194/we-20-153-2020

Casas, F., A. Benítez-López, J. García, C. Martín, J. Vióuela, and F. Mougeot. 2014. Assessing the short-term effects of capture, handling and tagging of Sandgrouse. Ibis 157:115–124. https://doi.org/10.1111/ibi.12222

Chiale, M. C., D. Montalti, and R. Maragliano. 2018. Age determination of captive Chilean Flamingo (Phoenicopterus chilensis) chicks based on plumage characteristics. Ornitología Neotropical 29:107–110. https://doi.org/10.58843/ornneo.v29i1.315

Childress, B., D. Harper, B. Hughes, and C. Ferris. 2005. Sex determination in the Lesser Flamingo (Phoenicopterus minor) using morphological measurements. Ostrich 76:148–153. https://doi.org/10.2989/00306520509485487

Colomina, I., and P. Molina. 2014. Unmanned aerial systems for photogrammetry and remote sensing: a review. ISPRS Journal of Photogrammetry and Remote Sensing 92:79–97. https://doi.org/10.1016/j.isprsjprs.2014.02.013

Colyn, R. B., T. A. Anderson, M. D. Anderson, E. T. Retief, E. J. Van der Westhuizen-Coetzer, and H. Smit-Robinson. 2024. The use of image classification to estimate flamingo abundance from aerial, drone and satellite imagery. Ostrich 95:188–199. https://doi.org/10.2989/00306525.2024.2325674

Conde, D. A., J. Staerk, F. Colchero, R. da Silva, J. Schöley, H. M. Baden, L. Jouvet, J. E. Fa, H. Syed, E. Jongejans, S. Meiri, J.-M. Gaillard, S. Chamberlain, J. Wilcken, O. R. Jones, J. P. Dahlgren, U. K. Steiner, L. M. Bland, I. Gomez-Mestre, J.-D. Lebreton, J. González Vargas, N. Flesness, V. Canudas-Romo, R. Salguero-Gómez, O. Byers, T. B. Berg, A. Scheuerlein, S. Devillard, D. S. Schigel, O. A. Ryder, H. P. Possingham, A. Baudisch, and J. W. Vaupe. 2019. Data gaps and opportunities for comparative and conservation biology. Proceedings of the National Academy of Sciences 116:9658–9664. https://doi.org/10.1073/pnas.1816367116

Dai, F., Y. Feng, and R. Hough. 2014. Photogrammetric error sources and impacts on modeling and surveying in construction engineering applications. Visualization in Engineering 2(2):(2014). https://doi.org/10.1186/2213-7459-2-2

Dai, F., and M. Lu. 2010. Assessing the accuracy of applying photogrammetry to take geometric measurements on building products. Journal of Construction Engineering and Management 136:242–250. https://doi.org/10.1061/(ASCE)CO.1943-7862.0000114

Dechaume-Moncharmont, F.-X., K. Monceau and, F. Cezilly. 2011. Sexing birds using discriminant function analysis: a critical appraisal. Auk 128:78–86. https://doi.org/10.1525/auk.2011.10129

Delestrade, A. 2001. Sexual size dimorphism and positive assortative mating in Alpine Choughs (Pyrrhocorax graculus). Auk 118:553–556. https://doi.org/10.1093/auk/118.2.553

Delfino, H. C., O. Aldana-Ardila, C. E. Fedrizzi, and C. J. Carlos. 2023. Multisource data reveals relevant trends in a Chilean Flamingo Phoenicopterus chilensis population at an important coastal wetland of Southern Brazil: implications for conservation and planning. Journal of Coastal Conservation 27:33. https://doi.org/10.1007/s11852-023-00963-x

Delfino, H. C., and C. J. Carlos. 2021. Behavioral repertoire of a population of wild Chilean Flamingos Phoenicopterus chilensis in southern Brazil. Journal of Natural History 55:1957–1981. https://doi.org/10.1080/00222933.2021.1978574

Delfino, H. C., and C. J. Carlos. 2022. Intra-annual variation in activity budgets of a wild Chilean Flamingo (Phoenicopterus chilensis) population in Southern Brazil. Austral Ecology 47:971–982. https://doi.org/10.1111/aec.13180

Delfino, H. C., and C. J. Carlos. 2024. Still standing on one leg: a systematic review of threats, priorities, and conservation perspectives for flamingos (Phoenicopteridae). Biodiversity and Conservation 33:1227–1268. https://doi.org/10.1007/s10531-024-02816-x

Donald, P. F. 2007. Adult sex ratios in wild bird populations. Ibis 149:671–692. https://doi.org/10.1111/j.1474-919X.2007.00724.x

Duarte, L. M. G. 2013. Impacts of capture and handling on wild birds. Dissertation. Cardiff University, Cardiff, UK. https://crbpo.mnhn.fr/IMG/pdf/these_duarte_pochons_cartons.pdf

Elphick, C. S., J. M. Reed, and D. J. Delehanty. 2007. Applications of reproductive biology to bird conservation and population management. Pages 367–399 in B. G. M. Jamieson, editor. Reproductive biology and phylogeny of Aves (Birds). Science Publishers, Enfield, New Hampshire, USA.

Fearnbach, H., J. W. Durban, L. G. Barrett‐Lennard, D. K. Ellifrit, and K. C. Balcomb, III. 2019. Evaluating the power of photogrammetry for monitoring killer whale body condition. Marine Mammal Science 36:359–364. https://doi.org/10.1111/mms.12642

Francis, R. J., R. T. Kingsford, and K. J. Brandis. 2022. Using drones and citizen science counts to track colonial waterbird breeding, an indicator for ecosystem health on the Chobe River, Botswana. Global Ecology and Conservation 38:e02231. https://doi.org/10.1016/j.gecco.2022.e02231

Fraser, C. S. 2001. Photogrammetric camera component calibration: a review of analytical techniques. Pages 95–121 in A. Gruen and T. S. Huang, editors. Calibration and orientation of cameras in computer vision. Springer, Berlin, Germany. https://doi.org/10.1007/978-3-662-04567-1_4

Gallant, A. L. 2015. The challenges of remote monitoring of wetlands. Remote Sensing 7:10938–10950. https://doi.org/10.3390/rs70810938

García-Moreno, J., R. P. Clay, and C. A. Ríos-Muñoz. 2007. The importance of birds for conservation in the Neotropical region. Journal of Ornithology 148:321–326. https://doi.org/10.1007/s10336-007-0194-5

Gaudioso, V., E. Sanz-Ablanedo, J. M. Lomillos, M. E. Alonso, L. Javares-Morillo, and P. Rodríguez. 2014. “Photozoometer”: a new photogrammetric system for obtaining morphometric measurements of elusive animals. Livestock Science 165:147–156. https://doi.org/10.1016/j.livsci.2014.03.028

Gilmore, T. E., F. Birgand, and K. W. Chapman. 2013. Source and magnitude of error in an inexpensive image-based water level measurement system. Journal of Hydrology 496:178–186. https://doi.org/10.1016/j.jhydrol.2013.05.011

González, P. M. 2017. Comparación de métodos para la determinación del sexo en el Flamenco austral (Phoenicopterus chilensis). Bachelor’s thesis. Universidad CAECE, Buenos Aires, Argentina.

Gray, M. J., M. J. Chamberlain, D. A. Buehler, and W. B. Sutton. 2013. Wetland wildlife monitoring and assessment. Pages 265–318 in J. T. Anderson and C. A. Davis, editors. Wetland techniques: volume 2: organisms. Springer, Dordrecht, The Netherlands. https://doi.org/10.1007/978-94-007-6931-1_7

Griffiths, R. 2000. Sex identification in birds. Seminars in Avian and Exotic Pet Medicine 9:14–26. https://doi.org/10.1016/S1055-937X(00)80012-2

Helfenstein, F., E. Danchin, and R. H. Wagner. 2004. Assortative mating and sexual size dimorphism in Black-Legged Kittiwakes. Waterbirds: The International Journal of Waterbird Biology 27:350–354. https://doi.org/10.1675/1524-4695(2004)027[0350:AMASSD]2.0.CO;2

Heal, M., B. A. Hoover, and J. J Waggitt. 2021. Using rangefinder binoculars to measure the behaviour and movement of European Shags Phalacrocorax aristotelis in coastal environments. Bird Study 68(1):135–140. https://doi.org/10.1080/00063657.2021.1960947

Humphries, G. R. W., T. Fail, M. Watson, W. Houghton, R. Peters-Grundy, M. Scott, R. Thomson, K. Keogan, and A. Webb. 2023. Aerial photogrammetry of seabirds from digital aerial video images using relative change in size to estimate flight height. Marine Biology 170:18. https://doi.org/10.1007/s00227-022-04161-5

Jiang, R., D. V. Jáuregui, and K. R. White. 2008. Close-range photogrammetry applications in bridge measurement: literature review. Measurement 41:823–834. https://doi.org/10.1016/j.measurement.2007.12.005

Kemper, G., A. Weidauer, and T. Coppack. 2016. Monitoring seabirds and marine mammals by georeferenced aerial photography. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLI-B8:689–694. https://doi.org/10.5194/isprsarchives-XLI-B8-689-2016

Kilpatrick, A. M., J. R. Hoyt, R. A. King, H. M. Kaarakka, J. A. Redell, J. P. White, and K. E. Langwig. 2020. Impact of censusing and research on wildlife populations. Conservation Science and Practice 2:e264. https://doi.org/10.1111/csp2.264

Labocha, M., and J. Hayes. 2012. Morphometric indices of body condition in birds: a review. Journal of Ornithology 153:1–22. https://doi.org/10.1007/s10336-011-0706-1

Lambertucci, S. A., M. Carrete, K. L. Speziale, F. Hiraldo, and J. A. Donázar. 2013. Population sex ratios: another consideration in the reintroduction-reinforcement debate? PLoS ONE 8(9):e75821. https://doi.org/10.1371/journal.pone.0075821

Leorna, S., T. Brinkman and T. Fullman. 2022. Estimating animal size or distance in camera trap images: photogrammetry using the pinhole camera model. Methods in Ecology and Evolution 13:1707–1718. https://doi.org/10.1111/2041-210X.13880

Lieury, N., S. Devillard, A. Besnard, O. Gimenez, O. Hameau, C. Ponchon, and A. Millon. 2017. Designing cost-effective capture-recapture surveys for improving the monitoring of survival in bird populations. Biological Conservation 214:233–241. https://doi.org/10.1016/j.biocon.2017.08.011

Lovich, J. E., and J. W. Gibbons. 1992. A review of techniques for quantifying sexual size dimorphism. Growth Development Aging 56:269–281.

Luhmann, T., C. Fraser, and H.-G. Maas. 2016. Sensor modelling and camera calibration for close-range photogrammetry. ISPRS Journal of Photogrammetry and Remote Sensing 115:37–46. https://doi.org/10.1016/j.isprsjprs.2015.10.006

Lürig, M., S. Donoughe, E. Svensson, A. Porto, and M. Tsuboi. 2021. Computer vision, machine learning, and the promise of phenomics in ecology and evolutionary biology. Frontiers in Ecology and Evolution 09:642774. https://doi.org/10.3389/fevo.2021.642774

Martins, B. H., A. Soriano-Redondo, A. M. A. Franco, and I. Catry. 2024. Age mediates access to landfill food resources and foraging proficiency in a long-lived bird species. Animal Behaviour 207:23–36. https://doi.org/10.1016/j.anbehav.2023.10.011

Medina, J. J., J. M. Maley, S. Sannapareddy, N. N. Medina, C. M. Gilman, and J. E. McCormack. 2020. A rapid and cost-effective pipeline for digitization of museum specimens with 3D photogrammetry. PLOS ONE 15:e0236417. https://doi.org/10.1371/journal.pone.0236417

Mendes, A. S., S. J. Paixão, R. R. Sikorski, D. V. Bonamigo, M. G. Morello, and R. A. R. Ponzoni. 2016. Photogrammetry: a non-Invasive and objective method for detecting locomotion problems in broiler chickens. Revista Brasileira de Ciência Avícola 18:573–580. https://doi.org/10.1590/1806-9061-2015-0079

Mokkink, L. B., H. de Vet, S. Diemeer, and I. Eekhout. 2023. Sample size recommendations for studies on reliability and measurement error: an online application based on simulation studies. Health Services and Outcomes Research Methodology 23:241–265. https://doi.org/10.1007/s10742-022-00293-9

Montalti, D., M. G. Grilli, R. Maragliano, and G. Cassini. 2012. The reliability of morphometric discriminant functions in determining the sex of Chilean flamingos Phoenicopterus chilensis. Current Zoology 58:851–855. https://doi.org/10.1093/czoolo/58.6.851

Napoli, C., N. Hirtle, J. Stepanuk, F. Christiansen, E. I. Heywood, T. J. Grove, A. Stoller, F. Dodds, M. Glarou, M. H. Rasmussen, G. L. Lonati, K. T. A. Davies, S. Videsen, Malene J. Simon, T. K. Boye, A. Zoidis, S. K. Todd, and L. H. Thorne. 2024. Drone-based photogrammetry reveals differences in humpback whale body condition and mass across North Atlantic foraging grounds. Frontiers in Marine Science 1–18. https://doi.org/10.3389/fmars.2024.1336455

Owens, I. P. F., and I. R. Hartley. 1998. Sexual dimorphism in birds: why are there so many different forms of dimorphism? Proceedings of the Royal Society B: Biological Sciences 265:397–407. https://doi.org/10.1098/rspb.1998.0308

Perdue, B. M., D. E. Gaalema, A. L. Martin, S. M. Dampier, and T. L. Maple. 2011. Factors affecting aggression in a captive flock of Chilean Flamingos (Phoenicopterus chilensis). Zoo Biology 30:59–64. https://doi.org/10.1002/zoo.20313

Piratelli, A. J., G. R. Favoretto, and M. F. de Almeida Maximiano. 2015. Factors affecting escape distance in birds. Zoologia (Curitiba) 32:438–444. https://doi.org/10.1590/s1984-46702015000600002

Pohl, M. B., and L. Farr. 2010. A comparison of foot arch measurement reliability using both digital photography and calliper methods. Journal of Foot and Ankle Research 3:14. https://doi.org/10.1186/1757-1146-3-14

Price, M. 2008. The impact of human disturbance on birds: a selective review. Pages 163–197 in D. Lunney, A. Munn, and W. Meikle, editors. Too close for comfort: contentious issues in human-wildlife encounters. Royal Zoological Society of New South Wales, Mosman, New South Whales, Australia. https://doi.org/10.7882/FS.2008.023

R Core Team. 2023. RStudio: integrated development for R. RStudio, PBC, Boston, Massachusetts, USA. https://posit.co/download/rstudio-desktop/

Rezende, S. L., and F. Leeuwenberg. 1987. Ecological studies of Lagoa do Peixe. Final report to World Wildlife Fund. WWF, Washington, D.C., USA.

Richter, N. A., and G. R. Bourne. 1990. Sexing Greater Flamingos by weight and linear measurements. Zoo Biology 9:317–323. https://doi.org/10.1002/zoo.1430090407

Richter, N. A., G. R. Bourne, and E. N. Diebold. 1991. Gender determination by body weight and linear measurements in American and Chilean flamingos, previously surgically sexed: within-sex comparison to Greater Flamingo measurements. Zoo Biology 10:425–431. https://doi.org/10.1002/zoo.1430100506

Rose, P., and L. Soole. 2020. What influences aggression and foraging activity in social birds? Measuring individual, group and environmental characteristics. Ethology 126:900–913. https://doi.org/10.1111/eth.13067

Rowley, L. D., M. Felgueiras, G. Réthoré, F. Bragança, R. J. Thomas, E. Huckle, and R. Medeiros. 2020. Reliability of morphological criteria for sexing birds during ringing, assessed using molecular methods - a study of thirteen species of passerines and near passerines. Ringing and Migration 35:83–93. https://doi.org/10.1080/03078698.2021.2009544

Royer, E. A., and M. J. Anderson. 2014. Evidence of a dominance hierarchy in captive Caribbean flamingos and its relation to pair bonding and physiological measures of health. Behavioural Processes 105:60–70. https://doi.org/10.1016/j.beproc.2014.03.005

Sapirstein, P. 2016. Accurate measurement with photogrammetry at large sites. Journal of Archaeological Science 66:137–145. https://doi.org/10.1016/j.jas.2016.01.002

Schmitz, R. A., and G. A. Baldassarre. 1992. Contest asymmetry and multiple bird conflicts during foraging among nonbreeding American Flamingos in Yucatan, Mexico. Condor 94:254–259. https://doi.org/10.2307/1368814

Schneider, C. A., W. S. Rasband, and K. W. Eliceiri. 2012. NIH image to ImageJ: 25 years of image analysis. Nature Methods 9:671–675. https://doi.org/10.1038/nmeth.2089

Schüßler, D., M. B. Blanco, N. K. Guthrie, G. M. Sgarlata, M. Dammhahn, R. Ernest, M. R. Evasoa, A. Hasiniaina, D. Hending, F. Jan, B. le Pors, A. Miller, G. Olivieri, A. N. Rakotonanahary, S. J. Rakotondranary, R. Rakotondravony, T. Ralantoharijaona, V. Ramananjato, B. Randrianambinina, N. N. Raoelinjanakolona, E. Rasoazanabary, R. M. Rasoloarison, D. W. Rasolofoson, S. Rasoloharijaona, E. Rasolondraibe, S. H. Roberts, H. Teixeira, T. van Elst, S. E. Johnson, J. U. Ganzhorn, L. Chikhi, P. M. Kappeler, E. E. Louis, Jr., J. Salmona, and U. Radespiel. 2024. Morphological variability or inter-observer bias? A methodological toolkit to improve data quality of multi-researcher datasets for the analysis of morphological variation. American Journal of Biological Anthropology 183:60–78. https://doi.org/10.1002/ajpa.24836

Smith, A. F., P. Bongi, and S. Ciuti. 2020. Remote, non‐invasive photogrammetry for measuring physical traits in wildlife. Journal of Zoology 313:250–262. https://doi.org/10.1111/jzo.12858

Stevens, J. 2002. Applied multivariate statistics for the social sciences. Fourth edition. Lawrence Erlbaum, Mahwah, New Jersey, USA.

Studer-Thiersch, A. 2007. Tarsus length as an indication of sex in the flamingo genus Phoenicopterus. International Zoo Yearbook 24:240–243.

Tablado, Z., Y. Bötsch, V. Bókony, F. Angelier, Á. Z. Lendvai, S. Jenni-Eiermann, and L. Jenni. 2021. Factors modulating the behavioral and physiological stress responses: do they modify the relationship between flight initiation distance and corticosterone reactivity? Hormones and Behavior 132:104979. https://doi.org/10.1016/j.yhbeh.2021.104979

Torres, W. I., and K. Bierlich. 2020. MorphoMetriX: a photogrammetric measurement GUI for morphometric analysis of megafauna. Journal of Open Source Software 5(45):1825 https://doi.org/10.21105/joss.01825

Venables, W. N., and B. D. Ripley. 2013. Modern applied statistics with S-PLUS. Springer Science and Business Media, New York, New York, USA.

Volis, S., and T. Deng. 2020. Importance of a single population demographic census as a first step of threatened species conservation planning. Biodiversity and Conservation 29:527–543. https://doi.org/10.1007/s10531-019-01897-3

Wärmländer, S. K. T. S., H. Garvin, P. Guyomarc'h, A. Petaros, and S. B. Sholts. 2019. Landmark typology in applied morphometrics studies: what’s the point? Anatomical Record 302:1144–1153. https://doi.org/10.1002/ar.24005

Watson, H., M. Bolton, B. J. Heidinger, W. Boner, and P. Monaghan. 2016. Assessing the effects of repeated handling on the physiology and condition of semi-precocial nestlings. Ibis 158:834–843. https://doi.org/10.1111/ibi.12402

Webster, M., and H. D. Sheets. 2010. A practical introduction to landmark-based geometric morphometrics. Paleontological Society Papers 16:163–188. https://doi.org/10.1017/S1089332600001868

Wedekind, C. 2012. Managing population sex ratios in conservation practice: how and why. Pages 81–96 in T. Povilitis, editor. Topics in conservation biology. IntechOpen, London, UK. https://doi.org/10.5772/37601

Winkler, D. W., S. M. Billerman, and I. J. Lovette. 2020. Flamingos (Phoenicopteridae). Version 1.0. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.phoeni1.01

Yang, S., R. J. Francis, M. Holding, and R. T. Kingsford. 2024. Aerial photography and machine learning for estimating extremely high flamingo numbers on the Makgadikgadi Pans, Botswana. Global Ecology and Conservation 53:e03011. https://doi.org/10.1016/j.gecco.2024.e03011

Yosef, R. 2000. Individual distances among Greater Flamingos as indicators of tourism pressure. Waterbirds: International Journal of Waterbird Biology 23:26–31. https://doi.org/10.2307/1522143

Zemanova, M. A. 2020. Towards more compassionate wildlife research through the 3Rs principles: moving from invasive to non-invasive methods. Wildlife Biology 1:2020. https://doi.org/10.2981/wlb.00607