Evaluating techniques for determining elasmobranch body size: a review of current methodologies

Photogrammetry using distance to the individual and bar photogrammetry

Photogrammetry using distance to the individual involves estimating the size and shape of elasmobranchs by measuring the camera-to-subject distance. It requires precise distance measurement tools to ensure accurate scaling of the images. Distances to objects on the sea surface can be calculated from photographic images by measuring the angle of dip from the horizon to the object in images taken from a known height with a calibrated lens. This method is subject to common sources of error including lens distortion, angle and distance inaccuracies, and environmental influences (Gordon, 2001; Lacey et al., 2010). In the study by Lacey et al. (2010), several key sources of error were identified. The swell-induced vertical displacement, the vertical movement of both the vessel and shark due to swell, was a significant factor. A simulation study indicated that a 0.5 m swell produced a standard deviation (SD) of error at 0.068 relative to the distance, and for more typical conditions with a 0.25 m swell, the SD was calculated to be 0.034. The shark body flexing, especially during swimming, introduced inaccuracies in length measurements. This flexing was modeled as a sine wave (based on dogfish swimming) with a flex amplitude resulting in a mean length underestimation of approximately 5%. Errors from lens calibration inconsistencies and variations in the shark’s body angle relative to the camera added further variance to the measurements. Variation in measurements within image clusters, assumed to reflect measurement error from the images themselves, yielded a mean SD of 0.09 m. Combined with the model of body flexing and swell effects, the overall error was approximated at 10% of the shark’s total length.

Bar photogrammetry uses a calibrated bar of known length placed in the camera’s field of view, parallel to the animal, serving as a reference scale for measuring the size of the elasmobranchs captured in the images. Errors associated with this method are introduced by refraction and photographic alignment. In a study by Govender, Kistnasamy & Van der Elst (1991), observations revealed that light refraction through water, air, and tank glass led to a consistent overestimation of shark length by approximately 5%. To correct for this, a compensation factor was applied to measurements. In the mentioned study, measurements were taken as sharks swam closest to a reference window, with the alignment minimized for distortion. By capturing images from this closest point (50 cm from the window), distortions due to angle or depth changes were reduced.

Both mentioned techniques require little equipment and have very low cost, and they rely heavily on stable calibration and controlled imaging angles to minimize distortions and measurement biases. They additionally require the animal to be at the surface (in the first method), and parallel and close to the bar (in the latter), which significantly limits its application. Although being the first steps towards developing photogrammetry methodologies with low-cost applications, these two techniques are obsolete and less efficient compared to modern digital and automated methods and, therefore, are no longer in use.

Laser photogrammetry

Laser photogrammetry utilizes parallel and calibrated laser pointers projected onto the elasmobranch, forming a scale of known distance on the animal’s body. The spacing between the laser points provides a reference for accurate size measurements. Images should include the whole animal perpendicular to the camera (Rohner et al., 2011; Araujo et al., 2014).

Laser photogrammetry for measuring sharks and rays can be subject to several notable errors. One primary error source is image distortion due to water refraction, particularly if there is any misalignment between the lasers and the camera setup. This refraction often causes measurement overestimates, though correction factors can be applied to mitigate these distortions. The technique’s accuracy is also affected by the animal’s orientation and movement relative to the camera, as even minor deviations from perpendicular can introduce errors of <3% (e.g., Barker & Williamson, 2010; Leurs et al., 2015) and up to 5% (e.g., Bansemer & Bennett, 2009; Rogers, Cambiè & Kaiser, 2017). Consistent parallel alignment of lasers is crucial; misalignment increases error in size estimates, especially at varied distances (Jeffreys et al., 2013; Leurs et al., 2015). Field calibration, such as using fixed measurement grids or objects, can reduce these errors and improve measurement reliability for in-water applications like monitoring shark and ray morphometry and growth rates (Perry et al., 2018).

The limitations of laser photogrammetry, such as the requirement for individuals to be perpendicular to the device, at the same level as the observer, and without body flexion, make sampling complex and time-consuming to obtain high-quality images for measurement. Overcoming these challenges necessitates investing in expensive equipment to ensure the lasers’ distance, parallel alignment, and advanced image processing software. This investment reduces the primary advantage of this technique, which is its low cost. New technologies provide a three-dimensional and more realistic approach thus, laser photogrammetry will probably be replaced with more advanced methods (Pulido Mantas et al., 2023). No article using this technique was found published in the last five years, corroborating this assertion.

Stereo-DOV

Based on the articles and techniques described, and considering their advantages and disadvantages, it is evident that stereo-DOVs have a vast field of application when the sampling depth is within a diver’s reach and the diver’s presence has minimal impact on the target species. The first use of Stereo-DOV was reported in 1983 by Klimley & Brown (1983), and this technique is currently widely applied. It comprises two synchronized cameras mounted at a fixed distance and angle apart, capturing stereoscopic footage that enables precise measurements of length and size without physically handling the animals (Harvey et al., 2010). By allowing researchers to calculate distance and depth within the video frame, stereo-DOVs can effectively gather morphometric data, monitor species behavior, and assess elasmobranch population structure (Langlois et al., 2010). Stereo-DOV systems, while powerful for underwater measurements, face several limitations and sources of error. One common issue is related to camera alignment and calibration stability, which can shift due to environmental conditions like temperature and pressure changes, affecting measurement accuracy (Harvey et al., 2003). Movement and orientation of the target organism also contribute to error, as accurate length measurements require the fish to be within a specific alignment range relative to the camera. Additionally, parallax issues arise if cameras are not perfectly synchronized, particularly with fast-swimming species, leading to slight inaccuracies (Harvey & Shortis, 1996). Furthermore, variations in water clarity and lighting can degrade image quality, making it harder to achieve precise measurements in natural, variable underwater conditions. Studies using stereo-DOV systems for measuring elasmobranchs report variable error rates, often influenced by environmental conditions and species behavior. For instance, error percentages for fish length measurements can be low, with some studies reporting mean errors around 1–3% when fish are positioned within an optimal range of 3–5 m from the camera and orientation is carefully managed. In such conditions, errors typically stay below 5% for most species (Harvey et al., 2003; Delacy et al., 2017). Accuracy can decrease with increased distance from the camera and fish movement, reaching up to 10% error under suboptimal conditions, such as low visibility or rapid swimming, which affects stereo alignment and measurement precision (Langlois et al., 2010). The presence of divers restricts stereo-DOV systems to depths accessible with SCUBA equipment, typically up to 30–40 m, depending on location and environmental conditions. This constraint limits their use in surveying deeper habitats and elasmobranch populations that occur beyond safe diver-operable depths (Langlois et al., 2010).

Stereo-BRUV

This method employs stereo camera systems, typically mounted on baited frames, which are deployed either downward or horizontally to lure elasmobranchs into the cameras’ field of view. By capturing stereoscopic images, this setup facilitates accurate size measurements and behavioral observations of these species. Unlike stereo-DOV systems, it can be utilized at greater depths and for prolonged durations, as it operates independently of divers. However, Stereo-BRUV may yield biased results, particularly concerning abundance when analyzed alongside body measurements, due to the use of bait and potential disturbances from additional lighting. Sample limitations mirror those outlined in the previous section on stereo-DOV, with several studies indicating challenges arising from individuals not being fully visible in the cameras (Ryan et al., 2015; Pinte et al., 2020) or, in more severe cases, the cameras becoming partially obstructed after deployment (Goetze & Fullwood, 2013; Yon et al., 2020).

Stereo-ROV

Stereo-ROV refers to a stereo imaging system mounted on an ROV that captures high-resolution, three-dimensional images of elasmobranchs in their natural habitat. This technology enables researchers to accurately measure the size and assess the behavior of these species in situ. By utilizing dual cameras positioned at specific angles, stereo-ROVs provide precise stereoscopic measurements, allowing for enhanced data collection in environments that may be challenging or unsafe for divers to access (McLean et al., 2019).

A similar situation to stereo-BRUV applies to stereo-ROV, where accurate measurements require a higher investment in equipment and training. These robust devices are mostly designed to sample at greater depths and according to the literature, usually for additional survey goals than biometrics, like visual assessments of fish assemblages (Sward, Monk & Barrett, 2019) or industry inspections and opportunistic sightings (Todd et al., 2020). The extremely high costs of ROVs are unlikely to lead to their adoption over stereo BRUVs and DOVs, which are considerably more affordable. This makes the latter the preferred choice for student projects and other initiatives with limited budgets that do not require extensive deep sampling (Schramm et al., 2020).

Aerial photogrammetry

Aerial surveys using drones sample extremely large areas, require considerable altitude, water clarity, and clear skies, and analyze animals close to the surface. This recent technique is being used to assess species that best fit these requirements (mostly marine mammals) but has mainly been applied to studies of abundance, distribution, behavior, and shark-human interactions (Oleksyn et al., 2020; Setyawan et al., 2022). In 2021, Butcher et al. (2021) published a review focusing on how drones are currently being used for shark monitoring, estimating total lengths, and filling knowledge gaps around fundamental shark behaviors or movements, social interactions, and predation across multiple species and scenarios. This technique has also been used with fair results to estimate elasmobranchs’ length (e.g., Whitehead et al., 2022 for R. typus estimates) using a known length object as a reference (one-meter-long pole) and previously testing the methodology measuring a known size object (paddleboard) at different altitudes, reporting <2% errors. However, estimates of UAV altitude and aquatic animal depth, if uncorrected, will lead to inaccurate measurements despite high-resolution images or video frame grabs (Rex et al., 2024). Aerial photogrammetry requires specialized, costly equipment to achieve precise length measurements of elasmobranchs, making it a resource-intensive approach for accurate data collection.

Although not 100% accurate, the methodologies listed represent excellent alternatives to traditional methods, involving the capture or killing of individuals (e.g., Merly et al., 2019). It is noteworthy that these approaches can be used for various purposes such as studies of population’s abundance or structure (Barley, Meekan & Meeuwig, 2017; Tuya et al., 2021), photo-identification (Bansemer & Bennett, 2009; O’Connell & Leurs, 2016; Moreno, Solleliet-Ferreira & Riera, 2021), growth rates (e.g., Perry et al., 2018), reproductive ecology (e.g., Deakos, 2012), and morphometric traits (Carrier, Heithaus & Simpfendorfer, 2018; Rastoin-Laplane et al., 2020). They have been applied to other marine species, e.g., blue whales (Durban et al, 2016), pilot whales (Wong & Auger-Méthé, 2018), , bottlenose dolphins (Aswegen et al., 2019) and various teleost fish (Harvey et al., 2003; Schramm et al., 2020).

The initial costs associated with video techniques are often considered high compared with more traditional biometric data collection methods (especially if the cost of animal capture in more traditional methods is excluded) and may probably be the cause of their scarce application in some research fields. However, several authors mentioned that the costs may be offset by a reduction in field time, staff needed, observer bias, and the provision of a permanent visual record that allows archived images to be revisited and compared against time series data for the detection of spatial and temporal variability of a vast range of species (e.g., Bennett et al., 2016; Langlois et al., 2020). Moreover, reducing cameras’ costs due to advances in technology and monitoring procedures made the use of video techniques more affordable to researchers in recent decades (Bennett et al., 2016; Delacy et al., 2017; Pulido Mantas et al., 2023). The stereo-DOV technique in particular has potential for use in managed environments such as aquariums, where many of the issues described become less relevant: good visibility of the individuals, desensitization of animals to their environment and the diver’s presence, reducing flight risk, and the availability of technical and financial resources, although results extrapolation is limited (Ferreira et al., 2024). Non-invasive techniques for length estimates are extremely important for elasmobranch studies as explored in the present review. They represent accurate, accessible, and versatile methodologies that contribute effectively to elasmobranchs’ management and conservation.