Comparing the performance of two scientific tools for obtaining fish length measurements

Comparing the performance of two scientific tools for obtaining fish length measurements | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Comparing the performance of two scientific tools for obtaining fish length measurements Alberto García Baciero, Carlos Robalino-Mejía, César Peñaherrera-Palma, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5501285/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 27 Jun, 2025 Read the published version in Marine Biology → Version 1 posted 6 You are reading this latest preprint version Abstract Precise descriptions of size structure are crucial for the adaptive management of marine fish populations influenced by human activity and environmental factors. Stereo-video systems are powerful tools for monitoring fish populations. Yet, due to the high investment required in software and equipment, stereo-videography can have some financial issues. This study compared the performance of the commercial software EventMeasure from SeaGIS and the open-access R package StereoMorph. We evaluated the effect of distance to the system, rotation movement of the object, and true length on the accuracy and precision of each software. Additionally, we used a diver-operated stereo-video system to obtain in situ measurements of reef fish species. EventMeasure was generally more accurate (error = 0.53%) and precise (CV = 0.35%) than StereoMorph (error = 4.54%; CV = 0.65%). However, the latter showed errors < 5% when measuring objects at distances up to 4 m and close to the plane axis. The precision for in situ measurements for EventMeasure (CV = 15.0%) was similar to that of StereoMorph (Mean CV = 15.8%). We found a high correlation (ρ = 0.93, p < 0.001) between paired fish length estimation from both software, although StereoMorph was slightly more precise than EventMeasure (Mean CV = 2.07% and Mean CV = 2.11%, respectively). This open-access software provided suitable accuracy and precision results despite some limitations, offering the option to reduce software costs without compromising accuracy for affordability. Stereo-videography Fish Monitoring Open-access Measurements Length estimates Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Introduction Reef-associated fish species are experiencing remarkable changes around the globe due to habitat degradation, pollution, and overexploitation (Brandl et al. 2020 ; Strona et al. 2021 ). Quantitative assessments are then required to evaluate and monitor their populations to understand their trends and status (Heenan et al. 2017 ; Tortolero-Langarica et al. 2022 ), including both fishery-dependent and fishery-independent approaches (Pennino et al. 2016 ; Howard et al. 2023 ). While the former has gear-specific sampling biases since they may target specific taxa, size classes, and trophic levels (King 2007 ), fishery-independent methods can prove to be a valuable and cost-effective information source for assessing fish populations in coastal coral reef habitats (Brock 1954 ; Murphy and Jenkins 2010 ). Non-destructive and non-extractive techniques, such as underwater visual censuses (UVC) (Brock 1954 ), and video recording-based methods, are crucial in Marine Protected Areas (MPAs) for preserving ecosystem integrity and reducing pressure on fish populations (Murphy and Jenkins 2010 ; Goetze et al. 2015 ; García-Baciero et al. 2024 ). Although UVC methods using strip-transects are probably the most frequently used tool to collect data on fish assemblages, they show biases for reliable long‐term monitoring and estimating the length of fish and sample areas (Smith 1988 ; Harvey et al. 2004 ). Moreover, UVC methods require divers to be trained in taxonomic identification and sampling techniques and are subject to natural inter-observer variability. Additionally, UVC methods do not allow the collected data to be validated by a second observer, as it is possible with video recordings (Goetze et al. 2019 ). In this context, researchers have begun implementing other monitoring techniques, such as video-based techniques, to overcome the different sampling biases and enhance data accuracy and precision (Harvey and Shortis 1998 ; Harvey et al. 2002 ; Cappo et al. 2004 ; Harvey et al. 2004 ; Whitmarsh et al. 2017 ). Underwater video is primarily used in ecological studies to monitor fish populations (Caldwell et al. 2016 ; Marra et al. 2016 ) examine fish behavior (Ajemian et al. 2016 ), compare methodologies (Schramm et al. 2020 ; Jessop et al. 2024 ), and study length-frequency distributions (Arachchige-Weerarathne et al. 2021 ; Williams et al. 2022 ; García-Baciero et al. 2024 ). The use of stereo-video systems for monitoring fish assemblages has grown in popularity in the last two decades as it allows for more accurate measurement of fish lengths in field conditions (Harvey et al. 2004 ; Goetze et al. 2015 ; Schramm et al. 2020 ). Accurate size structure descriptions are crucial for the adaptive management of marine fish populations facing anthropogenic and environmental pressures (Barnett et al. 2017 ; Queirós et al. 2018 ). Knowing the length of sampled fish is vital to many modern stock assessments and population demographic methods (Ono et al. 2015 ). Effective monitoring programs must measure the lengths of a sufficient number of individuals within each fish population (Arachchige-Weerarathne et al. 2021 ). Size structure data offers valuable insights into the reproductive potential, growth, and stability of reef fish populations (Hixon et al. 2014 ; van Overzee and Rijnsdorp 2015 ). A scarcity of smaller fish may indicate recruitment issues, while a low occurrence of larger fish could suggest high mortality rates among mature individuals (Neumann and Allen 2007 ). Increased fishing pressure and warming temperatures drive higher mortality, altering fish stock size structure (Tu et al. 2018 ) reducing large predator abundance (Angelini et al. 2025 ). Accuracy in length measurements require a proper calibration of the stereo-video system. During the calibration, the two cameras film an object of known dimensions to account for lens distortion and their location in space relative to each other (Harvey and Shortis 1998 ). There are two calibration approaches used in underwater ecological applications with different procedures for calculating camera parameters. One method involves using a 3D calibration object marked with several points of known positions, and it requires a photogrammetric network solution to calculate the parameters (Clarke and Fryer 1998 ). The other method uses a two-dimensional (2D) chequerboard pattern consisting of known-size squares to generate the camera parameters. In the former method, the solution is determined using collinearity and least squares estimation (LSE) to accurately model the network geometry and the random errors in the measurements. In contrast, in the latter method, a four-step algorithm based on the Direct Linear Transformation (DLT) method (Abdel-Aziz 1971 ; Heikkila and Silvén 1997 ) is used to provide the initial estimates of the parameters, followed by a network solution based on collinearity. The DLT method assigns each calibrated camera a set of coefficients that help to establish a relationship between unique 3D coordinates in calibration space and their corresponding (non-unique) 2D-pixel coordinates in that specific camera view. Although both network and DLT solutions differ in the camera calibration models, they produce distinct yet transformable mathematical descriptions of the relative orientation. The 3D calibration generally yields more accurate measurements across a broader range of distances compared to 2D calibration (Boutros et al. 2015 ), with average in-water errors typically below 1% versus 1.5–10% for 2D (Williams et al. 2010 ; Boutros et al. 2015 ). However, 2D calibration can achieve average errors below 5% for stereo-video measurements up to 8 m (Boutros et al. 2015 ) and < 3% when compared to tape-measured lengths in the field (Delacy et al. 2017 ). Besides accuracy, there is also a variation in price between the two calibration approaches mentioned. The cost for a single license and 2D hardware and software calibration object ranges from $ 500 to $ 2150 (source: www.mathworks.com , accessed on April 22, 2025). On the other hand, the 3D CAL and EventMeasure software, along with the 3D cube calibration hardware, cost nearly $ 7,600 for academic/research use (source: www.seagis.com.au , accessed on April 22, 2025). Thus, 2D calibration techniques may be more cost-effective, easier to set up and execute, and practically deployable in the field compared to constructing a complex cage that is unwieldy to maneuver underwater. Although stereo-video systems have multiple advantages for surveying fish, economic resources for monitoring marine systems are limited. Many institutions or organizations cannot afford the purchase of commercial software licenses in addition to conducting research. A few free and open-source applications for length estimation have emerged recently, such as the R package StereoMorph (Olsen and Westneat 2015 ; Olsen and Haber 2022 ) and VidSync (Neuswanger et al. 2016 ) that may benefit low-budget institutions and NGOs conducting fish monitoring programs. StereoMorph has features for processing images, calibrating cameras, digitizing images, and reconstructing landmarks and curves. This enables users to transform any digitized landmarks from both camera perspectives into a 3D reconstruction (Olsen and Westneat 2015 ). VidSync is a free Mac application for scientific video analysis that allows users to efficiently record, organize, and navigate complex 2D or 3D measurements of fish and their physical habitats (Neuswanger et al. 2016 ). VidSync is a valuable free alternative to EventMeasure and StereoMorph, especially for measuring fish communities. However, it requires a Mac computer, which may be a barrier for institutions with limited budgets. The StereoMorph R package has been recently used to estimate marine species’ lengths (García-Baciero et al. 2024 ), showing accurate results compared to true-length data (Delacy et al. 2017 ; Garner et al. 2021 ). Nevertheless, their differences in obtaining accurate and precise measurements compared to commercial software have not been assessed yet. Here, we compare the performance of the StereoMorph R package and the commercial software EventMeasure for estimating the length of known-size objects in controlled conditions, underwater, and of fish individuals in the field. 2. Material and methods 2.1. The stereo-video system We used a diver-operated stereo-video system (stereo-DOV) composed of a pair of GoPro Hero7 with underwater camera housing. The cameras were mounted with a convergence angle of 4°, separated by 800 mm, on a steel base bar provided with handles. This stereo configuration allows for accurate fish measurements up to 8 m distance (Boutros et al. 2015 ; Letessier et al. 2015 ). The cameras were set up to 60 frames per second, 1080p resolution (1920×1080 pixels), and linear field of view (FOV). 2.2. Calibration procedures Before sampling, we calibrated the stereo-DOV using a 3D cube from SeaGIS (1000×1000×500 mm), and with an 8×6 square checkerboard (total dimensions 472×314 mm, with each square 52.45×52.45 mm) created using the function drawCheckerboard from StereoMorph. The laser-printed checkerboard was placed between two acrylic sheets (2 mm thick) with their borders sealed with silicon, which do not affect the accuracy of length estimations. We performed the calibration in a clear pool immobilizing the camera system underwater while we manually moved the cube and checkerboard at different angles and positions along the field of view of both cameras for three to five minutes each. To synchronize the videos, we clamped in front of the cameras before and after moving the cube and the checkerboard. In the 3D calibration method, the cube was placed at an approximate fixed distance of 2 m and then rotated in 20 different positions according to the standard calibration protocol described by the software CAL 3.3 ( https://www.seagis.com.au/bundle.html ). This software uses the locations of the targets on the cube, as measured in the video images, to calculate the internal characteristics of the two cameras (e.g., the principal distance and lens distortion) and the 3D orientations of the cameras relative to one another. In the 2D calibration method, we moved the checkerboard along the entire calibration volume, following procedures described in Olsen and Westneat ( 2015 ), and subsequently used the function calibrateCameras from StereoMorph. This function provides a completely automated workflow for stereo-camera calibration, performing 3 main steps: checkerboard corner detection, undistortion coefficient estimation, and calibration coefficient estimation. StereoMorph determines the DLT coefficients using the internal corners automatically detected from the checkerboard pattern, which is photographed from both camera views at different positions and angles within the calibration space (Olsen and Westneat 2015 ). Rather than estimate the DLT coefficients directly, StereoMorph estimates the six transformation parameters (3 translations, 3 rotations) required to transform the first checkerboard into each subsequent checkerboard in 3D space by minimizing the reconstruction error. These transformation parameters are then used to generate 3D coordinates (the equivalent of a calibration object) and calculate the DLT coefficients (Olsen and Westneat 2015 ). 2.3. Data collection We first collected measurements from known-length objects in pool and ocean environments. Then, we collected video recordings in a shallow reef to obtain fish measurements in real oceanic conditions. Measurements of known-length objects were collected first in controlled conditions in a swimming pool and then in a shallow rocky reef in the southern Gulf of California, located in front of “El Saltito” beach (24°15′06.16” N, 110°09′18.17” W) near La Paz, Baja California Sur, Mexico. In the pool environment, we obtained measurements from two known-length objects to increase the range of measurements. The first object (manufactured by SeaGIS; Fig. 1a) was a 1 m long scale bar marked with circular, reflective targets, allowing measurements at three different lengths: 345 (short), 575 (medium), and 915 mm (long). The second object (an 80 cm long PVC tube, Fig. 1b) had six marks separated by 100 mm each, allowing the collection of measurements at 100, 200, 300, 400, 500, and 600 mm. The recordings were collected in the morning at 9:00, right after the pool was cleaned, to ensure optimal water transparency. We repeated the filming procedure of the two objects by rotating them around their Y- and Z-axis at a given degree (Fig. 2 ), and at distances separated by 1 m until reaching the end of the pool at 6 m. Thus, we obtained measurements at distances of 2, 3, 4, and 5 m from the cameras. The ocean environment consisted of patchy rocky areas of various sizes in the surroundings of two small islets. To ensure the best visibility possible, we collected the data at midday between 12:00 to 13:00 using the stereo-DOV system right after calibration. We filmed the PVC object for 5 minutes at approximately 3 m and using the same angles as in the pool environment. We estimated measurements only from the PVC object to not compromise the SeaGIS-manufactured object. Finally, reef fish were measured over the same rocky reef swimming at constant low speed for 20 minutes at a depth of 8 m while pointing the cameras straight ahead along the reef to film as many fish individuals as possible. Fish species were identified on-site and corroborated using available fish identification guides for the area. 2.4. Length measurements We analyzed the video footage obtained with the stereo-DOV by using EventMeasure software (SEAGIS Pty Ltd) and the StereoMorph package in the R environment to estimate the length of the two objects and fish individuals. Although StereoMorph does not have a tool to measure 3D lengths, an estimation can be done by selecting landmarks to define the two endpoints that will produce the measure of interest (Olsen and Westneat 2015 ). We reconstructed the selected landmarks into 3D using the function reconstructStereoSets and we then calculated the lengths by measuring the distance between the resulting 3D marker positions. In the pool environment, we obtained 300 measurements of the SeaGIS object with each software tool by taking five replicates of each true length with each rotation movement and distance combination. Likewise, we obtained 600 measurements per software of the PVC object, consisting of 5 measurements of each true length for the same rotation movement and distance combinations. In the ocean environment, we obtained 150 measurements per software of the PVC object, consisting of 5 measurements of each true length for the five rotation movements at a fixed distance of 3 m. Moreover, we measured the standard length (SL) of 47 fish individuals belonging to three species: Thalassoma lucasanum (1 individual), Abudefduf troschelii (6 individuals), and Mulloidichthys dentatus (40 individuals). We used three different frames of the same individual as replicates, accounting for 141 measurements obtained with each software. Frames in the pool and ocean environments were the same for both software to improve accuracy comparisons. Fish standard lengths were used rather than total length as it is more easily definable across a range of species. 2.5. Data analysis We assessed the performance of both software by comparing the accuracy and precision of the estimated measures. Accuracy is defined as the closeness of an estimated measurement to the real or true values, in this case of the known-lengths objects assessed, whereas precision refers to the consistency of repeated measurements of the same object (Harvey and Shortis 1995 ). The less the differences among repeated measures the larger the measurement precision. We evaluated the software’s accuracy in the pool and the ocean environment by comparing differences in the Relative Error (RE), a metric estimated from the ratio of the absolute error (true length minus measured length) to the true length, as shown in the following formula: $$\:RE=\:\frac{|measured\:length-true\:length|}{true\:length}\times\:100$$ We then calculated the mean RE (MRE) by averaging the RE obtained per rotation movement and distance, respectively. For illustration purposes, RE vs. true length were plotted by software type for each rotation movement and distance to the system, including linear trends for guidance. A multiple linear regression model was then fitted to explore the relationship between the RE (response variable) with the rotation movement (0, Z1, Z2, Y1 and Y2), distance (2, 3, 4, and 5 m), software (EventMeasure, StereoMorph) and true length (100, 200, 300, 345, 400, 500, 575, 600 and 915 mm) as explanatory variables for the pool environment, and rotation movement, software, and true length (100, 200, 300, 400, 500, and 600 mm) for the ocean environment. We evaluated the precision of each software in the pool environment by calculating the coefficient of variation (CV, %), which is the ratio of the standard deviation to the mean. The CV also measures the dispersion of the data regardless of the measurement units and magnitude of the values and is widely used for assessing measure precision in field research (Harvey et al. 2002 ). In the ocean environment, we first calculated the CV of the measurements from the PVC object at different angles at a fixed distance (3 m) using five replicates with each parameter. Then, we compared the precision of each software to estimate fish lengths by comparing the CV of the mean SL obtained from three replicates. Additionally, we used a Spearman correlation analysis to evaluate the relationship of paired fish length estimations between software. We performed all statistical analyses and graphs in the R environment and associated packages (R Core Team 2024 ). 3. Results 3.1. Known-length object measurements 3.1.1. Pool environment Multiple linear regression model showed significant effect of software output with StereoMorph having lower measurements accuracy (Table 1). While the RE increased with distance for both systems and differed according to the rotation of the object (Estimate = 0.50, p < 2.16×10 -5 ), the RE was much more noticeable in StereoMorph on objects with rotation movement of 315º around the Y-axis (Estimate = 5.54, p < 2.2×10 -16 ). The accuracy of the measurements was also affected significantly by true length (Estimate = 0, p < 2.16×10 -5 ). Table 1. Multiple linear regression model ranking results of relative errors (RE) for each measurement versus the type of software (categorical), distance to the system (continuous), the true length of the object (continuous), and rotation movement (categorical). Notations: Z1 and Z2 (rotation around Z-axis 45° and 315°, respectively) and Y1 and Y2 (rotation around Y-axis 45° and 315°, respectively). RE Predictors Estimates CI p (Intercept) -2.06 -2.55 – -1.56 <0.001 StereoMorph 4.14 3.92 – 4.36 <0.001 Distance 0.50 0.40 – 0.60 <0.001 Z1 0.30 -0.05 – 0.66 0.093 Z2 0.24 -0.12 – 0.59 0.191 Y1 1.86 1.51 – 2.21 <0.001 Y2 5.54 5.19 – 5.89 <0.001 True length -0.00 -0.00 – -0.00 <0.001 Observations 1800 R 2 adjusted 0.609 EventMeasure was more accurate when estimating measures of known-length objects compared to StereoMorph (MRE EM = 0.53%; MRE SM = 4.54%). MRE increased by distance and differed by the object’s rotation movement. In contrast, both parameters decreased with increasing true lengths in both measuring tools, although this effect was more accentuated in StereoMorph (Supplementary Table 1). EventMeasure exhibited a MRE <1% at every distance and close to 1% with all rotation movements and for each true length, while StereoMorph showed MRE values under 5% only when the object was not moved or was rotated around the Z-axis. Error estimations for each software were almost negligible at closer distances (< 3 m) and when objects were not rotated or were only rotated over their z-axis (Fig. 3). RE decreased for larger measurements in both tools, except when measuring with StereoMorph at 3 and 4 m distances and when the object was rotated 45º around the Y-axis (Y1 in Fig. 3). Similarly, to the accuracy metrics, EventMeasure measurements (CV = 0.35%) were more precise than StereoMorph (CV = 0.65%). CV was below 1% for measurements obtained with each software regardless of the distance, rotation movement, or true length, except when the object’s length was 100 mm long (CV EM = 1.16%; CV SM = 1.34%; Table 2) and when using StereoMorph with a rotation of 315° around the Y axis (CV = 1.27%). Precision decreased when the objects were rotated around the Y-axis with both measuring software as well as with increasing distance but increased with object size (Table 2). Table 2. Mean standard length (SL, mm) with standard error (SE) and mean coefficient of variation (CV, %) of estimated measurements by software at different rotation movements, distances (m), and true lengths (mm). Notations: 0 (no rotation), Z1 and Z2 (rotation 45° and 315°, respectively, around Z-axis) and Y1 and Y2 (rotation 45° and 315°, respectively, around Y-axis). EventMeasure StereoMorph Rotation movement Mean SL ± SE CV Mean SL ± SE CV 0 437.97 ± 0.20 0.14 431.38 ± 0.36 0.22 Z1 437.89 ± 0.25 0.17 428.55 ± 0.92 0.67 Z2 438.06 ± 0.27 0.19 429.49 ± 0.49 0.30 Y1 437.67 ± 0.63 0.48 418.85 ± 1.17 0.77 Y2 438.32 ± 0.89 0.77 389.69 ± 1.72 1.27 Distance 2 436.87 ± 0.20 0.13 429.27 ± 0.62 0.42 3 438.41 ± 0.45 0.34 423.11 ± 0.51 0.51 4 438.06 ± 0.44 0.32 415.04 ± 1.55 0.80 5 438.59 ± 0.71 0.61 410.95 ± 1.19 0.86 True length 100 100.32 ± 0.52 1.14 95.49 ± 0.55 1.34 200 201.84 ± 0.46 0.50 192.23 ± 0.74 0.89 300 300.72 ± 0.56 0.41 286.42 ± 0.93 0.75 400 400.32 ± 0.46 0.26 381.18 ± 1.01 0.61 500 500.79 ± 0.45 0.20 476.67 ± 1.24 0.60 600 601.08 ± 0.54 0.20 573.01 ± 1.33 0.53 345 345.86 ± 0.31 0.20 332.13 ± 0.67 0.46 575 572.44 ± 0.34 0.13 554.73 ± 0.84 0.34 915 918.47 ± 0.41 0.10 884.47 ± 1.09 0.28 3.1.2. Ocean environment The results from the regression analysis showed that StereoMorph affected significantly and negatively the accuracy of the measurements (Estimate = 4.71, p = 2×10 -16 ). The accuracy of the measurements decreased significantly when the object was rotated around the Y-axis, but it was not affected by true length (Table 3). Table 3. Multiple linear regression model ranking results of relative errors (RE) for each measurement versus type of software (categorical), true length (continuous), and rotation movement (categorical). Z1 and Z2 refer to rotating the object 45° and 315° around the Z-axis, respectively, and Y1 and Y2 to 45° and 315° around the Y-axis, respectively. RE Predictors Estimates CI p (Intercept) -1.12 -2.34 – 0.09 0.069 StereoMorph 4.71 3.95 – 5.47 <0.001 Z1 -0.16 -1.36 – 1.04 0.788 Z2 -0.00 -1.20 – 1.20 0.994 Y1 3.29 2.09 – 4.49 <0.001 Y2 8.74 7.54 – 9.94 <0.001 True length -0.00 -0.00 – 0.00 0.320 Observations 300 R 2 adjusted 0.607 In the ocean environment, EventMeasure exhibited greater accuracy than StereoMorph (MRE EM = 0.85% vs . MRE SM = 5.56%), both when comparing by object’s true length and by rotation movement (Supplementary Table 2). However, MRE grouped by true length showed values ranging between 0.21 and 1.34% for EventMeasure, and between 0.36-1.81% for StereoMorph for rotations around the Z-axis (Supplementary Table 2). There was also a noticeable increase in the MRE when measuring objects with StereoMorph, especially when they were rotated around the Y-axis (Fig. 4). EventMeasure and StereoMorph exhibited MRE under 1% when the objects were not moved nor rotated, excepting when rotation was over the Y-axis (Fig. 4). The CV decreased as true length increased, especially in EventMeasure (Table 4). We obtained the smallest CV when true lengths were higher than 400 mm, whereas the biggest CV corresponded to the shortest true lengths for both measuring tools. The mean CV was significantly higher with rotations around the Y-axis, especially for StereoMorph (CV angle Y1, Y2 = 1.49%) compared to EventMeasure (CV angle Y1 =0.89% and CV angle Y2 = 1.29%). Table 4 . Mean coefficient of variation (CV, %) for each software with different rotation movements and true lengths (mm) of the PVC object in the ocean environment. Notations: 0 (no rotation), Z1 and Z2 (rotation around Z-axis 45° and 315°, respectively) and Y1 and Y2 (rotation around Y-axis 45° and 315°, respectively) EventMeasure StereoMorph Rotation movement Mean SL ± SE CV (%) Mean SL ± SE CV (%) 0 352.0 ± 0.33 0.29 347.09 ± 0.34 0.28 Z1 351.36 ± 0.47 0.23 350.19 ± 0.48 0.40 Z2 352.22 ± 0.30 0.47 346.22 ± 0.39 0.44 Y1 351.55 ± 1.08 0.90 321.93 ± 1.79 1.49 Y2 352.15 ± 1.23 1.29 289.07 ± 1.47 1.49 True Length 100 101.10 ± 0.78 1.71 94.32 ± 0.70 1.70 200 202.66 ± 0.64 0.70 190.72 ± 0.78 0.96 300 301.66 ± 0.60 0.45 283.17 ± 0.85 0.70 400 401.17 ± 0.72 0.40 377.10 ± 1.08 0.67 500 502.00 ± 0.60 0.27 472.41 ± 1.08 0.53 600 602.54 ± 0.76 0.28 567.69 ± 0.88 0.37 3.2. Reef fish measurements We estimated the standard length (SL) of 47 fish individuals, obtaining a total of 153 measurements with each software (Supplementary Table 3). We obtained very similar length estimates when using EventMeasure (Mean SL = 223.2 mm, range = 143.2-281.4 mm) and StereoMorph (Mean SL = 227.2 mm, range = 141.1-302.9 mm). Precision for EventMeasure (Mean CV = 15.0%) was similar to that of StereoMorph (Mean CV = 15.8%). We found a strong monotonic correlation between paired mean length estimates by EventMeasure and StereoMorph (Spearman’s ρ= 0.93, p < 0.001; Fig. 5). StereoMorph showed a slightly better precision in estimating measurements for paired individuals (Mean CV = 2.07%, range = 0.20 - 9.25%), than EventMeasure (Mean = 2.11%, range = 0.19 - 6.42%) (Supplementary Table 3). 4. Discussion The accuracy and precision of measurements obtained with the open-access tool StereoMorph and with the commercial software EventMeasure were particularly similar when objects or fish were close to the plane axis. StereoMorph also performed well when measuring objects ranging in size from 100 to 915 mm, at distances up to 4 m. Nevertheless, EventMeasure provided more accurate and precise results than StereoMorph when measuring objects rotated over their y-axis. Generally, errors increased with increasing distances and decreased with increasing object size, yet these effects were further accentuated in StereoMorph. However, with errors below 2.5% when measuring objects at 3 m in the ocean, StereoMorph may improve precision compared to traditional visual size estimation, as it requires length-measurement training to obtain accurate in situ measurements. As the cost of living and research surge worldwide (Woolston 2023 ), cost-effective tools are needed to implement or maintain long-term monitoring programs dedicated to assessing the biodiversity and biomass of marine ecosystems. These results provided scientific evidence to support the use of this open-access tool for estimating fish lengths, particularly under some specific circumstances discussed below. Calibration is a critical step when estimating fish lengths using stereo-video systems. Average errors for in-water measurements are usually < 1% for 3D calibrations while for 2D calibrations, errors have ranged from 1.5–10% (Boynton and Voss 2006 ; Williams et al. 2010 ; Boutros et al. 2015 ). 2D and 3D calibration techniques might also show different outcomes due to the approaches they used to calculate the camera parameters, as there is a greater variability introduced through the calibration process in the 2D system compared to the 3D approach (Boutros et al. 2015 ). For example, Boutros et al. ( 2015 ) reported that stereo-video measurements obtained after using a 2D calibration approach produced measurement errors below 5% on average to distances up to 8 m from the cameras, while Delacy et al. ( 2017 ) reported errors below 3% in comparison with tape-measured lengths of sharks in the field. Our results using StereoMorph produced a MRE of 4.54%, similar to that reported by Letessier et al. ( 2015 ), who measured ten fixed target lengths (50 to 900 mm) at six distances (1 to 7 m) in a pool environment and obtained errors of < 5% at angles less than 10° to the optical axis and a distance closer to 3 m. Those results are larger than what was obtained by using the 3D calibrated software EventMeasure in our study (MRE = 0.53%), other studies using EventMeasure (Harvey et al. 2010 ; Boutros et al. 2015 ), and when using other 3D calibrated software such as VidSync (Neuswanger et al. 2016 ; López-Macías et al. 2023 ). Such differences in error estimations have been previously reported by Boutros et al. ( 2015 ) and Neuswanger et al. ( 2016 ), and are expected to occur even in measurements done with the same equipment but with different software and measuring approaches (Shafait et al. 2017 ; López-Macías et al. 2023 ). In the case of the 2D calibration approach, differences in MRE can arise from choosing the right checkerboard square size and the number of corners. Successful calibrations depend upon the size of desired target observations and water conditions (Boutros et al. 2015 ; Olsen and Westneat 2015 ). We used a checkerboard of 472 × 314 mm and obtained MREs between 0.7 and 3.5% for objects measured perpendicular to the optical axis at distances from 2 to 5 m in pool conditions. Our results improved on some previous 2D checkerboard-based calibrations, such as those performed by Olsen and Westneat ( 2015 ), who achieved an MRE of 11.6% using a smaller calibration checkerboard (270 × 210 mm). Contrarily, these authors obtained smaller errors (1.7%) when calibrating using a checkerboard of 360 × 280 mm in size. Similarly to us, Wehkamp and Fischer ( 2014 ) reported MREs of 2.5% and less after calibrating the cameras using a checkerboard of dimensions 290 × 210 mm (presumed). Delacy et al. ( 2017 ) obtained MREs of < 1% at angles up to 20°, whereas Garner et al. ( 2021 ) achieved MREs within the ± 5% error threshold they selected at angles ≤ 10° at several distances. However, our checkerboard size was smaller than the one used by Delacy et al. ( 2017 ) (571 × 317 mm) and by Garner et al. ( 2021 ) (610 × 457 mm), and it was detected by StereoMorph to a maximum distance of 6 m. Therefore, we recommend using a larger checkerboard, which improves calibration accuracy (Boutros et al. 2015 ). Olsen and Westneat ( 2015 ) recommended using a checkerboard image at least 40 pixels wide with as many internal corners as possible to maximize calibration accuracy since more internal corners can increase calibration accuracy by providing more data points, particularly for correcting lens distortion. Additionally, camera resolution and water conditions (e.g. poor visibility) can negatively affect the reconstructed 3D space within a software since they can influence the fish silhouette identification and, therefore, the range and angles at which fish can be correctly identified and accurately measured (Harvey and Shortis 1995 ; Savina et al. 2018 ; Goetze et al. 2019 ). Most studies using expensive high-end camera technology are limited to distances of 8–10 m to achieve accurate length estimates (Goetze and Fullwood 2013 ; Santana-Garcon et al. 2014 ; Delacy et al. 2017 ), even if water visibility exceeds 20 m. For the case of GoPro system configurations, it has been suggested that measurements be restricted to when the target is at distances closer than 5 m (Letessier et al. 2015 ) to overcome limitations related to the distance, encounter rate, and angles at which fish can be seen from the stereo-video system (Santana‐Garcon et al. 2014; García-Baciero et al. 2024 ). Our setup used GoPro cameras in an environment with uncontrolled conditions, showing error values slightly higher in the ocean than in the pool trial for both measuring tools. Freshwater and saltwater have different densities, affecting diffraction and distortion angles, which can influence measurement quality. Moreover, given that calibration was performed under high light levels in a pool, we recommend investigating the impact of lower light conditions (e.g., deeper water, cloudy days) on measurement accuracy. However, while there is a big difference in the overall error between both software, the MRE ranged from 0.36 to 1.81% for StereoMorph in angles perpendicular to the camera’s optical axis and for all object’s true lengths. These values were obtained at distances less than 3 m, which are well within the suggested distance for measuring fish in ocean conditions. Monitoring fish poses further challenges as length estimations may be affected by random factors such as water turbidity, fish movement, low light, fish camouflage with the background, or even fish size (Shortis et al. 2013 ). This makes it more challenging to place the reference points and, thus, get more accurate and precise measurements than in a controlled environment (Harvey et al. 2010 ; Neuswanger et al. 2016 ). Stereo-video systems have proved to detect changes in fish length-frequency distributions, and that size structure information is similar to other sampling methods such as line and trap data (Langlois et al. 2012 ). The present study did not intend to compare fish size structures across software because we collected too few fish measurements per species to represent their size structure accurately (Arachchige-Weerarathne et al. 2021 ). Nonetheless, we found that StereoMorph produced length estimates similar to those obtained with EventMeasure for common reef fish species in the Mexican Pacific, including the Mexican goatfish ( Mulloidichthys dentatus ). Measurements of this species with StereoMorph showed similar precision (Mean CV = 15.8%) to EventMeasure (Mean CV = 15.0%), indicating that both software can be used interchangeably for measuring fish of comparable sizes, assuming measurements are made when fish are laterally positioned relative to the camera. Additionally, given that M. dentatus can reach up to 40 cm in total length and that a related species grows at an average rate of ~ 0.30 cm yr − 1 (Mehanna et al. 2018 ), StereoMorph is well-suited for detecting annual growth rate changes in reef fish species, even those with small growth rates. Given the rapid decline and threat of extinction facing many fish populations, accurate length data is crucial for guiding conservation and management efforts, particularly in low-income areas where invasive sampling is unsuitable due to low population levels. Currently, the main drivers affecting fish populations are climate change(Brander 2007 ; Free et al. 2019 ; Whitfield et al. 2023 ), causing species distributions to move toward higher latitudes and deeper areas to track optimum temperature ranges (Perry et al. 2005 ; Pinsky et al. 2020 ), and overfishing (Jackson et al. 2001 ; Myers and Worm 2003 ; Mullon et al. 2005 ). Data on fish length estimates provide insights into the reproductive potential, growth, and stability of fish populations (Hixon et al. 2014 ; van Overzee and Rijnsdorp 2015 ; Audzijonyte et al. 2020 ). Moreover, length composition analysis can provide insights into the dynamics of fish populations (Pet et al. 1997 ; Audzijonyte et al. 2020 ). A lack of smaller fish may indicate recruitment challenges, while a low number of larger fish could signal high mortality rates among mature individuals (Neumann and Allen 2007 ). Overfishing impacts population size structure by removing large and valuable individuals (Fenberg and Roy 2008 ; Shantz et al. 2020 ), leading to increased mortality that shortens lifespans and reduces the average size of individuals (Kuparinen et al. 2016 ; Alonso-Fernández et al. 2021 ). Overexploited areas thus exhibit smaller average body sizes in targeted species, which negatively impacts populations by decreasing productivity and delaying maturation (Hamilton et al. 2007 ; Conover et al. 2009 ). Given the shifts in fish populations due to anthropogenic stressors, obtaining accurate fish length estimates is therefore essential for implementing management and conservation strategies. 5. Conclusion Generally, EventMeasure demonstrated to be a more accurate and precise tool than StereoMorph. However, this open-access 2D calibration-based software can produce similar errors to commercial 3D calibration-based software when objects or fish are close to the plane axis and at distances up to 4 m. Additionally, both measuring tools produced similar precision results when estimating the lengths of paired fish individuals. This suggests that StereoMorph performs well under equal circumstances. As a result, this free and open-access tool can be considered highly valuable for low-budget institutions conducting reef fish monitoring programs aiming at estimating fish size trends without having to spend significant economic resources. The stereo-video system produced measurements where the error increased with increasing distances and decreased with increasing object size. With errors below 2.5% when measuring objects at 3 m in the ocean, this may improve efficiency compared to traditional visual size estimation when conducting UVC methods, as it requires length-measurement training to obtain accurate in situ measurements. Until now, those wanting to use stereo-video have generally had to invest a significant amount of money in calibration equipment and commercial software, which may discourage researchers in developing countries from using stereo-video as a survey tool. As this open-access software reduces the cost of software and calibration equipment, we encourage more researchers to use stereo-video systems to enhance length estimate data. To achieve this, we recommend measuring fish close to the plane axis to avoid adding lens distortion bias into measurements. Furthermore, it is recommended to use this configuration to provide precise measurements for fish not less than 100 mm in size unless they are closer than 3 m to the cameras. These considerations will allow using StereoMorph to estimate fish length-frequency distribution using accurate and precise length estimations while avoiding error magnification during the calculations (Harvey et al. 2002 ). Finally, by following an appropriate camera configuration and 2D calibration, StereoMorph could readily be used with stereo-DOVs to detect changes in the growth rates of reef fish species annually, even for those species with small growth rates. Declarations Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. We would like to thank the NGOs Pelagios Kakunjá and MigraMar for letting us access the calibration and measuring software from SeaGIS. We would also like to extend our gratitude to Tu Polideportivo La Paz for allowing us to calibrate the stereo-video system in their sports facilities. Data availability Data will be made available on request. References Abdel-Aziz YI (1971) Direct linear transformation from comparator coordinates into object space coordinates in close range photogrammetry. Proc Amer Soc Photogrammetry 1971:1–19. https://doi.org/10.14358/PERS.81.2.103c Ajemian M, Kenworthy M, Sánchez-Lizaso JL, Cebrian J (2016) Aggregation dynamics and foraging behaviour of striped red mullet Mullus surmuletus in the western Mediterranean Sea. J Fish Biol 88:2051–2059. https://doi.org/10.1111/jfb.12932 Alonso-Fernández A, Otero J, Bañón R (2021) Supplementary Material Indicators of body size variability in a highly developed small-scale fishery: ecological and management implications https://doi.org/10.1016/j.ecolind.2020.107141 Angelini R, Lima MAL, Lira AS, Lucena-Frédou F, Frédou T, Bertrand A, Giarrizzo T, Steenbeek J, Coll M, Keppeler FW (2025) The projected impacts of climate change and fishing pressure on a tropical marine food web. Mar Environ Res 204:106909. https://doi.org/10.1016/j.marenvres.2024.106909 Arachchige-Weerarathne I, Monk J, Barrett N (2021) Sample-size requirements for accurate length-frequency distributions of mesophotic reef fishes from baited remote underwater stereo video. Ecol Ind 122:107262. https://doi.org/10.1016/j.ecolind.2020.107262 Audzijonyte A, Richards SA, Stuart-Smith RD, Pecl G, Edgar GJ, Barrett NS, Payne N, Blanchard JL (2020) Fish body sizes change with temperature but not all species shrink with warming. Nat Ecol Evol 4:809–814. https://doi.org/10.1038/s41559-020-1171-0 Barnett LA, Branch TA, Ranasinghe RA, Essington TE (2017) Old-growth fishes become scarce under fishing. Curr Biol 27:2843–2848 e2842 doi. http://dx.doi.org/10.1016/j.cub.2017.07.069 Boutros N, Shortis MR, Harvey ES (2015) A comparison of calibration methods and system configurations of underwater stereo-video systems for applications in marine ecology. Limnol Oceanography: Methods 13:224–236. https://doi.org/10.1002/lom3.10020 Boynton GC, Voss KJ (2006) An underwater digital stereo video camera for fish population assessment. University of Miami, Tech Rep Brander KM (2007) Global fish production and climate change. Proceedings of the national academy of sciences 104: 19709–19714 https://doi.org/10.1073/pnas.0702059104 Brandl SJ, Johansen JL, Casey JM, Tornabene L, Morais RA, Burt JA (2020) Extreme environmental conditions reduce coral reef fish biodiversity and productivity. Nat Commun 11:3832. https://doi.org/10.1038/s41467-020-17731-2 Brock VE (1954) A preliminary report on a method of estimating reef fish populations. J Wildl Manag 18:297–308. https://doi.org/10.2307/3797016 Caldwell ZR, Zgliczynski BJ, Williams GJ, Sandin SA (2016) Reef fish survey techniques: assessing the potential for standardizing methodologies. PLoS ONE 11:e0153066. https://doi.org/10.1371/journal.pone.0153066 Cappo M, Speare P, De'ath G (2004) Comparison of baited remote underwater video stations (BRUVS) and prawn (shrimp) trawls for assessments of fish biodiversity in inter-reefal areas of the Great Barrier Reef Marine Park. J Exp Mar Biol Ecol 302:123–152. https://doi.org/10.1016/j.jembe.2003.10.006 Clarke TA, Fryer JG (1998) The development of camera calibration methods and models. Photogram Rec 16:51–66. https://doi.org/10.1111/0031-868X.00113 Conover DO, Munch SB, Arnott SA (2009) Reversal of evolutionary downsizing caused by selective harvest of large fish. Proceedings of the Royal Society B: Biological Sciences 276: 2015–2020 https://doi.org/10.1098/rspb.2009.0003 Delacy CR, Olsen A, Howey LA, Chapman DD, Brooks EJ, Bond ME (2017) Affordable and accurate stereo-video system for measuring dimensions underwater: a case study using oceanic whitetip sharks Carcharhinus longimanus. Mar Ecol Prog Ser 574:75–84. https://doi.org/10.3354/meps12190 Dunbrack RL (2006) In situ measurement of fish body length using perspective-based remote stereo-video. Fish Res 82:327–331. https://doi.org/10.1016/j.fishres.2006.08.017 Fenberg PB, Roy K (2008) Ecological and evolutionary consequences of size-selective harvesting: how much do we know? Mol Ecol 17:209–220. https://doi.org/10.1111/j.1365-294X.2007.03522.x Free CM, Thorson JT, Pinsky ML, Oken KL, Wiedenmann J, Jensen OP (2019) Impacts of historical warming on marine fisheries production. Science 363:979–983. https://doi.org/10.1126/science.aau1758 García-Baciero A, Borges-Souza JM, Palomares-García JR, Rodríguez-Sánchez R, Rubio-Rodríguez U, Villalobos H (2024) Monitoring fish populations using stereo-DOV-based surveys and open-access tools in the Gulf of California. 103926. Regional Studies in Marine Science https://doi.org/10.1016/j.rsma.2024.103926 Garner SB, Olsen AM, Caillouet R, Campbell MD, Patterson WF III (2021) Estimating reef fish size distributions with a mini remotely operated vehicle-integrated stereo camera system. PLoS ONE 16:e0247985. https://doi.org/10.1371/journal.pone.0247985 Goetze J, Fullwood L (2013) Fiji’s largest marine reserve benefits reef sharks. Coral Reefs 32:121–125. https://doi.org/10.1007/s00338-012-0970-4 Goetze J, Jupiter S, Langlois TJ, Wilson S, Harvey ES, Bond T, Naisilisili W (2015) Diver operated video most accurately detects the impacts of fishing within periodically harvested closures. J Exp Mar Biol Ecol 462:74–82. https://doi.org/10.1016/j.jembe.2014.10.004 Goetze JS, Bond T, McLean DL, Saunders BJ, Langlois TJ, Lindfield S, Fullwood LA, Driessen D, Shedrawi G, Harvey ES (2019) A field and video analysis guide for diver operated stereo-video. Methods Ecol Evol 10:1083–1090. https://doi.org/10.1111/2041-210X.13189 Hamilton SL, Caselle JE, Standish JD, Schroeder DM, Love MS, Rosales-Casian JA, Sosa-Nishizaki O (2007) Size-selective harvesting alters life histories of a temperate sex‐changing fish. Ecol Appl 17:2268–2280. https://doi.org/10.1890/06-1930.1 Harvey E, Fletcher D, Shortis MR, Kendrick GA (2004) A comparison of underwater visual distance estimates made by scuba divers and a stereo-video system: implications for underwater visual census of reef fish abundance. Mar Freshw Res 55:573–580. https://doi.org/10.1071/MF03130 Harvey E, Shortis M (1995) A system for stereo-video measurement of sub-tidal organisms. Mar Technol Soc J 29:10–22 Harvey E, Shortis M, Stadler M, Cappo M (2002) A comparison of the accuracy and precision of measurements from single and stereo-video systems. Mar Technol Soc J 36:38–49. https://doi.org/10.4031/002533202787914106 Harvey ES, Goetze J, McLaren B, Langlois T, Shortis MR (2010) Influence of range, angle of view, image resolution and image compression on underwater stereo-video measurements: high-definition and broadcast-resolution video cameras compared. Mar Technol Soc J 44:75–85. https://doi.org/10.4031/MTSJ.44.1.3 Harvey ES, Shortis MR (1998) Calibration stability of an underwater stereo-video system: implications for measurement accuracy and precision. Mar Technol Soc J 32:3–17 Heenan A, Williams ID, Acoba T, DesRochers A, Kosaki RK, Kanemura T, Nadon MO, Brainard RE (2017) Long-term monitoring of coral reef fish assemblages in the Western central pacific. Sci Data 4:1–12. https://doi.org/10.1038/sdata.2017.176 Heikkila J, Silvén O (1997) A four-step camera calibration procedure with implicit image correction Proceedings of IEEE computer society conference on computer vision and pattern recognition. IEEE, pp 1106–1112 Hixon MA, Johnson DW, Sogard SM (2014) BOFFFFs: on the importance of conserving old-growth age structure in fishery populations. ICES J Mar Sci 71:2171–2185. https://doi.org/10.1093/icesjms/fst200 Howard RA, Ciannelli L, Wakefield WW, Haltuch MA (2023) Comparing fishery-independent and fishery-dependent data for analysis of the distributions of Oregon shelf groundfishes. Fish Res 258:106553. https://doi.org/10.1016/j.fishres.2022.106553 Jackson JB, Kirby MX, Berger WH, Bjorndal KA, Botsford LW, Bourque BJ, Bradbury RH, Cooke R, Erlandson J, Estes JA (2001) Historical overfishing and the recent collapse of coastal ecosystems. science 293: 629–637 doi https://doi.org/9601 Jessop SA, Saunders BJ, Goetze JS, Barrett NS, Harvey ES (2024) A comparison of the behavioural responses of fishes to a remotely operated vehicle and diver-based stereo-video sampling. Estuar Coast Shelf Sci 298:108621. https://doi.org/10.1016/j.ecss.2024.108621 King M (2007) Fisheries biology, assessment and management Kuparinen A, Boit A, Valdovinos FS, Lassaux H, Martinez ND (2016) Fishing-induced life-history changes degrade and destabilize harvested ecosystems. Sci Rep 6:22245. https://doi.org/10.1038/srep22245 Langlois TJ, Fitzpatrick BR, Fairclough DV, Wakefield CB, Hesp SA, McLean DL, Harvey ES, Meeuwig JJ (2012) Similarities between line fishing and baited stereo-video estimations of length-frequency: novel application of kernel density estimates. PLoS ONE 7:e45973. https://doi.org/10.1371/journal.pone.0045973 Letessier TB, Juhel J-B, Vigliola L, Meeuwig JJ (2015) Low-cost small action cameras in stereo generates accurate underwater measurements of fish. J Exp Mar Biol Ecol 466:120–126. https://doi.org/10.1016/j.jembe.2015.02.013 López-Macías J, Bravo-Ormaza E, Chinacalle-Martínez N, Miranda C, Murillo-Posada J, Vallejo F, Peñaherrera-Palma C (2023) Comparison of Two Stereo-Video Software for the Assessment of Marine Resources. Thalassas: Int J Mar Sci 39:395–404. https://doi.org/10.1007/s41208-022-00507-4 Marra S, Coppa S, Camedda A, Mazzoldi C, Wrachien F, Massaro G, de Lucia GA (2016) Recovery trends of commercial fish: the case of an underperforming mediterranean marine protected area. PLoS ONE 11:e0146391. https://doi.org/10.1371/journal.pone.0146391 Mehanna SF, Osman AGM, Farrag MMS, Osman YAA (2018) Age and growth of three common species of goatfish exploited by artisanal fishery in Hurghada fishing area, Egypt. J Appl Ichthyol 34:917–921. https://doi.org/10.1111/jai.13590 Mullon C, Fréon P, Cury P (2005) The dynamics of collapse in world fisheries. Fish Fish 6:111–120. https://doi.org/10.1111/j.1467-2979.2005.00181.x Murphy HM, Jenkins GP (2010) Observational methods used in marine spatial monitoring of fishes and associated habitats: a review. Mar Freshw Res 61:236–252. https://doi.org/10.1071/MF09068 Myers RA, Worm B (2003) Rapid worldwide depletion of predatory fish communities. Nature 423:280–283. https://doi.org/10.1038/nature01610 Neumann RM, Allen MS (2007) Analysis and interpretation of freshwater fisheries data Neuswanger JR, Wipfli MS, Rosenberger AE, Hughes NF (2016) Measuring fish and their physical habitats: versatile 2D and 3D video techniques with user-friendly software. Can J Fish Aquat Sci 73:1861–1873. https://doi.org/10.1139/cjfas-2016-0010 Olsen A, Haber A (2022) StereoMorph: stereo camera calibra-tion and reconstruction. Version 1.6. 4 Olsen AM, Westneat MW (2015) StereoMorph: An R package for the collection of 3D landmarks and curves using a stereo camera set-up. Methods Ecol Evol 6:351–356. https://doi.org/10.1111/2041-210X.12326 Ono K, Licandeo R, Muradian ML, Cunningham CJ, Anderson SC, Hurtado-Ferro F, Johnson KF, McGilliard CR, Monnahan CC, Szuwalski CS (2015) The importance of length and age composition data in statistical age-structured models for marine species. ICES J Mar Sci 72:31–43. https://doi.org/10.1093/icesjms/fsu007 Pennino MG, Conesa D, López-Quílez A, Munoz F, Fernández A, Bellido JM (2016) Fishery-dependent and-independent data lead to consistent estimations of essential habitats. ICES J Mar Sci 73:2302–2310. https://doi.org/10.1093/icesjms/fsw062 Perry AL, Low PJ, Ellis JR, Reynolds JD (2005) Climate change and distribution shifts in marine fishes. science 308: 1912–1915 doi https://doi.org/03081912 Pet J, Van Densen W, Machiels M, Sukkel M, Setyohadi D, Tumuljadi A (1997) Length-based analysis of population dynamics and stock identification in the sardine fisheries around East Java, Indonesia. Fish Res 31:107–120. https://doi.org/10.1016/S0165-7836(97)00006-4 Pinsky ML, Selden RL, Kitchel ZJ (2020) Climate-driven shifts in marine species ranges: scaling from organisms to communities. Annual Rev Mar Sci 12:153–179. https://doi.org/10.1146/annurev-marine-010419-010916 Queirós AM, Fernandes J, Genevier L, Lynam CP (2018) Climate change alters fish community size-structure, requiring adaptive policy targets. Fish Fish 19:613–621. https://doi.org/10.1111/faf.12278 R Core Team (2024) R: A language and environment for statistical computing. R Foundation fot Statistical Computing. Vienna, Austria Santana-Garcon J, Braccini M, Langlois TJ, Newman SJ, McAuley RB, Harvey ES (2014) Calibration of pelagic stereo‐BRUV s and scientific longline surveys for sampling sharks. Methods Ecol Evol 5:824–833. https://doi.org/10.1111/2041-210X.12216 Savina E, Krag LA, Madsen N (2018) Developing and testing a computer vision method to quantify 3D movements of bottom-set gillnets on the seabed. ICES J Mar Sci 75:814–824. https://doi.org/10.1093/icesjms/fsx194 Schramm KD, Harvey ES, Goetze JS, Travers MJ, Warnock B, Saunders BJ (2020) A comparison of stereo-BRUV, diver operated and remote stereo-video transects for assessing reef fish assemblages. J Exp Mar Biol Ecol 524:151273. https://doi.org/10.1016/j.jembe.2019.151273 Shafait F, Harvey ES, Shortis MR, Mian A, Ravanbakhsh M, Seager JW, Culverhouse PF, Cline DE, Edgington DR, Browman H (2017) Towards automating underwater measurement of fish length: a comparison of semi-automatic and manual stereo–video measurements. ICES J Mar Sci 74:1690–1701. https://doi.org/10.1093/icesjms/fsx007 Shantz AA, Ladd MC, Burkepile DE (2020) Overfishing and the ecological impacts of extirpating large parrotfish from Caribbean coral reefs. Ecol Monogr 90:e01403. https://doi.org/10.1002/ecm.1403 Shortis MR, Ravanbakskh M, Shaifat F, Beyerer J, Puente León F, Harvey ES, Mian A, Seager JW, Culverhouse PF, Cline DE, Edgington DR (2013) A review of techniques for the identification and measurement of fish in underwater stereo-video image sequences Videometrics. Range Imaging, and Applications XII; and Automated Visual Inspection Smith ML (1988) Effects of observer swimming speed on sample counts of temperate rocky reef fish assemblages. Mar Ecol progress Ser Oldendorf 43:223–231 Strona G, Beck PS, Cabeza M, Fattorini S, Guilhaumon F, Micheli F, Montano S, Ovaskainen O, Planes S, Veech JA (2021) Ecological dependencies make remote reef fish communities most vulnerable to coral loss. Nat Commun 12:7282. https://doi.org/10.1038/s41467-021-27440-z Tortolero-Langarica JA, Rodríguez‐Troncoso AP, Cupul‐Magaña AL, Morales‐de‐Anda DE, Caselle JE, Carricart‐Ganivet JP (2022) Coral calcification and carbonate production in the eastern tropical Pacific: The role of branching and massive corals in the reef maintenance. Geobiology 20:533–545. https://doi.org/10.1111/gbi.12491 Tu C-Y, Chen K-T, Hsieh C-h (2018) Fishing and temperature effects on the size structure of exploited fish stocks. Sci Rep 8:7132 van Overzee HM, Rijnsdorp AD (2015) Effects of fishing during the spawning period: implications for sustainable management. Rev Fish Biol Fish 25:65–83. https://doi.org/10.1007/s11160-014-9370-x Wehkamp M, Fischer P (2014) A practical guide to the use of consumer-level digital still cameras for precise stereogrammetric in situ assessments in aquatic environments. Underw Technol 32:111–128. https://doi.org/10.3723/ut.32.111 Whitfield AK, Gillanders BM, Able KW (2023) Climate change effects on fish populations Climate change and estuaries. CRC, pp 475–506 Whitmarsh SK, Fairweather PG, Huveneers C (2017) What is Big BRUVver up to? Methods and uses of baited underwater video. Rev Fish Biol Fish 27:53–73. https://doi.org/10.1007/s11160-016-9450-1 Williams EH, Shertzer KW, Bacheler N (2022) Estimating length composition of fish observed with stereo-video cameras: A simulation study with application to red snapper (Lutjanus campechanus). Fish Res 254:106424. https://doi.org/10.1016/j.fishres.2022.106424 Williams SB, Pizarro O, Jakuba MV, Mahon I, Ling SD, Johnson CR (2010) Repeated AUV surveying of urchin barrens in North Eastern Tasmania 2010 IEEE International Conference on Robotics and Automation. IEEE, pp 293–299 Woolston C (2023) Facing inflation: lab heads tighten supplies budgets. Nature 613:601–602. https://doi.org/10.1038/d41586-023-00088-z Supplementary Files Supplementarymaterial.docx Cite Share Download PDF Status: Published Journal Publication published 27 Jun, 2025 Read the published version in Marine Biology → Version 1 posted Reviewers agreed at journal 30 Apr, 2025 Reviewers invited by journal 30 Apr, 2025 Editor assigned by journal 25 Apr, 2025 First submitted to journal 24 Apr, 2025 Editorial decision: Accept 22 Apr, 2025 Editorial decision: Editor Decision - Provisional Accept 22 Apr, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5501285","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":450197296,"identity":"2a95792f-f9b7-4a0b-8198-1a6409d73c64","order_by":0,"name":"Alberto García Baciero","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0003-4972-8169","institution":"Instituto Politécnico Nacional Centro Interdisciplinario de Ciencias Marinas: Instituto Politecnico Nacional Centro Interdisciplinario de Ciencias Marinas","correspondingAuthor":true,"prefix":"","firstName":"Alberto","middleName":"García","lastName":"Baciero","suffix":""},{"id":450197297,"identity":"85f19054-c1e9-4ec5-a6b5-ba13d0dd31de","order_by":1,"name":"Carlos Robalino-Mejía","email":"","orcid":"","institution":"Instituto Politécnico Nacional Centro Interdisciplinario de Ciencias Marinas: Instituto Politecnico Nacional Centro Interdisciplinario de Ciencias Marinas","correspondingAuthor":false,"prefix":"","firstName":"Carlos","middleName":"","lastName":"Robalino-Mejía","suffix":""},{"id":450197298,"identity":"b279cdbe-3d73-4c56-8277-32e30effd85c","order_by":2,"name":"César Peñaherrera-Palma","email":"","orcid":"","institution":"Migramar","correspondingAuthor":false,"prefix":"","firstName":"César","middleName":"","lastName":"Peñaherrera-Palma","suffix":""},{"id":450197299,"identity":"58d3b897-27d6-4b49-96d5-52be612d3188","order_by":3,"name":"Héctor Villalobos","email":"","orcid":"","institution":"Instituto Politécnico Nacional Centro Interdisciplinario de Ciencias Marinas: Instituto Politecnico Nacional Centro Interdisciplinario de Ciencias Marinas","correspondingAuthor":false,"prefix":"","firstName":"Héctor","middleName":"","lastName":"Villalobos","suffix":""}],"badges":[],"createdAt":"2024-11-22 03:25:39","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5501285/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5501285/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00227-025-04682-9","type":"published","date":"2025-06-27T16:05:43+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81841409,"identity":"9ced58de-933f-42cb-b808-850ff1301f56","added_by":"auto","created_at":"2025-05-02 16:16:32","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":333392,"visible":true,"origin":"","legend":"\u003cp\u003eObjects measured in the pool: a) SeaGis-manufactured object; b) PVC object.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-5501285/v1/48d459d859cc265ee1ad4167.png"},{"id":81842014,"identity":"c934a486-4374-410e-ad2a-b542c5ca4ec4","added_by":"auto","created_at":"2025-05-02 16:24:32","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":164246,"visible":true,"origin":"","legend":"\u003cp\u003eRotation movements of the known-length objects with respect to the stereo-DOV system. a) 0: the object is not rotated; b) Z1: the object is rotated 45° around the Z-axis; c) Z2: the object is rotated 315° around the Z-axis; d) Y1: the object is rotated 45° around the Y-axis; e) Y2: the object is rotated 315° around the Y-axis.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-5501285/v1/9d542825226d5c3e302f3029.png"},{"id":81842336,"identity":"84780700-9382-4b52-880f-71bb8664b565","added_by":"auto","created_at":"2025-05-02 16:32:32","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":496382,"visible":true,"origin":"","legend":"\u003cp\u003eScatter plots of the relative error (RE, %) variation in relation to the object’s true length (mm) and grouped by rotation movement and distance (m) for each software. Lines represent a fitted linear regression model for each scatterplot. Colored bands indicate the confidence region. Notations: 0 (no rotation), Z1 and Z2 (rotation around Z-axis 45° and 315°, respectively) and Y1 and Y2 (rotation around Y-axis 45° and 315°, respectively).\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-5501285/v1/237e1fc5a35961f2430a7637.png"},{"id":81842016,"identity":"02d79d0f-ea65-4bb8-a836-421326477396","added_by":"auto","created_at":"2025-05-02 16:24:32","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":64822,"visible":true,"origin":"","legend":"\u003cp\u003eScatter plots of the relative error (RE, %) variation with object’s true length (mm) and grouped by rotation movement for each software in ocean conditions. Lines represent a fitted linear regression model for each scatterplot. Colored bands indicate the confidence region. Notations: 0 (no rotation), Z1 and Z2 (rotation around Z-axis 45° and 315°, respectively) and Y1 and Y2 (rotation around Y-axis 45° and 315°, respectively). All measurements were carried out at 3 m distance.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-5501285/v1/c3b85822735dc4e1bf4646f7.png"},{"id":81842017,"identity":"c6dd7742-4fec-4543-9306-5138fc24fc05","added_by":"auto","created_at":"2025-05-02 16:24:32","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":22954,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of paired measurements of fish lengths with EventMeasure and StereoMorph. The diagonal is the 45º line denoting no differences between measuring tools.\u003c/p\u003e","description":"","filename":"Picture5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5501285/v1/adb934e66f96fc0c8d756585.jpg"},{"id":85686495,"identity":"89b60730-bf74-423a-8413-90fc5659b673","added_by":"auto","created_at":"2025-06-30 16:07:59","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1872084,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5501285/v1/e1fc6832-d077-4fe6-bdeb-2cad22093909.pdf"},{"id":81842889,"identity":"7a74ceee-8e5b-48ce-8e4d-d099eca0cdf7","added_by":"auto","created_at":"2025-05-02 16:40:32","extension":"docx","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":22457,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-5501285/v1/f67e8038902157b00a8d7cfa.docx"}],"financialInterests":"","formattedTitle":"Comparing the performance of two scientific tools for obtaining fish length measurements","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eReef-associated fish species are experiencing remarkable changes around the globe due to habitat degradation, pollution, and overexploitation (Brandl et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Strona et al. \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Quantitative assessments are then required to evaluate and monitor their populations to understand their trends and status (Heenan et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Tortolero-Langarica et al. \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), including both fishery-dependent and fishery-independent approaches (Pennino et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Howard et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). While the former has gear-specific sampling biases since they may target specific taxa, size classes, and trophic levels (King \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), fishery-independent methods can prove to be a valuable and cost-effective information source for assessing fish populations in coastal coral reef habitats (Brock \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1954\u003c/span\u003e; Murphy and Jenkins \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Non-destructive and non-extractive techniques, such as underwater visual censuses (UVC) (Brock \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1954\u003c/span\u003e), and video recording-based methods, are crucial in Marine Protected Areas (MPAs) for preserving ecosystem integrity and reducing pressure on fish populations (Murphy and Jenkins \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Goetze et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Garc\u0026iacute;a-Baciero et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Although UVC methods using strip-transects are probably the most frequently used tool to collect data on fish assemblages, they show biases for reliable long‐term monitoring and estimating the length of fish and sample areas (Smith \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Harvey et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Moreover, UVC methods require divers to be trained in taxonomic identification and sampling techniques and are subject to natural inter-observer variability. Additionally, UVC methods do not allow the collected data to be validated by a second observer, as it is possible with video recordings (Goetze et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn this context, researchers have begun implementing other monitoring techniques, such as video-based techniques, to overcome the different sampling biases and enhance data accuracy and precision (Harvey and Shortis \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Harvey et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Cappo et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Harvey et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Whitmarsh et al. \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Underwater video is primarily used in ecological studies to monitor fish populations (Caldwell et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Marra et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) examine fish behavior (Ajemian et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), compare methodologies (Schramm et al. \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Jessop et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and study length-frequency distributions (Arachchige-Weerarathne et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Williams et al. \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Garc\u0026iacute;a-Baciero et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The use of stereo-video systems for monitoring fish assemblages has grown in popularity in the last two decades as it allows for more accurate measurement of fish lengths in field conditions (Harvey et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Goetze et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Schramm et al. \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Accurate size structure descriptions are crucial for the adaptive management of marine fish populations facing anthropogenic and environmental pressures (Barnett et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Queir\u0026oacute;s et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Knowing the length of sampled fish is vital to many modern stock assessments and population demographic methods (Ono et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Effective monitoring programs must measure the lengths of a sufficient number of individuals within each fish population (Arachchige-Weerarathne et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Size structure data offers valuable insights into the reproductive potential, growth, and stability of reef fish populations (Hixon et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; van Overzee and Rijnsdorp \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). A scarcity of smaller fish may indicate recruitment issues, while a low occurrence of larger fish could suggest high mortality rates among mature individuals (Neumann and Allen \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Increased fishing pressure and warming temperatures drive higher mortality, altering fish stock size structure (Tu et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) reducing large predator abundance (Angelini et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAccuracy in length measurements require a proper calibration of the stereo-video system. During the calibration, the two cameras film an object of known dimensions to account for lens distortion and their location in space relative to each other (Harvey and Shortis \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e1998\u003c/span\u003e). There are two calibration approaches used in underwater ecological applications with different procedures for calculating camera parameters. One method involves using a 3D calibration object marked with several points of known positions, and it requires a photogrammetric network solution to calculate the parameters (Clarke and Fryer \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e1998\u003c/span\u003e). The other method uses a two-dimensional (2D) chequerboard pattern consisting of known-size squares to generate the camera parameters. In the former method, the solution is determined using collinearity and least squares estimation (LSE) to accurately model the network geometry and the random errors in the measurements. In contrast, in the latter method, a four-step algorithm based on the Direct Linear Transformation (DLT) method (Abdel-Aziz \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1971\u003c/span\u003e; Heikkila and Silv\u0026eacute;n \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e1997\u003c/span\u003e) is used to provide the initial estimates of the parameters, followed by a network solution based on collinearity. The DLT method assigns each calibrated camera a set of coefficients that help to establish a relationship between unique 3D coordinates in calibration space and their corresponding (non-unique) 2D-pixel coordinates in that specific camera view. Although both network and DLT solutions differ in the camera calibration models, they produce distinct yet transformable mathematical descriptions of the relative orientation. The 3D calibration generally yields more accurate measurements across a broader range of distances compared to 2D calibration (Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), with average in-water errors typically below 1% versus 1.5\u0026ndash;10% for 2D (Williams et al. \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). However, 2D calibration can achieve average errors below 5% for stereo-video measurements up to 8 m (Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and \u0026lt;\u0026thinsp;3% when compared to tape-measured lengths in the field (Delacy et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Besides accuracy, there is also a variation in price between the two calibration approaches mentioned. The cost for a single license and 2D hardware and software calibration object ranges from \u003cspan\u003e$\u003c/span\u003e500 to \u003cspan\u003e$\u003c/span\u003e2150 (source: \u003cspan class=\"ExternalRef\"\u003e \u003cspan class=\"RefSource\"\u003e \u003ca href=\"https://orcid.org/0000-0003-4972-8169\" target=\"_blank\"\u003ewww.mathworks.com\u003c/a\u003e \u003c/span\u003e \u003cspan address=\"http://www.mathworks.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e \u003c/span\u003e, accessed on April 22, 2025). On the other hand, the 3D CAL and EventMeasure software, along with the 3D cube calibration hardware, cost nearly \u003cspan\u003e$\u003c/span\u003e7,600 for academic/research use (source: \u003cspan class=\"ExternalRef\"\u003e \u003cspan class=\"RefSource\"\u003e \u003ca href=\"https://orcid.org/0000-0003-4972-8169\" target=\"_blank\"\u003ewww.seagis.com.au\u003c/a\u003e \u003c/span\u003e \u003cspan address=\"http://www.seagis.com.au\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e \u003c/span\u003e, accessed on April 22, 2025). Thus, 2D calibration techniques may be more cost-effective, easier to set up and execute, and practically deployable in the field compared to constructing a complex cage that is unwieldy to maneuver underwater.\u003c/p\u003e \u003cp\u003eAlthough stereo-video systems have multiple advantages for surveying fish, economic resources for monitoring marine systems are limited. Many institutions or organizations cannot afford the purchase of commercial software licenses in addition to conducting research. A few free and open-source applications for length estimation have emerged recently, such as the R package StereoMorph (Olsen and Westneat \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Olsen and Haber \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and VidSync (Neuswanger et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) that may benefit low-budget institutions and NGOs conducting fish monitoring programs. StereoMorph has features for processing images, calibrating cameras, digitizing images, and reconstructing landmarks and curves. This enables users to transform any digitized landmarks from both camera perspectives into a 3D reconstruction (Olsen and Westneat \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). VidSync is a free Mac application for scientific video analysis that allows users to efficiently record, organize, and navigate complex 2D or 3D measurements of fish and their physical habitats (Neuswanger et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). VidSync is a valuable free alternative to EventMeasure and StereoMorph, especially for measuring fish communities. However, it requires a Mac computer, which may be a barrier for institutions with limited budgets.\u003c/p\u003e \u003cp\u003eThe StereoMorph R package has been recently used to estimate marine species\u0026rsquo; lengths (Garc\u0026iacute;a-Baciero et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), showing accurate results compared to true-length data (Delacy et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Garner et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Nevertheless, their differences in obtaining accurate and precise measurements compared to commercial software have not been assessed yet. Here, we compare the performance of the StereoMorph R package and the commercial software EventMeasure for estimating the length of known-size objects in controlled conditions, underwater, and of fish individuals in the field.\u003c/p\u003e"},{"header":"2. Material and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. The stereo-video system\u003c/h2\u003e \u003cp\u003eWe used a diver-operated stereo-video system (stereo-DOV) composed of a pair of GoPro Hero7 with underwater camera housing. The cameras were mounted with a convergence angle of 4\u0026deg;, separated by 800 mm, on a steel base bar provided with handles. This stereo configuration allows for accurate fish measurements up to 8 m distance (Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Letessier et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). The cameras were set up to 60 frames per second, 1080p resolution (1920\u0026times;1080 pixels), and linear field of view (FOV).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Calibration procedures\u003c/h2\u003e \u003cp\u003eBefore sampling, we calibrated the stereo-DOV using a 3D cube from SeaGIS (1000\u0026times;1000\u0026times;500 mm), and with an 8\u0026times;6 square checkerboard (total dimensions 472\u0026times;314 mm, with each square 52.45\u0026times;52.45 mm) created using the function \u003cem\u003edrawCheckerboard\u003c/em\u003e from StereoMorph. The laser-printed checkerboard was placed between two acrylic sheets (2 mm thick) with their borders sealed with silicon, which do not affect the accuracy of length estimations.\u003c/p\u003e \u003cp\u003eWe performed the calibration in a clear pool immobilizing the camera system underwater while we manually moved the cube and checkerboard at different angles and positions along the field of view of both cameras for three to five minutes each. To synchronize the videos, we clamped in front of the cameras before and after moving the cube and the checkerboard. In the 3D calibration method, the cube was placed at an approximate fixed distance of 2 m and then rotated in 20 different positions according to the standard calibration protocol described by the software CAL 3.3 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.seagis.com.au/bundle.html\u003c/span\u003e\u003cspan address=\"https://www.seagis.com.au/bundle.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). This software uses the locations of the targets on the cube, as measured in the video images, to calculate the internal characteristics of the two cameras (e.g., the principal distance and lens distortion) and the 3D orientations of the cameras relative to one another. In the 2D calibration method, we moved the checkerboard along the entire calibration volume, following procedures described in Olsen and Westneat (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), and subsequently used the function \u003cem\u003ecalibrateCameras\u003c/em\u003e from StereoMorph. This function provides a completely automated workflow for stereo-camera calibration, performing 3 main steps: checkerboard corner detection, undistortion coefficient estimation, and calibration coefficient estimation. StereoMorph determines the DLT coefficients using the internal corners automatically detected from the checkerboard pattern, which is photographed from both camera views at different positions and angles within the calibration space (Olsen and Westneat \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Rather than estimate the DLT coefficients directly, StereoMorph estimates the six transformation parameters (3 translations, 3 rotations) required to transform the first checkerboard into each subsequent checkerboard in 3D space by minimizing the reconstruction error. These transformation parameters are then used to generate 3D coordinates (the equivalent of a calibration object) and calculate the DLT coefficients (Olsen and Westneat \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Data collection\u003c/h2\u003e \u003cp\u003eWe first collected measurements from known-length objects in pool and ocean environments. Then, we collected video recordings in a shallow reef to obtain fish measurements in real oceanic conditions. Measurements of known-length objects were collected first in controlled conditions in a swimming pool and then in a shallow rocky reef in the southern Gulf of California, located in front of \u0026ldquo;El Saltito\u0026rdquo; beach (24\u0026deg;15\u0026prime;06.16\u0026rdquo; N, 110\u0026deg;09\u0026prime;18.17\u0026rdquo; W) near La Paz, Baja California Sur, Mexico. In the pool environment, we obtained measurements from two known-length objects to increase the range of measurements. The first object (manufactured by SeaGIS; Fig.\u0026nbsp;1a) was a 1 m long scale bar marked with circular, reflective targets, allowing measurements at three different lengths: 345 (short), 575 (medium), and 915 mm (long). The second object (an 80 cm long PVC tube, Fig.\u0026nbsp;1b) had six marks separated by 100 mm each, allowing the collection of measurements at 100, 200, 300, 400, 500, and 600 mm. The recordings were collected in the morning at 9:00, right after the pool was cleaned, to ensure optimal water transparency.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe repeated the filming procedure of the two objects by rotating them around their Y- and Z-axis at a given degree (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e2\u003c/span\u003e), and at distances separated by 1 m until reaching the end of the pool at 6 m. Thus, we obtained measurements at distances of 2, 3, 4, and 5 m from the cameras.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe ocean environment consisted of patchy rocky areas of various sizes in the surroundings of two small islets. To ensure the best visibility possible, we collected the data at midday between 12:00 to 13:00 using the stereo-DOV system right after calibration. We filmed the PVC object for 5 minutes at approximately 3 m and using the same angles as in the pool environment. We estimated measurements only from the PVC object to not compromise the SeaGIS-manufactured object. Finally, reef fish were measured over the same rocky reef swimming at constant low speed for 20 minutes at a depth of 8 m while pointing the cameras straight ahead along the reef to film as many fish individuals as possible. Fish species were identified on-site and corroborated using available fish identification guides for the area.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Length measurements\u003c/h2\u003e \u003cp\u003eWe analyzed the video footage obtained with the stereo-DOV by using EventMeasure software (SEAGIS Pty Ltd) and the StereoMorph package in the R environment to estimate the length of the two objects and fish individuals. Although StereoMorph does not have a tool to measure 3D lengths, an estimation can be done by selecting landmarks to define the two endpoints that will produce the measure of interest (Olsen and Westneat \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). We reconstructed the selected landmarks into 3D using the function \u003cem\u003ereconstructStereoSets\u003c/em\u003e and we then calculated the lengths by measuring the distance between the resulting 3D marker positions.\u003c/p\u003e \u003cp\u003eIn the pool environment, we obtained 300 measurements of the SeaGIS object with each software tool by taking five replicates of each true length with each rotation movement and distance combination. Likewise, we obtained 600 measurements per software of the PVC object, consisting of 5 measurements of each true length for the same rotation movement and distance combinations. In the ocean environment, we obtained 150 measurements per software of the PVC object, consisting of 5 measurements of each true length for the five rotation movements at a fixed distance of 3 m. Moreover, we measured the standard length (SL) of 47 fish individuals belonging to three species: \u003cem\u003eThalassoma lucasanum\u003c/em\u003e (1 individual), \u003cem\u003eAbudefduf troschelii\u003c/em\u003e (6 individuals), and \u003cem\u003eMulloidichthys dentatus\u003c/em\u003e (40 individuals). We used three different frames of the same individual as replicates, accounting for 141 measurements obtained with each software. Frames in the pool and ocean environments were the same for both software to improve accuracy comparisons. Fish standard lengths were used rather than total length as it is more easily definable across a range of species.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Data analysis\u003c/h2\u003e \u003cp\u003eWe assessed the performance of both software by comparing the accuracy and precision of the estimated measures. Accuracy is defined as the closeness of an estimated measurement to the real or true values, in this case of the known-lengths objects assessed, whereas precision refers to the consistency of repeated measurements of the same object (Harvey and Shortis \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). The less the differences among repeated measures the larger the measurement precision.\u003c/p\u003e \u003cp\u003eWe evaluated the software\u0026rsquo;s accuracy in the pool and the ocean environment by comparing differences in the Relative Error (RE), a metric estimated from the ratio of the absolute error (true length minus measured length) to the true length, as shown in the following formula:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:RE=\\:\\frac{|measured\\:length-true\\:length|}{true\\:length}\\times\\:100$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWe then calculated the mean RE (MRE) by averaging the RE obtained per rotation movement and distance, respectively. For illustration purposes, RE vs. true length were plotted by software type for each rotation movement and distance to the system, including linear trends for guidance. A multiple linear regression model was then fitted to explore the relationship between the RE (response variable) with the rotation movement (0, Z1, Z2, Y1 and Y2), distance (2, 3, 4, and 5 m), software (EventMeasure, StereoMorph) and true length (100, 200, 300, 345, 400, 500, 575, 600 and 915 mm) as explanatory variables for the pool environment, and rotation movement, software, and true length (100, 200, 300, 400, 500, and 600 mm) for the ocean environment.\u003c/p\u003e \u003cp\u003eWe evaluated the precision of each software in the pool environment by calculating the coefficient of variation (CV, %), which is the ratio of the standard deviation to the mean. The CV also measures the dispersion of the data regardless of the measurement units and magnitude of the values and is widely used for assessing measure precision in field research (Harvey et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). In the ocean environment, we first calculated the CV of the measurements from the PVC object at different angles at a fixed distance (3 m) using five replicates with each parameter. Then, we compared the precision of each software to estimate fish lengths by comparing the CV of the mean SL obtained from three replicates. Additionally, we used a Spearman correlation analysis to evaluate the relationship of paired fish length estimations between software. We performed all statistical analyses and graphs in the R environment and associated packages (R Core Team \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003ch2\u003e3.1. Known-length object measurements\u0026nbsp;\u003c/h2\u003e\n\u003ch3\u003e3.1.1. \u0026nbsp; Pool environment\u003c/h3\u003e\n\u003cp\u003eMultiple linear regression model showed significant effect of software output with StereoMorph having lower measurements accuracy (Table 1). While the RE increased with distance for both systems and differed according to the rotation of the object (Estimate = 0.50, p \u0026lt; 2.16\u0026times;10\u003csup\u003e-5\u003c/sup\u003e), the RE was much more noticeable in StereoMorph on objects with rotation movement of 315\u0026ordm; around the Y-axis (Estimate = 5.54, p \u0026lt; 2.2\u0026times;10\u003csup\u003e-16\u003c/sup\u003e). The accuracy of the measurements was also affected significantly by true length (Estimate = 0, p \u0026lt; 2.16\u0026times;10\u003csup\u003e-5\u003c/sup\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e Multiple linear regression model ranking results of relative errors (RE) for each measurement versus the type of software (categorical), distance to the system (continuous), the true length of the object (continuous), and rotation movement (categorical). Notations: Z1 and Z2 (rotation around Z-axis 45\u0026deg; and 315\u0026deg;, respectively) and Y1 and Y2 (rotation around Y-axis 45\u0026deg; and 315\u0026deg;, respectively).\u003c/p\u003e\n\u003cdiv\u003e\n \u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"421\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29.6912%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 70.0713%;\"\u003e\n \u003cp\u003eRE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 0.23753%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29.6912%;\"\u003e\n \u003cp\u003ePredictors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23.9905%;\"\u003e\n \u003cp\u003eEstimates\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25.6532%;\"\u003e\n \u003cp\u003eCI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003ep\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003e(Intercept)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23.9905%;\"\u003e\n \u003cp\u003e-2.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25.6532%;\"\u003e\n \u003cp\u003e-2.55 \u0026ndash; -1.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eStereoMorph\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23.9905%;\"\u003e\n \u003cp\u003e4.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25.6532%;\"\u003e\n \u003cp\u003e3.92 \u0026ndash; 4.36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eDistance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23.9905%;\"\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25.6532%;\"\u003e\n \u003cp\u003e0.40 \u0026ndash; 0.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eZ1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23.9905%;\"\u003e\n \u003cp\u003e0.30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25.6532%;\"\u003e\n \u003cp\u003e-0.05 \u0026ndash; 0.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003e0.093\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eZ2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23.9905%;\"\u003e\n \u003cp\u003e0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25.6532%;\"\u003e\n \u003cp\u003e-0.12 \u0026ndash; 0.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003e0.191\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eY1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23.9905%;\"\u003e\n \u003cp\u003e1.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25.6532%;\"\u003e\n \u003cp\u003e1.51 \u0026ndash; 2.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eY2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23.9905%;\"\u003e\n \u003cp\u003e5.54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25.6532%;\"\u003e\n \u003cp\u003e5.19 \u0026ndash; 5.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eTrue length\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23.9905%;\"\u003e\n \u003cp\u003e-0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25.6532%;\"\u003e\n \u003cp\u003e-0.00 \u0026ndash; -0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 20.6651%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eObservations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"4\" valign=\"top\" style=\"width: 70.3088%;\"\u003e\n \u003cp\u003e1800\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.6912%;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e adjusted\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"4\" valign=\"top\" style=\"width: 70.3088%;\"\u003e\n \u003cp\u003e0.609\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eEventMeasure was more accurate when estimating measures of known-length objects compared to StereoMorph (MRE\u003csub\u003eEM\u003c/sub\u003e = 0.53%; MRE\u003csub\u003eSM\u003c/sub\u003e = 4.54%). MRE increased by distance and differed by the object\u0026rsquo;s rotation movement. In contrast, both parameters decreased with increasing true lengths in both measuring tools, although this effect was more accentuated in StereoMorph (Supplementary Table 1). EventMeasure exhibited a MRE \u0026lt;1% at every distance and close to 1% with all rotation movements and for each true length, while StereoMorph showed MRE values under 5% only when the object was not moved or was rotated around the Z-axis. Error estimations for each software were almost negligible at closer distances (\u0026lt; 3 m) and when objects were not rotated or were only rotated over their z-axis (Fig. 3). RE decreased for larger measurements in both tools, except when measuring with StereoMorph at 3 and 4 m distances and when the object was rotated 45\u0026ordm; around the Y-axis (Y1 in Fig. 3).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSimilarly, to the accuracy metrics, EventMeasure measurements (CV = 0.35%) were more precise than StereoMorph (CV = 0.65%). CV was below 1% for measurements obtained with each software regardless of the distance, rotation movement, or true length, except when the object\u0026rsquo;s length was 100 mm long (CV\u003csub\u003eEM\u003c/sub\u003e = 1.16%; CV\u003csub\u003eSM\u003c/sub\u003e = 1.34%; Table 2) and when using StereoMorph with a rotation of 315\u0026deg; around the Y axis (CV = 1.27%). Precision decreased when the objects were rotated around the Y-axis with both measuring software as well as with increasing distance but increased with object size (Table 2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2.\u0026nbsp;\u003c/strong\u003eMean standard length (SL, mm) with standard error (SE) and mean coefficient of variation (CV, %) of estimated measurements by software at different rotation movements, distances (m), and true lengths (mm). Notations: 0 (no rotation), Z1 and Z2 (rotation 45\u0026deg; and 315\u0026deg;, respectively, around Z-axis) and Y1 and Y2 (rotation 45\u0026deg; and 315\u0026deg;, respectively, around Y-axis).\u003c/p\u003e\n\u003cdiv align=\"center\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 33px;\"\u003e\n \u003cp\u003eEventMeasure\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 33px;\"\u003e\n \u003cp\u003eStereoMorph\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003eRotation movement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003eMean SL \u0026plusmn; SE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003eCV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003eMean SL \u0026plusmn; SE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003eCV\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e437.97 \u0026plusmn; 0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e431.38 \u0026plusmn; 0.36\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003eZ1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e437.89 \u0026plusmn; 0.25\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e428.55 \u0026plusmn; 0.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.67\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003eZ2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e438.06 \u0026plusmn; 0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e429.49 \u0026plusmn; 0.49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.30\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003eY1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e437.67 \u0026plusmn; 0.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e418.85 \u0026plusmn; 1.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.77\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003eY2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e438.32 \u0026plusmn; 0.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e389.69 \u0026plusmn; 1.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e1.27\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003eDistance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e436.87 \u0026plusmn; 0.20\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e429.27 \u0026plusmn; 0.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.42\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e438.41 \u0026plusmn; 0.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e423.11 \u0026plusmn; 0.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.51\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e438.06 \u0026plusmn; 0.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e415.04 \u0026plusmn; 1.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.80\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e438.59 \u0026plusmn; 0.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.61\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e410.95 \u0026plusmn; 1.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.86\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003eTrue length\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e100\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e100.32 \u0026plusmn; 0.52\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e1.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e95.49 \u0026plusmn; 0.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e1.34\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e200\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e201.84 \u0026plusmn; 0.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e192.23 \u0026plusmn; 0.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.89\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e300.72 \u0026plusmn; 0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e286.42 \u0026plusmn; 0.93\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.75\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e400\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e400.32 \u0026plusmn; 0.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e381.18 \u0026plusmn; 1.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.61\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e500\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e500.79 \u0026plusmn; 0.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e476.67 \u0026plusmn; 1.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.60\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e600\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e601.08 \u0026plusmn; 0.54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e573.01 \u0026plusmn; 1.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.53\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e345\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e345.86 \u0026plusmn; 0.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e332.13 \u0026plusmn; 0.67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.46\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e575\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e572.44 \u0026plusmn; 0.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e554.73 \u0026plusmn; 0.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.34\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 32px;\"\u003e\n \u003cp\u003e915\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e918.47 \u0026plusmn; 0.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24px;\"\u003e\n \u003cp\u003e884.47 \u0026plusmn; 1.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8px;\"\u003e\n \u003cp\u003e0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003ch3\u003e3.1.2. \u0026nbsp; Ocean environment\u003c/h3\u003e\n\u003cp\u003eThe results from the regression analysis showed that StereoMorph affected significantly and negatively the accuracy of the measurements (Estimate = 4.71, p = 2\u0026times;10\u003csup\u003e-16\u003c/sup\u003e). The accuracy of the measurements decreased significantly when the object was rotated around the Y-axis, but it was not affected by true length (Table 3).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 3.\u003c/strong\u003e Multiple linear regression model ranking results of relative errors (RE) for each measurement versus type of software (categorical), true length (continuous), and rotation movement (categorical). Z1 and Z2 refer to rotating the object 45\u0026deg; and 315\u0026deg; around the Z-axis, respectively, and Y1 and Y2 to 45\u0026deg; and 315\u0026deg; around the Y-axis, respectively.\u003c/p\u003e\n\u003cdiv\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"441\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 256px;\"\u003e\n \u003cp\u003eRE\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003ePredictors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003eEstimates\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 105px;\"\u003e\n \u003cp\u003eCI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003ep\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003e(Intercept)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e-1.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 105px;\"\u003e\n \u003cp\u003e-2.34 \u0026ndash; 0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.069\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003eStereoMorph\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e4.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 105px;\"\u003e\n \u003cp\u003e3.95 \u0026ndash; 5.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003eZ1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e-0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 105px;\"\u003e\n \u003cp\u003e-1.36 \u0026ndash; 1.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.788\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003eZ2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e-0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 105px;\"\u003e\n \u003cp\u003e-1.20 \u0026ndash; 1.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.994\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003eY1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e3.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 105px;\"\u003e\n \u003cp\u003e2.09 \u0026ndash; 4.49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003eY2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e8.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 105px;\"\u003e\n \u003cp\u003e7.54 \u0026ndash; 9.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003eTrue length\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e-0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 105px;\"\u003e\n \u003cp\u003e-0.00 \u0026ndash; 0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e0.320\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003eObservations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 256px;\"\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 185px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e adjusted\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 256px;\"\u003e\n \u003cp\u003e0.607\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eIn the ocean environment, EventMeasure exhibited greater accuracy than StereoMorph (MRE\u003csub\u003eEM\u003c/sub\u003e = 0.85% \u003cem\u003evs\u003c/em\u003e.\u0026nbsp;MRE\u003csub\u003eSM\u003c/sub\u003e = 5.56%), both when comparing by object\u0026rsquo;s true length and by rotation movement (Supplementary Table 2). However, MRE grouped by true length showed values ranging between 0.21 and 1.34% for EventMeasure, and between 0.36-1.81% for StereoMorph for rotations around the Z-axis (Supplementary Table 2). There was also a noticeable increase in the MRE when measuring objects with StereoMorph, especially when they were rotated around the Y-axis (Fig. 4). EventMeasure and StereoMorph exhibited MRE under 1% when the objects were not moved nor rotated, excepting when rotation was over the Y-axis (Fig. 4).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe CV decreased as true length increased, especially in EventMeasure (Table 4). We obtained the smallest CV when true lengths were higher than 400 mm, whereas the biggest CV corresponded to the shortest true lengths for both measuring tools. The mean CV was significantly higher with rotations around the Y-axis, especially for StereoMorph (CV\u003csub\u003eangle Y1, Y2\u003c/sub\u003e = 1.49%) compared to EventMeasure (CV\u003csub\u003eangle Y1\u003c/sub\u003e =0.89% and CV\u003csub\u003eangle Y2\u003c/sub\u003e = 1.29%).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 4\u003c/strong\u003e. Mean coefficient of variation (CV, %) for each software with different rotation movements and true lengths (mm) of the PVC object in the ocean environment. Notations: 0 (no rotation), Z1 and Z2 (rotation around Z-axis 45\u0026deg; and 315\u0026deg;, respectively) and Y1 and Y2 (rotation around Y-axis 45\u0026deg; and 315\u0026deg;, respectively)\u003c/p\u003e\n\u003cdiv\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eEventMeasure\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eStereoMorph\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eRotation movement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMean SL \u0026plusmn; SE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCV (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMean SL \u0026plusmn; SE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCV (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e352.0 \u0026plusmn; 0.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e347.09 \u0026plusmn; 0.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eZ1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e351.36 \u0026plusmn; 0.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e350.19 \u0026plusmn; 0.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.40\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eZ2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e352.22 \u0026plusmn; 0.30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e346.22 \u0026plusmn; 0.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.44\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eY1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e351.55 \u0026plusmn; 1.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e321.93 \u0026plusmn; 1.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.49\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eY2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e352.15 \u0026plusmn; 1.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e289.07 \u0026plusmn; 1.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.49\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTrue Length\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e100\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e101.10 \u0026plusmn; 0.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e94.32 \u0026plusmn; 0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.70\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e200\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e202.66 \u0026plusmn; 0.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e190.72 \u0026plusmn; 0.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.96\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e301.66 \u0026plusmn; 0.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e283.17 \u0026plusmn; 0.85\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e400\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e401.17 \u0026plusmn; 0.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e377.10 \u0026plusmn; 1.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.67\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e500\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e502.00 \u0026plusmn; 0.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e472.41 \u0026plusmn; 1.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.53\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e600\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e602.54 \u0026plusmn; 0.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e567.69 \u0026plusmn; 0.88\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.37\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003ch2\u003e3.2. Reef fish measurements \u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eWe estimated the standard length (SL) of 47 fish individuals, obtaining a total of 153 measurements with each software (Supplementary Table 3). We obtained very similar length estimates when using EventMeasure (Mean SL = 223.2 mm, range = 143.2-281.4 mm) and StereoMorph (Mean SL = 227.2 mm, range = 141.1-302.9 mm). Precision for EventMeasure (Mean CV = 15.0%) was similar to that of StereoMorph (Mean CV = 15.8%). We found a strong monotonic correlation between paired mean length estimates by EventMeasure and StereoMorph (Spearman\u0026rsquo;s \u0026rho;= 0.93, p \u0026lt; 0.001; Fig. 5). StereoMorph showed a slightly better precision in estimating measurements for paired individuals (Mean CV = 2.07%, range = 0.20 - 9.25%), than EventMeasure (Mean = 2.11%, range = 0.19 - 6.42%) (Supplementary Table 3).\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThe accuracy and precision of measurements obtained with the open-access tool StereoMorph and with the commercial software EventMeasure were particularly similar when objects or fish were close to the plane axis. StereoMorph also performed well when measuring objects ranging in size from 100 to 915 mm, at distances up to 4 m. Nevertheless, EventMeasure provided more accurate and precise results than StereoMorph when measuring objects rotated over their y-axis. Generally, errors increased with increasing distances and decreased with increasing object size, yet these effects were further accentuated in StereoMorph. However, with errors below 2.5% when measuring objects at 3 m in the ocean, StereoMorph may improve precision compared to traditional visual size estimation, as it requires length-measurement training to obtain accurate in situ measurements. As the cost of living and research surge worldwide (Woolston \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), cost-effective tools are needed to implement or maintain long-term monitoring programs dedicated to assessing the biodiversity and biomass of marine ecosystems. These results provided scientific evidence to support the use of this open-access tool for estimating fish lengths, particularly under some specific circumstances discussed below.\u003c/p\u003e \u003cp\u003eCalibration is a critical step when estimating fish lengths using stereo-video systems. Average errors for in-water measurements are usually\u0026thinsp;\u0026lt;\u0026thinsp;1% for 3D calibrations while for 2D calibrations, errors have ranged from 1.5\u0026ndash;10% (Boynton and Voss \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Williams et al. \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). 2D and 3D calibration techniques might also show different outcomes due to the approaches they used to calculate the camera parameters, as there is a greater variability introduced through the calibration process in the 2D system compared to the 3D approach (Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). For example, Boutros et al. (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) reported that stereo-video measurements obtained after using a 2D calibration approach produced measurement errors below 5% on average to distances up to 8 m from the cameras, while Delacy et al. (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) reported errors below 3% in comparison with tape-measured lengths of sharks in the field. Our results using StereoMorph produced a MRE of 4.54%, similar to that reported by Letessier et al. (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), who measured ten fixed target lengths (50 to 900 mm) at six distances (1 to 7 m) in a pool environment and obtained errors of \u0026lt;\u0026thinsp;5% at angles less than 10\u0026deg; to the optical axis and a distance closer to 3 m. Those results are larger than what was obtained by using the 3D calibrated software EventMeasure in our study (MRE\u0026thinsp;=\u0026thinsp;0.53%), other studies using EventMeasure (Harvey et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), and when using other 3D calibrated software such as VidSync (Neuswanger et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; L\u0026oacute;pez-Mac\u0026iacute;as et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Such differences in error estimations have been previously reported by Boutros et al. (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and Neuswanger et al. (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), and are expected to occur even in measurements done with the same equipment but with different software and measuring approaches (Shafait et al. \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; L\u0026oacute;pez-Mac\u0026iacute;as et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the case of the 2D calibration approach, differences in MRE can arise from choosing the right checkerboard square size and the number of corners. Successful calibrations depend upon the size of desired target observations and water conditions (Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Olsen and Westneat \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). We used a checkerboard of 472 \u0026times; 314 mm and obtained MREs between 0.7 and 3.5% for objects measured perpendicular to the optical axis at distances from 2 to 5 m in pool conditions. Our results improved on some previous 2D checkerboard-based calibrations, such as those performed by Olsen and Westneat (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), who achieved an MRE of 11.6% using a smaller calibration checkerboard (270 \u0026times; 210 mm). Contrarily, these authors obtained smaller errors (1.7%) when calibrating using a checkerboard of 360 \u0026times; 280 mm in size. Similarly to us, Wehkamp and Fischer (\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) reported MREs of 2.5% and less after calibrating the cameras using a checkerboard of dimensions 290 \u0026times; 210 mm (presumed). Delacy et al. (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) obtained MREs of \u0026lt;\u0026thinsp;1% at angles up to 20\u0026deg;, whereas Garner et al. (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) achieved MREs within the \u0026plusmn;\u0026thinsp;5% error threshold they selected at angles\u0026thinsp;\u0026le;\u0026thinsp;10\u0026deg; at several distances. However, our checkerboard size was smaller than the one used by Delacy et al. (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) (571 \u0026times; 317 mm) and by Garner et al. (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) (610 \u0026times; 457 mm), and it was detected by StereoMorph to a maximum distance of 6 m. Therefore, we recommend using a larger checkerboard, which improves calibration accuracy (Boutros et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Olsen and Westneat (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) recommended using a checkerboard image at least 40 pixels wide with as many internal corners as possible to maximize calibration accuracy since more internal corners can increase calibration accuracy by providing more data points, particularly for correcting lens distortion.\u003c/p\u003e \u003cp\u003eAdditionally, camera resolution and water conditions (e.g. poor visibility) can negatively affect the reconstructed 3D space within a software since they can influence the fish silhouette identification and, therefore, the range and angles at which fish can be correctly identified and accurately measured (Harvey and Shortis \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Savina et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Goetze et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Most studies using expensive high-end camera technology are limited to distances of 8\u0026ndash;10 m to achieve accurate length estimates (Goetze and Fullwood \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Santana-Garcon et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Delacy et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), even if water visibility exceeds 20 m. For the case of GoPro system configurations, it has been suggested that measurements be restricted to when the target is at distances closer than 5 m (Letessier et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) to overcome limitations related to the distance, encounter rate, and angles at which fish can be seen from the stereo-video system (Santana‐Garcon et al. 2014; Garc\u0026iacute;a-Baciero et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Our setup used GoPro cameras in an environment with uncontrolled conditions, showing error values slightly higher in the ocean than in the pool trial for both measuring tools. Freshwater and saltwater have different densities, affecting diffraction and distortion angles, which can influence measurement quality. Moreover, given that calibration was performed under high light levels in a pool, we recommend investigating the impact of lower light conditions (e.g., deeper water, cloudy days) on measurement accuracy. However, while there is a big difference in the overall error between both software, the MRE ranged from 0.36 to 1.81% for StereoMorph in angles perpendicular to the camera\u0026rsquo;s optical axis and for all object\u0026rsquo;s true lengths. These values were obtained at distances less than 3 m, which are well within the suggested distance for measuring fish in ocean conditions.\u003c/p\u003e \u003cp\u003eMonitoring fish poses further challenges as length estimations may be affected by random factors such as water turbidity, fish movement, low light, fish camouflage with the background, or even fish size (Shortis et al. \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). This makes it more challenging to place the reference points and, thus, get more accurate and precise measurements than in a controlled environment (Harvey et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Neuswanger et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Stereo-video systems have proved to detect changes in fish length-frequency distributions, and that size structure information is similar to other sampling methods such as line and trap data (Langlois et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). The present study did not intend to compare fish size structures across software because we collected too few fish measurements per species to represent their size structure accurately (Arachchige-Weerarathne et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Nonetheless, we found that StereoMorph produced length estimates similar to those obtained with EventMeasure for common reef fish species in the Mexican Pacific, including the Mexican goatfish (\u003cem\u003eMulloidichthys dentatus\u003c/em\u003e). Measurements of this species with StereoMorph showed similar precision (Mean CV\u0026thinsp;=\u0026thinsp;15.8%) to EventMeasure (Mean CV\u0026thinsp;=\u0026thinsp;15.0%), indicating that both software can be used interchangeably for measuring fish of comparable sizes, assuming measurements are made when fish are laterally positioned relative to the camera. Additionally, given that \u003cem\u003eM. dentatus\u003c/em\u003e can reach up to 40 cm in total length and that a related species grows at an average rate of ~\u0026thinsp;0.30 cm yr\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Mehanna et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), StereoMorph is well-suited for detecting annual growth rate changes in reef fish species, even those with small growth rates.\u003c/p\u003e \u003cp\u003eGiven the rapid decline and threat of extinction facing many fish populations, accurate length data is crucial for guiding conservation and management efforts, particularly in low-income areas where invasive sampling is unsuitable due to low population levels. Currently, the main drivers affecting fish populations are climate change(Brander \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Free et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Whitfield et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), causing species distributions to move toward higher latitudes and deeper areas to track optimum temperature ranges (Perry et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Pinsky et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and overfishing (Jackson et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Myers and Worm \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Mullon et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Data on fish length estimates provide insights into the reproductive potential, growth, and stability of fish populations (Hixon et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; van Overzee and Rijnsdorp \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Audzijonyte et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Moreover, length composition analysis can provide insights into the dynamics of fish populations (Pet et al. \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Audzijonyte et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). A lack of smaller fish may indicate recruitment challenges, while a low number of larger fish could signal high mortality rates among mature individuals (Neumann and Allen \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Overfishing impacts population size structure by removing large and valuable individuals (Fenberg and Roy \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Shantz et al. \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), leading to increased mortality that shortens lifespans and reduces the average size of individuals (Kuparinen et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Alonso-Fern\u0026aacute;ndez et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Overexploited areas thus exhibit smaller average body sizes in targeted species, which negatively impacts populations by decreasing productivity and delaying maturation (Hamilton et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Conover et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Given the shifts in fish populations due to anthropogenic stressors, obtaining accurate fish length estimates is therefore essential for implementing management and conservation strategies.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eGenerally, EventMeasure demonstrated to be a more accurate and precise tool than StereoMorph. However, this open-access 2D calibration-based software can produce similar errors to commercial 3D calibration-based software when objects or fish are close to the plane axis and at distances up to 4 m. Additionally, both measuring tools produced similar precision results when estimating the lengths of paired fish individuals. This suggests that StereoMorph performs well under equal circumstances. As a result, this free and open-access tool can be considered highly valuable for low-budget institutions conducting reef fish monitoring programs aiming at estimating fish size trends without having to spend significant economic resources.\u003c/p\u003e \u003cp\u003eThe stereo-video system produced measurements where the error increased with increasing distances and decreased with increasing object size. With errors below 2.5% when measuring objects at 3 m in the ocean, this may improve efficiency compared to traditional visual size estimation when conducting UVC methods, as it requires length-measurement training to obtain accurate in situ measurements. Until now, those wanting to use stereo-video have generally had to invest a significant amount of money in calibration equipment and commercial software, which may discourage researchers in developing countries from using stereo-video as a survey tool. As this open-access software reduces the cost of software and calibration equipment, we encourage more researchers to use stereo-video systems to enhance length estimate data. To achieve this, we recommend measuring fish close to the plane axis to avoid adding lens distortion bias into measurements. Furthermore, it is recommended to use this configuration to provide precise measurements for fish not less than 100 mm in size unless they are closer than 3 m to the cameras. These considerations will allow using StereoMorph to estimate fish length-frequency distribution using accurate and precise length estimations while avoiding error magnification during the calculations (Harvey et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Finally, by following an appropriate camera configuration and 2D calibration, StereoMorph could readily be used with stereo-DOVs to detect changes in the growth rates of reef fish species annually, even for those species with small growth rates.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eDeclaration of Competing Interest\u003c/h2\u003e \u003cp\u003eThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThis research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. We would like to thank the NGOs Pelagios Kakunj\u0026aacute; and MigraMar for letting us access the calibration and measuring software from SeaGIS. We would also like to extend our gratitude to Tu Polideportivo La Paz for allowing us to calibrate the stereo-video system in their sports facilities.\u003c/p\u003e\u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eData will be made available on request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbdel-Aziz YI (1971) Direct linear transformation from comparator coordinates into object space coordinates in close range photogrammetry. Proc Amer Soc Photogrammetry 1971:1\u0026ndash;19. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.14358/PERS.81.2.103c\u003c/span\u003e\u003cspan address=\"10.14358/PERS.81.2.103c\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAjemian M, Kenworthy M, S\u0026aacute;nchez-Lizaso JL, Cebrian J (2016) Aggregation dynamics and foraging behaviour of striped red mullet Mullus surmuletus in the western Mediterranean Sea. J Fish Biol 88:2051\u0026ndash;2059. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jfb.12932\u003c/span\u003e\u003cspan address=\"10.1111/jfb.12932\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlonso-Fern\u0026aacute;ndez A, Otero J, Ba\u0026ntilde;\u0026oacute;n R (2021) Supplementary Material Indicators of body size variability in a highly developed small-scale fishery: ecological and management implications \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ecolind.2020.107141\u003c/span\u003e\u003cspan address=\"10.1016/j.ecolind.2020.107141\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAngelini R, Lima MAL, Lira AS, Lucena-Fr\u0026eacute;dou F, Fr\u0026eacute;dou T, Bertrand A, Giarrizzo T, Steenbeek J, Coll M, Keppeler FW (2025) The projected impacts of climate change and fishing pressure on a tropical marine food web. Mar Environ Res 204:106909. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.marenvres.2024.106909\u003c/span\u003e\u003cspan address=\"10.1016/j.marenvres.2024.106909\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArachchige-Weerarathne I, Monk J, Barrett N (2021) Sample-size requirements for accurate length-frequency distributions of mesophotic reef fishes from baited remote underwater stereo video. Ecol Ind 122:107262. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ecolind.2020.107262\u003c/span\u003e\u003cspan address=\"10.1016/j.ecolind.2020.107262\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAudzijonyte A, Richards SA, Stuart-Smith RD, Pecl G, Edgar GJ, Barrett NS, Payne N, Blanchard JL (2020) Fish body sizes change with temperature but not all species shrink with warming. Nat Ecol Evol 4:809\u0026ndash;814. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41559-020-1171-0\u003c/span\u003e\u003cspan address=\"10.1038/s41559-020-1171-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarnett LA, Branch TA, Ranasinghe RA, Essington TE (2017) Old-growth fishes become scarce under fishing. Curr Biol 27:2843\u0026ndash;2848 e2842 doi. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.1016/j.cub.2017.07.069\u003c/span\u003e\u003cspan address=\"10.1016/j.cub.2017.07.069\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoutros N, Shortis MR, Harvey ES (2015) A comparison of calibration methods and system configurations of underwater stereo-video systems for applications in marine ecology. Limnol Oceanography: Methods 13:224\u0026ndash;236. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/lom3.10020\u003c/span\u003e\u003cspan address=\"10.1002/lom3.10020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoynton GC, Voss KJ (2006) An underwater digital stereo video camera for fish population assessment. University of Miami, Tech Rep\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrander KM (2007) Global fish production and climate change. Proceedings of the national academy of sciences 104: 19709\u0026ndash;19714 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.0702059104\u003c/span\u003e\u003cspan address=\"10.1073/pnas.0702059104\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrandl SJ, Johansen JL, Casey JM, Tornabene L, Morais RA, Burt JA (2020) Extreme environmental conditions reduce coral reef fish biodiversity and productivity. Nat Commun 11:3832. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-020-17731-2\u003c/span\u003e\u003cspan address=\"10.1038/s41467-020-17731-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrock VE (1954) A preliminary report on a method of estimating reef fish populations. J Wildl Manag 18:297\u0026ndash;308. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2307/3797016\u003c/span\u003e\u003cspan address=\"10.2307/3797016\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCaldwell ZR, Zgliczynski BJ, Williams GJ, Sandin SA (2016) Reef fish survey techniques: assessing the potential for standardizing methodologies. PLoS ONE 11:e0153066. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0153066\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0153066\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCappo M, Speare P, De'ath G (2004) Comparison of baited remote underwater video stations (BRUVS) and prawn (shrimp) trawls for assessments of fish biodiversity in inter-reefal areas of the Great Barrier Reef Marine Park. J Exp Mar Biol Ecol 302:123\u0026ndash;152. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jembe.2003.10.006\u003c/span\u003e\u003cspan address=\"10.1016/j.jembe.2003.10.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClarke TA, Fryer JG (1998) The development of camera calibration methods and models. Photogram Rec 16:51\u0026ndash;66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/0031-868X.00113\u003c/span\u003e\u003cspan address=\"10.1111/0031-868X.00113\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConover DO, Munch SB, Arnott SA (2009) Reversal of evolutionary downsizing caused by selective harvest of large fish. Proceedings of the Royal Society B: Biological Sciences 276: 2015\u0026ndash;2020 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1098/rspb.2009.0003\u003c/span\u003e\u003cspan address=\"10.1098/rspb.2009.0003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDelacy CR, Olsen A, Howey LA, Chapman DD, Brooks EJ, Bond ME (2017) Affordable and accurate stereo-video system for measuring dimensions underwater: a case study using oceanic whitetip sharks Carcharhinus longimanus. Mar Ecol Prog Ser 574:75\u0026ndash;84. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3354/meps12190\u003c/span\u003e\u003cspan address=\"10.3354/meps12190\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDunbrack RL (2006) In situ measurement of fish body length using perspective-based remote stereo-video. Fish Res 82:327\u0026ndash;331. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fishres.2006.08.017\u003c/span\u003e\u003cspan address=\"10.1016/j.fishres.2006.08.017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFenberg PB, Roy K (2008) Ecological and evolutionary consequences of size-selective harvesting: how much do we know? Mol Ecol 17:209\u0026ndash;220. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1365-294X.2007.03522.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-294X.2007.03522.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFree CM, Thorson JT, Pinsky ML, Oken KL, Wiedenmann J, Jensen OP (2019) Impacts of historical warming on marine fisheries production. Science 363:979\u0026ndash;983. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aau1758\u003c/span\u003e\u003cspan address=\"10.1126/science.aau1758\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarc\u0026iacute;a-Baciero A, Borges-Souza JM, Palomares-Garc\u0026iacute;a JR, Rodr\u0026iacute;guez-S\u0026aacute;nchez R, Rubio-Rodr\u0026iacute;guez U, Villalobos H (2024) Monitoring fish populations using stereo-DOV-based surveys and open-access tools in the Gulf of California. 103926. Regional Studies in Marine Science\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rsma.2024.103926\u003c/span\u003e\u003cspan address=\"10.1016/j.rsma.2024.103926\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarner SB, Olsen AM, Caillouet R, Campbell MD, Patterson WF III (2021) Estimating reef fish size distributions with a mini remotely operated vehicle-integrated stereo camera system. PLoS ONE 16:e0247985. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0247985\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0247985\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoetze J, Fullwood L (2013) Fiji\u0026rsquo;s largest marine reserve benefits reef sharks. Coral Reefs 32:121\u0026ndash;125. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00338-012-0970-4\u003c/span\u003e\u003cspan address=\"10.1007/s00338-012-0970-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoetze J, Jupiter S, Langlois TJ, Wilson S, Harvey ES, Bond T, Naisilisili W (2015) Diver operated video most accurately detects the impacts of fishing within periodically harvested closures. J Exp Mar Biol Ecol 462:74\u0026ndash;82. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jembe.2014.10.004\u003c/span\u003e\u003cspan address=\"10.1016/j.jembe.2014.10.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoetze JS, Bond T, McLean DL, Saunders BJ, Langlois TJ, Lindfield S, Fullwood LA, Driessen D, Shedrawi G, Harvey ES (2019) A field and video analysis guide for diver operated stereo-video. Methods Ecol Evol 10:1083\u0026ndash;1090. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/2041-210X.13189\u003c/span\u003e\u003cspan address=\"10.1111/2041-210X.13189\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHamilton SL, Caselle JE, Standish JD, Schroeder DM, Love MS, Rosales-Casian JA, Sosa-Nishizaki O (2007) Size-selective harvesting alters life histories of a temperate sex‐changing fish. Ecol Appl 17:2268\u0026ndash;2280. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1890/06-1930.1\u003c/span\u003e\u003cspan address=\"10.1890/06-1930.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarvey E, Fletcher D, Shortis MR, Kendrick GA (2004) A comparison of underwater visual distance estimates made by scuba divers and a stereo-video system: implications for underwater visual census of reef fish abundance. Mar Freshw Res 55:573\u0026ndash;580. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1071/MF03130\u003c/span\u003e\u003cspan address=\"10.1071/MF03130\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarvey E, Shortis M (1995) A system for stereo-video measurement of sub-tidal organisms. Mar Technol Soc J 29:10\u0026ndash;22\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarvey E, Shortis M, Stadler M, Cappo M (2002) A comparison of the accuracy and precision of measurements from single and stereo-video systems. Mar Technol Soc J 36:38\u0026ndash;49. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4031/002533202787914106\u003c/span\u003e\u003cspan address=\"10.4031/002533202787914106\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarvey ES, Goetze J, McLaren B, Langlois T, Shortis MR (2010) Influence of range, angle of view, image resolution and image compression on underwater stereo-video measurements: high-definition and broadcast-resolution video cameras compared. Mar Technol Soc J 44:75\u0026ndash;85. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4031/MTSJ.44.1.3\u003c/span\u003e\u003cspan address=\"10.4031/MTSJ.44.1.3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarvey ES, Shortis MR (1998) Calibration stability of an underwater stereo-video system: implications for measurement accuracy and precision. Mar Technol Soc J 32:3\u0026ndash;17\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeenan A, Williams ID, Acoba T, DesRochers A, Kosaki RK, Kanemura T, Nadon MO, Brainard RE (2017) Long-term monitoring of coral reef fish assemblages in the Western central pacific. Sci Data 4:1\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/sdata.2017.176\u003c/span\u003e\u003cspan address=\"10.1038/sdata.2017.176\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHeikkila J, Silv\u0026eacute;n O (1997) A four-step camera calibration procedure with implicit image correction Proceedings of IEEE computer society conference on computer vision and pattern recognition. IEEE, pp 1106\u0026ndash;1112\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHixon MA, Johnson DW, Sogard SM (2014) BOFFFFs: on the importance of conserving old-growth age structure in fishery populations. ICES J Mar Sci 71:2171\u0026ndash;2185. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/icesjms/fst200\u003c/span\u003e\u003cspan address=\"10.1093/icesjms/fst200\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoward RA, Ciannelli L, Wakefield WW, Haltuch MA (2023) Comparing fishery-independent and fishery-dependent data for analysis of the distributions of Oregon shelf groundfishes. Fish Res 258:106553. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fishres.2022.106553\u003c/span\u003e\u003cspan address=\"10.1016/j.fishres.2022.106553\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJackson JB, Kirby MX, Berger WH, Bjorndal KA, Botsford LW, Bourque BJ, Bradbury RH, Cooke R, Erlandson J, Estes JA (2001) Historical overfishing and the recent collapse of coastal ecosystems. science 293: 629\u0026ndash;637 doi https://doi.org/9601\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJessop SA, Saunders BJ, Goetze JS, Barrett NS, Harvey ES (2024) A comparison of the behavioural responses of fishes to a remotely operated vehicle and diver-based stereo-video sampling. Estuar Coast Shelf Sci 298:108621. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ecss.2024.108621\u003c/span\u003e\u003cspan address=\"10.1016/j.ecss.2024.108621\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKing M (2007) Fisheries biology, assessment and management\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuparinen A, Boit A, Valdovinos FS, Lassaux H, Martinez ND (2016) Fishing-induced life-history changes degrade and destabilize harvested ecosystems. Sci Rep 6:22245. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/srep22245\u003c/span\u003e\u003cspan address=\"10.1038/srep22245\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLanglois TJ, Fitzpatrick BR, Fairclough DV, Wakefield CB, Hesp SA, McLean DL, Harvey ES, Meeuwig JJ (2012) Similarities between line fishing and baited stereo-video estimations of length-frequency: novel application of kernel density estimates. PLoS ONE 7:e45973. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0045973\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0045973\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLetessier TB, Juhel J-B, Vigliola L, Meeuwig JJ (2015) Low-cost small action cameras in stereo generates accurate underwater measurements of fish. J Exp Mar Biol Ecol 466:120\u0026ndash;126. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jembe.2015.02.013\u003c/span\u003e\u003cspan address=\"10.1016/j.jembe.2015.02.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026oacute;pez-Mac\u0026iacute;as J, Bravo-Ormaza E, Chinacalle-Mart\u0026iacute;nez N, Miranda C, Murillo-Posada J, Vallejo F, Pe\u0026ntilde;aherrera-Palma C (2023) Comparison of Two Stereo-Video Software for the Assessment of Marine Resources. Thalassas: Int J Mar Sci 39:395\u0026ndash;404. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s41208-022-00507-4\u003c/span\u003e\u003cspan address=\"10.1007/s41208-022-00507-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarra S, Coppa S, Camedda A, Mazzoldi C, Wrachien F, Massaro G, de Lucia GA (2016) Recovery trends of commercial fish: the case of an underperforming mediterranean marine protected area. PLoS ONE 11:e0146391. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0146391\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0146391\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMehanna SF, Osman AGM, Farrag MMS, Osman YAA (2018) Age and growth of three common species of goatfish exploited by artisanal fishery in Hurghada fishing area, Egypt. J Appl Ichthyol 34:917\u0026ndash;921. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jai.13590\u003c/span\u003e\u003cspan address=\"10.1111/jai.13590\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMullon C, Fr\u0026eacute;on P, Cury P (2005) The dynamics of collapse in world fisheries. Fish Fish 6:111\u0026ndash;120. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1467-2979.2005.00181.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1467-2979.2005.00181.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurphy HM, Jenkins GP (2010) Observational methods used in marine spatial monitoring of fishes and associated habitats: a review. Mar Freshw Res 61:236\u0026ndash;252. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1071/MF09068\u003c/span\u003e\u003cspan address=\"10.1071/MF09068\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMyers RA, Worm B (2003) Rapid worldwide depletion of predatory fish communities. Nature 423:280\u0026ndash;283. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nature01610\u003c/span\u003e\u003cspan address=\"10.1038/nature01610\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeumann RM, Allen MS (2007) Analysis and interpretation of freshwater fisheries data\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeuswanger JR, Wipfli MS, Rosenberger AE, Hughes NF (2016) Measuring fish and their physical habitats: versatile 2D and 3D video techniques with user-friendly software. Can J Fish Aquat Sci 73:1861\u0026ndash;1873. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1139/cjfas-2016-0010\u003c/span\u003e\u003cspan address=\"10.1139/cjfas-2016-0010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlsen A, Haber A (2022) StereoMorph: stereo camera calibra-tion and reconstruction. Version 1.6. 4\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlsen AM, Westneat MW (2015) StereoMorph: An R package for the collection of 3D landmarks and curves using a stereo camera set-up. Methods Ecol Evol 6:351\u0026ndash;356. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/2041-210X.12326\u003c/span\u003e\u003cspan address=\"10.1111/2041-210X.12326\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOno K, Licandeo R, Muradian ML, Cunningham CJ, Anderson SC, Hurtado-Ferro F, Johnson KF, McGilliard CR, Monnahan CC, Szuwalski CS (2015) The importance of length and age composition data in statistical age-structured models for marine species. ICES J Mar Sci 72:31\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/icesjms/fsu007\u003c/span\u003e\u003cspan address=\"10.1093/icesjms/fsu007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePennino MG, Conesa D, L\u0026oacute;pez-Qu\u0026iacute;lez A, Munoz F, Fern\u0026aacute;ndez A, Bellido JM (2016) Fishery-dependent and-independent data lead to consistent estimations of essential habitats. ICES J Mar Sci 73:2302\u0026ndash;2310. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/icesjms/fsw062\u003c/span\u003e\u003cspan address=\"10.1093/icesjms/fsw062\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerry AL, Low PJ, Ellis JR, Reynolds JD (2005) Climate change and distribution shifts in marine fishes. science 308: 1912\u0026ndash;1915 doi https://doi.org/03081912\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePet J, Van Densen W, Machiels M, Sukkel M, Setyohadi D, Tumuljadi A (1997) Length-based analysis of population dynamics and stock identification in the sardine fisheries around East Java, Indonesia. Fish Res 31:107\u0026ndash;120. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0165-7836(97)00006-4\u003c/span\u003e\u003cspan address=\"10.1016/S0165-7836(97)00006-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinsky ML, Selden RL, Kitchel ZJ (2020) Climate-driven shifts in marine species ranges: scaling from organisms to communities. Annual Rev Mar Sci 12:153\u0026ndash;179. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1146/annurev-marine-010419-010916\u003c/span\u003e\u003cspan address=\"10.1146/annurev-marine-010419-010916\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQueir\u0026oacute;s AM, Fernandes J, Genevier L, Lynam CP (2018) Climate change alters fish community size-structure, requiring adaptive policy targets. Fish Fish 19:613\u0026ndash;621. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/faf.12278\u003c/span\u003e\u003cspan address=\"10.1111/faf.12278\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR Core Team (2024) R: A language and environment for statistical computing. R Foundation fot Statistical Computing. Vienna, Austria\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSantana-Garcon J, Braccini M, Langlois TJ, Newman SJ, McAuley RB, Harvey ES (2014) Calibration of pelagic stereo‐BRUV s and scientific longline surveys for sampling sharks. Methods Ecol Evol 5:824\u0026ndash;833. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/2041-210X.12216\u003c/span\u003e\u003cspan address=\"10.1111/2041-210X.12216\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSavina E, Krag LA, Madsen N (2018) Developing and testing a computer vision method to quantify 3D movements of bottom-set gillnets on the seabed. ICES J Mar Sci 75:814\u0026ndash;824. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/icesjms/fsx194\u003c/span\u003e\u003cspan address=\"10.1093/icesjms/fsx194\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchramm KD, Harvey ES, Goetze JS, Travers MJ, Warnock B, Saunders BJ (2020) A comparison of stereo-BRUV, diver operated and remote stereo-video transects for assessing reef fish assemblages. J Exp Mar Biol Ecol 524:151273. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jembe.2019.151273\u003c/span\u003e\u003cspan address=\"10.1016/j.jembe.2019.151273\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShafait F, Harvey ES, Shortis MR, Mian A, Ravanbakhsh M, Seager JW, Culverhouse PF, Cline DE, Edgington DR, Browman H (2017) Towards automating underwater measurement of fish length: a comparison of semi-automatic and manual stereo\u0026ndash;video measurements. ICES J Mar Sci 74:1690\u0026ndash;1701. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/icesjms/fsx007\u003c/span\u003e\u003cspan address=\"10.1093/icesjms/fsx007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShantz AA, Ladd MC, Burkepile DE (2020) Overfishing and the ecological impacts of extirpating large parrotfish from Caribbean coral reefs. Ecol Monogr 90:e01403. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ecm.1403\u003c/span\u003e\u003cspan address=\"10.1002/ecm.1403\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShortis MR, Ravanbakskh M, Shaifat F, Beyerer J, Puente Le\u0026oacute;n F, Harvey ES, Mian A, Seager JW, Culverhouse PF, Cline DE, Edgington DR (2013) A review of techniques for the identification and measurement of fish in underwater stereo-video image sequences Videometrics. Range Imaging, and Applications XII; and Automated Visual Inspection\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith ML (1988) Effects of observer swimming speed on sample counts of temperate rocky reef fish assemblages. Mar Ecol progress Ser Oldendorf 43:223\u0026ndash;231\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStrona G, Beck PS, Cabeza M, Fattorini S, Guilhaumon F, Micheli F, Montano S, Ovaskainen O, Planes S, Veech JA (2021) Ecological dependencies make remote reef fish communities most vulnerable to coral loss. Nat Commun 12:7282. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-021-27440-z\u003c/span\u003e\u003cspan address=\"10.1038/s41467-021-27440-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTortolero-Langarica JA, Rodr\u0026iacute;guez‐Troncoso AP, Cupul‐Maga\u0026ntilde;a AL, Morales‐de‐Anda DE, Caselle JE, Carricart‐Ganivet JP (2022) Coral calcification and carbonate production in the eastern tropical Pacific: The role of branching and massive corals in the reef maintenance. Geobiology 20:533\u0026ndash;545. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/gbi.12491\u003c/span\u003e\u003cspan address=\"10.1111/gbi.12491\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTu C-Y, Chen K-T, Hsieh C-h (2018) Fishing and temperature effects on the size structure of exploited fish stocks. Sci Rep 8:7132\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Overzee HM, Rijnsdorp AD (2015) Effects of fishing during the spawning period: implications for sustainable management. Rev Fish Biol Fish 25:65\u0026ndash;83. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11160-014-9370-x\u003c/span\u003e\u003cspan address=\"10.1007/s11160-014-9370-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWehkamp M, Fischer P (2014) A practical guide to the use of consumer-level digital still cameras for precise stereogrammetric in situ assessments in aquatic environments. Underw Technol 32:111\u0026ndash;128. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3723/ut.32.111\u003c/span\u003e\u003cspan address=\"10.3723/ut.32.111\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWhitfield AK, Gillanders BM, Able KW (2023) Climate change effects on fish populations Climate change and estuaries. CRC, pp 475\u0026ndash;506\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWhitmarsh SK, Fairweather PG, Huveneers C (2017) What is Big BRUVver up to? Methods and uses of baited underwater video. Rev Fish Biol Fish 27:53\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11160-016-9450-1\u003c/span\u003e\u003cspan address=\"10.1007/s11160-016-9450-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilliams EH, Shertzer KW, Bacheler N (2022) Estimating length composition of fish observed with stereo-video cameras: A simulation study with application to red snapper (Lutjanus campechanus). Fish Res 254:106424. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fishres.2022.106424\u003c/span\u003e\u003cspan address=\"10.1016/j.fishres.2022.106424\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilliams SB, Pizarro O, Jakuba MV, Mahon I, Ling SD, Johnson CR (2010) Repeated AUV surveying of urchin barrens in North Eastern Tasmania 2010 IEEE International Conference on Robotics and Automation. IEEE, pp 293\u0026ndash;299\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoolston C (2023) Facing inflation: lab heads tighten supplies budgets. Nature 613:601\u0026ndash;602. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/d41586-023-00088-z\u003c/span\u003e\u003cspan address=\"10.1038/d41586-023-00088-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"marine-biology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mabi","sideBox":"Learn more about [Marine Biology](https://www.springer.com/journal/227)","snPcode":"227","submissionUrl":"https://submission.nature.com/new-submission/227/3","title":"Marine Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Stereo-videography, Fish, Monitoring, Open-access, Measurements, Length estimates","lastPublishedDoi":"10.21203/rs.3.rs-5501285/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5501285/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePrecise descriptions of size structure are crucial for the adaptive management of marine fish populations influenced by human activity and environmental factors. Stereo-video systems are powerful tools for monitoring fish populations. Yet, due to the high investment required in software and equipment, stereo-videography can have some financial issues. This study compared the performance of the commercial software EventMeasure from SeaGIS and the open-access R package StereoMorph. We evaluated the effect of distance to the system, rotation movement of the object, and true length on the accuracy and precision of each software. Additionally, we used a diver-operated stereo-video system to obtain in situ measurements of reef fish species. EventMeasure was generally more accurate (error\u0026thinsp;=\u0026thinsp;0.53%) and precise (CV\u0026thinsp;=\u0026thinsp;0.35%) than StereoMorph (error\u0026thinsp;=\u0026thinsp;4.54%; CV\u0026thinsp;=\u0026thinsp;0.65%). However, the latter showed errors\u0026thinsp;\u0026lt;\u0026thinsp;5% when measuring objects at distances up to 4 m and close to the plane axis. The precision for in situ measurements for EventMeasure (CV\u0026thinsp;=\u0026thinsp;15.0%) was similar to that of StereoMorph (Mean CV\u0026thinsp;=\u0026thinsp;15.8%). We found a high correlation (ρ\u0026thinsp;=\u0026thinsp;0.93, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) between paired fish length estimation from both software, although StereoMorph was slightly more precise than EventMeasure (Mean CV\u0026thinsp;=\u0026thinsp;2.07% and Mean CV\u0026thinsp;=\u0026thinsp;2.11%, respectively). This open-access software provided suitable accuracy and precision results despite some limitations, offering the option to reduce software costs without compromising accuracy for affordability.\u003c/p\u003e","manuscriptTitle":"Comparing the performance of two scientific tools for obtaining fish length measurements","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-02 16:16:28","doi":"10.21203/rs.3.rs-5501285/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2025-04-30T12:37:58+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-04-30T12:17:57+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-04-25T07:16:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Marine Biology","date":"2025-04-24T23:15:12+00:00","index":"","fulltext":""},{"type":"decision","content":"Accept","date":"2025-04-22T06:41:11+00:00","index":"","fulltext":""},{"type":"decision","content":"Editor Decision - Provisional Accept","date":"2025-04-22T06:41:11+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"marine-biology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mabi","sideBox":"Learn more about [Marine Biology](https://www.springer.com/journal/227)","snPcode":"227","submissionUrl":"https://submission.nature.com/new-submission/227/3","title":"Marine Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"8b499798-ddd4-4b4c-aee5-d53cf7562816","owner":[],"postedDate":"May 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-06-30T16:07:14+00:00","versionOfRecord":{"articleIdentity":"rs-5501285","link":"https://doi.org/10.1007/s00227-025-04682-9","journal":{"identity":"marine-biology","isVorOnly":false,"title":"Marine Biology"},"publishedOn":"2025-06-27 16:05:43","publishedOnDateReadable":"June 27th, 2025"},"versionCreatedAt":"2025-05-02 16:16:28","video":"","vorDoi":"10.1007/s00227-025-04682-9","vorDoiUrl":"https://doi.org/10.1007/s00227-025-04682-9","workflowStages":[]},"version":"v1","identity":"rs-5501285","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5501285","identity":"rs-5501285","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}