The aim of this study was to assess the suitability of different vertebrae staining techniques for the visualization and counting of growth bands in tropical species of batoids (
El objetivo de este estudio fue evaluar la efectividad de diferentes técnicas de tinción de vértebras en la visualización y el conteo de bandas de crecimiento en especies tropicales de batoideos (
Due to their life history characteristics (
Age studies in elasmobranchs are based on counting pairs of growth bands in hard structures (e.g. vertebrae, spines) which are formed periodically. As the visualization of these band pairs in some cases is not simple, over time several methods, such as X-rays (
Of the aforementioned processes, staining has been the most widely used, mainly because it is the least complex and least expensive procedure, and it is even suggested as a first step before applying more demanding and expensive methods (
Furthermore, few studies have reported in detail the procedures applied to select the most appropriate staining method for each species (e.g.
For the above reasons, and because studies based on radiocarbon dating, histology and X-rays are more restrictive in tropical countries, this study aimed to evaluate different staining techniques in vertebrae of nine tropical elasmobranch species to determine those that facilitate the visualization and counting of the growth bands, contributing to future age estimates and population assessments of the evaluated tropical species.
A total of 428 individuals were collected between 2007 ans 2015 from fisheries of the Colombian Pacific region (79°44′W, 5°45′N; 77°12′W, 2°15′N). These individuals belong to four batoid species,
Once each specimen was identified at the species level, its sex was determined and total length of sharks or disc width of rays was recorded. From each individual, a section of 14 to 16 vertebrae was extracted from under the first dorsal fin (in sharks) and from the abdominal region (in rays), which were labelled and frozen until laboratory analysis. The excess tissue in the vertebrae was removed using a scalpel. Because affinity of the stains and visualization of the growth bands could be affected by the individual size, vertebrae were separated into three size intervals: small, medium and large diameter. This was done for the species whose sample size allowed it.
As the thickness of the vertebrae sections influences the visualization and reading of the growth bands, size ranges of the vertebrae with different section thicknesses (0.3 to 0.7 mm) were combined per individual. For this purpose, each vertebra was fixed to a slide with Crystalbond 509 and cut sagittally with an Isomet Buehler low-speed saw cutter with two diamond head blades (Buehler, Lake Bluff, IL, USA) to obtain bow-tie sections (
In order to assess the effect of the stains used, sections were randomized and stained with alizarin red (0.05%), methylene blue (0.001%), crystal violet (0.001%), basic fuchsin (0.001%), acid fuchsin (0.001%), Bismarck brown (0.05%), light green (0.05%), silver nitrate (1%) and the Dahl staining (alizarin red 0.01% and light green 0.05%). Each dye was applied in successive intervals of one minute until reaching its saturation point in each structure. This was done as a preliminary test for a sub-sample of each species (n=5).
Subsequently, the best three staining times for each dye were qualitatively established and applied to the vertebrae of the individuals selected for the analysis. Biases given by individual variations were avoided using several vertebrae per individual to apply the treatments (stain + staining time). Sections (with and without staining) were mounted on a slide, observed under a microscope using transmitted light and photographed for each treatment. Additionally, vertebral sections without any staining were used to evaluate the effect of immersion oil and distilled water (imbibing each in a drop of the substance) in the visualization of the growth bands. The images obtained were analysed with the Image Pro Plus 7.0 software (Media Cybernetics) in which two skilled readers performed the quantitative counting of the growth bands independently in each sample. Readings were carried out twice per reader, who did not know the details of the sex, size or previous reading of each vertebra. Based on this information, analyses were carried out to establish the accuracy and bias among readers and subsequently to establish the most efficient technique for observing and counting growth bands per species.
In order to assess the degree of precision in the vertebral band readings among readers, the index of average percentage error, the coefficient of variation, the percentage of agreement between readers, Bowker’s symmetry test and the percentage of vertebrae read were calculated and analysed as follows.
The index of average percentage error (IAPE) provided information on the accuracy of age estimations among readers; small values indicated more precise readings (
where
Coefficient of variation (CV) measured reading accuracy (
where
Percentage of agreement between readers (PA) allowed us to establish the variation in the reading of bands among readers (
Bowker’s symmetry test determined, using a chi-square test, whether differences between the readers were systematic (p<0.05) or random (p>0.05;
Percentage of vertebrae read (RV) estimated the proportion of vertebrae that could be read successfully considering the entire sample; higher values indicated better performance of the assessed technique.
An evaluation of the results of all the tests applied was carried out in order to determine the most efficient technique for observing and estimating the age in each species studied. Subsequently, the final decision per species was taken by comparing the results of each of the treatments as a whole and by vertebrae size.
The qualitative evaluation of the section thickness for batoids showed that sections of 0.4 mm were most suitable, except for
For the leopard electric ray, treatments were applied for three vertebrae size intervals: small (diameters of 1.30 to 2.13 mm), medium (2.14 to 2.99 mm) and large (3.00 to 4.84 mm). There was no systematic bias in any of the treatments used for visualization of growth bands, except for small vertebrae stained with alizarin red for 14 min (Supplementary material Table S2).
Treatment analysis by vertebrae size showed that the highest PA (±0 bands) values were found in unstained large vertebrae (91.3%), medium vertebrae stained with basic fuchsin for 7 min (73.1%), and small vertebrae stained with methylene blue for 1 min (75%) followed by alizarin red for 16 min (73.1%) (Table S2). The three vertebrae sizes evaluated showed percentages of read vertebrae (RV) higher than 70% in all treatments. The highest RV values were found for large vertebrae stained with methylene blue and basic fuchsin for 3 min (90%); for medium vertebrae without staining or immersed in oil, basic fuchsin for 7 min or alizarin red for 14 min (86.7%); and for small vertebrae immersed in distilled water or in alizarin red for 14 and 16 min (86.7%) (Table S2). IAPE and CV showed relatively low values in large and medium vertebrae and high values in small vertebrae, except in those stained with alizarin red for 16 min, which showed the best values for small and medium intervals.
In conclusion, unstained large vertebrae and the medium and small vertebrae treated with alizarin red for 16 min (
Species | Vertebra size | Treatment (stain + staining time) | IAPE | CV | PA | RV | p |
---|---|---|---|---|---|---|---|
Small | Alizarin red 16′ | 4.9 | 7.0 | 73.1 | 86.7 | 0.14 | |
Medium | Alizarin red 16′ | 3.3 | 4.6 | 64.0 | 83.3 | 0.22 | |
Large | No staining | 0.7 | 1.7 | 91.3 | 76.7 | 0.09 | |
Small | Light green 5′ | 5.1 | 7.3 | 64.7 | 76.5 | 0.16 | |
Medium | Methylene blue 20′ | 1.2 | 1.7 | 84.2 | 89.5 | 0.32 | |
Large | Methylene blue 10′ | 2.2 | 3.1 | 80.0 | 100.0 | 0.39 | |
Small | Bismarck brown 1′ | 2.2 | 3.1 | 93.3 | 100 | 0.32 | |
Bismarck brown 2.0′ | 2.2 | 3.1 | 93.3 | 100 | 0.32 | ||
Light green 0.5′ | 2.2 | 3.1 | 93.3 | 100 | 0.32 | ||
Light green 1′ | 2.2 | 3.1 | 93.3 | 100 | 0.32 | ||
Medium | Basic fuchsin 2′ | 2.2 | 3.1 | 93.3 | 100% | 0.32 | |
Large | Light green 7′ | 0.0 | 0.0 | 100.0 | 100.0 | – | |
All | Alizarin red 7′ | 0.1 | 1.3 | 95.5 | 91.7 | 0.16 | |
All | Bismarck brown 15′ | 1.8 | 2.5 | 57.1 | 77.8 | 0.39 | |
All | Crystal violet 35′ | 0.5 | 0.7 | 85.7 | 63.6 | 0.42 | |
All | Immersion oil | 7.0 | 2.4 | 89.8 | 94.2 | 0.14 | |
All | Crystal violet 20′ | 0.0 | 0.0 | 100.0 | 55.6 | 0.16 | |
All | Light green 40′ | 0.0 | 0.0 | 100.0 | 78.6 | 0.30 |
For the Panamic stingray, treatments were also applied to three vertebrae size: small (diameters of 0.70 to 1.59 mm), medium (1.60 to 2.49 mm) and large (2.50 to 3.30 mm). Only the treatment with basic fuchsin for 1 min in medium vertebrae showed bias in the reading; for all others, the differences were due to random errors (Table S3).
The percentage of read vertebrae was higher than 70% in all treatments for the three size interval assessed, reaching a maximum of 100% with methylene blue for 10 min and alcohol in small vertebrae, with alizarin red for 15 min in medium vertebrae, and with almost all the treatments in the large vertebrae, except with crystal violet (Table S3). The highest PA (±0 bands) between readers occurred with alizarin red for 15 min in small vertebrae, methylene blue for 20 min in medium vertebrae, and methylene blue for 10 min in large vertebrae.
The lowest CV and IAPE values were found with methylene blue for 10 min in large vertebrae (
Vertebra sections of
PA values varied between 80.0% (no staining, basic fuchsin for 1 min and light green for 2 min) and 100% (light green for 7 min) in large vertebrae; between 53.9% (basic fuchsin for 3 min) and 93.3% (basic fuchsin for 2 min) in medium vertebrae, and from 73.3% (basic fuchsin for 1 and 3 min) to 93.3% (Bismarck brown for 1 and 2 min, light green for 0.5 and 1 min) in small vertebrae (Table S4). Previously selected stains showed variable IAPE and CV values within and among vertebrae sizes; the lowest values were found for small vertebrae stained with Bismarck brown for 1 and 2 min and light green for 0.5 and 1 min, for medium vertebrae stained with Bismarck brown for 2 and 3 min, and for large vertebrae stained with light green for 7 min (Table S4). In general, vertebrae with no treatment showed higher IAPE and CV values than vertebrae that received staining. All vertebrae sections of this species were read (100% RV in all cases).
Considering all the results, the best treatment was light green for 7 min for large vertebrae of
Age bias analysis of Magdalena river stingray showed no systematic bias in readings or treatments (Table S5). The percentage of read vertebrae varied between 52.9% for light green for 40 min and 95% for methylene blue for 30 and 40 min, the latter being followed by alizarin red for 7 min, with 91.7%. The lowest IAPE and CV values and the highest PA (±0 bands) values between readers were found with crystal violet for 60 min and light green for 40 min (0 IAPE, 0 CV, 100% PA each), followed by alizarin red for 7 min (0.1 IAPE, 1.3 CV, 95.5% PA). The group analysis of the treatments showed that vertebrae treated with light green and crystal violet require too much time for staining (≥40 min), while alizarin red showed the second-best values, with only 7 min of staining (
Vertebrae of the pelagic thresher shark involved high difficulty in observing and counting a growth band pattern, generating high variation in the precision tests and percentages of read vertebrae (Table S5). None of the treatments showed systematic bias in their readings (P>0.05 in all cases). Vertebrae stained with alizarin red for 5 and 7 min showed the best results in the precision analyses (IAPE, CV and PA), but the percentages of RV with these treatments were very low. Conversely, basic fuchsin for 45 min showed the highest RV (91.7%) but a low PA value (±0 bands=36.4%). Considering the values of all the tests, staining with Bismarck brown for 15 min showed a good performance, occupying the third place in IAPE, CV, and RV and the fourth place in PA values (
None of the treatments analysed for the silky shark showed systematic bias in the readings (P>0.05 for all cases). The highest PA (±0 bands) values between readers were found with silver nitrate for 2 and 3 min (100%) and crystal violet for 35 min (85.7%). Similarly, the lowest IAPE and CV values were obtained with silver nitrate for 2 and 3 min, crystal violet for 35 min and acid fuchsin for 50 min. However, the percentage of RV ranged from a very low value (16.7%) for silver nitrate for 2-3 min to 100% for light green for 35 min (Table S5). Vertebrae stained with crystal violet for 35 min showed the second-best values of IAPE (0.5), CV (0.7) and PA (85.7%), as well as an acceptable RV value (63.6%), being considered the most appropriate treatment for this species (
For the scalloped hammerhead none of the treatments analysed showed systematic bias in the readings (except using Bismarck’s brown for 25 min, P=0.02; Table S6), and all showed a high percentage of RV (>84%) but also relatively high IAPE and CV values. The best values of PA (±0 bands), IAPE and CV were found with immersion oil and methylene blue for 35 and 30 min, which was the best treatments for the species. Therefore, immersion oil was suggested as the reagent for treating the vertebrae of
For the scalloped bonnethead none of the treatments analysed showed systematic bias in the readings (P>0.05 in all cases) and all had a percentage of RV higher than 50%, reaching a maximum of 77.8% (with immersion oil and light green for 35 min). The highest PA values (±0 bands) were found with crystal violet for 20 min (100%), methylene blue for 25 min (81.8%) and crystal violet for 25 min (80.0%; Table S6); similarly, IAPE and CV were lower with crystal violet for 20 and 25 min, as well as with methylene blue for 25 min. Vertebrae treated with crystal violet for 20 min showed the best combination of precision and bias values, establishing it as the best treatment for this species (
For the smooth-hound shark, none of the treatments analysed showed systematic bias in the readings (P>0.05) and the percentage of RV was higher than 60% in all treatments, reaching a maximum of 100% with crystal violet for 30 and 40 min. The highest PA values (±0 bands) and lowest IAPE and CV values were found with light green for 40 min (100%; 0 and 0, respectively) and 30 min (91.7%; 1.2 and 1.7, respectively), followed by crystal violet for 20 min (71.4%; 3.2 and 4.5; Table S6). From the group analysis of the treatments assessed, light green for 40 min showed the best combination of values and is therefore the most appropriate treatment for the species (
There is sufficient evidence that the visualization of the growth bands in hard structures of elasmobranchs varies substantially among species (
Additionally, systematic development of sectioning and staining of the vertebrae and visualizing and counting the growth bands with different arrangements to establish the best combination is a great training exercise for the researchers, which will result in a higher precision in identifying and performing growth band counts and therefore a better estimation of the age of an individual.
Despite the evidence on the importance of developing this procedure by species to establish age and growth of elasmobranchs, many papers do not include detailed information on the procedures performed. This may be due to limited word constraints in publications (usually limiting them to one sentence) or to the fact that these procedures were not performed (and a thickness-stain-staining time combination already established in other studies was chosen; e.g.
Although the bias in the band reading was almost null, the greatest difficulty in the counts was related to the definition of the birthmark and the reading of the bands located towards the edges of the vertebrae. Consequently, the greatest differences between readers occurred in areas near the focus and on the edge of the vertebral sections. Other studies that used similar methods in sagittal vertebrae sections have experienced problems with sharpness in these same areas (e.g.
There are also contrasting results with those found in this study regarding cut thickness, since wider or thinner thicknesses have been proposed as the most suitable for counting growth bands in other species such as
These procedures that require few logistical resources and significant investment of time, substantially improve the results obtained in terms of experience, learning and quality of the data. However, as seen in this study, this is not a general rule, and some shark and ray species have vertebral structures that can be easily visualized without staining procedures (e.g. large vertebrae of
From the results obtained in this study we conclude that it is essential to consider several precision indexes (PA, IAPE, and CV) to define, with the best possible criteria, the combination of section thickness, stain and staining time that is most suitable for the species of interest (Supplementary material, Species comments section). It is a mistake to make this decision based on a single index and to assess the performance of a single technique without contrasting or comparing treatments. In this regard, most studies that used staining techniques have found better performance with stained vertebrae than without staining.
All the examples mentioned above illustrate how the implementation of techniques to enhance the visualization of growth bands significantly influences the accuracy of the age estimation, thus supporting our results. Even the results obtained versus the bibliographic references reviewed show how studies carried out within a same species in different geographical locations (latitudinal differences) identified different techniques as the most appropriate. Furthermore, considering only tropical species, a great variation is identified in the techniques selected for the visualization of the growth bands. This reaffirms the importance of evaluating different combinations (thickness, stain, staining time) for each species to be studied (
Species | Study area | Techniques used | Stainings used | Technique defined | Reference |
---|---|---|---|---|---|
Batoids | |||||
Kane’ohe Bay on Oahu, Hawai’i, USA | Sagittal sectioning, OTC | None | Sagittal sectioning | |
|
Cleveland Bay, Great Barrier Reef, Australia | Sagittal sectioning | None | 0.4 - 0.6 mm sagittal sectioning | |
|
Ningaloo Reef Marine Park, Australia | Sagittal sectioning | None | 0.35 mm sagittal sectioning | |
|
Eastern Atlantic coast, Rio Grande do Norte, Brazil | Sagittal sectioning | None | 0.2 mm sagittal sectioning | |
|
North-east Australia | Sagittal sectioning | None | 0.2 - 0.3 mm sagittal sectioning. | |
|
Ningaloo Reef Marine Park, Australia | Sagittal sectioning | None | 0.35 mm sagittal sectioning | |
|
North-east Australia | Sagittal sectioning | None | 0.2 - 0.3 mm sagittal sectioning. | |
|
North-east Australia | Sagittal sectioning | None | 0.2 - 0.3 mm sagittal sectioning | |
|
Ningaloo Reef Marine Park, Australia | Sagittal sectioning | None | 0.35 mm sagittal sectioning | |
|
Cleveland Bay, Australia | Sagittal sectioning | None | 0.4 - 0.6 mm sagittal sectioning | |
|
Cleveland Bay, Australia | Sagittal sectioning | None | 0.4 - 0.6 mm sagittal sectioning | |
|
Cleveland Bay, Australia | Sagittal sectioning | None | 0.4 - 0.6 mm sagittal sectioning | |
|
Ningaloo Reef Marine Park, Australia | Sagittal sectioning | None | 0.35 mm sagittal sectioning | |
|
Central Pacific coast, Colombia | Sagittal sectioning and staining | Light green (0.05%) methylene blue (0.001%) |
0.4 mm sagittal sectioning staining using light green for small vertebrae and methylene blue for medium and large vertebrae | |
|
Tehuantepec Gulf, Southeast Pacific, Mexico | Sagittal sectioning | None | 0.3 - 0.5 mm sagittal sectioning | |
|
Eastern Atlantic, Pernambuco, Brazil | Sagittal sectioning | None | 0.3 mm sagittal sectioning | |
|
Central Pacific coast, Colombia | Sagittal sectioning | Several (not indicated) |
0.4 mm sagittal sectioning | |
|
Sharks | |||||
Java Sea, Indonesia | Sagittal sectioning | None | 0.3 mm sagittal sectioning | |
|
Northeastern Taiwan | X-ray radiography and staining | Silver nitrate | X-ray radiography | |
|
Atlantic Ocean* | Sagittal sectioning and staining | Crystal violet | 0.5 mm sagittal sectioning and staining | |
|
Queensland, Australia* | Sagittal sectioning | None | 0.4 mm sagittal sectioning | |
|
Queensland, Australia* | Sagittal sectioning | None | 0.4 mm sagittal sectioning | |
|
Veracruz and Campeche, Mexico | Sagittal sectioning | None | 0.3 - 0.6 mm sagittal sectioning | |
|
Eastern Lombok, Indonesia | Sagittal sectioning | None | 0.4 mm sagittal sectioning | |
|
Queensland, Australia* | Sagittal sectioning | None | 0.4 mm sagittal sectioning | |
|
Northeastern Taiwan | Staining of sagittal and longitudinal sectioning, X-radiography of sagittal and longitudinal sectioning | Eosin, haematoxylin | X-radiography | |
|
Eastern Atlantic coast, Maranhão, Brazil | Sagittal sectioning, staining and cedarwood oil | Alizarin red cedarwood oil |
Sagittal sectioning and staining | |
|
Northern Australia | Staining. Protein stains mercurochrome and ninhydrin. Histology, Radiography X-ray, spectrometry, image analysis and examination of sectioned vertebrae under transmitted, reflected, interference and polarized light OTC |
Silver nitrate, Alizarin Red S, crystal violet cobalt nitrate, ammonium sulphide, mercurochrome ninhydrin |
Staining with ninhydrin | |
|
Northern Australia | Staining. Protein stains mercurochrome and ninhydrin. Histology, Radiography X-ray, spectrometry, image analysis and examination of sectioned vertebrae under transmitted, reflected, interference and polarized light OTC |
Silver nitrate, Alizarin Red S, crystal violet, cobalt nitrate, ammonium sulphide, mercurochrome and ninhydrin | Staining with ninhydrin | |
|
Queensland, Australia* | Sagittal sectioning | None | 0.4 mm sagittal sectioning | |
|
Australian east coast* | Sagittal sectioning and staining | Unstained, crystal violet, silver nitrate | 0.15 mm sagittal sectioning | |
|
Queensland, Australia* | Sagittal sectioning | None | 0.4 mm sagittal sectioning | |
|
Eastern Atlantic coast, Maranhão, Brazil | Sagittal sectioning and staining | Alizarin Red S | Sagittal sectioning and staining | |
|
Western and central North Pacific Ocean* | Shadowing (half-cut centra) staining, soft X-radiography (whole or half-cut centra) |
Alizarin red, silver nitrate | Shadowing on half-cut centra | |
|
Western and central Atlantic* | Sagittal sectioning | None | 0.3 mm sagittal sectioning | |
|
Southern Indian Ocean* | Sagittal sectioning and soft X-ray |
None | 1-1.44 mm sagittal sectioning and soft X-ray | |
|
Eastern Atlantic, Brazil | Sagittal sectioning | None | 1 mm sagittal sectioning | |
|
Baja California Peninsula, Mexico | Sagittal sectioning and staining |
Silver nitrate | Whole vertebra stained with silver nitrate and 0.5 mm sagittal sectioning | |
|
South Pacific Ocean* | Transversal sectioning and soft X-ray radiograph |
None | Transversal sectioning and soft X-ray radiograph |
|
|
Northwestern Pacific, Taiwan* | Transversal sectioning and X-ray radiograph | None | Transversal sectioning and X-ray radiograph | |
|
Northwestern Pacific, Taiwan and Pakistan* | Sagittal sectioning Bomb radiocarbon |
None | Sagittal sectioning Bomb radiocarbon |
|
|
Northeastern coast, Australia | Sagittal sectioning | None | 0.4-0.6 mm sagittal sectioning | |
|
Western Atlantic, Senegal | Sagittal sectioning and staining |
Acetic acid + Toluidine blue | Sagittal sectioning and staining | |
|
Eastern Atlantic coast, Maranhão, Brazil | Sagittal sectioning | None | 0.3 mm sagittal sectioning | |
|
Eastern Atlantic coast, Maranhão, Brazil | Sagittal sectioning | None | 0.3 mm sagittal sectioning | |
|
Cleveland Bay, Australia | Vertebrae grounding | None | 0.2-0.4 mm grounding | |
|
Northeastern Taiwan | Sagittal sectioning | None | 0.2 mm sagittal sectioning | |
|
Michoacan, Mexico | Sagittal sectioning and staining |
Crystal violet (0.01%) | Sagittal sectioning and staining | |
|
Java Sea, Indonesia | Sagittal sectioning | None | 0.3 mm sagittal sectioning | |
This variation could determine the calcification of the vertebrae and the affinity of the stains to these structures (based on the carbonate and/or phosphate concentrations), influencing the nature of the dye (basophilic or acidophilic) on the quality of the dye and hence the visualization of the growth bands. However, it is recommended to further this analysis in order to establish whether there is any influence of the type of dye on its ability to improve the visibility of the growth bands. In this regard, it is necessary to carry out a study on the possible causes of the accumulation and type of compounds that make possible changes in birthmarks, e.g. ontogenetic changes in the diet, temperature or reabsorption of materials accumulated in the vertebrae (
Parameter estimation from age and growth studies has profound implications in population assessments based on demography, directly affecting the estimation of demographic parameters and thus the potential management of the species based on their life history traits. Hence, an incorrect specification of the bands and their deposition frequency could lead to the under- or overestimation of the growth coefficient and the asymptotic lengths of the populations. Therefore, any effort made to reduce the bias in the visualization, counting and analysis of the growth bands is an indirect but essential contribution to the management and subsequent conservation of elasmobranch species.
This study was financed by the Departamento Administrativo de Ciencia, Tecnología e Innovación - Colciencias (Contract No: RC 156-2010), PADI Foundation, and Iniciativa de Especies Amenazadas (Scholarship IEA-2012-04-12-18-55-26).
The following supplementary material is available through the online version of this article and at the following link:
Table S1. – Age and growth publications in tropical elasmobranchs (Spreadsheet in MS Excel format available at:
Table S2. – Results of precision and bias tests between readers, for each of the treatments applied to the small, medium and large vertebrae of Narcine leoparda. IAPE, index of average error percentage; CV, coefficient of variation; PA (±0 bands), percentage of total agreement; RV, percentage of read vertebrae; P, p-values of the Bowker symmetry test. Best results of each test for each vertebrae size are highlighted in bold and the selected treatments are shaded in gray.
Table S3. – Results of precision and bias tests between readers, for each of the treatments applied to the small, medium and large vertebrae of Urotrygon aspidura. IAPE, index of average error percentage; CV, coefficient of variation; PA (±0 bands), percentage of total agreement; RV, percentage of read vertebrae; P, p-values of the Bowker symmetry test. Best results of each test for each vertebrae size are highlighted in bold and the selected treatments are shaded in gray.
Table S4. – Results of precision and bias tests between readers, for each of the treatments applied to the small, medium and large vertebrae of Hypanus longus. IAPE, index of average error percentage; CV, coefficient of variation; PA (±0 bands), percentage of total agreement; RV, percentage of read vertebrae; P, p-values of the Bowker symmetry test. Best results of each test for each vertebrae size are highlighted in bold and the selected treatments are shaded in gray.
Table S5. – Results of precision and bias tests between readers for each of the treatments applied to Potamotrygon magdalenae, Alopias pelagicus and Carcharhinus falciformis vertebrae. IAPE, index of average error percentage; CV, coefficient of variation; PA (±0 bands), percentage of total agreement; RV, percentage of read vertebrae; P, p-values of the Bowker symmetry test. Best results of each test are highlighted in bold and the selected treatment for each species.is shaded in gray.
Table S6. – Results of precision and bias tests between readers for each of the treatments applied to Sphyrna lewini, Sphyrna corona and Mustelus lunulatus vertebrae. IAPE, index of average error percentage; CV, coefficient of variation; PA (±0 bands), percentage of total agreement; RV, percentage of read vertebrae; P, p-values of the Bowker symmetry test. Best results of each test are highlighted in bold and the selected treatment for each species.is shaded in gray.
Species comments.