#install and load necessary library packages to conduct data anlysis.
#install.packages("tidyverse")
#install.packages("readr")
#install.packages("knitr")
#install.packages("modeldata")
#install.packages("tibble")
#install.packages("dbplyr")
#install.packages("ggplot2")
#install.packages(readxl)
library(tidyverse)
library(readr)
library(modeldata)
library(tibble)
library(dbplyr)
library(readxl)Off the hook: Blotching in Carcharinus Perezii (Caribbean Reef Sharks)
Abstract
The research for indicators of stress in sharks is sparse and inconclusive, due to the limited methods of studying them. Current studies indicate that there is evidenced physiological and hematological changes to the Carcharinus genus, with most academia referencing the metamorphism after capture; irrespective of mitigation techniques. Nonetheless, there is a gap within the research that explores an additional indicator of stress. The phenomenon of ‘blotching’ that has been observed on the ventral surface of ‘reef sharks’, Carcharinus perezii, which are often situated in the Western Atlantic Ocean.
This research aims to explore a comprehensive data set of reef sharks, to examine the severity of blotching and welfare of the sharks. The method of data collection required the capture of reef sharks on two instances for a repeated measures design experiment. It can be argued that the method of data collection has the possibility to cause ethical discourse of the sharks’ welfare. The results show we can rule out there being a correlation between air and water that affects blotching. But there is a sound association between multiple capture and time taken for the ventral surface to blotch. As well as there being a possibility to predict blotching time. We are yet to develop equations to conclusively assess by how much of an increase there is and at which threshold this stops. The research on physiological and hematological changes post capture and mortality of Carribean Reef sharks has a long way to go in terms of exploration into the above, particularly due to the ethical guidelines in place, but it is necessary to ensure our understanding of these beings.
Keywords Elasmobranchii ~ Pelagic sharks ~ Anthropogenic stress ~ Ventral surface ~ Blotching ~ Fisheries
This review will build on the foundational concepts of shark science which form the limited research on Carcharhinus Perezi. Exploring how human induced stress caused by hook and line fishing for biological research, presents itself as blotching on the ventral surface of Caribbean reef sharks. Stress, in this instance will be defined as a ‘physiological response to a stressor’ (Parkhurst 2011), which is the concept that is frequently found in literature, irrespective of discipline. Physiology, being the functions of a shark. It should be noted that there are occurrences of stress being operationalized through behavior (Carrier, et al 2010). Nonetheless, stress can be divided into acute and chronic, instant exhibition or constant, respectively (Ibid). Although they can be distinguished from each other, it is not possible to separate the two. One cannot be certain on where to draw the line as acute stress has the potential to result in chronic presentation. For example, post capture stress causing changes to ‘reproductive physiology’ (Wheeler et al 2020). It can also be argued that the two can be occurring at the same time. There is considerably more literature on the hematological analysis of stress in sharks, where the endocrine is understood through biochemical analysis of plasma (Skomal and Bernal 2010 Wosnick et al 2017).
Methods and Ethics
Study Site
The Caribbean Sea, an oceanic basin on the Caribbean plate stretches between 9°and 22° degrees N and 89° and 60° W. With its deepest point in the Cayman Trench at approximately 7,600 meters, it is estimated to be up to 180 million years old.
Sampling
Convenience sampling of the Carcharhinus Perezi was deemed the most viable considering we did not want to hold the animals in a captive environment, as this would have contributed to the trauma experienced by the sharks. Therefore, it would have been an additional factor to consider in the methods, results and discussions. Using opportunistic sampling for this experiment allowed us to remove the causational argument of captivity increases blotching. Besides convenience and practicality, as we are focusing solely on the Caribbean reef shark, we can assume that the data collected is representative of the whole shark population. As the capture method of sharks requires a multidisciplinary team, I would like to acknowledge the chance of selection bias from the researchers. Sharks could have been overlooked if they were thought to have not been ‘ideal’. Thus, presenting a possibility that the population is not as representative as it could have been.
The method of capture was ‘hook and line’, which is a typical longline fishing method. Two sizes of hooks were used to capture the sharks. 5/0 hooks were used to catch sharks up to 165cm and 12/0 hooks were used for sharks above this length. The method was used to capture Caribbean reef sharks on two occasions- exactly 30 days after the first occasion. With there being limited data on mortality after sharks are released, and a large portion of them dying, it was necessary to monitor and allow the shark to recover after the first capture event. This was to allow the best chance of success for a second capture. As we set out to have to sets of data from the same sharks, it would allow for a repeated measures design. This experimental design allows each shark to be a control, thus reducing variability in the data. On the contrary, the order in which the shark experiences each of the test could influence the data collection. Therefore, it will be necessary that all sharks undergo the same testing procedures in the order of identification, determining sex, blotching, BPM (heart rate), weight and length and metabolic analysis. Then ambient air and water temperatures at the surface will be taken using a YSI meter. Depth will be estimated from the length of the line. With this design, there is the possibility that the sharks may respond in a way that shows they have adapted to the trauma and time taken for blotching on the ventral surface is decreased as they have experienced the stress repeatedly.
Table 1 Showing the variables which the sharks were tested in, providing a brief description, the data type and the units used to measure.
| Variable | Description | Data type | Units |
|---|---|---|---|
| Sex | Sex of individual shark, determined by the presence of Claspers | Categorical | Male/Female |
| Blotch | Time taken for blotching to cover 30% of ventral surface | Continuous | Seconds |
| BPM | Heart rate (via surgically implanted electrodes) | Continuous | Beats per minute |
| Weight | Total body weight - measured by hoisting the animal into a specialized sling | Continuous | kg |
| Length | Total body length, measured from tip if the snout to the tip of the upper lode of the tail fin. | Continuous | cm |
| Air | Ambient air temperature | Continuous | Celsius |
| Water | Water temperature at surface at time of processing | Continuous | Celsius |
| Meta | Measurement of stress hormone Cortisol via blood sample | Continuous | mcg/dl |
| Depth | Depth at which the animal was hooked (estimated from length of line) | Continuous | meters (m) |
Ethical principles were considered to ensure the welfare of the subjects, as their autonomy was removed, we aimed to minimize pain and distress. To decrease susceptibility to post release mortality, sea water was used as a lubricant to allow the sharks to maintain homeostasis, while the data (from Table 1) was collected. Every effort was made to remove the hooks from the sharks and ensure the wound was free from debris to discourage growth of infection. Though it is inevitable that the sharks would have encountered a degree of harm, it was justified by the necessity to create groundbreaking, vital data to shape exploration of blotching on the ventral surface of Caribbean Reef Sharks. A new phenomenon that had not yet been observed or researched.
Results
These results were formulated from the experiment that occurred in March and April. The first round of sampling (of 500 sharks) took place in the evening of March, as this is a month when the water is typically calmer, and visibility improved. The time being restricted to the night is necessary as this is when Caribbean Reef Sharks are most active. The second sampling event took place in April under the same conditions, however of the 500 sharks we only had success with collecting data for 50 sharks. One criticism of this may have been down to the choice of tag we used to identify the sharks. We had found the most cost-efficient method would be to use agricultural-style Dalton Roto tags on the dorsal fin. As these were able to display ID, sex, and the research project information. However, the lack of infrastructure for monitoring the whereabouts of these animals with acoustic tracking and receivers denotes the low recapture rate. Additionally, the low recapture rate is possibly down to shark mortality, with sharks having been caught as by-catch or from the initial experiment. The fact of the matter is regardless of this, in the space of 30 days, we are unable to review where the sharks had been and their location at the point of the second data collection. Thus, we were unable to collect repeated data for all sharks. The justification for this method of tag alongside economic efficiency, was externally collected secondary indicated that Caribbean reef sharks did not typically migrate outside of the area they were captured. But it is important to note that though they don’t experience seasonal migration, they are free to roam a large area of the ocean indicating that there is a possibility for temporal population changes.
This section is split into three questions, to explore the effects of blotching on Carcharhinus Perezi:
Is there a correlation between the variables air and water?
Does multiple capture influence blotching time?
Is it possible to predict blotching time?
Is there a correlation between the variables air and water?
Initially, the display of Fig 2. highlights there is no correlation between the ambient air temperature and the water temperature at the surface during the time of processing the sharks. The line of regression is almost parallel to the x-axis and there is no grouping surrounding it showing a lack of strength. The conflict of a linear regression on the graph is that it assumes a normal distribution, but as the data fails to show this the data points are scattered across the whole graph illustrating an absence of a pattern in the data. Therefore, visually, we can assume there is no influence on air by water or vice versa. However, there is the possibility of additional variables that we did not consider such as the El Niño–Southern Oscillation and typical local climate data.
library(readxl) #loads package for reading excel files
sharks <- read_excel("sharks.xlsx") #stores the data as sharks variable
View(sharks) #opens a seperate display of sharks data
str(sharks) #This can be used to identifty structure of data, which will identify the varibales the sharks are tested against.library(readxl)
sharksub <- read_excel("sharksub.xlsx")
View(sharksub)
str(sharks)
#these steps loads the data so it can be called during the data analysis Fig 2 A scatter plot displaying the correlation between air (x- axis) and water (y-axis) in °c
# ggplot2 is already lodaded for creating the plot
library(ggplot2)
ggplot(sharks, aes(x = air, y = water)) + #Create the scatter plot with air on x-axis and water on y-axis
geom_point() + # Add data points to the plot
geom_smooth(method = "lm", se = TRUE) + #Add linear regression line with confidence interval
labs(
x = 'air °c',
y = 'water °c',
title = "Scatter plot displaying the correlation between air (x- axis) and water (y-axis) in °c"
) +
theme(axis.text = element_text(size = 16), # Set axis text size
axis.title = element_text(size = 16)) # Set axis title size`geom_smooth()` using formula = 'y ~ x'
Fig 3 A quantile-quantile plot displaying sampled air temperature quantiles against theoretical quantiles.
#Create QQ plot for 'air'
ggplot(sharks, aes(sample = air)) + #uses 'air' from shark data to create a qq plot
stat_qq() + #adds data points to qq plot against theoretically normal values
stat_qq_line() + #adds reference line of 'normall' data
labs( #sets the labels
x = "Theoretical Quantiles", #defines axis
y = "Sample Quantiles °c",
title = "QQ Plot for Air Temperature"
) +
theme_bw() + #sets the theme
theme(
axis.text = element_text(size = 16),
axis.title = element_text(size = 16)
)
The visual assessment of Fig 3. highlights a S-shape, implying there is not a normal distribution of data. Additionally, to the skewness of the S-shape, there is a deviation in the tail highlight the comparison to the normal distribution of data.
Fig 4 A quantile-quantile plot displaying sampled water temperature quantile against the theoretical quantiles.
#Create QQ plot for 'water'using same steps as above changing the variable
ggplot(sharks, aes(sample = water)) +
stat_qq() +
stat_qq_line() +
labs(
x = "Theoretical Quantiles",
y = "Sample Quantiles °c",
title = "QQ Plot for Water Temperature"
) +
theme_bw() +
theme(
axis.text = element_text(size = 16),
axis.title = element_text(size = 16)
)
Similarly to Fig 3. the visual analysis of Fig 4. depicts a deviation from the normal distribution in the form of the S-shaped curve. There is also an increase in the deviation of the upper tail suggesting the data is considerably skewed to the right. Both Fig 3. and Fig 4. display the lack of normally distributed residuals. To allow a conclusive decision on whether there is a correlation between air and water, we used the Shapiro-Wilk Test for normality. As the null hypothesis concludes that the data is of a normal distribution
Fig 5 A display of the linear regression summary of water being a response to air
#linear regression expression
model <- lm(water ~ air, data = sharks)
summary(model) # this lists cooefficients, p value and r- sqaured
Call:
lm(formula = water ~ air, data = sharks)
Residuals:
Min 1Q Median 3Q Max
-3.03472 -1.47563 0.09925 1.38700 3.06356
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 25.31781 1.86221 13.596 <2e-16 ***
air -0.06465 0.05236 -1.235 0.218
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 1.67 on 498 degrees of freedom
Multiple R-squared: 0.003052, Adjusted R-squared: 0.00105
F-statistic: 1.524 on 1 and 498 DF, p-value: 0.2176Fig 6 Shapiro-Wilk test on water component
#Use the inear regression model
model <- lm(water ~ air, data = sharks)
residuals <- residuals(model) #Extract residuals from the data
shapiro_test_result <- shapiro.test(residuals) #Code for Shapiro-Wilk test
print(shapiro_test_result)
Shapiro-Wilk normality test
data: residuals
W = 0.96102, p-value = 3.078e-10Fig 7 Spearman’s rank coefficient
library(stats)
spearman_correlation <- cor(sharks$air, sharks$water, method = "spearman") #Code for Spearman's rank correlation
cat("Spearman's rank correlation coefficient:", spearman_correlation, "\n") #Print the resultSpearman's rank correlation coefficient: -0.05637344 Fig 5. and 6. display the linear regression and Shapiro-Wilk test of water being a response to air. Fig 6. is the residual values of such data the difference between actual and theoretical data, divided by standard deviation. Fig 6. shows W = 0.96102 signaling a normal distribution. As values closer to 1 typically indicates this. However, the p-value = 3.078e-10 which is less than 0.05. Providing sound evidence for the rejection of the null hypothesis. However, as noted there are deviations which may affect the validity strength of this decision. Therefore, the use of Spearman’s rank will assist in confirming the rejection of the null hypothesis. Fig 7. concludes the Spearman’s rank correlation coefficient to be -0.05637344 portraying an inverse monotonic relationship between air and water temperature, thus providing statistical evidence on the fact that there is no correlation between air and water. Thus, we can reject the null hypothesis.
Does multiple capture influence blotching time?
The null hypothesis for this is set out as: there will be no effect on blotching time when sharks are captured twice. Fig 8. visually presents a positive linear correlation, we can assume there is an influence. The strength of which is witnessed with the regression line being proportional to the data plots. It is key to note that although the data points are closely clustered across the regression line, there is a degree of variability with Blotch 1 data displaying outliers. Suggesting that other factors are at play in addition to multiple capture, requiring secondary investigation.
Fig 8 Scatter plot illustrating blotching time of Sharks and Sharksub data
# Create the scatter plot, using same steps as previous
ggplot(sharksub, aes(x = blotch1, y = blotch2)) +
geom_point() +
geom_smooth(method = "lm", se = TRUE) +
labs(
x = "Blotch 1",
y = "Blotch 2",
title = "Relationship between Blotch 1 and Blotch 2"
) +
theme_bw() `geom_smooth()` using formula = 'y ~ x'
Fig 9 Violin plot illustrating blotching time of Sharks and Sharksub data
sharksub_long <- pivot_longer(sharksub, cols = c(blotch1, blotch2),
names_to = "variable", values_to = "duration") #pivot_longer changes the data to make it easier to plot. This part of the code specifes data needed to be reshaped and creates a new coloum to store the above categories.
ggplot(sharksub_long, aes(x = variable, y = duration, fill = variable)) + #this creares the plot using the new data and decides the axis and variables will determine colour
geom_violin() + #this csreates the plot
geom_boxplot(width = 0.1) +
labs( #sets the labels
x = "Blotch",
y = "Duration (seconds)",
title = "Distribution of Blotch Durations"
) +
scale_fill_manual(values = c("blotch1" = "red", "blotch2" = "blue")) + #determines the colours
theme_bw() #creates a simple theme to allow readibility
The violin plot shows the distributions of blotch duration for Sharks and Sharksub data in Fig 9. Upon comparison blotch 1 appears somewhat more symmetrical blotch 2. Suggesting the durations were similar as opposed to blotch 2. In contrast, they both share the right leaning skew, which is evidence to support the notion that both sets of data contains outliers. With blotch 2 having an increased distribution of data for longer durations, highlighting the possibility of multiple capture increasing blotch time. Arguably, implying the sharks can cope with the stress of the event.
Fig 10 Blotch 1 statistics
blotch1_summary <- summary(sharksub$blotch1) #create a data frame
print(blotch1_summary) Min. 1st Qu. Median Mean 3rd Qu. Max.
32.49 34.38 34.94 35.03 35.90 37.07 blotch1_median <- median(sharksub$blotch1)
blotch1_sd <- sd(sharksub$blotch1)
blotch1_q1 <- quantile(sharksub$blotch1, 0.25)
blotch1_q3 <- quantile(sharksub$blotch1, 0.75) #This is how to calculate median, standard deviation and quantiles
cat("Median for blotch1:", blotch1_median, "\n")Median for blotch1: 34.93777 cat("Standard Deviation for blotch1:", blotch1_sd, "\n")Standard Deviation for blotch1: 1.095959 cat("First Quartile (Q1) for blotch1:", blotch1_q1, "\n")First Quartile (Q1) for blotch1: 34.38008 cat("Third Quartile (Q3) for blotch1:", blotch1_q3, "\n") #This displays the results Third Quartile (Q3) for blotch1: 35.8953 Fig 11 Blotch 2 statistics
#This code is created following the same sequence as aboove changing the category to show blotch2 data
blotch2_summary <- summary(sharksub$blotch2)
print(blotch2_summary) Min. 1st Qu. Median Mean 3rd Qu. Max.
33.47 35.31 35.94 35.96 36.78 38.18 blotch2_median <- median(sharksub$blotch2)
blotch2_sd <- sd(sharksub$blotch2)
blotch2_q1 <- quantile(sharksub$blotch2, 0.25)
blotch2_q3 <- quantile(sharksub$blotch2, 0.75)
cat("Median for blotch2:", blotch2_median, "\n")Median for blotch2: 35.94079 cat("Standard Deviation for blotch2:", blotch2_sd, "\n")Standard Deviation for blotch2: 1.16283 cat("First Quartile (Q1) for blotch2:", blotch2_q1, "\n")First Quartile (Q1) for blotch2: 35.3086 cat("Third Quartile (Q3) for blotch2:", blotch2_q3, "\n")Third Quartile (Q3) for blotch2: 36.77562 The basic statistics (Fig 10 and 11), of both data sets are relatively similar. The medians are similar with 34.94s and 35.94. highlighting the comparability of the data sets. The same is true for standard deviation, allowing us to use these conclusions to make assumptions over a generalized population.
Fig 12 Wilcoxon signed-rank test calculation for blotch 1 and 2
mw_test <- wilcox.test(sharksub$blotch1,sharksub$blotch2) #This calls the blotch times from blotch 1 and 2 of sharksub data set, to complete wilcoxon test
print(mw_test) #shows results
Wilcoxon rank sum test with continuity correction
data: sharksub$blotch1 and sharksub$blotch2
W = 732, p-value = 0.0003603
alternative hypothesis: true location shift is not equal to 0The Wilcoxon signed-rank test concluded that the p-value = 0.0003603 Fig 12., is less than 0.05, thus we will reject the null hypothesis. This non parametric test was useful for the paired data sets of sharksubs. Concluding that the data is statistically significant and there is sufficient evidence to suggest that multiple occurrences of capture can increase the time taken for blotching to cover 30% of the ventral surface.
Is it possible to predict blotching time?
As the evidence in Fig 8 and 9 suggests that blotch time is increased upon multiple capture and Fig 12 proves that the data is statistically significant. The linear regression of the scatterplot in Fig 8, may be the foundation for determining the prediction of blotching time. But the accuracy of this may vary as there are noted outliers within the distribution of results (Fig 12,13), therefore the findings may not be appropriate for other shark groups. There is the potential for other variables to be governing blotching time, which were not investigated during this study.
Fig 13 A box plot displaying the distribution of duration for blotching of sharksub data
# Create the box plo, using the same sequence as previous expressed.
sharksub_long <- pivot_longer(sharksub, cols = c(blotch1, blotch2),
names_to = "variable", values_to = "duration")
ggplot(sharksub_long, aes(x = variable, y = duration, fill = variable)) +
geom_boxplot() +
labs(
x = "Blotch",
y = "Duration (seconds)",
title = "Distribution of Blotch Durations"
) +
scale_fill_manual(values = c("blotch1" = "red", "blotch2" = "blue")) +
theme_bw() 
However, it goes without saying to provide a conclusive answer to this question there would have to be an increased sample size, along with a longitudinal study. This would allow for better generalization as well as strengthening the possibility of prediction, which may be able to go as far as creating bands for blotching time after so many capture events. The most effective way to complete this would be to attach acoustic tags to allow the tracking of the animals to record if they have died before the next part of the study, and potential what caused the death. It would also be useful for locating these animals to ensure sample size requirements are met. To conclude from the basis of this data it is possible to predict that blotching time increases as capture events increases. But this study is unable to provide the equation which determines how much of an increase there will be and at which point there is no longer an increase.
Discussion
It is generally accepted that there is a decline in the populations of elasmobranch sub class and in this instance the Carcharhinus Perezi (ICUN; Carlson et al 2021; Clementi et al 2021), despite foundation population data being unknown (Baum and Myers 2004; Musick et al 2011; Talwar et al 2022). Arguably, this is the result of their natural tendency to produce K selected offspring, combined with the unregulated fishing industry- which has seen exponential growth (Brooks et al 2011; Whitney et al 2021). Making conservation efforts pertinent to restore population projections. Beyond that, predicting the initial population can possibly lead to the forecasting of how ‘sharks and their relatives’ (Carrier, et al 2010) will respond to anthropogenic stress in their habitat (Skomal and Bernal 2010).
Typically, the stress responses in fish are arranged into ‘primary, secondary and tertiary’ (Marshall et al 2012). Occurring as a chain reaction, these can be defined as a modification of the pituitary secretion of corticosteroids and catecholamines. Resulting in the change of homeostasis and osmoregulation that can affect mass and lifespan (Kjartansson et al 1988; Mazeaud1977; Parkhurst 2011; Schreck and Tort 2016). Though this form of research is commonly conducted on teleosts’ such as salmon, it is applicable as sharks also present primary and secondary reactions to stress with metabolic changes. Therefore, we can assume that these reactions exhibited by teleost’ to a degree is the same as in sharks. The phenomenon of blotching may be present in other fish but has gone unnoticed as it was an unnamed indicator, thus this study will be a first of its kind influencing further investigation.
As benthopelagic sharks typically sit at the upper of food webs, they play a vital role in maintaining their respective marine ecosystems (Hammerschlag et al 2019). Therefore, the trophic relationship between Carcharhinus and teleost’s has the potential to alter the ecosystem. But it would not cause the degradation that pragmatic scenarios have suggested (Bond et al 2018). Thus, it is important we understand the adaptability of sharks to anthropogenic stress, considering they are frequently caught as bycatch, as their death has the potential to disrupt essential ecosystem environments. To achieve this, the infrastructure and technology available need to be developed to ensure conclusive data collection. As well as this there would need to be a cooperative effort from conservation and ecological consultancies to ensure the welfare of sharks and other fish that may exhibit blotching.
Reference lists
Baum, J., Myers, R., 2004. Shifting baselines and the decline of pelagic sharks in the Gulf of Mexico. Ecology letters, 7(2) 135-145 doi.org/10.1111/j.1461-0248.2003.00564.x
Bond M E., et al., 2018 The trophic ecology of Caribbean reef sharks (Carcharhinus perezi) relative to other large teleost predators on an isolated coral atoll. Marine Biology, 165 (67)
Brooks, E J., et al., 2011. Validating the use of baited remote underwater video surveys for assessing the diversity, distribution and abundance of sharks in the Bahamas. Endangered Species Research, 13(1), 231-243 DOI: 10.3354/esr00331
Carlson J., et al 2021. Carcharhinus Perezi [online] ICUN. Available at: https://www.iucnredlist.org/species/60217/3093780 [Accessed 18 December 2024 ]
Carrier J C., Musick J A., Heithaus M R., eds., 2010. Sharks and Their Relatives II : Biodiversity, Adaptive Physiology, and Conservation. 1st ed. Florida: Taylor and Francis Group
Clementi G., et al., 2021 Anthropogenic pressures on reef-associated sharks in jurisdictions with and without directed shark fishing. Marine Ecology Process series, 661:175-186. DOI.org/10.3354/meps13607
Darekk2., 2015. Caribbean Sea Gulf of Mexico shaded relief bathymetry land map Available via: Wikipedia Commons [Accessed 21 December 2024]. [Fig 1.].
Hammerschlag N., et al., 2019. Ecosystem Function and Services of Aquatic Predators in the Anthropocene. Trends in Ecology and Evolution 34(4), 369-383
Kjartansson H., et al., 1988. Effects of different stocking densities on physiological parameters and growth of adult Atlantic salmon (Salmo salar L.) reared in circular tanks. Aquaculture, 73(1-4), 261-274
Marshall H., et al., 2012. Hematological indicators of stress in longline-captured sharks. Comparative Biochemistry and Physiology, Part A, 162(2) 121-129
Mazeaud M M., et al 1977. Primary and Secondary Effects of Stress in Fish: Some New Data with a General Review. Transactions of the American Fisheries Society. 106(3), 201-212 DOI.org/10.1577/1548-8659(1977)106<201:PASEOS>2.0.CO;2
Musick, J A., et al., 2011. Management of sharks and their relatives (Elasmobran-chii). Fisheries, 25(3) 9-13 DOI.org/10.1577/1548-8446(2000)025
Parkhurst N W,. 2011. The endocrinology of stress in fish: An environmental perspective. General and Comparative Endocrinology. 170(2), 265-275
Schreck C B., Tort L., 2016. 1 - The Concept of Stress in Fish. Fish Physiology, 35(1) 1-34
Skomal, G., Bernal D., 2010. Physiological Responses to Stress in Sharks. In : Carrier J C., Musick J A., Heithaus M R., eds. Sharks and Their Relatives II : Biodiversity, Adaptive Physiology, and Conservation, 2010, pp 459 - 491
Talwar, B S., et al., 2022. Estimated life-history traits and movements of the Caribbean reef shark (Carcharhinus perezi) in The Bahamas based on tag-recapture data. Marine Biology, 169(5), 2 -19
Wheeler, C R., et al., 2020. Anthropogenic stressors influence reproduction and development in elasmobranch fishes. Reviews in Fish Biology and Fisheries, 30(2), 373-386. DOI:10.1007/s11160-020-09604-0
Whitney N M ., et al., 2021. Connecting post-release mortality to the physiological stress response of large coastal sharks in a commercial longline fishery. PLoS One, 16(9), 1-23 DOI:10.1371/journal.pone.0255673
Wosnick, N., et al ., 2017. Talking to the dead: using Post-mortem data in the assessment of stress in tiger sharks (Galeocerdo cuvier) (Pe´ron and Lesueur, 1822). Fish Physiology and Biochemistry, 43(1), 165-178 DOI:10.1007/s10695-016-0276-5
Supplementary Information
body { font-family: "Times New Roman" }
@page {
@bottom-right {
content: counter(page);
}
}