Sharks Website

Quarto Formatting & Libraries

For this analysis, we will be using the “sharks” and “sharksub” CSV files acquired from NOW.

In this chunk of code we will ensure the tidyverse and corrplot packages are loaded.

Code

library(tidyverse)

── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.1.4     ✔ readr     2.1.5
✔ forcats   1.0.0     ✔ stringr   1.5.1
✔ ggplot2   3.5.1     ✔ tibble    3.2.1
✔ lubridate 1.9.3     ✔ tidyr     1.3.1
✔ purrr     1.0.2     
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

Code

library(corrplot)

corrplot 0.95 loaded

Now we will load the ggplot library:

Code

library(ggplot2)

Let’s make sure the sharks and sharksub datasets are loaded and in the correct place:

Code

# Get working directory
getwd()

[1] "/Users/becca/Desktop/Classes/Research Methods/Research Methods Summative"

Code

# Check the current working directory
getwd()

[1] "/Users/becca/Desktop/Classes/Research Methods/Research Methods Summative"

Code

# List files in the working directory
list.files()

[1] "journal articles"         "Shark Journal.docx"      
[3] "sharks.csv"               "sharksub.csv"            
[5] "SharksWebsite.qmd"        "SharksWebsite2_files"    
[7] "SharksWebsite2.html"      "SharksWebsite2.qmd"      
[9] "SharksWebsite2.rmarkdown"

Code

# Check if the file exists in the working directory
file.exists("sharks.csv")

[1] TRUE

Code

#Set the working directory
setwd("/Users/becca/Desktop/Classes/Research Methods/Research Methods Summative")
sharks <- read.csv("sharks.csv")

Code

# Check the current working directory
getwd()

[1] "/Users/becca/Desktop/Classes/Research Methods/Research Methods Summative"

Code

# List files in the working directory
list.files()

[1] "journal articles"         "Shark Journal.docx"      
[3] "sharks.csv"               "sharksub.csv"            
[5] "SharksWebsite.qmd"        "SharksWebsite2_files"    
[7] "SharksWebsite2.html"      "SharksWebsite2.qmd"      
[9] "SharksWebsite2.rmarkdown"

Code

# Check if the file exists in the working directory
file.exists("sharksub.csv")

[1] TRUE

Code

#Set the working directory
setwd("/Users/becca/Desktop/Classes/Research Methods/Research Methods Summative")
sharksub <- read.csv("sharksub.csv")

Now we will view the “sharks” and “sharksub” data, which opened up in a separate tab.

Code

view (sharks)
view (sharksub)

Data Visualisation

Sharks (By Sex)

Now that we have confirmed that both the shark and sharksub datasets are working properly, let’s begin to visualise the data. First we will do a ggplot of male and female sharks plotted by weight and length, organized by color and shape.

Code

ggplot(sharks,
       aes(x = length, y = weight)) +
  geom_point(aes(color = sex, shape = sex)) +
  scale_color_manual(values = c("purple","cyan4")) +
  labs(
    title = "Shark Comparison: Sex by Weight",
    
    x = "Length (cm)", y = "Weight (kg)",
    color = " ",
    shape = " ", 
    fill = " "
  ) +
 theme(
    axis.text = element_text(size = 10),
    axis.title = element_text(size = 10),
    plot.title = element_text(size = 12, face = "bold", hjust = 0.5))

Air & Water Correlation

We can see from the aforementioned graphic that there is no correlation between length and weight of females vs. males. Let’s perform the same action but for correlations between air and water.

Code

sharks %>%
  ggplot(aes(
    x = water, 
    y = air,
    shape = sex,
    color = sex,
    fill = sex
  )) +
  geom_point() +
  labs(
    title = "Correlation: Air & Water",
    x = "Water (Celsius)", 
    y = "Air (Celsius)",
    color = " ",
    shape = " ", 
    fill = " "  
  ) +
  theme(
    axis.text = element_text(size = 10),
    axis.title = element_text(size = 10),
    plot.title = element_text(size = 12, face = "bold", hjust = 0.5)
  )

There does not appear to be a correlation here. Let’s try to change the x and y axes.

Code

sharks %>%
  ggplot(aes(
    x = water, 
    y = air,
    shape = sex,
    color = sex,
    fill = sex
  )) +
  geom_point() +
  labs(
    title = "Correlation: Air & Water",
    x = "Air (Celsius)", 
    y = "Water (Celsius)",
    color = " ",
    shape = " ", 
    fill = " "  
  ) +
  theme(
    axis.text = element_text(size = 10),
    axis.title = element_text(size = 10),
    plot.title = element_text(size = 12, face = "bold", hjust = 0.5)
  )

Air Temperature & Catch Number

Still nothing. This may not be the correct type of graph, so let’s try something different by way of a histogram.

Code

data("sharks")

Warning in data("sharks"): data set 'sharks' not found

Code

sharks %>%
  group_by(sex) %>% 
  ggplot(aes(x = air, color = sex, fill = sex)) +
  geom_histogram() +
  labs(
    title = "Air Temperature by Sex",     
    x = "Air Temperature (°C)",                            # X-axis label
    y = "Count",                                       # Y-axis label
    color = "Sex",                                   # Legend title for color
    fill = "Sex"                                     # Legend title for fill
  ) +
  theme(
    axis.text = element_text(size = 10),
    axis.title = element_text(size = 10),
    plot.title = element_text(size = 12, face = "bold", hjust = 0.5)
  )

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

At this point, I believe we can safely say there does not appear to be a relevant correlation between air temperature and number of catches.

Blotching Correlation

Let’s move into blotching. First let’s see if there is a correlation between male and female blotching.

Code

ggplot(sharks, aes(x = sex, y = blotch, fill = sex)) +
  geom_boxplot() +
  labs(
    title = "Comparison of Blotching Time by Sex",
    x = "Shark Sex",
    y = "Blotching Time (seconds)"
  ) +
  theme(
    axis.text = element_text(size = 10),
    axis.title = element_text(size = 10),
    plot.title = element_text(size = 12, face = "bold", hjust = 0.5)
  ) +
theme_minimal() +
  scale_fill_manual(values = c("purple", "cyan4"))

Review of the data shows that males begin blotching slightly after females, but the difference is not substantial. Let’s look at the same data for males and females caught twice.

Code

ggplot(sharksub, aes(x = blotch1, y = blotch2, fill = sex)) +
  geom_boxplot() +
  labs(
    title = "Comparison of Blotching Time by Sex",
    x = "Sex",
    y = "Blotching Time (Seconds)"
  ) +
  theme_minimal() +
  scale_fill_manual(values = c("purple", "cyan4")) +
  theme(
    plot.title = element_text(face = "bold", hjust = 0.5)
  )

This chart doesn’t look quite right, so let’s try a different variation. Here, we will utilise columns for the first and second blotch times to better visualise the data. co

Code

library(tidyr)

sharks_long <- sharksub %>%
  pivot_longer(cols = c(blotch1, blotch2), names_to = "catch_number", values_to = "blotching_time")

ggplot(sharks_long, aes(x = catch_number, y = blotching_time, fill = sex)) +
  geom_boxplot() +
  labs(
    title = "Blotch Time by Catch Number and Sex",
    x = "Catch Number",
    y = "Blotch Time (seconds)"
  ) +
  theme_minimal() +
  scale_fill_manual(values = c("purple", "cyan4")) + 
  theme(
    plot.title = element_text(face = "bold", hjust = 0.5)
  )

Review of this data shows that for both males and females, blotch time increases by about one second with the second catch.

Let’s see if we can find any other data correlations before we continue much further.

Code

# Select only numeric columns for correlation test
shark_numeric <- sharks[, sapply(sharks, is.numeric)]

Code

# Compute the correlation matrix
cor_matrix <- cor(shark_numeric, method = "pearson")

Code

# Visualize the correlation matrix
corrplot(cor_matrix, 
         method = "circle",
         title = "Correlation Matrix", 
         mar = c(2, 2, 2, 2) 
         )

Blotch Time & Depth Correlation

It appears from the aforementioned graph that there is a correlation between depth and blotch time. We will now create a visual image of this to confirm, using a scatterplot comparing the depth to the blotch time and differentiating sex:

Code

ggplot(sharks, aes(x = blotch, y = depth, color = sex)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE, color = "red") +
  labs(
    
    x = "Blotch Time (Seconds)",  # X-axis label with units
    y = "Depth (Metres)",        # Y-axis label with units
    color = "Sex", 
    title = "Linear Regression"
  ) +
  theme_minimal()+
  theme(
    plot.title = element_text(face = "bold", hjust = 0.5)
  )

`geom_smooth()` using formula = 'y ~ x'

Blotch Predictors

Because we can clearly see a linear correlation between depth and blotch time, in order to predict bloch time we should follow a linear model.

Let’s try a simple linear model:

Code

model_simple <- lm(blotch ~ depth, data = sharks)
summary(model_simple)


Call:
lm(formula = blotch ~ depth, data = sharks)

Residuals:
     Min       1Q   Median       3Q      Max 
-2.81869 -0.65427 -0.01035  0.58825  2.83116 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept)  9.82178    1.11207   8.832   <2e-16 ***
depth        0.50467    0.02216  22.772   <2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 1 on 498 degrees of freedom
Multiple R-squared:  0.5101,    Adjusted R-squared:  0.5091 
F-statistic: 518.6 on 1 and 498 DF,  p-value: < 2.2e-16

Review of the information tells us the following: 1. The depth coeficcients tell us that for every 1 metre deeper a shark is caught from, blotching time increases by ~.50467 seconds. In other words, the deeper a shark is caught from, the longer it takes to blotch.

Because the p value is a statistically significant negative value, this tells us that it is very likely that the correlation between blotch time and depth is not by chance, which we saw in our linear model above, as well.
Residual values tell us that most predictions are within 2.8 units of the actual values.
Based on this model, approximately 51% of blotch time is explained, which is high for ecological and biological studies. However, because 49% is still unexplained, it is possible that depth combined with other factors such as stress, water/air temperature, or sex may improve the ability to predict blotch time.

As such, let’s add in additional variables to determine if we can further predict blotch time:

Code

model_multiple <- lm(blotch ~ depth + meta + air + water + weight + length + sex, data = sharks)
summary(model_multiple)


Call:
lm(formula = blotch ~ depth + meta + air + water + weight + length + 
    sex, data = sharks)

Residuals:
     Min       1Q   Median       3Q      Max 
-3.01869 -0.65130 -0.00238  0.62722  2.91144 

Coefficients:
              Estimate Std. Error t value Pr(>|t|)    
(Intercept) 10.7578141  1.7925782   6.001  3.8e-09 ***
depth        0.5036788  0.0220705  22.821  < 2e-16 ***
meta        -0.0011650  0.0025656  -0.454 0.649972    
air         -0.0295655  0.0314362  -0.940 0.347427    
water       -0.0146552  0.0267927  -0.547 0.584638    
weight       0.0017049  0.0033122   0.515 0.606964    
length       0.0013489  0.0009576   1.409 0.159577    
sexMale      0.3082889  0.0890047   3.464 0.000579 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.9906 on 492 degrees of freedom
Multiple R-squared:  0.5252,    Adjusted R-squared:  0.5184 
F-statistic: 77.73 on 7 and 492 DF,  p-value: < 2.2e-16

These results give us a bit more information, as follows:

Male sharks have a blotch time that is .31 seconds longer than female sharks (this was also seen in the box plots comparing male and female blotch time).
None of the other pedictors are statistically significant and therefore it can be assumed that there are not any other correlating factors in predicting blotch time.

Let’s do one last scatter plot to show these relationships (depth, blotch time, and sex):

Code

ggplot(sharks, aes(x = depth, y = blotch, color = sex)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE) +
  labs(title = "Blotch Time vs Depth by Sex",
       x = "Depth (Metres)",
       y = "Blotch Time (Seconds)",
       color = "Sex") +
  theme_minimal()

`geom_smooth()` using formula = 'y ~ x'

We can now say the most significant identifers in this review are the following: 1. There is no correlation between weight, length, and sex.
2. There is no correlation between air and water. 3. There is no correlation between air temperature and number of catches. 4. There are no other correlations aside from depth and blotch. 5. Males begin blotching slightly later than females. 6. Multiple capture does have an effect on blotching time, which increases by about one second with the second catch. 7. It is possible to predict blotching time.