Activity: Sea Otter Foraging Behavior Study

Background:
Sea otters play a crucial role in marine ecosystems, especially in maintaining the health of kelp forests. Their foraging behavior can provide insights into the overall health and balance of the marine environment.

Objective:
The primary objective of this study is to understand the variability in foraging success of sea otters across different coastal regions and within specific coves of these regions. As part of this, we account for the random effect introduced by replicate measurements taken from 10 otters across the regions and coves.

Dataset Overview:
The dataset provides information on the foraging success of sea otters across three main coastal regions: North, Central, and South.

Observations were made in 9 distinct coves spread across the three regions: Flufftail Inlet, Seagrass Haven, Shellshine Shore, Kelpie Nook, Otterly Cove, Whisker Bay, Paddlepaw Bay, Seaweed Serenade, and Splashsnout Strand.

Data was collected for 10 individual sea otters: Brook, Drizzle, Finn, Luna, Marina, Paddle, Ripple, Sway, Tidbit and Whisk.

The mesaured variable, Foraging Success, reflected the number of prey items (e.g., sea urchins, crabs) captured during a specified observation period, normalized to day (units = prey items/day).

Steps

i. Download the data

Follow the link to Download the sea otter foraging dataset.

ii. Set up your script

As always, initiate your R script with the necessary libraries and set your working directory appropriately.

1. Import and Describe the Dataset

Useful commands: read.csv(), head(), summary()

also, set your factors of YOURDATA, and re-level for plotting reasons (optional code included here, but feel free to change/reorder).

YOURDATA$Otter_ID <- 
  as.factor(YOURDATA$Otter_ID)

YOURDATA$Region <-
  factor(
    YOURDATA$Region, 
    levels = c(
      "North", 
      "Central", 
      "South"
      )
    )

YOURDATA$Cove <-
  factor(
    YOURDATA$Cove,
    levels = c(
      "Flufftail Inlet",
      "Seagrass Haven",
      "Shellshine Shore",
      "Kelpie Nook",
      "Otterly Cove",
      "Whisker Bay",
      "Paddlepaw Bay",
      "Seaweed Serenade",
      "Splashsnout Strand"
    )
  )

2. Visualize the Data

Create a plot to visualize the foraging success across different regions and coves.

3. Initial Observations and Discussion

Based on your plot, how might foraging vary across coastal regions? And how might conditions vary within a region (between coves)? No need to discuss statistical significance here - just observe the plot. This is good practice.

4. Perform the Nested ANOVA in R

Run the nested ANOVA model considering coves nested within regions.

Based on this output, are you able to disregard the effects of cove and perform a simple one-way anova of foraging success across regions? Why? Be specific, report significance levels

5. Perform ANOVA with Random Effects

Generate ANOVA models with and without random effectof Otter.

model_otterID <-
  aov(DEPENDENT_VARIABLE ~ INDEPENDENT_VARIABLE +
        Error(RANDOM_VARIABLE),
      data = YOURDATA)

model_noOtterID <-
  aov(DEPENDENT_VARIABLE ~ INDEPENDENT_VARIABLE, 
      data = YOURDATA)

Compare the F values of Region effects in the model without the random effect (model_noOtterID) and model with the random effect (model_otterID). Which F-Value is larger, and what does this reveal about the importance of accounting for nested or random effects in data? If we were to only consider the model without the Otter_ID random effect (model_noOtterID), what potential misconceptions might arise regarding the effects of Region on foraging success?

If you are having trouble answering these questions, check out the Random effects section, in addition to other resources. I personally recommend the fantastic tutorials by Bodo Winter. They use the lme4 package, but the concepts are the same as random effects applied to aov() models.

6. Interpret and Discuss

Provide a short paragraph explaining and interpreting the results. Consider both the statistical findings and their broader ecological implications. This paragraph should included mention of:
- Which sources of variability were important and which were not?
- Interpret the graph you made above, and the results of the ANOVAs, discuss which regions or coves stood out in terms of foraging success and discuss potential environmental or anthropogenic reasons for variability in otter foraging success in these regions
- any other points/observations that seem relevant