Welcome to Sample Reliability and Capture-Recapture!
Have you ever wondered how scientists count all the fish in a huge lake, or how ecologists know how many wild tigers are left in a forest? They certainly don't line them up one by one! Instead, they use a clever statistical technique known as the Petersen capture-recapture method.
In this chapter, we will look at how to estimate the size of a hidden population, how to calculate it step-by-step, and how to judge whether our results are actually reliable. Don't worry if this seems tricky at first—we will break it down into simple, manageable bites!
---Part 1: Understanding Sample Reliability
Before we look at wild animals, let's think about reliability in general statistics. When we take a sample from a large group (a population), we want our findings to be trustworthy and repeatable.
What Makes a Sample Reliable?
A sample is considered reliable if repeating the process under the same conditions would give similar, consistent results. Here are the key factors:
• Sample Size: Larger samples generally give more reliable results because they reduce the effect of random flukes or anomalies.
• Unbiased Selection: Every member of the population should have an equal chance of being selected.
• Randomness: Using random sampling methods avoids human bias and ensures the sample represents the whole group.
Analogy: Tasting Soup
Imagine tasting a pot of vegetable soup. If you take a tiny drip from the top without stirring, you won't get a true sense of the flavour. But if you stir the pot well (random mixing) and take a proper spoonful (adequate sample size), your taste test is reliable!
Key Takeaway: Large, well-mixed, and unbiased samples produce the most reliable statistical estimates.
---Part 2: The Petersen Capture-Recapture Method
The Petersen method (also called mark-release-recapture) is a famous statistical tool used to estimate the total size of a mobile population, such as birds, fish, or insects.
How Does It Work? (The 4 Steps)
Step 1: Capture and Mark — Catch a sample of animals, count them, and carefully mark each one.
Step 2: Release — Release the marked animals back into their habitat and give them enough time to mix thoroughly with the rest of the population.
Step 3: Recapture — Catch a second sample of animals from the same population.
Step 4: Count — Count the total number in this second sample, and count how many of them have marks!
The Formula
The idea is based on simple proportions. We assume that the proportion of marked animals in our second sample is equal to the proportion of marked animals in the entire population:
\( \frac{\text{Number marked in 2nd sample (Recaptured)}}{\text{Total number in 2nd sample}} = \frac{\text{Number marked in 1st sample}}{\text{Total Population Size}} \)
Using letters to make it easier:
• \( N \) = Estimated total population size (what we want to find)
• \( M \) = Number of animals caught and Marked in the 1st sample
• \( n \) = Total number of animals caught in the 2nd sample
• \( R \) = Number of marked animals Recaptured in the 2nd sample
The formula is written as:
\( \frac{R}{n} = \frac{M}{N} \)
Rearranging to find \( N \):
\( N = \frac{M \times n}{R} \)
Memory Aid: "Multiply the catches, divide by the matches!"
Take the total caught the first time (\( M \)), multiply by the total caught the second time (\( n \)), and divide by the number that have a matching mark (\( R \)).
---Part 3: Step-by-Step Worked Example
Example: A marine biologist wants to estimate the number of trout in a lake.
• On Monday, she catches \( 60 \) trout, marks them with harmless tags, and releases them back into the lake.
• On Friday, she returns and catches \( 40 \) trout. She finds that \( 8 \) of these trout have tags.
• Estimate the total trout population in the lake.
Solution:
Step 1: Identify the values
• \( M = 60 \) (marked in 1st catch)
• \( n = 40 \) (total in 2nd catch)
• \( R = 8 \) (recaptured with tags)
Step 2: Apply the formula
\( N = \frac{M \times n}{R} \)
\( N = \frac{60 \times 40}{8} \)
Step 3: Calculate
\( N = \frac{2400}{8} = 300 \)
Conclusion: The estimated population of trout in the lake is \( 300 \).
---Part 4: Assumptions and Limitations
In your exam, you will often be asked to discuss the validity or reliability of an estimate. The Petersen method is only accurate if several key assumptions hold true!
Key Assumptions:
1. Population is Closed: There are no births, deaths, immigration (animals moving in), or emigration (animals leaving) between the two catches.
2. Marks are Permanent and Safe: The marks do not fall off, wash away, or harm the animal.
3. Equal Chance of Capture: Being marked does not make an animal easier for predators to spot, nor does it make them more "trap-shy" (avoiding traps) or "trap-happy" (seeking food in traps).
4. Sufficient Mixing Time: The marked animals have had enough time to spread out randomly among the rest of the population.
Did You Know?
If tags fall off the animals between catches, the value of \( R \) will be too small. Looking at our formula \( N = \frac{M \times n}{R} \), dividing by a smaller number makes \( N \) much bigger—meaning you will overestimate the population!
---Part 5: Common Mistakes to Avoid
• Mixing up \( n \) and \( R \): Remember that \( n \) is the total number caught in the second sample, while \( R \) is only the marked ones inside that second sample. Therefore, \( R \) can never be bigger than \( n \).
• Forgetting to round appropriately: You cannot have half an animal! If your answer gives a decimal like \( 245.6 \), round to the nearest whole integer (e.g. \( 246 \)).
• Assuming it gives an exact answer: Capture-recapture produces an estimate, not an exact count.
Quick Review Summary
• Petersen Formula: \( N = \frac{M \times n}{R} \)
• Key terms: \( M \) = 1st catch marked, \( n \) = 2nd catch total, \( R \) = recaptured with marks, \( N \) = total population estimate.
• Crucial requirements: Adequate mixing time, no changes to the population (births/deaths), and marks that stay on without affecting behaviour.