Practical: Investigating Biodiversity Using Sampling
Random Sampling: A grid is established using tape measures, and coordinates are selected via a random number generator. A quadrat (square frame) is placed at these coordinates to record data.
Systematic Sampling: Used when there is a clear environmental gradient (e.g., moving from a forest into a field). A transect (line) is laid out, and samples are taken at regular intervals along it.
Line Transect: A string or tape where every organism touching the line is recorded; this is useful for visualizing changes in species distribution.
Belt Transect: Quadrats are placed at regular intervals along the transect line to provide more detailed data on both distribution and abundance.
| Feature | Random Sampling | Systematic Sampling |
|---|---|---|
| Primary Goal | Estimate total population/biodiversity | Study changes along a gradient |
| Site Selection | Random number generator | Fixed intervals (e.g., every 5m) |
| Best Use Case | Uniform habitats (e.g., open field) | Zonation (e.g., rocky shore) |
| Bias Risk | Lowest (removes human choice) | Higher (may miss periodic patterns) |
Scaling Up: To estimate total population, calculate the mean number of individuals per quadrat and multiply by the total area divided by the quadrat area: .
Consistency: Always specify that the same size quadrat must be used for all samples to ensure the data is comparable and valid.
Justification: If asked why random sampling is used, always mention 'to avoid bias' and 'to ensure the sample is representative'.
Verification: Check if the sample size is sufficient by plotting a running mean; when the mean levels off, enough samples have likely been taken.
Subjective Placement: Students often place quadrats where they see 'more plants'. This is biased and leads to overestimation of biodiversity.
Edge Effects: Organisms partially inside a quadrat can lead to counting errors. A standard rule (e.g., 'only count if touching the top or right edge') must be applied consistently.
Misinterpreting Richness: High species richness does not always mean high biodiversity if one species is dominant (low evenness). Always consider both metrics.
Inappropriate Scale: Using a 10cm quadrat to sample a forest or a 1m quadrat to sample microscopic soil organisms will yield inaccurate and unmanageable data.