# Sampling to Determine Biodiversity ## Introduction to sampling - Measuring biodiversity within an ecosystem can be challenging - In large and complex ecosystems it is simply impossible to find, identify and count every organism - **Sampling** can be used to make an **estimate** of the biodiversity in an area - Sampling involves measuring biodiversity in **multiple small areas** within a habitat and using this information to **represent the habitat as a whole**

Sampling is the process of measuring biodiversity in small portions of a habitat to estimate the biodiversity of the entire habitat

- Sampling can be used to measure: - **distribution**: where organisms live within a habitat - **abundance**: how many organisms are present - In order to be truly representative, a sample must be **large enough** to eliminate the effects of chance - Missing a species simply because it didn't happen to occur at the sample sites assessed - **The more samples collected, the more likely the data will be representative** of the habitat ## Types of sampling - Sampling methods can be: - Random - Non-random ## Random sampling - In random sampling the sampling sites are selected **at random** - This avoids **bias** that might result from a researcher choosing the sites to assess - E.g. unconsciously choosing areas that are easier to access or that look more interesting

Random sampling prevents researcher bias by using chance to select sampling locations

### How to carry out random sampling 1. A **grid system** is set up by laying out two long tape measures along the outer edges of a habitat 2. A **random number generator** is used to generate two random numbers 3. The numbers are used like **graph coordinates**, and the sample site is placed at the point of intersection 4. This process is repeated to generate multiple sample sites ### When to use random sampling - When a sampling area is **reasonably uniform** or has **no clear pattern** of species distribution, then random sampling is the best choice - For example, sampling plant species across a uniform grassland meadow ## Non-random sampling - Random sampling is not always possible, or may not yield the type of results required - In these situations **non-random sampling** may be more appropriate - There are three main types of non-random sampling: ### 1. Opportunistic sampling - Sampling is carried out on the basis of **opportunity** - Sample sites that can be reached **quickly and safely** are chosen - For example, choosing to sample areas close to a car park rather than hiking deep into a forest - This is the least rigorous method but may be necessary for practical reasons ### 2. Stratified sampling - The number of random samples taken within each habitat type is **proportional to the area covered** by each habitat type - For example, if 10% of a dense woodland contains grassy clearings then: - 90% of the sample locations should have tree cover - 10% will be in clear areas - This ensures all habitat types are fairly represented in the sample

Stratified sampling allocates sample sites proportionally across different habitat types within the study area

### 3. Systematic sampling - Samples are taken **across a habitat** with reference to the **changes in habitat conditions** - Sample sites may be taken at: - Increasing altitude (e.g. up a mountain slope) - Increasing distance from the edge of a woodland - Increasing distance from the shore line - Systematic sampling involves the use of **transects** (covered in detail below) ## Sampling methods and equipment ### Quadrat sampling - A quadrat is a **square frame** that is placed within the area to be studied to provide a sample - Quadrats are suitable for sampling **plants** or **slow-moving animals** - For example, herbaceous plants in a grassland or limpets on a rocky shore

Quadrats work best for organisms that don't move away when you approach them

#### Quadrat sizes - Quadrats can be **different sizes** depending on the species being studied - A **1 m² quadrat** can be used to study small organisms such as: - Herbaceous plants in a grassland - Barnacles on a rocky shore - A **400 m² quadrat** can be used to study large organisms such as trees - Quadrats like this will usually be marked out with string rather than a frame! #### Quadrat placement - Quadrats can be placed in a habitat: - **Randomly** (using random number generators as described above) - Along a **transect** (for systematic sampling) #### Measurements within quadrats Measurements within a quadrat might include: | Measurement type | Description | When to use | |-----------------|-------------|-------------| | Number of species present | Count how many different species are found | Assessing species richness | | Number of individuals | Count how many individuals of each species are present (species abundance) | When individuals can be easily distinguished | | Presence/absence | Simply record whether a species is present or absent | Quick survey method | | Percentage cover | Estimate the percentage of the quadrat area covered by each species | When it's difficult to count individual organisms (e.g. grass, moss) | #### Percentage cover technique - Usually used for **plants** where individuals are hard to distinguish - Place a **grid quadrat** (a square frame split into 100 smaller squares) at the sampling location - Each square of a grid quadrat is equivalent to **1% cover** - Count the number of squares in which the species occupies **over half the square** - If 30 squares contain the species, the percentage cover is 30%

Percentage cover estimation is subjective; the same person should make all estimates to control this variable

### Sweeping nets - Sweeping (or sweep) nets are **large, strong nets** made from a material with very small holes - The nets can be swept across vegetation to catch: - Flying insects - Insects that live on leaves - For example, sweeping across tall grass or through tree branches to catch butterflies, moths, and beetles - After sampling, the insects in the net can be counted and identified ### Pitfall traps - Pitfall traps are cans or jars that are **sunk into the ground** - **Ground-dwelling invertebrates** fall into the trap as they walk along the ground, and then cannot climb out again - For example, beetles, spiders, and woodlice - A **lid** ensures that the trap does not fill with rainwater - The lid should be raised slightly above the trap to allow animals to fall in - Newspaper or leaf litter can be placed in the trap to allow smaller insects to hide from larger predatory insects

Check pitfall traps regularly to minimize harm to trapped organisms and to prevent them from eating each other

### Pooters - Pooters are small plastic or glass containers with **two tubes** extending from the lid - The tubes can be used to **suck up small invertebrates** - The first tube is placed over the insect - The second tube is used by the scientist to create suction by breathing in - A piece of gauze or mesh prevents the insects from being sucked into the scientist's mouth! - Once inside the container, the sampled organisms can be easily **viewed and identified** - For example, collecting tiny spiders, aphids, or ants for closer examination ### Transects - A transect is a **line along which samples can be taken** - They are used during **systematic sampling** to assess how abundance and distribution of organisms is affected by changes in **abiotic factors** across a habitat - Transect lines are laid out across a habitat along the **gradient** that is to be investigated

A transect is a sampling line used to study how species distribution changes across an environmental gradient

#### When to use transects Examples of where transects might be used: - At right angles to the edge of a field (to study edge effects) - Vertically up a slope (to study effects of altitude) - At right angles to a water course (to study effects of distance from water) - From a beach into the sea (to study effects of water depth) #### Types of transect **Line transect:** 1. Lay out a measuring tape in a straight line across the sample area 2. At **regular distances** along the tape (e.g. every 1 m), record the identity of the organisms that **touch the line** 3. This gives a quick snapshot of species distribution **Belt transect:** 1. Lay out a measuring tape in a straight line across the sample area 2. Place quadrats at **regular intervals**, or **continuously**, along the tape 3. Record the **abundance or percentage cover** of each species within each quadrat 4. This provides more detailed information than a line transect ## Comparison of sampling methods | Sampling method | Best used for | Advantages | Limitations | |----------------|---------------|------------|-------------| | Random quadrats | Uniform habitats with no clear patterns | Avoids bias; statistically valid | May miss rare species; time-consuming | | Systematic transects | Studying changes across environmental gradients | Shows clear patterns of change | May miss variation away from the transect line | | Opportunistic | Quick surveys or difficult terrain | Fast; practical | Highly biased; not statistically valid | | Stratified | Habitats with distinct zones | Ensures all habitat types sampled fairly | Requires prior knowledge of habitat structure | ## Practical considerations ### Sample size - Use a **running mean** to determine the number of samples required - Calculate the mean for the first two samples - Then calculate the mean of the first three, then four, and so on - Continue until there is **no further significant change** in the mean - This indicates you have taken enough samples **Example: Using running mean to determine adequate sample size** Imagine sampling a grassland and counting the number of different plant species in each 1 m² quadrat. The running mean shows when you've taken enough samples:

- In this example, the running mean **stabilizes around 9.5 species** after about 8-9 samples - Taking additional samples beyond this point causes little change to the mean - This indicates that **8-10 samples** would be sufficient for this habitat ### Avoiding bias - Always use **random number generators** for random sampling, not personal choice - Use **standardized methods** for all measurements - The same person should carry out all subjective measurements (like percentage cover) - Sample at the **same time of day/year** to control for temporal variation ### Ethical considerations - Return organisms to their original location after identification - Minimize damage to habitats during sampling - Handle organisms carefully to avoid injury - Check and release pitfall trap contents promptly