English

Quadrat Sampling

High schoolBiologySeeded virtual plant habitat with explicit quadrat counts and independent random grid sampling

Click the meadow to move a quadrat, or drag its orange handles. Record real counts, then compare manual placement with independently sampled random grid cells.

3D

Drag the apparatus to adjust it. Drag the background to rotate. Scroll to zoom.

Measurement notebook

Change the setup, test a prediction and save the result. Up to 12 snapshots are kept.

About this experiment

Move a measuring frame across a three-dimensional meadow, count the plants inside it and build a recorded survey. Compare deliberate placement with random grid samples in uniform or clustered habitats.

Learning goals

  • Calculate density from a quadrat count and area.
  • Estimate the population of a larger habitat from recorded samples.
  • Distinguish random placement from manual placement and repeated inspection of one frame.

How to use it

  1. 1Click the field or drag a handle to position a quadrat and record its count.
  2. 2Move the frame and confirm the recorded count remains unchanged.
  3. 3Reset the survey and sample ten random grid cells.
  4. 4Compare clustered and uniform habitats with the same field population.

The model

The population estimator averages sampled densities.
The displayed standard error uses the sample variance of independently randomized densities.

Assumptions

  • The field is 10 m by 10 m. Plants are fixed seeded points, not growing organisms. Clustered positions use reflected Gaussian clusters.
  • A plant is counted by its centre. Shared cell boundaries use a consistent half-open convention, with the outer field boundary included.
  • Only 1 m and 2 m sides are offered, so each random sampling grid partitions the entire field equally. Cells are sampled independently with replacement.
  • The estimate is field area times mean sampled density. Standard error requires at least two recorded samples whose placements were independently randomized.
  • Manual positioning is recorded as manual. A random placement is consumed by its first recorded count; recording that unchanged frame again is not a new random replicate. Editing the habitat clears the survey.

Try this in class

  1. Q1Why can choosing the densest-looking patch bias the estimate?
  2. Q2Why does a larger quadrat often reduce variability in a clustered habitat?

Frequently asked questions

Why is standard error hidden for manual samples?

The random-grid sampling calculation is not justified for deliberately selected or repeated unchanged placements.

Do random samples avoid previously sampled cells?

No. Independent sampling with replacement can return to the same cell. Its seed and placement sequence make the result reproducible.