Lay a tape from woodland into meadow, throw your quadrats, read the sensors, then find out whether your data really shows a range of tolerance.
01 · The idea
One line, one gradient
A transect runs along a gradient.Quadrats at regular intervals sample abundance; a sensor records the abiotic factor at each one.
Every species has a range of tolerance.Optimum in the middle, stress towards the edges, intolerance beyond.
Correlation is not causation.Abiotic factors travel together: shade is also damp and cool.
Range of tolerance: abundance peaks in the optimum zone and falls to zero in the zones of intolerance.
02 · Predict first
Meet the locals
Three plants live on this woodland edge. Before you sample, commit: how will each one respond as light intensity increases from the shady woodland floor into the open meadow?
03 · Into the field
Lay the tape
Drag the two pink handles to position your transect. Choose a quadrat size and interval, then sample. Try laying the tape across the edge instead of along it and see what happens.
◀ WOODLAND · shade30 m × 12 m siteMEADOW · open ▶
Quadrat size
Interval between quadrats: 2 m · 15 quadrats
Ready. Your tape runs 28 m across the site.
QUADRAT VIEW
Sample a transect to look inside each quadrat.
No data yet. Your results table fills as you sample.
04 · Make sense of it
Kite, scatter, Spearman
The kite diagram shows the most recent transect: the width of each kite is the density at that point along the tape.
Distance along the latest transect (m) →. Kite width ∝ density (plants m⁻²).
Now pool every quadrat you've sampled and test for a correlation.
optimumzone of stresszone of intolerancetrue tolerance curve
rs = –
Sample at least five quadrats to calculate Spearman's rank correlation coefficient.
05 · The reveal
Where's the limit?
In this simulated site each plant responds to light alone. Reveal the true tolerance curve on your scatter graph, then check your predictions.
Your moisture correlation is real, and misleading. Switch the x-axis to soil moisture: wood sorrel still correlates strongly, yet in this model it never "feels" moisture at all. The canopy makes the ground both dark and damp, so the two variables move together. A transect shows correlation; separating causes needs a controlled experiment.
Why red campion fools Spearman. Its density rises then falls across the edge: a hump, not a slope. Spearman's rank only detects monotonic trends, so rs comes out weak (or even points one way) even though light is controlling where it grows. Always look at the graph before you trust the number.