Sampling Ecosystems & Comparing Evidence — Print · Grade 7 Science
Chapter: Ecological Evidence & Environmental DesignSampling Ecosystems & Comparing Evidence
Learning goals
- I can explain sampling ecosystems & comparing evidence.
- I can use evidence, a model, and stated assumptions to support a biological explanation.
Before this unit
Use the supplied vocabulary, distinguish an observation from an explanation, and identify any units or model assumptions. Revisit a connected earlier unit if needed; new terms are explained here.
Use three flexible 40–45-minute sessions: core background and Parts 1–2 with their checks; Parts 3–4 and the worked application; then remaining practice, a labeled model, revision, and the exit check. Checks completed during reading can count toward the practice set. Read one short part at a time. Investigations and chapter review/quiz have separate sessions, and flexible weeks allow more time.
Nobody counts all the daisies in a field. For almost any ecological question, counting everything is impossible, so ecologists count part of it and scale up. A sample is the part counted, and everything downstream depends on whether it represents the whole. That makes sampling a method with a built-in risk, and the risk has a name: sampling bias is any way the sample systematically differs from the thing it is meant to represent.
Two tools cover most situations. A quadrat is a frame of known area, usually a square, placed on the ground so that everything inside it can be counted. Several quadrats give a count per unit area, which can be scaled to the whole habitat. A transect is a line across an area along which samples are taken at intervals, and it is the tool for a question about change across a gradient — up a shore, out from a path, down a hillside — where an average over the whole area would hide exactly what you want to see.
What a quadrat gives you is density: individuals per unit area. Count several quadrats, average the counts, divide by the quadrat area, and multiply by the habitat area to estimate the total. The arithmetic is easy, and the result is an estimate whose quality depends entirely on decisions made before any counting happened — above all on where the quadrats went. Choosing attractive-looking spots produces a number that is precise, reproducible and wrong, which is why placement is decided in advance and at random.
Build your understanding
Where this fits
Every population figure and diversity measure in the last two chapters came from somewhere, and almost none of them came from counting everything. This unit is how those numbers are actually produced, what makes an estimate trustworthy, and the one decision — where to put the plots — that quietly determines the answer.
Use three flexible 40–45-minute sessions: core background and Parts 1–2 with their checks; Parts 3–4 and the worked application; then remaining practice, a labeled model, revision, and the exit check. Checks completed during reading can count toward the practice set. Read one short part at a time. Investigations and chapter review/quiz have separate sessions, and flexible weeks allow more time.
Part 1 · Matching the tool to the question
A quadrat answers questions about how much of something there is across an area: how dense a plant is, how cover differs between two fields, whether a species is more abundant here than there. The counts from scattered quadrats are averaged, and the scatter between them also tells you how variable the area is.
A transect answers questions about change along a direction. Vegetation up a beach, species composition outward from a path, plants along a gradient from wet to dry — in each case the pattern is the finding, and averaging over the whole area would destroy it. Samples are taken at stated intervals along the line, and the result is plotted against distance.
Choosing the wrong tool wastes the work. Averaging quadrats thrown across a shore gives a single meaningless figure for a place where everything depends on height up the beach. Running a transect across a uniform field produces a flat line that a handful of quadrats would have established faster. The question comes first and the tool follows from it.
Pause and explain
You want to know how plant species change as you move away from a footpath. Quadrats scattered at random, or a transect?
Check your reasoning after answering
A transect, because the question is about change along a direction. Samples taken at intervals on a line running away from the path show how composition shifts with distance, which is the finding. Quadrats scattered at random would be averaged into a single figure for the area and would hide the pattern entirely.
Part 2 · Deciding where to sample before you look
Placement is where bias enters, and it usually enters innocently. Put quadrats where the ground looks interesting and you oversample what caught your eye. Put them near the path because it is easier and you sample trampled ground. Avoid the nettles and you systematically exclude whatever grows with nettles. None of these requires any intention to mislead.
The defence is to decide placement before examining the area, and to use a procedure rather than a judgement. Random coordinates generated beforehand, or a systematic grid laid out regardless of what is in each cell, both remove the choice from the moment of looking. Write the procedure down before starting, and follow it even when a quadrat lands somewhere unpromising — especially then, since the unpromising spots are exactly what a biased sample leaves out.
Sample size matters alongside placement. Four quadrats in a variable area give an average that could easily be a long way from the truth, and the scatter between them is the clue: if four counts are 2, 4, 3 and 3 the area is fairly uniform, and if they are 0, 11, 1 and 0 the average of three means very little. Report the individual counts, not only the mean, which is the point the data unit in Chapter 1 made.
Pause and explain
Why should quadrat positions be decided before looking at the area?
Check your reasoning after answering
Because choosing positions while looking introduces bias without any intention to mislead — attractive patches get oversampled, difficult ground gets avoided, and convenient spots near a path are not typical. A procedure fixed in advance, such as random coordinates or a systematic grid, removes the choice from the moment of looking and makes the sample defensible.
Part 3 · Calculating density and scaling up
The calculation has three steps and each should be written out. Add the counts and divide by the number of quadrats to get a mean count per quadrat. Divide by the quadrat area to get density per unit area. Multiply by the habitat area to get an estimated total.
Work the model. Four quadrats of 1 m² contain 2, 4, 3 and 3 plants. The total is 12 across 4 m², so the density is 12 ÷ 4 = 3 plants per square metre. The habitat is 20 m², so the estimated total is 3 × 20 = 60 plants.
Then look at the scatter before trusting it. The four counts range from 2 to 4, a spread of 2 around a mean of 3, which suggests a reasonably uniform distribution and makes the estimate of 60 plausible. Had the counts been 0, 0, 1 and 11, the same mean of 3 would have come from a clumped distribution, and an estimate of 60 would depend entirely on how many clumps the 20 m² happened to contain.
Pause and explain
Four 1 m² quadrats contain 2, 4, 3 and 3 plants, and the habitat is 20 m². Calculate the density and the estimated total.
Check your reasoning after answering
The total is 12 plants in 4 m², so the density is 12 ÷ 4 = 3 plants per square metre. Multiplying by the habitat area gives 3 × 20 = 60 plants as the estimate. The counts range only from 2 to 4, which suggests a fairly even distribution and makes the scaling more trustworthy than it would be for a clumped one.
Part 4 · Saying that it is an estimate, and what it assumes
Sixty is an estimate and should be written as one. It assumes the four quadrats represent the whole 20 m², that the counting detected every individual present within them, that the species was correctly identified, and that the habitat area is what was stated. Each assumption can fail, and naming them is what distinguishes a scientific figure from a guess with arithmetic attached.
Detection is the assumption most often forgotten. Small, hidden or partly buried individuals are missed, and if they are missed consistently the estimate is biased low in a way more sampling will not fix — repeating a method that misses things simply produces the same error more precisely. Only a change of method addresses it.
So report the figure with its workings and its limits: four 1 m² quadrats placed at random coordinates fixed in advance, counts of 2, 4, 3 and 3, mean density 3 per square metre, estimated total 60 individuals in 20 m², assuming the quadrats represent the area and that detection was complete. That sentence is checkable, reusable and honest, and it is the standard this chapter has been working towards.
Pause and explain
Why does taking more quadrats not fix a detection problem?
Check your reasoning after answering
Because a method that consistently misses small or hidden individuals makes the same error in every quadrat. More sampling reduces the random scatter around the figure but not the systematic shortfall, so the estimate becomes more precise and remains biased low. Only changing the method — closer searching, a different technique — addresses it.
Apply the whole unit · Estimate a plant density
Illustrative sample, not surveyed data: four quadrats of 1 m² contain 2, 4, 3 and 3 plants of one species. The habitat area is 20 m².
Your task
Find the sample density, calculate an extrapolated total, and state the assumptions that prevent it being called a complete census.
What your response should show
- Explain the relevant biological process accurately.
- Use the supplied evidence and show calculation steps with units where needed.
- Connect a labeled model or mechanism with the claim.
- State an assumption, evidence limitation, or useful next observation.
Response guidance — after your attempt
Total the counts: 2 + 4 + 3 + 3 = 12 plants across 4 m² of quadrats.
Calculate density: 12 ÷ 4 = 3 plants per square metre.
Scale to the habitat: 3 × 20 = 60 plants as the estimated total.
Name the assumptions: that the quadrats represent the whole area, that every individual present in them was detected, that the species was identified correctly, and that the habitat is 20 m² as stated.
The sample density is 3 plants per square metre and the extrapolated total is 60 plants. The four quadrats contain 2 + 4 + 3 + 3 = 12 plants across 4 m², giving 12 ÷ 4 = 3 per square metre, and multiplying by the 20 m² habitat gives 3 × 20 = 60. The word estimate matters and should appear in any report of it, because four assumptions stand between the calculation and a count. First, that the four quadrats represent the whole area — which depends on how they were placed, and the figure is only defensible if positions were fixed by a procedure before the area was examined. Second, that every individual inside each quadrat was detected; small, hidden or partly buried plants are easily missed, and a method that misses them consistently biases the estimate low in a way that more quadrats will not correct. Third, that the species was identified correctly throughout, which matters where similar species grow together. Fourth, that the habitat really is 20 m² and is reasonably uniform. The scatter is reassuring here: the counts run from 2 to 4 around a mean of 3, which suggests a fairly even distribution and makes scaling reasonable. Counts of 0, 0, 1 and 11 would have given the same mean and a very different situation, where the total would depend on how many clumps the area happened to hold. Reported properly: four 1 m² quadrats at predetermined random positions, counts 2, 4, 3, 3, density 3 per m², estimated total 60 individuals in 20 m², assuming representative placement and complete detection.
Scientific references for this unit
Parent / teacher teaching notes
Ask for a because-statement and a labeled relationship, not just a memorized word. Work through one calculation step at a time, permit oral or diagram responses, and use flexible sessions for difficult reading. Compare the learner’s actual evidence with the supplied model; do not reward copying or agreement with a personal preference. No private family, genetic, or medical records are required.
The key idea
Ecologists often sample because counting every organism across a large habitat is impractical. A quadrat defines an area for selected observations; a transect organizes observations along a line. The method must fit the organism and question. A plant-count method does not automatically detect moving animals equally well.
Worked examples
Example 1 · Judge an estimate from its scatter
- Survey X: total 20 in 5 m², density 4 per m², estimate 4 × 50 = 200.
- Survey Y: total 20 in 5 m², density 4 per m², estimate 4 × 50 = 200 — the identical figure.
- Compare the scatter: X ranges from 2 to 5, while Y is four zeros and a 20, which indicates a strongly clumped distribution.
Identical means can come from completely different distributions. Always look at the individual counts.
Example 2 · Find the bias in a procedure
- Identify the decision rule actually used: positions chosen while looking, selected for apparent variety.
- Predict the direction of the error: the richest patches are oversampled, so richness and diversity will come out higher than the meadow's average.
- Replace judgement with procedure: generate random coordinates before entering, or lay a systematic grid, and sample every position the procedure gives.
Ask what decided each position. If the answer involves looking first, the sample is biased.
Avoid these mistakes
- Mistake: an extrapolated total is a count. Four quadrats giving 60 plants in 20 m² is an estimate assuming representative placement and complete detection; say so.
- Mistake: quadrats can be placed where the ground looks interesting. That oversamples what caught your eye; positions must be fixed by a procedure before the area is examined.
- Mistake: more quadrats fix any problem. More sampling reduces random scatter but not a systematic shortfall from missing hidden individuals — that needs a different method.
- Mistake: the mean is the result. Counts of 2,4,3,3 and 0,0,1,11 give the same mean and completely different confidence in scaling it up.
Check your understanding
Why specify whether sightings or individuals are counted?
Discuss the answer
Repeated sightings can involve one individual; the quantities answer different questions.
Can a mean remove biased site selection?
Discuss the answer
No. A biased set can still produce a precise-looking mean; location selection must be addressed.
Words to know
- sample
- The part of an area actually counted, from which the whole is estimated; everything depends on whether it represents the whole.
- quadrat
- A frame of known area placed on the ground so everything inside can be counted, giving a density that can be scaled up.
- transect
- A line across an area with samples taken at intervals, used when the question is about change along a gradient rather than an average.
- density
- Individuals per unit area; mean count per quadrat divided by quadrat area, then multiplied by habitat area to estimate a total.
- sampling bias
- Any systematic difference between the sample and the whole, usually introduced by choosing where to sample while looking at the area.
Extend your scientific explanation
Compare a second supplied example or the linked educational reference. Identify a connection, a model limit, and an additional observation that would test the explanation.
In the real world
Use a clear question, reliable observations, and a biological mechanism to evaluate a claim. State what remains uncertain before applying a classroom model to a different setting.
Exit check
Does sample density establish the exact habitat total?
Check your explanation after answering
No. Extrapolation depends on sample representativeness, area definitions, and detection.
Explain your reasoning without notes. If a key idea is missing, revisit the lesson and try again.
Problem 1
Answer and explanation
Repeated sightings can involve one individual; the quantities answer different questions.
Problem 2
Answer and explanation
No. A biased set can still produce a precise-looking mean; location selection must be addressed.
Problem 3
Answer and explanation
Plot areas can differ; count per area accounts for that boundary.
Problem 4
Answer and explanation
No. Extrapolation depends on sample representativeness, area definitions, and detection.
Problem 5
Answer and explanation
The sample density is 3 plants/m² and the extrapolated count is 60 plants. It is an estimate assuming the plots represent the 20 m² habitat and detection is adequate. Patchiness, omitted zones, or missed plants could change the result. The supplied data do not establish an exact total.
Total count: 12 across 4 m².
Density: 12/4=3 plants/m².
Estimated larger count: 3×20=60.
Response criteria
- Explain the relevant biological process accurately.
- Use the supplied evidence and show calculation steps with units where needed.
- Connect a labeled model or mechanism with the claim.
- State an assumption, evidence limitation, or useful next observation.
Problem 6
Answer and explanation
No. Different plot areas require count-per-area calculations or an appropriately weighted comparison using comparable detection methods.
Response criteria
- Explain the relevant biological process accurately.
- Use the supplied evidence and show calculation steps with units where needed.
- Connect a labeled model or mechanism with the claim.
- State an assumption, evidence limitation, or useful next observation.
Problem 7
- Only sample favorite patches
- Use a predefined representative location plan
- Delete low counts
- Change plot size secretly
Answer and explanation
Use a predefined representative location plan
A planned sampling method helps address selection bias, though other limits remain.
Problem 8
Answer and explanation
3 plants/m²
18/6=3 plants/m² under the stated detection method.