Experimental techniques practice questions.

A student plots the cumulative number of species recorded against the number of quadrats laid. The curve rises steeply at first and then flattens. The appropriate number of quadrats to use for the survey is

• the number at which the curve first flattens off
• the largest number that the available time allows
• the number at which the curve is rising most steeply
• ten, since that is the convention for vegetation surveys

Answer: the number at which the curve first flattens off

Once the curve flattens, further quadrats add almost no new species, so effort beyond that point buys nothing while stopping before it leaves species unrecorded. Running this plot during a pilot study is how sample size is justified rather than asserted.

Pitfall traps are set to sample the ground invertebrates of two woodland plots. The chief bias in the data they produce is that

• the traps catch only species that are active at night
• active, fast-moving species are caught far more often than sedentary ones
• the traps under-record species that are larger than the trap rim
• the catch depends entirely on the depth at which each of the traps was sunk

Answer: active, fast-moving species are caught far more often than sedentary ones

A pitfall trap samples activity rather than abundance: an animal is caught because it walked into the trap, so a fast-ranging beetle appears numerous and a slow-moving snail appears rare even where their densities are equal. Comparing two plots is still valid if the trapping effort is identical, but the counts must not be read as densities.

A student records seaweed abundance on a five-point abundance scale of the ACFOR type rather than counting individuals. The main disadvantage of this choice is that

• recording takes considerably longer in the field
• the scale cannot be applied to species that grow in continuous mats
• the categories are subjective and are not measured numbers
• the results cannot be displayed on a kite diagram

Answer: the categories are subjective and are not measured numbers

The scale is quick and works for species that cannot be counted as individuals, which is why it is used, but one observer's abundant may be another's common, and the categories are ordered rather than numerical, so calculating a mean of them is not meaningful. Percentage cover measured with a gridded quadrat is the more defensible alternative.

To compare invertebrate communities at two points on a river, a student must also record the flow velocity at each. The most appropriate simple field method is

• estimating the speed of the water by eye from the bank
• measuring the width and the depth of the channel only
• recording the time of day at which each of the samples was taken
• timing a float over a measured distance, repeated several times

Answer: timing a float over a measured distance, repeated several times

A float and a measuring tape give a repeatable quantitative value, and repeating it allows a mean and an estimate of variability. Velocity has to be measured because it is the abiotic factor most likely to explain a difference in the invertebrate community between two river sites.

A student surveying zonation on a rocky shore lays the belt transect parallel to the water's edge, at a constant height above the low-water mark, and records the seaweeds inside each frame along it. The effect of this layout on the recorded pattern is that

• no change in species composition will appear along the transect, so the zonation the survey was designed to detect is missed
• the transect will cross every zone twice, so each species appears in two separate bands and its abundance is recorded twice over
• the abundance of every species will be overestimated, because neighbouring frames along the transect overlap one another
• only the species of the splash zone will be recorded, because the transect lies above the tide line

Answer: no change in species composition will appear along the transect, so the zonation the survey was designed to detect is missed

Zonation is a change with height on the shore, so a transect must run up that gradient; laid along a contour it stays inside a single band and records only the variation within it. Note that zonation is a spatial gradient, not the change over time that succession describes.

In a capture-mark-recapture study of a beetle population the paint used as a mark makes the beetles conspicuous to birds, so more marked than unmarked beetles are eaten between the two sampling occasions. The estimate of population size obtained will be

• unaffected, because predation removes marked and unmarked beetles from the population in the same proportion over the interval between the two samples
• too high, because fewer marked beetles are recaptured and the recapture figure is the divisor in the estimate
• too low, because the number of beetles marked on the first occasion no longer represents the population
• too low, because the total caught on the second occasion falls while the number marked stays the same

Answer: too high, because fewer marked beetles are recaptured and the recapture figure is the divisor in the estimate

In N = (n1 x n2) / n3 the marked recaptures n3 are the divisor, so anything that removes marked individuals inflates N. A valid estimate requires that the mark itself changes neither survival nor behaviour.