Published wildlife population figures are frequently reported as single numbers. The underlying estimates almost always carry a range, and understanding where that range comes from changes how the figure should be read.
Nobody counts every animal
Complete counts are possible only for species that are large, conspicuous and concentrated in known locations, such as breeding colonies of certain seabirds.
For everything else, a portion of the range is surveyed and the result is extrapolated, which introduces uncertainty proportional to how variable the unsurveyed area is.
The extrapolation assumes the sampled areas are representative, and where survey effort concentrates in accessible or protected places, that assumption is usually optimistic.
Detection probability is estimated, not known
Observers miss animals, and the proportion missed varies with vegetation, weather, observer experience and the animal's own behaviour.
Methods such as distance sampling estimate that proportion from the data itself, using how detections fall off with distance from the observer.
Because detection is estimated rather than measured, its uncertainty propagates into the final figure, and it is often the largest single contributor to the range.
Different methods produce different numbers
Aerial counts, ground transects, camera arrays and genetic sampling each have distinct biases, and applying two methods to the same population routinely yields different estimates.
This is not evidence that one is wrong. It reflects that each measures a slightly different quantity under different assumptions about movement and visibility.
Comparisons over time therefore require the same method applied consistently, and a change of method can produce an apparent trend that is entirely methodological.
Trends are more reliable than absolute numbers
A consistent survey repeated over years shares its biases across all years, so the direction and rate of change is much better constrained than the level.
This is why conservation assessments emphasise rate of decline rather than population size, and why a species can be listed without a confident total.
It also explains why headline figures shift when methods improve, without the underlying population having changed at all.
How to read a published figure
A stated range indicates the analysts' confidence, and a wide range signals genuine uncertainty rather than carelessness in the study.
The survey area matters equally, since a national figure and a regional one are frequently reported in ways that make them look comparable.
Where a single number appears with no range and no method, it has usually been simplified in reporting, and the original source will carry both.