Species–Area Relationships and Ecological Calculations

Why species richness changes with sampled area

When naturalists explore a larger area, they typically encounter more habitat patches, more individual organisms and often more ecological niches. They therefore tend to record more species. Alexander von Humboldt noticed this pattern during his investigations of South American landscapes. The species–area relationship formalises the observation and helps ecologists compare sites and interpret habitat loss.

At small scales, a short walk through a meadow may encounter common grass species but miss rare herbs occupying wetter corners. Expanding the survey can add those species. At very large scales, including several mountain ranges or islands brings together species with different evolutionary histories and habitat requirements. The rate at which richness increases is not constant across all scales.

The source illustrates a rising, progressively flattening species–area curve and a straight line when both variables are plotted on logarithmic scales. It calls the rising curve a rectangular hyperbola; the quantitative model written alongside it is specifically a power function, S = C × Aᶻ, which should not be confused with a mathematical rectangular hyperbola or assumed to have a strict saturation limit.

The power-law model and every variable

The empirical relationship is S = C × Aᶻ. Here S means species richness in the area surveyed; A means the surveyed area; C is a fitted constant that depends on how area is measured and on the organisms studied; and z is the exponent that determines how rapidly recorded richness increases with area. The exponent is dimensionless because it expresses the response to proportional changes in area.

If z is positive, larger areas predict greater species richness. A z of 0.2 does not mean species richness rises by exactly 20 for every additional square kilometre. It means a multiplicative increase in area creates a proportional response, S₂ / S₁ = (A₂ / A₁)ᶻ. This ratio form is especially useful because C cancels out when predicting changes for the same system under the same model.

An empirical fit describes observed data within its sampled range. Predictions beyond that range can be unreliable: habitat quality, fragmentation, species movements, evolutionary history and sampling intensity can change the relationship. Area alone does not determine all species richness.

Why logarithms turn the relationship into a straight line

Taking logarithms of both sides gives log S = log C + z log A. This is the familiar straight-line form y = b + mx if y = log S and x = log A. The intercept is log C, and the slope is z. A log–log plot is therefore a useful way to estimate the exponent from field measurements spanning different areas.

The slope can be calculated from two measurements as z = [log S₂ − log S₁] / [log A₂ − log A₁], provided both points belong to the same approximately linear log–log relationship. Any consistent logarithm base gives the same slope. The intercept depends on the chosen area units and sampling definition, but the ratio prediction uses only z.

The source reports typical fitted z values around 0.1–0.2 for many organism groups within regions, and steeper values around 0.6–1.2 when analysing very large geographical areas such as continents. Its example of tropical fruit-eating birds and mammals on different continents gives z about 1.15. These are reported empirical ranges, not universal constants.

Worked example: what does a fourfold increase in area imply?

Suppose a regional plant survey records 50 species in a 100-hectare area and the fitted exponent is z = 0.2. For a comparable 400-hectare region, the area ratio is 400 ÷ 100 = 4. Thus S₂ = 50 × 4⁰·². Because 4⁰·² is approximately 1.320, predicted richness is about 66 species. The area quadruples, but richness rises by only about 32%, not by 300% or by a fixed addition of four species.

Now imagine the same fourfold area ratio in a separate continental-scale comparison with z = 0.8. The richness ratio would be 4⁰·⁸, approximately 3.03. A steeper slope means a given proportional increase in area is associated with a larger proportional increase in richness. It does not show that every individual habitat is more productive or that species are equally distributed within the area.

For conservation, an area–richness model can help predict possible species losses from habitat reduction. But it cannot independently identify which species disappear, and extinction may be delayed by population dynamics and habitat fragmentation. Interpreting model output as an immediate exact extinction count would overstate what the equation can establish.

Identifying and correcting common graph errors

A straight line is expected when both richness and area are logarithmically transformed, not necessarily on ordinary axes. On ordinary axes, the power law with 0 < z < 1 produces a curved line that rises but becomes less steep. The source figure conveys this contrast, even though the terminology for the untransformed curve is imprecise.

The logarithmic intercept is log C, not C itself. If a fitted line crosses the vertical axis when log A = 0, then A equals 1 in the chosen area unit and log S equals log C. Changing from hectares to square kilometres changes the numeric intercept. Finally, a high slope does not mean 'more species everywhere'; it specifically means richness changes more strongly with proportional changes in area.

Quick Quiz

Practice Problems