Both the area or landscape of interest and the sites within it may be of very different sizes in different situations, and no consensus has been reached on what spatial scales are appropriate to quantify alpha diversity.
[3] It has therefore been proposed that the definition of alpha diversity does not need to be tied to a specific spatial scale: alpha diversity can be measured for an existing dataset that consists of subunits at any scale.
[4] The subunits can be, for example, sampling units that were already used in the field when carrying out the inventory, or grid cells that are delimited just for the purpose of analysis.
If results are extrapolated beyond the actual observations, it needs to be taken into account that the species diversity in the subunits generally gives an underestimation of the species diversity in larger areas.
[5][6] Ecologists have used several slightly different definitions of alpha diversity.
[1][2] It has been argued that defining alpha diversity as a mean across all relevant subunits is preferable, because it agrees better with Whittaker's idea that total species diversity consists of alpha and beta components.
This measure allows weighting rare and abundant species in different ways, just as the diversity indices collectively do, but its meaning is intuitively easier to understand.
Then alpha diversity can be calculated in two different ways that give the same result.
The second approach is to calculate the species diversity for each subunit separately, and then take a weighted generalized mean of these.
If the second approach is used, the equation is: This also equals a weighted generalized mean but with exponent 1 - q.
Large values of q lead to smaller alpha diversity than small values of q, because increasing q increases the effective weight given to those species with the highest proportional abundance and to those subunits with the lowest species diversity.
[4][13] Alpha diversity can be calculated in both extinct and extant landscapes.