Question 3.3

National indices are combined to supranational ones. To this end, missing yearly indices of a particular country (national all-sites totals) are imputed from data of countries within the same region (West, North, East/Central, South, South East). Is it justified to do this? Can we e.g. really expect that data from France can be used to estimate missing values in Spain?


The assumption behind this approach is that countries in the same region have shared similar changes in socio-economic developments (rural development, urbanisation) and are subject to similar environmental pressures (pollution, climate change). As a consequence, bird species are expected to have had similar changes in population numbers.
This may not be true, but we regard it a better approximation of missing indices than alternatives. An alternative would be to extrapolate a species trend in a country beyond the years covered by the data, but this may lead to very unrealistic indices for missing years. Another alternative would be to assume that indices in years without data were equal to years with data, which is not very plausible either.