Seeing shapes in clouds: the fallacy of deriving ecological hypotheses from statistical distributions

We used worldwide nationality boundaries to aggregate data by standard boundaries that are comparable across disparate data sets (and because publicly available species and non-species data commonly are aggregated by national boundaries). We chose species data sets where abundances were available by country (amphibians, ants, birds, freshwater fish, human diseases, Lepidopteran species, mammals, mosquitoes, non-natives, rare, reptiles, trees, crops). We chose non-species data sets that parallel species (where individuals are aggregated by their group membership) [airlines, cheeses, cities, colleges, hospitals, imports, industries, journals, livestock, newspaper companies, political parties]. For example, the number of newspapers printed in a country would not be parallel, but the number of newspaper companies printing newspapers is parallel to the number of species containing individual organisms. We used several data sources where ecological and non-ecological data were aggregated by the same country boundaries: · wikipedia.org · www.worldatlas.com · www.nationmaster.com · www.ebird.org · www.iucnredlist.org · fish.mongabay.com/biotope_countries.htm · bovbeers.wordpress.com/2011/07/02/number-of-breweries-worldwide · www.scimagojr.com/journalrank.php · www.iucnredlist.org · www.antwiki.org/wiki/Diversity_by_Country · vectormap.si.edu/MosquitoCountryList.htm · reptile-database.reptarium.cz · www.invasive.org/species.cfm.