200 years ago, French American zoologist Nicolas Hentz estimated global
spider diversity to be above 2,000 species. Current estimates
certainly closer to reality, but the wide range of estimates also shows
that we are still far from having a reliable estimate of spider
For this reason I have started several long-term projects with the aim of producing reasonable estimates of the actual pholcid diversity in several megadiverse regions: the Atlantic Forest in Brazil (ongoing project, funded by AKG and Brazilian funding agencies), the Greater Antilles in the Caribbean, tropical Africa (previous DFG project: HU980/9-1), and Southeast Asia (current DFG project: HU980/11-1). The basic idea is simple: extrapolations from species per locality, endemism levels, and distribution patterns are combined with vegetation data to produce a first hypothesis of overall diversity based on more than just speculation.
Below are accumulation curves of species (y-axis) with field days (x-axis) from the four megatransects. Notably, none of the curves levels off; it is just the slope that differs, depending on various factors (e.g. it is more difficult to move in Africa than anywhere else).
The curves below show percentages of new species (as of 2008) (y-axis) with field days (x-axis). After large initial variation due to chance events, they all stabilize at about 65-85%. A rough preliminary extrapolation from these values suggests that there are worldwide about 4000 species of Pholcidae.
More detailed information about the Brazilian transect is found here; details about the expeditions are found here.