ZOOGEOGRAPHY 375 



Williams, 1941). This means that there is competition for the smaller food ani- 

 mals between the species of predators in the same size class. To use a hypotheti- 

 cal example in which only food is used: if there are ten pounds of food in X time, 

 and the predators require one pound each in X time, then the number of 

 species of predators will lie between ten (monoecious or parthenogenetic) 

 species each consisting of one individual, or one species with ten individuals. 



The 13,000,000 species derived above assumed optimal food and no compe- 

 tition within the community. This is patently not the case. The only study in 

 which neotropical Pselaphidae have been quantitatively collected in relation 

 to all other animals of their size range is that of Williams (1941) on the forest 

 floor of Barro Colorado Island. Williams analysed eighteen small quadrats 

 twenty-five centimeters square and eleven large quadrats one meter square, 

 both manually and with Berlese samples. This analysis produced 498 speci- 

 mens of Coleoptera, including 26 specimens of Pselaphidae. These beetles were 

 of small size, averaging between one to two millimeters. They belonged chiefly 

 to seven characteristic families (Carabidae, Staphylinidae, Pselaphidae, His- 

 teridae, Ptilliidae, Tenebrionidae, and Scarabaeidae). The first five families 

 are either wholly or partially carnivorous and compete for the same food in all 

 probability. Not all of these beetles were identified by experts in time for 

 publication so the number of competing species of beetles must be arrived 

 at in some other way. There were eight species of pselaphids with an average 

 length of 1.7 millimeters. If we reorganize the Williams data, then the simple 

 proportion of 498 : 26 as X : 8 gives 153 species of possible coleopterous 

 competitors. 



In addition to these beetles there were numerous animals present in the 

 leaf mold, many of which were predaceous. These included species in the same 

 size range of one to three millimeters, small Oniscidae, Araneida, Chilopoda, 

 Dermaptera, Reduviidae, and ants. The ants are characteristic, abundant, 

 omnivorous and at least some species are arhythmic (Park, 1941, 1941a). 

 These and the other animals are important competitors. Williams' tables list 

 about 102 possible competing species other than beetles. This brings the num- 

 ber of competitors to 255 species. 



In addition another hundred were not identified. Assuming thirty of 

 these small forms to compete, the total possible competing species is 285 to 8 

 species of Pselaphidae, or 35 species to one pselaphid species. 



This gives a competing ratio within the community but does not take 

 into account inequalities in intra-community distribution within the same 

 stratum, namely the forest floor. Both in this and the nearctic region I have 

 found some nights to have a high yield to lights while other nights produced 

 no pselaphids; some logs on Berlesing a sample have a high yield, nearby 

 logs few or no pselaphids. This must be taken into account. Williams found 

 that pselaphids occurred on 60 per cent of all quadrats. 



Therefore: 13,000,000 x 60% gives 222,857 species. 

 35 



