228 The Evolution of Communities 



ators which collectively would maintain a steady depression of the 

 prey species. In the parasites of the Mediterranean fruit fly in 

 Hawaii, just discussed, the populations of each parasite varied 

 greatly from 1915 to 1933, but the total parasitism (upper line) 

 fluctuated much less (Fig. 98). It is highly likely that the popula- 

 tion of the fruit fly would have periodically reached higher levels 

 if it had been parasitized by only one species of wasp rather than 

 by four. In contrast the observations of Burnett (1960) concern- 

 ing population fluctuations of the white fly Trialeurodes vaporario- 

 rum, parasitized by a single species of wasp Encarsia formosa, indi- 

 cate that even one parasite may, under special conditions, keep 

 the prey species at low population levels for long periods of time. 



A different dimension of the same principle is illustrated by other 

 studies made in Hawaii on the parasitism of a tephritid gall fly 

 Procecidochares titilis by three parasitic wasps. Bess and Haramoto 

 (1959) found that the proportionate abundance of each parasitic 

 species varied greatly in different vegetational and rainfall zones 

 but that in each zone the total parasitism was remarkably similar 

 (Fig. 99). This example indicates that multiple parasitism tends 

 to equalize host densities along ecological gradients as well as in 

 time. 



Observed low population densities in natural habitats frequently 

 bear out the absence, or at least apparent absence, of competition 

 between coexisting species. At no time during the ten years of 

 collecting for Erythroneura on sycamore in Illinois were trees 

 found to be heavily infested with the leafhoppers. Of the six syca- 

 more leafliopper species collected, five were predominant in at least 

 some collections, but the sixth, E. hella, was never predominant 

 (Ross, 1957). It seems probable that vicissitudes of weather, para- 

 sites, and predators normally deplete the populations of these leaf- 

 hoppers to a point far below the competitive level, at least in some 

 parts of the range, and that this has resulted in the long-term 

 survival of a species of possibly lower competitive value, such as 

 E. hella, in coexistence with the others. 



Two maxims appear from these data concerning the coexistence 

 of species in direct competition: 



(1) If populations over the entire community range are habitually 

 at levels producing inter-specific competition, coexistence 

 will be possible only for species best adapted to some recur- 

 ring variant of the ecological pattern of the community. 



(2) If populations are habitually or locally below lexels produc- 

 ing interspecific competition, any number of potentially but 

 not actually competing species may coexist. 



