The experiments reported here were designed to test the hypothesis that numerical responses of soil fauna to simulated rainfall would occur only in soils having an adequate leaf litter cover. The soil microarthropods responded to artificial rainfall as hypothesized, that is, the most abundant taxa responded to litter quantity and moisture but not to moisture in the absence of sufficient substrate.

Desert soil fauna, eg. microarthropods and nemntodes, arc patchily distributed with thodivcl'\lity and density varying as a !unction or leaf litter distribution and soil organic matter (S.INTOS el al. 1978; FRECKIIAN & MANK.\U 1977). The accumulution or plant debris in patelws or varying size •1>JJCal'\l to be a factor in the breakdown of surface litter in desert ecosystems. W111TFORD elal. (1982) and W1mPORD ti al. (1980) round relationships between quantities or surface litter aceumulations and rate of litter breakdown. Wn1TFOt,tD et al. (1982) also round that SUJ>pression ol soil microartbropod, by chemical means reduced the rate or litter disap1>earance and coneluded tlrnt soil microfauml were import.int in litter break down. Microarthropods are netive in surface litter accumulations for 2-4 hour􀌴 in the early morning even during 1>eriods when the soil has no measureable moisture (1V'mTFORO et al. J.981), but when the soil and litter ore wet, more indhridnals and more taxa can be extracted trom both soil nud litter. Jn our earlier work (WaITFORO et at. 1981) we found that simulated rainfall atrected the proportions ol the nematode J)opulation that were anbydrobiotic but found no changes in numbers following wat􀌵r addition such as that exhibited by soil rnirroarthropods. However tlrnt study was based on a single si1uulnted rainfall event (W111TYOH1) et al. 1981) which resulted iu large numerical responses of microartbropods iu 1mrfac·e litter and in soil under litter accumulations. Nov-MEIR (1074) cmphasir.cd that water is the factor of overriding importanc·e in desert eco,yste111s and stated that rainlall pulses must be i111portant stimulators ol soil biotic activity.The studies or distributions or soil laun11 (SAxTOS ct 111. 1978; FnAxc" cl al. 1979) and Jitter disappearance - soil launn relationships (SAN1'0S & W111wo1m 198L; WHITFORD et al. 1982; W111noao el al. 1983) suggest that lor soil organisms, an adequate energy ancl nutrient impply in the form of l)lant litter is more important than waler as a lactor arr('{'ting numeric·al responses of soil fauna. The e::q1eriment.s reported here were designed to t('st the hypothesis that numerical responses. of soil fauna to simulated rainfall would otrnr only in soils having an adequate leaf litter cover.