The hypothesis, that the epifauna present in Avicennia mangroves round Gazi Bay, Kenya (mainly deposit feeding fiddler crabs (Uca sp.) and grazing gastropods like Terebralia palustris, Littorina scabra) has a regulating influence (trophic, like predation/competition, or structural (bioturbation)) on the meiobenthos, is investigated in this work. A field exclusion experiment was test to test this null hypothesis. Therefore, a randomized block design was used, consisting of nine experimental units. Three replicates were taken for each treatment (Cage, Partial cage and Blanco respectively). In the Cage, the epifauna was excluded; in the Partial cage (the first control introduced in our experiment), the epifauna was not excluded, so only the influence of the cage itself was investigated; in the Blanco (the second control in our experimental design, without cage) the epifauna remained well in order to follow the natural changes in time of the meiobenthos in the Avicennia mangal. Three monthly meiobenthic samples were collected for each treatment; together with those meiobenthic samples, abiotic factors (grain size analysis, curtosis, skewness, median particle diameter, porosity, % organic material, temperature, redox potential, pigment analysis (measurements of chlorophyl a, fucoxanthin), nutrient analysis (NQ,- NO3, NH4+, and Si), salinity, pH and dissolved oxygen) were measured once a month. Both the meiofaunal and abiotic results were statistically processed. In the meantime, seven five-day samples were taken out of the Cage only, to confirm certain tendencies seen in the results of the monthly samples, or to discover new trends. Also a comparative study was performed between the results of the Avicennia mangal and the results of a comparable caging experiment, which took place in the Ceriops mangal in Gazi Bay (STEYAERT, 1993). After a correlation analysis, ANalysis Of VAriance (global effects and contrast analysis) and multivariate analyses (such as TWINSPAN-analysis, Correspondence Analysis, Detrended Correspondence Analysis and Canonical Correspondence Analysis), the following results were obtained: 1.The pigments (which, in our case, point in the direction of diatoms present in our plots), especially chlorophyl a, showed in the Cage a clear increase in the third month for Avicennia, and a comparable increase in the second month for Ceriops. This exclusion effect is most probably the consequence of the removal of the gastropods, which feed on micro-algae and on leafs of mangrove trees. The other abiotic factors show a quite irregular and unclear fluctuation. 2. The meiofauna showed no remarkable exclusion effect no matter what, as well in Avicennia as in Ceriops. In the three monthly samples, the meiofauna seems to show no reaction to the exclusion of the epifauna, and not even to the increase of the diatoms, which are an important food source for a lot of meiobenthic taxa in many habitats. The results of the five-day samples in the cage show a strange tendency, both in Avicennia and Ceriops a minimal meiofaunal density ( particularity Nematodes and Turbellaria) occurs in the deepest layers during the fifth five-day sampling period, accompagnied by a relatively high density in the upper layer. This could mean that there is an exclusion effect involved, if the disappearance of the epifaunal activities causes a stable upper region in the sediment, followed by a vertical, upwards migration of the Nematodes and their predators, the Turbellaria. But this possible vertical migration could be caused by other factors as well (for instance low oxygen concentrations in the deeper layers). And as there are no Blanco or Partial cage five-day samples processed, this trend can just as well be the consequence of a natural fluctuation. The next conclusions can be drawn from the results obtained in this study: 1. The null hypothesis is false and should be rejected; 2. The null hypothesis is true and should not be rejected; some possible causes for the unsuccesful results of this cage experiment are: 1) The sampling period was too short; the monthly samples were taken over a period too small to see a clear increase of meiofaunal densities; 2) Because of the (already) high densities present in the two mangals (normal densities lay round 3000 ind/10cm2), competition and predation in and among meiofaunal taxa could be too high to obtain a high and/or quick increase of the meiofaunal density; 3) Other, unknown factors (such as for example a longer generation time of tropical meiobenthos as is the case for many "bigger" nematodes, the negative influence of toxic tannins, or - a combination of - other limiting factors) in the mangroves of Gazi Bay make sure that the total density of the infauna doesn’t increase.