SUMMARYsoil. These effects of large-bodied grazers on soil properties in tidal marshes has beenlargely neglected so far, but need to be considered when we want to introduce livestockto ungrazed marshes.Next to above-ground grazing, we also find small grazers that grub for below-groundstorage organs in tidal marshes. This type of grazing removes the entire plant and it istherefore much more difficult for the vegetation to regenerate. This grubbing behaviorgenerally causes bare patches to form within the marsh vegetation. In extreme cases, itcan lead to large marsh areas to become degraded. We studied the regeneration of theselocal bare patches created by grubbing Greylag geese (chapter 5). Additionally, we studiedecosystem development on a landscape scale. Within our study site, a large populationof Greylag geese is reducing in size and we hypothesized that geese grubbing on alocal scale or ecosystem development on a landscape scale, is in fact reducing their foodsupply. We found that the bare patches regenerated back to a similar vegetation typewithin about 12 years. Therefore, the geese do reduce their food supply, but only shortlyand very locally. Very interestingly, we found an increase in plant diversity as young-successionalplant species established in these bare patches during regeneration. On a landscapescale, we found a very high accretion rate that outpaced the rising sea-level. Theirpreferred food source, Bolboschoenus maritimus, reduced in cover, while Elytrigia ather!ica was increasing in cover. Bolboschoenus maritimus is generally limited to the lowerelevated depressions between the creek banks, whereas Elytrigia atherica dominates optop of the creek banks. An increase in elevation would allow Elytrigia atherica to expandtowards the depressions and Bolboschoenus maritimus to reduce in cover. Therefore, weconcluded that the natural development of the marsh on a landscape scale is causing theGreylag geese to get evicted from the ecosystem and not due to local degradation by thegeese themselves. Additionally, and in contrast to other studies showing the negativeeffect of grubbing small grazers, we concluded that grubbing geese can positively affectplant diversity through a local set-back of the plant succession.THE IMPACT OF LARGE GRAZERS AND HETEROGENEITY ONSMALL"GRAZER ABUNDANCEIn tidal marshes, small grazer abundance is known to change with increasing product -ivity. At young successional stages, where productivity is still low, limited biomass productionlimits the amount of small grazers that can forage within the system. As themarsh develops and productivity increases, then the abundance of small grazersincreases as well. However, at mature marshes the cover of nutritious plant species getreplaced by unpalatable ones and the small grazers reduce in abundance again. Largegrazers are known to facilitate for small grazers by bringing back the nutritious plant species again. In line with these previous studies, we also found an increase in the abundanceof small above-ground grazers when cattle were introduced in mature marshes.However, we did not find any evidence for the cattle facilitating for below-ground grubbers(chapter 6). As mentioned previously, we found a reducing population of Greylaggeese in our study site due to natural succession. In contrast to the small above-groundgrazers, we did not find an increase in the abundance of small below-ground grubberswhen cattle were present.Next to the effect of the large grazers, we also found an effect by small-scale heterogeneityin the marsh platform on small grazer presence (chapter 4). We studied a smallscaletopographic heterogeneity (of a few square metres) that consisted of higherelevated hummocks alternating with lower elevated depressions. By comparing the soilcharacteristics underlying the heterogeneity in four European marshes, we concludedthat this pattern is formed in the pioneer stage, before marsh formation starts and finegrainedsediment accumulates on the marsh platform (box 1). Throughout ecosystemdevelopment, ranging from pioneer to mature marshes, this heterogeneity increasedplant diversity (chapter 4). Additionally, we found a very high grazing pressure by hareon top of the higher elevated hummocks, which was especially higher in the youngsuccessional stages of 15 and 30 yrs-old marsh compared to homogeneous marsh (chapter 4). Although limited primary production in young marshes generally limits theabundance of grazers, presence of higher elevated hummocks increased local primaryproduction and the hare could profit from these local elevated patches.MANAGEMENT IMPLICATIONSMany European marshes are grazed by wildlife as well as by livestock, to maintain plantdiversity. For management purposes, we need to understand the impact of these differenttypes of grazers on the vegetation and on important marsh dynamics. The results inthis thesis showed that small above-ground grazers reduced the vegetation height, buttheir effects on important marsh dynamics (that were studied in this thesis) were verylimited (chapters 2 and 3). Below-ground grubbing geese had a large impact on the vegetationcomposition and this increased plant diversity in tidal marshes, although thiseffect was only present for a limited period of time until the bare patches regenerated(chapter 5). With respect to cattle, these large-bodied animals had a very large effectthrough trampling of the soil. Trampling by cattle reduced the marsh accretion rates(chapter 2), while it enhanced the carbon sequestration rate in the marsh soil (chapter3). Whether the positive impact of large grazers on biodiversity and carbon sequestration,outweighs the negative impact on marsh accretion rate will be very site specific,depending on the problems the marsh are facing. Marshes with high sedimentation rates will be less affected by a reduced accretion rate. Hence, livestock grazing on thesesites could be used to provide increased carbon sequestration rates next to an increasedbiodiversity. These impacts of grazers on both vegetation and soil characteristics shouldbe taken into account in future studies that use models to estimate whether coastal habitatscan cope with an accelerated sea-level rise.