We wish to clarify that the boundary condition on the lower boundary of the domain was assumed to be a constant pressure head, as described in paragraph 11. The term “impermeable layer” in paragraph 10 is thus improper and was mistakenly left from a previous version of the manuscript. We apologize for this inconsistency.
 Wilson and Gardner  propose an interesting contribution to the study of salt-marsh subsurface hydrology and a useful numerical experiment supporting our finding that subsurface hydrology plays a major role in the ecogeomorphological dynamics of salt-marsh environments also as a result of previously undetected feedbacks between vegetation transpiration and saturated/unsaturated water flow [Ursino et al., 2004]. Indeed, even though theoretical and numerical characterizations of salt-marsh water table dynamics had previously been proposed [e.g., Harvey et al., 1987; Li et al., 1997, 2000], these focus on saturated flow, and no attempt had been made to describe the time evolution of the aerated soil layer. In this view, we believe that our results allow new approaches to the ecohydrological study of salt marshes, until recently dominated by the traditional tenet that the spatial distribution of halophytic vegetation species be linked just to the distribution of soil elevation, of the hydroperiod (ratio of marsh submersion durations to the duration of the entire time interval of reference) or of marsh inundation frequency [e.g., Chapman, 1964; Olff et al., 1988; Ungar, 1991; Kadlec and Knight, 1996; Morris et al., 2002].
 Wilson and Gardner's comment begins by stating that it is “commonly observed” that marsh grass productivity decreases with distance from the creek network. This is not correct and it is related to a much more complex picture than Wilson and Gardner imply. The salt-marsh ecology literature, in fact, reports cases in which biomass production increases away from the channels [e.g., Bertness, 1991] as well as cases in which a decrease is observed [e.g., Morris et al., 2002]. This is linked, in synthesis, to the fact that some vegetation species are better adapted than others to prolonged anoxic conditions and that, depending on their competitive abilities, both increased and decreased biomass productions may be observed on a negative oxygen availability gradient [e.g., Bockelmann et al., 2002; Silvestri et al., 2005; Silvestri and Marani, 2004]. Ursino et al.'s findings show that subsurface water flow characteristics have a major effect on root aeration and thus reconcile these observations with a new conceptual interpretation which may explain the different observed vegetation patterns on the basis of different subsurface hydrology regimes induced by variable soil characteristics (and boundary conditions, as exemplified by Wilson and Gardner and discussed later). We therefore believe that the presumed contrast between observations and our findings does not hold and that our results improve our understanding of salt-marsh biogeomorphodynamics.
 A point raised by Wilson and Gardner  regards the type of boundary conditions adopted. Ursino et al.  show that the detailed dynamics of saturated/unsaturated water flow within a marsh soil is dictated by tidal fluctuations, plant transpiration, flow domain geometry and soil characteristics. In this framework, Wilson and Gardner's results, indicating that a no-flux lower boundary condition leads to shorter root aeration periods with different spatial patterns from those previously determined is not surprising.
 On a more general, methodological, line we emphasize that our analyses were not aimed at describing a specific salt-marsh configuration, but rather at showing, for the first time, that the interplay between saturated/unsaturated subsurface flow, marsh elevation and vegetation transpiration gives rise to complex soil aeration patterns. We thus focused our analyses, in the framework of fixed domain configuration and boundary conditions, on determining the relative roles and interactions of vegetation transpiration, soil properties, marsh elevation and subsurface water flow, and we selectively simplified some of the detailed properties of real salt-marsh systems in order to explore their general behavior. It seems to us we have reached our objective and that the biohydrological importance of saturated/unsaturated flow in salt-marsh systems is established. Wilson and Gardner “agree that groundwater processes likely play an important role in tidal marsh ecology”: this can now be stated on the quantitative basis first proposed by Ursino et al. , rather than on one's personal judgment. Once the general conceptual framework and the importance of vegetation/hydrology feedbacks in salt-marsh environments has been established, the analysis of solutions obtained under a wide variety of assumptions, of which Wilson and Gardner's is one, may then be helpful in understanding the wide variety of observations from different tidal environments and even from different sites within a same tidal environment.
 Another observation by Wilson and Gardner concerns the boundary conditions imposed on the creek bank. The presence of the creek is represented by Ursino et al.  by the introduction of a region with very large hydraulic conductivity (10-1 m/s) on whose external boundary (the right vertical boundary in their Figure 1) a zero pressure head is imposed above the instantaneous tidal level and a hydrostatic distribution below it. As Wilson and Gardner point out, a seepage face cannot arise within such a model configuration and this may affect the model evaluation of lateral drainage from the salt-marsh soil to the channel. However, even though this effect may somewhat change the persistence and detailed position of the aerated zone near the creek during part of the ebb phase, our conclusions regarding the interplay between ecology and hydrology at the marsh-scale remain unaffected.
 Finally, we agree with Wilson and Gardner's suggestion that interesting future modeling developments should include the simulation of the gas phase dynamics. It is nevertheless worth noting that, treating the gas phase as well connected to the atmosphere at any time (i.e. even when the marsh surface is flooded), leads to an underestimation of actual aeration times, because it assumes that the gas is instantaneously displaced by water infiltrating from above. This implies that the effect of vegetation transpiration on aeration times is even greater than estimated by our present model. This mechanism will thus likely produce a more persistent aerated layer and constitutes a further motivation for the relevance of our findings.
 In conclusion we think that Wilson and Gardner's analyses are an interesting addition to our results, exploring different settings and showing the potential of our approach. In general, developments of subsurface hydrology models will in the future help to further clarify the role of previously neglected physical factors affecting salt-marsh plant ecology and to provide a basis for modeling the coupled evolution of vegetation and morphology in these environments. In specific applications, mathematical modeling of subsurface hydrology may allow the interpretation of observations which could not be previously reconciled within the traditional framework of salt-marsh ecology.
 Funding from TIDE EU project (EVK3-CT-2001-00064) and by COFIN 2002 Morfodinamica Lagunare project are acknowledged.