2.7.2.5 'Floodplain, terrestrial GDE' landscape group

The ‘Floodplain, terrestrial GDE’ landscape group consists of two landscape classes. Of these, the zone of potential hydrological change contains 2358 km² of ‘Terrestrial GDE, remnant vegetation’ and 75.2 km² of non-remnant ‘Terrestrial GDE’.

Four qualitative mathematical models were developed for the ‘Floodplain, terrestrial GDE’ landscape group. The models focused on the influence of surface water and groundwater hydrology on trees that support a woodland community, and create local conditions and a microclimate (i.e. shade, leaf litter and soil moisture) that favours mesic vegetation and suppresses xeric vegetation. During the process of developing the models at the qualitative modelling workshop, debate surrounded the role of deep-rooted trees in drawing groundwater to the surface where it can become available for shallow-rooted mesic vegetation. The strength or existence of this effect was deemed to be uncertain. Seasonal floods are generally shown to suppress xeric vegetation in the floodplain and to favour trees and other mesic species. Seasonal floods are important contributors to groundwater recharge, but excessive groundwater recharge could potentially contribute to saturated soil conditions where an anoxic root zone suppresses deep-rooted vegetation (i.e. trees connected to groundwater and mesic non-tree deep rooted vegetation). However, it is uncertain whether these floodplains could develop anoxic soil conditions, and thus the existence of this link was also uncertain. Four alternative models were developed to address the uncertainty for the links associated with trees drawing groundwater to the surface and the potential for groundwater to contribute to anoxic soils.

A single receptor impact model was developed based on expert elicitation during the quantitative modelling workshops. The model examined the response of woody vegetation to changes in flow regime and groundwater, thus indicating the dual influences of surface water and groundwater on this landscape group. The model used the annual mean percent foliage cover of woody vegetation as the receptor impact variable.