Prosopis velutina (velvet mesquite) forests are one of many types of a
ridland riparian ecosystems that are threatened by groundwater pumping
and other types of water development. Empirical models developed usin
g both hydrological and vegetational data sets have potential uses in
the management of these threatened ecosystems. To this end, we develop
ed models for Prosopis velutina stands across a xeric-to-mesic moistur
e gradient. The models expressed canopy height, basal area. leaf area
index, vegetation volume, and leaflet area as functions of plant water
potential, and they expressed plant water potential and riparian stan
d structure as functions of water table depth. These data indicated th
at stand structure was strongly related to water availability. Managem
ent applications of the models include the ability (1) to identify min
imum water-table depths for riparian stand maintenance and (2) to dete
ct stressful hydrological conditions, via water potential measurements
, before the onset of structural degradation.