Modelling spatial patterns of vegetation in desert sand dunes.

A stochastic numerical approach was developed to model the actual standing biomass in the sand dunes of the northwestern Negev (Israel) and probable boundary conditions that may be responsible for the vegetation patterns investigated in detail. Our results for several variables characteristic for the prevailing climate, geomorphology, hydrology and biologicy at four measurement stations along a transect from northwest to southeast allowed for the development of a stochastic model for biomass distribution over the entire sand dune field (mesoscale) and at Nizzana experimental station (microscale). With this equation it was possible to compute and interpolate a biomass index value for each grid point on the mesoscale and micro scale. The spatial distribution of biomass is negatively linked to distance from the sea, to rainfall and relief energy. [ABSTRACT FROM AUTHOR]