Deserts and semi-deserts are characterized as ecosystems of low primary productivity, but a more pertinent trait is the variability of plant production in space and time. In dryland ecosystems there are usually some locations on the landscape (e.g., sites receiving run-on moisture from adjacent areas) and some time periods where plant growth is lush in response to temporary favorable conditions.
Our objective in monitoring net primary production, therefore, is to understand the temporal and spatial patterns of production within ecosystems as well as to characterize the relative productivity of various ecosystems.
Our chief questions pertain to the impact of desertification (shrub encroachment in former semi-arid grasslands) on patterns of production. First, can we detect significant differences in productivity between shrub-dominated and grass-dominated systems? Second, given our evidence for greater spatial heterogeneity in soil resources at the plant scale in shrub systems, how does the spatial heterogeneity of aboveground production compare among ecosystems? Finally, how do seasonal and interannual patterns of plant production response to climate vary among ecosystems?
Our methodology for measuring aboveground net primary production has been adopted to allow explicit comparison of production among ecosystems of different structure and to facilitate assessment of spatial variation. We have designed a non-destructive method that samples vegetation of different structures with consistent methodology and intensity of sampling, and that allows the quantification of NPP for particular unit areas such that one can characterize spatial patterns in NPP.
In early 1989 grids of permanent 1 m2 quadrats were established in 15 sites: 3 each in Larrea shrubland, Bouteloua eriopoda grassland, Prosopis dune systems, Flourensia cernua alluvial flats, and grass-dominated dry lakes or playas. Sites were selected to represent the range of biomass and vegetation structure within each ecosystem type, rather than randomly selected among sites of that ecosystem type in the basin (Figure 1). Aboveground biomass is estimated for every species in 49 quadrats per site, three times per year, using non-destructive measures of plant size and applying regressions based on harvests from adjoining areas. These data provide quantitative measures of the abundance of all vascular plant species at each site every season. Productivity over an interval for a quadrat is estimated as the sum of all positive increments of biomass for all species in that quadrat.
At each of the 15 sites we are also measuring soil moisture (via neutron probe) monthly. Rain gauges at each site provide localized precipitation data to accompany weather data from the central meteorological station.
Annual ground-based photos from the 4 corners of each of the 15 NPP sites began in 1996 to provide a qualitative view.