Change in ecosystem structure due to widespread conifer mortality in Piñon-Juniper woodlands leads to rapid change in understory abundance detected using both field plots and remotely-sensed data

Background/Question/Methods

The semi-arid Southwestern US suffered a prolonged drought from 2000 to 2003 resulting in the selective, widespread mortality of piñon pine (Pinus edulis) across the region. This dramatic ecosystem restructuring significantly altered ecosystem processes such as carbon and water balance. Drought is expected to increase in frequency and severity in future climate scenarios, and our question is how the understory community structure in these disturbed ecosystems will change as a result. We quantified the understory vegetation in two piñon-juniper (Juniperus monosperma) (PJ) woodlands in central New Mexico <5 km apart, one intact woodland (control site) and one woodland where mature P. edulis trees were girdled in 2009 to simulate the large scale mortality event (girdled site). We used: 1) vegetation density and herbaceous cover or biomass in quadrats under ("canopy") and between ("interspace") mature living ("control") or girdled P. edulistrees one year before and up to two years after the manipulation; 2) basal vegetation cover along a 100m transect in each site recorded each spring and fall over multiple years following the manipulation; and 3) a time series of 36 RapidEye satellite images (5m spatial resolution) coincident in space with our plot level sampling campaigns for our analyses.

Results/Conclusions

Annual forb density in the girdled plots was half that of the control site (p<0.05) prior to manipulation, but switched to being twice that of the control site after two years (p<0.05). Although forb density increased following P. edulis mortality, we observed no significant differences in forb cover between the control and girdled sites after two years, suggesting the responses are due to increased germination at the girdled site. Contrasting with the rapid annual forb response to girdling, we observed no significant differences in perennial grass or CAM species' cover between the control and girdled sites. This contrast is likely due to life histories differences, where annuals can respond more quickly to changing conditions. Total basal cover of all groups was similar over time. Phenological assessment using the RapidEye time series showed that the relative magnitude of understory green-up at the girdled site has increased relative to the control. The apparent isolated green-up observed in pixels dominated by girdled vegetation corroborates the increase in forb germination at the girdle site. The overstory reduction from P. edulis mortality had significant impacts on the germination of annuals in these woodlands and will continue to alter ecosystem function.