In a large-scale grass removal experiment (NEAT), we identified several important thresholds that impact the conversion of grasslands to shrublands. Between 75%-100% grass loss, aeolian transport increases dramatically, but carbon and nitrogen in windborne sediment display another threshold between 50%-75% grass loss (Li et al. 2007). With lower grass cover, nutrient additions to the soil are overwhelmed by aeolian emissions, resulting in a net loss of soil nutrients (Li et al. 2008).
The sediment that is deposited downwind of the vegetation is both coarser (Li et al. 2009b) and lower in nitrogen (Li et al. 2009a) than the source sediment. Increased aeolian sediment flux downwind decreases grass cover and increases shrub cover (Alvarez et al. 2011). Wet years increased competition among grasses and decreased competition between grasses and shrubs (Alvarez et al. 2012). These data were used to develop and validate a model of aeolian sediment flux (Okin et al. 2006, Okin 2008).
Hypothesis: As bare gap sizes increase, a connectivity threshold level is reached that sets the stage for nonlinear increases in the spatial extent of shrub dominance owing to negative effects on grass persistence [1(a)] and positive feedbacks to shrub establishment and growth. This hypothesis will be tested on plots established in 2004 in the NEAT where 0, 50, 75, or 100% of original herbaceous cover was removed (and maintained thereafter) from 25 x 50 m plots in each of three blocks (Li et al. 2007). These removals generated varying levels of grass fragmentation with consequent effects on shrub expansion and grass loss in contiguous downwind plots that have been qualitatively observed (Alvarez et al. 2012). We propose to quantify (a) downwind effects at different fetch lengths, (b) vegetation feedbacks on gap size and aeolian transport, and (c) effects of interactions between climate and dust flux on plant mortality through time. We will continue to monitor aeolian sediment flux, vegetation composition (line-point intercept and gap-size distribution; Herrick et al. 2005), individual plants (sensu Alvarez et al. 2011), and soil C and N (sensu Li et al. 2008) and will relate dynamics in these variables to precipitation. Results will be analyzed by ANCOVA with flux, bare gap size, or herbaceous cover as continuous variables. We will also use the wind and vegetation dynamics components of our ENSEMBLE model to identify threshold effects of grass cover on connectivity by wind for different amounts of precipitation, and to determine feedbacks to shrub establishment and growth via changes in the deposition or erosion of soil and nutrients.