Responses of wind erosion to climate-induced vegetation changes on the Colorado Plateau

Abstract

Projected increases in aridity throughout the southwestern United States due to anthropogenic climate change will likely cause reductions in perennial vegetation cover, which leaves soil surfaces exposed to erosion. Accelerated rates of dust emission from wind erosion have large implications for ecosystems and human well-being, yet there is poor understanding of the sources and magnitude of dust emission in a hotter and drier climate. Here we use a two-stage approach to compare the susceptibility of grasslands and three different shrublands to wind erosion on the Colorado Plateau and demonstrate how climate can indirectly moderate the magnitude of aeolian sediment flux through different responses of dominant plants in these communities. First, using results from 20 y of vegetation monitoring, we found perennial grass cover in grasslands declined with increasing mean annual temperature in the previous year, whereas shrub cover in shrublands either showed no change or declined as temperature increased, depending on the species. Second, we used these vegetation monitoring results and measurements of soil stability as inputs into a field-validated wind erosion model and found that declines in perennial vegetation cover coupled with disturbance to biological soil crust resulted in an exponential increase in modeled aeolian sediment flux. Thus the effects of increased temperature on perennial plant cover and the correlation of declining plant cover with increased aeolian flux strongly suggest that sustained drought conditions across the southwest will accelerate the likelihood of dust production in the future on disturbed soil surfaces.

Fig. 1.

Map of the long-term vegetation plots and Western Regional Climate Center (WRCC)/National Oceanic and Atmospheric Administration (NOAA) weather stations in the four national park areas on the Colorado Plateau in southeastern Utah.

Fig. 2.

Regional mean annual temperature (°C) from 1988 to 2008 for all WRCC/NOAA weather stations in the national park areas with fitted piecewise regression (P < 0.0001) (A). Annual precipitation (mm) from 1988 to 2008 for all WRCC/NOAA weather stations in the national park areas and 20-y mean annual precipitation (MAP) (B).

Fig. 3.

One-hour peak wind speeds (m s^-1) at 3 m height at the Dugout Ranch near the study area from 1998 to 2008. Wind speeds of 15.0–25.0 m s^-1 used in this study shown by dashed lines. Two day time period in mid-April 2002 highlighted by gray box and corresponding hourly Sensit counts, which are an index of sediment movement. Data obtained from http://gec.cr.usgs.gov/info/sw/clim-met.

Fig. 4.

Dominant plant species and functional type canopy cover (A, C, E, and G) and modeled aeolian sediment flux (B, D, F, and H) at five wind speeds (15.0, 17.5, 20.0, 22.5, and 25.0 m s^-1) in relationship to mean annual temperature in the previous year in perennial grasslands (N = 6; A and B), shrublands dominated by Coleogyne ramosissima (N = 6; C and D), Artemisia tridentata and Sarcobatus vermiculatus (N = 4; E and F), and dwarf Atriplex species (N = 4; G and H). Significant (P < 0.05) regressions shown by an asterisk (*) for (A) Perennial grasses: y = -8.04x + 119.1, r² = 0.28; all perennial vegetation: y = -6.57x + 113.2, r² = 0.25; aeolian sediment flux: y = 0.02e^0.54x, r² = 0.26 (15.0 m s^-1); y = 0.77e^0.36x, r² = 0.22 (17.5 m s^-1); y = 3.65e^0.31x, r² = 0.23 (20.0 m s^-1); y = 9.96e^0.29x, r² = 0.24 (22.5 m s^-1), y = 20.7e^0.27x, r² = 0.25 (25.0 m s^-1). (C, Inset) Artemisia tridentata: y = -2.03x + 29.8, r² = 0.49; Sarcobatus vermiculatus: y = -4.60x + 70.5, r² = 0.51, all perennial vegetation: y = -7.09x + 108.7, r² = 0.57; aeolian sediment flux: y = 1.09e^0.40x, r² = 0.24 (22.5 m s^-1), y = 2.36e^0.38x, r² = 0.25 (25.0 m s^-1). (D) Atriplex spp.: y = -1.17x + 17.8, r² = 0.24; all perennial vegetation: y = -2.74x + 43.3, r² = 0.12; aeolian sediment flux: y = 5.35e^0.23x, r² = 0.11 (15.0 m s^-1), y = 31.8e^0.19x, r² = 0.12 (17.5 m s^-1), y = 88.7e^0.17x, r² = 0.12 (20.0 m s^-1), y = 181.3e^0.16x, r² = 0.12 (22.5 m s^-1), y = 313.8e^0.15x, r² = 0.12 (25.0 m s^-1).

Fig. 5.

Aeolian sediment flux following a 17.5 m s^-1 wind event in relation to percent perennial vegetation canopy cover with low (A), medium (B), and high (C) intensity soil disturbances on four plots in dwarf Atriplex shrublands. Plots varied in perennial vegetation cover and average BSC development class (1, low; 6, high).