Site characterization and change over time in semi-arid grassland and shrublands at three parks?Chaco Culture National Historic Park, Petrified Forest National Park, and Wupatki National Monument: Upland vegetation and soils monitoring 2007?2021

Abstract

This report presents results of upland vegetation and soil monitoring of semi-arid grasslands at three Parks by the Southern Colorado Plateau Inventory and Monitoring Network (SCPN) from 2007?2021. The purpose is to compare and contrast five grassland ecological sites and examine how they have changed during the first 15 years of monitoring. Crews collected data on composition and abundance of vegetation, both at the species level and by lifeform (e.g., perennial grass, shrub, forb) and soil aggregate stability and soil texture at 150 plots within five target grassland/shrubland communities delineated using NRCS ecological site (ecosite) classification (30 plots per ecosite). Soils in plots at Petrified Forest NP and Chaco Culture NHP were deeper than those at Wupatki NM. Undifferentiated soil crust comprised the largest component of the soil surface, except at Wupatki where surface gravel dominated. Cover of biological soil crust (cyanobacteria, lichen, and moss) was low. Soil aggregate stability was moderate. From 2007?2021, SCPN crews identified 283 unique plant species. Overall live foliar cover ranged from 12-24%. Four of five ecological sites were dominated by C4 grass species (>70% of total live foliar cover). Shrubs co-dominated at one site (WUPA L) and forbs were an overall small component of total vegetation cover but contributed most of the diversity in these sites. Less than 4% of species detected were nonnative. Russian thistle (Salsola tragus) was the most frequently sampled nonnative, occurring in > 50% of plots at Wupatki in the volcanic upland ecological site. Cheatgrass (Bromus tectorum) was the second most common invasive species but occurred in < 10% of the plots at all ecological sites. Vegetation cover was modeled using Bayesian hierarchical models and included seasonal climatic water deficits, year effects and topographic variables as covariates. Models revealed significant negative time trends (i.e., changes over time that were not explained by changes in seasonal deficit covariates included) in some modeled responses, particularly in the cover of perennial grass at all five ecological sites. Time trends in shrub and forb responses were mixed. Species richness showed variable effects by ecosite, decreasing at CHCU S, and increasing at PEFO S and WUPA V. Modeled responses were influenced by climate covariates, but direction of these effects varied. The most consistent effects were that greater July water stress and higher accumulated growing degree days (i.e., warmer spring temperatures) increased cover of perennial grasses and shrubs during the same year. However, greater water stress in the spring had a negative effect on many responses as expected. Decreasing cover of perennial grass and increasing cover of shrubs and weedy forbs has been predicted for southwestern grasslands in response to increasing aridification due to anthropogenic climate change. Perennial grass trends reported here correspond with these predictions with mixed results on shrub and forb community trends. Continued drought conditions will likely exacerbate negative changes in these systems.