Author : Kyle A. Taylor
Publisher :
ISBN 13 : 9781339440484
Total Pages : pages
Book Rating : 4.4/5 (44 download)
Book Synopsis Exploring Ecological Risks Facing Regional Sagebrush Ecosystems in the 21st Century by : Kyle A. Taylor
Download or read book Exploring Ecological Risks Facing Regional Sagebrush Ecosystems in the 21st Century written by Kyle A. Taylor and published by . This book was released on 2015 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: Big sagebrush (Artemisia tridentata) ecosystems have experienced loss over large portions of their historic range throughout the intermountain western United States over the last two centuries. Leading into the 21st century, much of this loss could be attributed to anthropogenic land conversion -- particularly from regional population growth, agricultural development, and the spread of invasive species. Over the course of the 21st century, climate change and invasive grassland conversion (Bromus tectorum) are expected to further alter the structure of regional sagebrush ecosystems, potentially contributing to range loss for big sagebrush and its many obligate species. My overall objective was to investigate the exposure of sagebrush ecosystems to risks from climate change and invasive species spread. I developed a series of species distribution models (SDMs) for three major sagebrush subspecies and compare the trajectory of regional sagebrush ecosystems under future climate conditions. I assessed my confidence in range projections for big sagebrush under future climate conditions using a sensitivity analysis to explore deficiencies in the underlying data behind my SDMs. Lastly, I attempted to assess the invasibility of climatically-stable sagebrush ecosystems by developing and deploying a process-based demographic model for cheatgrass at sites throughout the intermountain region. I implemented a series of big sagebrush species distribution models (SDMs) fit to individual subspecies records of big sagebrush. I used climate variables previously described as limiting for big sagebrush as explanatory data and extrapolate the models into mid-century and end-of-century conditions using an ensemble of recent global circulation model (GCM) data. I expected all three big sagebrush subspecies to respond negatively to increases in temperature and drier summer conditions forecasted for much of the intermountain region, assuming that future environmental conditions would warm beyond the environmental tolerances to which big sagebrush is physiologically adapted to. However, at the subspecies level, this did not appear to be the case. Wyoming big sagebrush (Artemisia tridentata ssp. wyomingensis) and basin big sagebrush (Artemisia tridentata ssp. tridentata) both demonstrated slight range expansion, largely along elevational gradients, under end-of-century forecasts across GCMs. Although experiencing loss within southern portions of their current range, the warmer winter conditions and slightly drier summer conditions forecasted for wyomingensis and tridentata are not so unusual that the subspecies don’t currently encounter these conditions within portions of their current range. Notably, mountain big sagebrush (Artemisia tridentata ssp. vaseyana) does respond negatively to future climate conditions, with winter temperatures increasing beyond what the subspecies experiences across its current range. Unlike ssp. tridentata and wyomingensis, vaseyana does not appear to have a large elevational gradient to retreat into under future climate conditions, which could result in potentially large losses of its current range if the subspecies is unable to adapt and compete under warmer conditions. In order to better indicate confidence in SDM predictions for big sagebrush, I used a Monte Carlo-based sensitivity analysis to demonstrate how uncertainties attributed to occurrence, absence, and climate data biases propagate through two commonly used SDM algorithms (Generalized Linear Models [GLMs] and Random Forests [RF]) for sagebrush ecosystems dominated by ssp. wyomingensis and tridentata. I derived predictive intervals for both GLM and RF and used the intervals to produce envelopes indicating areas of high and low predictive confidence in SDM predictions and explore how bias can contribute to latitudinal and elevational drift in suitability predictions. Both algorithms demonstrated sensitivities to climate and presence-record uncertainty that translated to large geographic uncertainty in predictions. I found that RF is more sensitive to bias than GLM, and that climate and presence record bias contributed to greater predictive ambiguity than absence record uncertainty. I found that uncertainties in predictions arising from data uncertainty could easily produce the same magnitude of range shift as those observed under mid-century (2050) predictions of distributional change for big sagebrush under climate change, suggesting that mid-century forecasts of change for big sagebrush are not so unusual that they couldn’t also be explained by biases affecting underlying data. However, many late-century (2100) predictions of suitability change appear to diverge from what was observed in our sensitivity analysis and deserve greater confidence. As roughly half of climatically suitable space available to big sagebrush throughout the intermountain region has experienced some degree of invasion by cheatgrass, it’s continued spread is an important factor to consider in assessing the long-term viability of big sagebrush ecosystems. Using a soil hydrology model (SOILWAT) and a meta-analysis conducted from a number of studies of cheatgrass’ demography and physiology, I built a process-based model that simulates cheatgrass population growth over time. I applied the model to current climate conditions at a number of sites sampled from sagebrush ecosystem subspecies strata across the intermountain region and found detectable differences in population suitability response. The cheatgrass demographic model demonstrates detectable differences in population response between montane sites (suitable for ssp. vaseyana) and basin sites (suitable for ssp. wyomingensis and tridentata) that suggest basin subspecies ranges are more suitable for cheatgrass than montane sites. I also demonstrate that basin subspecies are typically closer to known invasive annual grasslands throughout the intermountain region, and are probably more susceptible to cheatgrass propagule pressure. A synthesis of results across my studies suggests differing fates for regional sagebrush ecosystems. Although ssp. wyomingensis and tridentata dominated sagebrush ecosystems may be potentially robust to future climate change, they demonstrate a greater risk to cheatgrass invasion, which could challenge the long-term stability of sagebrush ecosystems in the basin interior of the intermountain region. Conversely, although montane sagebrush ecosystems dominated by ssp. vaseyana may be at comparatively less risk of cheatgrass invasion, these sagebrush ecosystems will face greater exposure to climate change stressors over the course of the 21st century.
