Mechanisms and Proactive Management of Woody Plant Encroachment on Southwestern Rangelands

Abstract

Woody plant encroachment and the transformation of grasslands and savannas to shrublands and woodlands is actively occurring on rangelands across the Southwest US and globally. Once established, woody plants may be long-lived and highly persistent. The early shrub life cycle is therefore a critical first step in the grassland-to-shrubland transformation process. Little is known regarding the relative importance of abiotic and biotic controls on shrub recruitment and how they might interact. Knowing what factors affect successful shrub recruitment can also inform where and under what conditions additional establishment might occur over time. Land managers use various brush management treatments alone or in combination to stem the tide of woody plant encroachment. These actions are costly and often short-lived, where reactive approaches to encroachment continues to be the primary approach on public and private rangelands. The overarching goal of this dissertation was to explore several potential constraints on shrub recruitment for informing the development of proactive brush management alternatives to reactionary methods.The first study used a model shrub, Prosopis velutina, and a controlled environment to quantify variation in morphological and physiological seedling traits under contrasting watering regimes. Extremely dry conditions reduced germination and slowed seedling growth but did not prevent early establishment. Morphological traits such as root volume, root surface area, root tips, tap root length, and leaf length were key predictors of seedling performance, and no physiological functional traits were found significant in predicting overall seedling performance or discriminating between moisture regimes. The second study expanded the controlled environment study and quantified P. velutina establishment in a 3-year factorial field experiment on the Sonoran Desert Santa Rita Experimental Range (SRER) that manipulated precipitation, seed/seedling predators, and livestock grazing (as simulated via herbaceous defoliation). Seed germination and early seedling survival readily occurred under a wide range of abiotic and biotic conditions. First-year seedling survival was only reduced under low rainfall event frequency with elevated seed/seedling predator abundance. Subsequent seedling survival was not influenced by rainfall treatment, including extreme drought, where herbaceous defoliation slightly decreased, but did not prevent, survival beyond the first year. No support was found for the commonly held assumption that shrub (P. velutina) recruitment in arid/semi-arid grasslands is episodic per se but can readily occur under a wide range of precipitation, herbaceous defoliation/heavy livestock grazing, and seed/seedling predator abundances. The third study built on the above plant- and plot-scale studies and used machine learning to predict landscape-level woody plant cover using geospatial and remote sensing data for developing a decision support framework to assess a site’s susceptibility to woody plant encroachment. Random forests models provided the smallest error and identified the most important determinants of woody cover: elevation, maximum clay percentage, cool-season rainfall, aspect, slope, and distance to nearest drainage/wash. Model-based maximum woody cover compared with current (e.g., a site’s potential to increase in cover) identified 45% of the SRER landscapes as having moderate to high risk for increased cover. The modeling framework identified sites most at-risk for encroachment and could be used to improve planning efforts with respect to the location, type (e.g., prescribed fire), and timing of brush management.