Southern Californian wildfires can influence climate in a variety of ways, including changes in surface albedo, emission of greenhouse gases and aerosols, and the production of tropospheric ozone. Ecosystem post-fire recovery plays a key role in determining the strength, duration, and relative importance of these climate forcing agents. Southern California's ecosystems vary markedly with topography, creating sharp transitions with elevation, aspect, and slope. Little is known about the ways topography influences ecosystem properties and function, particularly in the context of post-fire recovery.

We combined images from the USGS satellite Landsat 5 with flux tower measurements to analyze pre- and post-fire albedo and carbon exchanged by Southern California's ecosystems in the Santa Ana Mountains. We reduced the sources of external variability in Landsat images using several correction methods for topographic and bidirectional effects. We used time series of corrected images to infer the Net Ecosystem Exchange and surface albedo, and calculated the radiative forcing due to CO2 emissions and albedo changes. We analyzed the patterns of recovery and radiative forcing on north- and south-facing slopes, stratified by vegetation classes including grassland, coastal sage scrub, chaparral, and evergreen oak forest. We found that topography strongly influenced post-fire recovery and radiative forcing.

Field observations are often limited by the difficulty of collecting ground validation data. Current instrumentation networks do not provide adequate spatial resolution for landscape-level analysis. The deployment of consumer-market technology could reduce the cost of near-surface measurements, allowing the installation of finer-scale instrument networks. We tested the performance of the Microsoft Kinect sensor for measuring vegetation structure. We used Kinect to acquire 3D vegetation point clouds in the field, and used these data to compute plant height, crown diameter, and volume. We found good agreement between Kinect-derived and manual measurements.

The ability of ecosystems to recover from disturbance likely has limits, and shifting disturbance regimes or altered recovery conditions may cause ecosystem degradation and increase vulnerability to species invasion and type conversion. Human activities affect disturbance-prone ecosystems in many ways, including fire frequency, introduction of invasive species, and shifts in water, nutrients, and other plant resources. This dissertation investigates the effects of altered precipitation and nitrogen input on adjacent coastal sage scrub (CSS) and grassland communities in the Santa Ana Mountain foothills during post-fire recovery.

Chapter two examines the ecohydrology of CSS and grassland to determine their respective water use strategies and investigate how they coexist and differ under similar climates. Deeper roots and access to stable deep-water stores allow CSS to employ a conservative growth and water use strategy, whereas grassland follows a “live fast, die young” strategy with rapid, intense water withdrawal over a shallower depth range and a shorter growing season. While a mature CSS community tolerates the presence of invasive annuals, a disturbance pattern that kills CSS individuals completely and prevents post-fire resprouting may increase vulnerability to invasion.

Chapter three examines the impact of altered resources on CSS successional patterns. Altered water input is an important controller of successional rate for drought-adapted ecosystems, with reduced water decelerating and increased water accelerating succession. The combination of restricted water and added nitrogen may produce a fundamental shift in CSS succession away from a shrub-dominated community, leading to increased vulnerability of CSS to invasion by Eurasian grassland.

Chapter four investigates the transient maximum ecosystem productivity hypothesis. A transient productivity maximum occurs in the CSS independent of nitrogen or water inputs, indicating this phenomenon is likely driven by light availability and dominance by plant species that have been selected to rapidly capture space. Increased nitrogen and reduced water availability both shift the vegetation type responsible for the transient maximum from the native nitrogen-fixing subshrub Acmispon glaber to Eurasian grasses. Increases in drought and nitrogen deposition may reduce the importance of A. glaber during CSS fire recovery, with cascading effects on ecosystem function.

Global environmental change is impacting ecosystems. Drought, nitrogen deposition, and invasion by annual grasses (GL) are causing extensive loss of Southern California's coastal sage scrub (CSS) community. Warmer temperatures, changes in precipitation, and altered fire regimes are impacting interior Alaska’s boreal forest function and productivity. In this dissertation, I explore these two contrasting ecosystems with the goal of developing a mechanistic understanding of how ecosystems respond to multiple global environmental change drivers.

Chapter 2 examines an apparently stable vegetation boundary between CSS and GL. This chapter utilizes historical aerial photography and measures of above and belowground properties to examine how vegetation type impacts soil development. CSS had more deeply distributed root biomass, denser surface soil, and greater total stocks of soil organic carbon and nitrogen than GL. Vegetation type creates unique soil properties that may reinforce the stable boundary by preventing invasion of vegetation with different growth, rooting, and/or water-use strategies. Chapter 3 explores how CSS and GL respond to drought and added nitrogen. CSS was more sensitive to global change than GL and drought was the main driver of ecosystem change.

In CSS, drought and drought × +N reduced native shrub biomass and altered litter composition, consistent with a vegetation shift from CSS to GL. These changes were mirrored belowground, with reductions in soil bulk density, soil organic carbon content, and soil C/N ratio in CSS drought plots. The surface soil properties in CSS drought plots were more similar to GL than to CSS, indicating that drought induced vegetation shifts from CSS to GL impact ecosystem properties and processes, both above and belowground.

Chapter 4 combines measures of remotely-sensed NDVI with forest surveys to explore the drivers of recent NDVI trends. Recent fire led to sharp NDVI decreases, while early recovery of deciduous species led to NDVI increases. The mid-succession transition from deciduous to evergreen forests led to weak NDVI decreases, and the mid-to-late successional thinning of evergreen canopies to weak NDVI increases. Thus, both increasing and decreasing NDVI stands occur naturally across the landscape, and do not necessarily reflect a large-scale shift in boreal forest productivity.