Research

Differential Wildfire Damage Based on Home Price in Southern California Wildfires.

I am exploring the hypothesis that urban infill will benefit biodiversity (by leaving land for nature), mitigate wildfire damage (by reducing ignition in the wildland urban interface), and increase fire resiliency in socio-economically disadvantaged human communities. Building wildfire prediction maps, we found that low density residential development is at greater risk of wildfire damage in California (Syphard et al 2018, GEC). To explore whether economically disadvantaged communities experience greater fire damage, we combined remotely sensed structure loss data with high-resolution home price (Zillow) data. Correcting for risk factors (such as slope and road density), we found that lower priced homes are more likely to be damaged in a wildfire. The next step is to determine why this might be the case, exploring the potential for increasing wildfire to emerge as an additional environmental justice concern.

Integrating Remote-sensing and Ecological Forecasting to Optimize Multiple Benefits in Wetland Management in the Central Valley of California.

The California Central Valley, once an expansive network of seasonal wetlands, is now heavily managed, extensively irrigated agriculture with little available water for wildlife. Building on a successful management project where select farmers are paid to flood their fields to provide migratory birds with water (see press on Bird Returns project), I am combining remote-sensed (Landsat-derived) water maps, NDVI-derived crop yields (unharvested grain provides bird food), climate projections, and opportunistically collected (eBird) and rigorously designed (Pacific Flyway Survey and Nature Conservancy) bird data to predict seasonal spatially-explicit bird abundances. Our hypothesis is that by using real-time environmental data we will build models of bird abundance that can seasonally predict the impacts of drought, improving management of where and when farmers should flood their fields.

Understanding the limitations to upwards shift in treeline

Together with Alpine Tree Warming Experiment collaborators at UC Merced and the Lawrence Berkeley Lab, I analyzed five years of seedling recruitment in a common garden, artificial watering and warming experiment (by suspended infrared heaters) in the Colorado Rocky Mountains. We found that warming did not benefit any of our study species - Engelmann spruce, lodgepole pine, or limber pine - in plots across three elevations: an alpine, subalpine, and forested sites. Because watering was beneficial across sites, we hypothesize that seedlings were moisture limited. Combining our experimental results with mechanistic population models, we found that low elevation populations declined more faster than comparable populations emerged in high elevation sites, especially for Engelman spruce. Our results portend transient range contractions and potential species composition shifts with climate change. In the future, I would like to explore the combined effects of fire and experimental warming on lodgepole pine.

Linking Population Dynamics and Species Distribution Models

The direct impacts of climate change combine with indirect impacts (such as fire frequency and altered inter-specific competition) and other threats (such as land use change and exotic plant invasion) to impact biodiversity worldwide. I use species distribution models (SDMs), demographic models and models which combine these components to investigate the vulnerability of populations to multiple threats. Working in the Regan lab, in collaboration with Janet Franklin and Alexandra Syphard, we found that models that included demographic constraints showed larger impacts of climate change than predictions based on SDMs alone for threatened Engelmann oak (Quercus engelmannii). Further, in models for San Diego thornmint (Acanthomintha ilicifolia) we found that models were most sensitive to SDM choice, compared to choice of demographic rates. I used these models again to determine connectivity networks for focal wildlife species in Southern California, finding that many species are limited in their ability to disperse to keep pace with climate change. Working with an interdisciplinary SDSU team, this connectivity work is being incorporated into the State's climate-, fire-, and water-smart planning.

The impact of species and functional diversity on long-term restoration

At the Institute for Conservation Research, I worked with a UCSD student intern co-author to build spatially explicit fire predictions coupled with cactus wren population models to determine the optimal site (Lake Hodges) for habitat restoration for the coastal cactus wren. Because native shrubs at the site had been slow to recover following a massive fire in 2007, we wanted to know if restoration success would benefit from greater species and functional diversity. To explore this question, we manipulated the number of shrub species (1, 2, 4 or 8) transplanted across different plant traits (specifically, bloom time and drought deciduousness) in large scale restoration plots. Coming on or fifth year since planting, restoration success will be measured by transplant survival and native and exotic forb richness and prevalence.