MCA Pilot PUI: Evaluating Climate Drivers of Plant Demography

Measuring the effects of weather conditions on plant populations is critical to our collective understanding of ecological communities and how they will respond to climate change. Increasingly, researchers approach this problem by developing statistical models that predict plant survivorship, growth, and reproduction from weather data. The results can then be used to forecast population growth under different potential future climates.

Still, identifying the most significant ways that climate affects plants is challenging. Many different components of weather could be important, from average yearly temperature and rainfall to extreme heat or water stress over a short time. Many environmental changes are happening at the same time, and separating the effects of weather from other factors can be difficult.

Also, multiple potential methods could be used, and few studies have evaluated how the choice of method might influence the results. This project will analyze long-term data from multiple plant species in coastal California to compare methods for modeling the effects of weather variation on survival, reproduction, and population growth. The research will provide important new information about the strengths and limitations of different approaches for predicting weather effects on plant populations, as well as about the effects of climate changes such as increased drought on this ecosystem.

The project will train a mid-career scientist and undergraduate students in recently developed data analysis methods, expanding access to key skills.

This research will analyze 17–27-year long data sets for six species from the northern California Channel Islands (CI),10-years of data for a seventh species from a nearby mainland site, and artificially constructed data simulated from known parameters. The analyses will quantify correlations among different potential weather metrics, assess whether correlations are changing, and determine which metrics show trends that could be confounded with other causes of population decline.

Alternative statistical models linking weather variables to demographic rates will be developed for each of the focal plant species, with multiple methods grounded in likelihood-based metrics such as AIC or in cross-validation approaches. The results of these alternative methods will be compared based on which variables and time windows are included, the strength of relationships identified for different parameters, and the strength of support for the models.

Finally, the project will develop demographic population models for two of the species, testing sensitivity of model predictions under different climate scenarios to the methods used for parameterizing links with weather variables. Analyses will be used to determine which demographic rates are most important to variation in model predictions and uncertainty in driver estimation.

This work will provide valuable resources for researchers navigating the complex diversity of statistical approaches available to model weather effects on plant demography, in addition to new insights into the climate drivers most important to ongoing and future changes in coastal California plant communities.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.