Skip to main content

Oak woodland health assessment in Orange County metrics

Cite this dataset

Killian, Kyle (2021). Oak woodland health assessment in Orange County metrics [Dataset]. Dryad.


Oak woodlands are declining in Orange County (OC) due to habitat destruction and climate change, along with increased drought and weather extremes, fire frequency, and the recent invasion by goldspotted oak borers (GSOB). The Center for Environmental Biology (CEB) has been conducting a long-term oak woodland health monitoring program in OC for the goals of restoration. In this project, we analyzed oak monitoring data on demographics and pathogens collected from 2010 to 2020 and incorporated fire frequency and GSOB data to create a health-based oak woodland ranking system to make recommendations for restoration. Here we show our three-level recommendations of restoration priority: 1) Level 1: Blind Canyon (BLI) and Boy Scout Camp (BSC); 2) Level 2: Limestone Canyon (LSC) and Oak Flat (OF); 3) Level 3: Fremont Canyon (FMC), Gypsum Canyon (GYP), and Irvine Mesa (IRM). Increased fire frequency was positively correlated with declining health. We also assessed remote sensing data (Normalized Difference Vegetation Index, NDVI) against on the ground data for its potential to identify declining oak woodlands. NDVI was an imperfect indicator of declining oak woodland health, but may be useful for larger woodlands or with improved imagery resolution. Our work will inform local stakeholders in future restoration projects of OC oak woodlands.


Sampling protocol (from CEB Oak Monitoring Report Feb 29 2020): Within each woodland, random sections, or “reaches,” were identified for sampling. Each reach included clusters of trees that were monitored. Individual Coast Live Oak trees were sampled in “Y” shaped clusters that included three trees that surrounded one randomly selected “sentinel” tree. Clusters were located in woodlands in the coastal and inland protected open space within Orange CountyData processing:

Data from raw CEB data set were cleaned in Excel by removing inappropriate nonnumeric characters and correcting any other errors. Analysis using cleaned data set was performed in R.

Usage notes

Please see ReadMe file.