Dry-season grazing with small ruminants enhances native diversity in invaded oak woodlands
Data files
Oct 24, 2025 version files 150.98 KB
-
README.md
7.04 KB
-
species_data.csv
9.65 KB
-
veg_data.csv
134.28 KB
Abstract
Oak woodlands are biodiversity hotspots that provide critical ecosystem services and hold significant cultural value. However, these ecosystems are increasingly threatened by invasive grasses, which suppress native plant diversity, elevate fire risk, and disrupt ecosystem processes. While targeted grazing is gaining popularity as a management technique for controlling invasives, the ecological impacts of dry-season grazing are understudied, as are outcomes for native species.
We evaluated the effectiveness of targeted summer grazing with small ruminants for reducing invasive grass fuel loads and enhancing plant diversity in oak woodlands of California by comparing plant species richness and abundance in paired grazed and ungrazed plots over multiple years.
Grazing substantially increased richness and cover of native forbs. Nonnative cover was unaffected, though minor increases in nonnative forb numbers were observed. Ordination analysis revealed that grazing shifted plant community composition, largely driven by increases in native forb species and to a lesser extent nonnative forbs. Indicator species analysis revealed similar patterns, with increases in multiple forb species detected with grazing, highlighting the potential for grazing to support a diversity of plant species. These benefits likely stem from reduced thatch accumulation, improved light availability, and the creation of favorable microsites for seedling establishment.
Practical application: Dry-season grazing with goats and sheep presents a multi-benefit strategy for land stewardship, helping to reduce invasive grass thatch, enhance native plant diversity, and lower fire risk in invaded woodlands. While grazing alone is unlikely to effectively suppress invasive grasses and may encourage certain nonnative forbs, it can dramatically improve the ability of native species to coexist with nonnatives. To maximize benefits, practitioners should integrate grazing with complementary management strategies when controlling nonnative plants is a priority. Careful management of grazing intensity, timing, and duration, coupled with long-term monitoring, will aid in balancing these ecological trade-offs. Incorporating targeted dry-season grazing into a broader restoration framework could aid in the conservation of biodiversity in grass-invaded dryland ecosystems worldwide.
Dataset DOI: 10.5061/dryad.hhmgqnkvr
Description of the data and file structure
Vegetation monitoring was conducted at the Wantrup Preserve in Napa County, California—a 730-acre mosaic of blue oak (Quercus douglasii) woodlands and interspersed annual grasslands managed by the Land Trust of Napa County. The experiment was designed to assess the long-term effects of targeted grazing on native and nonnative plant communities within this invaded oak woodland ecosystem. Paired 1 m² plots (one grazed, one ungrazed) were established across the preserve, initially on an interior grid and later along accessible perimeter transects. Ungrazed plots were enclosed with fencing, while grazed plots remained open to livestock. Goats were used for grazing from 2016 to 2019 and sheep from 2020 to 2022, with treatments applied each summer to remove accumulated dry biomass.
Vegetation surveys were conducted annually each spring from 2019 to 2022 to quantify plant species richness and cover following the previous year’s grazing treatment. Percent cover and species identity were recorded for all vascular plants, with supplemental sampling later in the season to capture late-flowering forbs and geophytes. In total, data were collected from 176 grazed–ungrazed plot pairs (352 total plots). The resulting dataset provides a robust, multi-year record of plant community composition and structure under different grazing regimes within a Mediterranean-type oak woodland system.
Files and variables
File: veg_data.csv
Description: .csv file containing raw data on species abundance (i.e., percent cover), as well as abundance and richness by functional group (e.g., native or nonnative, forb or grass)
Plot definitions: definitions of plot properties (Columns A-G), including plot identifiers, plot type, year of sampling, type of ruminant used for grazing, and the grazing interval/history of the plot.
| Column name | Explanation |
|---|---|
| index_g | Unique identifier for each plot and sampling time |
| plot_type | Plot type (P = plots installed at the permimeter of fencing between grazed and ungrazed areas; W = original plots installed within the interior of the Wantrup Preserve's grazed area) |
| plot_name | Unique plot identifier |
| year | Year of vegetation sampling |
| treatment | Plot treatment (grazed or ungrazed) |
| grazer | Ruminat type used for grazing (sheep or goats) |
| interval | Grazing interval (i.e, newly added plots, plots grazed one year prior, or plots grazing two years prior) |
Vegetation definitions: definitions of vegetation metrics analyzed (Columns H-S), including the abbreviation used, the type of metric (plant functional group abundance or richness), nativity of the native functional group (native or nonnative), and type of plant functional group (all species, forbs, or grasses)
| abbreviation | metric | nativity | functional group |
|---|---|---|---|
| a_nat_all | abundance | native | all |
| a_nat_frb | abundance | native | forb |
| a_nat_grs | abundance | native | grass |
| a_non_all | abundance | non-native | all |
| a_non_frb | abundance | non-native | forb |
| a_non_grs | abundance | non-native | grass |
| r_nat_all | richness | native | all |
| r_nat_frb | richness | native | forb |
| r_nat_grs | richness | native | grass |
| r_non_all | richness | non-native | all |
| r_non_frb | richness | non-native | forb |
| r_non_grs | richness | non-native | grass |
File: species_data.csv
Description: .csv file containing raw data on species codes used in veg_data.csv file, scientific names, and information on origin (native or nonnative) and functional group (grass or forb).
species_data: definitions of all plant species' codes (gen_spec), the full genus and species (genus*species), and whether the species is native or not (is_native; TRUE or FALSE), a forb (is_forb; TRUE or FALSE), or a grass (is_grass: TRUE or FALSE)
Code/software
All analyses were conducted using R version 4.4.1 (R Core Team 2024), an open-source statistical computing environment freely available at https://www.r-project.org. The analyses can be fully reproduced using R and the following packages:
- lme4 and glmmTMB — for fitting generalized linear mixed models (GLMMs) with normal and negative binomial error structures.
- MuMIn — for information-theoretic model selection using Akaike’s Information Criterion (AICc).
- DHARMa and performance — for diagnostic checks of residuals, overdispersion, and model singularity.
- car, vcd, pscl, and fitdistrplus — for exploratory data analysis, distribution testing, and model assumption validation.
- vegan — for community ecology analyses, including non-metric multidimensional scaling (NMDS), environmental vector fitting, and permutational multivariate analysis of variance (PERMANOVA).
- indicspecies — for indicator species analysis using abundance-based association tests.
Data visualization, summary statistics, and model diagnostics were performed within RStudio (Posit, version 2024.04) using base plotting functions and ggplot2.
Access information
This is original data collected by the authors.
