Effects of proximity to abandoned livestock corrals on standing grass biomass, grass species diversity, and wildlife use in Olare Motorogi Conservancy, Maasai Mara, Kenya
Data files
Jun 11, 2026 version files 133.30 KB
-
Biomass-dry_and_wet.csv
1.88 KB
-
Boma_Diameter__perimeter.csv
605 B
-
dung_dry__wet_density.csv
3.62 KB
-
feeding_guilds_dry__wet__dung_density.csv
2.95 KB
-
grass_category__species_biomass_seasons.csv
60.30 KB
-
Livestock_corrals.R
48.20 KB
-
README.md
9.11 KB
-
Seasons___Biomass___dung_density.csv
2.10 KB
-
species_abundunace_dry__wet_seasons.csv
1.03 KB
-
species_category_biomass_seasons.csv
1.40 KB
-
table1_dry__wet.csv
2.11 KB
Abstract
Pastoralists in African savanna ecosystems are shifting from livelihoods based purely on livestock only to a blended approach of livestock keeping and wildlife conservation for eco-tourism. A popular feature of this model is the establishment of community conservancies that encourage co-existence of traditional livestock keeping with wildlife ecotourism. Pastoralists in these ecosystems periodically establish livestock corrals which they later abandon in search of fresh pasture. The impact of these abandoned corrals on trees and shrubs is fairly well documented, but few studies exist on grasses. To complement these studies, we examined the effects of proximity to five-year-old abandoned livestock corrals on standing grass biomass, grass species diversity, and wildlife use in Olare Motorogi Conservancy, Maasai Mara, Kenya, over a period of 6 months. Using a distance-based criterion, we compared close (5–55 m) and far (150–200 m) belts during wet and dry seasons. The standing grass biomass was consistently higher in close belt than far across both seasons. Biomass was significantly higher during wet season compared to dry season. Decreaser species showed the highest biomass close to corrals, especially during the wet season, indicating strong nutrient-driven productivity. However, grass species richness and Shannon diversity were higher in far belts (18 species; H′ = 2.03) than close belts (14 species; H′ = 1.43), indicating the dominance of a few competing species in close belt. Wildlife dung density, used as a proxy for habitat use, was significantly higher in close belt, especially in dry season. Biomass declined with increasing dung density (β = −0.0156 ± 0.0036 SE, p < 0.001), with stronger grazing effects close to corrals. These findings demonstrate that abandoned corrals function as ecological hotspots that enhance productivity, influence species composition, and shape wildlife distribution, highlighting their potential role in adaptive rangeland management and sustainable livestock–wildlife coexistence.
Dataset DOI: 10.5061/dryad.08kprr5j6
Description of the data and file structure
The data were collected to investigate how proximity to abandoned livestock corrals influences standing grass biomass, grass species diversity, and wildlife use in the Olare Motorogi Conservancy within the Maasai Mara ecosystem. Field sampling was conducted across close (5–55 m) and far (150–200 m) distance belts from livestock corrals during both wet and dry seasons. Measurements included standing grass biomass, grass species richness and diversity, forage category composition, and wildlife dung density as an indicator of habitat use. The study aimed to assess how nutrient enrichment associated with livestock corrals affects vegetation productivity, plant community structure, and wildlife distribution in savanna rangelands.The dataset focuses on the ecological impact of bomas (traditional livestock enclosures) on vegetation dynamics (grass biomass, species categories, and abundance) and wildlife dung density across different distance zones and seasons.
1. feeding_guilds_dry__wet__dung_density.csv
Description: This file contains data on the dung counts and dung density of various wildlife species (both herbivores and carnivores). The data is categorized by the species' feeding guilds and stratified by distance zones from the bomas and by season (dry and wet).
Variables:
-
Species: Common name of the observed animal species (e.g., Buffalo, Elephant, Impala, Wildebeest, Hyena).
-
Feeding guilds categories:
Grazers,Browsers,Mixed feeders,Carnivores. -
Belt: Distance zone from the center of the boma where the observation/count took place (m).
-
season: The season during which the data was collected.
-
dung counts: The total number of dung piles counted for the given species within the specified belt and season.
-
Dung density: The calculated density of dung piles per sampling unit or plot area.
2. grass_category__species_biomass_seasons.csv
Description: This file contains raw, plot-level biomass data for various grass species. The grasses are categorized by their ecological response to grazing pressure, and the data is stratified by distance from the bomas and season.
Variables:
-
Species: Two-letter abbreviation representing the specific grass species (e.g., ET, CD, TT, PM, SS, BI, AK, SA, CP, HS).
-
Category: Ecological category indicating the grass species' response to grazing pressure and disturbance.
-
Belt: Distance zone from the boma.
Close (5-55),Far (150-200). -
season: Season of data collection.
dry,wet. -
Biomass: standing grass biomass of the specific grass species (Unit: kg/ha ).
3. dung_dry__wet_density.csv
Description: This file provides similar data to the first file regarding wildlife dung counts and density across feeding guilds, distance zones, and seasons. However, it additionally includes the standard error for the calculated dung densities to provide a measure of statistical dispersion.
Variables:
-
Species: Common name of the animal species.
-
feeding guilds: Dietary classification of the species (
Grazers,Browsers,Mixed feeders,Carnivores). -
Belt: Distance zone from the boma (
Close (5-55),Far(150-200)). -
season: Season of data collection (
dry,wet). -
dung counts: Total number of dung piles counted.
-
Dung density: Calculated density of dung piles per sampling unit/plot.
-
SE: Standard Error of the mean dung density.
4. Biomass-dry_and_wet.csv
Description: This file contains total grass biomass measurements aggregated at various specific boma locations. The data is categorized by distance from the boma center and season.
Variables:
-
Location: Name of the specific boma location where data was collected (e.g., Ekisemei, Iseketa, Lepore, Matani).
-
season: Season of data collection
dry,wet. -
Belt: Distance zone from the boma.
-
Categories:
Close (5-55),Far (150-200). -
Biomass: Total measured grass biomass at the specific location and belt (Unit: kg/ha ).
5. Boma_Diameter__perimeter.csv
Description: This file contains the geographical coordinates and physical dimensions (diameter and perimeter) for the boma locations included in the study.
Variables:
-
Boma Location: Name of the boma.
-
GPS_Y: Latitude coordinate of the boma (Unit: Decimal degrees).
-
GPS_X: Longitude coordinate of the boma (Unit: Decimal degrees).
-
Approximate Diameter: Approximate diameter of the circular/oval boma enclosure (Unit: meters).
-
approximate perimeter: Approximate perimeter of the boma enclosure (Unit: meters).
6. species_category_biomass_seasons.csv
Description: This file contains aggregated biomass data for specific grass species, categorized by their ecological response to grazing pressure, stratified by distance from the bomas and by season. (Note: This file represents summarized values per species/belt/season combination, unlike the raw plot-level data in file 2).
Variables:
-
species: Two-letter abbreviation for the grass species.
-
category: Ecological category indicating the grass species' response to grazing pressure:
Increaser I,Increaser II,Decreaser. -
Belt: Distance zone from the boma.
Close (5-55),Far (150-200). -
seasons: Season of data collection.
dry,wet. -
Biomass: Measured above-ground biomass of the specific grass species (Unit: kg/ha ).
7. species_abundunace_dry__wet_seasons.csv
Description: This file records the abundance or count of specific grass species, further broken down by plant condition or phenological stage (mature plants, brown/senescent material, and green/live material) across different distance zones and seasons.
Variables:
-
species: Two-letter abbreviation for the grass species.
-
Belt: Distance zone from the boma.
Close (5-55),Far (150-200). -
season: Season of data collection.
dry,wet. -
mother: Count or abundance of mature/parent plants (main tussocks) of the species.
-
Brown: Count or abundance of brown (senescent/dry) stems or plant material.
-
Green: Count or abundance of green (live/actively growing) stems or plant material.
8. table1_dry__wet.csv
Description: This file serves as a summary table (likely corresponding to a primary table in an associated manuscript) that aggregates total grass biomass and overall dung density for each specific boma location, stratified by season and distance zone.
Variables:
-
Location: Name of the specific boma location.
-
Season: Season of data collection
Dry,Wet. -
Belt: Distance zone from the boma.
Close (5-55),Far (150-200). -
Biomass: Total measured grass biomass at the specific location and belt (Unit: kg/ha).
-
Dung: Total aggregated dung density at the location and belt (Unit: dung piles per sampling unit/plot).
9. Seasons___Biomass___dung_density.csv
Description: This file contains the exact same summary data as table1_dry& wet.csv, providing total grass biomass and dung density per boma location, season, and distance zone, but with slightly different column headers.
Variables:
-
Location: Name of the specific boma location.
-
Season: Season of data collection.
Dry,Wet. -
Belt: Distance zone from the boma.
Close (5-55),Far (150-200). -
Biomass: Total measured grass biomass at the specific location and belt (Unit: kg/ha ).
-
Dung density: Total aggregated dung density at the location and belt (Unit: dung piles per sampling unit/plot).
Code/software
The data can be viewed and analyzed using free and open-source software. Statistical analyses and data visualization were conducted in R statistical software (R version 4.3 or later) using RStudio Desktop (open-source edition).
Livestock_corrals.R
The main R packages used for data analysis included:
- tidyverse for data wrangling and visualization
- vegan for diversity indices and ecological analysis
- lme4 and glmmTMB for generalized linear mixed models
- MuMIn for model selection and AIC comparisons
- ggplot2 for graphical visualization
- car for statistical testing and diagnostics
The workflow involved importing field datasets into R, cleaning and organizing the data, calculating standing grass biomass and diversity indices, performing generalized linear mixed model analyses, and generating figures and tables for interpretation. Wildlife dung density data were analyzed as indicators of wildlife habitat use relative to livestock corrals and seasonality.
The script contains the commands used for data cleaning, statistical analyses, model selection, and figure generation. Scripts are organized sequentially to match the workflow from raw data processing through final analyses and visualization outputs. Please double-check the file names in the script as they may differ from what is uploaded here.
