Data from: Remote sensing reveals that wild herbivores limit senescent vegetation accumulation on dryland conservation reserves
Data files
Mar 03, 2026 version files 309.96 KB
-
DataFrom_WildHerbivoreGrazing.csv
307.28 KB
-
README.md
2.68 KB
Abstract
Wild herbivores threaten vegetation recovery on dryland conservation reserves globally. Monitoring herbivore impacts in remote drylands is difficult because vegetation biomass transitions between living and senescent states in response to irregular precipitation events. However, land managers need a detailed understanding of the impacts that wild herbivores have on vegetation to develop and refine herbivore management strategies. Remote sensing provides the ability to assess grazing impacts on living and senescent vegetation with high temporal and spatial resolution. Here, we use Sentinel-2 satellite imagery to investigate how grazing by kangaroos and rabbits impacted the fractional cover of photosynthetic (PV) and non-photosynthetic (NPV) vegetation over 7 years on three dryland reserves with experimental herbivore exclusion plots. We compared PV and NPV cover in plots that were accessible to all herbivores, accessible to kangaroos only, and inaccessible to both rabbits and kangaroos. Generalized linear mixed models were used to determine if the grazing impacts of rabbits and kangaroos varied from each other, between reserves, and in response to variable rainfall patterns. Grazing impacts varied between each herbivore, conservation reserve, and between PV and NPV. PV was only weakly limited by kangaroos across all reserves and antecedent rainfall totals. NPV was limited by rabbits and kangaroos, with grazing having stronger impacts on NPV than PV. The grazing impacts of rabbits and kangaroos varied spatially, with evidence that NPV was limited by kangaroos only, by rabbits only, and by both species across different reserves. Both herbivores had stronger impacts on NPV as antecedent rainfall decreased. Our results show that the impact of herbivores on vegetation biomass is greatest during periods of dry climatic conditions. These findings contribute to a growing body of evidence showing that grazing by wild herbivores can have detrimental impacts on dryland ecosystems by disrupting ecological processes supported by NPV. Our results highlight the importance of herbivore management during productive periods to ensure NPV is retained during periods of low rainfall.
Dataset DOI: 10.5061/dryad.j0zpc86rb
Description of the data and file structure
Files and variables
File: DataFrom_WildHerbivoreGrazing.csv
Description: This data set contains all data used in the analysis for "Remote sensing reveals that wild herbivores limit senescent vegetation accumulation on dryland conservation reserves". Fractional vegetation cover measurements were derived from Sentinel-2 reflectance data. Pre-processed fractional vegetation cover images for Mungo and Boolcoomatta were obtained from the Terrestrial Ecosystem Research Network here: TERN - Data.
Pre-proceesed Measurements for the Ikara-Flinders Ranges are not publicly available. We obtained these measurements by downloading raw Sentinel-2 images from the European Space Agency as Level 1C (version 02.04 system) here: doi.org/10.5270/S2_-742ikth. This data was processed using the same code as data held by TERN as described here: doi.org/10.5281/zenodo.7008343
Variables
- Location: Name of conservation reserve.
- Year: Year measurement was taken.
- Season: Season measurement was taken.
- Pair: Unique ID for paired Sp-Excl experimental treatments within each conservation reserve.
- Sp_Excl: All = Kangaroos and rabbits excluded from grazing, Kangaroo = Only kangaroos excluded (not
rabbits), none = control (both rabbits and kangaroos graze). - year_constructed: Year the fenced enclosure was constructed.
- FC_measurement: meanDead = Non-photosynthetic vegetation cover. meanGreen = Photosynthetic vegetation cover.
- value: Mean % cover of respective FC_measurement within each plot.
- prev12.months.rf: Cumulative total of rainfall over the previous 12 months.
- prev24.months.rf: Cumulative total of rainfall over the previous 24 months.
Code/software
All data analysis was performed in R v 4.3.1.
Raw Sentinel-2 imagery was processed using code deposited here: https://doi.org/10.5281/zenodo.7008343.
Access information
Other publicly accessible locations of the data:
- None
Data was derived from the following sources:
- Pre-processedSentinel-22 fractional cover measurements for Mungo and Boolcoomatta
- Raw sentinel-2 reflectence data used to produce fractional cover estimates for the Ikara-Flinders Ranges: doi.org/10.5270/S2_-742ikth
