Fencing the flux: Seasonal trends, environmental drivers, and mitigation opportunities of methane emissions from farm dams
Data files
Nov 25, 2025 version files 219.90 MB
-
Odebiri_et_al_2025_Global_Change_Biology.zip
219.90 MB
-
README.md
7.12 KB
Abstract
Farm dams are significant methane (CH4) sources in agricultural landscapes. Fencing them to limit livestock access reduces organic matter and nutrient inputs, thereby limiting CH4 production. However, existing studies on the benefits of fencing are constrained by short durations, omission of ebullitive fluxes, limited spatial and temporal coverage, and small sample sizes. Here, we report a large-scale, multi-season assessment of total CH4 (diffusive + ebullitive) and carbon dioxide (CO2) fluxes from fenced and unfenced farm dams, along key environmental drivers. We monitored 113 farm dams in temperate mainland south-eastern Australia over two years, amounting to 39,552 and 45,408 hourly observations of total CH4 and CO2 fluxes, respectively. We integrated field-measured emissions with Sentinel-2 indices, topo-climate variables, and geostatistical models to identify flux drivers, quantify temperature sensitivity, and spatially extrapolate mitigation potential across Local Government Authorities (LGAs). We found that fencing reduced CH4 fluxes by 66–82% across seasons while also significantly lowering the temperature sensitivity of CH4 fluxes, slowing the exponential rise in emissions under warming conditions. Specifically, CH4 fluxes in fenced dams increased by 71% per 10°C warming (Q10 = 1.71, EM = 0.4 eV), compared to unfenced dams increasing by 275% (Q10 = 3.75, EM = 0.98 eV). CH4 fluxes were driven by temperature, rainfall, and hydrological proxies (Modified Normalized Difference Water Index, MNDWI; Floating Algae Index, FAI), while CO2 fluxes responded to rainfall and Normalized Difference Water Index (NDWI). Extrapolating our findings across the study area (526,296 km²), fencing all farm dams could cut CH4~ fluxes by 1.16–1.35 kt yr⁻¹. By combining high-resolution emission data with scalable management strategies, this study offers a framework to improve greenhouse gas inventories and guide targeted climate mitigation in agriculture.
Running title: Fencing Farm Dams to Cut Methane Fluxes
1. Overview
This dataset (Odebiri_et_al_2025_Global_Change_Biology.zip) accompanies the article “Fencing the Flux: Seasonal Trends, Environmental Drivers, and Mitigation Opportunities of Methane Emissions from Farm Dams” published in Global Change Biology (Odebiri et al., 2025). The dataset provides the empirical and processed data used to quantify methane (CH₄) and carbon dioxide (CO₂) emissions from farm dams across temperate Southeastern Australia (2022–2024). It includes raw flux data, Sentinel-2 and terrain predictors, climate data, shapefiles, and analysis scripts used to assess seasonal patterns, environmental drivers, and emission mitigation potential via fencing interventions.
2. Contents of the Dataset
| File/Folder | Description |
|---|---|
| Data/ | Contains raw and processed datasets. |
| All_DataWeather_sum_Merged2.csv | Raw flux data including CH₄ and CO₂ concentrations and ancillary field measurements. |
| GCB_PAPER_DATA.csv | Comprehensive dataset integrating raw flux data with Sentinel-2, terrain, and climate predictors. |
| Python_extract_code/ | Contains Extract_pixel_script_GCB.py for extracting Sentinel-2 pixel values and terrain metrics using Google Earth Engine. |
| R_code/ | Contains GCB_Paper_R_Sript.R, the R script used for data cleaning, mixed-effects analysis, and mapping. |
| Updated_ALL_Farm_dams.shp | Shapefile representing farm dam polygons across study regions. |
| Updated_LCA_shapefile.shp | Shapefile defining Local Government Areas used for spatial aggregation and emission mapping. |
| Final_Metadata_Dictionary.xlsx | Metadata table describing all variables, units, transformations, and sources. |
| README.txt | Documentation of data, scripts, and methods (this file). |
| Licence.txt | CC0 1.0 Public Domain Dedication. |
3. Data Description
Key Variables
| Variable | Description | Units / Type | Source / Derivation |
|---|---|---|---|
| Name | Unique farm dam identifier | Text | Field metadata |
| Treatment | Management condition (FD = Fenced, UD = Unfenced) | Categorical | Field observation |
| Season | Sampling season (Summer, Autumn, Winter, Spring) | Categorical | Derived from Date_R |
| Gas | Measured gas type (CH₄, CO₂, N₂O) | Text | Field measurement |
| mgm2day_mean | Mean flux rate | mg m⁻² day⁻¹ | Field measurement |
| NDVI, NDWI, ARVI, MNDWI, FAI, GCI | Spectral indices | Unitless | Sentinel-2 L2A |
| B1–B12 | Sentinel-2 reflectance bands | Unitless (scaled) | Sentinel-2 MSI |
| Slope, Aspect, TWI, DEM | Terrain variables | Degrees / m / Unitless | DEM-S (GA) |
| Temp_mean, Humidity_mean, Rainfall_mean | Climate predictors | °C / % / mm | ERA5 / BOM |
| Boltzmann_15C, EM, Q10 | Temperature sensitivity parameters | eV⁻¹ / eV / Dimensionless | Derived via Boltzmann fits |
4. Methods Summary
- Data Collection: Methane and CO₂ fluxes were measured using the Pondi chamber system at 130+ farm dams across Southeastern Australia between 2022–2024.
- Predictor Extraction: Sentinel-2 indices and DEM-S terrain variables were extracted using
Extract_pixel_script_GCB.pyin Google Earth Engine. - Data Integration: All extracted predictors were merged with raw gas fluxes into
GCB_PAPER_DATA.csv. - Statistical Analysis: Mixed-effects models (
nlmepackage) were fitted to examine treatment and seasonal effects, while emission reduction potential was spatially quantified using R (sf,ggplot2). - Mapping: Total and per-hectare CH₄ reduction potentials were mapped at the LGA level using the
Updated_ALL_Farm_dams.shpandUpdated_LCA_shapefile.shplayers.
5. File Dependencies
GCB_PAPER_DATA.csvis the main dataset used in all analyses.All_DataWeather_sum_Merged2.csvprovides unmerged raw flux data.Extract_pixel_script_GCB.pygenerates satellite and terrain predictors used in the R analysis.GCB_Paper_R_Sript.Rperforms the full statistical and mapping workflow.
6. Citation
Odebiri, O., Scheele, B. C., Lindenmayer, D. B., Smith, D. G., & Malerba, M. E. (2025). Fencing the Flux: Seasonal Trends, Environmental Drivers, and Mitigation Opportunities of Methane Emissions from Farm Dams. Global Change Biology. Data available at Dryad: [DOI: 10.5061/dryad.z612jm6rc]
7. License and Collaboration Note
License: CC0 1.0 (Creative Commons Zero Public Domain Dedication)
This dataset is released under the CC0 1.0 Public Domain Dedication, which permits users to copy, modify, distribute, and use the data for any purpose without restriction and without requiring attribution.
Collaboration Note:
While no attribution is legally required under CC0, researchers are warmly encouraged to contact the corresponding author (Odebiri et al.) if considering substantial reuse or collaborative extensions. This helps ensure consistent interpretation of the dataset and may foster opportunities for collaboration on related projects.
This research was conducted in temperate mainland south-eastern Australia (Fig. 1), covering the states of Victoria, New South Wales, and southern Queensland. The region spans approximately from –39.14° to –25.85° latitude and from 140.96° to 168.00° longitude, encompassing temperate zones across these states with diverse landscapes, including temperate grasslands, coastal plains, and montane environments, and has a temperate climate influenced by the Southern Ocean. Seasonal temperatures range from 25–30°C in summer (December–February), 20–25°C in autumn (March–May), 15–25°C in spring (September–November), and 6–10°C in winter (June–August). Annual rainfall, peaking in winter, typically varies from 400 to 1200 mm. Livestock grazing, particularly beef cattle and sheep, dominates the region’s agriculture. The study area spans 526,296 km2 and contains an estimated 249,305 farm dams, with an average total water surface area of about 24,000 hectares (ha) across 168 Local Government Authorities (LGAs) (Malerba et al., 2021).
Over a period of 24 months (October 2022–October 2024), we monitored 113 farm dams to analyse seasonal patterns of CH4 and CO2 emissions, their driving factors, and the effects of fencing to limit livestock access. These dams were spread across ten farming properties arranged along an approximately 1,498 km transect in temperate southeast Australia. Each property contained 2–25 farm dams, with treatments nearly balanced between fenced dams (n = 52; 685 monitoring days) and unfenced dams (n = 61; 963 monitoring days). Sampling spanned all four seasons, including summer, autumn, winter, and spring, with individual dams monitored for periods ranging from one week to four months in one or multiple seasons, yielding a total of 1,648 dam-day observations.
