Plant production decreases more than litter decomposition with rising aridity in drylands
Data files
Jun 17, 2025 version files 21.61 KB
-
ANPP_ANPP.csv
8.59 KB
-
Decomposition_Litter_transplantation.csv
4.16 KB
-
Decomposition_Local_litter.csv
2.67 KB
-
README.md
4.37 KB
-
Traits_Leaf_traits.csv
1.06 KB
-
Traits_Litter_traits.csv
758 B
Abstract
Climate change increases aridity in many drylands worldwide, which has significant consequences for ecosystem functioning and may reduce carbon sequestration. However, responses of major carbon cycle processes in drylands, including primary production and decomposition, to increasing aridity remain poorly understood. In this study, we assessed the quantitative effects of precipitation and the underlying impacts of functional traits on aboveground net primary production (ANPP) and plant litter decomposition in herbaceous Mediterranean plant communities. A dataset encompassing a wide range of precipitation (~50-1000 mm) was generated by selecting four field sites along a natural precipitation gradient, applying rainfall manipulations, and conducting the experiment over 3-9 years. Across the entire dataset, ANPP and decomposition decreased non-linearly with declining precipitation, showing steeper reductions at low compared to high precipitation levels. Notably, above ~400 mm, the two processes followed a similar pattern, but below this threshold, ANPP dropped more rapidly, while decomposition decreased less and remained relatively high. Plant functional traits associated with low growth rates exacerbated the reduction of ANPP at the drier sites, whereas higher litter quality at these sites compared with the wetter sites enabled relatively high rates of litter decomposition. The latter findings were confirmed by a litter transplantation study, where litter from the arid site decomposed faster at all sites compared to litter from the wetter sites. In addition, dryland decay mechanisms, such as photodegradation and microbial degradation enabled by non-rainfall water sources might have mitigated some of the dryness-related decrease in decomposition. Increasing climate change-induced aridity in drylands may drive long-term shifts in herbaceous vegetation composition toward smaller, less productive species that produce more labile litter. This trend is expected to accelerate the decline in production more than the decline in decomposition, likely reducing carbon sequestration.
Dataset DOI: 10.5061/dryad.573n5tbm5
Study Description
The dataset was collected within a long-term field experiment along a precipitation gradient in Israel, covering four sites with mean annual rainfall between 90 and 780 mm. The study focused on herbaceous plant communities. At two sites, precipitation was experimentally manipulated (+30% and –30% rainfall), while the other two sites served as controls.
Aboveground net primary production (ANPP) was assessed annually by destructive biomass harvest in permanent plots over 9 years. Litter decomposition was measured using litterbags deployed in the field and in reciprocal transplantation experiments across sites. Plant and litter functional traits were determined for dominant species and representative litter samples. The data include biomass production, decomposition rates, plant and litter chemical and structural traits, and precipitation treatments.
All files are provided in unformatted, open-access CSV format for accessibility and ease of re-analysis.
Files and Variables
Files: Traits
Description: Community-weighted means of functional traits for dominant plant species and corresponding litter at each site and year.
File: Traits_Leaf_traits.csv
- Site – Site name (text)
- Year – Year of sampling (YYYY)
- Canopy Height (cm) – Mean plant canopy height (cm)
- SLA (cm²/kg) – Specific leaf area (cm² per kg)
- Leaf N concentration (%) – Leaf nitrogen concentration (%)
- Leaf C/N ratio – Carbon-to-nitrogen ratio (unitless)
File: Traits_Litter_traits.csv
- Site – Site name (text)
- Year – Year of sampling (YYYY)
- N% – Nitrogen content of litter (%)
- C/N – Carbon-to-nitrogen ratio in litter (unitless)
- Lignin % – Lignin content in litter (%)
- Lignin/N – Lignin to nitrogen ratio (unitless)
- P % – Phosphorus content in litter (%). Missing values are blanks.
Files: Local decomposition
Description: Litter decomposition data for both local and reciprocal litter decomposition experiments across sites, treatments, and years.
File: Decomposition_Local_litter.csv
- Site – Site name (text)
- Treatment – Rainfall manipulation treatment: Control, Dry, or Wet
- Year – Year of decomposition period (YYYY)
- Plot – Plot identifier (numeric label)
- Annual precipitation (mm) – Rainfall during decomposition (mm)
- Mass loss (%) – Percentage of initial litter mass lost (%)
File: Decomposition_Litter_transplantation.csv
- Period – Year(s) of decomposition experiment (e.g., 2002–2003)
- Decomposition site – Site where litter was incubated
- Litter source – Origin site of the litter material
- Replicated litterbag – Litterbag replicate number (numeric)
- Mass loss % – Percentage mass loss of each litterbag (%)
File: ANPP_ANPP.csv
Description: Aboveground net primary production (ANPP) data collected across sites, treatments, plots, and years.
Variables
- Site – Site name (text)
- Treatment – Rainfall manipulation treatment: Control, Wet, or Dry
- Plot – Plot identifier (numeric)
- Year – Year of sampling (YYYY)
- Rain to sampling (mm) – Cumulative rainfall from previous fall until peak biomass sampling date (mm; rounded to integer)
- ANPP (g/m²/yr) – Aboveground net primary production (grams per square meter per year)
Site Abbreviations
The following site abbreviations are used consistently across all datasets:
| Abbreviation | Full Site Name |
|---|---|
| Y | Ein Ya'akov |
| M | Matta |
| L | Lahav |
| B | Be'er Sheva region |
| Code | Description |
|---|---|
| C | Control |
| C-D | Control in 2002, then dry treatment |
| C-W | Control in 2002, then wet treatment |
| D | Dry treatment |
| W | Wet treatment |
Code and Software
All data files are provided in .csv (comma-separated values) format and are compatible with any text editor or statistical software such as R, Python (pandas), Excel, or SPSS.
