Riparian forests and macroinvertebrates support multiple ecosystem processes across temperate and tropical streams
Data files
Nov 03, 2025 version files 220.10 KB
-
dat.ergx.csv
210.84 KB
-
README.md
3.94 KB
-
SEM.csv
5.32 KB
Abstract
Ecosystems are interconnected, and ecological processes frequently transcend the physical boundaries that define them. Fluxes of energy, matter, and organisms not only form important ecosystem processes within but also between ecosystems. However, the role of biological drivers in simultaneously supporting multiple ecosystem processes at the interface between aquatic and terrestrial ecosystems (i.e., aquatic-terrestrial ecosystem processes) remains poorly understood, both locally and across regions. To assess the relative importance of riparian forests, detritus consumers, and leaf litter mixing on different ecosystem processes of freshwater detrital food webs, we used leaf litter bags to subsidise local terrestrial leaf litter to forested and non-forested headwater stream sites in a temperate and tropical region. We also manipulated macroinvertebrate access and the composition of leaf litter mixtures. We measured three key aquatic-terrestrial ecosystem processes: biomass accrual of aquatic fungi, nitrogen loss, and decomposition rates of local leaf litter. Across both temperate and tropical streams, ecosystem multifunctionality, i.e., the simultaneous sustaining of these processes, was positively associated with macroinvertebrates and riparian forests but not with leaf litter mixing. Especially, leaf litter nitrogen loss and decomposition rates were consistently higher when macroinvertebrates had access across all leaf litter species. Decomposition rates were also positively associated with other ecosystem processes. These findings highlight consistent, cross-regional effects of riparian forests and macroinvertebrate detritivores on freshwater detrital food webs. In a rapidly changing world, understanding ecosystem processes in headwater streams demands a holistic view that transcends ecosystem borders and incorporates cross-ecosystem interactions.
Dryad DOI: https://doi.org/10.5061/dryad.r7sqv9smr
Rebecca Oester, Paula M. de Omena, Larissa Corteletti, Marcelo S. Moretti, Florian Altermatt, Andreas Bruder
This READ ME file contains information on the R scripts and variables in the datasets used for the analyses.
Datafiles:
- dat.ergx
Main dataset with information on each leaf litter bag with the following variables:
Sample_ID: Unique identifier for each treatment (leaf litter in each bag); NA for Brazil because of naming differences
Site: Site identifier (Ending F indicating forested and ending A indicating non-forested)
Country: Study country (either Switzerland or Brazil)
Region: Region identifier for Switzerland and Brazil
Stream: Stream identifier
Vegetation: Riparian vegetation type (either forested or non-forested)
Mesh: Mesh size of leaf litter bag (either fine or coarse mesh)
LeafLeafLeaf: Leaf species (either Alder, Ash, Inga, or Miconia)
MonoMix: Indication of whether the species was in a single-species leaf litter bag (mono) or mixed (mix)
LeafLeaf: String combination of leaf species and mix
EA_N_Percent_corr: % Nitrogen in the leaf litter at the end of the experiment measured with an elemental analyzer
EA_C_Percent_corr: % Carbon in the leaf litter at the end of the experiment measured with an elemental analyzer
molar_CN: molar ratio between Nitrogen and Carbon
log_molar_CN: logged value of molar_CN
N_Loss: % Nitrogen lost based on calculations in Handa et al (2014)
C_Loss: % Carbon lost based on calculations in Handa et al (2014)
Fungal_Biom_Leaf_mgg_noRR_1: Fungal Biomass in mg / g leaf as proxy for fungal secondary production
k_corr1: the decay rate k based on the decay function incorporating initial and final mass and time progress expressed as Temperature degree days (dd -1).
Delta15_N_Air: the stable isotope signature of N (‰)
Delta13_C_VPDB: the stable isotope signature of C (‰)
Delta_logCN_abs: the difference in C:N ratio from initial and final leaves
ID: string combination of sample ID and leafleaf
Group: treatment combination (string combination of MonoMix, Vegetation, and Mesh)
- SEM
Dataset used to visualize and summarize effect sizes from the SEM (summary outputs from SEM.r)
model: Description of individual links of the SEM variables (Intercepts, NLoss ~ Mesh, etc.)
Estimate: Effect size of the links
Est.Error: estimated error
l95: lower 95% interval
u05: upper 95% interval
Rhat: R hat (indication of fit)
Bulk_ESS: Computes a bulk effective sample size estimate for a single variable. Bulk_ESS is useful as a diagnostic for the sampling efficiency.
Tail_ESS: Computes a tail effective sample size estimate for a single variable. Tail_ESS is useful as a diagnostic for the sampling efficiency in the tails of the posterior. It is defined as the minimum of the effective sample sizes for the 5% and 95% quantiles.
species: Leaf litter species (Alder, Ash, Inga, Miconia)
modelabc: Letters A-L indicating corresponding links in Figure 4
type: description of link type
R-scripts:
Run in R (version 4.1.2) on Linux.
- Multifunctionality
requires dat.ergx.csv
With getStdAndMeanFunctions(), the three functions (decomposition rate (k), N loss, and fungal secondary production) get scaled and averaged to calculate an overall multifunctionality score.
This score and the individual functions get analyzed and plotted depending on the leaf litter species and treatment effect separately using the brms() function.
- SEM
requires dat.ergx.csv
After data preparation, for each leaf litter separately, a big SEM with links specified in the manuscript is run and analyzed.
- Slopes
requires SEM.csv
From the SEM analyses, the summary outputs are aggregated and visualized to produce Figure 4 in the manuscript.