Data for: Rewilding soil and litter invertebrates and fungi increases decomposition rates and alters detritivore communities
Data files
Mar 07, 2024 version files 118.03 KB
Abstract
Habitat degradation and associated reductions in ecosystem functions can be reversed by reintroducing or ‘rewilding’ keystone species. Rewilding projects have historically targeted restoration of processes such as grazing regimes or top-down predation effects. Few projects focus on restoring decomposition efficiency, despite the pivotal role decomposition plays in global carbon sequestration and nutrient cycling. Here, we tested whether rewilding entire communities of detritivorous invertebrates and fungi can improve litter decomposition efficiency and restore detritivore communities during ecological restoration. Rewilding was conducted by transplanting leaf litter and soil, including associated invertebrate and fungal communities from species-rich remnant sites into species-poor, and geographically isolated, revegetated farmland sites in a temperate woodland region of southeastern Australia. We compared communities in sites under the following treatments: remnant (conservation area and source of litter transplant), rewilded revegetation (revegetated farmland site with litter transplant), and control revegetation (revegetated site, no transplant). In one ‘before’ and three ‘after’ sampling periods, we measured litter decomposition and the abundance and diversity of detritivorous invertebrates and fungi. We quantified the effect of detritivores on the rate of litter decomposition using piecewise Structural Equation Modelling. Decomposition was significantly faster in rewilding sites than in both control and remnant areas, and was largely driven by a greater abundance of invertebrate detritivores. Similarly, the abundance of invertebrate detritivores in rewilding revegetation sites exceeded the level of remnant communities, whereas there was little difference between control and remnant sites. In contrast, rewilding did not increase saprotrophic fungi relative abundance/diversity and there was no strong relationship between decomposition and fungal diversity. Our findings suggest the relatively simple act of transplanting leaf litter and soil can increase functional efficiency during restoration and alter community composition. Our methods may prove important across a range of contexts where other restoration methods have failed to restore ecosystem processes to pre-degradation levels.
README: Data for: Rewilding soil and litter invertebrates and fungi increases decomposition rates and alters detritivore communities
https://doi.org/10.5061/dryad.ncjsxkt1r
This dataset contains a suite of environmental measurements, decomposition measurements, and community matrices for invertebrate and fungal detritivores. Data were collected at four time points, once pre-rewilding, and three times post-rewilding. We found that rewilding detritivore communities via soil and litter transplants increased the number of invertebrate detritivores, which had cascading effects on decomposition rates.
Description of the data and file structure
The dataset contains various .csv files that link to the code.R file. Within the code.R file, there are lines of codes that detail the analyses of each figure in the manuscript.
Metadata for field sites can be found in the metadata.csv file, which documents geographic location of each site in WGS 84.
The first portion of this code is the analyses of decomposition rates found in Figure 3 of the manuscript. This code is working off the file labelled decomposition.csv.
Columns are labelled as:
- "Site" : Site replicates for each treatment, there are 6 sites within each treatment
- "Treatment" : One of three treatments: "Control", "Rewilding transplant", "Remnant", see manuscript for definitions
- "Distance" : The distance the decomposition measurements were taken from the experimental grid. Either 'Near' (2 m), or 'Far' (10 m)
- "Time" : One of 3 sessions the measurements were taken in: 3 months post-rewilding, 19 months post-rewilding, or 35 months post-rewilding
- "Mass Loss" : The % mass lost from each litter bag after 3 months in the field
The second portion of this code is the analyses of abundance and diversity of invertebrates and fungi found in Figure 4 of the manuscript. This code is working off two files. The first concerns invertebrate detritivores and is labelled invertebrate_detritivores.csv.
Columns are labelled as:
- "Site" : Site replicates for each treatment, there are 6 sites within each treatment
- "Treatment" : One of three treatments: "Control", "Rewilding transplant", "Remnant", see manuscript for definitions
- "Time" : One of 4 sessions the measurements were taken in: Pre-rewilding, 3 months post-rewilding, 19 months post-rewilding, or 35 months post-rewilding
- "Invertebrate abundance" : A count of invertebrate detritivores found at each site
- "Invertebrate richness" : The number of invertebrate detritivore species found at each site.
The second file concerns fungal saprotrophs and is labelled fungal_saprotrophs.csv.
Columns are labelled as:
- "Site" : Site replicates for each treatment, there are 6 sites within each treatment
- "Treatment" : One of three treatments: "Control", "Rewilding transplant", "Remnant", see manuscript for definitions
- "Time" : One of 4 sessions the measurements were taken in: Pre-rewilding, 3 months post-rewilding, 19 months post-rewilding, or 35 months post-rewilding
- "Fungi abundance" : Relative abundance of fungal saprotrophs found at each site
- "Fungi diversity" : Diversity of fungal saprotrophs found at each site.
The third portion of this code is the HMSC analyses of invertebrate morphospecies responses found in Figure 5 of the manuscript. This code is working off four files, each set up in the same manner: community_matrix_prerewilding.csv, community_matrix_3months.csv, community_matrix_19months.csv, community_matrix_35months.csv.
Columns are labelled as:
- "Site" : Site replicates for each treatment, there are 6 sites within each treatment
- "Treatment" : One of three treatments: "Control", "Rewilding transplant", "Remnant", see manuscript for definitions
- "Temperature" : The temperature in Celcius at each site over 3 months
- "Humidity" : The % humidity at each site over 3 months
- Remaining columns are species data
To construct Figure S3, repeat the code above but use the files found in the fungal ASV communities.
The fourth portion of this code is the Structural Equation Modelling analyses found in Figure 6 of the manuscript. This code is working off the sem_data.csv file.
Columns are labelled as:
- "Site" : Site replicates for each treatment, there are 6 sites within each treatment
- "Treatment" : One of three treatments: "Control", "Rewilding transplant", "Remnant", see manuscript for definitions
- "Treatment ordinal" : Treatments as ordinal variables - Control = 1, Rewilding transplant = 2, Remnant = 3.
- "Time" : One of 3 sessions the measurements were taken in: 3 months post-rewilding, 19 months post-rewilding, or 35 months post-rewilding
- "Temp" : The temperature in Celcius at each site over 3 months
- "Humidity" : The % humidity at each site over 3 months
- "Abundance" : The abundance of invertebrate detritivores found at each site
- "Breakdown" : The % mass lost from each litter bag after 3 months in the field
- "Litter cover" : The % of litter cover found at each site
- "sap diversity" : Diversity of fungal saprotrophs found at each site
- "Meso abundance" : Abundance of mesoarthropod detritivores found at each site (only present for 19 months post-rewilding)
Code/Software
The code required for analyses of each section of this manuscript is documented in the R file code.R.
Methods
This dataset is part of a long term study, examining the effect of litter and soil transplants on invertebrates and microbes. To process invertebrate detritivores, we took litter samples and ran them through tullgren funnels at 3 different time points after the initial litter and soil transplants. To process fungal saprotrophs, we took soil samples in conjuction with the invertebrate samples. All steps in data processing are listed in the 'code.R' file.