Dataset of plant communities treated by different herbivore dung quality, quantity and type of grazing
Data files
Feb 28, 2026 version files 114.23 KB
-
app.R
6.12 KB
-
dataset.csv
102.78 KB
-
README.md
5.33 KB
Abstract
The dataset was derived from a 3-year mesocosm experiment designed to examine the effects of herbivore dung quantity, quality, and grazing type on plant community productivity and diversity. It includes data on five levels of dung quantity from three herbivore species (rabbit, horse, and European bison), each differing in dung quality (N:P ratio). The experiment also incorporated three grazing treatments: no grazing, bulk grazing, and selective grazing of legume plants. Measurements include plant community productivity, species diversity, light availability, and the abundance of specific plant types. The dataset captures interactions between dung type, grazing, and plant responses, providing insights into the complex roles of herbivores in shaping terrestrial ecosystems.
https://doi.org/10.5061/dryad.kkwh70sf4
Description of the data and file structure
The dataset includes five levels of dung quantity (ranging from 12.5 to 200 mg P per mesocosm) from three herbivore species: European bison, horse, and rabbit, each differing in dung quality (N:P ratio). The experiment also incorporated three grazing treatments: no grazing, bulk grazing (clipping all plants to 5 cm), and selective grazing (clipping only legume plants). The dataset encompasses measurements of plant community productivity, species abundance, and light penetration to the soil surface, collected annually over three years using a modified point-intercept method and photosynthetically active radiation (PAR) measurements.
Statistical analyses, including Mixed-design ANOVA and Generalized Effect Size (GES), were used to assess the effects of dung type, quantity, grazing, and time on plant community dynamics. Statistical analyses were performed using R version 4.3.2. All analysis code is contained in the app.R file, which is a Shiny web application that can be run directly after updating the working directory.
Files and variables
File: dataset.csv, app.R
Description:
Variables (column names) in the dataset
- pots: the number of mesocosms (1-90)
- herbivore: the dung from three different herbivore species (bison, horse, rabbit)
- Pgradient: 5 levels of dung supply according to the P concentration in the herbivore dung (0.0125, 0.025, 0.05, 0.1 or 0.2 g P per mesocosm)
- grazing: three types of grazing treatments to simulate plant consumption (No, Bulk, Selective)
- replicate: 2 replicates (1,2)
- days: day of data collection from the day of the experimental setup (150days, 515days, 880days)
- moisture: Moisture content of mixed-soil per mesocosm
- light availability: the rate of light penetration to the soil surface per mesocosm (0-1)
- abcent: absolute abundance of Centaurea jacea in each mesocosm
- abagr: absolute abundance of Agrostis capillaris in each mesocosm
- abcar: absolute abundance of Carex flava in each mesocosm
- abtri: absolute abundance of Trifolium pratense in each mesocosm
- abpla: absolute abundance of Plantago lanceolata in each mesocosm
- abhol: absolute abundance of Holcus lanatus in each mesocosm
- abrum: absolute abundance of Rumex acetosella in each mesocosm
- ablot: absolute abundance of Lotus corniculatus in each mesocosm
- totab: total absolute abundance per mesocosm
- shannon: shannon diversity index per mesocosm
- pielous evenness: pielous evenness index per mesocosm
- rabcent: relative abundance of Centaurea jacea in each mesocosm
- rabagr: relative abundance of Agrostis capillaris in each mesocosm
- rabcar: relative abundance of Carex flava in each mesocosm
- rabtri: relative abundance of Trifolium pratense in each mesocosm
- rabpla: relative abundance of Plantago lanceolata in each mesocosm
- rabhol: relative abundance of Holcus lanatus in each mesocosm
- rabrum: relative abundance of Rumex acetosella in each mesocosm
- rablot: relative abundance of Lotus corniculatus in each mesocosm
- rablegum: relative abundance of two leguminous forbs (Trifolium pratense, Lotus corniculatus) in each mesocosm
- rabmychnlegum: relative abundance of two non-leguminous forbs with mycorrhiza (Plantago lanceolata, Centaurea jacea) in each mesocosm
- rabnmnl: relative abundance of two non-mycorrhizal species (Rumex acetosella, Carex flava) in each mesocosm
- rabgrass: relative abundance of two grasses (Agrostis capillaris, Holcus lanatus) in each mesocosm
- ablegum: absolute abundance of two leguminous forbs (Trifolium pratense, Lotus corniculatus) in each mesocosm
- abmychnlegum: absolute abundance of two non-leguminous forbs with mycorrhiza (Plantago lanceolata, Centaurea jacea) in each mesocosm
- abnmnl: absolute abundance of two non-mycorrhizal species (Rumex acetosella, Carex flava) in each mesocosm
- abgrass: absolute abundance of two grasses (Agrostis capillaris, Holcus lanatus) in each mesocosm
- pincent: pin hits counted by the point-intercept method of Centaurea jacea in each mesocosm
- pinagr: pin hits counted by the point-intercept method of Agrostis capillaris in each mesocosm
- pincar: pin hits counted by the point-intercept method of Carex flava in each mesocosm
- pintri: pin hits counted by the point-intercept method of Trifolium pratense in each mesocosm
- pinpla: pin hits counted by the point-intercept method of Plantago lanceolata in each mesocosm
- pinhol: pin hits counted by the point-intercept method of Holcus lanatus in each mesocosm
- pinrum: pin hits counted by the point-intercept method of Rumex acetosella in each mesocosm
- pinlot: pin hits counted by the point-intercept method of Lotus corniculatus in each mesocosm
- totpin: total pin hits counted by the point-intercept method per mesocosm
- logP: Log-converted values for 5 levels of dung supply according to the P concentration in the herbivore dung (12.5, 25, 50, 100 or 200 mg P per mesocosm)
Dung and Soil Collection
Dung from European bison, horse, and rabbit was collected on 25 February 2020 in the National Park Zuid-Kennemerland, Netherlands, a coastal dune ecosystem with nutrient-poor sandy soils. Fresh dung was collected directly after excretion (horse and bison) or from areas with high rabbit density. Samples from at least five individuals per species were stored at 4 °C. Moisture content and concentrations of total carbon (C), nitrogen (N), and phosphorus (P) were determined. Soil from the top 10 cm layer was also collected, sieved, mixed, and used to inoculate mesocosms with soil microbes.
Mesocosm Experiment
In April 2020, 90 mesocosms (30 L) were set up in Brussels. Each mesocosm was filled with quartz sand and 1 L of collected soil. Dung from rabbit, horse, or bison was added at five levels (12.5–200 mg P per mesocosm). Grazing treatments included no grazing, bulk grazing (clipping all plants to 5 cm), or selective grazing (clipping only legumes). Treatments were replicated twice. Plant communities consisted of eight grassland species, and light penetration and plant abundance were measured annually for three years using a point-intercept method and PAR measurements.
Calculations and Statistical Analyses
Dung quantity was log-transformed, and plant abundance data were log(x+1)-transformed. Pielou’s evenness index was calculated from Shannon’s diversity index. Mixed-design ANOVA tested the effects of dung type, quantity, grazing, and time on plant abundance, light penetration, and evenness. Generalized Effect Size (GES) assessed factor impacts. Differences in dung nutrients across species were analyzed using one-way ANOVA. All analyses were performed in R.