Moʻorea plant herbivory damage type data
Data files
Jun 29, 2026 version files 315.37 KB
-
dt_metrics_all_species.csv
24.57 KB
-
plant_DT_network_matrix.csv
11.49 KB
-
plant_metrics_all_species.csv
4.01 KB
-
raw_data_leaves.csv
264.23 KB
-
raw_functional_breadth_data.csv
2.07 KB
-
raw_location_data.csv
1.92 KB
-
README.md
7.09 KB
Abstract
This dataset contains standardized records of insect and fungal herbivory traces on leaf litters collected in the ʻŌpūnohu tropical lowland rainforest, Moʻorea, French Polynesia. Each row represents a single observation of herbivory on an individual leaf and includes the following variables: leaf_ID (unique identifier), tree (collection tree), species (plant species name), host_plant_category(biogeographic category: native, naturalized, or invasive), Func_feed_group (functional feeding group, e.g., hole feeding, mining), damage_type (standardized damage type code largely following the Damage Type–Functional Feeding Group system of Labandeira et al. 2007), and PAD(percent area damage, ranging from 0 to 1). Instances of no observable damage are included with NA values for feeding group and damage type and a PAD of 0.
This dataset enables detailed analyses of herbivory diversity, intensity, and specificity at both individual and species levels and is reusable for studies in community ecology, invasion biology, trophic interactions, and trait-based herbivory assessments. It is fully anonymized and contains no sensitive human or environmental data. All plant species are non-threatened and identified using field and herbarium references. Reuse is permitted under standard academic attribution guidelines.
Dataset DOI: 10.5061/dryad.4b8gthtrz
Comments and requests should be addressed to Maia Supple: msupple@ucla.edu
Description of the data and file structure
The data were collected to quantify patterns of insect and fungal herbivory across native, naturalized, and invasive plant species within the ʻŌpūnohu rainforest on Moʻorea, French Polynesia. Leaf samples were studied for herbivory damages using the Damage Type–Functional Feeding Group (DT–FFG) classification system. The data collection involved detailed documentation of herbivory diversity, intensity, and specificity through standardized microscopic examination of leaf damage types. This dataset supports analyses of herbivore–plant interactions, herbivory host specificity, and the ecological impacts of plant invasions in tropical forest ecosystems.
Files and variables
File: raw_data_leaves.csv
Description:
This dataset contains individual leaf-level herbivory observations collected from multiple plant species across different sites in the ʻŌpūnohu rainforest. Each row represents a unique leaf sample and its associated herbivory metrics.
Variables:
- leaf_ID: Unique identifier for each leaf sample
- tree: Individual tree identifier
- species: Plant species name (scientific binomial)
- host_plant_category: Plant classification group (native(N), naturalized(R), invasive(I))
- Func_feed_group: Functional feeding group of the herbivory damage (e.g., hole feeding, mining, skeletonization, etc)
- damage_type: Specific damage type code observed on the leaf as determined by Labandeira et al. (2007)
- PAD: Quantitative measure of herbivory intensity on the leaf (e.g., percent leaf area damaged)
File: raw_location_data.csv
Description:
This dataset contains geographic coordinates for the sampling locations where leaf herbivory data were collected.
Variables:
- Name: tree number
- Latitude: Geographic latitude in decimal degrees
- Longitude: Geographic longitude in decimal degrees
File: raw_functional_breadth_data.csv
Description:
This dataset provides functional breadth (host specificity) information for different damage types used in herbivory assessment.
Variables:
- damage_type: Code identifying the specific damage type
- occurrences: Number of observed occurrences of this damage type across samples
- num_plant: Number of plant species on which this damage type was observed
- functional_breadth: Numerical score assigned to the damage type based on how many plants it occurs on (1 = generalized, 3 = specialized; see more in S4 File of the manuscript)
- Func_feed_group: Broad herbivory category (e.g., hole feeding, mining, skeletonization)
Network analysis
File: plant_DT_network_matrix.csv
Description: Edge list used to build the plant–damage type network. Each row represents an observed interaction between a plant species and a damage type.
File: dt_metrics_all_species.csv
File: plant_metrics_all_species.csv
Description: Calculated network metrics summarized at the damage type (DT) level in the dt_metrics_all_species.csv file and calculated network metrics at the plant species level in the plant_metrics_all_species.csv file.
Variables:
- Plant: The scientific name of the plant species.
- Degree: The total number of unique partner species (e.g., pollinators or herbivores) interacting with this plant.
- normalised.degree: The degree divided by the total number of possible partner species in the entire network.
- nested rank: The rank of the plant species based on its position in a nested interaction matrix.
- betweenness: Counts how often the plant node falls on the shortest path between other nodes in an unweighted network.
- weighted.betweenness: Betweenness centrality adjusted for the frequency or weights of the interactions.
- Closeness: How close a plant node is to all other nodes in the unweighted network based on shortest paths.
- weighted.closeness: Closeness centrality adjusted for interaction weights.
- species.strength: The sum of the dependencies of all partner species on this plant (indicates the plant's overall importance to its partners).
- interaction.push.pull: Measures interaction asymmetry; shows whether the plant exerts more influence on partners ("push") or vice versa ("pull").
- PSI: (Pollination Service Index) Quantifies the plant’s specific contribution to the network-wide interaction flow or service.
- PDI: (Paired Difference Index) A measure of specialization quantifying the differences in interaction strength across all partners.
- species.specificity.index: An index quantifying how specific or selective the plant is toward its interaction partners.
- node.specialisation.index.NSI: A node-level specialization index measuring how specialized the plant's interaction niche is.
- d_prime: Blüthgen’s selectiveness index. It measures specialization by taking into account the overall availability/abundance of partner species, ranging from 0 (generalized) to 1 (highly selective).
- resource.range: The fraction of available partner resources/species that this plant interacts with.
- Fisher.alpha: A diversity index expressing the relationship between the number of partner species and the number of interaction instances.
- partner.diversity: The Shannon-Wiener diversity index of the plant's interactions.
- EffectivePartners: The effective number of partner species, calculated as e^{H'}.
- ProportionalSimilarity: A measure of niche overlap, indicating how similar this plant's interaction network is compared to others.
- ProportionalGenerality: A measure of a species' generality relative to the total available interaction partner abundance.
Code/software (hosted on Zenodo)
This repository contains three main analysis scripts focused on insect herbivory and plant interactions on Moʻorea:
Herbivory_metric_analysis_by_host_plant_categories.R
A comprehensive workflow for analyzing leaf herbivory across multiple plant taxa. Major components include:
- Data cleaning and scoring of herbivory types
- Calculation of herbivory metrics (damage type richness, herbivory frequency, percent area damage)
- Specificity score analyses
- analysis across functional feeding groups and classes
- Rarefaction and sampling-effort standardization
- Visualization of herbivory diversity and intensity
This script primarily uses the tidyverse for data manipulation and ggplot2 for figure generation.
Network_analysis.r
Performs network-level analyses of plant–insect herbivory interactions, including:
- Construction of DT × host-plant interaction matrices
- Calculation of network metrics (connectance, specialization, modularity)
- Network visualization and comparison across plant categories
Freshly fallen leaf litter was collected from the ʻŌpūnohu rainforest on Moʻorea, French Polynesia across 12 plant species, representing native, naturalized, and invasive species. At each site, 50 recently fallen leaves were randomly selected within a 3-meter radius plot around individual plants of the focal species. Only confidently identifiable leaves were retained for analysis. Each leaf was assigned a field identification number and brought to the Richard B. Gump South Pacific Research Station for detailed study.
Traces of insect and fungal herbivory were assessed using the Damage Type–Functional Feeding Group (DT–FFG) system (Labandeira et al. 2007). Each distinct damage type (DT) was categorized into a functional feeding group (FFG), such as hole feeding, margin feeding, skeletonization, mining, and galling. For every leaf, the presence and type of herbivory were recorded, along with a percent area damaged (PAD; between 0 and 1). Leaves with no visible herbivory were included with NA values for damage_type and Func_feed_group and a PAD of 0.
Herbivory diversity and intensity were quantified using three metrics: DT richness (total number of distinct DTs), DT frequency (proportion of damaged leaves; presence/absence for each leaf), and percent area damaged (PAD). The DT richness is cumulatively measured per tree (out of 50 leaves), whereas DT frequency and PAD represent mean values of the 50 leaf litter specimens. Species-level metrics were calculated as the mean across five individual trees per species. For comparisons among host plant categories (native, naturalized, invasive), values were averaged across 20 individual trees—four species per category.
Data were entered, cleaned, and analyzed in R.