Functional traits and habitat use: investigating community assembly in a montane community (Carabidae: Nebria)
Data files
Jun 20, 2024 version files 1.49 MB
-
ambient_temp_resampled.csv
81.89 KB
-
body_size_resampled.csv
280.91 KB
-
carbon_resampled.csv
81.21 KB
-
counts_by_site_adjusted_by_density.csv
3.92 KB
-
Desiccation.csv
59.46 KB
-
Dij_abundance.csv
1.65 KB
-
distance_resampled.csv
107.17 KB
-
elevation_resampled.csv
70.12 KB
-
FS1.csv
190.56 KB
-
isotope_data.csv
57.49 KB
-
Morphology_complete.csv
42.78 KB
-
nebria_tp.csv
63.91 KB
-
nitrogen_resampled.csv
77.66 KB
-
peaks.csv
428 B
-
Pop_prop.csv
4.43 KB
-
presence_absence.csv
3 KB
-
pronotum_ratio_resampled.csv
187.49 KB
-
README.md
7.77 KB
-
substrate_resampled.csv
32.78 KB
-
substrate.csv
468 B
-
surface_temp_resampled.csv
70.33 KB
-
temperature_pref_resampled.csv
66.15 KB
Abstract
The processes that influence community assembly, such as competition for resources and environmental filtering, are often scale-dependent and vary across ecotones. Trait-based ecology provides a useful framework for testing which ecological processes most strongly influence local community composition, especially across environmental gradients where species diversity varies. Where environmental filtering dominates, species distributions are expected to be defined by strong turnover along environmental gradients, with more similar species occupying more similar habitats. Where interspecific competition dominates, species are expected to diverge in relative abundance and resource utilization at sites, so species can co-occur. Here, we integrate measurements of functional traits, microhabitat usage, isotopic composition (δ15N and δ13C), and abundance to test the importance of environmental filtering and resource/habitat partitioning in shaping a montane ground beetle species assemblage (Carabidae: Nebriini: Nebria) in the isolated, volcanic peaks of the northern Cascades Range, U.S.A. Across species of Nebria, body size, pronotal shape, temperature preference, and isotopic enrichment varied across habitats [gravel, rocks 10 cm – 50 cm in diameter), large rocks (>50 cm in diameter), vegetation-covered rocks, and alpine (snowfields and talus)], and habitat/microhabitat features were reliable predictors of species presence. Resource consumption among mid-elevation species on Mt. Rainier – the peak with the greatest species diversity – is highly overlapping. Species turnover and nestedness varied significantly across habitat gradients and peaks throughout this region and varied nearly significantly across sites. Across habitat types and sites, more similar species are more likely to coexist. These results suggest that environmental filtering is the primary process structuring this species assemblage, although we find detailed evidence for microhabitat niche partitioning among species of Nebria at the site-scale.
https://doi.org/10.5061/dryad.vmcvdnd02
Description of the data and file structure
In some files, “eschscholtzii” is misspelled “eschschultzii” and “mannerheimii” is misspelled “mannerheimi”.
In general, rows are specimens, columns are variables. Any empty cells in datasets can be replaced with “NA”.
FS1 includes environmental variables measured for each specimen collected during ecological surveys. Environmental variables included: Elevation (m), location (name of the river system), river name, aspect (north, south, etc.) of the mountain, latitude and longitude (WGS84), distance from the river’s edge (m) where the specimen was found, surface temperature (C), habitat (substrate where the specimen was found), species identity, average ambient temperature during the survey (C), relative humidity during the survey (%), substrate level based on substrate type (1-5), notes, distance from the river’s edge (in), ambient temperature at the beginning of the survey (C), ambient temperature at the end of the survey (C), relative humidity at the beginning of the survey (%), relative humidity at the end of the survey (%), and lastly a unique specimen identifier was included in the final column.
Morphology complete: variables include (all cm): species name, antennal scape length, elytral length, elytral width, narrowest distance between eyes, femur length, pronotal width ratio (widest point divided by the base), dorsal-ventral thickness, tibia length, a color identifier for plotting in R, an index identifier, a unique specimen identifier, the elevation of the habitat that the specimen is typically found in (low, high, mid), and the type of habitat the specimen is typically found in (alpine, riparian).
isotope data: variables included: Sample ID (site where specimen was collected and specimen ID), peak (where specimen was collected), elevation (high, mid, low), latitude and longitude (WGS84), site ID, species name, index number, 13C enrichment, total C content (ug), 15N enrichment, total N content (ug), estimated trophic position, tray name (for laboratory purposes only), type of material (plant or beetle elytron), elevation of specimen collection (m), standardized 13C enrichment (using site-specific plant values), standardized 15N enrichment (using site-specific plant values), group id, community id.
Desiccation: variables include: the location where specimens were collected, the date they were collected, a unique specimen identifier, species name, initial mass (g) including capsule, mass of specimen after 8 hours (g) including capsule, mass of specimen after 24 hours (g) including capsule, mass of specimen after 48 hours (g) including capsule, and the specimen’s ending weight (g) including capsule. Next, the weight of the capsule the specimen was contained in (g), and subsequent corrected mass values accounting for the added weight of the capsule, then total weight loss and proportional weight loss (loss/initial).
nebria_tp: variables include: specimen identifier, species name, date, time of the start of the trial, trial number, lane number, time interval for recording (every 15 minutes), position of the specimen along the lane, temperature at that position (C), and a specimen identifier.
Functional trait variables include: morphology, stable isotope composition, desiccation tolerance, and temperature preference.
ambient temp resampled: The average ambient temperature where each specimen was found was recorded in degrees Celsius. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
body size resampled: Body size for each species was estimated using PC1 loadings from a PCA of all morphological features. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
carbon resampled: 13C standardized enrichment for each specimen. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
distance resampled: Distance of each specimen from the river’s edge. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
elevation resampled: Elevation where each specimen was collected. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
nitrogen resampled: 15N standardized enrichment for each specimen. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
pronotum ratio resampled: Pronotum width/pronotum base. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
surface temp resampled: Surface temperature (C) where specimens were found. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
temperature pref resampled: Average temperature where beetles were found during temperature preference assays. This data was sampled without replacement 100x for each species. The sample sizes were n-1, where n is the minimum sample size of each species in each environmental or functional trait variable.
counts by site adjusted by density: Estimates of species abundance at each site, adjusted by detection probability curves and width of the site. Variables include: latitude and longitude of site (WGS84), peak name, site elevation (m), altitude (low, low2, low3, mid, mid2, high, high2; all lows are equivalent to other lows, etc.), substrate most common at the site, width of the site from the river’s edge to the edge of the forest, detection probability of species on that substrate type, adjusted counts for species based on substrate type detection probabilities and actual observations for all species. Then, the average abundance (counts) of i when j is present, and the average abundance (counts) of i when j is absent.
Dij abundance: a matrix of species (i, rows) average abundances when others (j, columns) are present or absent.
peaks: species presence (1) or absence (0) on each peak.
Pop prop: same as counts by site adjusted by density, but instead of raw counts, cells are the proportion of the population at a site which is a given species. If the cell says 0.25 for N. acuta at the first site, then 0.25 of the population at the first site was N. acuta.
substrate: counts of species’ occurrences on various substrate types, the substrate types are 1-5, 1: gravel (<10 cm diameter), 2: medium rocks (10-50 cm diameter), 3: large rocks (>50 cm diameter), 4: vegetation, 5: alpine.
Code/Software
R was used for analysis. Packages used include: dplyr, corrplot, nnet, rcartocolor, pheatmap, caret, tibble, randomForest, SIBER, nlme, and betapart. The R code used for analysis is titled Functional_traits_and_habitat_use_investigating_community_assembly_in_a_montane_community_(Carabidae_Nebria).R
Ecological surveys were conducted during the first three hours after night fell. Specimens were collected along 100 m length transects, parallel to the riverbed. Surveys were conducted by walking in parallel lines, normal to the 100 m axis of the transect, from the edge of the river to the edge of the forest habitat. For thermal preference assays, beetles acclimated to laboratory conditions for two weeks on a 12-hour light cycle in a 5°C refrigerator. They were allowed to traverse a lane on a table with a thermal gradient in a dark room. Their position was tracked using a thermal imaging camera. Desiccation assays were performed by starving insects for 48 hours (with access to water), and then placing the beetles in a plastic vial sealed with a canvas cloth. This vial was placed in a 50 mL Falcon along with Dririte. Beetles were weighed at 0, 8, 24, and 48 hours. For stable isotope composition, beetles were collected and immediately placed in ethyl acetate. Their elytra were collected and pulverized before sending the UC Davis Stable Isotope Facility (per their specifications) for analysis.