The goal of impIndicator is to allow users to seamlessly calculate and visualise the impact of alien taxa and individual species in a given area. It calculates and visualises potential impact per site as a map. It takes in Global Biodiversity Information Facility (GBIF) occurrence cube and EICAT assessment data. It enables users to choose from various methods of calculating impact indicators based on different assumptions.
The impIndicator produces three main products:
compute_species_indicator(): The
species impact produces the trends of individual alien species,
enabling a species-specific impact attributed to invasions. This data
supports comparisons of individual species’ impacts, revealing their
roles and interactions within invaded area. The species impact is
invaluable for prioritising species-specific management efforts,
informing control and eradication strategies, and advancing research
on alien species’ ecological roles and adaptation patterns.compute_site_indicator(): The site
impact as a map serves as a visual and analytical tool to represent
the intensity of biological invasions across different parts of an
area. By enabling spatial comparisons—such as between provinces,
states, or conservation areas—, it highlights hotspots and areas at
risk of invasion impact. This spatial data is useful for prioritising
management actions, coordinating restoration projects, and fostering
cross-regional collaboration to address alien species impacts
effectively.compute_regional_indicator(): The
impact indicator offers a nuanced representation of the trends of
biological invasions of an area (local, regional, or global scales).
By tracking the increase and decrease of ecological threats over time,
this product provides insights into the dynamics of alien species
impacts, helping assess whether current management practices are
effective or need adjustment. The temporal analysis of impact
indicator enables targeted resource allocation, fostering proactive
interventions to mitigate biodiversity loss and ecosystem degradation.Install impIndicator in R:
install.packages("impIndicator", repos = c("https://b-cubed-eu.r-universe.dev", "https://cloud.r-project.org"))
You can install the development version from GitHub with:
# install.packages("remotes")
remotes::install_github("b-cubed-eu/impIndicator")
We demonstrate the computation and visualisation of impact indicator of
biological invasions using the impIndicator package:
compute_regional_indicator() to compute impact indicators of alien
taxa, compute_species_indicator() to compute impact indicators per
species, and compute_site_indicator() to compute impact indicators per
site. The functions require (1) a species occurrence cube processed by
the b3gbi::process_cube() and (2) Environmental Impact Classification
for Alien Taxa (EICAT) data of species. Go to
vignette("Background", package = "impIndicator") to read more about
these functions.
# Load packages
library(impIndicator)
library(b3gbi) # General biodiversity indicators for data cubes
library(tidyverse) # Visualisation
library(dubicube) # Data sensitivity and uncertainty estimation
The cube_acacia_SA is a GBIF occurrence cube of Acacia species in
South Africa, processed with the process_cube() function in b3gbi
package. The cube_acacia_SA grid is an extended quarter degree cell,
thus specified for the process_cube. The first_year and last_year
arguments define the temporal range of the cube, restricting the
processed occurrences to records between first_year and last_year.
# Process GBIF Acacia occurrence cube
acacia_cube <- process_cube(cube_name = cube_acacia_SA,
grid_type = "eqdgc",
first_year = 2010,
last_year = 2024)
acacia_cube
#>
#> Processed data cube for calculating biodiversity indicators
#>
#> Date Range: 2010 - 2024
#> Single-resolution cube with cell size 0.25degrees
#> Number of cells: 461
#> Grid reference system: eqdgc
#> Coordinate range:
#> xmin xmax ymin ymax
#> 8.00 31.75 -35.00 -22.75
#>
#> Total number of observations: 19431
#> Number of species represented: 36
#> Number of families represented: 1
#>
#> Kingdoms represented: Plantae
#>
#> First 10 rows of data (use n = to show more):
#>
#> # A tibble: 3,604 × 11
#> year cellCode taxonKey scientificName obs kingdom family
#> <dbl> <chr> <dbl> <chr> <dbl> <chr> <chr>
#> 1 2010 E018S32AD 2980425 Acacia cyclops 1 Plantae Fabaceae
#> 2 2010 E018S33CD 2979793 Acacia viscidula 1 Plantae Fabaceae
#> 3 2010 E018S33CD 2980425 Acacia cyclops 1 Plantae Fabaceae
#> 4 2010 E018S34AB 2978369 Acacia ulicifolia 1 Plantae Fabaceae
#> 5 2010 E018S34AB 2978604 Acacia pycnantha 1 Plantae Fabaceae
#> 6 2010 E018S34AB 2978552 Acacia saligna 6 Plantae Fabaceae
#> 7 2010 E018S34AB 2979232 Acacia implexa 4 Plantae Fabaceae
#> 8 2010 E019S34AA 2978604 Acacia pycnantha 1 Plantae Fabaceae
#> 9 2010 E020S33DC 2979775 Acacia mearnsii 3 Plantae Fabaceae
#> 10 2010 E029S30DD 2979775 Acacia mearnsii 2 Plantae Fabaceae
#> # ℹ 3,594 more rows
#> # ℹ 4 more variables: minCoordinateUncertaintyInMeters <dbl>, xcoord <dbl>,
#> # ycoord <dbl>, resolution <chr>
The Environmental Impact Classification for Alien Taxa (EICAT) assessment data is the reported impact of alien taxa based on EICAT method which is the International Union for Conservation of Nature (IUCN) standard. An assessed alien taxa with adequate data is classified into massive (MV), major (MR), moderate (MO), minor (MN), or minimal concern (MC) depending on the severity of the impact caused on recipient ecosystem. Additional information such as the mechanisms and location of impact are also recorded. An example of an EICAT dataset is:
# View EICAT data
head(eicat_acacia, 10)
#> # A tibble: 10 × 3
#> scientific_name impact_category impact_mechanism
#> <chr> <chr> <chr>
#> 1 Acacia saligna MC (1) Competition
#> 2 Acacia saligna MC (12) Indirect impacts through interaction …
#> 3 Acacia saligna MC (1) Competition
#> 4 Acacia saligna MC (1) Competition; (9) Chemical impact on th…
#> 5 Acacia mearnsii MC (6) Poisoning/toxicity
#> 6 Acacia longifolia MC (9) Chemical impact on ecosystems
#> 7 Acacia dealbata MC (9) Chemical impact on ecosystems
#> 8 Acacia dealbata MC (9) Chemical impact on ecosystems
#> 9 Acacia saligna MC (9) Chemical impact on ecosystems
#> 10 Acacia dealbata MC (12) Indirect impacts through interaction …
We compute the impact indicator per species by combining the impact
score and occurrence of individual species.
compute_species_indicator() use max (maximum), mean or max_mech
(sum of maximum score per mechanism) method to compute the impact per
species. Interval calculation is not implemented yet.
# Impact indicator per species
species_value <- compute_species_indicator(
cube = acacia_cube,
impact_data = eicat_acacia,
method = "mean",
ci_type = "none"
)
# Visualise species impact
plot(species_value)
The impact per site is a risk map that shows the impact score for each
site, where multiple species can be present. To compute the impact map
per site, aggregated scores across species at each site are needed. The
compute_site_indicator() uses max, sum and mean metrics to
aggregate impact scores within species and across species in a site as
proposed by Boulesnane-Guengant et al. (2025). The combinations of
within species aggregation metrics for each species and across species
for each site leads to five methods of calculating an impact indicator,
namely,
siteImpact <- compute_site_indicator(
cube = acacia_cube,
impact_data = eicat_acacia,
method = "mean_cum"
)
# Impact map
# Visualise last four years for readability
plot(x = siteImpact, region = southAfrica_sf, first_year = 2021)
To compute the impact indicator of alien taxa, we sum all the yearly
impact scores of each site of the study region. The impact indicator use
one of the methods named above in the impact per site. The ci_type
argument allows the function compute_regional_indicator() to calculate
the confidence interval of the regional impact indicator using a
bootstrapping method via the dubicube package. The seed = 123
makes the bootstrap to generate same random sample for reproducibility.
# Regional impact indicator
impactIndicator <- compute_regional_indicator(
cube = acacia_cube,
impact_data = eicat_acacia,
method = "mean_cum",
ci_type = "perc",
boot_args = list(seed = 123)
)
#> [1] "Performing whole-cube bootstrap."
# Visualise impact indicator
plot(impactIndicator, trend = "smooth")
To estimate how the impact indicator is sensitive to a taxon and
indicating how the taxon e.g., species can disproportionately influence
the impact indicator. We can use a leave-one-species-out
cross-validation (LOSO-CV) technique. This technique evaluates the
extent to which the regional indicator is driven by individual species
by systematically recalculating the indicator after excluding one
species at a time and comparing the resulting values to those obtained
from the full occurrence cube. The deviation between the left out
species indicator and the full-data indicator provides a measure of
error attributable to species omission and can be summarised using the
root mean square error (RMSE) for each year. RMSE values indicate
whether observed trends are robust or disproportionately influenced by
particular species. We use the cross_validation_cube() from
dubicube package to perform the LOSO-CV. Details and tutorial about
the LOSO-CV calculation can be found here
https://b-cubed-eu.github.io/dubicube/articles/group-level-sensitivity.html
The above plot shows the RMSE of the impact indicator of Acacia species is South Africa. The figure indicates the values from the year 2019 are sensitive or dominated by some individual species in the indicator.
The above figure shows the impact indicator value and LOSO-CV Error of each species from the year 2017. The LOSO-CV indicates that the impact indicator value will increase (if positive) or decrease(if negative) by the corresponding value of where the species lies in the y-axis. The impact indicator is majorly influenced by Acacia saligna and Acacia mearnsii from the year 2019.