--- title: "Retrieving Species Occurrence Data and IUCN Red List Status for Protected Areas" output: rmarkdown::html_vignette vignette: > %\VignetteIndexEntry{Retrieving Species Occurrence Data and IUCN Red List Status for Protected Areas} %\VignetteEngine{knitr::rmarkdown} %\VignetteEncoding{UTF-8} --- ```{r echo=FALSE,include=FALSE} library(redlist) library(kableExtra) library(dplyr) ``` ```{r, include = FALSE} knitr::opts_chunk$set( collapse = TRUE, eval=FALSE, comment = "#>" ) ``` ## Introduction Understanding the conservation status of species within protected areas is essential for effective biodiversity management and monitoring. This tutorial demonstrates how to obtain species occurrence records from the Global Biodiversity Information Facility (GBIF) for a specific geographic area and retrieve their corresponding IUCN Red List categories. We use Wari-Maro and Monts Kouffé National Parks in Benin as a case study, focusing on chordate species recorded since 2000. The workflow combines spatial filtering of occurrence data with automated queries to the IUCN Red List API, providing a reproducible approach for assessing the conservation status of species assemblages in any protected area. ## Setup and Dependencies We begin by loading the required packages. The `rgbif` package provides access to GBIF data, `redlist` interfaces with the IUCN Red List API, and `sf` handles spatial operations. The `dplyr` and `ggplot2` packages support data manipulation and visualization respectively. ```{r setup, warning=FALSE, message=FALSE} # Define required packages packages <- c("rgbif", "redlist", "sf", "dplyr", "ggplot2") # Install missing packages and load all for (pkg in packages) { if (!requireNamespace(pkg, quietly = TRUE)) { install.packages(pkg, repos = "https://packagemanager.posit.co/cran/latest", dependencies = TRUE) } library(pkg, character.only = TRUE) } ``` ## Downloading Occurrence Data from GBIF ### Setting Download Predicates GBIF allows users to filter occurrence data using predicates, which define specific criteria for the records to retrieve. For this analysis, we establish three predicates to ensure data quality and relevance. We limit records to those from 2000 onwards, restrict the basis of record to human and machine observations to exclude fossil and preserved specimens, and require valid geographic coordinates for spatial analysis. ```{r predicates} # Temporal filter: records from 2000 onwards year_pred <- pred_gte("year", 2000) # Record type filter: observational data only basis_of_record <- pred_in("basisOfRecord", c("MACHINE_OBSERVATION", "HUMAN_OBSERVATION")) # Spatial quality filter: only georeferenced records has_coordinate <- pred("hasCoordinate", TRUE) ``` ### Preparing the Spatial Boundary To restrict occurrence data to our study area, we need to prepare the park boundary in a format compatible with GBIF's download API. This involves several spatial processing steps. First, we import the shapefile and verify its geometric validity. Since the boundary consists of multiple polygons representing different park sections, we merge them into a single polygon using `st_union()`. To reduce complexity and file size, we simplify the boundary by retaining vertices at 200-meter intervals. ```{r spatial-prep} # Import park boundary shapefile wama_moko <- read_sf("wama_moko/wari-maro_monts kouffé.shp") # Verify polygon validity st_is_valid(wama_moko) # Merge into single polygon and simplify wama_moko_single <- st_union(wama_moko) %>% st_simplify(dTolerance = 200) %>% st_reverse() # flips the polygon vertices to follow counterclockwise ``` GBIF requires polygon coordinates in geographic (longitude/latitude) format rather than projected coordinate systems. We check the current coordinate reference system and transform it to WGS 84 (EPSG:4326), the standard geographic coordinate system. ```{r crs-transform, eval=FALSE} # Check current CRS st_crs(wama_moko_single)$input # Transform to geographic coordinates wama_moko_single <- st_transform(wama_moko_single, crs = 4326) ``` Finally, we convert the polygon to Well-Known Text (WKT) format, which GBIF uses to define spatial predicates. ```{r wkt-conversion, eval=FALSE} # Convert to WKT format for GBIF wama_moko_wkt <- st_as_text(wama_moko_single) # Create spatial predicate wama_moko_pred <- pred_within(wama_moko_wkt) ``` ### Configuring GBIF Authentication Downloading data from GBIF requires a free user account. You need to configure your credentials once by adding them to your `.Rprofile` file. This approach keeps your authentication information secure and persistent across R sessions. ```{r auth-setup} # Open .Rprofile file rl_open_file() ``` Add these lines to .Rprofile (replace with your actual credentials): ``` gbif_user = "your_username" gbif_pwd = "your_password" gbif_email = "your_email@example.com" ``` Save the file and restart R/RStudio ### Submitting and Retrieving the Download With all predicates defined and authentication configured, we submit the download request to GBIF. The download process is asynchronous, meaning GBIF processes the request on their servers while you can continue working. The function returns a download key that you use to check status and retrieve results. ```{r download-request} # Submit download request wama_moko_occ <- occ_download( year_pred, basis_of_record, has_coordinate, wama_moko_pred, format = "SIMPLE_CSV" ) ``` Monitor the download status using the download key provided in the output. Once complete, retrieve and import the data into R. ```{r download-retrieve} # Replace with your actual download key from, wama_moko_occ download_key <- "0007614-251120083545085" # Wait for download to complete occ_download_wait(download_key) # Retrieve and import data wama_moko_df <- occ_download_get(download_key) %>% occ_download_import() ``` ### Filtering and Processing Occurrence Records For this analysis, we focus on chordate species, which include vertebrates such as mammals, birds, reptiles, amphibians, and fish. We filter the downloaded records to this taxonomic group and remove duplicate records for the same species at identical coordinates, which often represent repeated sampling events rather than distinct occurrences. ```{r data-filtering} species_of_interest <- wama_moko_df %>% filter(phylum == "Chordata", species != "") %>% distinct(species, decimalLongitude, decimalLatitude, .keep_all = TRUE) %>% select(species, decimalLongitude, decimalLatitude, year) # View processed data species_of_interest ``` ```{r echo=FALSE, eval=TRUE} species_of_interest <- readRDS("gbif_iucn/wama_moko_species_of_interest.rds") ``` The resulting dataset contains `r nrow(species_of_interest)` occurrence records representing `r length(unique(species_of_interest$species))` unique species. We can visualize the temporal distribution of sampling effort to identify patterns in data collection intensity. ```{r install-ggplot2, warning=FALSE, message=FALSE, include=FALSE, eval=TRUE} if (!requireNamespace("ggplot2", quietly = TRUE)) { install.packages(pkg, dependencies = TRUE) } library(ggplot2) ``` ```{r temporal-viz, fig.width=7, fig.height=4, fig.alt='Temporal distribution', eval=TRUE} species_of_interest %>% count(year) %>% ggplot(aes(x = year, y = n)) + geom_col(fill = "gray20") + labs(x = "Year", y = "Number of Records") + scale_x_continuous(breaks = seq(2000, 2023, 1)) + theme_minimal() + theme( axis.text.x = element_text(angle = 90, vjust = 0.5), axis.text = element_text(size = 12), axis.title = element_text(size = 14) ) ``` ## Retrieving IUCN Red List Categories ### API Configuration The `redlist` package requires an IUCN Red List API token for authentication. Detailed instructions for obtaining and configuring your API token are available in the [package documentation](https://stangandaho.github.io/redlist/articles/get_data.html). ### Querying Species Assessments We systematically query the IUCN Red List API for each unique species in our occurrence dataset. The process includes rate limiting to avoid overwhelming the API and error handling to manage species that are not found in the database. For species with multiple assessments, we retain only the most recent evaluation to reflect current conservation status. ```{r iucn-query} # Initialize storage for unfound species not_found_sp <- list() avail_sp <- unique(species_of_interest$species) # Query IUCN status for each species genus_species <- lapply(avail_sp, function(x) { # Display progress progress <- round(which(avail_sp == x) * 100 / length(avail_sp), 2) message(paste0(progress, "%")) # Rate limiting to respect API constraints Sys.sleep(0.5) # Parse genus and species names splited <- strsplit(x, "\\s")[[1]] # Query Red List API with error handling search_result <- tryCatch({ rl_scientific_name( genus_name = trimws(splited[1]), species_name = trimws(splited[2]) ) %>% mutate(year_published = as.numeric(year_published)) %>% slice_max(year_published, n = 1, with_ties = FALSE) }, error = function(e) { not_found_sp[[x]] <<- x return(NULL) }) search_result }) ``` ```{r echo=FALSE, eval=TRUE} not_found_sp <- readRDS("gbif_iucn/not_found_sp.rds") genus_species <- readRDS("gbif_iucn/genus_species.rds") ``` ```{r eval=TRUE} # Compile species with IUCN assessments species_with_iucn <- bind_rows(genus_species) ``` ### Analyzing Conservation Status Some species may not be found in the IUCN database due to taxonomic revisions, synonymy issues, or lack of formal assessment. Species without IUCN status `r paste0("(n =", length(not_found_sp), "):")` `r paste0("*", paste(names(unlist(not_found_sp)), collapse = ", "), "*")`. The conservation status distribution provides insight into the vulnerability of the chordate assemblage within Wari-Maro and Monts Kouffé National Parks. We summarize the frequency of each Red List category to identify the proportion of threatened species requiring conservation attention. ```{r status-summary, eval=TRUE, include=FALSE} # Summary of IUCN Red List categories species_with_iucn %>% count(red_list_category_code) %>% arrange(desc(n)) ``` ```{r eval=TRUE, echo=FALSE} species_with_iucn %>% count(red_list_category_code) %>% arrange(desc(n)) %>% kableExtra::kable("html") %>% kableExtra::kable_styling("striped") %>% kableExtra::scroll_box(width = "100%") ``` ## Conclusion The alysis of chordate species in Wari-Maro and Monts Kouffé National Park revealed 54 species with available IUCN assessments. The majority (83%) are classified as Least Concern, indicating relatively stable populations. However, the presence of threatened species warrants attention: two species are Endangered (*Loxodonta africana*, and *Terathopius ecaudatus*), one is Vulnerable (*Panthera leo*), and five are Near Threatened, together representing 15% of the assessed assemblage. One species remains Data Deficient, highlighting ongoing knowledge gaps. This distribution suggests that while the parks harbor predominantly common species, they also provide critical habitat for species facing elevated extinction risk. It is important to note that since our data spans from 2000 onwards, some recorded species may no longer be present in the parks; however, their historical occurrence indicates that suitable habitat conditions existed and potentially remain, suggesting these areas could support recolonization or continued persistence if populations remain viable.