Skip to contents

Newsletter

newsletter.Rmd

CafriplotsR Newsletter

Updates on major milestones for the CafriplotsR package and the Central African forest plots database.

May 2026

New capabilities

Switch between the internal taxonomic backbone and WCVP

CafriplotsR now ships with support for a second taxonomic backbone: the World Checklist of Vascular Plants (WCVP). The internal Cafriplots backbone remains the default — it is the reference used to assemble the database, link individuals, and curate taxa-level trait data — but you can now operate in WCVP space whenever you need to, for example, compare with, or contribute to, a global reference.

The backbone is selected at query time via a single argument:

con <- call.mydb()
query_plots(country = "GABON", backbone = "wcvp", extr)
query_taxa(species = "Dacryodes edulis", backbone = "wcvp", check_synonymy = F)

When backbone = "wcvp", the standard taxonomy columns are replaced in place with the WCVP values. Two extra columns make the substitution transparent:

  • name_source — "wcvp" if the taxon resolved to a WCVP name, "internal" if it fell back silently to the Cafriplots backbone (no WCVP match)
  • alt_taxon_name — the original internal name preserved alongside, so you can always trace back to the Cafriplots reference

The current version of the WCVP can be checked using the following code line (this is the most recent one from early January 2026):

con.taxa <- call.mydb.taxa()
get_wcvp_status(con_taxa = con.taxa)

Internal IDs (idtax_n, idtax_good_n) are never overwritten; the WCVP analogues (wcvp_plant_name_id, wcvp_accepted_plant_name_id) are added as separate columns. This keeps every row anchored to both reference systems at once.

The same toggle is available in the interactive apps:

  • launch_taxonomic_match_app() — the backbone-selection modal now offers WCVP as an alternative whenever WCVP data are present in the taxa database.
  • launch_query_plots() — extraction can be run in WCVP space; the source of every taxonomic name is documented in the output tables via name_source.

The motivation is the same as for citation tracking: transparency about taxonomic provenance. Whenever data leave the database, the user knows exactly which reference system was used to interpret each name, and can switch backbones without re-importing or re-curating anything.

Aggregated taxa-level traits from individual measurements

The package now closes the second half of the virtuous circle presented in Barcelona: individual-level measurements (DBH, height, crown diameter, bark thickness…) recorded in plots can be aggregated by taxon and exposed as taxa-level traits, ready to be grafted onto any future query — exactly like the existing wood-density or seed-mass traits coming from external compilations.

Aggregation is declarative: a small configuration table (trait_aggregation_config) holds rules such as “take the 95th percentile of stem diameter, restricted to individuals identified at species level or below, with at least 5 measurements”. The pipeline reads the rules, computes the aggregated value per taxon, and writes the result into taxa_traits_measures:

add_trait_aggregation(
  source_trait_id = 1,         # stem_diameter
  method          = "percentile",
  method_param    = 95,
  min_n           = 5
)
rebuild_aggregated_taxa_traits(con, con_taxa)

Supported methods: mean, median, min, max, sum, sd, percentile, mode, concat, count.

To keep the transformation method visible — so that p95 of DBH and the raw DBH never get mixed up, and so that several methods on the same source can coexist — aggregated rows are written into derived traits named after the source plus the method (e.g. stem_diameter_p95, stem_diameter_max). If you would rather keep the original trait name, you can pass target_trait_id = source_trait_id explicitly.

Once written, these aggregated traits behave exactly like any other taxa-level trait: query_taxa_traits() returns them transparently, and query_individual_features() / query_plots() graft them onto your individuals through the usual taxa link. From the user’s perspective, there is no second pipeline to run.

Public access and data sovereignty. Aggregated traits are tagged with a dedicated, non-public citation (CafriplotsR_aggregated) and the row-level security policy on taxa_traits_measures is RESTRICTIVE: the public role cannot see them. They are visible only to authenticated users with appropriate permissions. This preserves the principle stated in the Barcelona presentation — only aggregated summaries circulate, raw plot data never leaves your network, and the public-facing portal sees neither.

March 2026

New data

Tervuren xylarium Wood Density Database (TWDD) integrated

A large wood density dataset — the TWDD — has been added to the database. It contains 13,332 samples spanning 2,994 species, 1,022 genera, and 156 plant families across six continents, with 72% of records from Africa.

The TWDD adds 1,164 species, 160 genera, and 8 plant families not previously documented.

Before integration, the taxonomy of the TWDD was standardized against the CafriplotsR taxonomic backbone to ensure consistency with the rest of the database. A total of 12,970 values from the TWDD were added.

Citation requirement: Any use of this dataset must cite the original publication:

Verbiest W.W.M., Hicter P., Beeckman H. et al. (2026). The Tervuren xylarium Wood Density Database (TWDD). Scientific Data, 13, 243. https://doi.org/10.1038/s41597-026-06563-2

New capabilities

Citation tracking for taxa-level traits

Taxa-level trait measurements stored in the database are now linked to their original sources — published studies or trait databases from which values were extracted. When you enrich a species dataset or explore traits through the interactive apps (launch_taxo_backbone_app() and launch_taxonomic_match_app()), a dedicated Data Sources panel lists the citations involved and the number of measurements/observations from each source.

This makes proper attribution straightforward and unambiguous: the panel lists the full reference for each source so that users know exactly what to cite when using these data. This new development reinforces one of the core goals of this package, which is to facilitate the integration and reuse of diverse datasets. Achieving that requires being as transparent as possible about this integration and therefore respecting the data-sharing requirements of the underlying publications and databases.

For R users: The query_taxa_traits() function now accepts include_citation = TRUE to attach citation information directly to the returned data frame. Both wide and long output formats are supported.

Public access mode for two interactive Shiny apps

Two apps in the CafriplotsR toolkit can now be launched without database credentials — useful for collaborators, reviewers, or anyone who wants to explore the data without requesting a personal account:

  • Taxonomic backbone browser (launch_taxo_backbone_app()): Browse the full taxonomic reference, search by name or identifier, explore the taxonomic hierarchy, and extract taxa-level traits as wide or long tables with associated citations.

  • Taxonomic harmonisation & trait enrichment app (launch_taxonomic_match_app()): Upload a species list, standardise names against the CafriplotsR backbone, and enrich with traits from the database — all in a guided, step-by-step interface.

In public mode the apps connect automatically using read-only credentials, so no login is required. Both apps display a clear banner indicating that you are operating in public mode with read-only access.