Extinction has increased as human activities impact ecosystems. Conservation assessments for the IUCN red list are a fundamental tool in aiding the prevention of further extinction, yet, relatively few species have been thoroughly assessed. To increase the efficiency of assessments, novel approaches are needed to highlight threatened species that are currently data deficient. Many Madagascan plant species currently have extremely narrow ranges, but this may not have always been the case. To assess this, we used high-throughput DNA sequencing for 2-5 individuals of each species - reflecting the paucity of samples available for rare species. We estimated effective population size (N_e) for each species and compared this to census population (N_c) sizes when known. In each case, N_e was an order of magnitude larger than N_c – a signature of rapid, recent population decline. We then estimated the demographic history of each species, tracking changes in N_e over time. Five out of ten species displayed significant population declines towards the present (68–90% decreases). Our results for palm trees indicate that it is possible to predict extinction risk, particularly in the most threatened species. We performed simulations to show that our approach has the power to detect population decline during the Anthropocene, but performs less well when less data is used. Similar declines to those in palms were observed in data deficient species or those assessed as of least concern. These analyses reveal that Madagascar’s narrow endemics were not always rare, having experienced rapid decline in their recent history. Our approach offers the opportunity to target species in need of conservation assessment with little prior information, particularly in regions where human modification of the environment has been rapid.

The "fasta_genepop.zip" folder contains outputs of the STACKS program "populations" that was run on each of the 10 species in the study. The output fasta files were used to estimate effective population size.

The scripts presented here are used in the pipeline to simulate demographic declines.

"format_sims.sh" runs the simulations using "temp_sims.par" as a template (which can be run in batches using "batch_sims.sh").

"fsc26" is the program fastsimcoal2, required for simulations.

"haps.sh" will generate haplotypes from the simulations output with the help of "RAD_Haplotypes.R".

Files with the pattern dp*.txt are example files that provide parameters to haps.sh

The output haplotypes can be inserted into an xml file (e.g. dp.xml) and used with the software program BEAST.