The dead forest on Trindade Island was not monospecific, says the wood.
The first reports of a dead forest on Trindade Island are from the 18th century. Since then, the tentative identifications of the trees with red wood included Caesalpinia, Acacia, Rapanea, Pisonia, Eugenia and Colubrina, the latter having been confirmed by three independent wood anatomists familiar with Brazilian woods. In the 1960s Johann Becker was the last to sample a live Colubrina glandulosa Perkins var. reitzii on Trindade, which was presumed to be a remnant of the extinct forest. Based on this information, along with the eradication of feral goats from the island in 2005, thousands of C. glandulosa seedlings were reintroduced to Trindade. These trees, which grew well at first, are now collectively dying, less than two decades after planting. Their wood colour is much lighter than that of the dead trees, raising doubts about the latter's correct identification. Herein we report the first detailed descriptions of two wood types from the extinct forest of Trindade, confirming the presence of C. glandulosa and reporting the presence of Paratecoma peroba (Bignoniaceae), a novel occurrence for the island. Radiocarbon dating of a dead C. glandulosa tree confirms that it belongs to the forest which died three centuries ago. The preserved wood proves that the extinct forest was not monospecific and suggests that further sampling of the remaining dead wood may enhance the floristic knowledge of the forest which once covered most of the island with additional species.