Panama is a small country, but it is entirely in the moist tropics and has substantial climatic and topographic variability. It is thus one of the richest areas for plant species in the world, and since it was at one time (before human intervention) nearly entirely covered in dense forest, it is rich in tree species Condit et al. (2011). A checklist of Panama's flora was published in D'Arcy (1987), including a flag for growth form, and as early as 1990, R.B. Foster (pers. comm.) extracted those marked as trees. Since, Foster, R. Pérez, S. Aguilar, and I have updated that checklist of trees regularly, and it formed the basis of A Field Guide to the Trees of Panama and Costa Rica (Condit et al. 2011). The Guide, however, includes only a quarter of the species. In the past two years, I have made exhaustive updates to the tree checklist and estimated the geographic range of every species.
In the past 10 years, the BIEN database has provided a new tool for examining neotropical flora and their ranges. It includes 11 million records of individual specimens, including a vetted Latin name and the geographic coordinates. After studying the most recent monographs on each taxonomic group, especially the Flora Mesoamericana and the Manual de Plantas de Costa Rica, I matched the Panama tree list against BIEN. In addition, BIEN provided more potential tree species names in Panama which I examined for validity.
This exercise has immediate relevance to the current attempt to assess the conservation status of all the world's tree species.
The most important source for updating the Panama checklist is the Flora Mesoamericana, with three volumes published 2009-2015. They are organized by family and provide details on every species known in Central America, including a thorough list of synonyms, or alternative names for each. But only xxx of the xxx families from the 50-ha plot are published, as the Flora is still in the works. For those families not yet finished in the Flora Mesoamericana, another important recent source is the Manual de Plantas de Costa Rica, volumes 1-6 published 2007-2012. Once again, however, the Manual is not yet finished, and since the focus is Costa Rica, some Panama species are omitted.
For those families not covered in the two major works, we located the most recent monographs. Several important families have been thoroughly reviewed in the past 20 years. Moreover, there are cases where taxonomy has changed since Flora Mesoamericana. In all cases, we have done are best to track down papers on taxonomy of the 50-ha plot, and we repeatedly make use of the Missouri Botanical Garden online database, Tropicos, to resolve names. A full list of genera and the monographs we found is provided (Appendix 1).
The fact that tropical flora are so diverse has limited the taxonomic work on individual groups, but this offers an ironic advantage. Many groups have only one expert, so there is not much disagreement. Nevertheless, there are cases where two taxonomists publish alternative versions of species names, forcing us to choose. Whenever available, the Flora Mesoamericana was usually our final arbiter, and we (almost) always followed the Tropicos website (and since both the Flora and the website come from the Missouri Botanical Garden, they nearly always agree). In the Appendix of species we include notes on names we selected in any unusual or debatable cases.
The BIEN database assembles many different sources, including GBIF (Global Biodiversity Information Facility) and Tropicos (Missouri Botanical Garden data), in one database (Maitner et al. 2017), and the data are readily available for download. Nearly all the BIEN records are herbarium specimens, but there are tree plots and other species records. A majority have latitude-longitude associated, though in some cases it was estimated from a location's description. The taxonomic updates we have maintained since the plot started in 1982 provide a large number of old names by which the tree species have been known. When those out-of-date names appeared in the BIEN database, we were able to match to the current name so that the specimens were not lost.
Collecting together all those records with coordinates, we constructed the minimum convex polygon around each species' observations, subtracting large bodies of water, and calculated its area. This is a description of a species' range that requires few assumptions, and we carefully designate it the 'observed range' as opposed to the 'geographic range'. The polygon around observations can easily underestimate the true range, but can also overestimate, but it requires a minimum set of assumptions, in particular, nothing about the environmental conditions a species requires is assumed. We are presenting the realized range of each species, as opposed to the potential range.
For species lacking any coordinates at all, or having just one or two records with coordinates, we estimated a range extent using the area of provinces and countries in which they were observed.