Cassipourea toroensis

MONRO, A. K., SANTAMARÍA-AGUILAR, D., GONZÁLEZ, F., CHACÓN, O., SOLANO, D., RODRÍGUEZ, A., ZAMORA, N., FEDELE, E. & CORREA, M., 2017, A first checklist to the vascular plants of La Amistad International Park (PILA), Costa Rica-Panama, Phytotaxa 322 (1), pp. 448-450 : 448-450

Cassipourea toroensis ined ( Rhizophoraceae )

Clusia sp. B . ( Clusiaceae )

Clusia sp. C . ( Clusiaceae )

Columnea sp. A . ( Gesneriaceae )

Dieffenbachia sp. A . ( Araceae )

Graffenrieda sp. A . ( Melastomataceae )

Inga sp. B . ( Fabaceae )

Marila sp. A . ( Calophyllaceae )

Phyllanthus sp. A . ( Phyllanthaceae )

Pouteria sp. A . ( Sapotaceae )

Psilochilus sp. A . ( Orchidaceae )

Sacoglottis sp. A . ( Humiriaceae )

Stenospermation sp. A . ( Araceae )

Styrax sp. B . ( Styracaceae )

Biodiversity of vascular plants

We record 3,046 vascular plant species, 26 of which are lycopods, 433 are ferns and 2,586 are seed plants. The twenty most species-rich families are listed in Table 2 and the twenty most species-rich genera in Table 3. Tables 4 lists species diversity by life-form and habit and Table 1 by biodiversity zone and elevation.

The twenty most species-rich families account for 57% of the species diversity of PILA. These largely follow the twenty most species-rich families recognized for Costa Rica (Hammel, pers. comm.) with the notable exception of Fabaceae and Euphorbiaceae , which are remarkably absent from the list of the twenty most species-rich genera in PILA. Of the twenty most species-rich families, ten are predominantly composed of herbaceous species, four are predominantly composed of shrubby species and a single family, Lauraceae is predominantly composed of tree species. Of these families ten are predominated by terrestrial species and six by epiphytic species and four are ferns.

The twenty most species-rich genera account for 24% of the species diversity of PILA ( Table 3). It is surprising that a fern and not an orchid genus, was the most species-rich in a tropical forested area of this size. Indeed, amongst the twenty most species-rich genera, a total of five are ferns, Elaphoglossum Schott ex Smith (1842: 148) , Asplenium Linneaus (1753: 1078) , Thelypteris Schmidel (1763: 45) , Diplazium Swartz (1801: 61) and Hymenophyllum Smith (1793: 418) and only three orchids, Epidendrum Linneaus (1763: 1347) , Pleurothallis R. Brown (1813: 211) and Maxillaria Ruiz & Pavón (1794: 116) . Of the twenty genera, fourteen are predominantly composed of herbaceous species and four of woody species. In terms of habit, nine are predominantly composed of terrestrial species and seven of epiphytic species.

At both the rank of family and genus there is a dominance of herbaceous terrestrial over herbaceous epiphytic groups and when species composition is categorized by plant form and habit ( Table 4) these trends are confirmed to some extent. Terrestrial species account for 78% of the species, whilst woody species are more evenly matched in species number with herbaceous species, 52% and 48% respectively.

The biodiversity zones identified by Monro et al. (2009) are congruent with Holdridge’s ‘life zones’ albeit the resolution is finer. When species diversity is compared by biodiversity zone cloud forest (1,601 –2,100 m) has the highest recorded number of species, followed by low elevation oak forest (2,101 –2,600 m), mixed forest in transition to cloud forest (1,201 –1,600 m) and mixed forest in transition to high elevation oak forest (2,601 –3,100 m). There is therefore a clear trend of species diversity increasing with elevation up to 1,601 –2,100 m and then dropping sharply between 2601 and 3500 m ( Fig. 1, Table 1). Such a ‘mid elevation species bulge’ is observed by Moran (2008) and Christenhusz (2009) for ferns and is suggested by Rahbek (2005) as a common phenomenon in animal and plant diversity where distribution is restricted by a lower (sea) and upper (sky, glacier) limit. However, it is not a phenomenon commonly observed for vascular plant floras in the tropics.

It is also possible that this trend is a consequence of sample bias, the upland parts of PILA being more accessible than the lowland parts. A consequence of access into PILA being much easier from the Pacific than the Caribbean side meaning that low elevation Caribbean areas may have been sampled less frequently. Nevertheless, two observations make this scenario unlikely. Firstly, Hammel et al. (2004, Fig. 4) compared sampling intensity in Costa Rica as measured by number of collections and this indicated that low-elevation (0–800 m) Caribbean slopes are as well sampled as high elevation areas (1,501–3,800 m). Secondly, an analysis of the sampling effort which generated 42% of the records used in this checklist demonstrates that the smallest sample effort was applied to low elevation forest, low elevation oak forest, high elevation oak forest and Páramo (Monro et al., in press). It is also possible that this trend is a product of the area of habitat in each elevational range, the larger an area the greater the number of possible niches. Figure 1 compares species number and surface area for each elevation. If species number were to reflect the area of available habitat then the trend for species number would be expected to match surface area and this is largely what can be observed with the exceptions of the 701–1,200, 1,601–2,100 and 3001–3500 m elevation range where diversity remains high despite a reduced surface area .

Species endemic to PILA