Manual de Plantas de Costa Rica 211 
on the ATTA database is from Standley’s Pacific tierra caliente, the greatest number 
of species is recorded from the tierra templada middle elevations (Fig. 4). Although 
these data are by no means complete, nor based on rigorous statistical sampling, they 
do bear on discussions concerning the so-called mid-elevation “bulge” in species rich- 
ness in the tropics (cf., Boyle, 1996; Gentry, 1988). Boyle (1996, and pers. comm.), 
whose transects counting woody species of = 2.5 cm DBH demonstrate peak diversity 
between 750 and 1000 m elevation in Costa Rica (Caribbean slope, Braulio Carrillo 
National Park), has also shown that local diversity (based on rigorous statistical sam- 
pling) is strongly correlated with regional diversity (based on simple floristic check- 
lists). Our data on the total vascular flora of Costa Rica, separated into Standley’s re- 
gions, confirm Boyle’s observations, i.e., that the highest diversity is at the middle 
elevations. However, the ferns and fern allies, counted separately, do not follow the 
same pattern; instead, they show a gradual increase in diversity upward from tierra 
caliente to tierra fria. On the other hand, Lellinger (1985) found that in Panama the 
highest diversity of ferns in montane forests is between 500 and 1500 m (see also 
Moran, 1995). 
The estimates of habit-type composition of the four regions are shown graphically 
in Fig. 5 (p. 213). The ATTA database was queried for the number of species/elevational 
region in the three most diverse plant families in each habit type, as calculated from the 
total flora, and the number of species/habit type was estimated as the product of the 
number of species per region and the fraction of the family represented by the habit type 
in the total flora. For example, the number of epiphytic Orchidaceae in the tierra tem- 
plada is calculated as 451 (number of orchid species collected in the 800—1500 m band) 
xX 0.86 (percentage of orchid species in the Costa Rican flora that are epiphytes) = 388. 
The clearest (and perhaps only) trend suggested by these graphics (Fig. 5) is that 
epiphyte species richness is greatest in the tierra templada, i.e., the middle elevations. 
However, Standley (1937b: 22) suggested that the upper limit of this region, on the Pa- 
cific slope, is sharply marked by “the line at which trees begin to be heavily infested 
with orchids, mosses and other epiphytes.” He felt that in the tierra templada, epiphytes 
were “much less diversified than in the tierra fria’” (Standley, 1937b: 25). Although our 
data indicate that the greatest number of species occurs in the tierra templada, again we 
should point out that the greatest number of collections in the database of the largely 
epiphytic family Orchidaceae is also from that region. A definite mid-elevation bulge 
in epiphyte diversity has been noted elsewhere (Gentry & Dodson, 1987), and those 
same authors stated that epiphyte biomass seemed to be highest at somewhat higher el- 
evations than epiphyte diversity. Thus, given Boyle’s studies of trees, shrubs, and lianas 
showing highest species diversity at middle elevations, the overall diversity of vascular 
plants is certainly highest in the tierra templada, considering its rich epiphyte flora. For 
information on theories to explain the bulge (e.g., that a bulge in species diversity 
somewhere between sea level and the highest peak is the expected or null hypothesis), 
see Colwell & Hurtt (1994). 
