Cloud cover, however, may be greater at this time
compared to other times of the year (Fig. 2.3).
Ascending air masses at the ITCZ create a surface
low pressure that must be replaced by air masses from
regions to the north and south. This produces the
surface trade winds associated with global-scale
Hadley cell circulation (Riehl 1979). When the ITCZ
is located to the south of Costa Rica during the tran-
sition and dry seasons, northeasterly trade winds
deliver moist air from the Caribbean Sea to the low-
lands on the Caribbean side of Costa Rica. Maximum
mean wind velocity occurs from sea level up to 2800-
3000 m, above which velocities decrease. When these
air masses encounter the Cordilleras, they are forced
to ascend (orographic uplift), and adiabatic cooling
causes condensation and stratus or stratocumulus
cloud formation. Cloud base is typically at 1400-1700
m during these two seasons. Cloud immersion on the
upper slopes and ridges along the Continental Divide
(particularly on the Brillante Ridge and above Rio
Cafio Negro) may approach 20-25% of the time,
which is similar to the duration of cloud immersion
in some montane forests in northeastern United States
(Vong et al. 1991).
Compared to the wet season, cloud immersion and
precipitation depths across Monteverde are more vari-
able during the transition and dry seasons and depend
strongly on topographic position and exposure to the
trade winds. Slopes and ridges on the east-facing (wind-
ward) side of the Continental Divide typically receive
amounts of precipitation that are more characteristic
of the Caribbean side of Costa Rica, primarily due to
wind-driven precipitation inputs during the transition
season (see Clark and Nadkarni, "Microclimate Vari-
ability," pp. 33-34, Fig. 2.6). Cloud immersion and
wind-driven precipitation occur with decreasing fre-
quency and duration on east-facing slopes and ridges
as the dry season progresses. Slopes and valleys on the
west-facing (leeward) side of the Continental Divide
have precipitation regimes that are more similar to
those on the northern and central Pacific side of Costa
Rica (Fig. 2.6). The San Luis valley, for example, ex-
periences relatively small amounts of wind-driven pre-
cipitation and minimal cloud immersion during the
transition and dry seasons. Wind-driven cloud water
and precipitation inputs represented approximately
a 20% increase over annual precipitation input for the
leeward cloud forest site in the MCFP that was im-
mersed in cloud only about 7% of the time. Intermit-
tent data collected at La Ventana on the Brillante, how-
ever, suggest that much higher inputs to windward
cloud forests occur (Clark 1994; see Clark and Nadkarni,
"Microclimate Variability," pp. 33-34).
A similar pattern for cloud water and precipitation
has been documented across a transect in northern
Panama (Cavelier et al. 1996). The input of "mist" that
augmented precipitation collected with standard rain
gauges ranged between <200 mm (500 m elevation)
and 2295 mm (on a ridge at 1270 m), which repre-
sented 2-60% of total hydrologic inputs. Other esti-
mates of wind-driven cloud water and precipitation
in Central American and northern South American
montane forests range from 70 to 940 mm/yr (7-48%
of total hydrologic inputs; Baynton 1968, Vogelmann
1973, Cavelier and Goldstein 1989, Asbury et al.
1994).
Using a regional hydrologic balance method,
Zadroga (1981) estimated cloud water and precipita-
tion inputs for the headwaters of Rio Pefias Blancas.
He divided the total annual runoff by the total annual
rainfall for streams in the San Carlos drainage basin
(which includes the Rio Pefias Blancas) on the Atlan-
tic slope, and compared these ratios with streams in
the Bebedero basin, which flows from the Pacific
slope. For the San Carlos basin on the Atlantic slope,
the amount of runoff compared to precipitation was
102%, which indicates that more water ran off than
fell as direct precipitation. Runoff exceeded monthly
precipitation, which was attributed either to an un-
derestimation of rainfall due to an insufficient num-
ber of precipitation gauges, or to an underestimation
of the total precipitation depth due to inadequate
sampling of cloud water and wind-driven precipita-
tion (Zadroga 1981). An average annual cloud water
and precipitation depth of up to 9000 mm was calcu-
lated for the upper Pefias Blancas watershed. In con-
trast, runoff was 34.5% of rainfall for the Bebedero
basin on the Pacific slope. Runoff exceeded precipi-
tation only in the dry season months (January-April)
when the river flows were maintained by water re-
leased from bank storage (Zadroga 1981).
Wind-driven hydrologic inputs to tropical cloud
forests may be greater than those to temperate mon-
tane and coastal forests. Studies in the United States
suggest that although cloud water deposition equals
approximately 20-30% of total hydrologic inputs,
absolute depths are less than in tropical cloud for-
ests (Bruijnzeel and Proctor 1993, Dingman 1994,
Cavelier et al. 1996). Tropical cloud forests likely
receive greater wind-driven hydrologic inputs due to
relatively higher tradewind velocities and "wetter"
cloud events, although cloud immersion may be of
similar duration (Vong et al. 1991, Bruijnzeel and
Proctor 1993, Asbury et al. 1994).
Quantification of wind-driven cloud water and pre-
cipitation inputs in Monteverde is a major future chal-
lenge. A transect of cloud water and precipitation
collectors from the Pefias Blancas valley to the San
Luis valley and measurements of other meteorologi-
cal variables are needed to estimate hydrologic inputs.19 The Physical Environment