relative contribution of each of these sources to tropi-
cal forest communities is unknown. From an ecosys-
tem standpoint, it is important to distinguish between
them in order to assess the vulnerability of a forest to
disruptions in external and internal nutrient cycles.
Precipitation inputs. A major pathway of nutrient in-
put to forest ecosystems is via precipitation; nutrients
derived from atmospheric sources (aerosols, dissolved
chemicals and gases, dust, and particulates) as well
as pollutants can arrive as wet or dry fallout as allo-
chthonous sources. These can be intercepted, retained,
modified, and conducted within the ecosystem as
authochthonous sources before being stored in or
exiting the forest.
Precipitation and cloud water chemistry, and
atmospheric deposition and retention at tropical mon-
tane forest sites have been an increasing focus of re-
search (Veneklaas 1990, Asbury et al. 1994). The pre-
cipitation pathway is especially important for tropical
montane forests. Cloud water accounts for a signifi-
cant portion of the total ion deposition in montane
forests because (a) ion concentrations in cloud water
are 3-10 times greater than those in precipitation
(Weathers et al. 1988), (b) montane areas are fre-
quently immersed in cloud (Bruijnzeel and Proctor
1995), and (c) foliage, branches, and epiphytic vege-
tation are aerodynamically rough surfaces. Ion depo-
sition is typically greater in montane cloud forests
when compared to lower elevation forests, which re-
ceive only wet and dry deposition (Fowler et al. 1988,
Johnson and Lindberg 1992).
Increased combustion of biomass and fossil fuels,
and changes in land use practices have led to an in-
crease in the emissions of nitric oxide and ammonia
at tropical latitudes (Keller et al. 1991). Reports from
lowland sites have indicated regional effects of bio-
mass burning and conversion of forests to pasture on
precipitation and air chemistry (Lewis 1981, Andreae
et al. 1988, Keller et al. 1993). These activities poten-
tially lead to greater concentrations of NO 3 _ and NH 4 +
in cloud water and precipitation in tropical montane
forests.
At the leeward cloud forest study site, cloud water
and precipitation depth were estimated using an arti-
ficial surface of plastic "foliage" mounted above a
funnel connected to a tipping bucket rain gauge. The
gauge was mounted on a boom on a 27-m meteoro-
logical tower in a small gap in the forest. Bulk cloud
water and precipitation (BCWP) samples for ion
analyses were collected with a passive cloud collec-
tor (Falconer and Falconer 1980). Bulk precipitation
(BP) was collected with a polypropylene funnel and
bottle. Windspeed (u) was measured with a 3-cup
anemometer. Events were categorized as u < 2 m/s or
u > 2 m/s. All instruments were connected to an auto-
mated datalogger.
Average weighted mean concentrations of H, NO 3 ~,
and NH 4 + in cloud water were 7, 14, and 17 times
greater (respectively) than those in precipitation with
windspeeds >2 m/s and 10, 30, and 45 times greater
for precipitation where wind was <2 m/s (Table 9.8).
The molar ratio of sodium and magnesium in cloud
water was similar to that in seawater, indicating that
these ions originated primarily from wind-driven
marine aerosols (Blanchard 1983).
Cloud water samples collected toward the end of
the dry season (late February to early May) had greater
concentrations of NO 3 ~, NH 4 +, K, and Ca than those
collected at the end of the transition season and in the
early dry season (December and January). Sampling
periods toward the end of the dry season coincided
with noticeable haze layers, presumably due to burn-
ing of agricultural and forest biomass. It is likely that
ion enrichment in cloud water samples collected to-
ward the end of the dry season resulted from the in-
corporation of gases and particles derived from burn-
ing (Andreae et al. 1988, Lobert et al. 1990).
Concentrations of inorganic nitrogen in cloud
water and mist measured in Monteverde were highTable 9.8. Ion concentrations of canopy water sources.
Ion Concentration (umol/liter)Sample Type
Cloud waterMistPrecipitation
iz> 2 m/s
Precipitation
u< 2 m/sH+
132.1
(2.5-489.8)
30.8
(0.2-251.2)
19.0
(3.7-91.2)
13.8
(2.8-55.0)N0 3 -
102.8
(32.1-382.9)
37.3
(10.8-106.1)
7.5
(1.3-47.1)
3.4
(bdl-46.8)NH 4 +
148.9
(47.2-739.4)
45.3
(2.8-131.1)
8.9
(0.6-87.2)
3.3
(bdl-127.8)Na+
365.0
(144.4-713.4)
183.0
(45.7-839.5)
75.4
(7.4-765.6)
7.8
(2.6-107.0)Mg+2
63.4
(18.1-121.8)
21.9
(4.9-137.4)
8.8
(0.8-104.9)
1.2
(bdl-8.2)Ca+2
35.2
(8.7-70.1)
16.8
(6.0-78.6)
6.6
(3.0-49.4)
3.4
(bdl-11.7)K+
30.7
(7.7-76.5)
13.5
(2.3-71.4)
4.1
(1.3-94.6)
2.1
(bdl-27.6)
Data are volume-weighted mean ion concentrations (umol/liter) in cloud water (n = 15), mist (n = 32), precipitation where mean u> 2 m/s (N
= 54), and precipitation where mean u< 2 m/s (N= 47). Ranges are in parentheses; bdl = below detection limit324 Ecosystem Ecology and Forest Dynamics