Table 9.22. Mean nutrient concentration (mg/g ± SEM) of
standing crop and litterfall of epiphytic components for 29
collection intervals between 16 May 1988 and 10 January 1989.ComponentNutrient
N Ca Mg K
Standing crop, TotalLitterfall
Vascular plantsBryophytesDead organic matter18.4
(0.7)10.7
(0.7)
15.8
(0.8)
15.5
(1.3)1.3
(0.003)0.8
(0.5)
1.1
(0.04)
0.7
(0.02)11.9
(1.2)58.6
(0.9)
8.4
(0.5)
10.8
(1.3)1.9
(0.1)2.5
(0.3)
1.7
(0.06)
1.2
(0.1)6.0
(0.4)0.7
(0.1)
0.3
(0.02)
0.2
(0.03)litter biomass (7.5 tons/ha/yr; Nadkarni and Matelson
1992b). The nutrient transfer via EM litterfall is up to
8% of the annual nutrient transfer in total fine litterfall
(Table 9.21). If we assume that this primary forest is in
a steady state, then over an annual cycle, litter decom-
position equals litter deposition (Olson 1963). The
annual decay rate (KJ of this material is the annual
litter input divided by the forest floor pool, and was
calculated as 1.3 for EM biomass (Table 9.21). The frac-
tional turnover time (1/KJ for EM biomass would be
0.8, or approximately 10 months. We found a more
rapid decay rate and shorter turnover time for EM than
measured for terrestrially rooted material at the same
time and place (Ka = 0.7, l/Ka = 1.4; Nadkarni and Matel-
son 1991). However, we found a much slower decay
rate for nutrients in EM litter (except potassium) than
for nutrients in terrestrially rooted fine litter. Turnover
time for all nutrients except potassium were four to six
times longer for EM than terrestrially rooted material.
The turnover time for potassium was 10-fold faster in
EM than terrestrially rooted material (Table 9.21).
These results indicate that at least a portion of EM
is recalcitrant and highly resistant to decomposition
and mineralization. However, certain components of
EM decompose very rapidly. Input of bryophytes was
76% of fallen EM, but only 22% of the EM standing
crop, which suggests that this component decomposes
quickly. The calculated decay rate for bryophytes is
4.3, with a turnover time of only 0.23 yr. Conversely,
dead organic matter appears to have a much slower
decay rate, comprising only 13% of input, but 58%
of the EM standing crop; Ka for this material is 0.28,
and the turnover time is 3.6 yr. Vascular plants are
intermediate; Ka is 0.68 and turnover time is 1.5 yr.
Further studies of particular components are needed
to determine the timing of nutrient mineralization and
release from this material.
The deposition of EM is patchy; over half of the EM
fell in less than 2% of the collections. One implication
for plants rooted in the forest floor is that nutrient depo-
sition from EM concentrates input in unpredictable
locations. This contrasts to TM litter, which is distrib-
uted fairly evenly across the forest floor. Second, nu-
trients deposited in EM that ride down treefalls and
large branch falls co-occur with higher levels of light
associated with resulting gaps. This pulse of nutrients
released from EM may alter nutrient availability in the
immediate vicinity of regenerating gap species. An-
other factor is that there may be delays in the death of
epiphytes and subsequent release of their nutrients
after they fall to the forest floor (see Matelson et al.,
"Longevity of Fallen Epiphytes," pp. 344-345).
Fallen EM has been either overlooked or inaccu-
rately measured in nearly all nutrient cycling studies.
In a tropical evergreen lowland forest in Cameroon,
Songwe et al. (1988) found 105 kg/ha/yr from epiphytic
mosses and ferns (0.8% of total fine litter). In a Jamai-
can montane forest, Tanner (1980a) reported 4-180 kg/
m2/yr (0-3% of total fine litter) as EM (mainly brome-
liads). In a lower montane rain forest, Veneklaas (1991)
documented 220 kg/ha/yr of fallen vascular and non-
vascular EM (3% of total fine litter) using 0.5 x 0.5 m
wire frames. If we had collected only the epiphytic
material identified in our TM litterfall collectors, we
would have underestimated EM by 72%, as only 14 g/
m^2 /yr in the bryophyte category fell into our collectors.
It is thus critical to use collecting areas of the appro-
priate sizes for fallen EM.9.2.4. Decomposition Fluxes
Terrestrially rooted material. The slower rate of lit-
ter decomposition in tropical montane forests, com-
pared to tropical lowland forests, has been attributed
to the high frequency of misty conditions, low air and
soil temperatures, lack of drying-rewetting cycles, a
high degree of sclerophylly, and waterlogged soils
(Leigh 1975, Tanner 1981, Vitousek 1984). Nutrients334 Ecosystem Ecology and Forest DynamicsP