Content uploaded by Erich Fischer
Author content
All content in this area was uploaded by Erich Fischer on Oct 02, 2017
Content may be subject to copyright.
ORIGINAL ARTICLE
Bizarre Cecropia pachystachya (Urticaceae) hemiepiphytic growth
on palms in the ‘‘Pantanal’’ wetland
Christiane E. Corre
ˆa
1
•Erich Fischer
2
Received: 25 August 2016 / Accepted: 21 November 2016
Botanical Society of Sao Paulo 2016
Abstract In this study, we describe for the first time fac-
ultative hemiepiphytism in Cecropia pachystachya Tre
´cul
growing on palms in the ‘‘Pantanal’’ wetland of Brazil and
investigate ecological factors associated with this unex-
pected phenomenon. We sampled C. pachystachya seeds
lodged in palm stems and recorded the seed rain to low
stems, high stems and to the ground below and away from
palms. We also tested seed germination on soil and stem
substrates and measured the chemical composition of both
substrates. Compared to freestanding conspecifics,
hemiepiphytes of C. pachystachya were rare and displayed
wider trunks but similar height. All hemiepiphytes were
fixed on low stems by aerial roots encircling the palms and
coalesced branches rooted into the ground. Most palm
stems contained C. pachystachya seeds but a few concen-
trated seeds massively. Dispersal of seeds was 15 times
greater to lower than higher portions of stems, and seven
times higher below palms than elsewhere. Germination
rates of C. pachystachya seeds did not differ between soil
and stem substrates. The clayey stem substrate presented
more Zn, K, P, Ca, Mg and Al than the sandy soil, which in
turn presented more Fe and Cu. The exceptional occurrence
of C. pachystachya as a facultative hemiepiphyte in the
‘‘Pantanal’’ wetland results from massive seed rain to lower
regions of palm stems, where germination, mechanical
stability and access to nutrients are conducive to estab-
lishment and subsequent growth. However, the rarity of
hemiepiphytism in C. pachystachya indicates that oppor-
tunities for successful establishment rely on a combination
of uncertain spatiotemporal conditions.
Keywords Aerial roots Attalea phalerata Bat dispersal
Hemiepiphytes Soil composition
Introduction
Accumulation of litter and sediment on vegetation struc-
tures may provide microsites for seed deposition and
growth of a variety of other plants, resulting in the evo-
lution of different epiphytic life forms (Benzing 1990;
Nadkarni and Haber 2009; Hao et al. 2010,2011; Corre
ˆa
et al. 2012). Hemiepiphytism has evolved on multiple
occasions in several angiosperm families, markedly in
Moraceae and Clusiaceae tropical trees (Nadkarni et al.
2001; Wanek et al. 2002). Hemiepiphytes germinate on
other plants (phorophytes) and have no access to the
ground during early growth, deriving water and nutrients
only from the host plant substratum at this stage (Zotz
2013). Aerial roots initially encircle the phorophyte, pro-
viding essential fixation and feeding, and then grow
downward to the soil. Hemiepiphytes may become free-
standing after coalescence of numerous aerial roots and
phorophyte decay. Therefore, availability of water and
nutrients on phorophyte substratum is critical for the initial
establishment of hemiepiphytes, and abundance of suit-
able host plants may limit them. In addition, the success of
Electronic supplementary material The online version of this
article (doi:10.1007/s40415-016-0339-y) contains supplementary
material, which is available to authorized users.
&Christiane E. Corre
ˆa
correa.ce@gmail.com
1
Programa de Po
´s-Graduac¸a
˜o em Ecologia e Conservac¸a
˜o,
Universidade Federal de Mato Grosso do Sul,
Campo Grande, Mato Grosso do Sul 79070-900, Brazil
2
Centro de Cie
ˆncias Biolo
´gicas e da Sau
´de, Universidade
Federal de Mato Grosso do Sul, Campo Grande,
Mato Grosso do Sul 79070-900, Brazil
123
Braz. J. Bot
DOI 10.1007/s40415-016-0339-y
the hemiepiphytic growth form depends on consistent
interactions with seed dispersers to attain appropriate
phorophytes. In general, hemiepiphytic tree species show
effective seed dispersal through zoochoric diaspores
swallowed and defecated by vertebrates, but also by sec-
ondary dispersal by ants (Kaufmann et al. 1991; Passos and
Oliveira 2002; Teixeira et al. 2009; Munin et al. 2011).
Cecropia Loefl. (Urticaceae) is a conspicuous Neotrop-
ical genus of pioneer trees originated in northern Andean
regions of Colombia and Ecuador (Berg et al. 2005). Spe-
cies of Cecropia can occur densely in large second-growth
areas or scattered in treefall gaps of old-growth forests, in
different habitats from wet to dry (Parolin 2002; Damas-
ceno Ju
´nior et al. 2004; Faxina et al. 2015). The presence of
adventitious roots is common in the genus and these can
function for aeration and mechanical support, especially for
individuals growing in swamps (Barlow 1986; Berg et al.
2005; Parolin et al. 2010). Cecropia obtusifolia Bertol. in
Colombia is an extraordinary reported case, as its roots
grow toward the nutrient-rich crown centre of nearby palms
and tightly encircle their stems and petioles often killing the
palms in manner similar to the behaviour of hemiepiphytic
strangler figs (Bernal and Balslev 1996). However, to the
best of our knowledge, the hemiepiphytic growth has not
been described for species of Cecropia (Gentry 1993; Berg
et al. 2005). Here, we describe the occurrence of facultative
hemiepiphytism in Cecropia pachystachya Tre
´cul on At-
talea phalerata Mart. ex Spreng. palms in the ‘‘Pantanal’’
wetland of Brazil. Aside from reporting this phenomenon
for the first time, we also addressed its magnitude in the
local population and evaluated some critical ecological
factors that could allow its occurrence. To address the
magnitude, we asked how frequent and large hemiepiphytic
individuals of C. pachystachya are, compared to free-
standing conspecifics. With regard to ecological factors, we
assessed whether seed arrival and germination differ
between forest ground and palm stems, and whether avail-
able nutrients differ between forest soil and sediment
accumulated on palm stems.
Methods
Study site and species
We sampled C. pachystachya in forest patches (0.5–5 ha) in
the ‘‘Pantanal’’ floodplain, region of Miranda (19340S,
57010W), Southern Mato Grosso do Sul State, Brazil. The
‘‘Pantanal’’ is a major world wetland (160,000 km
2
) in the
centre of South America that includes the upper Paraguay
River basin (Almeida et al. 2015). Climate is classified as Aw
of Ko
¨ppen (Peel et al. 2007), with a marked dry season from
April to September and a rainy season from October to
March. Annual rainfall varies between years, from 800 to
1400 mm. Seasonal flood pulses alternate with severe
drought periods, and the landscape is mainly composed of
fields, temporary or permanent lakes, channels, riparian
forests and forest patches on slightly elevated terrains
(Almeida et al. 2015). Inundation intensity varies between
years and zones, from three to six months in length and from
one to four metres in depth (Moura
˜o et al. 2013; Cunha et al.
2014). The spatiotemporal heterogeneity of the wetland and
the influence of different surrounding vegetation types—
mainly ‘‘Cerrado’’, Chaco, Amazon and Atlantic Forest—
support a distinctive and rich biodiversity in the ‘‘Pantanal’’
(Pott and Pott 1994; Alho et al. 2011; Moura
˜o et al. 2013).
Cecropia pachystachya is a dioecious tree widely dis-
tributed from southern Amazon to northern Argentina
(Berg et al. 2005), and an important invader in Southeast
Asia (Raphael et al. 2015). It is the only Cecropia in the
‘‘Pantanal’’ floodplain out of five congeners reported in the
surrounding uplands (Pott and Pott 1994). Flowering and
fruiting are year round, but most trees fruit during the wet
rather than in the dry season. Seeds are endozoochorous
dispersed by vertebrates, markedly bats at our study site
(Gonc¸alves et al. 2007; Teixeira et al. 2009). Seeds gen-
erally have high longevity and commonly form seed banks
in forests (Dalling et al. 1998). Cecropia pachystachya
tolerates flooding, depending on plant maturity (height) and
the duration of waterlogging (Damasceno Ju
´nior et al.
2004; Faxina et al. 2015). Seeds that are submerged up to
30–45 days germinate faster, but germination ability
decreases after longer submersion (C. Corre
ˆa and E. Fis-
cher, unpublished data). Individuals growing under flooded
condition exhibit lower photosynthetic rates, slower growth
and an increased number of stilt roots and lenticels than
those in unflooded condition (Batista et al. 2008).
Attalea phalerata (Arecaceae) is the commonest of nine
palm trees in the ‘‘Pantanal’’ (Pott and Pott 1994). Indi-
viduals occur sparsely inside forests and often densely
along forest borders, which are subject to longer and deeper
flood durations than forest interiors. Leaves have smooth
petioles with broad and invaginate sheaths joined at the
stem apex. Sheaths of dead leaves are persistent and cover
the entire stem, providing numerous microsites for the
deposition of sediment and organic material. Attalea
phalerata stems possess 70–500 persistent sheaths
depending on stem length (0.8–4.0 m) (Corre
ˆa et al. 2012).
It is the main host plant used by obligate hemiepiphyte figs
in the ‘‘Pantanal’’.
Data collection and analyses
To assess the proportion of C. pachystachya individuals
established on palm stems, we sampled all adult trees and
all palms in 1.9 ha of forests. Trees taller than 4 m were
C. E. Corre
ˆa, E. Fischer
123
considered adults, as this is approximately the minimal
height of flowering C. pachystachya. In addition, palm
individuals were inspected for the presence of seedlings or
juveniles of C. pachystachya. We also measured the height
(to ±0.1 m) and trunk diameter (to ±0.1 cm) at 1.3 m
height of all hemiepiphytic individuals and their nearest
freestanding conspecific. To evaluate the seed bank of C.
pachystachya in persistent sheaths of A. phalerata, and the
vertical distribution of seeds deposited along palm stems,
we extracted data from Corre
ˆa et al. (2012). They consisted
of the number of C. pachystachya seeds sampled in five
persistent sheaths below 1.5 m plus five sheaths above
1.5 m from 64 stems at our study site.
To evaluate the dispersal of C. pachystachya seeds (seed
rain) to palm stems and to the forest ground, we used seed
traps consisting of half-cylinders of PVC with 20 cm depth
and semicircular openings of 10 cm radius (314 cm
2
). The
shape and dimension of seed traps allowed their physical
attachment to palm stems and thus the simulation of an
enlarged persistent sheath (Fig. 1 in Supplementary mate-
rial). We perforated the traps at the bottom and internally
lined them with tissue to retain seeds and allow passage of
rainwater. At the beginning of rainy season (October), we
set up 24 blocks of four treatments among 24 forest patches
at least 150 m apart from each other. In each forest patch,
we haphazardly selected three individuals of A.phalerata
(C3 m tall) at least 50 m apart. We then attached one trap
to a high stem ([1.5 m) of the first palm (Fig. 1 in Sup-
plementary material) followed by one trap to a low stem
(*0.5 m) of the second palm (Fig. 2 in Supplementary
material), and a third trap was partly buried in forest
ground below the canopy of the third palm (Fig. 3 in
Supplementary material). In addition, in each forest patch,
a fourth seed trap was buried 2–5 m away from palm
canopies, at least 50 m distant from other traps. The
openings of ground traps were 10 cm above the surface of
the soil to avoid trapping adjacent material. We installed 96
seed traps in total, four per forest patch, but peccaries and
cattle removed eight ground traps before samples could be
collected. Thus, the final number of traps was 88, of which
24 fixed in high stems, 24 in low stems, 20 on ground
below palm canopies and 20 on ground apart from palms.
We collected all trapped material three times, in Novem-
ber, January and March, in order to avoid decreasing of
trap efficiency due to initial accumulation of material and
to cover the wet season when most fruiting trees occur. The
collected material was initially stored in paper bags and
then inspected in laboratory for counting of C.pachys-
tachya seeds.
For two different purposes, investigating available
nutrients in substrates and conducting experiments of seed
germination, we collected eight samples of 500 g of forest
soil (5 cm depth) and eight samples of 500 g of sheath
sediment, in eight forest patches. Samples of sediment
were from one palm per forest patch and samples of soil
were collected 50 cm away from the palms. For the ger-
mination experiment, we subsampled 50 g out of each
500 g sample and mixed the subsamples of the same type,
giving 400 g of each substrate assembled from eight forest
patches. We then separated 200 g of soil and 200 g of
sediment and washed them in slow-running clean water for
2 min in order to reduce the possibility of potential inhi-
bitors of germination. The washed substrates were then
placed in trays for atmospheric drying for two days, as well
as the remaining non-washed part of both substrates. In
each tray containing a different treatment—washed soil,
washed sediment, non-washed soil and non-washed sedi-
ment—we spread 200 intact seeds of C. pachystachya
randomly selected among seeds trapped in the field and
inspected trays daily to record germinated seeds during
40 days on a glasshouse bench. All treatments were equally
watered with substrates kept near field capacity. The
remaining soil and sediment samples from the eight forest
patches were individually sent for chemical–physical
analyses carried out by Age
ˆncia Estadual de Defesa Sani-
ta
´ria Animal e Vegetal (IAGRO) (cf. Anonymous 1997).
Measured substrate traits included texture, proportion of
organic matter, pH, exchangeable quantities of Al and
quantities of essential macronutrients (P, K, Ca, Mg) and
micronutrients (Fe, Mn, Cu, Zn).
To test for differences in mean height and diameter
between hemiepiphytic and ground-established trees of C.
pachystachya, we used the student’s ttest. To test for
differences in the frequency of C. pachystachya seeds in
palm seed banks between high ([1.5 m) and low (\1.5 m)
stem height classes, we used the Chi-squared test of equal
frequencies. The same test was used for the evaluation of
differences of seed frequency among seed trap treatments
in the field and for differences among germination treat-
ments. To evaluate the similarity of chemical composition
between substrates, we used principal component analysis
(PCA) based on standardized values of pH, Al and eight
nutrients.
Results
Hemiepiphytic growth
In 1.9 ha of semideciduous forest, we recorded seven
hemiepiphytes (3.7 ind ha
-1
) and 134 freestanding
(70.5 ind ha
-1
) trees of C. pachystachya, plus 153 indi-
viduals of A. phalerata (80.5 ind ha
-1
) free of hemiepi-
phytes. We found no seedlings or juveniles of C.
pachystachya growing on palm stems and did not survey
them among ground-established plants. At least four
Bizarre Cecropia pachystachya (Urticaceae) hemiepiphytic growth on palms in the ‘‘Pantanal’’…
123
hemiepiphytes of C. pachystachya were females as they
presented infructescences. Trunk diameter was larger
(P\0.001, t=7.86, df =6) for hemiepiphytes than
ground-established conspecifics (38.9 ±9.43 cm, n=7,
and 22.8 ±6.48 cm, n=7, respectively), but height did
not differ (P=0.69, t=-0.42, df =6) between them
(8.8 ±1.11 m, n=7, and 8.9 ±0.84 m, n=7, respec-
tively). All individuals of C. pachystachya found as
hemiepiphytes started growth on low (\1.5 m) stems of A.
phalerata (Figs. 1–3). Several tiny roots of these individ-
uals developed horizontally around palm stems within the
remaining sheaths, in contrast to stilt roots that grow
obliquely downwards in freestanding C. pachystachya
trees. In addition to tiny roots around stems, other root
branches were geotropically oriented, coalesced and pro-
vided a strong functional link to the ground at the opposite
side that the plants were fixed to the palms (Figs. 1–3). All
aerial roots of C. pachystachya hemiepiphytes departed
from their low trunks that, in turn, were extensively free
upward and usually inclined in relation to the A. phalerata
stems; canopies of all C. pachystachya hemiepiphytes
surpassed those of their phorophytes in their height.
Seed bank and seed rain
Most A. phalerata stems (84%; 54/64) contained seeds of
C. pachystachya deposited on at least one out of 10 per-
sistent sheaths. In total, 1594 seeds were recovered from all
stems and these were distributed among 480 (75%) of the
640 persistent sheaths sampled. In spite of such a wide
occurrence of seeds among stems and among persistent
sheaths, few palm individuals concentrated seeds mas-
sively (Fig. 2). Among the 64 palm stems, the frequency of
C. pachystachya seeds did not differ (P=0.10, Yates’
corrected Chi squared =2.71) between high (n=671)
and low (n=923) stems, but extremely high seed abun-
dances occurred only on low stems (Fig. 4).
During seed trapping in the field, we recorded 3354
seeds of C. pachystachya among 88 seed traps, a general
frequency of more than 38 seeds per trap. Seven traps
contained 3122 seeds and the remaining traps had only 232
seeds, a pattern similar to the distribution of C. pachys-
tachya seeds among palm stems. Differential dispersal of
seeds among traps was consistent throughout the three
sampling dates, as those traps with more seeds at the first
Figs. 1–3. 1 Hemiepiphyte
Cecropia pachystachya tree on
(*7 m height) Attalea
phalerata palm (*4 m height)
in the ‘‘Pantanal’’ wetland,
Brazil. 2,3Aerial roots of the
tree involved the low stem,
coalesced and fixed to the
ground at the opposite side
C. E. Corre
ˆa, E. Fischer
123
inspection continued trapping increased number of seeds in
the second and third inspections. The frequency of trapped
seeds differed among treatments (Fig. 5) and was 15 times
higher in traps attached to low than to high A. phalerata
stems (97.5 and 6.1 seeds per trap, respectively), and seven
times higher in ground traps below palm canopies than in
ground traps away from palm plants (37.9 and 5.4 seeds per
trap, respectively).
Seed germination and substrate compositions
During 40 days of the glasshouse experiment, 420 out of
800 (52.5%) C. pachystachya seeds germinated. The fre-
quency of germination was higher (P\0.0001, v
2
tests)
on previously washed substrates (90% on soil and 70% on
stem sediment) than on non-washed substrates (25% on soil
and 26% on stem sediment), but the frequency of germi-
nation did not differ between soil and stem sediment
(Fig. 6). Germination started earlier on washed than non-
washed substrates with the frequency after 13 days on
washed substrates being similar to the frequency after
30 days on non-washed substrates (Fig. 6).
Chemical–physical traits differed between forest soil
and sediment from A. phalerata stems mainly due to dif-
ferences of pH, Zn, Cu and Fe, as evidenced through
loadings on PCA Component 1 (Fig. 7; Table 1). Forest
soil was sandy and presented neutral pH and less organic
matter than the clayey and moderately acid palm stem
sediment (Table 1). Compared to forest soil, stem sediment
showed one order of magnitude more Zn, almost five times
more K and two times more P, Ca, Mg and Al. The
amounts of Mn were approximately similar between the
substrates. Only Fe and Cu were proportionally more
abundant in forest soil, with the former being five times
higher in soil than in stem sediment. In addition to these
major differences between substrates, there was consider-
able variation among forest patches (samples) for the same
substrate, mainly due to differences of pH, Al and Mn
(Fig. 7; Table 1).
Discussion
Our results show that C. pachystachya can grow as
hemiepiphytes in the ‘‘Pantanal’’ region of Brazil, a phe-
nomenon unreported for Cecropia but common for Cous-
sapoa Aubl. in the tribe Cecropieae (Berg et al. 1990).
Hemiepiphytes of C. pachystachya, however, were rare as
they accounted for only 5% of adult trees even though A.
phalerata phorophytes were common at the sites we sam-
pled. Furthermore, hemiepiphytic growth of C. pachys-
tachya is unusual because only the base of each tree is fixed
to palm stems. This contrasts with obligate hemiepiphytes
that firmly attach through numerous roots to their phoro-
phytes (Lo
´pez-Acosta and Dirzo 2015). Aerial roots of C.
pachystachya do not depart from branches or along trunk
extension as occurs in true hemiepiphytes (Harrison 2006;
Hao et al. 2011; Jim 2014). Therefore, the presence of
aerial roots in C. pachystachya normally adapted to
Fig. 4 Distributions of Cecropia pachystachya seed abundances on
high (a) and low (b) stems of Attalea phalerata palms in the
‘‘Pantanal’’ wetland, Brazil
Fig. 5 Frequency of Cecropia pachystachya seeds trapped in four
microhabitats: high on Attalea phalerata palm stems (high stem), low
on palm stems (low stem), on the ground below palm canopies (below
canopy) and on the ground out of palm canopy projections (outside
canopy), in the ‘‘Pantanal’’ wetland, Brazil. Different letters above
bars indicate significant (P\0.01) difference of seed arrival
frequencies (v
2
tests)
Bizarre Cecropia pachystachya (Urticaceae) hemiepiphytic growth on palms in the ‘‘Pantanal’’…
123
temporary flooding (Sposito and Santos 2001; Parolin et al.
2010) served the function of attaching the trees to palms.
The numerous coalescent roots in most hemiepiphytes
differ from typical freestanding C. pachystachya. If such
variation has a genetic basis, then only part of the popu-
lation may be able to grow on phorophytes.
Gravitational destabilization may also contribute to the
rarity of hemiepiphytes and we noted two unsuccessful
cases in our study region. One C. pachystachya tree had
fallen along with the host palm because the palm stem
broke at its base, and another had fallen alone and exhib-
ited roots partially detached from the palm. Such instability
can intensify if the height of attachment of C. pachystachya
on the palm stem increases, and thus it may constrain the
hemiepiphytes to lower palm stems. Individuals fixed to
low stems can root into the ground early and develop
strong coalesced roots that provide gravitational compen-
sation force at the opposite side of attachment to palm stem
(Soethe et al. 2006). In spite of the rarity and unusual
growth of C. pachystachya hemiepiphytes, they exhibited
heights similar to and trunk diameter larger than free-
standing conspecifics, indicating that hemiepiphytes are as
vigorous as or even more vigorous than ground-established
trees. Cecropia pachystachya decreases photosynthetic
rates during the floods (Batista et al. 2008) and hemiepi-
phytism reduces the time that trees are exposed to inun-
dation, which could favour more vigour.
Cecropia pachystachya forms a remarkable seed bank
on the stems of A. phalerata in the ‘‘Pantanal’’, as
demonstrated by the presence of seeds in more than 80% of
the surveyed stems. Indeed, C. pachystachya is the most
abundant of the 75 species that occur as seeds on A.
phalerata stems at the study site, including hemiepiphyte
figs (Corre
ˆa et al. 2012). Thus, the widespread occurrence
and heavy accumulation of C. pachystachya seeds create
plenty of opportunities for seedlings to establish on A.
phalerata in the ‘‘Pantanal’’. As the frequency of C.
pachystachya seeds in palm seed banks did not differ sig-
nificantly between high and low palm stems, the height of
seeds does not seem to prevent the occurrence of individ-
uals fixed at high stems, as strangler figs do (Harrison
2006; Jim 2014). However, the frequency of C. pachys-
tachya seed dispersal to traps was nearly two orders of
magnitude higher in low than high stems. This indicates
that the primary dispersers deposit seeds to a much greater
extent to low than high stems. Since seed frequency in low
and high stems differed in seed rain but not in their seed
bank, secondary dispersers are likely carrying seeds
upward along stems or removing seeds predominantly from
low stem to elsewhere. Ants may carry seeds upward or
away from palms, and flooding pulses may remove seeds
from low stems (Passos and Oliveira 2002; Corre
ˆa et al.
2012). On the other hand, the pattern of distribution of C.
pachystachya seeds among palm individuals was similar
between the seed rain and seed bank. Most seeds were
concentrated in a few palm individuals, while many palms
had a small number of seeds. This pattern may contribute
to the rarity of C. pachystachya hemiepiphytes if the
chance of growing on a palm depends largely on whether
massive amounts of seeds are present. Furthermore, if
germination proves to be higher among newly than old
Fig. 6 Accumulated number of Cecropia pachystachya seeds that
germinated in greenhouse during 30 days of tests on two substrates
(forest soil or sediment accumulated on sheaths along Attalea
phalerata stems), each one in two situations (natural or washed in
slow-running water prior to the tests). Significant (P\0.0001)
differences of germination frequencies occurred only between washed
and natural treatments, irrespective of whether on soil or stem
sediment (v
2
tests)
Fig. 7 Principal component analysis and 95% confidence ellipses for
the chemical composition (pH, P, K, Ca, Mg, Fe, Mn, Cu, Zn, Al) of
samples from forest soil (circles) and accumulated sediment on
Attalea phalerata stems (squares) in the ‘‘Pantanal’’ wetland, Brazil.
Components 1 and 2 explain 62% of variance (45 and 17%,
respectively)
C. E. Corre
ˆa, E. Fischer
123
dispersed seeds, the continuous arrival of seeds toward low
stems would favour hemiepiphytes fixed at low rather than
high stems.
Dispersal of C. pachystachya seeds is directed to A.
phalerata palms in forest patches of the ‘‘Pantanal’’, as
demonstrated by the much higher seed trapping on stems or
below palm canopies than elsewhere on the forest floor.
The common fruit bats of ‘‘Pantanal’’, Artibeus planirostris
(Spix 1823) and Platyrrhinus lineatus (E. Geoffroy 1810),
most likely cause this pattern. Both species use C.
pachystachya fruits as a core food item (Teixeira et al.
2009; Munin et al. 2012) and feed on fruits while hanging
underneath A. phalerata canopies nearby fruit sources in
the ‘‘Pantanal’’ (Marinho Filho 1992). The palms are
abundantly distributed and provide suitable roosts with
dense canopies that are freely accessible to bats. These bats
get and leave feeding roosts many times each night and
perform U-shaped flights when entering or leaving roosts.
As they defecate in flight and during fruit handling as well
(Munin et al. 2011), stools are released in variable direc-
tions under palm canopies and consistently to palm stems.
Although bat individuals use different feeding roosts
through time, they show a high fidelity to a few canopies
(Marques and Fischer 2009). This behaviour probably
concentrates seeds of C. pachystachya to a small number of
the total available palms, consistent with the patterns we
recorded for seed banks and seed trapping among palm
individuals at our study site.
Our glasshouse experiment indicated that the germina-
tion of C. pachystachya seeds does not differ between the
substrate types, forest soil or stem sediments. However,
both substrates likely accumulate inhibitors of germination,
as indicated by the higher germination success of C.
pachystachya in previously washed than non-washed sub-
strates. If flood pulses in the ‘‘Pantanal’’ reduce inhibitors,
the germination of C. pachystachya seeds would be
stimulated in low stem parts freshly emerged after the
flooding peak, whereas high stem parts not subjected to
inundation would accumulate inhibitors. In addition, if
stem inundation increases germination, only part of palm
individuals would propitiate this condition at a given time
since A. phalerata occupies different topographic levels
and the flooding pulses are quite variable in the
‘‘Pantanal’’.
The clay sediment on A. phalerata stems is richer in
nutrients than the sandy soil of forest patches of the
‘‘Pantanal’’, like other palms reported with rich substrates
for true hemiepiphytes and C. obtusifolia (Bernal and
Balslev 1996;Lo
´pez-Acosta and Dirzo 2015). Our results
demonstrate that the sediments accumulated on persistent
sheaths have considerably higher amounts of organic
matter and of six nutrients (P, K, Ca, Mg, Mn and Zn) than
forest soils, where only one micronutrient (Fe) was sig-
nificantly more available. Therefore, palm sediments can
offer nutritional benefits that may contribute to the vigor-
ous growth observed for C. pachystachya hemiepiphytes in
the ‘‘Pantanal’’. In addition, our observations of fruiting
hemiepiphytes support the hypothesis that A. phalerata
stems provide high-quality microhabitat suitable for female
reproductive function (Fischer and Santos 2001; Guillon
et al. 2006). The values of aluminium concentration in both
substrates were apparently low to cause negative effects on
plant growth (Steiner et al. 2012). Moreover, the acid pH of
stem sediment can prevent the accumulation of aluminium
in growing plants, although it may reduce seed longevity
compared to the proximate neutral pH of forest soil (Jansen
et al. 2002; Pakeman et al. 2012). Therefore, the mainte-
nance of viable seed banks in palm stems might depend
upon a continual arrival of seeds compared to soil seed
banks in forest patches of ‘‘Pantanal’’.
In conclusion, a variety of ecological factors may con-
tribute to the facultative occurrence of C. pachystachya
Table 1 Chemical composition
of forest soil and sediment
accumulated on stems of Attalea
phalerata palms, in the
‘‘Pantanal’’ wetland of Brazil,
and loadings of elements on
Components 1 and 2 of a
principal component analysis
(PCA) to evaluate the similarity
between substrates (see Fig. 7)
Elements Stem sediment (n=8) Forest soil (n=8) PCA loadings
Mean SD Range Mean SD Range Comp. 1 Comp. 2
pH 5.4 0.36 4.8-5.9 6.8 0.74 5.6–8.1 -0.31 0.49
Organic matter (%) [5 – – 3.2 0.67 2.3–4.1 – –
Al (meq 100 cm
-3
) 0.6 0.12 0.4–0.8 0.4 0.12 0.2–0.6 0.27 -0.48
P(lgcm
-3
) 42.3 11.85 30.2–61.7 22.8 19.62 9.3–68.8 0.33 0.23
K (meq 100 cm
-3
) 1.4 0.59 0.6–2.6 0.3 0.08 0.2–0.4 0.33 -0.30
Ca (meq 100 cm
-3
) 10.7 3.03 7.9–17.5 6.8 2.75 4.1–11.8 0.33 0.28
Mg (meq 100 cm
-3
) 5.8 4.41 1.3–12.1 3.8 5.14 0.4–13.2 0.16 0.12
Fe (mg dm
-3
) 7.7 5.02 2.2–17.4 39.2 29.54 1.8–87.0 -0.37 -0.24
Mn (mg dm
-3
) 70.4 31.88 27–118.8 55.0 26.02 27.5–107.0 0.18 0.40
Cu (mg dm
-3
) 0.3 0.04 0.3–0.4 0.4 0.06 0.3–0.5 -0.36 -0.24
Zn (mg dm
-3
) 29.2 9.79 11.5–40.5 2.5 0.76 1.1–3.6 0.43 -0.14
Bizarre Cecropia pachystachya (Urticaceae) hemiepiphytic growth on palms in the ‘‘Pantanal’’…
123
hemiepiphytes in the ‘‘Pantanal’’. Of particular importance
is the consumption of C. pachystachya fruits by dense
populations of fruit bats (Teixeira et al. 2009; Munin et al.
2012), whose behaviour provides continuous seed dispersal
to A. phalerata palms. Nutrient availability and a shorter
duration under flooded conditions would aid fast and vig-
orous growth of C. pachystachya on palms. However, the
attachment of hemiepiphytes is apparently limited to low
palm stems, where gravity stabilization of trees appears to
be feasible. In addition, seed arrival is very high and
inhibitors of germination may be somewhat reduced by
flooding pulses in low parts of the palm stems. However,
only a small proportion of the palms receives C. pachys-
tachya seeds massively and the potential benefits after
inundation are rather ephemeral. These factors, summed to
potentially intrinsic limitation of most C. pachystachya
individuals in producing sufficient aerial roots to efficiently
fix on palms and then reach the forest floor, could explain
the rarity of hemiepiphytic trees at the study site.
Acknowledgements We thank J Celso, LC Corre
ˆa, LF Carvalho, N
Cunha, N Penatti and S Ferreira for field assistance in the ‘‘Pantanal’’,
F Gonc¸alves, LC Corre
ˆa and S Ferreira for assistance during the lab
work, and SCH Barrett for comments on early draft. This work was
supported by Fundac¸a
˜o de Apoio ao Desenvolvimento do Ensino,
Cie
ˆncia e Tecnologia do Estado de Mato Grosso do Sul (23/200.009/
2008), Coordenac¸a
˜o de Aperfeic¸oamento de Pessoal de Nı
´vel Supe-
rior (130180/2003-3) and Conselho Nacional de Desenvolvimento
Cientı
´fico e Tecnolo
´gico (311001/2012-2).
References
Alho CJR, Camargo G, Fischer E (2011) Terrestrial and aquatic
mammals of the Pantanal. Braz J Biol 71:297–310
Almeida TIR, Penatti NC, Ferreira LG, Arantes AE, Amaral CH
(2015) Principal component analysis applied to a time series of
MODIS images: the spatio-temporal variability of the Pantanal
wetland, Brazil. Wetl Ecol Manag 23:737–748
Anonymous (1997) Manual de me
´todos de ana
´lise de solo. Empresa
Brasileira de Pesquisa Agropecua
´ria. Centro Nacional de
Pesquisa de Solos, Rio de Janeiro
Barlow PW (1986) Adventitious roots of whole plants: their forms,
functions, and evolution. In: Jackson MB (ed) New root
formation in plants and cuttings. Martinus Nijhoff Publishers,
Dordrecht, pp 67–110
Batista CUN, Medri ME, Bianchini E, Medri C, Pimenta JA (2008)
Tolera
ˆncia a
`inundac¸a
˜odeCecropia pachystachya (Cecropiaceae):
aspectos ecofisiolo
´gicos e morfoanato
ˆmicos. Acta Bot Bras
22:91–98
Benzing DH (1990) Vascular epiphytes. Cambridge University Press,
Cambridge
Berg CC, Akkermans RWAP, van Heusden ECH (1990) Cecropiaceae:
Coussapoa and Pourouma, with an introduction to the family.
Flora Neotrop Monogr 51:1–208
Berg CC, Rosselli PF, Davidson DW (2005) Cecropia. Flora Neotrop
94:1–230
Bernal R, Balslev H (1996) Strangulation of the palm Phytelephas
seemannii by the pioneer tree Cecropia obtusifolia: the cost of
efficient litter trapping. Ecotropica 2:177–184
Corre
ˆa CE, Fischer E, Santos FAM (2012) Seed banks on Attalea
phalerata (Arecaceae) stems in the Pantanal wetland, Brazil.
Ann Bot 109:729–734
Cunha NL, Fischer E, Lorenz-Lemke AP, Barrett SCH (2014) Floral
variation and environmental heterogeneity in a tristylous clonal
aquatic of the Pantanal wetlands of Brazil. Ann Bot
114:1637–1649
Dalling JW, Swaine MD, Garwood NC (1998) Dispersal patterns and
seed bank dynamics of pioneer trees in moist tropical forest.
Ecology 79:564–578
Damasceno Ju
´nior GA, Semir J, Santos FAM, Leita
˜o Filho HF (2004)
Tree mortality in a riparian forest at Rio Paraguai, Pantanal,
Brazil, after an extreme flooding. Acta Bot Bras 18:839–846
Faxina C, Fischer E, Pott A (2015) Flora of inland Atlantic riparian
forests in southwestern Brazil. Biota Neotrop 15:1–12. doi:10.
1590/1676-06032015008313
Fischer E, Santos FAM (2001) Demography, phenology and sex of
Calophyllum brasiliense (Clusiaceae) trees in the Atlantic forest.
J Trop Ecol 17:903–909
Gentry AH (1993) A field guide to the families and genera of woody
plants of northwest South America (Colombia, Ecuador, Peru)
with supplementary notes on herbaceous taxa. The University of
Chicago, Chicago
Gonc¸alves F, Munin R, Costa P, Fischer E (2007) Feeding habits of
Noctilio albiventris (Noctilionidae) bats in the Pantanal, Brazil.
Acta Chiropterol 9:535–538
Guillon JM, Julliard R, Leturque H (2006) Evolution of habitat-
dependent sex allocation in plants: superficially similar to, but
intrinsically different from animals. J Evol Biol 19:500–512
Hao G, Sack L, Wang A, Cao K, Goldstein G (2010) Differentiation of
leaf water flux and drought tolerance traits in hemiepiphytic and
non-hemiepiphytic Ficus tree species. Funct Ecol 24:731–740
Hao G, Goldstein G, Sack L, Holbrook NM, Liu Z, Wang A, Harrison
RD, Su Z, Cao K (2011) Ecology of hemiepiphytism in fig
species is based on evolutionary correlation of hydraulics and
carbon economy. Ecology 92:2117–2130
Harrison RD (2006) Mortality and recruitment of hemi-epiphytic figs
in the canopy of a Bornean rain forest. J Trop Ecol 22:477–480
Jansen S, Broadley MR, Robbrecht E, Smets E (2002) Aluminum
hyperaccumulation in angiosperms: a review of its phylogenetic
significance. Bot Rev 68:235–269
Jim CY (2014) Ecology and conservation of strangler figs in urban
wall habitats. Urban Ecosyst 17:405–426
Kaufmann S, McKey DB, Hossaert-McKey M, Horvitz CC (1991)
Adaptations for a two-phase seed dispersal system involving
vertebrates and ants in a hemiepiphytic fig (Ficus microcarpa:
Moraceae). Am J Bot 78:971–977
Lo
´pez-Acosta JC, Dirzo R (2015) Aspectos relevantes sobre la
historia natural de las plantas hemiepı
´fitas estranguladoras.
Interciencia 40:190–197
Marinho Filho JS (1992) Ecologia e histo
´ria natural das interac¸o
˜es entre
palmeiras, epı
´fitas e frugı
´voros na regia
˜o do Pantanal Matogros-
sense. Thesis, Universidade Estadual de Campinas, Campinas
Marques MCM, Fischer E (2009) Effect of bats on seed distribution
and germination of Calophyllum brasiliense (Clusiaceae).
Ecotropica 15:1–6
Moura
˜o GM, Calheiros DF, Oliveira MD, Padovani C, Fischer E,
Tomas W, Campos Z (2013) Respostas ecolo
´gicas de longo
prazo a variac¸o
˜es plurianuais das enchentes do Pantanal. In:
Tabarelli M, Rocha CFD, Romanowski HP, Rocha O, Lacerda
LD (eds) PELD-CNPq: dez anos do Programa de Pesquisas
Ecolo
´gicas de Longa Durac¸a
˜o do Brasil: achados, lic¸o
˜es e
perspectivas. Editora Universita
´ria da UFPE, Recife, pp 90–116
Munin RL, Costa PC, Fischer E (2011) Differential ingestion of fig
seeds by a Neotropical bat, Platyrrhinus lineatus. Mamm Biol
76:772–774
C. E. Corre
ˆa, E. Fischer
123
Munin RL, Fischer E, Gonc¸alves F (2012) Food habits and dietary
overlap in a phyllostomid bat assemblage in the Pantanal of
Brazil. Acta Chiropterol 14:195–204
Nadkarni NM, Haber WA (2009) Canopy seed banks as time capsules
of biodiversity in pasture-remnant tree crowns. Conserv Biol
23:1117–1126
Nadkarni NM, Merwin MC, Nieder J (2001) Forest canopies, plant
diversity. In: Levin SA (ed) Encyclopaedia of biodiversity 3.
Academic Press, London, pp 27–40
Pakeman RJ, Small JL, Torvell L (2012) Edaphic factors influence the
longevity of seeds in the soil. Plant Ecol 21:57–65
Parolin P (2002) Life history and environment of Cecropia latiloba in
Amazonian floodplains. Rev Biol Trop 50:531–545
Parolin P, Lucas C, Piedade MTF, Wittmann F (2010) Drought
responses of flood-tolerant trees in Amazonian floodplains. Ann
Bot 105:129–139
Passos L, Oliveira PS (2002) Ants affect the distribution and
performance of Clusia criuva seedlings, a primarily bird-
dispersed rainforest tree. J Ecol 90:517–528
Peel MC, Finlayson BL, McMahon TA (2007) Updated world map of
the Koppen-Geiger climate classification. Hydrol Earth Syst Sci
11:1633–1644
Pott A, Pott VJ (1994) Plantas do Pantanal. Empresa Brasileira de
Pesquisa Agropecua
´ria, Brası
´lia
Raphael MB, Chong KY, Yap VB, Tan HTW (2015) Comparing
germination success and seedling traits between exotic and
native pioneers: Cecropia pachystachya versus Macaranga
gigantea. Plant Ecol 216:1019–1027
Soethe N, Lehmann J, Engels C (2006) Root morphology and
anchorage of six native tree species from a tropical montane
forest and an elfin forest in Ecuador. Plant Soil 279:173–185
Sposito TC, Santos FAM (2001) Architectural patterns of eight
Cecropia (Cecropiaceae) species of Brazil. Flora 196:215–226
Steiner F, Zoz T, Pinto Ju
´nior AS, Castagnara DD, Dranski JAL
(2012) Effects of aluminum on plant growth and nutrient uptake
in young physic nut plants. Cienc Sem Agrar 33:1779–1788
Teixeira RC, Corre
ˆa CE, Fischer E (2009) Frugivory by Artibeus
jamaicensis (Phyllostomidae) bats in the Pantanal, Brazil. Stud
Neotrop Fauna Environ 44:7–15
Wanek W, Arndt SK, Huber W, Popp M (2002) Nitrogen nutrition
during ontogeny of hemiepiphytic Clusia species. Funct Plant
Biol 29:733–740
Zotz G (2013) ‘Hemiepiphyte’: a confusing term and its history. Ann
Bot 111:1015–1020
Bizarre Cecropia pachystachya (Urticaceae) hemiepiphytic growth on palms in the ‘‘Pantanal’’…
123
