UvA-DARE (Digital Academic Repository)
Vascular epiphytes in Taiwan and their potential response to climate change
Hsu, C.C.
Link to publication
Citation for published version (APA):
Hsu, R. C. C. (2013). Vascular epiphytes in Taiwan and their potential response to climate change
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),
other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating
your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask
the Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,
The Netherlands. You will be contacted as soon as possible.
UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)
Download date: 16 Jun 2017
Vascular epiphytes in Taiwan and their
potential response to climate change
Rebecca C.-C. Hsu
2013
UNIVERSITEIT VAN AMSTERDAM
Cover: The epiphytic fern Polypodium formosanum growing on a blue glass ball (representing the Earth) to express that species' response to climate change is an entangled question.
Back cover: A common scenery of montane cloud forests in the afternoon on Taiwan.
Vascular epiphytes in Taiwan and their
potential response to climate change
Hsu,R.C.‐C.2013.VascularepiphytesinTaiwanandtheirpotentialresponseto
climatechange.
PhDthesis,UniversityofAmsterdam,TheNetherlands
Coverlayout:RebeccaC.‐C.Hsu
Coverillustration:RebeccaC.‐C.Hsu
ISBN:978‐94‐91407‐12‐3
Vascular epiphytes in Taiwan and their
potential response to climate change
ACADEMISCHPROEFSCHRIFT
terverkrijgingvandegraadvandoctor
aandeUniversiteitvanAmsterdam
opgezagvandeRectorMagnificus
prof.dr.D.C.vandenBoom
tenoverstaanvaneendoorhetcollegevoorpromotiesingesteldecommissie,
inhetopenbaarteverdedigenindeAgnietenkapel
opdinsdag10september2013,te10.00uur
door
RebeccaChia‐ChunHsu
geborenteHsinChu,Taiwan
Promotiecommissie
Promotores: Copromotores:
Overigeleden: Prof.dr.G.R.deSnoo
Prof.dr.J.H.D.Wolf
Dr.J.G.B.Oostermeijer
Dr.W.L.M.Tamis
Dr.J.F.Duivenvoorden
Prof.dr.H.Hooghiemstra
Prof.dr.S.B.J.Menken
Dr.N.Raes
Prof.dr.P.H.vanTienderen
FaculteitderNatuurwetenschappen,WiskundeenInformatica
CONTENTS
1
General Introduction
6
2
Diversity and phytogeography of vascular epiphytes in a tropical‐subtropical
transition island, Taiwan
(Flora 204(8), 2009, pp. 612‐627)
16
3
Regional and altitudinal patterns in vascular epiphyte richness on an East Asian
island
30
4
Canopy CO2 concentrations and crassulacean acid metabolism in Hoya carnosa
in a subtropical rain forest in Taiwan: consideration of CO2 availability and the
evolution of CAM in epiphytes
(Photosynthetica 44(1), 2006, pp. 130‐135)
48
5
Comparative photosynthetic capacity of abaxial and adaxial leaf sides as related
to exposure in an epiphytic fern in a subtropical rainforest in northeastern
Taiwan
(American Fern Journal 99(3), 2009, pp. 145-154)
60
6
Adaptation of a widespread epiphytic fern to simulated climate‐change
conditions
70
7
Simulating climate change impacts on forests and associated vascular epiphytes
in a subtropical island of East Asia
(Diversity and Distributions 18(4), 2012, pp. 334-347)
86
8
Conclusions
108
SUMMARY
115
13 9
149
159
總結
163
SAMENVATTING
169
175
REFERENCES
APPENDIX 1
APPENDIX 2
ACKNOWLEDGEMENTS
Chapter 1
General Introduction
The medium (category‐2) typhoon Morakot
(2009) brought 2777 mm rainfall in 72 h, causing catastrophic damage. Numerous uprooted trees and associated epiphytes were brought by floods to the coast. CHAPTER 1
General Introduction
The distribution of vascular epiphytes
Epiphytic plants are a characteristic component of the tropical wet forest (Benzing, 1990). In
this thesis, epiphytes are defined, following Barkman (1958), as organisms that grow on plants
without extracting water or nutrients from hosts’ living tissues. Unlike parasitic plants,
epiphytic plants are autotrophic depending only on their hosts for anchorage whilst obtaining
essential resources by intercepting dry and wet depositions (e.g. dust, litter, rainfall and fog). It
is not rare to find so‐called accidental epiphytes growing on other plants; however, those are
mostly unable to reproduce in the canopy (Moffett, 2000).
The focus in this thesis is on vascular epiphytic plants, whereas many other organisms
such as bryophytes and lichens can also be found growing as epiphytes in the forest canopy.
Based on their life history, vascular epiphytes can be classified into true epiphytes
(holo‐epiphytes) and hemi‐epiphytes (Schimper, 1888). The former complete their entire life
cycle without contacting the forest floor, whilst individuals of the latter spend part of their life
cycle as terrestrial plants. The epiphytic life‐form is a successful adaptation of plants to the
forest canopy, comprising ca. 29,000 species, or approximately 10% of all vascular plants, in 83
different families and 876 genera (Gentry and Dodson, 1987a). Epiphytes are unevenly
distributed over taxonomic groups and geographic locations. The epiphyte flora is
concentrated in monocotyledons, especially orchids, bromeliads and aroids, and in ferns and
fern‐allies (Benzing, 1990). With few exceptions, epiphytic vascular plants are mainly found in
the tropical region (< 23.5° latitude). In contrast to the Neotropics, paleotropical areas lack
several species‐rich epiphytic families (e.g. Bromeliaceae, Cactaceae and Marcgraviaceae) and
have received less attention from botanists. Especially from Asia, epiphyte inventories are still
rare (Wolf and Flamenco‐S, 2003).
Whereas epiphyte richness generally decreases with latitude, numerous studies have
reported a different pattern in richness along the altitudinal gradient on mountains. Apparently
epiphytes (and many other organisms) achieve greatest diversity at mid‐elevations although the
altitudinal position of the diversity maximum may vary among geographical areas (Wolf and
Flamenco‐S, 2003; McCain, 2004; Kromer et al., 2005; Cardelus et al., 2006; Laurance et al.,
2011). Some studies (e.g. Cardelus et al., 2006) suggest that the observed hump‐shape in species
richness can best be explained by a null distribution (i.e. the mid‐domain effect, MDE). The
MDE arises from geographic constraints on species ranges within a bounded domain (e.g.
7 INTRODUCTION from coasts to mountaintops) whereby the null model predicts that overlapping species’ ranges
lead to a peak in species richness at mid‐elevation (Colwell and Lees, 2000). However, it has
been argued that the MDE has been overstated in the past and that climatic factors are closely
related to species richness patterns (Kessler et al., 2011). Moreover, island and continental
systems seem to demonstrate different elevational diversity patterns. For instance, whereas
MDE provided a reasonable explanation for bryophyte richness in the continental Andes,
MDE underestimated the species richness at mid‐elevation in an Indian Ocean island
(Ah‐Peng et al., 2012). The exceptional high species richness at mid‐elevation on this island
reflects the presence of a large number of species with a small range size, presumably due to
climatic compression. Another explanation for the unbalanced proportion of narrow‐ranged
species on geologically young islands postulates that nonequilibrium communities here have a
higher speciation rate owning to frequent habitat disturbance by, among others, volcanic
activity, cyclones, and landslides (Whittaker, 2000). Finally, the physiological preference of
different taxonomic groups probably accounts for their distinctively elevational richness along
the same gradient (Krömer et al., 2013; Krömer et al., 2005; Rahbek, 1995).
Epiphyte salient features
Without access to a buffering supply of water and nutrients in the forest soil, the arboreal
habitat for epiphytes is extremely dynamic in terms of moisture and nutrient availability.
Accordingly, epiphytic plants have evolved morphologically and physiologically to deal with
the typical water and nutrient‐stress conditions in the forest canopy. Bromeliads form a good
example to illustrate how epiphytic plants have adapted to their arboreal habitat. Many
bromeliads have developed a rosette growth form that serves as a reservoir for water and litter.
Uptake is facilitated by the presence of specialized structures (trichomes) on the leaf surface.
Moreover, many bromeliads possess a water‐saving metabolic pathway (Crassulacean Acid
Metabolism, CAM). To reduce evapotranspiration during the day, CAM plants acquire CO2
mostly at night, and the pre‐collected CO2 is stored as an intermediate product of malic acid in
special water storage tissue that permits this two‐step carbohydrate fixation process. CAM
presumably evolved as an adaptation to arid conditions. CAM occurs in about 4% of the
vascular flora and a majority is epiphytic plants, such as bromeliads (Martin, 1994; Winter and
Smith, 1996). Interestingly, CAM plants are also found in aquatic environments, extremely wet
forests or shaded understory in (sub‐)tropical areas (Pierce et al., 2002; Craig, 2005; Skillman
and Winter, 1997). This has led to the speculation that CAM might have evolved in response
to CO2 availability instead of drought stress (Benzing, 1990; Keeley, 1996).
Epiphytic bromeliads occur exclusively in the Neotropics. In the paleotropical area, a
similar niche is apparently occupied by the widespread Asplenium species, the ‘bird’s nest fern’.
The nickname bird’s nest is derived from its rosette growth form, which traps fallen leaves and
8 CHAPTER 1 other debris. The clumped plant bases are composed of fibrous roots and trapped humus,
which sponge up rainwater to facilitate successful establishment in the forest canopy. Adult
plants may reach 300 cm in diameter, creating conspicuous populations in the forest canopy.
The genetic differentiation in bird's nest ferns such as Asplenium antiquum at the large spatial
scale of East Asia is high (Murakami et al., 1999). However, regional variation for species such
as A. nidus may also be high and presumably leads to many cryptic species (Zhang et al., 2010).
Epiphytes in a changing environment
The steep latitudinal gradient of epiphyte distributions shows that most epiphyte species occur
in (sub‐)tropical areas. Compared with temperate species, tropical species experience limited
annual thermal variability and the thereby resulting thermal specialization might render tropical
species amongst the most imperilled species on Earth due to global warming (Laurance et al.,
2011; Cunningham and Read, 2003). Physiologically, a rising temperature may directly alter the
metabolic and evaporation rates of plants. For instance, non‐vascular epiphytes in tropical
mountains show an optimum photosynthetic rate at mean habitat temperature, suggesting an
adaptation to the local thermal regime (Wagner et al., 2013). A rising temperature might
increase the evaporation rate, causing dehydration and carbon loss. Under manipulated
warming conditions, epiphytic orchids also showed a reduction in biomass, by 30%, and a
shortening of the flower spikes (Vaz et al., 2004). Notably, the warming trend is more
pronounced at night, leading to a decrease of the diurnal temperature range (Solomon et al.,
2007). For CAM species, a small change above the nightly temperature optimum would
decrease air humidity, and in turn markedly reduce stomatal conductance and the amount of
CO2 available for the synthesis of malic acid, which may cause carbon loss (Martin, 1994).
On the other hand, atmospheric CO2 has been expected to exceed 550 ppm by the
second half of this century, a doubling of the preindustrial concentration of 280 ppm
(Solomon et al., 2007). Since the photosynthesis rate is not carbon saturated under current
atmospheric CO2 level, plants may uptake more carbon under increased ambient CO2
concentration which leads to a faster growth (Körner, 2000). However, response of plants to
elevated CO2 does not merely depend on assimilation rate, water‐use efficiency is also crucial.
The slow‐growth CAM plant demonstrates the trade‐off between water transpiration and
biomass accumulation. In addition, even desiccation tolerant plants (bryophytes and lichens)
show a long‐term reduced photosynthesis in response to elevated CO2, and the short‐term
positive reaction is relatively small and taxon‐specific (Tuba et al., 1999; Monteiro et al., 2008).
Some future climate change scenarios project an increase in annual rainfall and in the
number of dry days, which might increase the frequency and/or intensity of extreme weather
events (e.g. floods, drought; Solomon et al., 2007). Water availability is a determining factor for
9 INTRODUCTION epiphyte distribution (Gentry and Dodson, 1987a), yet rainfall seasonality rather than total
rainfall is more relevant to epiphyte survival (Zotz and Hietz, 2001). Hence, the redistribution
of seasonal rainfall under climate change might have a bigger influence on epiphytes than an
increased rainfall.
Another important source of water for epiphytes, the so‐called horizontal
precipitation (i.e. fog/cloud water), has also been projected to change due to global warming.
Rising sea surface temperatures presumably would cause a lifting of the cloud base, thereby
decreasing ground‐level clouds. A decrease in the frequency of cloud immersions may threaten
the survival of tropical montane cloud forests (TMCF, Still et al., 1999). TMCF, estimated to
represent about 14.2% of all tropical forests, are characterized by a conspicuous epiphyte
community, contributing up to 35% of the floristic diversity and up to 60% of total abundance
(Gentry and Dodson, 1987a; Wolf and Flamenco‐S, 2003; Mulligan, 2010). Under continuously
high air humidity, many TMCF epiphytes are poikilohydric (e.g. bryophytes,
Hymenophyllaceae, Grammitidaceae) that react strongly to humidity change. Epiphytes
intercept a disproportional high amount of rainfall and cloud droplets in view of their
contribution to total forest biomass, and thus epiphytes may have a relatively large influence on
the hydrology and nutrient cycles of TMCF (Hofstede et al., 1993; Hsu et al., 2002; Pypker et al.,
2006). A dieback of TMCF epiphytes would not only affect the hydrological cycle, but also
have a negative impact on associated fauna (e.g. canopy amphibians), leading a possible
cascading effect in this unique ecosystem (Benzing, 1998; Foster, 2001). Since the level of
endemism in TMCF is high, presumably due to geographical isolation and narrow climatic
conditions, the loss of TMCF species is especially detrimental (Ponce-Reyes et al., 2013; Leo,
1995).
Occupying the most climate‐defined space in the forest, epiphytes are often assumed
to be more sensitive to atmospheric change than the soil‐based flora. Non‐vascular epiphytes,
such as lichens and bryophytes, have long been used as indicators for air pollution and acid
rain (Farmer et al., 1992; Szczepaniak and Biziuk, 2003). Therefore, it is not surprising that
ecologists have been aware of the potential of epiphytes for monitoring anthropogenic climate
change (Lugo and Scatena, 1992). In response to the warming climate of the past two decades,
several (sub‐)tropical epiphytic bryophytes and lichens species have invaded Europe (Frahm,
2001; van Herk et al., 2002). However, up to now, little physiological and ecological
information is available for vascular epiphytes to predict their likely response to on‐going
global climatic change.
10 CHAPTER 1 Simulating species responses to climate change
The survival risk from climate change varies amongst biomes and, similarly, the response to
climate change is also species‐specific (Thomas et al., 2004; Loarie et al., 2009). Species
responses to changing environments include migration, genetic adaptation and tolerance
(phenotypic plasticity). If the intensity and velocity of change is beyond the ability of a species
to cope, extinction is inevitable. Therefore, the ability of dispersal and establishment may well
determine the vulnerability of a species to climate change, and so is the extent of species
genetic adaptation (Hedderson and Longton 2008). Widespread (generalist) species that occur
across a broad range of environmental gradients, comprising several climatically‐adapted
populations, may be less vulnerable since the ability of these species to occupy diverse habitats
may permit populations to persist in spatially and temporally changing environments (Silander
1985). Therefore, generalist species are likely to demonstrate broader tolerances to climate
change than specialists that are geographically restricted (Broennimann et al. 2006; Aitken et al.
2008). Accordingly, the range size of a species might be an indicator to assess species
vulnerability to climate change. Furthermore, the altitudinal range of a species in mountain
areas might be interpreted as the degree of thermal specialization (Janzen, 1967; Huey, 1978).
To evaluate species sensitivity to manipulated climate‐change conditions, in vitro experiments
or in situ reciprocal transplant field experiments may be used (Nadkarni and Solano 2002; Vaz
et al., 2004; Song et al. 2012). For epiphytes, field studies on the adaptability to simulated
conditions of climate change are still rare.
In field experiments, warming conditions may be simulated by transplanting species
to lower latitudes or elevations. These studies, however, are logistically complex and time
consuming, which is one of the reasons why species distribution models (SDMs) are
increasingly used to estimate the potential range shift of species under future climate change
conditions (Bakkenes et al., 2002; Broennimann et al., 2006; Hijmans and Graham, 2006;
Thuiller et al., 2006; Carnaval and Moritz, 2008; Fitzpatrick et al., 2008; Jensen et al., 2008).
SDMs attempt to recognize species’ realized niche by relating species occurrences with values
of predictor variables across a series of observation sites (Guisan and Thuiller, 2005). This
modelling‐based approach provides valuable first‐order assessments of potential climatic
change impacts on biodiversity (Huntley et al., 2010). Another advantage of SDMs is the
visualization of the predicted distribution of each species, cumulating in an one‐picture
overview to demonstrate regional patterns.
Notwithstanding the general acceptance of SDM’s as a valuable tool, several aspects
of purely climate‐based models have been criticized (Heikkinen et al., 2006; Austin and Van
Niel, 2011). For example, SDMs tend to overestimate the area of suitable habitats, particularly
for those species with a strong dependency on other species, such as epiphytic plants (Huntley
11 INTRODUCTION et al., 2010). Epiphyte distribution relies on the presence and characteristics of host trees.
Precisely, it is the composition of and structure of the host trees in a forest that significantly
influence epiphyte assemblages, which might be attributed to the microclimate associated with
forest types (Benavides, 2010). Numerous studies have indicated the importance of
incorporating biotic interactions for improving the accuracy of SDMs (Leathwick et al., 1996;
Araújo and Luoto, 2007; Preston et al., 2008). Another criticism to climate SDM’s is that the
velocity of climate change often outpaces the migration capacity of many species (Svenning et
al., 2008; Thuiller et al., 2008). Therefore, it is crucial to take dispersal limitation into account
when simulating species distributions (Engler and Guisan, 2009). However, because it is
difficult to obtain reliable dispersal data, most studies assume either unlimited or no dispersal
for the target species. Other debates concern species persistence under unfavourable climatic
conditions (Loehle and LeBlanc, 1996). Common sense dictates that many species, especially
long‐lived trees, will not immediately perish after climate changes and this justifies including
species persistence in SDM’s. Finally, a general shortcoming of SDM’s is the disregard for
intraspecific variation in the modelled species. Many species are projected to tolerate climate
change better when sub‐taxon information is considered (Pearman et al. 2010; Benito Garzón
et al. 2011; Oney et al. 2013).
Study Area
Taiwan (formerly known as Formosa) is a continental island, separated from Southeast China
by the ca. 200 km wide Taiwan Strait. With an area of 36,000 km2, the Tropic of Cancer
crosses through the middle of the southern half of the island (Chapter7; Fig. 1). Taiwan owes
its existence to a collision of the Philippines Sea plate with the Eurasian continental margin
some five million years ago (Ho, 1988). Subsequent active orogenesis promoted the creation of
an extensive mountain system on the island; about 70% of the area is covered by (> 1,000 m
asl) mountains. Mt. Jade (3952 m) is the highest peak within the more than 50 peaks above
3,000 m asl in Taiwan.
There is no distinct dry season in Taiwan. Unlike most areas at the tropic of Cancer
or Capricorn that are relatively arid, Taiwan has a humid climate thanks to the high mountains
that induce cloud formation from high‐humidity oceanic winds. The annual rainfall ranges
from 1,000 mm to over 6,000 mm, and generally falls during the NE monsoon (October–
January), spring rain (February–April), plum rain (May–June) and typhoon‐induced heavy rain
events (July–September). The NE monsoon accounts for 45% of the total annual rainfall,
mainly in east Taiwan (Kao et al., 2004). On average, 3.7 typhoons hit Taiwan in summer (July
to September) each year, of which about 80% land on the east coast with westbound tracks
(Wu and Kuo, 1999). The dominating central range on the island often has a complex
interaction with the typhoon circulation and could decrease the intensity of typhoons by an
12 CHAPTER 1 average of over 40% within 12h after the storm centre reaches the island (Wu and Kuo, 1999).
Therefore, the western part of the island on the lee side of the central range receives less
impact from typhoons.
Taiwan floristic diversity is high, comprising ca. 4077 species (Hsieh, 2003). The
dramatic biodiversity is driven by habitat heterogeneity and the diversity in biomes on this
island, ranging from alpine tundra to tropical rain forests. In addition, the (sub‐)tropical
location enables Taiwan to receive tropical and temperate species from adjacent regions; for
example, ca. 25% of the pteridophyte flora consists of “marginal species” (i.e. tropical or
temperate species at their northern or southern limit of the geographical distribution,
respectively, Moore, 2000). Finally, the oceanic climate facilitated Taiwan as a refuge during
Quaternary glaciations, a presumable explanation for the high number of epiphytic species in
Taiwan. Tropical islands provide an ideally natural laboratory for testing ecological and
evolutional theories, for their compressed biodiversity and simplified ecosystem within a
relatively small land area. The mountainous island of Taiwan is therefore a well‐located site for
describing distribution patterns and testing corresponding hypotheses.
Aims and contents of this thesis
The overall goal of this thesis is to get insight in the relative vulnerability of epiphytic species
and associated forest types of Taiwan to global climate change. To achieve this goal,
descriptive, experimental (laboratory and field), and modelling approaches are all applied.
Chapter 2 is a descriptive study of the epiphyte flora of Taiwan. After consultation of the
herbarium specimens and own field observations, for the first time, a checklist of the vascular
epiphytes in Taiwan and several associated islets is organised. The systematic composition of
the epiphyte flora of Taiwan and its phytogeographical connection with adjacent areas is
described and explanations are proposed.
Chapter 3 is also largely descriptive. Thanks to the wealth of information available in
herbarium collections and published literature, an epiphyte database is compiled, comprising
ca. 39,000 unique records. Based on the assembled collections, regional and altitudinal patterns
of epiphyte distribution are described in more detail. Altitudinal richness patterns are tested
with the mid‐domain effect theory. Next, a species distribution modelling (SDM) approach,
MaxEnt, is used to obtain a more complete picture for the distribution of species, and a
summarizing richness map is constructed. The relative importance of environmental factors
for epiphyte distribution is furthermore explored with ordination analysis (canonical
correspondence analysis). Special attention is paid to the influence of typhoons on epiphytes
and the altitudinal distribution of thermal specialists for the reason that thermal specialist may
be more vulnerable to global warming.
13 INTRODUCTION In chapter 4, the influence of environmental factors on epiphytes is further evaluated in both
field and laboratory experiments that focus on the uptake of atmospheric CO2 by the CAM
epiphyte Hoya carnosa. Diurnal acid accumulation and leaf carbon isotope are compared
between dense and open forests to evaluate the effect of increased CO2 availability.
Chapter 5 is another ecophysiological study. Using a common bird's nest fern, Asplenium
nidus, the photosynthetic rates of its leaf blades that receive different levels of sunlight on both
surfaces are compared to assess the photosynthetic plasticity of the epiphytic fern in relation to
microclimate variation.
In Chapter 6, another species of bird's nest fern, Asplenium antiquum, is studied in the field to
understand the intraspecific variation of this altitudinally widespread epiphytic fern in response
to climate change. Climate change is simulated by reciprocally transplanting individuals
between high (1950 m asl), mid (1100 m asl) and low (600 m asl) elevations. Mortality and
growth rates of transplanted juvenile plants are monitored over two years.
Finally, in chapter 7 again a SDM modelling approach is used to assess climate change impacts
on forests and epiphytes in Taiwan. Here, a novel hierarchical model, tailored for epiphytes
was developed by incorporating dispersal limitation, tree persistence, and non‐climatic factors
and by considering biotic interactions between epiphytes and host trees. The model is used to
identify certain forest types and species that are relatively more sensitive to projected scenarios
of climate change. For identified areas that fall outside current conservation reserves and with
a relatively high number of vulnerable species, additional human disturbance is likely to
exacerbate the effect of climate change, thus deserving prioritised conservation measures.
14 15 Chapter 2
Diversity and phytogeography of vascular epiphytes in a tropical‐subtropical transition island, Taiwan
The Taiwanese epiphyte flora is dominated by
Pteridophytes (i.e. ferns and fern allies, 171
spp.) followed by orchids (120 spp.).
CHAPTER 2
Diversity and phytogeography of vascular epiphytes in
a tropical-subtropical transition island, Taiwan
Rebecca C.-C. Hsu & Jan H. D. Wolf
Flora 204, no. 8 (2009): 612-627
Abstract
We present the first checklist of vascular epiphytes in Taiwan, based on herbarium specimens,
literature records and field observations. Epiphyte phytogeography was analyzed using
Takhtajan’s modified division in floristic regions. We ascertain the presence of 336 species of
vascular epiphytes (24 families, 105 genera) in Taiwan. Pteridophytes contribute most species
(171 species), followed by orchids (120 species). Epiphytes contribute eight percent to
Taiwanese floristic diversity and epiphyte endemism is near 21.3%. The extensive mountain
system is probably the most effective driver for epiphyte diversification and endemicity in
Taiwan. Phytogeographically, Taiwanese epiphytes exhibit equal affinity to the Malesian region,
southern China and Indo-China, and Eastern Asiatic regions. However, some species have a
disjunctive distribution between Taiwan and SW China and/ or E Himalaya, presumably
related to low habitat similarity with adjacent China and/or the legacy of Late Quaternary
climate change. Vascular epiphyte distribution patterns corroborate the phytogeographical
separation of the island of Lanyu from the main island of Taiwan along Kanto’s Neo-Wallace
Line.
Introduction
The conspicuous vascular epiphyte community in the canopy of wet tropical forests has
attracted botanists as early as 1888, especially during the second half of the last century
(Benzing, 1990; Gentry and Dodson, 1987a; Johansson, 1974; Kress, 1986; Madison, 1977;
Richards, 1952). These studies have shown that the epiphytic life-form is a successful
adaptation of plants to conditions in the canopy, comprising ca. 29,000 species, or
approximately 10% of all vascular plants, in 83 different families and 876 genera (Gentry and
Dodson, 1987a). Whereas the number of epiphyte inventories is gradually increasing,
inventories from the paleotropics are still rare and especially from Asia few inventories are
available (Wolf and Flamenco-S, 2003). In addition, little is known about epiphytes in tropicalsubtropical transition zones. Consequently, the differences in vascular epiphyte diversity and
17 EPIPHYTE DIVERSITY AND PHYTOGEOGRAPHY composition between temperate and tropical areas and between paleotropics and neotropics
remain ambiguous and lack generally accepted explanations (Benzing, 1987; Gentry and
Dodson, 1987a; Zotz, 2005).
Fig. 1 Location of Taiwan, Lanyu, Lutao, and the Neo‐Wallace Line (Kanto 1993) 18 CHAPTER 2 Taiwan (formerly known as Formosa) is a continental island, separated from
Southeast China by the ca. 200 km wide Taiwan Strait, which reaches a depth of 70 meters. The
Tropic of Cancer crosses through the middle of the southern half of the island, and about 70%
of the total area is covered by mountains. Taiwan owes its existence to a collision of the
Philippines Sea plate with the Eurasian continental margin some five million years ago, which
induced orogenesis (Ho, 1988). In contrast to many other regions at the tropic of Cancer or
Capricorn, Taiwan has a humid climate thanks to the high mountains that induce cloud
formation in high-humidity oceanic winds. Frequent typhoons in summer and NE monsoon in
winter provide most precipitation throughout the year.
Taiwan floristic diversity is high, comprising ca. 4077 species (Hsieh, 2003). Being a
mountainous island, species diversity is the result of great habitat heterogeneity. Furthermore,
situated at the transition from tropics to subtropics, in Taiwan many tropical plant species
reach their northern limit (Hsueh and Lee, 2000), whereas temperate species are found in the
high mountains (Hosokawa, 1958). Phytogeographically, Taiwan belongs to the Eastern Asiatic
region (Takhtajan, 1986). Yet the south end of Taiwan, Henchun Peninsula, and two small
volcanic islands, Lanyu and Lutao, located in the south-eastern Taiwan, are pertained to
Malesian region (Fig. 1, Fig. 2). The vegetation of Lanyu is characterized by tropical rain
forests, and its flora and fauna have more in common with the Philippines than with Taiwan.
On this basis, Kanto (1933) proposed the Neo-Wallace Line by extending the boundary of
Dickerson and Merrill’s Line (Dickerson, 1928) from northern Luzon to Lanyu through the
middle sea of Lanyu and Lutao (Fig. 1). Kanto’s proposal was corroborated by several
subsequent biogeological studies (Hosokawa, 1958; Kanehira, 1935; Yen et al., 2003).
In this study we describe the epiphyte flora of Taiwan for the first time. Specifically,
we address the following research questions: (i) is species richness, endemism, and familial
makeup similar to that of other floristic regions such as tropical and temperate areas in the
neotropics, (ii) what is the phytogeographical affinity of epiphytes and several sub-categories,
(iii) do epiphytes corroborate the Neo-Wallace Line?
Materials and Methods
Study Site
Taiwan is situated between 21˚45'-25˚56'N and 119˚18'E-124˚34'E with an area of 36,000 km2
(Fig. 1). The Central Ridge of Taiwan comprises over 200 peaks higher than 3000 meters asl,
and Yushan is the highest (3952 m) peak in this island. The annual rainfall ranges from 1000 to
over 6000 mm (data from 1949-2004). Mean monthly temperature in the lowlands ranges from
15 to 20Ԩ, and is about 28Ԩ in summer. Based on bioclimatic analyses, Taiwan can be
classified into seven climatic regions, and Lanyu is separated independently (Su, 1984, 1992).
19 EPIPHYTE DIVERSITY AND PHYTOGEOGRAPHY Lanyu (ca. 46 km2, also known as Botel Tobago, Kotosho, and Orchid I.) and Lutao (ca. 16
km2, Green I., Kwasyoto I., and Samasana I.) are small tropical islands located at 22˚03’N,
121˚32’E and 22˚40N, 121˚29E, respectively. During summer and early autumn, typhoons
frequently hit Taiwan, which have less impact in western Taiwan, sheltered by the Central
Ridge.
Fig. 2 Takhtajan’s floristic regions. Numbers indicated: 2, Eastern Asiatic region; 2‐20, Ryukyu islands; 2‐25, SW China; 2‐27, E Himalaya; 12, Sudano‐Zambezian region; 15, Madagascan regions; 16, Indian region; 17, Indochinese region; 18, Malesian region; 18‐104, Philippines; 19, Fijian region; 20, Polynesian region; 22, Neocaledonian region; 29, NE Australian region. Regions that not covered in above map but with Taiwanese epiphyte occurrence are: 3, North American Atlantic region; 4, Rocky Mountain region; 6, Mediterranean region; 8, Iran‐Turanian region; 9, Madrean region; 10, Guineo‐Congolian region; 21, Hawaiian region; 23, Caribbean region; 24, Guayana Highlands; 25, Amazonian region; 26, Brazilian region; 27, Andean region. The figure was modified from Takhtajan (1986). 20 CHAPTER 2 Epiphyte definition
We define epiphytes as organisms that grow on plants without extracting water or nutrients
from hosts’ living tissues, following Barkman (1958). In this paper, focus is on vascular plants,
but many other epiphytic organisms are found in the canopy of the forest. In addition, it is not
rare to find accidental epiphytes growing on other plants, which are unable to reproduce in the
canopy (Moffett, 2000). We excluded accidental epiphytes from our checklist and classified
vascular epiphytes in following sub-categories:
i. Holo-epiphytes: epiphytes that complete their entire life cycle without contacting the forest
floor (Benzing, 1990).
ii. Hemi-epiphytes: epiphytes that complete part of their life cycle as terrestrial plants. Primary
hemi-epiphytes begin their life cycle as epiphytes and eventually send their roots to the ground
(e.g. strangler figs), whereas secondary hemi-epiphyte seedlings germinate terrestrially to
become epiphytic secondarily when their rooting shoots decompose (e.g. aroids).
iii. Facultative epiphytes: species in which some individuals are terrestrial.
Epiphyte checklist
Botanically, Taiwan is one of the best explored regions in the tropics. The national database
houses over 200,000 botanical records (ca. 60% of herbarium collections). We gratefully made
use of this wealth of information, scrutinizing for epiphytes in well-known epiphytic
taxonomic groups (Benzing, 1990). In addition, we used epiphyte records in published plant
inventories and floras. Nomenclature follows the 2nd edition of the Flora of Taiwan (Boufford
et al., 2003). To compile this checklist, species listed in Flora of Taiwan were examined one by
one, and the approximate number of epiphytes was ascertained.
Phytogeography analyses
We assessed the presence of Taiwanese vascular epiphytes in Takhtajan’s floristic regions
(Takhtajan, 1986). The floristic provinces, SW China, E Himalaya, Ryukyu and Philippines
under Eastern Asiatic and Malesian regions of Takhtajan’s system, were recognized
independently (Fig. 2). Species geographical distributions were consulted flora of Taiwan and
collections in the global biodiversity information facility (GBIF) online database. For smaller
floristic provinces, such as SW China and Ryukyu, the floras of Japan and China were
consulted to determine the specific occurrence locations.
21 EPIPHYTE DIVERSITY AND PHYTOGEOGRAPHY Table 1 Contribution of vascular epiphytes to the flora of Taiwan in various taxonomic categories (data Flora of Taiwan, Boufford et al., 2003). All vascular plants Ferns & allies Angiosperm Dicotyledons Monocotyledons Families 24/235(10%) 12/37(32%) 12/190(6%) 10/151(7%) 2/39(5%) Genera 105/1419(7%) 48/145(33%) 57/1257(5%) 16/901(2%) 41/356(12%) Species 336/4077(8%)* 171/629(27%) 165/3420(5%) 40/2410(2%) 125/1010(12%) *Epiphyte‐Quotient Results
Species richness, family makeup, and endemism
There are 336 species of vascular epiphytes in 105 genera and 24 families in Taiwan and two
subsidiary isles, Lanyu and Lutao (Appendix 1). Obligate holo-epiphytes comprise 271 (81%)
species, 41 (12%) species are facultative holo-epiphytes, and 7 (2%) and 17 (5%) species are
primary and secondary hemi-epiphytes, respectively.
The Taiwanese epiphyte flora is dominated by Pteridophytes, i.e. ferns and fern allies,
comprising 171 species (Table 1). The number of orchids is also substantial, 120 species (Fig.
3). The ten most species-rich families contain 89% of all epiphytes and the remaining plant
families with epiphytic representatives only contribute about 11% to total epiphyte richness
(Fig. 3). At the genus level also, epiphytism is concentrated in few taxa. Only five percent of
the genera contain more than 10 species and 54 (51%) genera are represented with a single
species only in the region. More than a quarter of native Pteridophytes (Table 1) and 36% of
native orchids are epiphytes. In contrast, the Epiphyte-Quotient (Ep.-Q, Hosokawa, 1950), i.e.
the proportion of epiphytic species in the flora, is only approximately 8% (Table 1).
Of the 336 epiphytes, 75 are endemic species. Sixty-nine species are confined to
Taiwan, and one disjunctively occurs in Taiwan and Lanyu. Despite the small size of Lanyu
and Lutao, 5 species are confined here (4 species are endemic to Lanyu, and one species is
shared by both). The proportion of Taiwan endemic epiphytes (21.3 %, Table 2) is less than
that in the entire flora (26.2 %, Hsieh, 2003). Most endemic epiphytes are orchids (54.2 %)
despite overall higher number of epiphytic Pteridophytes in Taiwan. Of all 114 epiphytic
orchids, 38 species (33.3%) are endemic to Taiwan, as opposed to 19 species (11.2%) of
Pteridophytes (Table 2).
22 CHAPTER 2 Table 2 Floristic affinity of Taiwan epiphyte flora with phytogeographical regions, following Takhtajan (1986). Given is the proportion (%) and number of Taiwanese species, in parentheses, of epiphytic Taiwanese species per region. Floristic Regions Taiwan (324) Pteridophytes (170) Orchids (114) 38.9 (126) 50.7 (35) 64.0 (16)
48.8 (83) 25.4 (29) China, Japan, Korea
27.2 (88) 31.2 (53) 20.2 (23) E. Himalaya & S.W. China
13.0 (42) 4.4 (3) 0.0 (0) 13.5 (23) 13.2 (15) Ryukyu
13.0 (42) 29.0 (20) 24.0 (6) 18.2 (31) 6.1 (7) 40.9 (132) 71.0 (49) 72.0 (18)
51.8 (88) 25.4 (29) Malay archipelago 31.2 (101) 49.3 (34) 64.0 (16)
42.4 (72) 14.0 (16) 8.0 (2) 9.4 (16) 11.4 (13) Indo‐China 39.2 (127) 46.4 (32) 60.0 (15)
43.5 (74) 35.1 (40) India & Sirilanka 23.5 (76) 29.0 (20) 52.0 (13)
28.8 (49) 14.9 (17) Melanesia & Hawaii 12.0 (39) 26.1 (18) 44.0 (11)
20.0 (34) 1.8 (2) Africa 4.9 (16) 8.7 (6) 8.0 (2) 7.1 (12) 0.9 (1) Australia 9.0 (29) 18.8 (13) 36.0 (9) 12.4 (21) 1.8 (2) Neotropis 2.5 (8) 5.8 (4) 4.0 (1) 3.5 (6) 0.0 (0) Holarctis other than E.A. 1.5 (5) 2.9 (2) 0.0 (0) 2.9 (5) 0.0 (0) 21.3 (69) 5.8 (4) 0.0 (0) 11.2 (19) 33.3 (38) Eastern Asiatic Region Malesian Region Philippines
Endemicity 9.6 (31) Lanyu (69) 21.7 (15) 40.0 (10)
21.7 (15) 23 Lutao (25) EPIPHYTE DIVERSITY AND PHYTOGEOGRAPHY Epiphyte phytogeography
With respect to phytogeographical region, about 41% of epiphytes in Taiwan also occur in the
Malesian region, including 10% of species shared with only the Philippines (Table 2). About
39% of species are shared with Indo-China, and about the same proportion is shared with
Eastern Asiatic regions, which cover temperate E Asia, E Himalaya, SW China, and Ryukyu.
The islands Lutao and Lanyu share most species (over 70%) with the Malesian region, whilst
Lutao has a high proportion (40%) of species that also occur in temperate E Asia. Only Lanyu
shares an exceptional high proportion (22%) of species with the Philippines (Table 2).
Overall, epiphytic ferns shared more species with other floristic regions than total
epiphytic species (Table 2). Over forty percent of Taiwanese epiphytic ferns also occurred in
Eastern Asiatic, Malesian, and Indochinese regions. Epiphytic orchids exhibited the highest
affinity (35%) to Indo-China, yet shared no species with Neotropical and Holarctic areas,
except E. Asia.
Fig. 3 Ten most species‐rich epiphytic families and their contribution to total epiphyte flora in Taiwan. Numbers in parentheses are species numbers. Shading indicates Pteridophyta. 24 CHAPTER 2 Discussion
Species richness and taxonomic distribution
For a paleotropical region, the island of Taiwan is with 336 species rich in epiphytes (Table 1).
There is no distinct dry season in Taiwan and abundant rainfall and warm climate promote
epiphyte species richness and growth. Another reason why epiphyte richness is high may be
that Taiwan served as a refuge during Late Quaternary climate change, which has been used to
explain the exceptionally high diversity in Taiwan (4077 plant species; further discussed below).
In view of this high floristic diversity, Taiwan may even be considered relatively poor in
vascular epiphytes. The contribution of vascular epiphytes to total vascular flora is only eight
percent, whilst the EP.-Q worldwide is near ten percent. Moreover, about 36% of orchids are
epiphytic in Taiwan, which is far less than the 70% worldwide level (Atwood 1986). Possibly
frequent tropical storms have reduced epiphyte diversity in Taiwan. On average, five typhoons
hit Taiwan each year (data from 1958 to 2007, Central Weather Bureau). Typhoons may have a
dramatic impact on forest canopies and cause understory light levels to increase to 30% of
outside levels (Lin et al., 2003). Similarly, low epiphyte diversity in Puerto Rico has been
attributed to island isolation and large-scale hurricane disturbances (Migenis and Ackerman,
1993).
Epiphyte richness in neotropical areas, moreover, is generally higher. For example,
Wolf and Flamenco-S (2003) report 1173 species for the state of Chiapas (75,000 km2, 1618˚N). Typical for any epiphyte flora, the diversity is concentrated in few taxa (Fig. 3, Table 1).
In contrast to the Neotropics, paleotropical areas lack particularly species-rich epiphyte
families (e.g. Bromeliaceae, Cactaceae and Marcgraviaceae) and genera in the orchids (e.g.
Pleurothallis, 1500 spp.; Epidendrum, 720 spp.; Maxillaria, 570 spp.; Stelis, 540 spp.) and in the
aroids (Anthurium 600 spp.; Philodendron (350 spp.) (Benzing, 1990). In Taiwan, the most
abundant epiphytes are ferns, and in this respect Taiwanese epiphyte flora is typical for
temperate regions. However, in comparison with other vegetation types, ecosystems, and
floristic regions, the relative proportion of epiphytic ferns and orchids of Taiwan is not
dramatically different, showing a transition from tropical to temperate regions (Table 3). A
high proportion of ferns and fern allies is probably due to the presence of temperate
mountains in Taiwan that favour epiphytic ferns over, for example, orchids (Kessler et al.,
2001, Zotz, 2005). In Taiwan, no epiphytic orchids are found above approximately 2300
meters asl (Gastrochilus hoii, pers. comm.) in contrast to epiphytic ferns with ultimate altitudes of
ca. 3000 meters asl (e.g. Pyrrosia spp., Lepisorus spp., Mecodium wrightii, pers. observ.).
25 EPIPHYTE DIVERSITY AND PHYTOGEOGRAPHY Epiphyte endemism
Many islands are considered global biodiversity hotspots because of high endemicity of insular
biota (Kreft et al., 2008). Taiwan is no exception, having extraordinary plant endemicity. More
than one thousand vascular plant species are endemic to the island, comprising 26% of the
entire flora. The strikingly high flora endemism can be explained by Taiwan’s extensive
mountain system. Taiwan was formed from the collision between the Philippines Sea plate and
the Eurasian continental margin and gave rise to the Central Ridge of Taiwan in the Mid
Pliocene (3 Ma) (Ho, 1988). Active orogenesis induced a massive earthquake in central Taiwan
as recent as 1999. Orogenesis results in greater microhabitat differentiation of mountainous
regions, which promotes island-wide biodiversity and endemicity. Kreft et al. (2008) concluded
that in continental islands, geographic isolation from the mainland may contribute less to
species diversity than mountain isolation. Our data are in agreement with this conclusion. For
example, several epiphytic genera of mountainous regions, Bulbophyllum (24 spp.), Gastrochilus (9
spp.) and Oberonia (7 spp.), show exceptionally high endemicity of nearly 50 percent.
Furthermore, Goodyera, a mid-elevation (ca. 1500-2000 m asl) species, evolved three epiphytic
species, including two endemics. This is the first report of epiphytism in this genus. Finally,
endemicity increases with altitude in Taiwan up to nearly 60% above 3500 meters asl
Yet, vascular epiphytes show lower endemism (21.3%) than terrestrial plants (Table
2). This may be due to their superior dispersal ability; 89 percent of vascular epiphytes in
Taiwan disperse by wind. The arboreal habitat and dust-like seeds and diaspores enable longdistance dispersal. Overall, ferns show wider ranges and lower endemicity than angiosperms
(Gentry and Dodson, 1987a; Kelly et al., 2004) (Table 2). In contrast with epiphytic seed plants,
most large epiphytic fern genera are preponderantly pantropical (Gentry and Dodson, 1987a).
Apart from dispersal ability, historical factors may also explain species geographical range
(Lester et al., 2007). Kelly et al. (2004) reported that in the tropical Andes species endemism
increased from primitive to advanced taxonomic groups (bryophytes < Pteridophytes <
angiosperms). Furthermore, taxa with narrow geographical range are often considered to have
high speciation rates (Kelly et al., 2004). In this view, the high endemism (33%) in Taiwanese
epiphytic orchids in Taiwan relates to their highly specific pollination system, which, together
with the fragmented canopy habitat, promotes rapid speciation (Benzing, 1987; Gentry, 1982;
Gentry and Dodson, 1987a; Gravendeel et al., 2004).
Epiphyte phytogeography
Taiwan has a relatively unique vascular epiphyte flora. The regions with closest affinity are the
Malesian region, Indo-China, and Eastern Asiatic regions; ca. 40% of Taiwanese species are
shared with those regions. Interestingly, about 13% of vascular epiphytes have a disjunctive
distribution between Taiwan and SW China and/or E Himalayan regions (Table 2). This
26 CHAPTER 2 floristic disjunction is consistent with Hosokawa’s (1958) finding that Taiwan’s flora, especially
of the highland, is more closely related to SW China and E Himalaya than to adjacent coastal
provinces of mainland China. Kuo (1985) indicated similar observations on Taiwanese
Pteridophyte flora. He found that the Pteridophytes of warm-temperate forests (500 to 1800
meters asl) were closely related to SW China and the Himalayan regions, whilst lowland species
showed higher affinity to Ryukyu, south-eastern China and Indo-China.
The simplest explanation for the lower epiphyte affinity of Taiwan with adjacent
coastal regions of south-eastern China is lack of suitable habitats (Kuo, 1985). Due to long
term population pressure and associated agricultural activities, south-eastern China has
endured extensive habitat change. Since epiphytes are most diverse and abundant in oldgrowth forests (Cascante-Marin et al., 2006; Köhler et al., 2007; Wolf, 2005), epiphyte diversity
is especially affected. Furthermore, lowland south-eastern China shows little habitat similarity
with Taiwan mountain areas.
Late Quaternary climate change offers another explanation. On an evolutionary timescale, epiphytism is relatively recent, occurring in evolutionary advanced families of ferns and
seed plants. Orchidaceae did not evolve until the Quaternary (1.6 Ma ago) (Benzing, 1990).
Zotz (2005) discussed the possibility that the Pleistocene extinction was one of the limits of
epiphytism in temperate zones, whilst few temperate areas (e.g. Chile, New Zealand,
Himalayas, Japan) have a high number of epiphytes for being Tertiary refugia. The common
feature of the flora in these areas is a high proportion of autochthonous and monotypic taxa.
During the ice age in the Quaternary, the sea level in the Taiwan Strait dropped, connecting
Taiwan with mainland Eurasia. According to the projected vegetation map of Last Glacial
Maximum (LGM, 18,000 ago), Eurasia had relatively scarce tree cover with scattered areas of
close forests in the uplands across south-western China and along the south-eastern coast of
Eurasia (Ray and Adams, 2001). Presumably, the oceanic climate facilitated Taiwan as a refuge
during Quaternary glaciations. Moreover, apart from high endemicity, more than half of plant
genera in Taiwan are monotypic (Hsieh, 2003). There is an endemic monotypic epiphyte genus
Haraella (Orchidaceae) in Taiwan. Thus, we propose that Late Quaternary climate change helps
explain the disjunctive distribution of many vascular epiphytes between Taiwan and southwestern China as well as eastern Himalayan regions.
Interestingly, the epiphyte flora of Lanyu and Lutao is phytogeographically distinct.
Lanyu has more affinity with the Philippines (22%) in the Malesian region than Lutao (8%),
whereas Lutao shares more species with China, Japan and Korea in the Eastern Asiatic Region
(40%) than Lanyu (22%) (Table 2). This pattern is in agreement with the proposed NeoWallace Line based on insect distributions (Kanto, 1933).
27 EPIPHYTE DIVERSITY AND PHYTOGEOGRAPHY In summary, this one of the few epiphyte inventories in Asia shows that the Taiwanese
epiphyte flora is rich in species and has an extraordinarily high endemicity. Regional mountain
isolation is probably the most effective driver for epiphyte diversification in Taiwan. Regarding
the proportional contribution of epiphytic ferns and orchids, Taiwan is transitional between
tropical and temperate zones. The disjunctive distribution of epiphytes between Taiwan and
SW China as well as E Himalaya suggests low habitat similarity to adjacent China and/or a
legacy of Late Quaternary climate change. Taiwanese vascular epiphyte distributions are in
agreement with the Neo-Wallace Line.
28 29 Chapter 3
Regional and altitudinal patterns in vascular epiphyte richness on an East Asian island
In Taiwan, active orogenesis
has created an extensive mountain system with high vegetation heterogeneity, providing diverse habitats for epiphyte growth. CHAPTER 3
Regional and altitudinal patterns in vascular epiphyte
richness on an East Asian island
Rebecca C.-C. Hsu, Jan H.D. Wolf & Wil L.M. Tamis
Abstract
The distribution of species on mountains has been related to various predictor variables,
especially temperature. Thermal specialization, which is presumed to be more pronounced on
tropical mountains than on temperate mountains, accounts for the elevational pattern of
species richness and varies between organisms and geographic areas. In this study, the
elevational and regional distribution patterns of 331 epiphyte species in Taiwan were explored
using 39,084 unique botanic collections, mostly from herbaria. Species richness showed a peak
in elevation between 500 and 1500 m. This peak could not be explained by a null model, the
mid-domain effect, suggesting that environmental variables accounted mostly for the
distribution of species on the mountains. Next, species distributions were modelled (with 30
predictor variables) to assess epiphyte regional and altitudinal distribution patterns. The model
results not only corroborated the position of the mid-elevation peak in richness, they also
identified two mountain areas on the island with exceptionally high species richness. These
areas of high epiphyte diversity coincide with areas of high rainfall in relation to the direction
of the prevailing winds. Moreover, a subsequent exploratory ordination analysis showed a
varied thermal preference between epiphyte subcategories (hemiepiphytes, dicotyledons,
orchids and ferns). In contrast to predictions by the Rapoport Effect hypothesis, ordination
analysis also showed that the degree of thermal specialization increased with elevation,
suggesting that highland species may be especially vulnerable to global warming. Finally, the
partial ordination analysis controlling for all other variables suggested that typhoons exert a
significant influence on the distribution of epiphytes.
31 EPIPHYTE DISTRIBUTION PATTERN Fig. 1 The geographical location of Taiwan and climatic zones in the island according to the Köppen‐
Trewartha climate system. Ar = tropical wet climate (coolest month > 18°C), south‐eastern peninsula (< 500 m); Aw = tropical savannah climate (winter dry > two dry months), southern lowlands (< 500 m); Cfa = wet subtropical climate (warmest month > 22°C, no distinct dry season), island‐wide (< 500 m); Cwa = wet subtropical ‐winter dry climate (warmest month > 22°C), south‐western inland hills (500–1000 m); GCfa = mountain climate (warmest month > 22°C, no distinct dry season), island‐wide (500–1500 m); GCfb = mountain climate (warmest month < 22°C, no distinct dry season), island‐wide (1500–3000m). The central range, with an altitude > 3000 m is unshaded. 32 CHAPTER 3 Introduction
The distribution of species on tropical mountains has received renewed interest since highelevation thermal specialists in the tropics could be among the most imperiled species on earth
due to global warming (Laurance et al., 2011). Compared with species in temperate areas,
species in the tropics experience limited annual thermal variability and presumed resulting
thermal specialization may explain the generally relatively low elevation-range of species on
tropical mountains (Janzen 1967). The degree of thermal specialization is nevertheless not
universal, varying among taxa, elevations and geographic locations. Hence, studies with various
species and from different areas are required to attain a complete picture. Moreover, the
assessment of thermal specialization is obscured because, in addition to thermal factors,
hydrological, biotic and other unknown factors may determine the distribution of species on
mountains (Bruijnzeel et al., 2010). Another arguably characteristic of species in tropical areas is
that mountain species have less thermal specialization than lowland species as an extension of
Rapoport’s latitudinal rule (Stevens 1992). Accordingly, on small continental islands such as
Taiwan, overall thermal specialists are relatively rare, largely due to a paucity of upper-zone
specialists (Laurance et al., 2011).
Species distribution patterns on mountains account for the variability in species
richness with elevation. Many organisms show a peak in species richness at mid-elevation
(Laurance et al., 2011), and this is also true for epiphytes (Wolf and Flamenco-S 2003, McCain
2004, Cardelus et al., 2006). In addition to environmental factors, such as temperature, rainfall
and fogs, and historical factors (Gentry and Dodson 1987a, Küper et al., 2004), the midelevation peak in species richness has been explained solely by applying a distribution null
model (i.e. the mid-domain effect). The mid-domain effect arises from geographic constraints
on species range within a bounded domain (Colwell and Lees 2000). Within a landmass
boundary (e.g. from coasts to mountain tops), the null model predicts a peak in species
richness at mid-elevation, simply based on overlapping species’ ranges. For epiphytic
bryophytes in Colombia, the mid-elevation maximum in species richness was indeed explained
by a mid-domain effect (Wolf 1993, Ah-Peng et al., 2012). In contrast, the richness of ferns on
mountains was best accounted for by climatic factors (Kessler et al., 2011).
Here, we present for the first time data on the elevational distribution of epiphytes in
Taiwan, an island in the western Pacific on the transition from tropical to subtropical latitudes.
In Taiwan, active orogenesis has created an extensive mountain system with diverse vegetation
types, ranging from alpine tundra to tropical rainforests. With approximately 4000 species of
vascular plants (including ca. 600 Pteridophytes), the floristic diversity of Taiwan is
exceptionally high compared with other (sub-) tropical islands (Dawson 1963, Reyes-Betancort
et al., 2008, Creese et al., 2011). Taiwan is also one of the botanically best explored regions in
33 EPIPHYTE DISTRIBUTION PATTERN Southeast Asia, and digitized herbarium collections contain over 200,000 records. Despite
having immense plant diversity, Taiwan may be considered relatively poor in epiphytes. There
have been ca. 350 species of vascular epiphytes reported for Taiwan, comprising only eight
percent of the total vascular flora (Hsu and Wolf 2009), which is less than the worldwide level
of 10 percent (Benzing 1990). In some wet tropical ecosystems, the native vascular flora may
consist of up to 35 percent epiphytic species (Gentry and Dodson 1987a). Epiphytes are also
poorly represented on tropical islands of the Caribbean, which has been attributed to
geographical isolation and large-scale disturbances by tropical cyclones (Migenis and Ackerman
1993). In south Florida, a single cyclone (hurricane) may reduce the population density of
epiphytic bromeliads by 12–43 percent (Oberbauer et al., 1996). Tropical cyclones (called
typhoon in Asia) may also have a dramatic influence on forest canopies, increasing understorey
light levels to 30 percent of outside levels (Lin et al., 2003).
The aim of this study was to assess patterns in the distribution of Taiwanese vascular
epiphytes. In particular, we tested the following hypotheses: (1) epiphytes show a mid-elevation
peak in species richness, (2) the peak in richness is explained by a mid-domain effect, (3)
environmental forcing accounts for areas with high species richness, (4) hemiepiphytes,
orchids, ferns and epiphytic dicotyledons have a different thermal preference on the mountain,
(5) upper-zone thermal specialists are relatively rare in comparison with those in the lower
zone, and (6) typhoons influence the distribution of epiphytes.
Methods
Study site
Taiwan is a 36,000 km2 tropical-subtropical transition island (21°45'–25°56'N and 119°18'E–
124°34'E). About 70 percent of the island is covered by mountains (> 1,000 m above sea level
[asl]; Fig. 1), including more than 50 peaks > 3000 m in altitude. Annual rainfall ranges from
1000 mm to > 6000 mm, and falls mainly during the north-east (NE) monsoon (October–
January) and during typhoon-induced heavy rain events (July–September). The NE monsoon
accounts for 45 percent of the total annual rainfall in north-eastern Taiwan (Kao et al., 2004).
On average, 3.7 typhoons hit Taiwan every summer (July–September), of which about 80
percent land on the east coast and track westbound (Wu and Kuo 1999). The torrential
precipitation associated with typhoons mainly causes catastrophic damage to human lives and
natural habitats. For example, the medium (category-2) typhoon Morakot (2009) brought 2777
mm rainfall in 72 h, triggering disastrous flooding, debris flows and landslides, especially in the
mountain area. The typhoon induced-heavy rain supplied by south-westerly monsoon flows
usually interact with the South China Sea summer monsoon (Xie and Zhang 2012). In addition,
the dominated central range on the island often has complex interaction with the typhoon
34 CHAPTER 3 circulation, altering its structure, intensity and path, producing significant mesoscale variations
in pressure, wind and precipitation distribution over Taiwan (Wang 1980). The central range
may decrease the intensity of typhoons by an average of > 40 percent within 12 h of the storm
centre reaching the island (Wu and Kuo 1999), thus typhoon impact is reduced in the western
part of the island on the lee side of the central range.
Table 1. Predictors that were used for modelling species distribution, including four temperature‐
related (1–4), 12 precipitation‐related (5–16), nine topographic variables (17–25), and five land‐
cover/vegetation indices (16–30). Predictor 1 2 3 4 Description
Unit Tmean* TcoldM TdryQ Tsd* Annual mean temperature °C Mean temperature of coldest month Mean temperature of driest quarter The standard deviation of the monthly mean No dimension temperatures 5 Pannual* Annual precipitation 6 PdryM Precipitation of driest month 7 PdryQ Precipitation of driest quarter Millimetre 8 PcoldQ Precipitation of coldest quarter 9‐14 P.1, P.4, P.5, P.6 P.7*, P.10* Monthly rainfall: January, April, May, June, July, October 15 Pdef Water deficiency: monthly precipitation minus Millimetre doubled monthly mean temperature minus °C 16 Pcv* The coefficient of variation of the monthly mean No dimension precipitation 17‐18 Eastness* Northness* Aspect transformed by sin(aspect rad) and Ordinal: 0–8 cos(aspect rad) 19 Soilcode Soil category Cardinal: 0–9 20 SoilPH Soil alkalinity Ordinal: 0–9 2
21 Estd The standard deviation of elevation within 1‐km No dimension 22 Elevation Altitude above sea level 23 Dto3000* The distance to the nearest location above 3000 m [asl] Metre 24 DtoSea The distance to the nearest coast 25 DtoRiver The distance to the nearest river 26 Landcover* Land‐cover classification Cardinal: 0–27 27‐30 EVI.1–4 Monthly enhanced vegetation index, EVI.1: spring (April to May), EVI.2: summer (June to September), EVI.3: NE monsoon initiation No dimension (October to November), EVI.4: slow growth season (December to March) * Predictors only used in model building for species with < 80 occurrences. 35 EPIPHYTE DISTRIBUTION PATTERN Data collection
We compiled occurrences of epiphytic species in herbarium records, published plant
inventories and our own botanical observations as a georeferenced epiphyte database that
finally comprised 39,084 records in 331 species (24 families, 105 genera). Pteridophytes
contributed most species (171), followed by orchids (120). The epiphyte species in the database
were divided into four subcategories based on life form and taxonomy: hemiepiphytes,
(abbreviation Hemis, e.g. Moraceae, Araceae), ferns and fern allies (Ferns), orchids (Orchids)
and dicotyledons (abbreviation Dicots). For more detailed information on the species in the
database, see Hsu and Wolf (2009).
To assess species richness patterns along the altitudinal gradient (Hypotheses 1 and 2),
we used the entire epiphyte collection database. We computed species accumulation curves
(sample-based rarefaction) and associated richness estimators (Chao 2005), using the freeware
program EstimateS 8.2 (Colwell 2011). The species range midpoints were tested against the
mid-domain effect hypothesis using Mid-Domain Null, a Monte Carlo based simulation
programme, applying 1000 permutations without replacement (McCain 2004).
To evaluate regional species richness, species thermal niches and the influence of
typhoons (Hypotheses 3 to 6), we subsequently assembled the plotless herbarium collection
localities as records in grid cells with a spatial resolution of 1 km2. Multiple occurrences of the
same species in a single cell were considered a ‘unique’ record. The final database comprised
28,693 records. A total of 252 species occurred in more than 10 cells, and the most widespread
species occurred in 1613 cells. It is well known that systematic botanists have a bias for certain
accessible localities and taxonomic groups, and the absence of species in cells is possibly due to
insufficient sampling. To endeavour to fill in the distribution gaps, we used species distribution
models (SDMs).
Epiphyte species distribution models
In our SDM, we used a maximum entropy method, MaxEnt (version 3.3.3k) (Phillips et al.,
2006). In MaxEnt, species’ occurrences are related by predictor variables across a series of
observation sites to recognize the realized niche of each species (Guisan and Thuiller 2005).
Statistically speaking, the MaxEnt model minimizes the relative entropy between probability
densities of species presence data and the background landscape (Elith et al., 2011). MaxEnt
uses species presence-data only and we entered all species in the model that occurred in at least
ten 1-km2 grid cells. Since MaxEnt puts no weight on the absence of a species, it is particularly
suited for high-canopy epiphytes, which are often difficult to detect from the ground (FloresPalacios and García-Franco 2001).
36 CHAPTER 3 Based on the ecological understanding of epiphytes, 30 environmental variables
(Table 1) were selected for building SDMs at a resolution of 1-km2, comprising 35,928 grid
cells in total. The predictors included nine topographic variables, 16 climatic (12 precipitationrelated and four temperature-related) variables and five land-cover/vegetation indices. The
land-cover classification (27 categories) was derived from data of a national vegetation
inventory and mapping programme (Chiou et al., 2009) combined with satellite data from a
global land cover facility (Hansen et al., 1998). An enhanced vegetation index (EVI) with
improved sensitivity in high biomass regions was obtained from NASA’s Land Processes
Distributed Active Archive Center (see URL http://reverb.echo.nasa.gov/reverb, averages
from year 2001 to 2010). The monthly EVI was further summarized and averaged to represent
annual patterns for spring (April–May), summer (June–September), NE monsoon initiation
(October–November) and slow growing season (December–March). There was a degree of
correlation among some of our predictors, such as mean temperature of coldest month and
annual mean temperature. However, visual maps of these factors indicated regional
heterogeneity in spatial patterns, despite a general similarity throughout most of the island.
Certain small regions may provide ‘unique’ environmental requirements for species with
restricted distribution. There is a reduced need for our modelling method MaxEnt to pre-select
predictors, since it is more stable than most methods when dealing with correlated variables
(Elith et al., 2011). In order to extract all possible information on regional spatial patterns, we
used all thirty candidate predictors to build our SDMs, unless species were represented by only
a few samples (< 80 records). For these rarer species, we pre-selected the candidate predictors
to avoid over-parameterizing models using correlation tests and the result of a preliminary run
of the model. Ten variables (Table 1) were indentified for modelling the distribution of species
with less than 80 samples. To avoid misinterpretation and ensure model reliability, we also
excluded species with fewer than ten records (79 species, 24%) from the SDMs. For different
sample sizes (numbers of records), we used different model validation approaches and varied
the MaxEnt settings for background samples and selection of features.
Background samples
In MaxEnt, a finite collection of points with associated covariates (environmental predictors)
from the geographic area (landscape) of interest is called a background sample (VanDerWal et
al., 2009). Conceptually, the landscape used for the background sample should include the full
environmental range required by the species, and exclude the areas where species are unlikely
to disperse. Areas that have not been surveyed because there is no suitable habitat for the
species should also be excluded (Elith et al., 2011). In this study, we used the MaxEnt
program’s default setting, randomly sampled 10,000 background locations from the given
35,928 grids of covariates covering the island for the common species (≥ 80 occurrences). We
restricted backgrounds sampled from preferred epiphyte habitats for species with few records
37 EPIPHYTE DISTRIBUTION PATTERN (< 80 occurrences) by deriving a set of 5000 backgrounds randomly sampled from the full set
of epiphyte occurrences (29,087).
Features selection
MaxEnt uses the term ‘feature’to describe the transformation of predictors. Currently, MaxEnt
has six feature classes: linear, product, quadratic, hinge, threshold and categorical. The
programme by default (i.e. using Auto features) restricts models to simple features if few
samples were introduced. When there are at least 80 training samples, all six feature types are
used; features are excluded as sample numbers decline (for example, for 15–79 samples, the
product and threshold features are excluded; for 10–14 samples, the hinge feature is excluded;
and for < 10 samples, only linear features are used). We ran preliminary models using 10-fold
cross-validation to estimate predictive performance via held-out data (Phillips 2008). For
species with ≥ 80 samples, the test statistic (the area under the receiver operating characteristic
curve [AUC]) was significantly higher when using all features than when using only the hinge
feature as suggested by Elith et al., (2011). However, for species with few (< 80) samples, using
auto features provided a significantly lower 10-fold cross-validated AUC than the linear and
hinge features (the last two features had statistically equal AUC values). Nevertheless, the hinge
feature exhibited many more violations of AIC (Akaike’s Information Criterion) values than
the linear feature for species with few samples. Therefore, the final models were fitted on the
full data sets (i.e. all samples for model training) using all feature types for species with ≥ 80
samples and using the linear feature for species with < 80 samples.
Model validation
We used several measures to validate the resulting SDMs. We calculated AIC values using
ENMtools 1.3 to determine whether the models had more parameters than samples, which
would have violated the assumptions of AIC (Warren and Seifert 2010). Three SDMs
(containing < 80 samples) were excluded from later analyses at this stage. Next, we used a null
method to test the model significance (Raes and ter Steege 2007). Models with the same
settings as described above were fitted on 29 sets (each with a thousand permutations) of
randomly chosen samples (with intervals of one for 5–30 records, intervals of five for 35–55
records, and intervals of ten for 60–80 records). We applied a curve-fit through the upper limit
of the 95% confidence interval (CI) on the MaxEnt generated AUC values (1000 values per set)
to identify which SDMs had a significantly higher AUC value than expected by chance alone (p
< 0.05). There were 94 SDMs (< 80 samples) excluded from later analyses at this stage. We did
not apply the null test to species with ≥ 80 samples because it had been found previously that
models based on more than 80 samples were rarely insignificant (Hsu et al., 2012).
38 CHAPTER 3 Creating an epiphyte richness map (Hypothesis 3)
We obtained 156 validated SDMs, comprising 68 relatively rare (< 80 samples) and 88
common (≥ 80 samples) species. To create an epiphyte richness map, we applied a threshold
of sensitivity-specificity sum maximization (Liu et al., 2005) to convert the MaxEnt probability
distribution to a predicted presence map for each species. We then overlaid every singlespecies map to produce a species richness map for epiphytes in Taiwan.
Ordination analysis (Hypothesis 4-6)
We used a direct gradient ordination analysis, canonical correspondence analysis (CCA) to
assess the thermal specialization of species and species groups and the influence of typhoons
(Braak and Smilauer 2002). For predictor variables, we used the same 30 environmental
variables that we used in MaxEnt (Table 1). For species, we used the same presence-absence
data from the 156 epiphyte SDMs that we used to create the species richness map. To avoid
multicollinearity, we performed a principal component analysis (PCA) on all variables. The first
extracted PCA component was highly correlated with temperature: Tmean (Loading [L]: –0.27),
TcoldM (L: –0.26), TwarmM (L: –0.26), TdryQ (L: –0.27) and Elevation (L: 0.26). The second
PCA component was highly correlated with rainfall: Pcv (L: –0.30), PdryM (L: 0.29), PdryQ (L:
0.28), PcoldQ (L: 0.29), P01 (L: 0.29), PTY (L: –0.27), and P06 (L: –0.26). The first and second
component together explained 52 percent of the variation. Both components were retained in
the CCA, as opposed to a third component that had little additional explanatory value (9%).
Next, the species and environmental variables (PCA component-1 and -2) were subjected to
CCA We tested the significance of the first extracted CCA axis using a Monte Carlo test (999
permutations). CCA not only generates the species scores on the axes, but also their standard
deviations (called tolerances in CCA), which may be seen as a measure of niche width (Lepš
and Smilauer 2003). Hence, on a generated axis that is highly correlated with elevation, the
species tolerance is a measure of thermal specialization.
We defined typhoon disturbance as the frequency of historical typhoons on the same
1 × 1 km grid (35,928 cells) as the other environmental predictors. Recorded traces and eyes of
typhoons from 1958 to 2006 were plotted as circles with radii of Beaufort scale 7 and 10 (wind
speed = 17.1 and 28.3 m/sec, respectively), and the accumulated numbers of typhoons per cell
were calculated (Lin et al., 2006). To establish whether typhoon influenced the distribution of
epiphytes, we performed a separate CCA analysis using our typhoon frequency data and 156
epiphyte SDMs, entering the PCA components of our 30 environmental predictors as
covariables. Due to the complex interaction between the central range and typhoon circulation,
which causes unpredictable changes in typhoon structure on the lee side of the central range,
we excluded cells from the western part of the island. Therefore, we only considered cells east
of the central ridge (10,725 grids) and past typhoons landing on the east coast of Taiwan to
39 EPIPHYTE DISTRIBUTION PATTERN explore the influence of typhoons on epiphyte vegetation. The ordination analyses were
performed in R with vegan, a community ecology package (Oksanen et al., 2012, R Core Team
2012).
Results
Epiphyte species richness showed a peak in species richness between 500 and 1500 m (Table 2;
Fig. 2). Although the difference between the species richness at < 500 m and richness at 500–
1000 m was small, it was significant (M = 306.03, SD = 8.53 and M = 308.33, SD = 9.47,
respectively; paired t(24740) = 20.1, p < 0.05). Species richness showed a more rapid decrease
above 1500 m. The elevation-species richness curve fell outside the 95% CI curves of the middomain null model curve (Fig. 3); hence the mid-elevation peak in richness could not be
explained by a mid-domain effect.
Fig. 2 Epiphyte species accumulation curves based on sample‐based rarefaction (software program EstimateS 7). Collections per altitudinal interval: 12,944 (< 500 m), 11,798 (500–1000 m), 7,198 (1000–1500 m), 4,024 (1500–2000 m), 2,197 (2000–2500 m) and 923 (> 2500 m). 40 CHAPTER 3 Table 2. Number of epiphyte species per altitudinal interval in Taiwan; n = number of records, Sobs = the number of observed species, singletons = number of species that were only recorded once and Schao = estimated number of species for 95% CI and standard deviations (SD). Altitudinal interval < 500m 500–1000m 1000–1500m 1500–2000m 2000–2500m > 2500m n Sobs Singletons Schao Schao 95% CI SD 12944 11798 7198 4024 2197 923 286 289 281 235 205 165 39 29 28 39 39 37 306.03 308.33 293.6 276.17 232.44 185.81 (294.99, 330.59) (296.79, 336.98) (285.91, 313.32) (253.47, 326.76) (217.43, 265.590 (174.2, 212.08) 8.53 9.47 6.42 17.56 11.56 9.06 Fig. 3 The species richness curve (with data points), based on 39,084 collections and 331 epiphytic species, and the 95% CI null model prediction curves sampled without replacement (software program Mid‐Domain Null, 1000 simulations). 41 EPIPHYTE DISTRIBUTION PATTERN The mid-elevation peak in species richness was also shown by the epiphyte SDMs:
most epiphyte species being found in the mild mountain climate of the GCfa KöppenTrewartha climate zone between 500 and 1500 m (Fig. 1). The SDMs also identified two
regions with high diversity on the western slope of the central range at mid-elevations:
HsuehShan in northern Taiwan and AliShan in central Taiwan (Fig. 4A). Both areas are
sheltered from east-coast landing typhoons and receive high amounts of annual rainfall under
the influence of NE monsoons and SW flows, respectively. SDMs also showed considerable
variation in distribution patterns between epiphyte subcategories. The most south-eastern tip
of Taiwan (Fig. 1, HenChun peninsula) is characterized by a tropical wet lowland climate (Ar),
and contained the highest percentage (11%) of hemiepiphytes (Hemis) of all the climatic zones.
Epiphytic ferns were most often found in the cool mountain climate of the Köppen-Trewartha
GCfb climate zone (67%), at an altitude of 1500–2500 m.
Fig. 4 (A) Modelled richness of pooled epiphytes (156 spp.) and two regions with exceptional high epiphyte diversity, namely HsuehShan and AliShan, both located at mid‐elevation (800–2000 m asl). (B) Richness pattern for the half of the modelled species (78 spp.) that had lower thermal tolerance values (i.e. the species with the highest thermal specialization). (C) Richness pattern for the other half of the modelled species that possessed higher thermal tolerance values. 42 CHAPTER 3 The thermal preference of the subcategories was also indicated by the CCA analysis.
The distribution of the species on a significant first axis (explained variance 15.0%, Monte
Carlo, p = 0.001) constrained by elevation, PCA component-1, indicated that hemiepiphytes
(e.g. Ficus spp.) were predominately found at lower elevations, whilst epiphytic ferns (e.g.
Crypsinus quasidivaricatus, Lepisorus clathratus and L. suboligolepidus) were the most prominent low
temperature specialists (Fig. 5). Epiphytic orchids were found from low elevations (e.g. Liparis
grossa and Appendicula reflexa) to upper mountains (Gastrochilus hoii). The thermal specialization
(inverse niche-width) of the 156 analysed species was higher with increasing elevation,
particularly for some epiphytic ferns and orchids (Pearson´s R = -0.39, p < 0.001, Fig. 6).
Moreover, the half of the species (78 species) possessing relatively high thermal specialization
(low tolerance values) exhibited an obvious mid-elevational pattern (Fig. 4B) in comparison to
the indistinct distribution of the other half of species with a higher tolerance value (Fig. 4C).
Fig. 5 CCA ordination diagram of the species scores (156 spp.) on the first two axes with epiphyte species arranged by subcategory: orchids (open squares), ferns (open triangles), hemiepiphytes (black diamonds) and dicotyledons (black circles). The first axis (eigenvalue 0.4534, explaining 15% of total variance; Monte Carlo p < 0.001), is highly correlated with temperature (elevation), with higher elevation shown towards the right. The second axis (eigenvalue 0.1173, explaining 4% of variance) is related to water availability, with reduced water availability shown towards the top. 43 EPIPHYTE DISTRIBUTION PATTERN Fig. 6 Species scores (156 spp.) on the first constrained canonical axis generated by CCA and standard deviations. The first axis is highly correlated with temperature (elevation), and thus its standard deviation (i.e. tolerance) may be interpreted as a measure of niche‐width. The thermal tolerance of the species is lower (i.e. higher thermal specialization) with increasing elevation towards the right (Pearson´s R = –0.39, p < 0.001). Epiphyte species are arranged by subcategory: orchids (open squares), ferns (open triangles), hemiepiphytes (black diamonds), and dicotyledons (black circles). Independent of thermal and rainfall influences (PCA component-1 and component2), partial CCA showed that typhoons also exert an influence on the distribution of epiphytes
(explained variance 1.6%, Monte Carlo p < 0.001; Fig.7). The two typhoon intensities that we
analysed (scales 7 and 10) had largely opposing effects on the epiphyte community.
Discussion
In agreement with many tropical epiphyte studies from the American continent (Gentry and
Dodson 1987a, Wolf 1993, Wolf and Flamenco-S. 2003, Krömer et al., 2005, but see Ibisch et
al., 1996), our analyses, which used both empirical collections and SDMs, showed that vascular
epiphytes had a peak in species richness at a mid-elevation on mountains. Recognizing that
botanic collections are essentially non-random, we nevertheless presume that, for our data, a
meaningful assessment of the observed distribution and diversity patterns is possible because
of the extremely high number of unique records in the database (39,084).
44 CHAPTER 3 Fig. 7 Partial CCA ordination diagram with typhoon scales (arrows) as variables and epiphyte species arranged by subcategory: orchids (open squares), ferns (open triangles), hemiepiphytes (black diamonds) and dicotylendons (black circles). Independent of temperature and water availability, typhoons exert a significant influence on epiphyte distributions (explaining 1.6% of variance; Monte Carlo p < 0.001). The mid-elevational peak in species richness could not explained by a null model. The
result indicated a substantially higher species richness, and a richness peak at slightly lower
elevations than expected by the null test. A similar pattern has also been described for an
Indian Ocean island (Ah-Peng et al., 2012). Such a pattern is probably explained by the
Massenerhebung effect (i.e. mountain mass elevation effect; Schroeter 1908, Bruijnzeel et al.,
1993). This phenomenon occurs on isolated, small coastal mountains, where floristically-similar
vegetation types tend to distribute at lower altitudes than on large mountain masses due to
climatic compression. Moreover, the exceptionally high species richness observed at island
mid-elevation may be augmented by a large number of species with a small range size in
relation to fine niche partitioning. For instance, the restricted altitudinal band of Chamaecyparisdominated cloud forest (1800–2500 m), characterized by low temperatures and continuously
moist and dim conditions, is inhabited by no less than 92 species of rare ferns (Moore 2000).
Environmental factors may thus account for the observed epiphyte distribution and, with this
in mind, our approach using SDMs to complement grid cells with absent species is not
unreasonable.
The SDMs identified two centres of epiphyte diversity, one in the north (HsuehShan)
and another in central Taiwan (AliShan). Both areas are at a mid-elevation (800–2500 m asl)
45 EPIPHYTE DISTRIBUTION PATTERN and are also subject to significant precipitation, being under the influence of the NE monsoon
in the winter or the SW rains that follow typhoons in the summer, respectively. Prevailing
winds probably import diaspores to these two regions, since the majority of epiphytes in
Taiwan (89%) are wind dispersed (Hsu and Wolf 2009); the ferns Asplenium hondoense and A.
pekinense, which have an affinity with temperate East Asia and Japan, are only found in small
regions under the influence of the NE monsoon (Moore 2000). Humid conditions and
accessibility probably both contribute to the presence of areas of high epiphyte richness and
endemism. Accordingly, the HsuehShan and AliShan regions merit special attention from
conservationists.
The SDMs also showed that whilst epiphytic ferns were relatively common in
northern Taiwan, central Taiwan has a relatively high number of epiphytic orchids. Central
Taiwan receives little influence from the NE monsoon and is therefore relatively dry and warm
in winter. Thus, in agreement with many other epiphyte studies (Gentry and Dodson 1987a,
Benzing 1990, Wolf 1994), the SDMs confirmed that, of all environmental predictors that were
used in the models, elevation and water availability accounted to great extent for the
distribution of epiphytes in Taiwan.
The relative importance of temperature (elevation) and water availability was also
demostrated by the exploratory multivariate ordination analysis, where temperature was
identified as the most important variable. In agreement with patterns in the neotropics (Wolf
and Flamenco-S. 2003, Benavides et al., 2010), hemiepiphytes such as aroids and Ficus species
dominated moist stream valleys in the lowlands (< 1000 m asl), especially in the south-eastern
peninsula where monthly mean temperature was > 18 ˚C. As in the Andes, epiphytic ferns
were particularly adapted to mountain climates (Kessler 2011); the epiphytic fern Crypsinus
quasidivaricatus (Hayata) Copel was recorded near the timberline (3500 m asl).
Interestingly, our results showed that thermal specialization or inverse thermal nichewidth was not uniform along the elevational gradient, but increased with altitudes. Epiphytic
ferns in Bolivia show a similar pattern (Kessler 2011). However, this pattern contrasts with the
Rapoport effect hypothesis, which suggests that the elevational ranges of species are greatest at
higher altitudes, and consequently thermal-specialist species are more likely to colonize lower
altitudes than higher altitudes (Stevens 1992, Laurance et al., 2011). Studies on thermal
specialization in tropical mountains are clearly not conclusive. Our study identified several
montane cloud forests at mid-elevations with many epiphytic thermal specialists. In midelevational cloud forests, the frequently occurring fog events lead to little diurnal (and seasonal)
thermal variation, which according to the Rapoport effect promotes thermal specialization and
susceptibility to global climate change (Foster 2001, Mulligan 2010, Ah-Peng et al., 2012).
46 CHAPTER 3 The ordination analysis also suggested for the first time that typhoons have a
significant influence on the distribution of epiphytes, independent of the temperature- and
humidity-related variables that were used in our analysis. Typhoons may directly blow
epiphytes of their hosts, or indirectly alter the microclimate through mechanical defoliation of
canopies (Mabry et al., 1998). Hemiepiphytes, such as strangler figs and aroids, seem relatively
resistant to direct wind-blow in terms of their tightly-attached adventitious roots on hosts,
which may explain why they are largely found on the south-eastern peninsula of Taiwan,
despite ca. 11 percent of all typhoons landing in this region. Whereas powerful typhoons can
be damaging to epiphytes, moderate typhoons are likely less damaging or may even promote
epiphyte proliferation because of the significant accompanying precipitation. This may explain
why Beaufort scale 10 and scale 7 typhoons have differing effects on epiphyte distributions.
Further field studies are necessary to properly identify how typhoon may affect epiphyte
distribution, which currently remains elusive.
47 Chapter 4
Canopy CO2 concentrations and crassulacean acid metabolism in Hoya carnosa in a subtropical rain forest in Taiwan: consideration of CO2 availability and the evolution of CAM in epiphytes
Hoya carnosa
Fushan is a subtropical rainforest with annual rainfall above 3.5 m. Although here the average daily humidity throughout the year typically approaches 95 %, our study indicated that the likely ecophysiological significance of CAM in H. carnosa remains water conservation rather than CO2
availability.
CHAPTER 4
Canopy CO2 concentrations and Crassulacean acid
metabolism in Hoya carnosa in a subtropical rain
forest in Taiwan: consideration of CO2 availability
and the evolution of CAM in epiphytes
R. C.-C. Hsu, T.-C. Lin, W.-l. Chiou, S.-H. Lin, K.-C. Lin & C. E. Martin
Photosynthetica 44, no. 1 (2006): 130-135
Abstract
The potential importance of CO2 derived from host tree respiration at night as a substrate for
nighttime CO2 uptake during CAM was investigated in the subtropical and tropical epiphytic
vine Hoya carnosa in a subtropical rainforest in northeastern Taiwan. Individuals were
examined within the canopies of host trees in open, exposed situations, as well as in dense
forests. Although nighttime CO2 concentrations were higher near the epiphytic vines at night,
relative to those measured during the day, presumably the result of CO2 added to the canopy
air by the host tree, no evidence for substantial use of this CO2 was found. In particular, stable
carbon isotope ratios of H. carnosa were not substantially lower than those of many other CAM
plants, as would be expected if host-respired CO2 were an important source of CO2 for these
CAM epiphytes. Furthermore, laboratory measurements of diel CO2 exchange revealed a
substantial contribution of daytime CO2 uptake in these vines, which should also result in
lower carbon isotope values than those characteristic of a CAM plant lacking daytime CO2
uptake. Overall, the results of this study indicate that host-respired CO2 does not contribute
substantially to the carbon budget of this epiphytic CAM plant. This finding does not support
the hypothesis that CAM may have evolved in tropical epiphytes in response to diel changes in
the CO2 concentrations within the host tree canopy.
Introduction
Of the three major photosynthetic pathways found among plants, Crassulacean acid
metabolism (CAM) is unique in that atmospheric CO2 is absorbed primarily at night via open
stomata (Kluge and Ting 1978, Osmond 1978, Winter 1985, Lüttge 1987). As a result of PEP
carboxylase activity, the absorbed CO2 is converted to oxaloacetate, which is then reduced to
malate. Throughout the night, the malate is stored in the vacuoles as malic acid. During the
49 CAM EVOLUTION IN EPIPHYTES day, the malic acid leaves the vacuoles and is decarboxylated. The resultant accumulation of
CO2 effects daytime stomatal closure while the CO2 is slowly reduced to carbohydrate via the
typical C3 photosynthetic machinery. By opening their stomata at night and closing them
during the hotter and drier day, CAM plants lose considerably less water during
photosynthesis, i.e., have much higher water-use efficiencies (WUE), relative to C3 and C4
plants (Kluge and Ting 1978, Osmond 1978, Winter 1985, Lüttge 1987). Therefore, it is not
surprising that many CAM plants grow in arid regions of the world. In addition, even more
CAM taxa, primarily in the form of epiphytes, are found in tropical and subtropical areas
(Winter 1985, Lüttge 1989, Winter and Smith 1996). This too is not surprising, given that
drought stress is frequently endured by epiphytic plants between periods of precipitation
(Winter 1985, Lüttge 1989, Benzing 1990, Martin 1994).
In the past decade, CAM has also been reported among various taxa of submerged
aquatic plants (Keeley 1996). Clearly, minimization of water loss (high WUE) was not a
selective force in the evolution of CAM in such plants. Instead, several studies have shown
that low daytime CO2 availability, coupled with high nighttime availability, constitutes an
important benefit of CAM in these aquatic plants (Keeley 1996).
As indicated above, the widespread occurrence of CAM among epiphytes is not
surprising as a result of short, but potentially frequent periods of drought stress. On the other
hand, reports of CAM in epiphytes that are found in the understory and/or dense canopies of
rain forests in tropical and subtropical regions with extremely abundant and frequent rainfall
and a short or no dry season are puzzling (e.g., Martin et al., 1981, 1985, Winter et al., 1983,
1986, Adams 1988, Adams et al., 1988, Griffiths 1988, Kluge et al., 1989, Carter and Martin
1994, Skillman and Winter 1997, Martin et al., 2005). Because atmospheric CO2 concentrations
in the C3 host canopies are higher at night due to respiration of the canopy leaves, relative to
the canopy atmosphere during the day when the host leaves are absorbing CO2 (see references
below), it is tempting to speculate that, as in aquatic CAM plants, CAM might have evolved in
such epiphytes in response to CO2 availability instead of drought stress (Knauft and Arditti
1969, Benzing 1990, Carter and Martin 1994). With this speculation in mind, the goal of the
present study was to determine the degree to which an epiphytic CAM plant utilizes CO2
respired by its host tree at night. Although evidence that extensive usage of such CO2 might
be construed as lending support to the hypothesis that CAM evolved in such plants in
response to CO2 availability, evidence to the contrary would be difficult to reconcile with this
hypothesis.
50 CHAPTER 4 Materials and methods
Study area and plants
The study site was located in a semi-disturbed area (allowing greater accessibility to the plants)
in the Fushan Experimental Forest, a subtropical rain forest at 600 m elevation, in northeastern
Taiwan (longitude 121°34’ E, latitude 24°26’N). Species of dominant trees at this site were
numerous and were primarily in the families Fagaceae and Lauraceae. Climatic conditions at
Fushan are subtropical, with monthly average air temperatures ranging from 10 to 25 °C and
monthly rainfall ranging from less than 10 cm to 50 cm, with maxima occurring in the summer
months (annual rainfall is 3.56 m). Even in January, the month with the lowest average rainfall,
humidities are very high, and rain falls on an average of 20 days of the month; thus, there is no
true dry season at Fushan. The average daily humidity throughout the year typically
approaches 95 %.
In mid-June 2001, twenty Hoya carnosa plants were selected, ten in trees growing in
intact, dense stands of forest (referred to here as “closed” canopies) and ten in trees in the
open with few neighbors (“open” canopies). The latter trees were often in clearings or near
sparsely used roads. Canopy openness and penetration of direct and diffuse irradiance into the
canopy were measured with a Nikon 4500 digital camera, Nikon FC-E8 fisheye lens, and the
software program Delta-T Hemiview Canopy Analysis Software (Cambridge, UK; Lin et al.,
2003). Photographs were taken at a height of two meters in the canopy of each tree by holding
the camera adjacent and perpendicular to H. carnosa individuals and recording exposures in
three directions outward from the tree trunk. These three measurements were then averaged
for each tree. In all cases, individuals of H. carnosa were epiphytic vines extending vertically
along most of the tree trunks.
Atmospheric CO2 concentrations
Air CO2 concentrations at mid-day (11:00 – 13:00 hours) and mid-night (00:00 – 03:00 hours)
were measured using a LI-COR (Lincoln, NE) LI-6400 Portable Photosynthesis System at a
height of two meters, the same height at which leaves were sampled for acidity and carbon
isotope ratios. Air was sampled within 10-20 cm of the H. carnosa plants. No rain fell during
the night and day of measurements; the day during which air was sampled was partly sunny
and warm. The air was calm during both sets of measurements.
Leaf titratable acidity
Shortly before sunset and again shortly after sunrise, a leaf was removed from each of the H.
carnosa plants and frozen (-10 °C) within five minutes of excision. The days on which the
51 CAM EVOLUTION IN EPIPHYTES leaves were sampled were partly sunny and warm. After two days in the freezer, the leaf was
thawed, and a portion was weighed and pulverized in distilled water in a mortar and pestle.
The resultant slurry was titrated to pH 7.0 using 0.01 N NaOH. After titrating, the water was
evaporated, and the dry mass of the tissue was obtained after a week in an oven at 70 °C.
Leaf carbon isotope ratios
Leaves of the H. carnosa individuals and of their host trees were collected and dried for at least
a week at 70 °C, then ground into a powder and combusted for determination of the stable
isotopic composition of their carbon at the University of Arkansas Stable Isotope Facility using
a Carlo Erba elemental analyzer (NA1500 CHN Combustion Analyzer, Carlo Erba
Strumentazione, Milan, Italy) coupled to a Finnigan Delta+ mass spectrometer (Finnigan MAT,
Bremen, Germany) via a Finnigan Conflo II interface. The spectrometer had been calibrated
using the PDB standard. The instrument error (twice the standard deviation) associated with
each measurement was ± 0.1 ‰.
Shoot gas exchange
Plants were collected at the study site, transported to the University of Kansas, and grown in
potted soil in a growth chamber under the following conditions: 50-100 μmol m-2 s-1
photosynthetic photon flux density (PPFD), 30/20 °C day/night air temperatures, 2.43/0.52
kPa day/night vapor pressure deficits, and a photo/thermoperiod of 12 hours. After three
years of growth under these conditions, plants were large, vigorous, and flowering. Shoots
with two to four leaves were sealed in gas exchange cuvettes, and net CO2 exchange was
measured continuously for three days under environmental conditions similar to those in the
growth chamber. The open gas exchange system comprised: a LI-COR LI-6262 differential
infrared gas analyzer; polycarbonate, water-jacketed gas exchange cuvettes with small fans for
air mixing; thermocouple and thermistor temperature sensors and meters for air and leaf
temperature measurements; a temperature-controlled humidifier; and a computer for data
collection. Further details of this system, as well as methods of data analysis, have been
previously described (Harris and Martin 1991, Gravatt and Martin 1992).
Statistics
In most cases, pairs of means were compared using the Student’s t-test or the Mann-Whitney
U-test whenever the data failed to meet the assumptions of the parametric t-test (Sokal and
Rohlf 1981). Day and night air CO2 concentrations measured in the closed and open canopies
were compared with a two-way analysis of variance, followed by the Tukey comparison-ofmeans test (Sokal and Rohlf 1981). In all tests, statistical significance was inferred when P ≤
0.05.
52 CHAPTER 4 Results and discussion
All individuals of H. carnosa performed CAM, as evidenced by significant and substantial
accumulations of acidity in the leaves at night (Fig. 1; also see gas exchange findings below).
Plants in the open canopies exhibited much higher acid accumulations, presumably a result of
the increased availability of irradiance in the more exposed trees (see Fig. 2), although the latter
conclusion is in direct contrast to the conclusions of past studies of sun/shade adaptations in
three species of Hoya, including H. carnosa (Winter et al., 1983, Adams et al., 1987, 1988). On
the other hand, some data from these studies support those of the current study; nocturnal
acid accumulations for plants growing in the field in Australia were greater in plants growing in
full sunlight, relative to values for plants growing in deep shade. Thus, it is clear that more
work is required before the sun/shade status of epiphytic species of Hoya is fully understood.
Leaf acidity [mol g-1]
2000
1500
1000
500
0
open
closed
Fig. 1 Mean morning (black bars), evening (light gray bars), and overnight increases in (dark gray bars) acidity of leaves of Hoya carnosa in open and closed host tree canopies in a subtropical rain forest in northeastern Taiwan. The error bars are standard deviations (n = 10). Differences between both sets of morning/evening means and between the mean overnight increases in leaf acidities are highly significant (P < 0.001). 53 CAM EVOLUTION IN EPIPHYTES 0.5
canopy/open
0.4
0.3
0.2
0.1
0.0
openness direct
diffuse
Fig. 2 Mean measures of irradiance penetration into the canopy of host trees at a height of two meters and adjacent to individuals of Hoya carnosa in a subtropical rain forest in northeastern Taiwan. The error bars are standard deviations (n = 10). Canopy openness is the fractional area of all gaps in the canopy; direct irradiance is the proportion of direct PPFD, relative to that in the open; and diffuse irradiance is the proportion of indirect (diffuse) PPFD, relative to that in the open. All values are based on photographic images and are relative to measurements taken in a fully exposed (open) location nearby. In all cases, the open canopies (black bars) allowed more irradiance (P < 0.001) to the H. carnosa vines than did the closed canopies (gray bars). The CO2 concentration of the atmosphere in the host tree canopies in the Fushan
rain forest was 40-60 μmol mol-1 higher at night than during the day, regardless of the closed
or open nature of the forest canopy (Fig. 3). In addition, nocturnal CO2 concentrations were
higher in the closed canopies (trees in dense forest), relative to the open canopies of the more
exposed trees. This was not the case during the day (Fig. 3). The day/night changes in canopy
CO2 concentrations at a height of two meters in this Taiwanese subtropical rain forest are not
unlike those found at this height in a temperate deciduous forest in Japan (Koike et al., 2001).
Although the latter study reported somewhat larger day/night differences (around 100 μmol
mol-1), CO2 concentrations were measured at dawn and dusk, as opposed to the mid-day and
mid-night measurements in the current study. At another site in Taiwan (approximately 400
km southwest of Fushan), the air CO2 concentrations at a height of two meters in the forest
ranged from diurnal minima around 345 μmol mol-1 to nocturnal maxima of approximately
390 μmol mol-1 (Cheng and Kuo 2004), a diel change of 45 μmol mol-1, which is in the range
of values reported here for the forest at Fushan.
54 CHAPTER 4 -1
Air CO2 concentration [mol mol ]
As expected for a CAM plant, the stable carbon isotope ratios of the epiphytes were
substantially less negative than those of the host trees (Fig. 4), reflecting, in large part, the
discriminatory properties of Rubisco versus PEP carboxylase (Kluge and Ting 1978, Holtum et
al., 1982, Winter 1985, Griffiths 1992, 1993). As was the case with the host trees, the stable
carbon isotope ratios of the epiphytes found in the closed canopies were not significantly
different from those of the epiphytes in the open canopies (Fig. 4). Values reported in the
current study fall between those reported for H. nicholsoniae growing in “deep shade” in an
Australian rain forest (-14.33 ‰; Winter et al., 1986) and those of H. carnosa growing at various
PPFD levels in a glasshouse also in Australia (-20.1 to -22.2 ‰; Adams et al., 1987).
480
440
400
360
320
O pen
Closed
Fig. 3 Mean mid‐night (black bars) and mid‐day (gray bars) CO2 concentrations of the air in the canopies of host trees at a height of two meters and adjacent to individuals of Hoya carnosa in a subtropical rain forest in northeastern Taiwan. The error bars are standard deviations (n = 10). The differences in mean CO2 concentrations between hosts with open and closed canopies were significant (P < 0.05), as were differences between night and day (P < 0.001). In addition, the interaction between canopy closure and time of day was significant (P < 0.05). 55 CAM EVOLUTION IN EPIPHYTES 30
20
13
12
Leaf  C/ C [- per mil]
40
10
0
host
(o p e n )
host
(c lo s e d )
H o ya
(o p e n )
H oya
(c lo s e d )
Fig. 4 Mean δ13C/12C values (‐ ‰) of leaves of Hoya carnosa (gray bars) and their host trees (black bars; open and closed canopies) in a subtropical rain forest in northeastern Taiwan. The error bars are standard deviations (n = 10). Differences between open and closed host canopies in both sets of plants were not significantly different (P > 0.05). The general goal of this study was to provide evidence that might be used in support
of or against the hypothesis that CAM might have evolved in tropical epiphytes in response to
increased CO2 availability at night. Air CO2 concentrations within the host tree canopies were
indeed substantially higher at night than during the day. Because the host trees were all C3
plants, this respired air would be highly depleted in 13C, and its stable carbon isotope value
would be similar to that of the host leaves, i.e., around -30 ‰ (Rundel et al., 1989, Griffiths
1993). Thus, the air surrounding the epiphytes should have an isotopic composition that
represents a mixture of air from above the canopy, presumably with a carbon isotope ratio
around -8 ‰ (Rundel et al., 1989, Griffiths 1993), with the host-respired air. As a result,
although the stable carbon isotope ratio of the canopy air was not measured, its value was
presumably more negative, and probably substantially so, than -8 ‰. Indeed, several studies
have reported more negative carbon isotope ratios of the air and plant tissues inside tropical
forests as a result of respired CO2 (Medina and Minchin 1980, Schleser and Jayasekera 1985,
Da Silveira et al., 1989, Medina et al., 1986, 1991). Therefore, the stable carbon isotope ratios
of the H. carnosa plants were predicted to be substantially lower than values for most CAM
plants, especially those growing in more exposed locations (for a study using a similar rationale
56 CHAPTER 4 and approach, see Treseder et al., 1995). Confirmation of this prediction would lend support
to the feasibility of the hypothesis about the evolution of CAM in epiphytes stated above.
Despite this expectation, the stable carbon isotope ratios of the H. carnosa epiphytes
(Fig. 4) were not substantially more negative than values typical of many terrestrial CAM plants
that grow in exposed habitats with few neighbors, e.g., desert succulents (Troughton et al.,
1974, Eickmeier and Bender 1976, Sutton et al., 1976, Ting 1989). In addition, laboratory
measurements of gas exchange using plants of H. carnosa collected at the study site revealed
substantial amounts of daytime CO2 uptake (Fig. 5), which would result in carbon isotope
ratios more negative than those of CAM plants lacking daytime CO2 uptake (Winter and
Holtum 2002). Furthermore, although nighttime CO2 concentrations were higher in the closed
canopies, the stable carbon isotope ratios of the epiphytes in these canopies were not more
negative than the values for the epiphytes in the open canopies. These findings indicate that H.
carnosa does not utilize CO2 respired by the host tree canopy, at least to a substantial degree, in
spite of the elevated CO2 concentrations in the canopy at night relative to during the day.
Although a different species, the relatively high (less negative) carbon isotope value reported
for H. nicholsoniae in a dense forest canopy (see above; Winter et al., 1986) also indicates a
minimal contribution of host-respired CO2 to the carbon budget of this epiphytic CAM plant.
It is surprising that host-respired CO2 does not contribute more to the carbon composition of
these epiphytic CAM plants. Although the weather was calm, and canopy air turbulence was
not observed during the measurements made in this study, it is possible that the canopy air is
typically less stagnant than this, mixing frequently with the atmosphere above the forest
canopy and thereby diluting the contribution of host-respired CO2 to the canopy air. Canopy
air turbulence measurements throughout the year are needed in order to test this explanation.
The results of this study do not lend support to the hypothesis that CAM might have
evolved in tropical/subtropical epiphytes in response to atmospheric CO2 availability. Of
course, different results might be obtained with other species of epiphytic CAM plants in other
forests. The potential for the utilization of substantial amounts of host-respired CO2 in tree
canopies should be investigated in more epiphytic CAM plants before definitive conclusions
are drawn. At least in the case of H. carnosa in this subtropical rain forest, the likely
ecophysiological significance of CAM remains water conservation, which should prove
beneficial to such plants despite the infrequent occurrence of drought in this rain forest.
57 0.8
-1
-1
Net CO2 exchange [nmol g s ]
CAM EVOLUTION IN EPIPHYTES 0.6
0.4
0.2
0.0
10
14
22
18
2
6
Tim e of day
Fig. 5 Net CO2 exchange for a shoot of an individual of Hoya carnosa collected in a subtropical rain forest in northeastern Taiwan, then grown and measured in Kansas. The thick, horizontal black line indicates the nighttime. Environmental conditions during measurements are provided in “Materials and methods.” 58 59 Chapter 5
Comparative photosynthetic capacity of abaxial
and adaxial leaf sides as related to exposure in an epiphytic fern in a subtropical rainforest in northeastern Taiwan
The Asplenium bird’s nest fern often forms a conspicuous layer in the forest canopy. Among three species in Taiwan, A. antiquum is usually found coexisting with A. nidus. The two species are morphologically very similar, but can be distinguished by the width of the scales on their leaf bases.
A. nidus
A. antiguum
CHAPTER 5
Comparative photosynthetic capacity of abaxial and
adaxial leaf sides as related to exposure in an
epiphytic fern in a subtropical rainforest in
northeastern Taiwan
Craig E. Martin, Rebecca C.-C. Hsu & Teng-Chiu Lin
American Fern Journal 99, no. 3 (2009): 145-154
Abstract
Photosynthetic gas exchange was measured in situ with either the adaxial or abaxial leaf surface
illuminated on vertical, horizontal, and angled leaves of Asplenium nidus, an epiphytic ferns in a
subtropical rain forest in northeastern Taiwan. Leaves for gas exchange measurements were
selected to ensure a diversity of different exposures of the two leaf surfaces to direct sunlight.
For most leaves of Asplenium nidus, photosynthetic rates were higher when the side of the leaf
that typically received more direct insolation was illuminated during the gas exchange
measurement. Higher rates of net CO2 uptake when one side of the leaf was illuminated,
relative to rates when the opposite side was illuminated, were attributable to a greater
biochemical capacity for photosynthesis, not to greater stomatal conductances. Based on the
results of this study, the photosynthetic capacity of the two sides of the leaves of epiphytic
ferns, for the most part, reflects the degree of exposure of each side of the leaf to direct
sunlight, as has been found in similar studies of terrestrial taxa.
61 PHOTOSYNTHETIC PLASTICITY IN EPIPHYTES Introduction
Most leaves are green and, thus, presumably capable of some level of photosynthetic activity,
even if just recycling respiratory CO2, on both their adaxial and abaxial surfaces (Moore et al.,
1998; Terashima, 1986). Work with terrestrial taxa has shown that the capacity for
photosynthesis is equal, or nearly so, when either leaf surface of vertically oriented leaves is
illuminated, as long as both surfaces intercept similar amounts of solar radiation during leaf
development (Syvertsen and Cunningham, 1979; DeLucia et al., 1991; Poulson and DeLucia,
1993). In contrast, if one side of a vertically oriented leaf typically receives more insolation
than the opposite side, the photosynthetic capacity of the leaf is greater when the normally
sunlit surface is irradiated during photosynthetic measurements, relative to photosynthesis
when the shaded side is irradiated (Poulson and DeLucia, 1993; but see Václavík, 1984)
Likewise, the photosynthetic activity of horizontally oriented leaves is greater when their
adaxial surface is illuminated than when their abaxial surface is illuminated (Syvertson and
Cunningham, 1979; Terashima, 1986; DeLucia et al., 1991) The latter applies only to the sun
leaves, not the shade leaves, of Sitka spruce (Leverenz and Jarvis, 1979).
Epiphytic vascular plants appear to have been excluded from such studies, yet are
ideal subjects for such investigations. Epiphytic vascular plants often exhibit a great diversity of
leaf orientations and exposures (Benzing, 1990). For example, epiphytes with a rosette growth
form often have leaves ranging from vertical to horizontal, and many have intermediate angles.
Most epiphytes also live in a complex light environment, being shaded by the host tree stem
and canopy, as well as surrounding trees, depending on the location of the sun at any point in
time. Given their leaf angles and the complexity of the light environment in which epiphytes
grow, it is difficult to predict how the photosynthetic capacity of the two sides of the leaves of
such plants compare and whether or not findings based on terrestrial taxa might apply to
epiphytes. Therefore, the goal of this study was to determine if photosynthesis in epiphytes,
particularly ferns, responds to leaf surface illumination in a similar manner as has been found
in terrestrial plants.
Materials and methods
Study site and species
Leaf photosynthetic parameters were measured for six individuals of Asplenium nidus L. in situ at
the Fushan Experimental Forest, a comparatively pristine tract of subtropical rainforest
(121°34′E, 24°46′N) at an elevation of ca. 600 m located 40 km southeast of Taipei in
northeastern Taiwan. For general climatic conditions at the Fushan site, see Martin et al. (2004).
Environmental conditions during the week of measurements (11-15 July 2005) were: 25.1° C
62 CHAPTER 5 average daily air temperature (29.8° C average daily maximum; 21.3° C average daily minimum),
4.2 mbar average daily vapor pressure deficit (vpd); and 20.0 mol m-2 day-1 average daily
Photosynthetic Photon Flux Density (PPFD).
Asplenium nidus was chosen for this investigation to ensure a diversity of different
exposures of the two leaf surfaces to direct sunlight. Plants were selected in a partially
disturbed section of the forest to allow easy access for in situ measurements of photosynthesis.
The study site included several walking trails and was tens of meters from a laboratory building.
Species of dominant trees at this site were numerous, primarily in the families Fagaceae and
Lauraceae; examples include Litsea acuminata (Bl.) Kurata (Lauraceae), Machilus zuihoensis Hayata
(Lauraceae), Castanopsis cuspidata (Thunb. ex Murray) Schottky var. carlesii (Hemsl.) Yamazaki
(Fagaceae) and Pasania hancei (Benth.) Schottky (Fagaceae).
All plants were large (plant diameter for A. nidus ≥ 0.5m) growing epiphytically on a
variety of host trees, including those listed above. Most plants had sporangia on some leaves at
the time of this study (sporangia-bearing portions of the leaves were avoided in all
measurements to avoid effects of sporangia on the measurements (Chiou et al. 2004). All leaves
were measured no higher than three to four meters from the ground, i.e., within arm’s reach
while standing, with or without a ladder. Only mature, non-senescent leaves lacking substantial
insect damage were sampled; very young and very old leaves were avoided. Leaves were
selected without regard to host tree species, height from the ground (except as noted), and
degree of canopy shade at the time of measurements.
Photosynthesis measurements
Photosynthesis was measured on three different leaves for each of six plants of A. nidus; the
three leaves were selected for measurements based primarily on the likelihood of exposure of
each leaf surface to direct sunlight. Horizontal leaves were older (based on size, presence of
sporangia, weathering, and phyllotaxy of the epiphyte) than the other two leaves selected for
measurements and grew perpendicular to and away from the host tree trunk. Such leaves
should intercept very little direct sunlight on their abaxial surface, whereas their adaxial surface
should intercept direct sunlight during much of a sunny day. Angled leaves grew at about a 45
degree angle from the tree trunk, so should occasionally intercept direct sunlight on both
surfaces of the leaf. Vertical leaves grew close to the trunk of the host tree, and, thus, were
shaded by the trunk much of the day. These leaves should intercept little light on their
outward-facing adaxial surface most of the day, but occasionally direct sunlight on their abaxial
surface, depending on the location of the sun.
Photosynthesis was measured with a LI-COR (Lincoln, NE) LI-6400 Portable
Photosynthesis System. Because all leaves measured were large, the area of leaf for which gas
63 PHOTOSYNTHETIC PLASTICITY IN EPIPHYTES exchange was measured matched the maximum area possible (6 cm2) in the gas exchange
chamber. Photosynthetic parameters were measured two different ways at the central portion
of each leaf: once with the adaxial surface illuminated and again adjacent to the same leaf
location with the abaxial surface illuminated. The exact same location on the leaf was not used
for both measurements to ensure that manipulation by inserting the leaf into the chamber and
clamping the chamber on the leaf for the first measurement did not influence the second
measurement. Although gas exchange was always measured for both sides of the leaf
simultaneously, the chamber was oriented such that only the adaxial or abaxial surface received
light from the blue and red diodes in the top half of the chamber. Very little ambient light
reached the opposing leaf surface during the measurements as a result of shading by parts of
the gas exchange chamber, the investigators, and nearby vegetation. Photosynthesis was
measured three times with illumination on one surface of a leaf at a low PPFD (100 µmol m-2
s-1), then three times at a high PPFD (1000 µmol m-2 s-1) Using the same leaf, the chamber was
then reversed to measure gas exchange with illumination (both PPFD levels) on the opposite
leaf surface. Net CO2 uptake in A. nidus saturated at approximately 500 µmol m-2 s-1
(determined with preliminary gas exchange measurements). Other environmental conditions
during all measurements were maintained by the LI-6400 system at the following values: air
CO2 concentration of 370 µmol mol-1, chamber (“block”) temperature of 30°C (leaf
temperatures were typically 0.5° C higher), vapor pressure deficit (vpd) of 0.9 mbar, and flow
rate of 200 μmol s-1. Lower vpd values resulted in exceedingly low transpiration rates, which
led to unrealistic values for Ci; any such data were discarded. For each gas exchange
measurement, data were recorded only when gas exchange rates were stable (Coefficient of
Variation of exchange rates of both gases and flow rates not varying by more than 0.2%
among successive measurements every 2-3 seconds), typically within 10 seconds of inserting
the leaf in the gas exchange chamber or after the previous measurement (for a total of three
repeated measurements). The gas exchange chamber remained clamped to a leaf for
approximately five minutes at each light level, allowing for stable readings, as well as steps
taken to ensure instrument accuracy (e.g., using the “match” function of the LI-6400 prior to
each measurement).
Statistical analyses
Means of gas exchange parameters (N=5; the value for each plant being a mean of three repeat
measurements; see above) for abaxial and adaxial surfaces at each light level were compared
with a paired Student’s t-test when the data met the assumptions of parametric statistics (Sokal
and Rohlf, 1981) or with a Mann-Whitney U-test otherwise.
64 CHAPTER 5 Results and discussion
The adaxial side of the vertical leaves growing out of the rosettes of A. nidus is unlikely to
receive direct radiation due to shading by the host tree trunk, whereas the exposed abaxial side
should at least occasionally intercept direct solar radiation. Thus, based on results with
terrestrial plants (Syvertson and Cunningham, 1979; Terashima, 1989; DeLucia et al. 1991;
Poulson and DeLucia 1993), it was predicted that the illumination of the abaxial side of the
vertical leaves of A. nidus would result in higher photosynthetic rates than when the adaxial
side of the same leaf is illuminated. Measurements of photosynthesis at both high and low
PPFD of plants in northeastern Taiwan did not, however, support this prediction (Fig. 1). In
contrast, although not statistically significant (high PPFD P= 0.28; low PPFD P = 0.17), the
trend in the data indicated the opposite of expectations, i.e., photosynthetic rates at either
PPFD appeared higher when the adaxial surface was illuminated. According to the statistical
analyses, however, photosynthetic rates at both light levels were equal regardless of which side
of the leaf was illuminated (Fig. 1).
Fig. 1 Mean (lines projecting from bars are standard deviations; n = 6 plants, three repeated measurements per leaf for each plant) rates of net CO2 exchange (positive values indicate CO2 uptake) for different leaves and with illumination at two light levels on either side of the leaves of the epiphytic fern Asplenium nidus measured in situ in a subtropical rain forest in northeastern Taiwan). Abbreviations for type and side of leaf are: “VR” = vertical, “HZ” = horizontal, “AN” = angled (45° from vertical); and “AD” indicates illumination (A, 100 µmol m‐2 s‐1; B, 1000 µmol m‐2 s‐1) provided to the adaxial side of the leaf during gas exchange measurements; “AB” indicates illumination (low and high PPFD as in AD) provided to the abaxial side of the leaf during measurements. The abaxial and adaxial means for two leaves at low PPFD are significantly different at P < 0.05 or P < 0.01 indicated by “*” or “**”, respectively, above each pair of means, while the other pairs of means are not significantly different (P > 0.05, indicated by “ns” above each pair of means). 65 0.4
ns
ns
ns
A
B
ns
0.3
ns
4
3
ns
0.2
2
0.1
1
0
0.0
VRAD VRAB HZAD HZAB ANAD ANAB
VRAD VRAB HZAD HZAB ANAD ANAB
Leaf Type & Side
Leaf Type & Side
Net H2O Exchange, mmol m-2 s-1
Net H2O Exchange, mmol m-2 s-1
PHOTOSYNTHETIC PLASTICITY IN EPIPHYTES Fig. 2 Mean (lines projecting from bars are standard deviations; n = 6 plants, three repeated measurements per leaf for each plant) rates of net H2O exchange (positive values indicate water vapor loss) for different leaves and with illumination at two light levels on either side of the leaves of the epiphytic fern Asplenium nidus measured in situ in a subtropical rain forest in northeastern Taiwan). Abbreviations for type and side of leaf are: “VR” = vertical, “HZ” = horizontal, “AN” = ‐2 ‐1
‐2 ‐1
angled (45° from vertical); and “AD” indicates illumination (A, 100 µmol m s ; B, 1000 µmol m s ) provided to the adaxial side of the leaf during gas exchange measurements; “AB” indicates illumination (low and high PPFD as in AD) provided to the abaxial side of the leaf during measurements. None of the abaxial and adaxial means at any leaf location are significantly different (P > 0.05, indicated by “ns” above each pair of means). Light interception of the two surfaces of the horizontal leaves of the epiphytic fern A.
nidus is quite different from that of the vertical leaves, and the prediction of comparative
photosynthetic capacities when the two sides of this leaf are illuminated is the opposite of that
of the vertical leaves of this fern. Because the adaxial surfaces of these leaves intercept more
direct solar radiation than do the abaxial surfaces, photosynthetic rates when the adaxial
surface of the horizontal leaves of this epiphyte are illuminated should be higher than those of
the leaf when the abaxial surface of the same leaf is illuminated. Measurements of
photosynthetic rates confirmed this prediction, although the higher photosynthetic rates when
the adaxial side of the leaves was illuminated, were statistically significant only when
measurements were made at the lower PPFD (Fig. 1). These higher net CO2 uptake rates were
accompanied by equal transpiration rates (Fig. 2) and stomatal conductances (Fig. 3), while
internal CO2 concentrations were significantly lower (Fig. 4). These gas exchange results
indicate that the higher photosynthetic rate was most likely the result of a greater biochemical
capacity for photosynthesis and not the result of greater stomatal opening and, hence, easier
gas diffusion into the leaf (Farquhar and Sharkey, 1982; Sharkey, 1985). In agreement with the
latter interpretation, it is possible, especially for the measurements made at high light, that
66 CHAPTER 5 illumination of the abaxial surface resulted in photoinhibition in this lateral half of the section
of leaf being measured. This possibility is supported by previous findings that the side of a leaf
that is typically less exposed to sunlight has chloroplasts and photosynthetic features typical of
shade-adapted leaves (Schreiber et al., 1977; Kulandaivelu et al., 1983; Terashima and Inoue,
1984; Terashima et al., 1986). Differences in photosynthetic capacity depending on which side
of the leaf is illuminated might also reflect other anatomical or optical (e.g., absorptance)
features of the two sides of the leaf (Terashima 1986; DeLucia et al., 1991). Such differences
would also be interpreted as non-stomatal and non-diffusional mechanisms responsible for
differences in photosynthesis between the two sides of the leaf, as was found in this study.
0.4
ns
ns
ns
A
0.3
B
*
ns
*
4
3
0.2
2
0.1
1
0.0
Conductance, mol m-2 s-1
Conductance, mol m-2 s-1
0
VRAD VRAB HZAD HZAB ANAD ANAB
VRAD VRAB HZAD HZAB ANAD ANAB
Leaf Type & Side
Leaf Type & Side
Fig. 3 Mean (lines projecting from bars are standard deviations; N = 6 plants, three repeated measurements per leaf for each plant) stomatal conductances for different leaves and with illumination at two light levels on either side of the leaves of the epiphytic fern Asplenium nidus measured in situ in a subtropical rain forest in northeastern Taiwan). Abbreviations for type and side of leaf are: “VR” = vertical, “HZ” = horizontal, “AN” = angled (45° from vertical); and “AD” indicates illumination (A,100 µmol m‐2 s‐1; B, 1000 µmol m‐2 s‐1 in Fig. 3B) provided to the adaxial side of the leaf during gas exchange measurements; “AB” indicates illumination (low and high PPFD as in AD) provided to the abaxial side of the leaf during measurements. The abaxial and adaxial means at two leaf locations at high PPFD are significantly different at P < 0.05, indicated by “*” above each pair of means, while the other pairs of means are not significantly different (P > 0.05, indicated by “ns” above each pair of means). 67 PHOTOSYNTHETIC PLASTICITY IN EPIPHYTES Fig. 4 Mean (lines projecting from bars are standard deviations; N = 6 plants, three repeated measurements per leaf for each plant) leaf internal CO2 concentrations (external CO2 concentration ‐1
was 370 µmol mol ) for different leaves and with illumination at two light levels on either side of the leaves of the epiphytic fern Asplenium nidus measured in situ in a subtropical rain forest in northeastern Taiwan). Abbreviations for type and side of leaf are: “VR” = vertical, “HZ” = horizontal, ‐2 ‐1
“AN” = angled (45° from vertical); and “AD” indicates illumination (A, 100 µmol m s ; B,1000 µmol ‐2 ‐1
m s ) provided to the adaxial side of the leaf during gas exchange measurements; “AB” indicates illumination (low and high PPFD as in AD) provided to the abaxial side of the leaf during measurements. The abaxial and adaxial means at several leaf locations are significantly different at P < 0.05 or P < 0.01, indicated by “*” or “**”, respectively, above each pair of means, while the other pairs of means are not significantly different (P > 0.05, indicated by “ns” above each pair of means). Both the adaxial and abaxial surfaces of the “angled” leaves of A. nidus should
intercept direct sunlight, at least for brief periods, throughout a day. Thus, one might predict
that the photosynthetic capacity of these leaves is comparable, regardless which surface is
illuminated (Syvertsen and Cunningham, 1979; Václavík, 1984; DeLucia et al., 1991; Poulson
and DeLucia, 1993). Based on measurements made in situ with this epiphytic fern in
northeastern Taiwan, this prediction was supported when gas exchange was measured at high
PPFD (Fig. 1), but the photosynthetic rate when the adaxial leaf surface was illuminated
exceeded that when the abaxial surface of the leaf was illuminated at low PPFD (Fig. 1). As
was the case with the horizontal leaves, the higher net CO2 exchange rate of the angled leaves
was apparently the result of a greater biochemical capacity for photosynthesis, generating a
lower leaf internal CO2 concentration (Fig. 4), and not due to a greater stomatal conductance
(Fig. 3; Farquhar and Sharkey, 1982; Sharkey, 1985). These findings contrast directly with
those for Sitka spruce by Leverenz and Jarvis (1979), who found that differences in
photosynthetic capacity of the leaves, depending on which side of the leaf was illuminated
could be ascribed to differences in stomatal conductance, not to the biochemical capacity of
the leaf.
68 CHAPTER 5 Overall, the results of in situ gas exchange measurements with A. nidus in a subtropical
rain forest in northeastern Taiwan lend considerable, but not complete, support to past
findings with terrestrial taxa (Syvertsen and Cunningham, 1979; Terashima, 1989; DeLucia et al.,
1991; Poulson and DeLucia, 1993). In most, but not all, cases, if a leaf is oriented such that
one side receives more direct solar radiation than the other, the leaf has a higher
photosynthetic capacity when the more exposed surface is illuminated. In addition, this higher
capacity reflects a greater biochemical capacity for photosynthesis and not easier diffusion of
CO2 into the leaf (Farquhar and Sharkey 1982; Sharkey, 1985).
69 Chapter 6
Adaptation of a widespread epiphytic fern to simulated climate‐change conditions
A. antiquum sporelings were transplanted to each of the sites where the spores were collected. Young ferns were planted in paper tea bags filled with peat moss, which were fixed to a 40 × 50 cm coconut mat. The mats, 10 per site, were nailed to a tree trunk at eye‐level height.
CHAPTER 6
Adaptation of a widespread epiphytic fern to
simulated climate-change conditions
Rebecca C.-C. Hsu, J. Gerard B. Oostermeijer & Jan H.D. Wolf
Abstract
The response of species to climate change is generally studied using ex situ manipulation of
microclimate or by modeling species range shifts under simulated climate scenarios. In
contrast, a reciprocal transplant experiment was used to investigate the in situ adaptive response
of the elevationally widespread epiphytic fern Asplenium antiquum to simulated climate change
conditions. Fern spores were collected at three elevations and germinated in a greenhouse. The
sporelings (juvenile ferns) were reciprocally transplanted to each collection site. Growth and
mortality rates were monitored for two years. Wild sporelings were monitored at two sites to
assess possible transplant effects. Habitat suitability, indicated by overall growth and survival
patterns, declined as elevation increased. Only the highland population showed significant
adaptation to the ‘home’ habitat, achieving the highest survival rates. Microclimate data suggests
that the presumed genetic adaptation at the highland site occurred mainly in response to
drought stress in winter. Based on our previous study on species distribution models, which
projected an expansion in the range of A. antiquum under future climate change scenarios, the
populations at the upper margins of the species’ elevational range may play an important role
during this expansion, given their improved adaptation to the shifting marginal conditions. The
study suggests that intraspecific variation should be considered when establishing the potential
impact of climate change on biodiversity.
Introduction
The response of species to global climate change is of great interest in conservation biology.
Vulnerability to climate change differs among biomes and is related to the ecological and
genetic properties of species (Root et al., 2003; Broennimann et al., 2006; Loarie et al., 2009).
Under changing climate conditions, inferred responses include phenotypic plasticity, genetic
adaptation and migration, and the most dramatic consequence is species extinction. Phenotypic
plasticity is the ability of a genotype to exhibit variable phenotypes in response to
environmental change, whereas genetic adaptation is an evolutionary process, during which
selection favours individuals with novel gene and allele combinations that either arise by sexual
recombination or by immigration from other populations (Nicotra et al., 2010). Adaptation and
71 EPIPHYTE ADAPTATION TO CLIMATE CHANGE phenotypic plasticity contribute to the ecological amplitude of a species if populations occur in
diverse habitats, and may permit populations to persist in spatially and temporally
heterogeneous environments (Silander 1985). Besides the capacity for dispersal and
establishment, the extent of adaptation may well determine the vulnerability of a species to
climate change (Hedderson and Longton 2008).
Widespread species (generalists) occur across a broad range of environmental
gradients and thus usually comprise several climatically-adapted populations (ecotypes). Hence,
generalist species are likely to demonstrate broader tolerances to climate change than specialists
that are geographically restricted (Broennimann et al., 2006; Aitken et al., 2008). The
identification of ecotypes of economically important plants that are genetically adapted to
future climates is viewed as a promising strategy for sustainable agriculture in the face of
climate change (Kreyling et al., 2012). The favored traits under climate changes could be
identified by comparing selection regimes in current environments to those in environments
similar to predicted future conditions, (Etterson 2004). However, climate change impacts on
species adaptability have been explored for only very few species. Therefore, population
studies on the genetic adaptation and phenotypic plasticity of species in relation to climate
change deserve special attention.
Epiphytes are presumed to be particularly sensitive to climate change since they have
no vascular connection to the ground or their host plants. They solely rely on the contact with
rain or cloud droplets for moisture input, hence respond rapidly to slight changes in ambient
climate (Benzing 1998; Zotz and Bader 2009). Contrary to common expectation, our recent
study (Hsu et al., 2012) using species distribution modeling (SDM) suggested that several
species (e.g. Asplenium antiquum) would expand its range size under future A2 and B2 climate
change scenarios (Nakicenovic et al., 2000). However, a general shortcoming of SDM is that it
does not consider the possibility of intraspecific variation of the modeled species along climate
gradients (Benito Garzón et al., 2011). Studies suggested when sub-taxon information (i.e.
subspecies) was incorporated into SDM, the species was projected to better tolerate climate
change (Pearman et al., 2010; Oney et al., 2013).
Reciprocal transplant experiments were previously conducted, using epiphyte mats
(i.e. the combined unit of living epiphytic plants and associated detrital matter), to evaluate
epiphyte sensitivity to manipulated climate-change conditions (Nadkarni and Solano 2002;
Song et al., 2012). In this study, we reciprocally transplanted juveniles of a single species,
Asplenium antiquum Makino, to three different elevations. The main aim of this study is to
evaluate if altitudinally separated populations of A. antiquum are adapted to their local
environment.
72 CHAPTER 6 Fig. 1 Locations of the three study sites within Taiwan. Circle: low elevation site at 600 m asl; diamond: mid elevation site at 1100 m asl; triangle: high elevation site at 1950 m asl. 73 EPIPHYTE ADAPTATION TO CLIMATE CHANGE Materials and Methods
Study species
A. antiquum is a widespread epiphytic fern native to China, Japan, Korea and Taiwan. Its
common name, ‘bird’s nest fern’, is derived from its rosette growth form, which traps fallen
leaves and other debris. Adult plants may reach 300 cm in diameter (pers. observation). The
clumped plant bases are composed of fibrous roots and trapped humus, which sponge up
rainwater to facilitate successful establishment in the forest canopy. A. antiquum is the
elevationally most widespread species, ranging from the coast to up to c. 2500 m above sea
level [asl], among the three species of bird’s nest ferns in Taiwan.
Study sites
We selected three remote sites in primary broad-leaved forests with thriving populations
comprising several hundreds of adult A. antiquum plants for spore collection and the reciprocal
transplant experiment. The sites were located at Fushan (lowland, 600 m asl), SiangBenShan
(midland, 1100 m asl) and PeiTungYenShan (highland, 1950 m asl) (Fig. 1). The horizontal
distances between the lowland and the midland and between the midland and the highland
sites were 39 km and 50 km, respectively. The lowland and midland sites are in north-eastern
Taiwan and dominated by Lauraceae trees (e.g. Machilus zuihoensis, Litsea acuminata, Machilus
japonica, Phoebe formosana) with an average height of 15 m. The highland site is situated at the
west side of the island central ridge, with a higher canopy (c. 20 m, dominant trees: Schima
superba and Castanopsis carlesii) than the two lower sites.
Climate measurements
The average annual rainfall recorded at the lowland, midland and highland sites is c. 3500, 3800
and 2500 mm, respectively (Central Weather Bureau). At each site we placed two data loggers
(model U23-001, HOBO Pro V2 Temp/RH Data logger, Onset computer corporation,
Bourne, MA, USA), and one visibility meter (model MiniOFS, Sten Löfving Optical Sensors,
Göteborg, Sweden) to record local temperature and relative humidity per hour and fog events
every 30 minutes during the course of the experiment. In addition, we recorded the phenology
of three adult A. antiquum individuals per site at monthly intervals.
74 CHAPTER 6 Fig. 2 Monthly mean temperature recorded at three elevation sites from Dec 2008 to Dec 2010. Bars indicate the monthly maximum and minimum temperatures being recorded. L: low elevation site (600m asl); M: mid elevation site (1100m asl); H: high elevation site (1950 m asl). Reciprocal transplant experiment
In November 2007, fertile leaves of ten A. antiquum individuals were sampled haphazardly with
a 5-m branch cutter from each of the three elevational sites and air-dried to collect the spores.
In January 2008, spores were bulked and germinated on sterilized soil in covered plastic boxes.
After six months, sporelings were replanted on Sphagnum peat substrate. In December 2008,
one-year-old sporelings were transplanted to each of the sites from which the spores were
collected, coinciding with the beginning of the north-eastern monsoon that brings rainwater to
help establish the plants. Nevertheless, all sporelings transplanted at the highland site died in
the first month due to very low moisture levels. Therefore, a second batch of sporelings from
the same bulked spore sample that had been germinated as backup in July 2008 was retransplanted in July 2009 to the highland site. Young ferns were planted in paper tea bags filled
with peat moss, which were fixed to a 40 × 50 cm coconut mat. The mats, 10 per site, were
nailed to a tree trunk at eye-level height, with 15 bags containing one plant each (i.e. five plants
per altitudinal origin). All transplants were randomly allocated a place on the mat; in total, there
were 50 sporelings per origin per location. The diameter and mortality of the transplanted
sporelings were recorded in the field each month. Dead and missing (due to animals, wind and
heavy rain) plants were tallied separately, based on their health condition at previous month’s
visit. Wild sporelings were also monitored at the high and low elevation sites that are relatively
environmental distinctive to assess possible transplant effects. A batch of sporelings planted in
75 EPIPHYTE ADAPTATION TO CLIMATE CHANGE pots was kept in a nursery. At the end of the experiment, their diameter was measured before
they were dried in an oven at 65 °C to assess the correlation between diameter and biomass.
Data analysis
A regression analysis was performed to test the correlation between the rosette diameter and
the dry weight (biomass) of the sporelings. The growth of local and foreign sporelings at each
site was compared using the relative growth rate, RGR (Hunt 1982), calculated using the
following equation:
RGR = (ln(D2) – ln(D1))/(t2 – t1)
in which D1 and D2 are plant diameters (mm) at times t1 and t2 (days).
Overall differences in mean RGR (i.e. averaged RGR for each individual sporeling between
visits) and final sporeling size among populations were tested with two-way ANOVA, using
initial size as covariate, to test effects of elevation, sporeling origin and the elevation-by-origin
interaction. The smaller second batch of transplanted sporelings at the high elevation site was
excluded from the final size comparison among sites. Mean sporelings’ RGR between visits
were correlated with local microclimates (i.e. mean temperature and relative humidity). Because
temperature and humidity were not independent, we calculated partial correlation coefficients,
controlling for one or the other variable. We tested the success of sporeling settlement at the
three elevations using Kaplan-Meier survival analysis (Kaplan and Meier 1958), and compared
the survival curves of populations among and within sites with a log rank test (Bland and
Douglas 2004). All analyses were performed using SPSS (version 13.0, IBM).
Results
Climate at study sites
Our climate dataloggers showed that mean temperatures decreased from the lowland to the
midland and highland site, having an annual mean temperature (± SE.) of 17.3 (4.27), 15.7
(4.35) and 13.1 (3.43) °C, respectively (Fig. 2). Unexpectedly, during the course of the
experiment, the temperature occasionally dropped below zero at the lowland site. The annual
mean diurnal temperature (i.e. the difference between daily maximum and minimum
temperature) ranged between 5.79 (1.17), 3.87 (0.2) and 5.57 (0.9) °C from the lowland to the
highland site, respectively. Monthly mean air humidity significantly decreased from the lowland
to the highland site (Fig. 3, ANOVA, p < 0.001). During the experimental period, the lowland
and midland sites were relatively dry in late spring (May), whilst the highland was quite dry in
the winter (Fig. 3). The frequency of mist (1 km < visibility < 2 km) and fog (visibility < 1 km)
76 CHAPTER 6 events also varied between the three sites. Foggy conditions were most frequent in the midland
site (1667 hrs/ year) in comparison to the lowland (116 hrs/ year) and highland (754 hrs/ year)
sites. In contrast to regular afternoon fogs at mid and high elevation sites, morning mist was
relatively common at the low elevation site, a typical characteristic of tropical lowland cloud
forest (Gehrig-Downie et al., 2011).
Table 2 Results of partial correlation analysis between the RGR of A. antiquum sporelings and local microclimate (mean temperature and relative humidity of between‐visit duration) at three transplant sites from Dec 2008 to Dec 2010. For comparison, data were also pooled to calculate (a) single and (b) partial correlation coefficients. T = mean temperature, RH = relative humidity; low elevation: 600 m asl, mid elevation: 1100 m asl and high elevation: 1950 m asl. Numbers in brackets indicate p values. Correlated (Controlling) factors T (RH) RH (T) Correlation coefficient Source Lowland origin Site pooled (a) Sites pooled (b) Low elevation Mid elevation High elevation Site (a)
pooled Site (b)
pooled Low elevation Mid elevation High elevation Midland origin Highland origin 0.265 (0.094) 0.406 (0.008)*** 0.265 (0.094) 0.183 (0.266) 0.247 (0.124) 0.144 (0.377) 0.723 (0.005)*** 0.506 (0.078) 0.527 (0.064)* 0.435 (0.137) 0.68 (0.015)** 0.223 (0.464) ‐0.114 (0.724) ‐0.502 (0.096) 0.252 (0.113) 0.492 (0.001)*** 0.333 (0.033)** 0.132 (0.424) 0.385 (0.014)** 0.251 (0.118) 0.229 (0.452) ‐0.536 (0.059)* 0.513 (0.073)* ‐0.072 (0.814) 0.568 (0.043)** 0.631 (0.021)** 0.15 (0.642) 0.604 (0.038)** 0.222 (0.488) Codes for significance: *** p < 0.01, ** p < 0.05, * p < 0.1. 77 ‐0.201 (0.53) EPIPHYTE ADAPTATION TO CLIMATE CHANGE Fig. 3 Monthly mean air humidity at three elevation sites from Dec 2008 to Dec 2010. L: low elevation site (600m asl); M: mid elevation site (1100m asl); H: high elevation site (1950 m asl). Plant phenology, growth and survival
At the lowland and midland sites, the monitored adult ferns produced new leaves in early
spring (February to March), and there was a second budding in autumn (September), yet the
plants produced no sporangia. At the highland site, new leaves appeared only once a year,
during the spring rain period in April. Although, sporeling growth estimated as diameter
increase varied greatly among different origin sites, the diameter of the sporelings was
significantly correlated with their dry weight, for sporelings of the same origin as well as for all
sporelings combined (Fig. 6).
After correcting for a significant effect of initial sporeling size (i.e. larger transplants
grew faster), the average RGR was significantly lower at the highest elevation (F(2, 78) = 22.1, p
< 0.001, Table 1). The origin of sporelings had no significant effect on the average RGR, but
there was an elevation-by-origin interaction (F(4,78) = 3.11, p = 0.02). Separate analyses for each
elevation showed a significant origin effect on RGR (F(2,49) = 3.98, p = 0.025) only at the low
elevation site (Table 1). Sporelings’ average final sizes were 57.26 (27.5), 58.81 (33.04) and
19.02 (9.30) mm (± SE.) at low, mid and high elevations, respectively (Fig. 5). Between the low
and mid elevation sites, sporelings’ final sizes were similar (F(1, 69) = 0.259, p = 0.612, Table 1).
However, the second batch of transplant sporelings at the high elevation site (i.e. a half year
78 CHAPTER 6 younger) were significantly smaller than those at the other two sites (F(2,78) = 7.468, p = 0.001).
Moreover, the origin of sporelings only had a marginally significant effect on the final size at
the low elevation (F(2,78) = 2.68, p = 0.079). This was mainly caused by a lower RGR (Fig. 4)
and smaller size (Fig. 5) of the highland sporelings. The wild sporelings generally had the same
growth pattern with transplanted plants (Fig. 5), suggesting the external (climatic)
environmental exert more influence on growth than transplant effect.
Fig. 4 Comparison of sporelings' mean RGR from Dec 2008 to Dec 2010 among three elevation sites. Bars indicate ± SE. Different letters (a, b, c) indicate significant difference at p < 0.05 (ANOVA). L: lowland sporelings (circle); M: midland sporelings (diamond); H: highland sporelings (triangle). 79 EPIPHYTE ADAPTATION TO CLIMATE CHANGE Fig. 5 Growth (represented by average rosette diameter) of transplanted and wild local sporelings of A. antiquum from Dec 2008 to Dec 2010 at three elevation sites. Wild sporelings were monitored for comparative purposes at the low and high elevation sites. Bars indicate ± SE. L: lowland sporelings; M: midland sporelings; H: highland sporelings; W: wild sporelings. 80 CHAPTER 6 Fig. 6 Regression analysis of the biomass‐diameter relationship of A. antiquum sporelings cultivated in a nursery for two years. The exponential equation is fitted for samples originating from three elevation sites. L: lowland sporelings (circle); M: midland sporelings (diamond); H: highland sporelings (triangle). The results suggested that temperature explained most of the variation in RGR at the
low-elevation site, whereas the RGR was correlated mainly with relative humidity at both the
mid and high elevation sites (Table 2). The RGR of lowland origins were only correlated (p =
0.015) with temperature at the low elevation site. For highland origins, air humidity was
significant correlated with their RGR at the mid elevation site (p = 0.021), and was marginally
correlated at the low elevation site (p = 0.073). The midland origins showed the highest
sensitivity to microclimates among sites, and demonstrated a significant correlation (p = 0.005)
with air temperature at the low elevation site.
81 EPIPHYTE ADAPTATION TO CLIMATE CHANGE Fig. 7 Kaplan‐Meier survival curves of reciprocally transplanted A. antiquum sporelings originating from three elevations, grouped by transplantation sites. L: lowland sporelings (circle); M: midland sporelings (diamond); H: highland sporelings (triangle). Across origins, the survivorship of sporelings was significantly affected by elevation,
and to some extent, by origin within some elevations (Fig. 7). Mean sporeling survival was
highest at the low elevation site (50%), followed by the mid (22.3%) and high elevation (10.3%)
sites (Mantel-Cox log-rank test, Chi-square = 119.1, df = 2, p ≤ 0.001). Across elevations, mean
survival of sporelings from all three origins was largely similar, although marginally better for
highland sporelings (23.8% for lowland, 16.1% for midland and 34% for highland sporelings;
Chi-square = 5.0, df = 2, p = 0.084). At the low elevation, the highland sporelings survived
significantly better (70%) than the lowland (50%) and midland (28%) sporelings (Chi-square =
82 CHAPTER 6 6.4, df = 2, p = 0.040). At the mid elevation site, the survival rates were not significantly
different among origins (lowland: 24.3%, midland: 22.6% and highland: 20.0%; Chi-square =
1.06, df = 2, p = 0.590). At the high elevation, the highland sporelings survived significantly
better (20%) than the midland (2.7%) and lowland sporelings (7.1%) (Chi-square = 16.2, df = 2,
p ≤ 0.001). Hence, an advantage for ‘home’ sporelings only existed at high elevations, but these
highland sporelings also survived better at low elevation, where they grew less and slower than
sporelings from the other two origins (Fig. 4, 5).
Discussion
The climatic differences between the three elevations were largely in accordance with
expectations for wet subtropical mountains (Walter 1985). Average daily temperatures dropped
with elevation. However, we recorded frost (daily minimum temperature < 0°C) not only at
the high elevation site but also at the low elevation site even in March. In comparison with the
mid elevation site, the low elevation site exhibited a relatively high seasonal and diurnal
temperature range. This pattern appears to be induced by variation in local topography and
associated regional climates. The low elevation site, situated in north-eastern island, is
intensively influenced by NE-monsoon in winter. NE-monsoon generally accounts for 45% of
the total annual rainfall in eastern Taiwan (Kao et al., 2004) and occasionally causes frost events
in early spring. For example in 2005, a relatively warm winter followed by a severe frost event
in March led to extensive second budding and re-foliating of plants in northern Taiwan. On
the contrary, the mid elevation site demonstrated less variation in temperature, which probably
can be attributed to its high frequency of fogs and associated reduced thermal radiation, a
characteristic of montane cloud forests worldwide (Jarvis and Mulligan 2011). The lowest mean
relative humidity occurred at the high elevation. Unlike the lowland and midland sites that
receive large amounts of monsoon rainfall in winter, the highland site is only slightly influenced
by the NE-monsoon for its location on the west side of the central ridge. In agreement, we
observed wrinkled fronds of adult A. antiquum plants in winter during the study period.
Moreover, we noticed delayed leaf budding at the highland site that may also be related to
water deficiency, since rapid elongation of fronds requires sufficient water (Freiberg and
Turton 2007). Low water availability at the high elevation site presumably accounted for the
failure of the first transplant experiment in Dec 2008.
It is also likely that the variation in climate between the elevational sites affected the
growth and survivorship of the A. antiquum sporelings. To estimate growth during the course
of the study, we measured the diameter of the plants in a non-destructive way. The significant
diameter-biomass correlation showed that the diameter of A. antiquum rosettes may be used to
measure individual growth. We found that site (elevation) had a significant effect on the final
transplant size, the relative growth rate (RGR) and the survival rate of sporelings. From low to
83 EPIPHYTE ADAPTATION TO CLIMATE CHANGE high elevation, the RGR decreased and the mortality increased. Growth rates were significantly
positively correlated with both temperature and relative humidity, especially for midland
sporelings that originated in the mid elevation site with a relatively stable microclimate. When
controlling for the correlation between these variables, we found that temperature had the
greatest influence on RGR variation at the low elevation site, but at the mid and high elevation
sites, relative humidity had the greatest influence. We postulate that the high mortality at the
high elevation site is better explained by low water availability rather than reduced temperature,
since sporelings at the low elevation site also experienced low temperatures, even frost, but
showed higher survivorship than those at the highland site. Based on differences in final size,
RGR and mortality of sporelings among sites, we conclude that the warm low elevation site
with prolonged moisture availability was the most suitable habitat for A. antiquum sporelings in
this study, whereas the high elevation site appeared to impose a more intense selection
pressure, mostly through drought stress in the winter.
Our study showed that significant differences among elevations occurred in A.
antiquum phenology, sporeling growth and mortality. Although the lack of replication among
elevations (owing to poor site accessibility) does not allow us to conclusively link these
differences to the elevation-specific climate conditions, the patterns observed suggest that such
links exist. At the high elevation site, the higher survivorship of local sporelings as compared
to the foreign ones (from lowland and midland origins) suggests a certain degree of genetic
adaptation, resulting in higher tolerance to drought stress and low temperatures. The highland
sporelings also had higher survivorship than the local and midland plants at low elevation,
which suggests that their significantly slower growth rates were partly adaptive. Slow growth, a
trait that is advantageous at high altitude (Oleksyn et al., 1998; Macek et al., 2009) was
maintained by highland sporelings at low elevations, indicating a genetic basis. Although
normally a disadvantage at lower elevations, a reduced RGR may have given them an
advantage under the extreme conditions encountered during the experiment (such as frost).
The advantage of highland sporelings during frost events at the low elevation site might have
been accidental. Slower growth might also have been beneficial to survive the transplantation
from greenhouse to the field (Wright et al., 2010). Generally though, the higher RGR of
lowland sporelings should increase their fitness at the relatively warm and humid low
elevations.
Our experiment has shown that there seems adaptive genetic differentiation among
populations of A. antiquum growing at different elevations, even though we expected that the
generally high dispersal ability of fern spores would prevent such differentiation. Regarding
gene flow, previous research has identified a high level of genetic differentiation in A. antiquum
at a larger spatial scale, namely within East Asia (Murakami et al., 1999). In West Java, the
closely-related A. nidus, a species likely to have similar patterns of spore dispersal and thus gene
84 CHAPTER 6 flow as A. antiquum, was observed to have separate rbcL-haplotypes that were clearly linked to
different elevations (Yatabe et al., 2002). A. nidus was also reported to have a diverse height of
attachment and habitat preference (hills versus swamps) for individuals of different sizes in
peninsular Malaysia, which was attributed to the existence of cryptic species (Zhang et al., 2010).
Supported by our experimental results, the literature suggests that a genetic differentiation of A.
antiquum among different elevations is likely. The differentiation is probably driven by
adaptation to the more extreme climate conditions at high elevations that led to selection
favouring slower growth and conservative use of resources. This selection appears to be
sufficient to counter any ‘diluting’ effects of gene flow from populations at lower elevations
(Gonzalo-Turpin and Hazard 2009).
Comparing this with the projected distributions of A. antiquum under climate change
scenarios (Hsu et al., 2012), we found that the greatest range expansion occurred in the southeastern lowlands (both A2 and B2 scenarios) and at higher elevations (A2 scenario, with high
rainfall increase). Like most species distribution models, our initial model assumed unlimited
dispersal. Based on these experimental results, we conclude that dispersal is poorer than
expected for the small lightweight spores. For the above-mentioned projected distributions, we
deduce that colonization to higher altitudes would occur mainly through the genetically
preadapted highland populations, whereas expansion into lower altitudes would be best
achieved by rapid-growing plants from lowland (and midland) populations. Thus, A. antiquum
is not expected to be negatively affected by climate change, owing to its wide distribution and
genetic adaptation at its range margin. However, some caution is in order, since we have only
considered sporeling growth and survival and have no data on the performance (e.g. growth,
mortality and reproduction) of adult plants or the establishment of the sporelings in situ (i.e.
germination and attachment).
In conclusion, our reciprocal transplant experiment showed a strong site effect on
both the growth and survivorship of juvenile A. antiquum, indicating that habitat suitability
differed substantially between the sites. At the more extreme climate conditions observed at
the high elevation site, the local plants were clearly better adapted, evidenced by their higher
survival. These highland plants also grew more slowly but survived more successfully at the
low elevation site. This may be an accidental consequence of the frost event that we recorded
during our experimental period, to which the highland sporelings were probably better adapted.
Under normal lowland conditions, frost will be rare and faster growth will probably be
advantageous, resulting in higher fitness. The present study demonstrates an integrated
approach to assess the biodiversity consequence of climate change. The field studies on
phenotypic plasticity and patterns of intraspecific adaptation provide complementary
information which is valuable in parameterizing statistical distribution models.
85 Chapter 7
Simulating climate change impacts on forests and associated vascular epiphytes in a subtropical island of East Asia Pleione formosana
Cypress forest is characterized by cool temperatures, continuously moist and dim conditions, typically enveloped in clouds during the afternoon. Many epiphytes with restricted distributions are specialized to this particular thermal and hydrological regime.
CHAPTER 7
Simulating climate change impacts on forests and
associated vascular epiphytes in a subtropical island
of East Asia
Rebecca C.-C. Hsu, Wil L.M. Tamis, Niels Raes, Geert R. de Snoo, Jan
H.D. Wolf, Gerard Oostermeijer & Shu-Hua Lin
Diversity and Distributions 18, no. 4 (2012): 334-347
Abstract
A hierarchical modelling approach incorporating forest migration velocity and forest typeepiphyte interactions with classical SDMs was used to model the responses of eight forest
types and 237 vascular epiphytes for the year 2100 under two climate change scenarios. Forest
distributions were modelled and modified by dominant tree species’ dispersal limitations and
hypothesized persistence under unfavourable climate conditions (20 years for broad-leaved
trees and 50 years for conifers). The modelled forest projections together with 16
environmental variables were used as predictors in models of epiphyte distributions. A null
method was applied to validate the significance of epiphyte SDMs and potential vulnerable
species were identified by calculating range turnover rates. For the year 2100, the model
predicted a reduction in the range of most forest types, especially for Picea and cypress forests,
which shifted to altitudes ca. 400 and 300 m higher, respectively. The models indicated that
epiphyte distributions are highly correlated with forest types, and the majority (77–78%) of
epiphyte species were also projected to lose 45–58% of their current range, shifting on average
to altitudes ca. 400 m higher than currently. Range turnover rates suggested insensitive
epiphytes were generally lowland or widespread species, whereas sensitive species were more
geographically restricted, showing a higher correlation with temperature-related factors in their
distributions. The hierarchical modelling approach successfully produced interpretable results,
suggesting the importance of considering biotic interactions and the inclusion of terrain-related
factors when developing SDMs for dependant species at a local scale. Long-term monitoring
of potentially vulnerable sites is advised, especially of those sites that fall outside current
conservation reserves where additional human disturbance is likely to exacerbate the effect of
climate change.
87 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES Introduction
Numerous studies indicate that climate change has already altered global patterns of
biodiversity by modifying the geographical distributions of species (Root et al., 2003; Walther et
al., 2005; Lenoir et al., 2008; Harsch et al., 2009; Woodall et al., 2009). In the field of climate
change impact research, species distribution models (SDMs) or ecological niche models
(ENMs) have been increasingly used to estimate potential species range shifts under
paleontological and/or future climate change conditions (Bakkenes et al., 2002; Broennimann et
al., 2006; Hijmans and Graham, 2006; Thuiller et al., 2006; Fitzpatrick et al., 2008; Jensen et al.,
2008; Carnaval and Moritz, 2008). SDMs attempt to recognize species’ realized niche, which is
used to construct potential geographic distributions by relating species occurrences with values
of predictor variables across a series of observation sites (Guisan and Thuiller, 2005). However,
purely climate-based models have been criticized in numerous studies because they may not
contain sufficient environmental parameters to assess climate change impacts (Heikkinen et al.,
2006; Austin and Van Niel, 2011a, b). For example, SDMs tend to overestimate the area of
suitable habitats, particularly for those species with a strong dependency on other species
(Huntley et al., 2010).
In wet tropics, epiphytes form a conspicuous layer in the forest canopy, and are
regarded as one of the groups most vulnerable to global climate change (Benzing, 1998;
Nadkarni & Solano, 2002; Zotz and Bader, 2009). Canopy-dwelling plants have no vascular
connection to the ground or their host plants, making them more sensitive to environmental
changes than their soil-rooted counterparts (Benzing, 2004). Two decades of monitoring the
lichen flora of the Netherlands indicated a dramatic change on the species composition and
abundance attributed to global warming (van Herk et al., 2002). Epiphyte performance relies on
the presence and characteristics of host trees. Although exceptions exist (Callaway et al., 2002),
most vascular epiphytes exhibit no clear host tree preference (Zimmerman and Olmsted, 1992;
Hsu et al., 2002; Martin et al., 2007), yet, the host tree (phorophyte) composition has a
significant influence on likely epiphyte assemblages (Benavides, 2010). Thus, assessing climate
change impacts on epiphytes requires information on not only the regional climate, but also
the microclimate associated with forest types and the specific epiphyte-tree biotic interactions.
Studies have indicated that the inclusion of biotic interactions significantly improved the
accuracy of SDMs (Leathwick et al., 1996; Araújo and Luoto, 2007; Preston et al., 2008). Other
studies have pointed out that the rate of climate change probably outpaces the migration
capacity of many species (Svenning et al., 2008; Thuiller et al., 2008). However, epiphytes are
adapted to highly dynamic forest canopies by producing many, mostly wind-dispersed, seeds or
spores (Benzing, 1990). Accordingly, the colonization of epiphytes on trees should be rapid,
which, in addition to short life-cycles, makes epiphytes suitable climate change indicators
88 CHAPTER 7 (Lugo and Scatena, 1992). For other forest plants, it is still crucial to take dispersal limitation
into account when simulating species distributions (Engler and Guisan, 2009); a study on Cape
Proteaceae indicated that, even with an optimistic migration rate scenario, the modelled species
range loss closely approximated null migration (Midgley et al., 2006). However, because it is
difficult to obtain reliable dispersal data, especially for the tail end of the leptokurtic
distribution, most studies assume either unlimited or no dispersal for the target species.
Fig. 1 Location and the contour altitudes of Taiwan. The Central Ridge runs north‐east to south along the island mountain chain. 89 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES Other debates are concerned with species persistence in unfavourable climatic
conditions (Loehle and LeBlanc, 1996). Common sense dictates that many species (especially
long-lived trees) will not immediately perish during climate changes. Long-lived dominant
canopy trees will be relatively resistant since they can tolerate years of slow growth, whilst early
successional species will die rapidly if their growth rate falls below a minimum (Loehle and
LeBlanc, 1996); this justifies the importance of including species persistence in SDMs. Despite
aforementioned limitations, SDMs do provide valuable first-order assessments of potential
climatic change impacts on biodiversity (Huntley et al., 2010). Pearson and Dawson (2003)
suggested a hierarchical framework for modelling species distributions at different geographical
scales in order to improve model reliability. We have also adopted this approach, and
incorporated a number of non-climatic factors (such as topography).
This study aims to assess the climate change impacts on forests and vascular
epiphytes in the subtropical island of Taiwan, using species distribution models. We propose a
stepwise hierarchical modelling approach, and aim to improve model accuracy and realism by
considering dispersal limitation, tree persistence and biotic interactions between epiphytes and
host trees. Our study specifically addresses two questions: (1) How do environmental factors
contribute to species distributions and their ecological interpretations? (2) What areas and
which species are potentially vulnerable to climate change?
Methods
Study site, species collections and forest types
Taiwan (situated between 21°45'–25°56'N and 119°18'E–124°34'E) is an island with an area of
36,000 km2 (Fig. 1). About 70% of the island area is covered by mountains (> 1,000 m above
sea level [asl]); Mt. Jade (3952 m) is the highest peak in Taiwan. The annual rainfall in Taiwan
ranges from 1,000 mm to over 6,000 mm, and generally falls during the NE monsoon
(October–January), spring rain (February–April), plum rain (May–June) and typhoon-induced
heavy rain events (July–September). The NE monsoon accounts for 45% of the total annual
rainfall, mainly in east Taiwan (Kao et al., 2004). Three hundred and thirty six species of
vascular epiphytes have been reported for Taiwan (Hsu and Wolf, 2009), of which 271 species
are holo-epiphytes (i.e. epiphytes that complete their entire life cycle without contacting the
forest floor). In this study, we applied SDMs on those 271 strictly arboreal species to assess the
impact of climate change under two projected scenarios.
90 >3200 m Abies kawakamii Wind 30 30 50 50 50 50 50 20 20 20 1548** 1548** 1857** 3096** 1548** 3715* 5173* 7759* 3096 3096 3715 6191 3096 4643 6466 9699 Maximum dispersal distance in 2100 (m) 2
91 USDA Forest Service (http://www.fs.fed.us/database/feis/plants/index.html); He and Mladenoff (1999); Verdú (2002); Engler and Guisan (2009). Engler and Guisan ( 2009). 3
Maximum dispersal distance for the year 2080 (*) and 2050 (**). 1
8 Abies forest 2500–3200 m Picea morrisonicola Wind 7 Picea forest 25 2500–3200 m Tsuga chinensis var. Wind formosana 30 20 15 10 15 Wind Large canopy birds, Macaques, Rodents
Large canopy birds, Macaques, Rodents
Large canopy birds, Macaques, Rodents
Wind 800–3000 m Pinus taiwanensis Chamaecyparis spp. Quercus spp. Machilu spp. Castanopsis spp. Ficus spp. Machilus spp. 6 Tsuga Forest 5 Pinus forest 1800–2500 m >1500 m Highland 3 broad‐leaved forest (BLH) Cypress forest 500–1500 m Midland 2 broad‐leaved forets (BLM) 4 <500 m Maximum Age of Altitudinal Persistence dispersal distance Dominant species Dispersal vector maturity range (yr) 2 for persistence (yr)1 trees (m)3 Lowland 1 broad‐leaved forest (BLL) Forest type (abbr. used) Table 1 The eight forest types, associated characters and the maximum dispersal distance at target years. We identified the locations of epiphytic species from herbarium records, published
plant inventories and our own botanical observations. We assigned species occurrences to 1
km2 grid cells; multiple occurrences within the same cell were considered as one ‘unique’
record. The final database comprised 18,239 records (occurrences ranged from five to 1,083)
including 237 species; 34 species with less than five unique localities were excluded from the
model. Over 90% of modelled species were either ferns or orchids (see Appendix 2).
According to the typology studies (Su, 1992; Chiou et al., 2009), the Taiwanese major forest
types can be grouped as: (1) lowland broad-leaved forest (BLL), (2) midland broad-leaved
forest (BLM), (3) highland broad-leaved forest (BLH), (4) cypress forest, (5) Pinus forest, (6)
Tsuga forest, (7) Picea forest and (8) Abies forest (see Table 1 for descriptions). Localities of the
forest types (dominant canopy trees, 11,700 unique records in total) were obtained from the
third national forest resource inventory, conducted by the Taiwan Forest Bureau in 1993
(Taiwan Forest Bureau, 1995).
Table 2. All environmental variables calculated in this study. Asterisks indicating variables used in model building. Environmental variable (Abbreviation) Unit Calculation Citation Annual mean temperature Average monthly mean °C (Tmean)* temperature 2 Annual precipitation (Pannual)* Millimetre
Average monthly precipitation Decimal The standard deviation of the 3 Temperature seasonality (Tsd)*
fraction monthly mean temperatures (Hijmans et Decimal The coefficient of variation of the al., 2005) 4 Precipitation seasonality (Pcv)*
fraction monthly mean precipitation Mean temperature of warmest 5 month The monthly mean temperature of the warmest or coldest month Mean temperature of coldest 6 month Mean temperature of wettest 7 The average monthly mean quarter °C temperature of the three wettest (Nix, 1986) Mean temperature of driest or driest contiguous months 8 quarter Mean temperature of warmest The average monthly mean 9 quarter temperature of the three warmest or coldest contiguous Mean temperature of coldest 10 months quarter ‐Table continued next page‐ 1 92 CHAPTER 7 11 Precipitation of wettest month 12 Precipitation of driest month 13 Precipitation of wettest quarter
14 Precipitation of driest quarter Precipitation of warmest quarter 16 Precipitation of coldest quarter
17 Temperature annual range Precipitation ratio of coldest 18 quarter The monthly precipitation of the wettest or driest month Millimetre
The total precipitation of the three warmest or coldest contiguous months 15 19 Warmth index 20 Total water deficiency (Pdef)* 21 Potential evapotranspiration ratio 22 Monthly rainfall (P01–P12) 23 Inclination (slope)* 24 Aspect (Eastness*, Northness*)
25 (Dto3000)* 26 Soil category (Soilcode)* The total precipitation of the three wettest or driest contiguous months °C Decimal fraction Variable 5 minus variable 6 Variable 17 as a percentage of variable 2 Sum of monthly mean °C (Kira, 1977) temperature above 5 °C Monthly precipitation minus Millimetre (Lee et al., doubled monthly mean minus °C 1997) temperature Mean annual biotemperature Decimal (Anderson divided by total annual fraction et al., 2002) precipitation P01*, P04*, P05*, P06*, P07*, Millimetre
P10* Average terrain slopes of 1 km2 Degree land area Transformed by sin(aspect rad), cos(aspect rad), and assigned Ordinal ordinals: 0: flat, 1: (–1)–(–0.75), numbers: 2: (–0.75)–(–0.5), 3: (–0.5) –(–
0~8 0.25), 4: (–0.25)–0, 5: 0–0.25, 6: 0.25–0.5, 7: 0.5–0.75, 8: 0.75–1 The distance to the nearest (Lee et al., Metre location above 3000 m asl 1997) No soil (0), Inceptisols (1), Oxisols Cardinal (2), Alfisols (3), Spodosols (4), (Guo et al., numbers: Mollisols (5), Entisols (6), Ultisols 2005) 0~9 (7), Andisols (8), Vertisols (9) 93 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES Environmental variables preparation
Present climate data were derived from an array of weather stations (data recorded from 1900
to 1990). Future projected climate data (for the years 2050, 2080 and 2100, determined by
decadal average) were obtained from the Intergovernmental Panel on Climate Change (IPCC)
Third Assessment Report (IPCC, 2001). By 2100, based on Taiwan regional averaging, a
greater temperature increase is predicted for scenario A2 (4.8 °C) than for scenario B2 (3.2 °C),
and predicted annual rainfall increases are 193 mm for A2 and 79 mm for B2. The simulated
climate data were statistically downscaled to a resolution of 1 km2 to match the resolution of
the present day data (35,928 grid cells in total; Wilby and Wigley, 1997, Lin et al., 2010) for the
purposes of regional assessment. Based on monthly temperature and rainfall data, we
calculated ecologically-relevant climate variables representing annual trends (such as mean
annual temperature), seasonality (for example temperature seasonality) and extreme or limiting
climatic factors (such as water deficiency) (Nix, 1986). To avoid multicollinearity (Heikkinen et
al., 2006), we applied correlation tests between variables to exclude highly correlated (Pearson’s
r > 0.75) factors. Along with one edaphic and four topographic factors, 16 environmental
variables with low correlation were eventually selected for model building (Table 2).
Modelling species distributions and model validation
The species distributions were modelled with the maximum entropy method (MaxEnt, version
3.3.3; http://www.cs.princeton.edu/~schapire/maxent/). This programme was developed for
modelling species’ geographic distributions with presence-only data, and has been shown to
outperform the majority of other modelling applications, especially when sample sizes are
small (Elith et al., 2006; Hijmans and Graham, 2006; Pearson et al., 2007; Graham et al., 2008;
Wisz et al., 2008; Mateo et al., 2010). MaxEnt is particularly suited for epiphytes, since most
epiphyte species (especially orchids) are notoriously rare, and it puts no weight on the absence
of an epiphyte in a forest, which is difficult to ensure, especially for high-canopy species.
MaxEnt calculates a probability distribution over the grid, which may be interpreted as an
index of habitat suitability for a species (Elith et al., 2011). The programme also gives an
estimate of the relative contribution of each environmental variable to the model by means of
iterative calculations (in this study, 500 times). Furthermore, the relative magnitudes of
environmental variables derived from one training set of data can be ‘projected’ to another set
of environmental data, which enables MaxEnt to model species distribution under different
climate conditions, such as future climate simulations (VanDerWal et al., 2009).
94 CHAPTER 7 Fig. 2 The stepwise hierarchical modelling approach used in this study. The procedure THRESHOLD removed species distributions below thresholds. The years 2050, 2080 and 2100 are target years for our models. Solid‐line arrows indicate SDM modelling; broken‐line arrows indicate SDM projection. Framed squares indicate our final SDMs. NL = needle forests: Abies, Picea, Tsuga, cypress and Pinus; BL = broad‐leaved forests: highland (BLH), midland (BLM) and lowland (BLL); ENVI VARs = environmental variables; EP = epiphyte. 95 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES We applied a stepwise hierarchical modelling approach to simulate forest and
epiphyte distributions under various climate change scenarios (Fig. 2). In the first step,
forest distributions were modelled under present climatic conditions and subsequently
projected on future scenarios (for the years 2050, 2080 and 2100). The modelled
forest projections at year 2050 and 2080 were used as intermediate steps (Fig. 2),
incorporating divergent persistence abilities for needle- (NL) and broad-leaved trees
(BL). We randomly selected 70% of the forest occurrences for model building, and
reserved the remaining 30% for model testing, calculating the area under the curve
(AUC) value (Phillips et al., 2006). In the second step, we included species dispersal
limitation as a factor affecting future forest distributions. Corlett (2009) pointed out
that most plant species, depending on their dispersal vectors, probably have maximum
dispersal distances of between 100 m and 1 km in tropical East Asia. In Taiwan,
annual typhoons may promote long distance (up to 1 km) dispersal of conifer winged
seeds (Engler and Guisan, 2009). After carefully reviewing earlier reports (Vittoz and
Engler, 2007; Engler et al., 2009) and considering dispersal vectors, for our model we
hypothesized a maximum horizontal dispersal distance of 1 km per year for each
forest type. We calculated the maximum expanded range of each forest type with the
age of the tree at maturity in target years (Table 1), and calibrated by average terrain
inclinations (both 14° below and 22° above 1500 m asl). In step three, we included the
persistence time of forests, being a measure of the time that trees can tolerate
unfavourable climate conditions. We hypothesized a persistence of 20 years for broadleaved trees (BLL, BLM and BLH) and 50 years for needle trees (Abies, cypress, Picea,
Pinus and Tsuga) (Table 1). Accordingly, we modified the projected forest distributions
at year 2100 by incorporating BL distributions at year 2080 and NL distributions at
year 2050 (Figs 2 and 3). The persistent/extended distributions were assigned
threshold values (i.e. minimum habitat suitability). In step four, the resulting eight
forest distributions (eight variables), together with the 16 abiotic variables (Table 2)
were used to model the distribution of 237 epiphyte species. For each species, we
simulated present day conditions and then modelled projections for the year 2100
under both A2 and B2 climate change scenarios.
To validate our model, we used a null method to test the significance of the
epiphyte SDMs (Raes and ter Steege, 2007). This analysis uses all presence records for
model building, which is an advantage because the sample sizes of most epiphyte
species were small. We created null-distributions (999 permutations) for 5–30 records
96 CHAPTER 7 (with intervals of one), 35–55 records (with intervals of five) and 60–100 records (with
intervals of 10), and then applied a curve-fit through the upper limit of the 95%
confidence interval AUC values (Fig. 4). We thus identified which epiphyte SDM had
a significantly higher AUC value than expected by chance (p < 0.05). Species with a
non-significant SDM were omitted from the analyses. Null analysis was not applied on
forest SDMs because each forest type had more than a hundred occurrences.
Fig. 3 An example of a model incorporating dispersal limitation and tree persistence. The modelled distributions (year 2050 = blue, and year 2100 = red) outside dispersal ranges (grey bubbles) were removed from the result. Black dots indicate present day plant occurrence. Considering tree persistence, the tree distribution in year 2050 (blue grids) was assigned a threshold value (lowest suitability) and added to the 2100 distribution (red grids). 97 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES Fig. 4 The area under the curve (AUC) values of species distribu on models (SDMs, ●) and the 95% confidence intervals. AUC values of the randomly drawn null‐models (ᇞ). Three curve fits indicated the consecutive modelling ‘features’ of MaxEnt (occurrence numbers range from 5–10, 10–15 and 15–100). The SDMs with AUC values above corresponding curves were judged to be significant in this study. Data analysis
We calculated the number of newly appearing, remaining and disappearing epiphytes in each
grid cell. The altitude of each grid cell was derived from digital terrain models (DTM). After
testing for normality (Shapiro-Wilk test), pairs of means of median altitudes of the projected
distributions were compared using one-way ANOVA, (SPSS, version 13.0, IBM). We
described the dissimilarity between present and projected distributions using the Jaccard
distance index (J’) and calculated the range turnover rate for each species. To create a species
richness map, we first applied a threshold of sensitivity-specificity sum maximization (Liu et al.,
2005) to convert the MaxEnt probability distribution to a predicted presence map for each
species. Next, every single-species map was overlaid to produce a species richness map for
epiphytes. The richness map was corrected for land-use change to eliminate species
distributions in urbanized regions (assuming this remains unchanged in 2100).
98 CHAPTER 7 Table 3 the range changes (%) and altitudinal change (in metres) for each forest type under two climate change scenarios (scenarios A2 and B2; Ipcc, 2001), the area under curve (AUC) values for the forest models, and the top three factors sorted in descending order according to their relative contributions to each SDM. Broad‐leaved forest: lowland (BLL), midland (BLM), highland (BLH). Forest type Range change (%) A2 B2 Altitudinal change (m) A2 B2 AUC Top three factors Abies –46 –49 217 239 0.9595 Tmean, Dto3000, Eastness Picea –77 –81 403 428 0.9606 Dto3000, Tmean, Tsd Tsuga –48 –53 250 279 0.9124 Tmean, Tsd, Dto3000 Cypress –54 –52 322 329 0.9113 Tmean, Pcv, Pdef Pinus –29 –29 130 148 0.8878 Tmean, Dto3000, Pdef BLH –44 –34 378 282 0.8428 Tmean, Pdef, Tsd BLM –20 –2 578 364 0.8091 Tmean, slope, P10 BLL –12 37 470 268 0.8406 Tmean, P10, Eastness Results
Forest transitions
The SDM-generated forest distribution patterns agreed strongly with observed data (AUC
values ranging from 0.809 to 0.967; Table 3). Although our models suggested that the total
forest area would decrease by 27% and 4% (scenarios A2 and B2, respectively), most forest
types exhibited larger area reductions (Table 3), with the exception of the lowland broad-leaved
forest (BLL), which was projected to expand by 37% from its current extension under scenario
B2. The largest projected reductions in range were for the Picea forest, which decreased by 77%
in scenario A2 and 81% in scenario B2, followed by the cypress forest (–52% and –54%,
respectively). Moreover, projected forest distributions indicated a general tendency to move to
higher altitudes (Table 3). Picea, cypress and midland broad-leaved (BLM) forests showed more
significant movement to higher altitudes than other forest types under both scenarios, whereas
Pinus forests had the most stable distribution. According to the top-three factors contributing
to each forest model, all forest types were sensitive to annual mean temperature (Tmean; Table
3). We also found that the factor distance to elevations above 3,000 m (Dto3000) and
temperature-related factors (such as Tmean and Tsd) were relatively important for Picea forests,
while cypress forest was also sensitive to moisture-related factors (such as Pdef and Pcv).
99 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES October rainfall (P10) was a contributing factor to midland broad-leaved (BLM) forest
distributions.
In addition to the shifting distribution patterns of the eight forest types, the relative
extent of each forest type was also projected to change under climate change. Currently, the
ratio (in terms of area occupied) of broad-leaved forests to coniferous forests is nearly 1:1. By
the year 2100, our models suggest this ratio will be 1.5:1 under scenario A2 and 2:1 under
scenario B2. Vegetation maps (Fig. 5) provided a visual indicator of predicted changes in forest
type, notably in the north-east of Taiwan, especially under scenario A2 (Fig. 5b). Isolated Tsuga
and Picea forests at the southern end of the Central Ridge (Fig. 1) were projected to disappear
under both scenarios. The projections suggested a large decline and fragment of the cypress
forest on the eastern side of the Central Ridge under scenario B2.
Fig. 5 The potential distributions of eight forest types under present and climate change conditions (scenario A2, B2; IPCC, 2001) in Taiwan. Since more than one forest type exists within several altitudinal zones (Abies, Picea and Tsuga >2500 m; cypress, Pinus and BLH >1500 m; BLM and BLL < 1500 m), the resulting habitat suitability of grids are compared to visually present the major forest type. Slashed boundaries indicate reserves suggested for forest monitoring. A = Chi‐Lan reserve, B = Da‐Wu reserve. 100 CHAPTER 7 Fig. 6 The modelled number of species lost (5a), newly appearing (5b) and remaining stable (5c) under climate change conditions (values are the average of scenarios A2 and B2, IPCC, 2001). Figure 5a indicates boundaries of present reserves in Taiwan, and suggested monitoring sites for epiphytes: (1) Chi‐Lan reserve, (2) Mt. Chia‐Li, (3) Tai‐Chi Canyon and (4) Jin‐Shuei‐Ying reserve. Occurrence of high J’ and low J' species listed in Table 4 plotted as dots in 5a and 5c, respectively. Epiphyte transitions
After testing SDMs against null distributions, we excluded 26 non-significant SDMs (see Fig.
4). The 211 modelled epiphyte species consisted of 83 orchids, 111 ferns and 17 species
belonging to other taxa (see Appendix 2). We identified the top ten most and least sensitive
epiphytes to the two climate change scenarios by ranking their Jaccard distance index (J’, i.e.
range turnover rate) and the three most contributing factors to the modelled distribution of
each epiphyte (Table 4). Generally, relatively insensitive (low J’) species correlated with lowland
forest (BLL), whilst more sensitive (high J’) species were associated with mid-elevation forests
(cypress and BLM) and temperature-related factors (Tmean and Tsd). Insensitive epiphytes
were generally lowland or widespread species (those with greater occurrence; see Appendix 2),
whereas sensitive species were more geographically restricted (Fig. 6a, c). Under scenario A2,
83% of epiphyte species had shifted to higher altitudes by 2100; this figure was 90% for
scenario B2 (see Appendix 2). In our projections, high J’ species were more likely to shift to
higher altitudes than low J’ species (Table 4). The average median altitude increased by ca. 400
101 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES m under both climate change scenarios (Fig. 7). On average, 78% of epiphyte species were
projected to lose 58% of their currently occupied area under scenario A2 and 77% of species
were projected to lose 45% of their area under scenario B2 (see Appendix 2). Our models
showed that the remainder of the species (about 20%) expanded their range size by on average
210% and 170% under scenarios A2 and B2, respectively.
At the community level, projected altitudinal shifts in epiphyte distributions brought
about changing spatial patterns of epiphyte richness. At present, epiphyte diversity is highest at
1,000–1,500 m asl (nearly 100 species per 1 km2). Under climate change conditions, our model
indicated that this belt of maximum species richness would shift to 1,500–2,000 m asl (Fig. 8).
On average, 28 epiphytic species are projected to disappear from each grid cell under scenario
A2 and 24 species under scenario B2 (Fig. 6a); the most stable species number was generally
found at 1,000–1,500 m asl (Fig. 6c). Our models suggested a dramatic decrease of species
richness in the north of Taiwan, which was more pronounced under scenario A2 than B2 (Fig.
8). In general, newly appearing species occurred in the southern mountains (1,500–2,000 m asl)
of Taiwan (Fig. 6b).
Fig. 7 Box plot of median altitudes of 211 SDMs for present climate conditions and two scenarios of climate change (A2 and B2; IPCC, 2001). The plots present median, lower quartile, upper quartile, maximum and minimum observations. Different letters indicate significant differences (p < 0.01).
102 CHAPTER 7 Fig. 8 The species richness maps of epiphytes under present and climate change conditions (scenario A2, B2; IPCC, 2001). Discussion
Modelled species responses and possible ecological interpretations
The massive sample size of the tree occurrence data may partially contribute to the high quality
of the forest models. All forest types were highly sensitive to mean annual temperature; this is
expected because mean annual temperature strongly correlates with elevation, driving
vegetation stratification in Taiwan (Su, 1992). Our results showed that most species, both trees
and epiphytes, are projected to shift to higher altitudes. This would probably lead to increased
habitat fragmentation, since landscapes are dissected by deep ravines at higher elevations. In
Taiwan, Picea morrisonicola currently has a scattered distribution between 2,500 and 3,200 m asl,
and our model indicates that a major factor in Picea’s distribution is distance to elevations
above 3,000 m (Dto3000), producing two discrete populations separated by a depression in the
middle of the Taiwan Central Ridge (Fig. 5a). Variable Dto3000 is related to Massenerhebung
effect which explains the variation in altitudinal limits of forest types based on mountain sizes
and locations. In Taiwan, the forest type on the main ridges of major ranges generally have
higher altitudinal limits due to heat retention and wind shadowing; a phenomenon that has
often been noted on small coastal islands (Grubb, 1971; Foster, 2001). The relatively small and
fragmented population of Picea is thus more sensitive to global warming than the other forest
types.
103 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES Table 4 The top 10 (grey shading) and lowest 10 (white background) species for the two scenarios (A2 and B2; IPCC, 2001) sorted by their Jaccard distance index (J’), and the top three factors sorted in descending order according to their relative contributions to each SDM. Species Bulbophyllum chitouense* Elaphoglossum luzonicum Grammitis nuda* Dendrobium falconeri Mecodium oligosorum Goodyera bilamellata* Flickingeria tairukounia* Pyrrosia matsudae* Saxiglossum angustissimum Bulbophyllum electrinum Microtatorchis compacta Humata chrysanthemifolia Cleisostoma paniculatum Scleroglossum pusillum Gastrochilus raraensis* Mecodium badium Psilotum nudum Vaginularia paradoxa Davallia solida Thrixspermum fantasticum Vittaria taeniophylla Luisia cordata* Oberonia rosea Medinilla formosana* Oberonia gigantea* Pentapanax castanopsisicola*
Calymmodon cucullatus Pomatocalpa acuminata* Thrixspermum formosanum Hoya carnosa Schoenorchis vanoverberghii Altitudinal Altitudinal
shift (m) shift (m) Scenario
J' J' A2 B2 A2 B2 A2 A2, B2 A2, B2 A2 A2, B2 B2 B2 A2 A2 A2, B2 A2 A2 B2 B2 B2 B2 A2 A2 A2 A2, B2 A2, B2 A2, B2 A2, B2 A2 B2 B2 A2 B2 B2 B2 A2, B2 1.00
1.00
1.00
1.00
1.00
0.97
0.99
1.00
1.00
1.00
1.00
1.00
0.98
0.95
0.97
0.96
0.73
0.71
0.68
0.66
0.64
0.61
0.61
0.55
0.79
0.84
0.50
0.84
0.80
0.75
0.31
0.90 0.97 1.00 0.82 0.99 1.00 1.00 0.77 0.93 1.00 0.89 0.86 0.97 0.96 0.96 0.96 0.61 0.71 0.78 0.47 0.42 0.31 0.41 0.59 0.51 0.51 0.56 0.48 0.48 0.46 0.32 496 –654 944 –36 585 845 793 383 661 –731 830 549 745 –742 660 1151 385 –54 –53 9 –22 12 24 –99 535 158 –173 203 203 269 –3 430 –160 885 196 –162 530 715 270 –127 717 471 304 579 –775 738 742 197 –18 –128 56 22 23 11 227 431 243 –61 135 182 217 –17 104 Factor Tsd, BLM, Cypress BLM, Dto3000, soil_code BLM, Dto3000, Pdef P10, slope, soil_code Pdef, P01, slope Tmean, slope, Tsd BLL, Eastness, P05 P10, Dto3000, Pinus Northerness, Tsd, P10 BLM, Pdef, Pinus Tmean, BLL, Pdef P05, P07, BLH P10, BLH, slope Pdef, Pinus, Tsd Pdef, Cypress, BLL Tmean, Tsd, Pannual slope, Pdef, BLL P06, BLL, BLH Dto3000, Tsd, BLL Pdef, BLH, Pinus Pinus, slope, Tsuga soil_code, Dto3000, BLL Dto3000, Eastness, BLH BLL, Dto3000, P07 Pdef, Pinus, slope Pdef, BLL, P05 Dto3000, P06, BLM BLH, P06, Northerness BLH, Eastness, Dto3000 BLL, slope, BLH Dto3000, BLL, slope CHAPTER 7 The model indicated a distinct decline in cypress forest, a major component of
montane cloud forest in Taiwan. Cypress forest is characterized by cool temperatures,
continuously moist and dim conditions, typically enveloped in clouds during the afternoon (Lai
et al., 2006). Many epiphytes with restricted distributions are specialized to this particular
thermal and hydrological regime. Consistent with observations, the models indicated that
epiphyte distributions were strongly correlated with forest type (Table 4). Cypress forest was an
important factor in the distribution of two sensitive endemic orchids (Bulbophyllum chitouense
and Gastrochilus raraensis). Rainfall seasonality (Pcv) and water deficiency (Pdef) were apparently
the most contributing factors to cypress forest distribution. The climate change scenarios
suggest increased precipitation variability in time and space, and future weakening of the NE
monsoon (Lin et al., 2010), which accounts for a substantial proportion of Taiwan’s annual
rainfall, especially in the north-east. These factors are probably responsible for the projected
general decline of the cypress forests and associated epiphytic species (Fig. 5b, c). October is
the onset of the NE monsoon season, thus species distributions correlating closely with
October rainfall (P10) are projected to have high range turnover rates (high J') under future
climate conditions (Table 4).
The sensitivity of species to global climate change is often related to differences in
ecological properties (Broennimann et al., 2006). Past studies suggested that generalists (i.e.
species with wider niche breadths and hence larger range sizes on the environmental gradient)
are expected to demonstrate broader tolerances to climate changes than specialists (Brown et al.,
1995; Benzing, 1998; Thuiller et al., 2004; Broennimann et al., 2006). In other words, the species
with the more critical habitat demands are probably more sensitive to climate change and may
thus be suitable indicator species. Our model results confirm many sensitive (high J') epiphytes
presently have restricted distributions (for example Bulbophyllum chitouense, Grammitis nuda,
Flickingeria tairukounia and Saxiglossum angustissimum), whereas insensitive species (low J') are
widespread, and include several pantropical species (such as Psilotum nudum or Hoya carnosa)
(Table 4). Insensitive epiphytes are usually lowland species, distributed in southern Taiwan, and
less sensitive to temperature-related factors (Fig. 6c and Table 4). Conversely, our results
indicated temperature-related factors (Tmean and Tsd) had an important effect on the
modelled distributions of sensitive epiphytes that grew in the mid-elevation/montane cloud
forests (cypress, BLM and the lower altitude ranges of BLH; Fig. 6a and Table 4). Tropical
montane cloud forests are unique among terrestrial ecosystems for their particular hydroregime
(Still et al., 1999), and typically occur in narrow altitude belts characterized by high endemism
and abundant epiphytes (Foster, 2001). Accordingly, the epiphytes specialized in this ecotone
are probably thermal- or hydro-specialists. Among high J’ species, Mecodium badium is relatively
widespread, yet was projected to have a high range turnover rate under climate change
conditions. This filmy fern occurs widely at mid-altitudes, and its occurrence correlates
105 MODELLING CLIMATE CHANGE IMPACTS ON EPIPHYTES strongly with climate factors (i.e., annual mean temperature, Tmean; temperature seasonality,
Tsd; annual rainfall, Pannual) in the distribution model; thus this species may be more
susceptible to atmospheric drying in a warming climate.
In addition to climate variables, the models indicated that stable topographic or
edaphic factors should be considered when modelling species distributions under climate
change. Eastness was identified as an important predictor of Abies and lowland broad-leaved
forest distributions (Table 3). During the NE monsoon, precipitation (in the form of snow at
high altitudes) is greater on east-facing slopes than those of other aspects, exerting a significant
influence on forest distributions. Austin and Van Niel (2011a) noted similar climate regime
differences between north- and south- facing aspects in temperate latitudes. Soil category is a
contributing factor to some epiphyte distribution models (see Appendix 2); terrestrial soil
fertilities may affect nutrient availability in the canopy and hence epiphyte species
compositions (Gentry and Dodson, 1987a; Benner and Vitousek, 2007). SDMs using climateonly predictors often overestimate range reduction and fail to recognize potential landscapedefined refugia when assessing climate change impacts (Austin and Van Niel, 2011b). Our
study confirms terrain-related factors must be incorporated when projecting species response
to climate change at a local scale.
Migration velocity
Dispersal limitation and persistence induce a lag in modelled tree displacement but forest
transition may be unexpectedly rapid. Pollen records from the Andes indicated that during the
last glacial maximum, the forest belt shifted by ca. 1,000 m and a massive replacement of
ecotone forests occurred, implying a rapid altitudinal displacement of trees on tropical
mountains (Hooghiemstra and van der Hammen, 2004; Groot et al., 2010). Clark (1998)
combined field data with a population growth model to prove that plant dispersal was
compatible with the rapid spread shown by paleontological records. An investigation of 13 tree
species in the French mountains found that the low altitude limits of seedlings were on average
29m higher than the adult parent trees, in response to the warming trend of the past two
decades (Lenoir et al., 2009). An analysis of 60,000 long-term forest inventory plots in the
eastern USA suggested an approximately northward tree migration rate of 100 km per century
(Woodall et al., 2009); in the Alps, the altitudinal shift may have been as much as 340 m over
the past 50 years, this speed correlated with the species wind dispersal (Parolo and Rossi 2008).
However, for some species and areas, establishment limits distributions more than dispersal
(Alsos et al., 2007). Most epiphytes produce highly mobile propagules capable of long-distance
dispersal, yet recruitment of phorophyte-dependent epiphytes inevitably lags behind the trees,
particularly of those epiphytes that depend on old-growth trees for establishment. Considering
the biotic interaction between epiphytes and forest trees, we used forest habitat suitability as
106 CHAPTER 7 the predictor for epiphyte modelling. This approach had the additional advantage that the
range boundary of forests with low habitat suitability for trees, hence, for epiphytes, was also
assessed in the model.
Conservation implications
Our model indicated a considerable decline in the area extent of mid-elevation forests and
associated epiphytic species under the two climate change scenarios. Many mid-altitude species
fall outside current reserves because of their proximity to aboriginal villages (Fig. 6a). We
recommend establishing two long-term forest monitoring sites in the Chi-Lan and Da-Wu
reserves (Fig. 5). Chi-Lan is dominated by primitive cypress and Picea forests, and is an ideal
site for monitoring forest type change. In south Taiwan, Da-Wu reserve lies in a region where
coniferous forests are projected to be replaced by shadier broad-leaved forests, thus favouring
an increase in shade-tolerant epiphytes. We anticipate that tree and epiphyte populations will
change relatively rapidly at Da-Wu. Secondly, we recommend that three mid-altitude sites be
established for epiphyte monitoring, at Mt. Chia-Li, Tai-Chi Canyon and Jin-Shuei-Ying
reserve (Fig. 6a). Jin-Shuei-Ying reserve is characterized by a rich diversity of epiphytic ferns
and is thus an ideal site for monitoring climate-sensitive species, including two locally rare
epiphytic ferns (Elaphoglossum luzonicum and Grammitis nuda; Table 4). Mt. Chia-Li and Tai-Chi
Canyon are near human settlements, and currently lie outside conservation areas, but both
areas are rich in epiphytic orchids and contain the majority of the local sensitive species (Table
4). Long-term plots in these locations can be used to investigate the effects of anthropogenic
disturbance on sensitive epiphytes in a changing climate, thus evaluating the effectiveness of
the present conservation reserves.
Global warming effects seem to be less pronounced in undisturbed forests; human
disturbance may produce vacant niches for invasive species (Aptroot and van Herk, 2007).
Thus, conserving old-growth forests may be crucial in supporting species to resist climate
change (Ellis et al., 2009). Our climate change models showed that midlands are likely to
remain richer in epiphyte species than higher or lower altitudes (Fig. 6c), thus present centres
of species diversity will probably retain their importance into the future (Venter et al., 2010).
On a regional scale, a mountainous island such as Taiwan may act as a potential refuge during
climate change; high mountains provide the space for species migration, as most likely
occurred after the Quaternary glaciations (Hsu and Wolf, 2009). 107 Chapter 8
Conclusions
Chi‐Lan
Mt. Chia‐Li
Tai‐Chi Canyon
Da‐Wu
Jin‐Shuei‐Ying
The recommended sites in Taiwan for monitoring the influence of changing climates
on forests, Chi‐Lan and Da‐Wu reserves (blue boundaries) and on epiphytes, Mt.
Chia‐Li, Tai‐Chi Canyon and Jin‐Shuei‐Ying (orange dots). Two sites (dashed circle)
indicate areas with high epiphyte richness, HsuehShan and AliShan, located in the
northern and central part of the island, respectively.
CHAPTER 8
Conclusions
Patterns in composition and distribution of vascular epiphytes
To investigate and explain regional patterns in species richness along gradients is one of the
major challenges for ecological and biogeographical research. To date, few epiphyte studies are
available from the Paleotropics. Moreover, most studies focus on tropical areas, whilst
epiphyte research from sub-tropical areas remains scarce. In chapter 2, data is provided on the
distribution of epiphyte diversity in Taiwan, a subtropical mountainous island in East Asia, to
complement the perplexity of global patterns. Similar to the epiphyte flora in tropical areas, the
epiphyte diversity in Taiwan was dominated by few higher taxa (families), mostly
monocotyledons. Lacking several species-rich epiphyte families (e.g. Bromeliaceae and
Marcgraviaceae) that have evolved independently in the Neotropics, the most abundant
epiphytes in the checklist were ferns, followed by orchids. The taxonomic composition of the
epiphyte flora demonstrated the transitional aspect of Taiwan, incorporating both tropical and
sub-tropical regions, corroborating a trend of increasing proportion of epiphytic ferns and
fern-allies with latitudes (Wolf and Flamenco-S, 2003; Zotz, 2005).
In addition to the latitudinal gradient, the presence of an extensive mountain system
on the island provides an ideal opportunity for studying species richness patterns along an
elevational gradient. Using 39,084 unique botanical collections, in chapter 3 the epiphyte
richness was found to show a mid-elevation peak at ca. 1000 m asl. This often described
phenomenon of a hump-shaped curve in species richness could not be explained by the middomain effect, as observed in some other studies, but coincided with the richness pattern of
bryophytes on an island in the Indian Ocean (Cardelús et al., 2006; Ah-Peng et al., 2012). The
epiphyte pattern in species richness showed a peak of substantially higher species richness and
at slightly lower elevation than expected under the null model. The latter is presumably
explained by the Massenerhebung effect (i.e. mountain mass elevation effect, Bruijnzeel et al.,
1993). This phenomenon occurs on isolated, small coastal mountains, where floristically-similar
vegetation types tend to be distributed at lower altitude than on large mountain masses, due to
a steep lapse rate of temperature and cloud formation (Flenley, 1995).
The exceptional high species richness beyond the expectation of the null model can
probably be attributed to a large number of species with a small range size, related to fine niche
partitioning. For example, the restricted altitudinal band of Chamaecyparis-dominated “cypress”
cloud forest (1800-2500 m) is inhabited by no less than 92 species of rare ferns (Moore 2000).
109 CONCLUSIONS Moreover, extraordinary endemism has been observed in the mountains and, for instance,
several epiphytic orchid genera, Bulbophyllum (24 spp.), Gastrochilus (9 spp.) and Oberonia (7 spp.),
show a high endemicity of nearly 50 percent. Further analysis in chapter 3 on the altitudinal
ranges of species showed a higher degree of thermal specialization in the upper-zone of
mountains than in the lowlands, which is in contrast to the Rapoport Effect hypothesis
(Stevens, 1992). Interestingly, a transplant experiment suggested that also at the intraspecific
(Asplenium antiquum) level, there was more genetic adaptation of populations at higher
elevations (chapter 6). In summary, the results of above mentioned chapters suggest that
environmental factors mostly account for the observed epiphyte distribution in Taiwan. In this
light, the approach in chapter 7 to use species distribution models (SDMs) to assess potential
range change of epiphytes under future climate conditions is reasonable.
Common features of vulnerable epiphyte species and biomes to
climate change
In chapter 7, the SDMs indicated a large proportion of epiphytes that were projected to have a
high range turnover rate under climate change scenarios (referred to here as “sensitive”
species) presently have restricted distributions in the mountain area (e.g. Bulbophyllum chitouense,
Grammitis nuda, Flickingeria tairukounia, Saxiglossum angustissimum), whereas species with a low
range turnover (“insensitive”) are generally widespread lowland species, including several
pantropical species (e.g. Hoya carnosa, Psilotum nudum). Corroborated by the findings of chapter
3, presumably the sensitivity and/or vulnerability of species under climate change is mostly
correlated with thermal specialisation. The SDMs did indicate that temperature-related factors
(Tmean and Tsd, appendix) had an important effect on the modelled distributions of sensitive
epiphytes, and many of them occurred only in the mid-elevation cloud forests (e.g. cypress
forest). The SDMs projected a distinct decline of cypress forest under future climates, and
showed that here rainfall seasonality (Pcv) and water deficiency (Pdef) were the most
contributing factors to the distribution of cypress forest. This forest is typically enveloped in
clouds during the afternoon and characterized by cool temperatures and continuously moist
and dim conditions (Still et al., 1999; Lai et al., 2006). Many epiphytes with restricted
distributions (e.g. Bulbophyllum chitouense and Gastrochilus raraensis) are specialized to this
particular thermal and hydrological regime (chapter 3). In addition, the SDM result in chapter 3
showed that species with a narrow niche-width (specialists) often have a scattered distribution.
Under future warming scenarios, most species were projected to shift to higher altitudes
(chapter 7), which may result in increased habitat fragmentation due to isolation of deep
ravines at higher elevations. High elevation specialist species, having a relatively small range
size and fragmented population, are likely most susceptible to global warming.
110 CHAPTER 8 In agreement with many prior studies, the ordination analysis in chapter 3 suggested
that, next to light conditions, temperature and water availability were most crucial for epiphyte
distribution (Gentry and Dodson, 1987a, Benzing, 1990, Wolf, 1994). In chapter 4, the
experiment on Hoya carnosa indicated that even in a wet subtropical forest, water conservation
was the likely ecophysiological significance of CAM instead of CO2 availability. In fact, there
exists a positive correlation between air humidity and CAM/C3 species ratio (Monteiro et al.,
2008). The drought-tolerance adaptation might be beneficial under future climate change of
increasing seasonal variation and weather extreme. A contrasting example was represented by
the filmy fern Mecodium badium. Although presently widespread, this epiphytic fern was
projected to have a high range turnover rate under climate change scenarios (chapter 7). Since
the fern's frond consists of only a single layer of cells, absorbing moisture from the air, this
species may be particularly susceptible to atmospheric drying in a warming climate. However,
many epiphytes may be more tolerant to drought stress than usual expectations. For example,
the widespread and abundant population of bird's nest ferns on the island suggests their
successful adaptation to the canopy environment, especially in some dry forests at the
highlands. In chapter 5, an experiment to explore the physiological plasticity of Asplenium nidus
revealed its flexibility in photosynthetic capacity to diverse microclimates. The experiment
accidentally found that A. nidus lacked stomata on the adaxial surface of leaf blades, which
likely also is a morphological adaptation to drought stress. Moreover, drought tolerance may
vary amongst populations of the same species. A transplant experiment with altitudinally
widespread Asplenium antiquum indicated intraspecific variation of drought tolerance (chapter
6).
Species distribution modelling: what can we learn from the
MaxEnt approach?
Despite some uncertainties (Barry and Elith, 2006; Pearson et al., 2006), present studies
(chapter 3 and 7) demonstrated that species distribution modelling or ecological niche
modelling was a practical tool for assessing species richness patterns or evaluating the impact
of climate change. Consistent with field observations, the models indicated that epiphyte
distributions were strongly correlated with forest type, suggesting the importance of
considering biotic interactions for modelling dependent species such as epiphytes. Moreover,
SDMs confirmed that terrain-related factors (e.g. aspect, inclination) were influential when
projecting species’ response to climate change at a local scale. Incorporating stable topographic
or edaphic factors into models might prevent overestimating range reduction and may help to
recognize potential landscape-defined refugia when assessing climate change impacts (Austin
and Van Niel, 2011).
111 CONCLUSIONS Since the presence-only tool MaxEnt puts no weight on the absence of a species, it is
particularly suitable for modelling distribution of canopy epiphytes that are often difficult to
detect from the ground (Flores-Palacios and García-Franco, 2001). In addition, MaxEnt has
been proved to outperform most current SDM approaches, especially when a sample size is
small (Elith et al., 2006; Hernandez et al., 2006). Although a large sample size is beneficial for
accurately mapping of species ranges (Feeley and Silman, 2011), most species deserving special
attention in conservation are inherently rare (Pearson et al., 2007). Thus, caution should
especially be taken when modelling species with relatively few collections. In addition, it is
advisable that SDM’s adopt the natural boundary of the studied species, rather than an artificial
one, to prevent under-prediction of range sizes (Raes, 2012). In this respect, the island biome
of this study is therefore convenient for applying a SDM approach. Another particularly
important decision of presence-only SDMs is how to select background samples (pseudoabsences) for parameterizing models (VanDerWal et al., 2009). This issue becomes even more
relevant since more than half of the epiphyte species in this study were present with less than
50 collections. It is advised that background sample should include the full environmental
range required by the species, and exclude the areas where species might not disperse to or that
are unsuitable for the species (Elith et al., 2011). In chapter 3, by using a full set of data
comprising unique epiphyte occurrences (29,087 out of 35,928 of the total island area) for
backgrounds sampling, the model reliability was improved substantially.
MaxEnt is known to be more robust than most methods when dealing with
correlated variables, thus there is less necessity to pre-select predictors for this approach (Elith
et al., 2011). However, when correlated predictor variables are used, variable contributions
should be interpreted with care (Phillip, 2006). In addition, it is advisable to use correlation
tests or ordination analyses for pre-selection of the candidate predictors to avoid overparameterizing models if the sample size is small, thus providing limited information on the
distribution of species and their environment. The transformation of predictors in MaxEnt is
termed feature, which determines model complexity. Currently, MaxEnt has six feature classes:
linear, product, quadratic, hinge, threshold and categorical. The programme by default (i.e.
using Auto features) restricts models to simple features if few samples were introduced (linear
feature at < 10 samples, linear and quadratic at 10−14 samples, linear, quadratic and hinge at
15−79, all six features at > 79 samples). The hinge feature is recommended for substantially
improving model performance (Phillips and Dudik, 2008; Elith et al., 2011), however a simple
feature may be sufficient for an adequate sample as being found in chapter 3 (Syfert et al.,
2013).
For a species presence-only model such as MaxEnt, it is unclear what particularly
diagnostic tool should be used for model validation. It is advisable that with presence-only data
multiple evaluation measures are used to determine the accuracy of the produced models
112 CHAPTER 8 (Hernandez et al., 2006). MaxEnt also generates the commonly used statistical area under the
ROC curve (AUC) for the assessment of prediction errors in conventional presence/absence
models. Yet, instead of the standard commission rate (i.e. false positive rate or fraction of true
absences being predicted false present), the fraction of total predicted study area present is
used in MaxEnt. Therefore, the AUC values tend to be higher for species with narrow ranges
than widespread species, and a high AUC value necessarily does not suggest a better model.
Consequently, in chapter 3 and 7 a null method was adopted to test the significances of the
SDMs (Raes and ter Steege, 2007). A thousand random-permutation SDMs were generally
performed for sample sizes of less than 100 occurrences, and those SDMs that had a
significantly higher AUC value than expected by chance (p<0.05) were identified. In chapter 3,
additional AIC values (Akaike’s Information Criterion) were calculated to determine whether
the models had more parameters than samples, which would violate the assumptions of AIC
(Warren and Seifert, 2010). Finally, it is noted that MaxEnt provides several options for repeat
sampling and cross-validation, which makes it especially appropriate for small sample sizes,
avoiding the use single training/test splits.
Recommendations on conservation and management of forests
and associated epiphytes
The SDMs in chapter 7 projected a dramatic decline of several forests under future changing
climates. To confirm this result, the establishment of two permanent forest research sites is
recommended for monitoring the trend on forest composition change, for example, by
recording seedling establishment and the trunk diameter growth of adult trees. The two sites,
Chi-Lan (棲蘭) and Da-Wu (大武) reserves, are located in the northern and southern island at
similar elevations (ca. 1500-2000 m asl), yet comprising distinct tree species due to regional
climate dissimilarity (Fig. 1). The Chi-Lan site is a primary forest, dominated by old-growth
cypress and Picea trees, receiving substantial influence from NE monsoon. The future climate
change scenario projects weakening of the NE monsoon, and consequently a substantial
decrease in annual rainfall of this area, which might account for the prediction of major decline
for the local forests (Lin et al., 2010). Accordingly, Chi-Lan may provide the first-order
information on the climate change impacts on forests in the relatively near future. Despite
sharing a common feature of cloud forests with Chi-Lan, the Da-Wu site mainly comprises
broad-leaved trees (e.g. Fagaceae) due to less influence from NE monsoon and its relatively
southern latitude. Here the common coniferous tree Tsuga chinensis var. formosana was projected
to be replaced by broad-leaved forests which might in turn favour shade-tolerant epiphytic
species. Since broad-leaved trees and associated epiphytes are characterized by a shorter
generation time than coniferous trees, the Da-Wu site is a promising site for observing climate
change influence in a relatively short period of time.
113 CONCLUSIONS The SDMs of chapter 3 identified two areas, HsuehShan (雪山) and AliShan (阿里山),
with high epiphyte diversity and endemism that deserve special attention in conservation (title
figure). Receiving adequate precipitation from SW flows, these two epiphyte hotspots harbour
many epiphytic orchids. It was reported recently that terrestrial species in the Alishan area
showed an upward-shift of ca. 3.6 m yr-1 in their range by comparing recorded upper range
limits in 1906 with those of 2006 (Jump et al., 2012). However, no study concerning the
dynamic of abundant epiphyte populations here is known to date. In chapter 7, many sensitive
species that are projected to have a high range-turnover rate under climate change also
occurred in these two areas. Considering present species diversity and the severity of
anthropogenic disturbance (e.g. climate change, land-use change), three sites (title figure), Mt.
Chia-Li (棲蘭山), Tai-Chi Canyon (太極峽谷) and Jin-Shuei-Ying (浸水營), are prioritized for
epiphyte conservation, requiring in-depth investigations. The former two sites, Mt. Chia-Li (ca.
1000-2000 m asl) and Tai-Chi Canyon (ca. 1000 asl), are located in HsuehShan and AliShan
respectively. Both sites are near human settlements and currently lie outside conservation areas.
Consequently, both sites are prone to species extinction. Several rare epiphytic orchids here
(e.g. Bulbophyllum rubrolabellum, Bulbophyllum tokioi, Cymbidium floribundum, Dendrobium falconeri,
Eria javanica, Gastrochilus ciliaris, Gastrochilus fuscopunctatus, Gastrochilus raraensis, Thelasis pygmaea,
Pleione bulbocodioides, Thrixspermum pensile) are potential indicators for monitoring anthropogenic
influences, including climate change. The last site, Jin-Shuei-Ying (ca. 500−1500 m asl) is
located in the very south of Taiwan, characterized by a primary tropical montane cloud forest
and an high endemism in epiphytic ferns (e.g. Grammitis nuda, Crepidomanes palmifolium). Many
epiphytic ferns here are notorious rare and particularly hydro-sensitive (e.g. Hymenophyllaceae,
Grammitidaceae), thus Jin-Shuei-Ying is an ideal site for monitoring climate change influence
on water regime (e.g cloud formation) and population dynamics of climate-sensitive epiphytes.
Global warming effects seem to be less pronounced in undisturbed forests, and
present centres of species diversity might retain their importance into the future (Aptroot and
van Herk, 2007; Venter et al., 2010). Thus, conserving old-growth forests as in above
mentioned sites is crucial in supporting species to resist climate change, particularly those
epiphytes that depend on old-growth trees for establishment (Ellis et al., 2009). SDMs showed
that mid-elevation forests will remain relatively rich in epiphyte species, yet become more
fragmented under future climate change (chapter 7). The combination of anthropogenic
habitat disturbance and destruction, over-collection, and the ongoing climate change will likely
increase the risk for extinction, particularly for mid-elevational tree species and their associated
epiphytes. The present study on the ecology and distribution of epiphytes provides a
framework for future conservation strategies and highlights the urgency of conservation
actions under global climate change. Future conservation strategies should enable conservation
authorities to evaluate the effectiveness of conservation and management efforts.
114 REFERENCES
Adams III, W. (1988) Photosynthetic Acclimation and Photoinhibition of Terrestrial and Epiphytic CAM Tissues Growing in Full Sunlight and Deep Shade. Functional Plant Biology, 15, 123‐134. Adams, W., Terashima, I., Brugnoli, E. & Demmig, B. (1988) Comparisons of photosynthesis and photoinhibition in the CAM vine Hoya australis and several C3 vines growing on the coast of eastern Australia. Plant, Cell & Environment, 11, 173‐181. Adams, W.W., Osmond, C.B. & Sharkey, T.D. (1987) Responses of two CAM species to different irradiances during growth and susceptibility to photoinhibition by high light. Plant Physiology, 83, 213‐218. Ah‐Peng, C., Wilding, N., Kluge, J., Descamps‐Julien, B., Bardat, J., Chuah‐Petiot, M., Strasberg, D. & Hedderson, T.A. (2012) Bryophyte diversity and range size distribution along two altitudinal gradients: Continent vs. island. Acta Oecologica, 42, 58‐65. Aitken, S.N., Yeaman, S., Holliday, J.A., Wang, T. & Curtis‐McLane, S. (2008) Adaptation, migration or extirpation: climate change outcomes for tree populations. Evolutionary Applications, 1, 95‐111. Alsos, I.G., Eidesen, P.B., Ehrich, D., Skrede, I., Westergaard, K., Jacobsen, G.H., Landvik, J.Y., Taberlet, P. & Brochmann, C. (2007) Frequent long‐distance plant colonization in the changing arctic. Science, 316, 1606‐1609. Anderson, R.P., Peterson, A.T. & Gomez‐Laverde, M. (2002) Using niche‐based GIS modeling to test geographic predictions of competitive exclusion and competitive release in South American pocket mice. Oikos, 98, 3‐16. Aptroot, A. & van Herk, C.M. (2007) Further evidence of the effects of global warming on lichens, particularly those with Trentepohlia phycobionts. Environmental Pollution, 146, 293‐298. Araújo, M.B. & Luoto, M. (2007) The importance of biotic interactions for modelling species distributions under climate change. Global Ecology and Biogeography, 16, 743‐753. Atwood, J.T. (1986) The size of the Orchidaceae and the systematic distribution of epiphytic orchids. Selbyana, 9, 171‐186. 115 REFERENCES Austin, M.P. & Niel, K.P.V. (2011a) Impact of landscape predictors on climate change modelling of species distributions: a case study with Eucalyptus fastigata in southern New South Wales, Australia. Journal of Biogeography, 38, 9‐19. Austin, M.P. & Van Niel, K.P. (2011b) Improving species distribution models for climate change studies: variable selection and scale. Journal of Biogeography, 38, 1‐8. Bakkenes, M., Alkemade, J.R.M., Ihle, F., Leemans, R. & Latour, J.B. (2002) Assessing effects of forecasted climate change on the diversity and distribution of European higher plants for 2050. Global Change Biology, 8, 390‐407. Barkman, J.J. (1958) Phytosociology and ecology of cryptogamic epiphytes. Van Gorcum & Comp. N. V., Assen. Barry, S. & Elith, J. (2006) Error and uncertainty in habitat models. Journal of Applied Ecology, 43, 413‐423. Benavides, A.M. (2010) Distribution and succession of vascular epiphytes in Colombian Amazonia. PhD dissertation, Universiteit van Amsterdam, Amsterdam. Benavides, A.M., Duque, A.J., Duivenvoorden, J.F. ,Vasco, G.A. & Callejas, R. (2005) A first quantitative census of vascular epiphytes in rain forests of Colombian Amazonia. Biodiversity and Conservation, 14, 739‐758. Benito Garzón, M., Alía, R., Robson, T.M. & Zavala, M.A. (2011) Intra‐specific variability and plasticity influence potential tree species distributions under climate change. Global Ecology and Biogeography, 20, 766‐778. Benner, J.W. & Vitousek, P.M. (2007) Development of a diverse epiphyte community in response to phosphorus fertilization. Ecology Letters, 10, 628‐636. Benzing, D.H. (1987) Vascular Epiphytism: Taxonomic Participation and Adaptive Diversity. Annals of the Missouri Botanical Garden, 74, 183‐204. Benzing, D.H. (1990) Vascular epiphytes. Cambridge University Press, Cambridge. Benzing, D.H. (1998) Vulnerabilities of tropical forests to climate change: the significance of resident epiphytes. Climatic Change, 39, 519‐540. Benzing, D.H. (2004) Vascular Epiphytes. Forest canopies (ed. by M. D. Lowman and H.B. Rinker), pp. 175‐211. Elsevier Academic Press, Burlington, U.S.A. Bland, J.M. & Douglas, G.A. (2004) The logrank test. Statistics Notes, 328, 1073. 116 REFERENCES Boufford, D.E., Hsieh, C.F., Huang, T.C., Kuoh, C.S., Ohashi, H., Peng, C.I., Tsai, J.L. & Yang, K.C. (2003) Flora of Taiwan, 2 edn. Dep. of botany, NTU, Taipei, Taiwan. Braak, C.J.F. & Smilauer, P. (2002) CANOCO reference manual and CanoDraw for Windows user's guide: software for canonical community ordination (version 4.5). Microcomputer Power, Ithaca, NY. Broennimann, O., Thuiller, W., Hughes, G., Midgley, G.F., Alkemade, J.M.R. & Guisan, A. (2006) Do geographic distribution, niche property and life form explain plants' vulnerability to global change? Global Change Biology, 12, 1079‐1093. Brown, J.H., Mehlman, D.W. & Stevens, G.C. (1995) Spatial variation in abundance. Ecology, 76, 2028‐2043. Bruijnzeel, L.A., Scatena, F. & Hamilton, L.S. (2011) Tropical montane cloud forests: science for conservation and management. Cambridge University Press., Cambridge, UK. Bruijnzeel, L.A., Waterloo, M.J., Proctor, J., Kuiters, A.T. & Kotterink, B. (1993) Hydrological observations in montane rain forests on Gunung Silam, Sabah, Malaysia, with special reference to the `Massenerhebung' Effect. Journal of Ecology, 81, 145‐167. Callaway, R., Reinhart, K., Moore, G., Moore, D. & Pennings, S. (2002) Epiphyte host preferences and host traits: mechanisms for species‐specific interactions. Oecologia, 132, 221‐230. Cardelus, C.L., Colwell, R.K. & Watkins, J.E. (2006) Vascular epiphyte distribution patterns: explaining the mid‐elevation richness peak. Journal of Ecology, 94, 144‐156. Carnaval, A.C. & Moritz, C. (2008) Historical climate modelling predicts patterns of current biodiversity in the Brazilian Atlantic forest. Journal of Biogeography, 35, 1187‐
1201. Carter, J. & Martin, C. (1994) The occurrence of crassulacean acid metabolism among epiphytes in a high‐rainfall region of Costa Rica. Selbyana, 15, 104‐106. Cascante‐Marin, A., Wolf, J.H.D., Oostermeijer, J.G.B., den Nijs, J.C.M., Sanahuja, O. & Duran‐Apuy, A. (2006) Epiphytic bromeliad communities in secondary and mature forest in a tropical premontane area. Basic and Applied Ecology, 7, 520‐532. Chao, A. (2005) Species richness estimation. Encyclopedia of Statistical Sciences (ed. by N. Balakrishnan, C.B. Read and B. Vidakovic), pp. 7909‐7916. Wiley, New York. Cheng, C. & Kuo, Y. (2004) Forest CO2 gradient and its effect on photosynthetic characteristics of understory seedlings in the Nanjenshan forest. Taiwan Journal of Forest Science, 19, 143‐152. 117 REFERENCES Chiou, C.‐R., Chen, T.‐Y., Liu, H.‐Y., Wang, J.‐C., Yeh, C.‐L. & Hsieh, C.‐F. (2009) Atlas of Natural Vegetation in Taiwan (In Chinese). Taiwan Forestry Bureau, Taipei. Chiou, C.‐R., Hsieh, C.‐F., Wang, J.‐C., Chen, M.‐Y., Liu, H.‐Y., Yeh, C.‐L., Yang, S.‐Z., Chen, T.‐Y., Hsia, Y.‐J. & Song, G.‐Z.M. (2009) The first national vegetation inventory in Taiwan. Taiwan Journal of Forest Science, 24, 295‐302. Chiou, W.L., Martin, C.E., Lin, T.C., Hsu, C.C., Lin, S.H. & Lin, K.C. (2005) Ecophysiological Differences Between Sterile and Fertile Fronds of the Subtropical Epiphytic Fern Pyrrosia lingua (Polypodiaceae) in Taiwan. American Fern Journal, 95, 131‐140. Clark, J.S. (1998) Why trees migrate so fast: Confronting theory with dispersal biology and the paleorecord. American Naturalist, 152, 204‐224. Colwell, R.K. & Lees, D.C. (2000) The mid‐domain effect: geometric constraints on the geography of species richness. Trends in Ecology & Evolution, 15, 70‐76. Colwell, R.K. (2011) EstimateS, Version 8.2: Statistical Estimation of Species Richness and Shared Species from Samples. http://viceroy.eeb.uconn.edu/estimates/. Corlett, R.T. (2009) Seed dispersal distances and plant migration potential in tropical East Asia. Biotropica, 41, 592‐598. Creese, C., Lee, A. & Sack, L. (2011) Drivers of morphological diversity and distribution in the Hawaiian fern flora: Trait associations with size, growth form, and environment. American Journal of Botany, 98, 956‐966. Croat, T. (1978) Flora of Barro Colorado island. Stanford University Press, Standford, USA. Cunningham, S. & Read, J. (2003) Do temperate rainforest trees have a greater ability to acclimate to changing temperatures than tropical rainforest trees? New Phytologist, 157, 55‐64. Da Silveira, L., Sternberg, L., Mulkey, S.S. & Wright, S.J. (1989) Ecological interpretation of leaf carbon isotope ratios: influence of respired carbon dioxide. Ecology, 70, 1317‐
1324. Dawson, J. (1963) New Caledonia and New Zealand‐a botanical comparison. Tuatara, 11, 178‐193. De Lucia, E.H., Shenoi, H.D., Naidu, S.L. & Day, T.A. (1991) Photosynthetic symmetry of sun and shade leaves of different orientations. Oecologia, 87, 51‐57. Dickerson, R.E. (1928) Distribution of life in the Philippines. Bureau of Science Manila, Manila. 118 REFERENCES Eickmeier, W. & Bender, M. (1976) Carbon isotope ratios of crassulacean acid metabolism species in relation to climate and phytosociology. Oecologia, 25, 341‐347. Ek, R.C. (1997) Botanical diversity in the tropical rain forest of Guyana. PhD thesis, Universiteit Utrecht, Utrecht, The Netherlands. Elith, J., Graham, C.H., Anderson, R.P., Dudik, M., Ferrier, S., Guisan, A., Hijmans, R.J., Huettmann, F., Leathwick, J.R., Lehmann, A., Li, J., Lohmann, L.G., Loiselle, B.A., Manion, G., Moritz, C., Nakamura, M., Nakazawa, Y., Overton, J.McC., Peterson, A.T., Phillips, S.J., Richardson, K., Scachetti‐Pereira, R., Schapire, R.E., Soberon, J., Williams, S., Wisz, M.S. & Zimmermann, N.E. (2006) Novel methods improve prediction of species' distributions from occurrence data. Ecography, 29, 129‐151. Elith, J., Phillips, S.J., Hastie, T., Dudík, M., Chee, Y.E. & Yates, C.J. (2011) A statistical explanation of MaxEnt for ecologists. Diversity and Distributions, 17, 43‐57. Ellis, C.J., Yahr, R. & Coppins, B.J. (2009) Local extent of old‐growth woodland modifies epiphyte response to climate change. Journal of Biogeography, 36, 302‐313. Engler, R. & Guisan, A. (2009) MigClim: Predicting plant distribution and dispersal in a changing climate. Diversity and Distributions, 15, 590‐601. Etterson, J.R. (2004) Evolutionary potential of Chamaecrista fasciculata in relation to climate change. I. Clinal patterns of selection along an environmental gradient in the great plains. Evolution, 58, 1446‐1456. Farmer, A.M., Bates, J.W. & Bell, J.N.B. (1992) Ecophysiological effects of acid rain on bryophytes and lichens. Bryophytes and lichens in a changing environment (ed. by J.W. Bates and A.M. Farmer), pp. 284‐313. Clarendon Press, Oxford. Farquhar, G.D. & Sharkey, T.D. (1982) Stomatal conductance and photosynthesis. Annual review of plant physiology, 33, 317‐345. Feeley, K.J. & Silman, M.R. Keep collecting: accurate species distribution modelling requires more collections than previously thought. Diversity and Distributions, 17, 1132‐1140. Fitzpatrick, M.C., Gove, A.D., Sanders, N.J. & Dunn, R.R. (2008) Climate change, plant migration, and range collapse in a global biodiversity hotspot: the Banksia (Proteaceae) of Western Australia. Global Change Biology, 14, 1337‐1352. Flenley, J. (1995) Cloud forest, the Massenerhebung effect, and ultraviolet insolation. Tropical montane cloud forests, pp. 150‐155. Springer, New York. 119 REFERENCES Flores‐Palacios, A. & García‐Franco, J.G. (2001) Sample methods of vascular epiphytic plants: their effects on recording species richness and frequency. Selbyana, 22, 181‐191. Foster, P. (2001) The potential negative impacts of global climate change on tropical montane cloud forests. Earth‐Science Reviews, 55, 73‐106. Frahm, J.‐P. & Klaus, D. (2001) Bryophytes as indicators of recent climate fluctuations in Central Europe. Lindbergia, 26, 97‐104. Freiberg, M. & Freiberg, E. (2000) Epiphyte diversity and biomass in the canopy of lowland and montane forests in Ecuador. Journal of Tropical Ecology, 16, 673‐688. Freiberg, M. & Turton, S.M. (2007) Importance of drought on the distribution of the birds nest fern, Asplenium nidus, in the canopy of a lowland tropical rainforest in north‐
eastern Australia. Austral Ecology, 32, 70‐76. Gehrig‐Downie, C., Obregón, A., Bendix, J. & Gradstein, S.R. (2011) Epiphyte biomass and canopy microclimate in the tropical lowland cloud forest of French Guiana. Biotropica, 43, 591‐596. Gentry, A.H. & Dodson, C.H. (1987a) Diversity and biogeography of neotropical vascular epiphytes. Annals of the Missouri Botanical Garden, 74, 205‐233. Gentry, A.H. & Dodson, C.H. (1987b) Contribution of Nontrees to Species Richness of a Tropical Rain Forest. Biotropica, 19, 149‐156. Gonzalo‐Turpin, H. & Hazard, L. (2009) Local adaptation occurs along altitudinal gradient despite the existence of gene flow in the alpine plant species Festuca eskia. Journal of Ecology, 97, 742‐751. Graham, C.H., Elith, J., Hijmans, R.J., Guisan, A., Peterson, A.T. & Loiselle, B.A. (2008) The influence of spatial errors in species occurrence data used in distribution models. Journal of Applied Ecology, 45, 239‐247. Gravatt, D. & Martin, C. (1992) Comparative ecophysiology of five species of Sedum (Crassulaceae) under well‐watered and drought‐stressed conditions. Oecologia, 92, 532‐541. Gravendeel, B., Smithson, A., Slik , F.J.W. & Schuiteman, A. (2004) Epiphytism and pollinator specialization: drivers for orchid diversity? Phil. Trans. R. Soc. Lond. B, 359, 1523 ‐ 1535. 120 REFERENCES Griffiths, H. (1988) Carbon balance during CAM: an assessment of respiratory CO2 recycling in the epiphytic bromeliads Aechmea nudicaulis and Aechmea fendleri. Plant, Cell & Environment, 11, 603‐611. Griffiths, H. (1992) Carbon isotope discrimination and the integration of carbon assimilation pathways in terrestrial CAM plants. Plant, Cell & Environment, 15, 1051‐1062. Griffiths, H. (1993) Carbon isotope discrimination. Photosynthesis and Production in a Changing Environment: a Field and Laboratory Manual. (ed. by D.O. Hall, J.M.O. Scurlock, H.R. Bolhar‐Nordenkampf and R.C. Leegood, Long, S.P.), pp. 181‐192. Chapman & Hall, London. Groot, M.H.M., Bogotá, R.G., Lourens, L.J., Hooghiemstra, H., Vriend, M., Berrio, J.C., Tuenter, E., van der Plicht, J., van Geel, B., Ziegler, M., Weber, S.L., Betancourt, A., Contreras, L., Gaviria, S., Giraldo, C., González, N., Jansen, J.H.F., Konert, M., Ortega, D., Rangel, O., Sarmiento, G., Vandenberghe, J., van der Hammen, T., van der Linden, M. & Westerhoff, W. (2010) Rapid shifts in South American montane climates driven by pCO2 and ice volume changes over the last two glacial cycles. Climate of the Past, 6, 2117‐2158. Grubb, P.J. (1971) Interpretation of the 'Massenerhebung' effect on tropical mountains. Nature, 229, 44‐45. Guisan, A. & Thuiller, W. (2005) Predicting species distribution: offering more than simple habitat models. Ecology Letters, 8, 993‐1009. Guo, H.‐Y., Liu, C.‐H., Chu, C.‐L., Chiang, C.‐H., & Yei, M.‐C. (2005) Taiwan soil information system (http://taiwansoil.tari.gov.tw/Web.Net2008/index_1/main1‐1.aspx). Taiwan Agricultural Research Institute, Council of Agriculture. Gupta, R. (1968) Flora Nainitalensis: A Handbook of the Flowering Plants of Nainital. Navayug Traders, New Delhi, India. Haber, W.A. (2001) Number of species with different plant growth forms. Monteverde: ecology and conservation of a tropical cloud forest (ed. by N.M. Nadkarni and N.T. Wheelwright), pp. 519‐522. Oxiford University Press, New York, Oxford. Hansen, M., DeFries, R., Townshend, J.R.G. & Sohlberg, R. (1998) UMD Global Land Cover Classification, 1 Kilometer. In: 1981‐1994. Department of Geography, University of Maryland, Maryland. Harris, F.S. & Martin, C.E. (1991) Correlation between CAM‐Cycling and Photosynthetic Gas Exchange in Five Species of Talinum (Portulacaceae). Plant Physiology, 96, 1118‐1124. 121 REFERENCES Harsch, M.A., Hulme, P.E., McGlone, M.S. & Duncan, R.P. (2009) Are treelines advancing? A global meta‐analysis of treeline response to climate warming. Ecology Letters, 12, 1040‐1049. He, H.S. & Mladenoff, D.J. (1999) The effects of seed dispersal on the simulation of long‐
term forest landscape change. Ecosystems, 2, 308‐319. Hedderson, T. & Longton, R.E. (2008) Local adaptation in moss life histories: population‐
level variation and a reciprocal transplant experiment. Journal of Bryology, 30, 1‐
11. Heikkinen, R.K., Luoto, M., Araujo, M.B., Virkkala, R., Thuiller, W. & Sykes, M.T. (2006) Methods and uncertainties in bioclimatic envelope modelling under climate change. Progress in Physical Geography, 30, 751‐777. Hernandez, P.A., Graham, C.H., Master, L.L. & Albert, D.L. (2006) The effect of sample size and species characteristics on performance of different species distribution modeling methods. Ecography, 29, 773‐785. Hijmans, R.J. & Graham, C.H. (2006) The ability of climate envelope models to predict the effect of climate change on species distributions. Global Change Biology, 12, 1–10. Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis, A. (2005) Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology, 25, 1965‐1978. Ho, C.S. (1988) An introduction to the geology of Taiwan, explanatory text of the geologic map of Taiwan., 2 edn. Central Geological Survey, Ministry of Economic Affairs, Taipei. Hofstede, R.G.M., Wolf, J.H.D. & Benzing, D.H. (1993) Epiphytic biomass and nutrient status of a Colombian upper montane rain forest. Selbyana., 14, 37‐45. Holtum, J.A.M., O'Leary, M.H. & Osmond, C.B. (1982) Carbon isotope fractionation during dark CO2 fixation in CAM plants. Crassulacean Acid Metabolism (ed. by I.P. Ting and M. Gibbs), pp. 299‐300. American Society of Plant Physiologists, Rockville. Hooghiemstra, H. & van der Hammen, T. (2004) Quaternary Ice‐Age dynamics in the Colombian Andes: developing an understanding of our legacy. Philosophical Transactions of the Royal Society B, 359, 173‐181. Hosokawa, T. (1950) Epiphyte‐quotient. The Botanical Magazine, Tokyo, 63, 739‐740. Hosokawa, T. (1958) On the synchorological and floristic trends and discontinuities in regard to the Japan‐Liukiu‐Formosa area. Plant Ecology, 8, 65‐92. 122 REFERENCES Hsieh, C.‐F. (2003) Composition, endemism and phytogeographical affinities of the Taiwan flora. Flora of Taiwan (ed. by D.E. Boufford, Hsieh, C.‐F., Huang, T.‐C., Kuoh, C.‐S., Ohashi, H., Peng, C.‐I., Tsai, J.‐L., Yang, K.‐C.), pp. 1‐14, Taipei. Hsu, R.C.‐C. & Wolf, J.H.D. (2009) Diversity and phytogeography of vascular epiphytes in a tropical‐subtropical transition island, Taiwan. Flora, 204, 612‐627. Hsu, R.C.‐C., Horng, F.‐W. & Kuo, C.‐M. (2002) Epiphyte biomass and nutrient capital of a moist subtropical forest in north‐eastern Taiwan. Journal of Tropical Ecology, 18, 659‐670. Hsu, R.C.‐C., Tamis, W.L.M., Raes, N., de Snoo, G.R., Wolf, J.H.D., Oostermeijer, G. & Lin, S.‐
H. (2012) Simulating climate change impacts on forests and associated vascular epiphytes in a subtropical island of East Asia. Diversity and Distributions, 18, 334‐
347. Hsueh, M.‐L. & Lee, H.‐H. (2000) Diversity and distribution of the mangrove forests in Taiwan. Wetlands Ecology and Management, 8, 233‐242. Huey, R.B. (1978) Latitudinal pattern of between‐altitude faunal similarity: mountains might be" higher" in the tropics. The American Naturalist, 112, 225‐229. Hunt, R. (1982) Plant growth curves: the functional approach to plant growth analysis. E. Arnold Publishers, London. Huntley, B., Barnard, P., Altwegg, R., Chambers, L., Coetzee, B.W.T., Gibson, L., Hockey, P.A.R., Hole, D.G., Midgley, G.F., Underhill, L.G. & Willis, S.G. (2010) Beyond bioclimatic envelopes: dynamic species' range and abundance modelling in the context of climatic change. Ecography, 33, 621‐626. Ibisch, P.L., Boegner, A., Nieder, J. & Barthlott, W. (1996) How diverse are neotropical epiphytes? An analysis based on the "Catalogue of the flowering plants and gymnosperms of Peru". Ecotropica, 2, 13‐28. Ingram, S.W., Ferrell‐Ingram, K. & Nadkarni, N.M. (1996) Floristic composition of vascular epiphytes in a neotropical cloud forest, Monteverde, Costa Rica. Selbyana, 17, 88‐
103. IPCC (2001) Model output described in the 2001 IPCC Third Assessment Report (http://www.ipcc‐data.org/sres/gcm_data.html). Janzen, D.H. & Liesner, R. (1980) Annotated check‐list of plants of lowland Guanacaste Province, Costa Rica, exclusive of grasses and non‐vascular cryptogams. Lista anotada de plantas de las tierras bajas de la provincia de Guanacaste, exclusivamente pastos y criptógamas no vasculares. Brenesia, 18, 15‐90. 123 REFERENCES Janzen, D.H. (1967) Why mountain passes are higher in the tropics. The American Naturalist, 101, 233‐249. Jarvis, A. & Mulligan, M. (2011) The climate of cloud forests. Hydrological Processes, 25, 327‐343. Jensen, R.A., Madsen, J., O'Connell, M., Wisz, M.S., Tommervik, H. & Mehlum, F. (2008) Prediction of the distribution of Arctic‐nesting pink‐footed geese under a warmer climate scenario. Global Change Biology, 14, 1‐10. Johansson, D. (1974) Ecology of vascular epiphytes in West African rain forest. Acta Phytogeographica Suecica, 59, 1‐129. Jump, A.S., Huang, T.‐J. & Chou, C.‐H. (2012) Rapid altitudinal migration of mountain plants in Taiwan and its implications for high altitude biodiversity. Ecography, 35, 204‐210. Kanehira, R. (1935) The phytogeographical relationship between Botel Tabago(Kotosyo) and the Philippines on the basis of the ligneous flora Bulletin Biogeography Society Japan, 5, 209‐212. Kanto, T. (1933) Zoogeographical studies of Botel Tobago, with a consideration of the northern part of Wallace's line. Geogr. Rev. Jap., 9, nos. 5‐8. (In Japanese). Kao, S.‐J., Shiah, F.‐K. & Owen, J.S. (2004) Export of dissolved inorganic nitrogen in a partially cultivated subtropical mountainous watershed in Taiwan. Water, Air, and Soil Pollution, 156, 211‐228. Kao, S.‐J., Shiah, F.‐K. & Owen, J.S. (2004) Export of dissolved inorganic nitrogen in a partially cultivated subtropical mountainous watershed in Taiwan. Water, Air, and Soil Pollution, 156, 211‐228. Kaplan, E.L. & Meier, P. (1958) Nonparametric estimation from incomplete observations. Journal of the American Statistical Association, 53, 457‐481. Keeley, J.E. (1996) Aquatic CAM photosynthesis. Crassulacean Acid Metabolism. Biochemistry, Ecophysiology and Evolution (ed. by K. Winter and J.a.C. Smith), pp. 281‐295. Springer, Berlin. Kelly, D.L., O'Donovan, G., Feehan, J., Murphy, S., Drangeid, S.O. & Marcano‐Berti, L. (2004) The epiphyte communities of a montane rain forest in the Andes of Venezuela: patterns in the distribution of the flora. Journal of Tropical Ecology, 20, 643‐666. 124 REFERENCES Kessler, M. (2001) Patterns of diversity and range size of selected plant groups along an elevational transect in the Bolivian Andes. Biodiversity and Conservation, 10, 1897‐
1921. Kessler, M., Kluge, J., Hemp, A. & Ohlemüller, R. (2011) A global comparative analysis of elevational species richness patterns of ferns. Global Ecology and Biogeography, 20, 868‐880. Kessler, M., Parris, B.S. & Kessler, E. (2001) A comparison of the tropical montane pteridophyte floras of Mount Kinabalu, Borneo, and Parque Nacional Carrasco, Bolivia. Journal of Biogeography, 28, 611‐622. Kira, T. (1977) A climatological interpretation of Japanese vegetation zones. Vegetation science and environmental protection (ed. by A. Miyawaki and R. Tuxen), pp. 21‐
30. Maruzen, Tokyo. Kluge, M. & Ting, I.P. (1978) Crassulacean acid metabolism. Analysis of an ecological adaptation. Berlin, Heidelberg, New York.: Springer‐Verlag. Kluge, M., Friemert, V., Ong, B., Brulfert, J. & Goh, C. (1989) In situ studies of Crassulacean acid metobolism in Drymoglossum piloselloides, an epiphytic fern of the humid tropics. Journal of Experimental Botany, 40, 441‐452. Knauft, R.L. & Arditti, J. (1969) Partial identification of dark 14CO2 fixation products in leaves of Cattleya (Orchidaceae). New Phytologist, 68, 657‐661. Köhler, L., Tobón, C., Frumau, K. & Bruijnzeel, L. (2007) Biomass and water storage dynamics of epiphytes in old‐growth and secondary montane cloud forest stands in Costa Rica. Plant Ecology, 193, 171‐184. Koike, T., Kitao, M., Maruyama, Y., Mori, S. & Lei, T.T. (2001) Leaf morphology and photosynthetic adjustments among deciduous broad‐leaved trees within the vertical canopy profile. Tree Physiology, 21, 951‐958. Kolbeck, J. (1995) Notes on epiphytic communities in forests of North Korea. Preslia, Praha, 67, 41‐45. Körner, C. (2000) Biosphere responses to CO2 enrichment. Ecological Applications, 10, 1590‐1619. Kreft, H., Jetz, W., Mutke, J., Kier, G. & Barthlott, W. (2008) Global diversity of island floras from a macroecological perspective. Ecology Letters, 11, 116‐127. 125 REFERENCES Kreft, H., Koster, N., Kuper, W., Nieder, J. & Barthlott, W. (2004) Diversity and biogeography of vascular epiphytes in Western Amazonia, Yasuní, Ecuador. Journal of Biogeography, 31, 1463‐1476. Kress, W.J. (1986) The systematic distribution of vascular epiphyte: an update. Selbyana, 9, 2‐22. Kreyling, J., Thiel, D., Simmnacher, K., Willner, E., Jentsch, A. & Beierkuhnlein, C. (2012) Geographic origin and past climatic experience influence the response to late spring frost in four common grass species in central Europe. Ecography, 35, 268‐
275. Krömer, T., Acebey, A., Kluge, J. & Kessler, M. (2013) Effects of altitude and climate in determining elevational plant species richness patterns: A case study from Los Tuxtlas, Mexico. Flora ‐ Morphology, Distribution, Functional Ecology of Plants, 208, 197‐210. Krömer, T., Kessler, M., Robbert Gradstein, S. & Acebey, A. (2005) Diversity patterns of vascular epiphytes along an elevational gradient in the Andes. Journal of Biogeography, 32, 1799‐1809. Kulandaivelu, G., Noorudeen, A., Sampath, P., Periyanan, S. & Raman, K. (1983) Assessment of the photosynthetic electron transport properties of upper and lower leaf sides in vivo by fluorometric method. Photosynthetica, 17, 204‐209. Kuo, C.M. (1985) Taxonomy and phytogeography of Taiwanese pteridophytes. Taiwania, 30, 5‐100. Küper, W., Kreft, H., Nieder, J., Koster, N. & Barthlott, W. (2004) Large‐scale diversity patterns of vascular epiphytes in Neotropical montane rain forests. Journal of Biogeography, 31, 1477‐1487. Lai, I.‐L., Chang, S.‐C., Lin, P.‐H., Chou, C.‐H. & Wu, J.‐T. (2006) Climatic characteristics of the subtropical mountainous cloud forest at the Yuanyang lake long‐term ecological research site, Taiwan. Taiwania, 51, 317‐329. Laurance, W.F., Carolina Useche, D., Shoo, L.P., Herzog, S.K., Kessler, M., Escobar, F., Brehm, G., Axmacher, J.C., Chen, I.C., Gámez, L.A., Hietz, P., Fiedler, K., Pyrcz, T., Wolf, J., Merkord, C.L., Cardelus, C., Marshall, A.R., Ah‐Peng, C., Aplet, G.H., del Coro Arizmendi, M., Baker, W.J., Barone, J., Brühl, C.A., Bussmann, R.W., Cicuzza, D., Eilu, G., Favila, M.E., Hemp, A., Hemp, C., Homeier, J., Hurtado, J., Jankowski, J., Kattán, G., Kluge, J., Krömer, T., Lees, D.C., Lehnert, M., Longino, J.T., Lovett, J., Martin, P.H., Patterson, B.D., Pearson, R.G., Peh, K.S.H., Richardson, B., Richardson, M., Samways, M.J., Senbeta, F., Smith, T.B., Utteridge, T.M.A., 126 REFERENCES Watkins, J.E., Wilson, R., Williams, S.E. & Thomas, C.D. (2011) Global warming, elevational ranges and the vulnerability of tropical biota. Biological Conservation, 144, 548‐557. Leathwick, J.R., Whitehead, D. & McLeod, M. (1996) Predicting changes in the composition of New Zealand's indigenous forests in response to global warming: a modelling approach. Environmental Software, 11, 81‐90. Lee, P.‐F., Liao, C.‐Y., Lee, Y.‐C., Pan, Y.‐H., Fu, W.‐H. & Chen, H.‐W. (1997) An ecological and environmental GIS database for Taiwa. Council of agriculture, Taipei. Lenoir, J., Gegout, J.C., Marquet, P.A., de Ruffray, P. & Brisse, H. (2008) A significant upward shift in plant species optimum elevation during the 20th century. Science, 320, 1768‐1771. Leo, M. (1995) The importance of tropical montane cloud forest for preserving vertebrate endemism in Peru: the Rio Abiseo National Park as a case study. Tropical montane cloud forests (ed. by L.S. Hamilton, J.O. Juvik and F.N. Scatena), pp. 198‐205. Springer‐Verlag, New York. Lepš, J. & Smilauer, P. (2003) Multivariate analysis of Ecological data using CANOCO. Cambridge University Press. Lester, S.E., Ruttenberg, B.I., Gaines, S.D. & Kinlan, B.P. (2007) The relationship between dispersal ability and geographic range size. Ecology Letters, 10, 745‐758. Leverenz, J. & Jarvis, P. (1979) Photosynthesis in Sitka spruce. VIII. The effects of light flux density and direction on the rate of net photosynthesis and the stomatal conductance of needles. Journal of Applied Ecology, 16, 919‐932. Lin, S.‐H., Liu, C.‐M., Huang, W.‐C., Lin, S.‐S., Yen, T.‐H., Wang, H.‐R., Kuo, J.‐T. & Lee, Y.‐C. (2010) Developing a yearly warning index to assess the climatic impact on the water resources of Taiwan, a complex‐terrain island. Journal of Hydrology, 390, 13‐22. Lin, T.‐C., Hamburg, S.P., Hsia, Y.‐J., Lin, T.‐T., King, H.‐B., Wang, L.‐J. & Lin, K.‐C. (2003) Influence of typhoon disturbances on the understory light regime and stand dynamics of a subtropical rain forest in northeastern Taiwan. Journal of Forest Research, 8, 139‐145. Lin, Y.‐H., Tsai, J.‐M., Lai, J.‐T. & Lin, Y.‐C. (2006) The development and applications of the historical typhoon analysis system (HTAS)‐the disturbance regime of typhoons in Hengchun Peninsula. National Taiwan Museum Quarterly, 59, 50‐60. 127 REFERENCES Liu, C., Berry, P.M., Dawson, T.P. & Pearson, R.G. (2005) Selecting thresholds of occurrence in the prediction of species distributions. Ecography, 28, 385‐393. Loarie, S.R., Duffy, P.B., Hamilton, H., Asner, G.P., Field, C.B. & Ackerly, D.D. (2009) The velocity of climate change. Nature, 462, 1052‐1055. Loehle, C. & LeBlanc, D. (1996). Model‐based assessments of climate change effects on forests: a critical review. Ecological Modelling, 90, 1‐31. Lugo, A.E. & Scatena, R.N. (1992). Epiphytes and climate change research in the Caribbean: A proposal. Selbyana, 13, 123‐130. Lüttge, U. (1987) Carbon dioxide and water demand: crassulacean acid metabolism (CAM), a versatile ecological adaptation exemplifying the need for integration in ecophysiological work. New Phytologist, 106, 593‐629. Lüttge, U. (1989) Vascular plants as epiphytes: evolution and ecophysiology. Springer‐
Verlag, Berlin. Mabry, C.M., Hamburg, S.P., Lin, T.‐C., Horng, F.‐W., King, H.‐B. & Hsia, Y.‐J. (1998) Typhoon disturbance and stand‐level damage patterns at a subtropical forest in Taiwan. Biotropica, 30, 238‐250. Macek, P., Macková, J. & de Bello, F. (2009) Morphological and ecophysiological traits shaping altitudinal distribution of three Polylepis treeline species in the dry tropical Andes. Acta Oecologica, 35, 778‐785. Madison, M. (1977) Vascular epiphytes: their systematic occurrence and salient features. Selbyana, 2, 1‐13. Martin, C., Christensen, N. & Strain, B. (1981) Seasonal patterns of growth, tissue acid fluctuations, and 14CO2 uptake in the crassulacean acid metabolism epiphyte Tjllandsia usneoides L. (Spanish moss). Oecologia, 49, 322‐328. Martin, C.E. (1994) Physiological ecology of the Bromeliaceae. The Botanical Review, 60, 1‐
82. Martin, C.E., Lin, T.‐C., Hsu, C.‐C. & Lin, S.‐H. (2007) No effect of host tree species on the physiology of the epiphytic orchid Bulbophyllum japonicum in a subtropical rainforest in Northeastern Taiwan. Taiwan Journal of Forest Science, 22, 241‐251. Martin, C.E., McLeod, K.W., Eades, C.A. & Pitzer, A.F. (1985) Morphological and physiological responses to irradiance in the CAM epiphyte Tillandsia usneoides L. (Bromeliaceae). Botanical Gazette, 146, 489‐494. 128 REFERENCES Martin, C.E., Lin, T.‐C., Hsu, R.C.‐C., Lin, S.‐H., Lin, K.‐C., Hsia, Y.‐J. & Chiou., W.‐L. (2004) Ecophysiology and plant size in a tropical epiphytic fern, Aplenium nidus, in Taiwan. International Journal of Plant Science, 165, 65‐72. Martin, S.L., Davis, R., Protti, P., Lin, T.C., Lin, S.H. & Martin, C.E. (2005) The occurrence of Crassulacean acid metabolism in epiphytic ferns, with an emphasis on the Vittariaceae. International Journal of Plant Sciences, 166, 623‐630. Mateo, R.G., Croat, T.B., Felicísimo, Á.M. & Muñoz, J. (2010) Profile or group discriminative techniques? Generating reliable species distribution models using pseudo‐absences and target‐group absences from natural history collections. Diversity and Distributions, 16, 84‐94. McCain, C.M. (2004) The mid‐domain effect applied to elevational gradients: species richness of small mammals in Costa Rica. Journal of Biogeography, 31, 19‐31. Medina, E. & Minchin, P. (1980) Stratification of δ13C values of leaves in Amazonian rain forests. Oecologia, 45, 377‐378. Medina, E., Montes, G., Guevas, E. & Rokzandic, Z. (1986) Profiles of CO2 Concentration and δ13C Values in Tropical Rain Forests of the Upper Rio Negro Basin, Venezuela. Journal of Tropical Ecology, 2, 207‐217. Medina, E., Sternberg, L. & Cuevas, B. (1991) Vertical stratification of δ13C values in closed natural and plantation forests in the Luquillo mountains, Puerto Rico. Oecologia, 87, 369‐372. Midgley, G.F., Hughes, G.O., Thuiller, W. & Rebelo, A.G. (2006) Migration rate limitations on climate change‐induced range shifts in Cape Proteaceae. Diversity and Distributions, 12, 555‐562. Migenis, L.E. & Ackerman, J.D. (1993) Orchid‐phorophyte relationships in a forest watershed in Puerto Rico. Journal of Tropical Ecology, 9, 231‐240. Moffett, M.W. (2000) What's "Up"? A critical look at the basic terms of canopy biology. Biotropica, 32, 569‐596. Monteiro, J.A.F., Zotz, G. & Körner, C. (2009) Tropical epiphytes in a CO2‐rich atmosphere. Acta Oecologica, 35, 60‐68. Moore, R., Clark, W.D. & Vodopich, D.S. (1998) Botany. WCB/McGraw‐Hill New York. Moore, S.J. (2000) The biodiversity and conservation of Taiwanese pteridophytes. The 2000' Cross‐strait Symposium on Bio‐diversity and Conservation (ed by Y.S. Chou, F.R. Hrieh, S.H. Wu and W.H. Chou), pp. 331‐359. Taichung, Taiwan. 129 REFERENCES Mulligan, M. (2010) Modelling the tropics‐wide extent and distribution of cloud forest and cloud forest loss, with implications for conservation priority. Tropical Montane Cloud Forests. Science for Conservation and Management (ed. by L.A. Bruijnzeel, F. Scatena and L.S. Hamilton.), pp. 14‐38. Cambridge university, Cambridge, UK. Murakami, N., Watanabe, M., Yokoyama, J., Yatabe, Y., Iwasaki, H. & Serizawa, S. (1999) Molecular taxonomic study and revision of the three Japanese species of Asplenium sect. Thamnopteris. Journal of Plant Research, 112, 15‐25. Nadkarni, N. & Solano, R. (2002) Potential effects of climate change on canopy communities in a tropical cloud forest: an experimental approach. Oecologia, 131, 580‐586. Nakicenovic, N., Alcamo, J., Davis, G., de Vries, B., Fenhann, J., Gaffin, S., Gregory, K., Grubler, A., Jung, T.Y., Kram, T., La Rovere, E.L., Michaelis, L., Mori, S., Morita, T., Pepper, W., Pitcher, H.M., Price, L., Riahi, K., Roehrl, A., Rogner, H.‐H., Sankovski, A., Schlesinger, M., Shukla, P., Smith, S.J., Swart, R., van Rooijen, S., Victor, N. & Dadi, Z. (2000) Special report on emissions scenarios : a special report of Working Group III of the Intergovernmental Panel on Climate Change. Cambridge University Press, New York. Nicotra, A.B., Atkin, O.K., Bonser, S.P., Davidson, A.M., Finnegan, E.J., Mathesius, U., Poot, P., Purugganan, M.D., Richards, C.L., Valladares, F. & van Kleunen, M. (2010) Plant phenotypic plasticity in a changing climate. Trends in Plant Science, 15, 684‐692. Nieder, J., Prosperí, J. & Michaloud, G. (2001) Epiphytes and their contribution to canopy diversity. Plant Ecology, 153, 51‐63. Nix, H. (1986) A biogeographic analysis of Australian elapid snakes. Atlas of elapid snakes of Australia. (ed. by R. Longmore), pp. 4‐14, Australian Government Publishing Service, Canberra. Oberbauer, S.F., von Kleist, K., Whelan, K.R.T. & Koptur, S. (1996) Effects of Hurricane Andrew on epiphyte communities within cypress domes of Everglades National Park. Ecology, 77, 964‐967. Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R., Simpson, G.L., Solymos, M., Stevens, H.H. & Wanger, H. (2013) Vegan: Community Ecology Package. R package version 2.0‐6. http://CRAN.R‐project.org/package=vegan. Oleksyn, J., Modrzýnski, J., Tjoelker, M.G., Z∙ytkowiak, R., Reich, P.B. & Karolewski, P. (1998) Growth and physiology of Picea abies populations from elevational transects: common garden evidence for altitudinal ecotypes and cold adaptation. Functional Ecology, 12, 573‐590. 130 REFERENCES Oliver, W.R.B. (1930) New Zealand Epiphytes. The Journal of Ecology, 18, 1‐50. Oney, B., Reineking, B., O'Neill, G. & Kreyling, J. (2013) Intraspecific variation buffers projected climate change impacts on Pinus contorta. Ecology and Evolution, 3, 437‐449. Osmond, C. (1978) Crassulacean acid metabolism: a curiosity in context. Annual Review of Plant Physiology, 29, 379‐414. Parolo, G. & Rossi, G. (2008) Upward migration of vascular plants following a climate warming trend in the Alps. Basic and Applied Ecology, 9, 100‐107. Pearman, P.B., D'Amen, M., Graham, C.H., Thuiller, W. & Zimmermann, N.E. (2010) Within‐taxon niche structure: niche conservatism, divergence and predicted effects of climate change. Ecography, 33, 990‐1003. Pearson, R.G. & Dawson, T.P. (2003) Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful? Global Ecology and Biogeography, 12, 361‐371. Pearson, R.G., Raxworthy, C.J., Nakamura, M. & Peterson, A.T. (2007) Predicting species distributions from small numbers of occurrence records: a test case using cryptic geckos in Madagascar. Journal of Biogeography, 34, 102‐117. Phillips, S. (2008) A Brief Tutorial on Maxent. In. AT&T Research, Florham Park, New Jersey. Available at: http://www.cs.princeton.edu/schapire/maxent/tutorial/tutorial.doc. Phillips, S.J. & Dudík, M. (2008) Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography, 31, 161‐175. Phillips, S.J., Anderson, R.P. & Schapire, R.E. (2006) Maximum entropy modeling of species geographic distributions. Ecological Modelling, 190, 231‐259. Ponce‐Reyes, R., Nicholson, E., Baxter, P.W.J., Fuller, R.A. & Possingham, H. (2013) Extinction risk in cloud forest fragments under climate change and habitat loss. Diversity and Distributions, 19, 518‐529. Poulson, M.E. & DeLucia, E.H. (1993) Photosynthetic and structural acclimation to light direction in vertical leaves of Silphium terebinthinaceum. Oecologia, 95, 393‐400. Preston, K.L., Rotenberry, J.T., Redak, R.A. & Allen, M.F. (2008) Habitat shifts of endangered species under altered climate conditions: importance of biotic interactions. Global Change Biology, 14, 2501‐2515. 131 REFERENCES Pypker, T.G., Unsworth, M.H. & Bond, B.J. (2006) The role of epiphytes in rainfall interception by forests in the Pacific Northwest. II. Field measurements at the branch and canopy scale. Canadian Journal of Forest Research, 36, 819‐832. R Core Team (2012) R: A language and environment for statistical computing. R Foundation for Statistical Computing. Raes, N. & ter Steege, H. (2007) A null‐model for significance testing of presence‐only species distribution models. Ecography, 30, 727‐736. Raes, N. (2012) Partial versus full species distribution models. Natureza & Conservação, 10, 127‐138. Rahbek, C. (1995) The elevational gradient of species richness: a uniform pattern? Ecography, 18, 200‐205. Rauer, G. & Rudolph, D. (2001) Vaskuläre Epiphyten eines westandinen Bergregenwaldes in Ecuador. The flora of the Rio Guajalito mountain rain forest (Ecuador). Results of the Bonn‐Quito epiphyte project, funded by the Volkswagen foundation (ed. by J. Nieder and W. Barthlott), pp. 323‐470. Book on Demand GmbH, Bonn. Ray, N. & Adams, J.M. (2001) A GIS‐based vegetation map of the World at the Last Glacial Maximum (25,000‐15,000 BP). Internet Archaeology, 11. http://anthro.unige.ch/ lgmvegetation/ Reyes‐Betancort, J.A., Santos Guerra, A., Guma, I.R., Humphries, C.J. & Carine, M.A. (2008) Diversity, rarity and the evolution and conservation of the Canary Islands endemic flora. Anales del Jardín Botánico de Madrid, 65, 25‐45. Richards, P.W. (1952) The tropical rain forest. Cambridge University press, London, UK. Riveros, M. & Ramírez, C. (1978) Fitocenoses epífitas de la asociación Lapagerio‐
Aextoxiconetum en el fundo San Martín (Valdivia‐Chile). Acta Científica Venezolana, 29, 163‐169. Root, T.L., Price, J.T., Hall, K.R., Schneider, S.H., Rosenzweig, C. & Pounds, J.A. (2003) Fingerprints of global warming on wild animals and plants. Nature, 421, 57‐60. Rundel, P.W., Ehleringer, J.R. & Nagy, K.A. (1989) Stable isotopes in ecological research. Springer‐Verlag New York. Schaijes, M. & Malaisse, F. (2001) Diversity of Upper Katanga epiphytes (mainly orchids) and distribution in different vegetation units. Systematics and Geography of Plants, 71, 575‐584. 132 REFERENCES Schleser, G. & Jayasekera, R. (1985) δ13C‐variations of leaves in forests as an indication of reassimilated CO2 from the soil. Oecologia, 65, 536‐542. Schreiber, U., Fink, R. & Vidaver, W. (1977) Fluorescence induction in whole leaves: differentiation between the two leaf sides and adaptation to different light regimes. Planta, 133, 121‐129. Schröter, C., Brockmann‐Jerosch, M.C., Günthart, A. & Vogler, P. (1908) Das Pflanzenleben der Alpen: eine schilderung der hochgebirgsflora. A. Raustein, Zürich. Sharkey, T.D. (1985) Photosynthesis in intact leaves of C3 plants: physics, physiology and rate limitations. The Botanical Review, 51, 53‐105. Silander, J.A., Jr. (1985) The genetic basis of the ecological amplitude of Spartina patens. II. variance and correlation analysis. Evolution, 39, 1034‐1052. Skillman, J.B. & Winter, K. (1997) High photosynthetic capacity in a shade‐tolerant crassulacean acid metabolism plant. Implications for sunfleck use, nonphotochemical energy dissipation, and susceptibility to photoinhibition. Plant Physiology, 113, 441‐450. Sokal, R.R. & Rohlf, F.J. (1981) Biometry: the principles and practice of statistics in biological research. W.H. Freeman & Co., New York. Solomon, S. (2007) Climate change 2007‐the physical science basis: Working group I contribution to the fourth assessment report of the IPCC. Cambridge University Press, Cambridge. Song, L., Liu, W.‐Y. & Nadkarni, N.M. (2012) Response of non‐vascular epiphytes to simulated climate change in a montane moist evergreen broad‐leaved forest in southwest China. Biological Conservation, 152, 127‐135. Stevens, G.C. (1992) The elevational gradient in altitudinal range: an extension of Rapoport's latitudinal rule to altitude. The American Naturalist, 140, 893‐911. Still, C.J., Foster, P.N. & Schneider, S.H. (1999) Simulating the effects of climate change on tropical montane cloud forests. Nature, 398, 608‐610. Su, H.J. (1985) Studies on the climate and vegetation type of the natural forest in Taiwan (III). A scheme of geographical climatic region. Quarterly Journal of Chinese Forestry, 18, 33‐44. Su, H.‐J. (1992) Vegetation of Taiwan: Altitudinal vegetation zones and geographical climatic regions (in Chinese). Proceedings of the Institute of Botany, Academia Sinica Taiwan, 11, 39‐53. 133 REFERENCES Sutton, B., Ting, I. & Troughton, J. (1976) Seasonal effects on carbon isotope composition of cactus in a desert environment. Nature, 261, 42‐43. Svenning, J.‐C., Normand, S. & Skov, F. (2008) Postglacial dispersal limitation of widespread forest plant species in nemoral Europe. Ecography, 31, 316‐326. Syfert, M.M., Smith, M.J. & Coomes, D.A. (2013) The effects of sampling bias and model complexity on the predictive performance of MaxEnt species distribution models. PLoS ONE, 8, e55158. doi:10.1371/journal.pone.0055158. Syvertsen, J. & Cunningham, G. (1979) The effects of irradiating adaxial or abaxial leaf surface on the rate of net photosynthesis of Perezia nana and Helianthus annuus. Photosynthetica, 13, 267‐293. Szczepaniak, K. & Biziuk, M. (2003) Aspects of the biomonitoring studies using mosses and lichens as indicators of metal pollution. Environmental Research, 93, 221‐230. Taiwan Forest Bureau (1995) The third forest resource and land use inventory in Taiwan. Council of Agriculture, Nantou. Takhtajan, A. (1986) Floristic Regions of the World. University of California Press, Berkeley, California, U.S.A. Terashima, I. & Inoue, Y. (1984) Comparative photosynthetic properties of palisade tissue chloroplasts and spongy tissue chloroplasts of Camellia japonica L.: functional adjustment of the photosynthetic apparatus to light environment within a leaf. Plant and Cell Physiology, 25, 555‐563. Terashima, I. (1986) Dorsiventrality in photosynthetic light response curves of a leaf. Journal of Experimental Botany, 37, 399‐405. Terashima, I., Sakaguchi, S. & Hara, N. (1986) Intra‐leaf and intracellular gradients in chloroplast ultrastructure of dorsiventral leaves illuminated from the adaxial or abaxial side during their development. Plant and Cell Physiology, 27, 1023‐1031. Thomas, C.D., Cameron, A., Green, R.E., Bakkenes, M., Beaumont, L.J., Collingham, Y.C., Erasmus, B.F.N., Siqueira, M.F.d., Grainger, A., Hannah, L., Hughes, L., Huntley, B., Jaarsveld, A.S.v., Midgley, G.F., Miles, L., Ortega‐Huerta, M.A., Peterson, A.T., Phillips, O.L. & Williams, S.E. (2004) Extinction risk from climate change. Nature, 427, 145‐148. Thuiller, W., Albert, C., Araujo, M. B., Berry, P.M., Cabeza, M., Guisan, A., Hickler, T., Midgley, G.F., Paterson, J., Schurr, F.M., Sykes, M.T. & Zimmermann, N.E. (2008) Predicting global change impacts on plant species' distributions: Future challenges. Perspectives in Plant Ecology, Evolution and Systematics, 9, 137‐152. 134 REFERENCES Thuiller, W., Brotons, L., Araújo, M.B. & Lavorel, S. (2004) Effects of restricting environmental range of data to project current and future species distributions. Ecography, 27, 8. Thuiller, W., Lavorel, S., Sykes, M.T. & Araújo, M.B. (2006) Using niche‐based modelling to assess the impact of climate change on tree functional diversity in Europe. Diversity and Distributions, 12, 49‐60. Ting, I. (1989) Photosynthesis of arid and subtropical succulent plants. Aliso, 12, 387‐406. Treseder, K.K., Davidson, D.W. & Ehleringer, J.R. (1995) Absorption of ant‐provided carbon dioxide and nitrogen by a tropical epiphyte. Nature, 375, 137‐139. Troughton, J., Card, K. & Hendy, C. (1974) Photosynthetic pathways and carbon isotope discrimination by plants. Carnegie Institute of Washington Yearbook, 73, 768‐80. Tuba, Z., PROCTOR, M. & Takács, Z. (1999) Desiccation‐Tolerant Plants under Elevated Air CO2: A Review. Zeitschrift fur Naturforschung C, 54, 788‐796. Václavík, J. (1984) Photosynthetic CO2 uptake by Zea mays leaves as influenced by unilateral irradiation of adaxial and abaxial leaf surfaces. Biologia Plantarum, 26, 206‐214. Van Herk, C.M., Aptroot, A. & van Dobben, H.F. (2002) Long‐term monitoring in The Netherlands suggests that lichens respond to global warming. The Lichenologist, 34, 141‐154. VanDerWal, J., Shoo, L.P., Graham, C. & Williams, S.E. (2009) Selecting pseudo‐absence data for presence‐only distribution modeling: How far should you stray from what you know? Ecological Modelling, 220, 589‐594. Vaz, A.P.A., Figueiredo‐Ribeiro, R.d.C.L. & Kerbauy, G.B. (2004) Photoperiod and temperature effects on in vitro growth and flowering of P. pusilla, an epiphytic orchid. Plant Physiology and Biochemistry, 42, 411‐415. Venter, O., Laurance, W.F., Iwamura, T., Wilson, K.A., Fuller, R.A. & Possingham, H.P. (2010) Planning for biodiversity in future climates－response. Science, 327, 1453. Verdú, M. (2002) Age at maturity and diversification in woody angiosperms. Evolution, 56, 1352‐1361. Vittoz, P. & Engler, R. (2007) Seed dispersal distances: a typology based on dispersal modes and plant traits. Botanica Helvetica, 117, 109‐124. 135 REFERENCES Wagner, S., Zotz, G., Allen, N.S. & Bader, M.Y. (2013) Altitudinal changes in temperature responses of net photosynthesis and dark respiration in tropical bryophytes. Annals of Botany, 111, 455‐465. Walter, H. (1985) Vegetation of the earth and ecological systems of the geo‐biosphere. Springer‐Verlag, Heidelberg, Germany. Walther, G.‐R., Beißner, S. & Pott, R. (2005) Climate change and high mountain vegetation shifts. Mountain ecosystems (ed. by G. Broll and B. Keplin), pp. 77‐96. Springer, Heidelberg. Wang, S.‐T. (1980) Prediction of the behavior and strength of typhoons in Taiwan and its vicinity. National Science Council Research Report, Taipei, Taiwan. Warren, D.L. & Seifert, S.N. (2010) Ecological niche modeling in Maxent: the importance of model complexity and the performance of model selection criteria. Ecological Applications, 21, 335‐342. Whittaker, R.J. (2000) Scale, succession and complexity in island biogeography: are we asking the right questions? Global Ecology and Biogeography, 9, 75‐85. Wilby, R.L. & Wigley, T.M.L. (1997) Downscaling general circulation model output: a review of methods and limitations. Progress in Physical Geography, 21, 530‐548. Winter, K. & Holtum, J.A. (2002) How closely do the δ13C values of Crassulacean acid metabolism plants reflect the proportion of CO2 fixed during day and night? Plant Physiology, 129, 1843‐1851. Winter, K. & Smith, J.A.C. (eds.) (1996) Crassulacean Acid Metabolism. Biochemistry, Ecophysiology and Evolution. Springer, Berlin. Winter, K. (1985) Crassulacean acid metabolism. Photosynthetic Mechanisms and the Environment (ed. by J. Barber and N.R. Baker), pp. 329‐387. Elsevier Science, Amsterdam. Winter, K., Osmond, C. & Hubick, K. (1986) Crassulacean acid metabolism in the shade. Studies on an epiphytic fern, Pyrrosia longifolia, and other rainforest species from Australia. Oecologia, 68, 224‐230. Winter, K., Wallace, B.J., Stocker, G.C. & Roksandic, Z. (1983) Crassulacean acid metabolism in Australian vascular epiphytes and some related species. Oecologia, 57, 129‐141. Wisz, M.S., Tamstorf, M.P., Madsen, J. & Jespersen, M. (2008) Where might the western Svalbard tundra be vulnerable to pink‐footed goose (Anser brachyrhynchus) 136 REFERENCES population expansion? Clues from species distribution models. Diversity and Distributions, 14, 26‐37. Wolf, J.H.D. & Flamenco‐S, A. (2003) Patterns in species richness and distribution of vascular epiphytes in Chiapas, Mexico. Journal of Biogeography, 30, 1689‐1707. Wolf, J.H.D. (1993) Diversity patterns and biomass of epiphytic bryophytes and lichens along an altitudinal gradient in the northern Andes. Annals of the Missouri Botanical Garden, 80, 928‐960. Wolf, J.H.D. (1994) Factors controlling the distribution of vascular and nonvascular epiphytes in the northern Andes. Vegetatio, 112, 15‐28. Wolf, J.H.D. (2005) The response of epiphytes to anthropogenic disturbance of pine‐oak forests in the highlands of Chiapas, Mexico. Forest Ecology and Management, 212, 376‐393. Woodall, C.W., Oswalt, C.M., Westfall, J.A., Perry, C.H., Nelson, M.D. & Finley, A.O. (2009) An indicator of tree migration in forests of the eastern United States. Forest Ecology and Management, 257, 1434‐1444. Wright, S.J., Kitajima, K., Kraft, N.J.B., Reich, P.B., Wright, I.J., Bunker, D.E., Condit, R., Dalling, J.W., Davies, S.J., Díaz, S., Engelbrecht, B.M.J., Harms, K.E., Hubbell, S.P., Marks, C.O., Ruiz‐Jaen, M.C., Salvador, C.M. & Zanne, A.E. (2010) Functional traits and the growth–mortality trade‐off in tropical trees. Ecology, 91, 3664‐3674. Wu, C.‐C. & Kuo, Y.‐H. (1999) Typhoons affecting Taiwan: Current understanding and future challenges. Bulletin of the American Meteorological Society, 80, 67‐80. Xie, B. & Zhang, F. (2012) Impacts of typhoon track and island topography on the heavy rainfalls in Taiwan associated with Morakot (2009). American Meteorological Society, 140, 3379‐3394. Yatabe, Y., Darnaedi, D. & Murakami, N. (2002) Allozyme analysis of cryptic species in the Asplenium nidus complex from West Java, Indonesia. Journal of Plant Research, 115, 483‐490. Yen, S.H., Kitching, I.J. & Tzen, C.S. (2003) A New Subspecies of Hawkmoth from Lanyu Island, Taiwan, with a Revised and Annotated Checklist of Taiwanese Sphingidae (Lepidoptera). Zoological Studies, 42, 292‐306. Zapfack, L., Nkongmeneck, A., Villiers, J. & Lowman, M. (1996) The importance of pteridophytes in the epiphytic flora of some phorophytes of the Cameroonian semi‐deciduous rain forest. Selbyana, 17, 76‐81. 137 REFERENCES Zhang, L., Nurvianto, S. & Harrison, R. (2010) Factors affecting the distribution and abundance of Asplenium nidus L. in a tropical lowland rain forest in peninsular Malaysia. Biotropica, 42, 464‐469. Zimmerman, J.K. & Olmsted, I.C. (1992) Host tree utilization by vascular epiphytes in a seasonally inundated forest (Tintal) in Mexico. Biotropica, 24, 402‐407. Zotz, G. & Bader, M. (2009) Epiphytes in a changing world: Global change effects on vascular and non‐vascular epiphytes. Progress in Botany, 70, 47‐70. Zotz, G. & Hietz, P. (2001) The physiological ecology of vascular epiphytes: current knowledge, open questions. Journal of Experimental Botany, 52, 2067‐2078. Zotz, G. (2005) Vascular epiphytes in the temperate zones–a review. Plant Ecology, 176, 173‐183. 138 Appendix 1 The vascular epiphyte checklist of Taiwan
(Abbreviations: E: Epiphyte, FacuE: Facultative epiphyte, HemiE-P: Primary hemi-epiphytes,
HemiE-S: Secondary hemi-epiphyte, EndemicF: endemic species in Taiwan, EndemicL:
endemic species in Lanyu, EndemicG: endemic species in Lutao, floristic codes refer to Fig. 2
in Chapter2). no Family Species/ taxon Habit Floristic_Region Pteridophytes 1 Aspleniaceae Asplenium adiantoides FacuE 15, 18, 22, 29 2 Aspleniaceae Asplenium antiquum E 2 3 Aspleniaceae Asplenium australasicum E 18, 22, 29 4 Aspleniaceae Asplenium bullatum E 16, 17 5 Aspleniaceae Asplenium cuneatiforme E EndemicF 6 Aspleniaceae Asplenium ensiforme FacuE 2‐25, 17, 16 7 Aspleniaceae Asplenium griffithianum FacuE 2‐20, 16, 17 8 Aspleniaceae Asplenium incisum FacuE 2 9 Aspleniaceae Asplenium laciniatum E 2‐27 Asplenium neolaserpitiifolium E 2‐20, 17 E 2‐20, 17, 18, 19, 20, 21, 22, 23, 29, 15, 12 10 Aspleniaceae 11 Aspleniaceae Asplenium nidus 12 Aspleniaceae Asplenium normale FacuE 2, 15, 17, 18, 20, 29, 12, 21 13 Aspleniaceae Asplenium oldhami FacuE 2‐20, 17 14 Aspleniaceae Asplenium planicaule FacuE 2, 17, 18‐104 15 Aspleniaceae Asplenium prolongatum FacuE 16, 17, 2 16 Aspleniaceae Asplenium pseudolaserpitiifolium E 17 17 Aspleniaceae Asplenium ritoense 2, 17 FacuE 18 Davalliaceae Araiostegia parvipinnata E 2‐25 19 Davalliaceae Davallia formosana E 17 20 Davalliaceae Davallia mariesii E 2 21 Davalliaceae Davallia solida E 17, 18, 22 22 Davalliaceae Humata chrysanthemifolia E 18‐104 23 Davalliaceae Humata griffithiana E 2‐27, 2‐25 24 Davalliaceae Humata pectinata E 18, 20, 29 25 Davalliaceae Humata repens E 2, 15, 17, 18, 29 26 Davalliaceae Humata trifoliata E 2‐20, 17, 18 27 Davalliaceae Humata vestita E 17, 18 ‐CONTINUED‐ 139
APPENDIX 1
28 Davalliaceae Leucostegia immersa E 2‐27, 16, 17, 18 29 Grammitidaceae Calymmodon cucullatus E 16, 18, 22, 29 30 Grammitidaceae Calymmodon gracilis E 17, 18 31 Grammitidaceae Ctenopteris curtisii E 18 32 Grammitidaceae Ctenopteris merrittii E 18 33 Grammitidaceae Ctenopteris mollicoma E 18 34 Grammitidaceae Ctenopteris obliquata E 16, 17, 18 35 Grammitidaceae Ctenopteris subfalcata E 16, 17, 18 36 Grammitidaceae Ctenopteris tenuisecta E 18 37 Grammitidaceae Grammitis adspera E 18, 29 38 Grammitidaceae Grammitis congener E 17, 18 39 Grammitidaceae Grammitis fenicis E 18‐104 40 Grammitidaceae Grammitis intromissa E 18 41 Grammitidaceae Grammitis jagoriana E 18 42 Grammitidaceae Grammitis nuda E EndemicF 43 Grammitidaceae Grammitis reinwardtia E 18 44 Grammitidaceae Prosaptia contigua E 16, 18, 19, 20, 22, 29 45 Grammitidaceae Prosaptia urceolaris E 17, 18 46 Grammitidaceae Scleroglossum pusillum E 17, 18 47 Grammitidaceae Xiphopteris okuboi E 2, 17 48 Hymenophyllaceae Abrodictyum cumingii E 2, 18 49 Hymenophyllaceae Crepidomanes bilabiatum E 2‐20, 17, 18 50 Hymenophyllaceae Crepidomanes birmanicum E 2, 17, 16 51 Hymenophyllaceae Crepidomanes kurzii E 16, 17, 18, 29 52 Hymenophyllaceae Crepidomanes latealatum FacuE 2, 16, 17, 18 53 Hymenophyllaceae Crepidomanes latemarginale FacuE 2‐20, 16, 17, 18 54 Hymenophyllaceae Crepidomanes palmifolium E EndemicF 55 Hymenophyllaceae Crepidomanes schmidtianum var. FacuE latifrons 2‐27, 18‐104 56 Hymenophyllaceae Gonocormus minutus E 2, 16, 17, 18, 20, 22 57 Hymenophyllaceae Hymenophyllum barbatum E 2, 16, 17 58 Hymenophyllaceae Hymenophyllum devolii E EndemicF 59 Hymenophyllaceae Hymenophyllum fimbriatum E 18‐104 60 Hymenophyllaceae Hymenophyllum productum E 17, 18 61 Hymenophyllaceae Hymenophyllum simonsianum E 2‐27 62 Hymenophyllaceae Hymenophyllum taiwanense E EndemicF 63 Hymenophyllaceae Mecodium badium E 2, 16, 17, 18 ‐CONTINUED‐ 140
APPENDIX 1
64 Hymenophyllaceae Mecodium javanicum E 16, 18, 19 65 Hymenophyllaceae Mecodium oligosorum E 2 66 Hymenophyllaceae Mecodium polyanthos E 2, 15, 17, 18 67 Hymenophyllaceae Mecodium wrightii E 2, 4 68 Hymenophyllaceae Meringium blandum E 18 69 Hymenophyllaceae Meringium denticulatum FacuE 2‐20, 16, 17, 18, 19 70 Hymenophyllaceae Meringium holochilum FacuE 18 71 Hymenophyllaceae Microgonium bimarginatum FacuE 2‐20, 16, 17, 18, 20, 29 72 Hymenophyllaceae Microgonium motleyi FacuE 2‐20, 16, 17, 18, 20 73 Hymenophyllaceae Microgonium omphalodes FacuE 2‐20, 18, 20, 29 74 Hymenophyllaceae Microtrichomanes nitidulum E 16, 17, 18, 29 75 Hymenophyllaceae Pleuromanes pallidum E 16, 17, 18, 20 76 Hymenophyllaceae Vandenboschia auriculata E 2, 16, 17, 18, 20 77 Hymenophyllaceae Vandenboschia maxima FacuE 2‐20, 17, 18 78 Hymenophyllaceae
E 2‐27, 2‐20, 6, 12, 16, 17, 18, 23, 24, 25, 27 79 Lomariopsidaceae Elaphoglossum callifolium E 17, 18 80 Lomariopsidaceae Elaphoglossum commutatum E 10, 12, 15, 16, 18, 21, 25 81 Lomariopsidaceae Elaphoglossum luzonicum E 18 82 Lomariopsidaceae Elaphoglossum marginatum E EndemicF 83 Lomariopsidaceae Elaphoglossum yoshinagae E 2, 17 84 Lomariopsidaceae Lomariopsis spectabilis E 2‐20, 18 85 Lycopodiaceae Lycopodium carinatum E 2‐20, 17, 18, 20, 29 86 Lycopodiaceae Lycopodium cryptomerianum E 2 87 Lycopodiaceae Lycopodium cunninghamioides E 2 88 Lycopodiaceae Lycopodium fargesii E 2 89 Lycopodiaceae Lycopodium fordii E 2, 16, 17 90 Lycopodiaceae Lycopodium phlegmaria E 2, 18, 22, 29, 15, 12 91 Lycopodiaceae Lycopodium salvinioides E 2‐20, 18‐104 92 Lycopodiaceae Lycopodium sieboldii E 2 Vandenboschia radicans 93 Lycopodiaceae Lycopodium squarrosum E 2, 20, 18, 22 94 Lycopodiaceae Lycopodium taiwanense E 2‐27, 2‐20, 16 FacuE 2‐20, 16, 17, 18, 9, 23, 24, 25, 26, 27, 22, 21, 15, 29 FacuE 2‐20, 19, 20, 18, 23, 12, 15, 16, 10, 29, 27, 25 FacuE 2‐20, 16, 17, 18‐104 95 Oleandraceae 96 Oleandraceae 97 Oleandraceae Nephrolepis auriculata Nephrolepis biserrata Nephrolepis multiflora ‐CONTINUED‐ 141
APPENDIX 1
98 Oleandraceae Oleandra wallichii E 2‐25, 2‐27, 16, 17 99 Opioglossaceae Ophioderma pendula E 17, 18, 15, 21, 29 100 Polypodiaceae Aglaomorpha meyeniana E 18‐104 101 Polypodiaceae Arthromeris lehmanni E 2, 16, 17, 18‐104 102 Polypodiaceae Belvisia mucronata E 16, 18, 20, 22, 19, 29 103 Polypodiaceae Colysis hemionitidea FacuE 2‐27, 16, 17, 18‐104 104 Polypodiaceae Colysis pothifolia FacuE 2, 16, 17, 18‐104 105 Polypodiaceae Colysis shintenensis FacuE 2 106 Polypodiaceae Colysis wrightii FacuE 2‐20, 17 107 Polypodiaceae Crypsinus echinosporus E EndemicF 108 Polypodiaceae Crypsinus engleri E 2 109 Polypodiaceae Crypsinus hastatus FacuE 2, 18‐104 110 Polypodiaceae Crypsinus quasidivaricatus FacuE 2‐27, 16 111 Polypodiaceae Crypsinus taeniatus var. palmatus FacuE 112 Polypodiaceae Crypsinus taiwanensis FacuE EndemicF 113 Polypodiaceae Crypsinus yakushimensis FacuE 2‐20 114 Polypodiaceae Drymotaenium miyoshianum E 2 115 Polypodiaceae Drynaria fortunei E 17 18, 20 116 Polypodiaceae Lemmaphyllum diversum E 2 117 Polypodiaceae Lemmaphyllum microphyllum E 2 118 Polypodiaceae Lepisorus clathratus E 2, 8, 16 119 Polypodiaceae Lepisorus kawakamii E EndemicF 120 Polypodiaceae Lepisorus kuchenensis E 2‐25 121 Polypodiaceae Lepisorus megasorus E EndemicF 122 Polypodiaceae Lepisorus monilisorus E EndemicF 123 Polypodiaceae Lepisorus morrisonensis E 2‐25, 2‐27 124 Polypodiaceae Lepisorus obscurevenulosus E 2 125 Polypodiaceae Lepisorus pseudoussuriensis E EndemicF 126 Polypodiaceae Lepisorus suboligolepidus E 2 127 Polypodiaceae Lepisorus thunbergianus E 2, 18‐104 128 Polypodiaceae Lepisorus tosaensis E 2 129 Polypodiaceae Leptochilus decurrens FacuE 16, 17, 18, 20 130 Polypodiaceae Loxogramme confertifolia E EndemicF 131 Polypodiaceae Loxogramme formosana E 2‐25 132 Polypodiaceae Loxogramme grammitoides E 2 133 Polypodiaceae Loxogramme remotefrondigera E ‐CONTINUED‐ 142
EndemicF APPENDIX 1
134 Polypodiaceae Loxogramme salicifolia E 2, 17 135 Polypodiaceae Microsorium buergerianum E 2, 17 136 Polypodiaceae Microsorium dilatatum E 2‐20, 16, 17 137 Polypodiaceae Microsorium fortunei FacuE 2‐27, 2‐20 138 Polypodiaceae Microsorium membranaceum FacuE 2‐25, 2‐27, 16, 17, 18‐104 139 Polypodiaceae Microsorium punctatum E 16, 17, 22, 29 140 Polypodiaceae Microsorium rubidum E 2‐20, 16, 17, 18, 20 141 Polypodiaceae Polypodium amoenum E 2‐27, 17 142 Polypodiaceae Polypodium argutum E 2‐25, 2‐27, 17, 18‐104 143 Polypodiaceae Polypodium formosanum E 2‐20 144 Polypodiaceae Polypodium microrhizoma E 2‐25, 2‐27 145 Polypodiaceae Polypodium raishanense E EndemicF 146 Polypodiaceae Polypodium transpianense E EndemicF 147 Polypodiaceae Pseudodrynaria coronans E 2‐20, 2‐25, 2‐27, 17 148 Polypodiaceae Pyrrosia adnascens E 2‐20, 16, 17, 18, 20 149 Polypodiaceae Pyrrosia gralla E 2‐25 150 Polypodiaceae Pyrrosia linearifolia E 2 151 Polypodiaceae Pyrrosia lingua E 2, 17 152 Polypodiaceae Pyrrosia matsudae E EndemicF 153 Polypodiaceae Pyrrosia polydactylis E EndemicF 154 Polypodiaceae Pyrrosia sheareri E 17 155 Polypodiaceae Pyrrosia transmorrisonensis E EndemicF 156 Polypodiaceae Saxiglossum angustissimum E 2 E 2, 17, 18, 21, 22, 29, 10, 12, 15, 23, 9, 3, 25, 27, 26 157 Psilotaceae Psilotum nudum 158 Selaginellaceae Selaginella delicatula E 2, 16, 17, 18, 20 159 Selaginellaceae Selaginella involvens E 2, 16, 17, 18 160 Selaginellaceae Selaginella stauntoniana FacuE 2 161 Selaginellaceae Selaginella tamariscina FacuE 2, 16, 18 162 Vittariaceae Antrophyum formosanum FacuE 2‐20 163 Vittariaceae Antrophyum obovatum FacuE 2, 16, 17 164 Vittariaceae Antrophyum parvulum FacuE 2‐20, 18 165 Vittariaceae Antrophyum sessilifolium FacuE 18‐104 166 Vittariaceae Vaginularia paradoxa E 16, 18, 20, 21 167 Vittariaceae Vaginularia trichoidea E 18, 21 168 Vittariaceae Vittaria anguste‐elongata E 18 169 Vittariaceae Vittaria flexuosa E 2, 16, 17, 18 ‐CONTINUED‐ 143
APPENDIX 1
170 Vittariaceae Vittaria taeniophylla E 2‐27, 2‐25, 17, 18‐104 171 Vittariaceae Vittaria zosterifolia E 2‐20, 18, 20 172 Araliaceae Schefflera arboricola E 17 173 Asclepiadaceae Dischidia formosana E EndemicF&L 174 Asclepiadaceae Hoya carnosa E 2, 16, 17 EndemicF Dicotyledons 175 Ericaceae Rhododendron kawakamii E 176 Ericaceae Vaccinium dunalianum var. caudatifolium E 177 Ericaceae Vaccinium emarginatum E EndemicF 178 Gesneriaceae Aeschynanthus acuminatus E 2‐27, 16, 17, 18 179 Gesneriaceae Lysionotus pauciflorus E 2 180 Gesneriaceae Lysionotus pauciflorus var. ikedae E EndemicF EndemicL 181 Melastomataceae Medinilla formosana E EndemicF 182 Melastomataceae Medinilla hayataina E EndemicL EndemicF 183 Melastomataceae Pachycentria formosana E 184 Moraceae Ficus benjamina HemiE‐P 17, 18, 29 185 Moraceae Ficus caulocarpa HemiE‐P 2‐20, 17, 18, 16 186 Moraceae Ficus heteropleura HemiE‐P 2‐27, 18, 17 187 Moraceae Ficus microcarpa var. microcarpa HemiE‐P 2‐20, 18, 17, 16, 29 188 Moraceae Ficus microcarpa var. crassifolia HemiE‐P 18‐104 189 Moraceae Ficus pumila 190 Moraceae Ficus pumila L. var. awkeotsang HemiE‐S EndemicF 191 Moraceae Ficus sarmentosa var. henryi 192 Moraceae Ficus sarmentosa var. nipponica HemiE‐S 2 193 Moraceae Ficus superba var. japonica HemiE‐P 2, 16, 17, 18 194 Moraceae Ficus virgata HemiE‐P 2‐20, 16, 17, 18, 29, 22 195 Piperaceae Peperomia japonica E 2 196 Piperaceae Peperomia nakaharai E EndemicF E 2, 23, 26, 25, 21, 12, 10, 15, 25, 29 Peperomia rubrivenosa E 18‐104 EndemicF 197 Piperaceae 198 Piperaceae HemiE‐S 2, 16 Peperomia reflexa HemiE‐S 2 199 Piperaceae Peperomia sui E 200 Piperaceae Piper arborescens HemiE‐S 18 201 Piperaceae Piper betle HemiE‐S 18 202 Piperaceae Piper interruptum var. multinervum HemiE‐S
‐CONTINUED‐ 144
18 APPENDIX 1
203 Piperaceae Piper kadsura HemiE‐S 2 204 Piperaceae Piper kawakamii HemiE‐S EndemicF 205 Piperaceae Piper kwashoense HemiE‐S EndemicL&G 206 Piperaceae Piper sintenense HemiE‐S EndemicF 207 Piperaceae Piper taiwanense HemiE‐S EndemicF 208 Rubiaceae Psychotria serpens HemiE‐S 2, 17 209 Saxifragaceae Hydrangea integrifolia E 18‐104 210 Saxifragaceae Pileostegia viburnoides E 2‐20, 16, 17 211 Urticaceae Procris laevigata E 2‐25, 15, 16, 17, 18 Monocotyledons 212 Araceae Epipremnum formosanum HemiE‐S EndemicF 213 Araceae Epipremnum pinnatum HemiE‐S 2, 18, 20, 29 214 Araceae Pothoidium lobbianum HemiE‐S 18 215 Araceae Pothos chinensis 216 Araceae HemiE‐S 2 Remusatia vivipara E 2‐25, 15, 16, 17, 18, 12, 29, 10, 25 217 Orchidaceae Acampe rigida E 2‐27, 16, 17, 18 218 Orchidaceae Appendicula fenixii E EndemicL 219 Orchidaceae Appendicula reflexa E 17, 18 220 Orchidaceae Arachnis labrosa E 17 221 Orchidaceae Ascocentrum pumilum E EndemicF 222 Orchidaceae Bulbophyllum affine E 2‐27, 16, 17 223 Orchidaceae Bulbophyllum albociliatum E EndemicF 224 Orchidaceae Bulbophyllum aureolabellum E EndemicF 225 Orchidaceae Bulbophyllum chitouense E EndemicF 226 Orchidaceae Bulbophyllum drymoglossum E 2 227 Orchidaceae Bulbophyllum electrinum E 2‐25, 17 228 Orchidaceae Bulbophyllum hirundinis E 17 229 Orchidaceae Bulbophyllum insulsum E 17 230 Orchidaceae Bulbophyllum japonicum E 2 231 Orchidaceae Bulbophyllum macraei E 2, 16 232 Orchidaceae Bulbophyllum melanoglossum E EndemicF 233 Orchidaceae Bulbophyllum omerandrum E 2 234 Orchidaceae Bulbophyllum pectenveneris E 17 235 Orchidaceae Bulbophyllum pectinatum E 17 236 Orchidaceae Bulbophyllum pingtungense E EndemicF 237 Orchidaceae Bulbophyllum retusiusculum E 2‐27, 17, 16 ‐CONTINUED‐ 145
APPENDIX 1
238 Orchidaceae Bulbophyllum riyanum E 17 239 Orchidaceae Bulbophyllum rubrolabellum E EndemicF 240 Orchidaceae Bulbophyllum setaceum E EndemicF 241 Orchidaceae Bulbophyllum taitungianum E EndemicF 242 Orchidaceae Bulbophyllum taiwanense E EndemicF 243 Orchidaceae Bulbophyllum tokioi E EndemicF 244 Orchidaceae Bulbophyllum umbellatum E 2‐27, 16, 17 245 Orchidaceae Bulbophyllum wightii E 16 246 Orchidaceae Chiloschista segawai E EndemicF 247 Orchidaceae Cleisostoma paniculatum E 17 248 Orchidaceae Cleisostoma uraiensis E 2‐20, 18‐104 249 Orchidaceae Cymbidium dayanum E 2, 16, 17, 18 250 Orchidaceae Dendrobium catenatum E 2 251 Orchidaceae Dendrobium chameleon E 18‐104 252 Orchidaceae Dendrobium chryseum E 2, 16, 17 253 Orchidaceae Dendrobium crumenatum E 16, 17, 18 254 Orchidaceae Dendrobium equitans E 18‐104 255 Orchidaceae Dendrobium falconeri E 2‐27, 16, 17 256 Orchidaceae Dendrobium furcatopedicellatum E EndemicF 257 Orchidaceae Dendrobium goldschmidtianum E 18‐104 258 Orchidaceae Dendrobium leptocladum E EndemicF 259 Orchidaceae Dendrobium linawianum E 2 260 Orchidaceae Dendrobium moniliforme E 2 261 Orchidaceae Dendrobium somae E EndemicF 262 Orchidaceae Dendrochilum uncatum E 18‐104 263 Orchidaceae Diploprora championii E 2‐27, 16, 17 264 Orchidaceae Epigeneium fargesii E 2‐27, 17 265 Orchidaceae Epigeneium nakaharae E EndemicF 266 Orchidaceae Eria amica E 2‐25, 2‐27, 17 267 Orchidaceae Eria corneri E 2‐20, 17 268 Orchidaceae Eria japonica E 2‐20, 17 269 Orchidaceae Eria javanica E 2, 16, 17, 18 270 Orchidaceae Eria ovata E 2‐20, 18 271 Orchidaceae Eria robusta E 18 272 Orchidaceae Eria tomentosiflora E 18‐104 273 Orchidaceae Flickingeria comata E 18, 29, 19, 20, 22 274 Orchidaceae Flickingeria tairukounia E EndemicF ‐CONTINUED‐ 146
APPENDIX 1
275 Orchidaceae Gastrochilus ciliaris E 2 276 Orchidaceae Gastrochilus formosanus E 2 277 Orchidaceae Gastrochilus fuscopunctatus E EndemicF 278 Orchidaceae Gastrochilus hoii E EndemicF 279 Orchidaceae Gastrochilus japonicus E 2 280 Orchidaceae Gastrochilus linii E EndemicF 281 Orchidaceae Gastrochilus matsudai E EndemicF 282 Orchidaceae Gastrochilus rantabunensis E 2 283 Orchidaceae Gastrochilus raraensis E EndemicF 284 Orchidaceae Goodyera bilamellata E EndemicF 285 Orchidaceae Goodyera pendula E 2 286 Orchidaceae Goodyera nantoensis E EndemicF 287 Orchidaceae Haraella retrocalla E EndemicF 288 Orchidaceae Holcoglossum quasipinifolium E 2 289 Orchidaceae Liparis bootanensis E 2, 17, 18 290 Orchidaceae Liparis caespitosa E 17, 18, 16, 12, 15, 19, 20 291 Orchidaceae Liparis condylobulbon E 17, 18 292 Orchidaceae Liparis cordifolia FacuE 2‐27, 2‐25, 16 293 Orchidaceae Liparis elliptica E 2, 16, 17 294 Orchidaceae Liparis grossa E 17, 18‐104 295 Orchidaceae Liparis nakaharai E EndemicF 296 Orchidaceae Liparis somai E EndemicF 297 Orchidaceae Liparis viridiflora E 2‐27, 16, 17, 18 298 Orchidaceae Luisia cordata E EndemicF 299 Orchidaceae Luisia megasepala E EndemicF 300 Orchidaceae Luisia teres E 2 301 Orchidaceae Microtatorchis compacta E 18‐104 302 Orchidaceae Oberonia arisanensis E 2‐20 303 Orchidaceae Oberonia caulescens E 2‐25, 2‐27, 17 304 Orchidaceae Oberonia gigantea E EndemicF 305 Orchidaceae Oberonia japonica E 2 306 Orchidaceae Oberonia pumila E EndemicF 307 Orchidaceae Oberonia rosea E 17 308 Orchidaceae Oberonia seidenfadenii E EndemicF 309 Orchidaceae Papilionanthe taiwaniana E EndemicF 310 Orchidaceae Phalaenopsis aphrodite E 18‐104 311 Orchidaceae Phalaenopsis equestris E 18‐104 ‐CONTINUED‐ 147
APPENDIX 1
312 Orchidaceae Pholidota cantonensis E 17 313 Orchidaceae Phreatia caulescens E 18‐104 314 Orchidaceae Phreatia formosana E 2‐25, 17 315 Orchidaceae Phreatia morii E EndemicF 316 Orchidaceae Phreatia taiwaniana E EndemicF 317 Orchidaceae Pleione bulbocodioides FacuE 2 318 Orchidaceae Pomatocalpa acuminata E EndemicF 319 Orchidaceae Schoenorchis vanoverberghii E 18‐104 320 Orchidaceae Staurochilus luchuensis E 2‐20 321 Orchidaceae Sunipia andersonii E 2‐27, 16, 17 322 Orchidaceae Taeniophyllum complanatum E EndemicF 323 Orchidaceae Taeniophyllum glandulosum E 2, 17, 18, 29 324 Orchidaceae Thelasis pygmaea E 2‐27, 16, 17, 18 325 Orchidaceae Thrixspermum annamense E 17 326 Orchidaceae Thrixspermum eximium E 18‐104 327 Orchidaceae Thrixspermum fantasticum E 2‐20, 18‐104 328 Orchidaceae Thrixspermum formosanum E 17 329 Orchidaceae Thrixspermum laurisilvaticum E 2 330 Orchidaceae Thrixspermum merguense E 17, 18 331 Orchidaceae Thrixspermum pensile E 17, 18 332 Orchidaceae Thrixspermum saruwatarii E EndemicF 333 Orchidaceae Thrixspermum subulatum E 17, 18 334 Orchidaceae Trichoglottis rosea E 18‐104 335 Orchidaceae Tuberolabium kotoense E EndemicL 336 Orchidaceae Vanda lamellata E 2‐20, 18‐104 148
Appendix 2. The epiphyte species with significant SDMs (211 spp.) and their predicted
changes in median altitude and area under two climate change scenarios (A2 and B2; IPCC,
2001). Stars indicate species endemic to Taiwan. Samples denotes the number of occurrences
used in MaxEnt, J’ = Jaccard distance index, AUC = area under the curve value, and Dico =
Dicotyledons. Bold type numbers indicate species were projected to shift downward of median
altitudes or to expand range sizes.
No. Species 1 Acampe rigida Aeschynanthus acuminatus Aglaomorpha meyeniana Appendicula reflexa Araiostegia parvipinnata Arthromeris lehmanni
Ascocentrum pumilum* 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Asplenium antiquum Asplenium australasicum Asplenium bullatum Asplenium cuneatiforme* Asplenium laciniatum
Asplenium neolaserpitiifolium Belvisia mucronata Bulbophyllum albociliatum* Bulbophyllum aureolabellum* Bulbophyllum chitouense* Bulbophyllum drymoglossum Altitude Altitude Area Area shift shift change change AUC Taxa (m) (m) (%) (%) Samples
J' J' A2 B2 A2 B2 A2 B2 14 0.90 0.66
374 259 ‐33 5 0.9611 Orchid 343 0.90 0.80
656 545 ‐51 ‐44 0.8920 Dico 15 0.85 0.59
18 7 341 118 0.9802 Fern 6 0.83 0.71
503 263 134 108 0.9968 Orchid 34 0.99 0.71
505 484 ‐86 ‐46 0.9672 Fern 182 0.87 0.74
282 370 ‐11 ‐30 0.9278 Fern 43 0.96 0.55
416 ‐65 ‐92 ‐3 0.9327 Orchid 642 0.79 0.58
‐158 ‐138 99 30 0.8306 Fern 129 0.76 0.55
287 140 95 101 0.9550 Fern 26 0.95 0.63
661 447 ‐57 18 0.9580 Fern 180 0.96 0.90
826 643 ‐75 ‐70 0.9338 Fern 83 0.99 0.76
560 358 ‐92 ‐39 0.9578 Fern 215 0.86 0.71
447 432 54 17 0.9160 Fern 7 0.99 0.87
667 584 ‐73 ‐44 0.9935 Fern 16 0.92 0.74
394 349 ‐67 ‐60 0.9745 Orchid 12 0.76 0.60
‐78 ‐121 ‐40 ‐35 0.9158 Orchid 5 1.00 0.90
496 430 ‐53 256 0.9942 Orchid 26 0.96 0.83
779 658 ‐92 ‐69 0.9074 Orchid -CONTINUED-
149
APPENDIX 2
19 Bulbophyllum electrinum Bulbophyllum hirundinis Bulbophyllum insulsum Bulbophyllum japonicum Bulbophyllum macraei Bulbophyllum melanoglossum* Bulbophyllum pectenveneris Bulbophyllum pectinatum Bulbophyllum retusiusculum Bulbophyllum setaceum* Bulbophyllum taiwanense* Bulbophyllum tokioi* Calymmodon cucullatus Calymmodon gracilis 6 1.00 1.00
‐731 717 ‐96 ‐99 0.9619 Orchid 12 0.97 0.88
860 738 ‐86 ‐60 0.9449 Orchid 7 0.94 0.84
252 574 ‐76 ‐66 0.9411 Orchid 23 0.93 0.85
682 433 ‐74 ‐74 0.9377 Orchid 31 0.90 0.80
671 457 17 ‐8 0.9458 Orchid 53 0.80 0.74
424 426 117 ‐9 0.9453 Orchid 17 0.99 0.74
716 483 ‐96 ‐53 0.9713 Orchid 31 0.78 0.55
393 350 ‐15 ‐2 0.9304 Orchid 64 0.89 0.77
720 643 3 5 0.9082 Orchid 15 0.89 0.71
317 454 ‐47 ‐37 0.9664 Orchid 5 0.82 0.74
20 ‐106 164 60 0.9824 Orchid 15 0.95 0.63
445 376 ‐71 7 0.9437 Orchid 7 0.50 0.56
‐173 ‐61 ‐22 ‐19 0.9885 Fern 11 0.95 0.91
992 935 ‐87 ‐73 0.9605 Fern 11 0.97 0.53
249 218 ‐75 ‐43 0.9548 Orchid 38 0.98 0.97
745 579 ‐3 ‐85 0.9429 Orchid 107 0.98 0.88
879 800 ‐83 ‐60 0.9401 Fern 12 0.95 0.65
‐86 391 ‐72 18 0.9486 Fern 60 0.92 0.93
547 749 ‐71 ‐88 0.9444 Fern 87 0.86 0.73
471 554 ‐54 ‐44 0.9496 Fern 39 Ctenopteris curtisii 95 0.98 0.71
625 556 ‐88 ‐38 0.9502 Fern 40 Ctenopteris merrittii 5 0.98 0.94
1072 926 ‐88 ‐69 0.9788 Fern 41 Ctenopteris obliquata
Ctenopteris 42 subfalcata Ctenopteris 43 tenuisecta 80 0.89 0.91
554 385 ‐52 ‐86 0.9399 Fern 13 0.89 0.71
436 531 ‐46 ‐54 0.9442 Fern 10 0.87 0.91
‐37 ‐751 34 ‐50 0.9949 Fern 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Chiloschista segawai*
Cleisostoma 34 paniculatum Crepidomanes 35 birmanicum Crepidomanes 36 palmifolium* Crypsinus 37 echinosporus* 38 Crypsinus engleri -CONTINUED- 150
APPENDIX 2
44 Cymbidium dayanum Cymbidium 45 floribundum 46 Davallia formosana 52 0.86 0.68
376 417 ‐4 ‐19 0.8845 Orchid 8 0.98 0.88
658 603 ‐93 ‐63 0.9690 Orchid 117 0.99 0.76
496 387 ‐86 ‐51 0.9336 Fern 47 Davallia mariesii 48 Davallia solida Dendrobium 49 chameleon Dendrobium 50 chryseum 51 Dendrobium falconeri
Dendrobium 52 linawianum Dendrobium 53 moniliforme 54 Dendrobium somae* Diploprora 55 championii 56 Dischidia formosana*
Drymotaenium 57 miyoshianum 58 Drynaria fortunei Elaphoglossum 59 callifolium Elaphoglossum 60 luzonicum Elaphoglossum 61 marginatum* Elaphoglossum 62 yoshinagae 63 Epigeneium fargesii Epigeneium 64 nakaharae* 65 Eria corneri 410 0.85 0.71
622 550 ‐26 ‐13 0.8465 Fern 28 0.68 0.78
‐53 ‐128 89 314 0.9572 Fern 21 0.92 0.85
866 617 ‐78 ‐71 0.9458 Orchid 44 0.93 0.70
432 309 ‐80 ‐59 0.9182 Orchid 17 1.00 0.82
‐36 196 ‐93 ‐75 0.9921 Orchid 6 0.89 0.90
‐518 ‐492 ‐57 ‐79 0.9730 Orchid 119 0.88 0.81
‐11 ‐242 17 15 0.9105 Orchid 17 0.85 0.60
285 88 145 24 0.9648 Orchid 50 0.79 0.73
‐246 ‐300 136 86 0.9365 Orchid 132 0.89 0.80
631 551 ‐56 ‐48 0.9028 Dico 20 0.91 0.51
256 158 ‐55 ‐6 0.9485 Fern 85 0.98 0.77
1137 545 ‐62 ‐36 0.9517 Fern 12 0.86 0.72
492 461 ‐71 ‐45 0.9650 Fern 9 1.00 0.97
‐654 ‐160 ‐96 ‐95 0.9752 Fern 33 0.87 0.69
347 523 ‐42 ‐37 0.9571 Fern 59 0.95 0.91
964 951 ‐90 ‐86 0.9429 Fern 8 0.99 0.85 ‐1497 232 ‐87 ‐54 0.9432 Orchid 29 0.88 0.60
583 344 ‐67 ‐4 0.9171 Orchid 90 0.90 0.80
746 521 ‐23 ‐10 0.9105 Orchid 66 Eria japonica 29 0.99 0.83
603 579 ‐93 ‐69 0.9162 Orchid 67 Eria ovata 75 0.75 0.69
764 508 ‐23 ‐43 0.9249 Orchid 68 Eria tomentosiflora Flickingeria 69 tairukounia* 70 Gastrochilus ciliaris 53 0.77 0.52
258 235 1 ‐22 0.9339 Orchid 6 0.99 1.00
793 715 ‐49 ‐71 0.9966 Orchid 7 0.98 0.87
283 440 ‐86 ‐66 0.9817 Orchid -CONTINUED- 151
APPENDIX 2
71 72 73 74 75 76 77 78 79 Gastrochilus formosanus Gastrochilus fuscopunctatus* Gastrochilus hoii* Gastrochilus japonicus Gastrochilus matsudai* Gastrochilus rantabunensis Gastrochilus raraensis* Gonocormus minutus Goodyera bilamellata* Grammitis congener 62 0.99 0.65
360 362 ‐95 ‐37 0.8746 Orchid 46 0.88 0.57
497 308 ‐64 ‐5 0.9207 Orchid 7 0.93 0.86
‐255 17 65 ‐73 0.9899 Orchid 37 0.83 0.74
836 781 ‐17 13 0.9184 Orchid 21 0.98 0.75
434 233 ‐96 ‐41 0.9891 Orchid 12 0.96 0.78
856 589 ‐90 ‐77 0.9352 Orchid 14 0.97 0.96
660 738 ‐92 ‐94 0.9503 Orchid 44 0.97 0.89
1197 776 ‐74 ‐75 0.8816 Fern 16 0.97 1.00
845 530 ‐93 ‐99 0.9651 Orchid 22 0.96 0.94
902 476 ‐89 ‐93 0.9549 Fern 81 Grammitis fenicis 13 0.88 0.79
908 692 ‐68 ‐59 0.9610 Fern 82 Grammitis jagoriana 5 0.84 0.80
‐385 ‐388 ‐9 ‐66 0.9887 Fern 83 Grammitis nuda* Grammitis 84 reinwardtia Holcoglossum 85 quasipinifolium 86 Hoya carnosa Humata 87 chrysanthemifolia 88 Humata griffithiana 5 1.00 1.00
944 885 ‐83 ‐90 0.9982 Fern 13 0.97 0.95
1090 1000 ‐89 ‐87 0.9850 Fern 23 0.96 0.70
611 478 ‐85 ‐45 0.9461 Orchid 376 0.75 0.46
269 217 ‐8 7 0.8848 Dico 15 1.00 0.86
549 304 ‐85 ‐33 0.9641 Fern 80 37 0.81 0.56
0 325 ‐46 ‐15 0.9108 Fern 89 Humata repens 20 0.86 0.81
342 346 ‐69 ‐67 0.9290 Fern 90 Humata trifoliata 52 0.90 0.79
956 481 ‐66 ‐61 0.9060 Fern 91 Humata vestita Hydrangea 92 integrifolia Hymenophyllum 93 barbatum Hymenophyllum 94 fimbriatum Hymenophyllum 95 simonsianum Lemmaphyllum 96 diversum 8 0.76 0.51
‐65 19 109 2 0.9504 Fern 354 0.90 0.67
671 552 ‐63 ‐39 0.8978 Dico 73 0.88 0.78
507 280 ‐64 ‐70 0.9249 Fern 10 0.78 0.61
855 600 ‐4 ‐18 0.9354 Fern 10 0.92 0.68
522 489 ‐57 ‐20 0.9258 Fern 125 0.98 0.80 ‐1080 531 ‐89 ‐65 0.9555 Fern -CONTINUED- 152
APPENDIX 2
Lemmaphyllum microphyllum 98 Lepisorus clathratus 97 922 0.85 0.78
1045 871 ‐26 ‐22 0.8504 Fern 15 0.95 0.83
‐582 22 ‐80 ‐52 0.9841 Fern 99 Lepisorus kawakamii*
67 0.82 0.53
240 239 ‐58 ‐20 0.9130 Fern 100 Lepisorus megasorus*
Lepisorus 101 monilisorus* Lepisorus 102 morrisonensis Lepisorus 103 obscurevenulosus Lepisorus 104 pseudoussuriensis* Lepisorus 105 suboligolepidus Lepisorus 106 thunbergianus 107 Lepisorus tosaensis 87 0.91 0.71
468 439 ‐42 ‐36 0.9210 Fern 170 0.80 0.63
637 629 ‐33 ‐13 0.8911 Fern 60 0.89 0.63
75 290 ‐48 ‐39 0.9464 Fern 41 0.94 0.58
598 286 ‐76 9 0.9273 Fern 87 0.89 0.57
140 102 ‐28 8 0.9491 Fern 21 0.93 0.55
132 55 ‐30 ‐5 0.9850 Fern 362 0.95 0.80
923 879 ‐68 ‐56 0.8690 Fern 30 0.90 0.58
336 146 ‐61 ‐3 0.9449 Fern 108 Leucostegia immersa 26 0.94 0.70
468 379 ‐67 ‐39 0.9582 Fern 109 Liparis bootanensis 181 0.82 0.74
‐418 ‐299 47 1 0.9208 Orchid 110 Liparis caespitosa 42 0.79 0.73
‐359 ‐404 11 ‐9 0.9186 Orchid 111 Liparis condylobulbon
30 0.81 0.68
217 231 11 9 0.9170 Orchid 112 Liparis elliptica 62 0.88 0.85
709 544 31 ‐40 0.9028 Orchid 113 Liparis grossa 10 0.94 0.79
204 147 882 275 0.9711 Orchid 114 Liparis nakaharai* 85 0.89 0.85
970 1037 ‐3 ‐15 0.8945 Orchid 115 Liparis somai* 5 0.81 0.72
227 452 232 54 0.9678 Orchid 116 Liparis viridiflora Lomariopsis 117 spectabilis Loxogramme 118 formosana Loxogramme 119 grammitoides Loxogramme 120 remotefrondigera* Loxogramme 121 salicifolia 122 Luisia cordata* 9 0.95 0.65
694 343 ‐60 ‐1 0.9427 Orchid 26 0.78 0.66
42 49 126 90 0.9428 Fern 51 0.96 0.63
525 384 ‐89 ‐50 0.9308 Fern 14 0.85 0.54
‐71 86 ‐78 ‐46 0.9038 Fern 49 0.77 0.58
121 322 ‐22 ‐37 0.9120 Fern 166 0.87 0.74
926 679 ‐33 9 0.8925 Fern 7 0.61 0.31
12 23 ‐18 ‐4 0.9763 Orchid 11 0.85 0.62
164 333 30 48 0.9584 Orchid 123 Luisia megasepala* -CONTINUED- 153
APPENDIX 2
124 Luisia teres Lycopodium 125 carinatum Lycopodium 126 cunninghamioides 127 Lycopodium fargesii 25 0.75 0.66
300 418 92 55 0.9027 Orchid 14 0.77 0.61
456 385 85 59 0.9005 Fern 8 0.80 0.60
486 410 46 13 0.9378 Fern 61 0.83 0.68
556 556 ‐42 ‐32 0.8585 Fern 191 0.81 0.68
703 576 ‐25 ‐3 0.9024 Fern 31 0.82 0.62
14 58 ‐5 ‐25 0.9232 Fern 29 0.85 0.73
‐15 188 ‐72 ‐40 0.9294 Fern 27 0.91 0.89
158 533 98 ‐74 0.9195 Fern 19 0.96 0.96
1143 1086 ‐58 ‐88 0.9479 Fern 128 Lycopodium fordii Lycopodium 129 phlegmaria Lycopodium 130 salvinioides 131 Lycopodium sieboldii Lycopodium 132 squarrosum Lycopodium 133 taiwanense 134 Lysionotus pauciflorus
21 0.86 0.80
‐575 323 ‐52 ‐68 0.9391 Fern 323 0.84 0.70
652 565 ‐38 ‐32 0.8982 Dico 135 Mecodium badium 111 0.96 0.96
1151 742 ‐81 ‐84 0.9496 Fern 136 Mecodium javanicum
Mecodium 137 oligosorum Mecodium 138 polyanthos 139 Medinilla formosana*
Microsorium 140 buergerianum Microsorium 141 dilatatum Microsorium 142 punctatum Microtatorchis 143 compacta Microtrichomanes 144 nitidulum 145 Oberonia arisanensis 16 0.92 0.86
891 766 ‐78 ‐67 0.9566 Fern 23 1.00 0.99
585 ‐162 ‐98 ‐97 0.9514 Fern 283 0.86 0.75
915 780 ‐41 ‐38 0.8960 Fern 39 0.55 0.59
‐99 227 ‐32 ‐2 0.9865 Dico 649 0.90 0.79
923 768 ‐66 ‐46 0.8478 Fern 42 0.95 0.87
866 701 ‐90 ‐77 0.9011 Fern 145 0.85 0.68
87 1 47 68 0.9151 Fern 5 1.00 0.89
830 471 ‐84 ‐15 0.9836 Orchid 11 0.95 0.80
738 603 ‐80 ‐53 0.9324 Fern 51 0.90 0.79
599 592 ‐56 ‐55 0.8553 Orchid 146 Oberonia caulescens 71 0.83 0.61
660 470 ‐48 ‐17 0.8805 Orchid 147 Oberonia gigantea* 10 0.79 0.51
535 431 ‐23 ‐12 0.9191 Orchid 148 Oberonia japonica 17 0.83 0.71
780 655 ‐15 ‐27 0.9420 Orchid 149 Oberonia pumila* 7 0.81 0.56
291 394 41 9 0.9399 Orchid 150 Oberonia rosea 5 0.61 0.41
24 11 17 ‐14 0.9888 Orchid -CONTINUED- 154
APPENDIX 2
Oberonia seidenfadenii* 152 Oleandra wallichii 151 5 0.93 0.82
257 257 169 35 0.9939 Orchid 17 0.88 0.73
‐147 ‐10 45 45 0.9567 Fern 81 0.89 0.80
621 505 ‐23 ‐38 0.9300 Fern 104 0.94 0.82
‐104 213 ‐73 ‐26 0.9430 Dico 9 0.84 0.51
158 243 ‐65 ‐45 0.9073 Dico 276 0.81 0.57
540 444 ‐17 ‐10 0.8403 Dico 28 0.76 0.56
‐117 ‐143 44 53 0.9411 Dico 209 0.83 0.58
100 282 ‐46 ‐27 0.8976 Dico 159 Peperomia sui* Phalaenopsis 160 aphrodite 161 Pholidota cantonensis
28 0.99 0.76
‐168 151 ‐97 ‐64 0.9281 Dico 6 0.85 0.77
158 70 471 317 0.9829 Orchid 45 0.95 0.92
957 701 ‐82 ‐86 0.9275 Orchid 162 Phreatia formosana 13 0.81 0.60
158 167 50 ‐29 0.8922 Orchid 163 Phreatia morii* 15 0.94 0.71
813 538 ‐71 ‐17 0.9592 Orchid 164 Phreatia taiwaniana*
Pileostegia 165 viburnoides Pleuromanes 166 pallidum Polypodium 167 amoenum 168 Polypodium argutum Polypodium 169 formosanum Polypodium 170 microrhizoma Polypodium 171 raishanense* Polypodium 172 transpianense* Pomatocalpa 173 acuminata* 174 Procris laevigata 7 0.97 0.91
1020 822 ‐83 ‐64 0.9514 Orchid 390 0.82 0.71
747 695 ‐44 ‐42 0.8901 Dico 10 0.76 0.69
578 628 ‐58 ‐53 0.9488 Fern 139 0.95 0.81
361 492 ‐56 ‐51 0.9346 Fern 100 0.96 0.86
605 447 ‐65 ‐36 0.9670 Fern 71 0.88 0.86
985 813 ‐35 ‐41 0.8687 Fern 8 0.95 0.68
253 233 ‐25 36 0.9805 Fern 50 0.75 0.64
463 554 9 ‐31 0.9210 Fern 35 0.89 0.76
288 495 ‐54 ‐40 0.9480 Fern 12 0.84 0.48
203 135 ‐31 8 0.9084 Orchid 175 0.96 0.80
830 594 ‐75 ‐52 0.9109 Dico 101 0.92 0.82
912 796 ‐51 ‐41 0.9463 Fern 418 0.92 0.77
780 583 ‐52 ‐38 0.8644 Fern 137 0.73 0.61
385 197 ‐8 ‐7 0.9230 Fern 153 Ophioderma pendula Pachycentria 154 formosana* Pentapanax 155 castanopsisicola* 156 Peperomia japonica Peperomia 157 nakaharai* 158 Peperomia reflexa 175 Prosaptia contigua Pseudodrynaria 176 coronans 177 Psilotum nudum -CONTINUED- 155
APPENDIX 2
178 Pyrrosia adnascens 136 0.82 0.66
193 233 ‐16 87 0.9079 Fern 179 Pyrrosia gralla 50 0.99 0.67
106 358 ‐97 ‐59 0.9564 Fern 180 Pyrrosia linearifolia 117 0.94 0.77
140 324 ‐60 ‐47 0.9373 Fern 181 Pyrrosia lingua 591 0.91 0.73
819 718 ‐66 ‐37 0.8259 Fern 182 Pyrrosia matsudae* 30 1.00 0.77
383 270 ‐99 ‐65 0.9643 Fern 183 Pyrrosia polydactylis*
233 0.86 0.67
416 524 ‐3 6 0.9074 Fern 184 Pyrrosia sheareri Pyrrosia 185 transmorrisonensis* Rhododendron 186 kawakamii* Saxiglossum 187 angustissimum Schoenorchis 188 vanoverberghii Scleroglossum 189 pusillum 190 Selaginella delicatula 194 0.84 0.61
641 397 ‐59 ‐35 0.9356 Fern 21 0.87 0.68
‐188 91 ‐37 ‐3 0.9732 Fern 64 0.88 0.66
549 459 ‐45 ‐28 0.9062 Dico 14 1.00 0.93
661 ‐127 ‐89 ‐89 0.9907 Fern 9 0.31 0.32
‐3 ‐17 20 4 0.9574 Orchid 9 0.95 0.96
‐742 ‐775 ‐15 ‐81 0.9820 Fern 1083 0.86 0.65
724 440 ‐25 ‐17 0.8002 Fern 413 0.87 0.70
929 749 ‐58 ‐35 0.8660 Fern 22 0.90 0.76
‐36 ‐75 524 312 0.9638 Orchid 41 0.90 0.66
297 215 ‐62 ‐41 0.9498 Orchid 5 0.94 0.81
123 494 193 15 0.9750 Orchid 12 0.66 0.47
9 56 64 ‐14 0.9236 Orchid 27 0.80 0.48
203 182 15 38 0.9344 Orchid 8 0.87 0.74
264 153 ‐38 ‐7 0.9839 Orchid 11 0.87 0.62
238 237 20 61 0.9699 Orchid 16 0.97 0.86
14 60 ‐84 ‐69 0.9467 Orchid 8 0.78 0.68
304 248 ‐7 152 0.9578 Orchid 10 0.79 0.76
554 369 7 7 0.9816 Orchid 126 0.96 0.74
856 596 ‐70 ‐29 0.9323 Dico 210 0.95 0.78
651 473 ‐86 ‐65 0.9257 Dico 191 Selaginella involvens Staurochilus 192 luchuensis 193 Sunipia andersonii Thrixspermum 194 eximium Thrixspermum 195 fantasticum Thrixspermum 196 formosanum Thrixspermum 197 laurisilvaticum 198 Thrixspermum pensile
Thrixspermum 199 saruwatarii* Thrixspermum 200 subulatum 201 Trichoglottis rosea Vaccinium 202 dunalianum var. caudatifolium* Vaccinium 203 emarginatum* -CONTINUED- 156
APPENDIX 2
204 Vaginularia paradoxa
Vandenboschia 205 auriculata Vandenboschia 206 radicans Vittaria anguste‐
207 elongata 208 Vittaria flexuosa 7 0.71 0.71
‐54 ‐18 65 88 0.9884 Fern 360 0.88 0.69
617 566 ‐55 ‐37 0.8819 Fern 33 0.78 0.69
982 725 ‐27 ‐23 0.9012 Fern 205 0.85 0.70
320 322 ‐45 ‐35 0.8983 Fern 421 0.81 0.65
418 522 ‐22 ‐30 0.8615 Fern 209 Vittaria taeniophylla 58 0.64 0.42
‐22 22 ‐33 ‐17 0.9221 Fern 210 Vittaria zosterifolia 158 0.84 0.73
631 379 ‐25 ‐22 0.9070 Fern 211 Xiphopteris okuboi 88 0.88 0.84
945 880 ‐42 ‐58 0.9617 Fern 157
158
SUMMARY
From 1997 to 1998, a strong El Niño event caused a dramatic decline of epiphytic populations
in NE Taiwan (pers. observ.). Looking back, this incident has triggered my interest in the
response of epiphytes to climate change, which is the main topic of this dissertation. At the
time of the El Niño event, for Taiwan virtually no information was available yet on the floristic
composition of the epiphyte flora, the biogeography of epiphytes, and the regional epiphyte
distribution patterns. Moreover, the ecophysiology of Taiwanese epiphytes had been little
studied, in particular in relation to global warming with presumed accompanying changes in
CO2 availability and solar insulation. Hence, conservationists were far removed from making a
dependable assessment of the impact of climatic change on epiphyte communities in Taiwan.
Taiwan is a 36,000 km2 island in East Asia (21°45'–25°56'N and 119°18'E–124°34'E).
About 70% of the island is covered by mountains of 1000 up to 3952 m asl in height, with a
dominant central range along the island’s long axis. Annual rainfall ranges from 1,000 to over
6,000 mm depending on the prevailing wind directions.
The general aim of this study is to get insight in the response of Taiwanese epiphytes
to climate change. More in detail, the following hypotheses were tested: 1) the composition of
epiphyte flora is similar to other tropical areas; 2) the epiphyte flora is a mixture of that of
adjacent floristic regions, influenced by prevailing winds; 3) epiphytes show a mid-elevation
peak in richness that is better explained by environmental factors than by the mid-domain
effect; 4) The evolution of Crassulacean Acid Metabolism (CAM) in humid forest epiphytes
occurred in response to CO2 availability; 5) in Asplenium nidus, the photosynthetic capacity is
greater for the leaf surface that receives more insolation during a day; 6) there exists
intraspecific variation of a widespread epiphytic fern Asplenium antiquum which determines its
responses to changing climate; 7) epiphyte distribution are correlated with forest types; 8)
certain epiphytic species and forest types are relatively susceptible to climate change.
To test the various hypotheses, descriptive, experimental (laboratory and field), and
modelling studies were performed. A descriptive study, based on botanical collections in
herbaria, helped to obtain insight in the current floristic composition, distribution and richness
patterns of vascular epiphytes (hypothesis 1-3, chapters 2,3). Laboratory experiments and in situ
measurements gained insight in the hypothesized evolution of CAM in response to diurnal
changes in air CO2 concentration (Hoya carnosa) and in the photosynthetic capacity of fern
leaves (Asplenium nidus) under different conditions (hypothesis 4,5, chapters 4,5). A field
experiment assessed the occurrence of adaptation of populations of a widespread epiphytic
fern (Asplenium antiquum) to simulated climate-change conditions (hypothesis 6, chapter 6).
Finally, a modelling approach was performed to assess epiphyte distribution patterns
159 SUMMARY (hypothesis 3, chapter 3) and climate change impacts on forests and associated vascular
epiphytes (hypothesis 7,8, chapter 7).
Chapter 2. Composition and phytogeography of the epiphyte flora
By consulting herbarium specimens, literature records, and field observations, an epiphyte
checklist was compiled comprising 336 vascular species (105 genera of 24 families). The
Epiphyte-Quotient (i.e. the proportion of epiphytic species) was only 8%. Presumably,
frequent tropical storms (typhoons) have contributed to the reduced epiphyte diversity in
Taiwan. Similar to epiphytic flora’s in other tropical regions, our checklist is dominated by few
families, especially ferns (171 spp) and orchids (120 spp). Epiphyte endemism was high
(21.3%), with half of the endemic species being orchids. Regarding epiphyte phytogeography,
the total epiphyte flora exhibited a similar affinity to Malesian, Eastern Asiatic and Indochinese
regions, yet epiphytic orchids shared most species with Indo-China, which likely may be
attributed to prevailing winds.
Chapter 3. Epiphyte distribution pattern and explanatory factors
Using 39,084 unique botanical records, the elevational distribution pattern of over 300
epiphytic species was explored. The result showed a richness peak between 500 and 1500m asl
that could not be explained by the mid-domain effect, suggesting environmental factors mostly
accounting for epiphyte distribution. The overall epiphyte richness patterns were modelled
using species distribution models, software MaxEnt. The modelled result not only
corroborated the position of the mid-elevation peak in epiphyte richness, it also identified two
regions with exceptionally high species richness in mid-elevations. The epiphyte hotspots are
probably related to the direction of prevailing winds. Exploratory ordination analyses indicated
two factors, elevation-related temperature and precipitation, which were most influential for
epiphyte distribution. However, subcategories demonstrated different thermal preferences; for
instance, hemi-epiphytes were most abundant in the lowland tropical forest whilst epiphytic
ferns showed a preference for increasing elevations. In contrast to predictions by the Rapoport
Effect hypothesis, the ordination analysis also showed that the degree of thermal specialisation
increased with elevation, suggesting that highland species are especially vulnerable to global
warming. Finally, in a partial ordination analysis controlling for all other variables, typhoons
were shown to exert a significant influence on the distribution of epiphytes.
160 SUMMARY Chapter 4. CO2 availability and the evolution of CAM in the
epiphyte Hoya carnosa
Twenty CAM plants of Hoya carnosa were selected to compare the acid accumulations and
stable carbon isotope ratios of their leaves under two habitat conditions. Ten in host trees that
grow in intact, dense stands of forest (closed canopies), and ten in hosts with few neighbour
trees (open canopies). We found that the air CO2 concentration was significantly higher (40-60
µmol mol-1) at night than during the day, and was higher in closed canopies than in open
canopies at night, presumably the result of host-respired CO2 added to the canopy air.
However, the carbon isotope ratio of H. carnosa was not substantially lower than those of many
other CAM plants, suggesting that the surplus CO2 released by host trees to the atmosphere at
night was not an importance CO2 source for these CAM plants. In addition, in vitro experiment
showed an appreciable daytime CO2 uptake in H. carnosa, which should even lower the carbon
isotope values of the species. Overall, the results indicated that host-respired CO2 does not
contribute CO2 budget of canopy epiphytes, hence does not support the hypothesis that CAM
has evolved in epiphytes in response to diurnal changes in air CO2 concentration rather than
water conservation.
Chapter 5. Plasticity of photosynthetic capacity in the epiphytic
fern Asplenium nidus
CO2 exchange rates of leaves in an epiphytic ferns were measured in situ to compare the
difference of photosynthetic capacity between two leaf sides in relation to sunlight exposure.
Three orientations of leaves with different patterns in sunlight exposure were selected, namely,
vertical, angled and horizontal leaves spacing from inner to outer rings in Asplenium nidus, a
fern of a rosette growth form. Except the vertically oriented leaves, the results indicated that
photosynthetic rates were higher when the side of the leaf that typically received more direct
isolation was illuminated during the measurement. Judging from equal stomatal conductances
and accompanying transpiration rates, the higher CO2 uptake rates were attributed to a greater
biochemical capacity for photosynthesis. The study revealed the physiological plasticity within
epiphytes in relation to their diverse microclimate conditions.
Chapter 6. Adaptation of a widespread epiphytic fern, Asplenium
antiquum, to simulated climate change
A two-year reciprocal transplant field experiment along an altitudinal gradient was conducted
to investigate the adaptive response of juvenile plants of the widespread epiphytic fern
Asplenium antiquum to simulated climate change conditions. The experiment results showed a
161 SUMMARY strong site effect between the three altitudinal sites at 600, 1100 and 1950 m asl on both the
growth and survivorship of juvenile A. antiquum. Under the more extreme climate conditions at
the highland site, the local population was clearly better adapted, evidenced by their
significantly higher survival than the other two populations. The results suggested that
intraspecific genetic diversity should be considered when assessing the potential impact of
climate change on species.
Chapter 7. Modelling climate change impacts on forests and
associated epiphytes
Hierarchical species distribution models (SDMs) were used to assess climate change impacts
on forests and 237 vascular epiphyte species in Taiwan. By (1) incorporating dispersal
limitation, tree persistence, and non-climatic factors into models, and (2) considering biotic
interactions between epiphytes and host trees, a novel approach was developed to improve
SDMs' accuracy and realism. The modelled results suggested that epiphyte distributions highly
depended on forest compositions. In the model results, the annual means and the variances of
the climate variables exerted an equal influence on species distributions, and non-climatic
factors tended to retain their influence under climate change conditions. Our model also
indicated certain forest types (e.g. Cypress and Picea forests) and certain thermal- or hydrosensitive species are relatively more vulnerable to projected scenarios of climate change on the
island.
In conclusion, the descriptive study of epiphytes on Taiwan has shown that its epiphyte flora is
typical for tropical island biota, having relatively low diversity compared to mainland areas, yet
showing high endemism, and sharing a low number of dominating groups, especially ferns and
orchids. The relatively low diversity and the low contribution of epiphytes to the total vascular
flora is, presumably, at least partly, explained by frequent large scale typhoon disturbances.
Modelling showed that the altitudinal epiphyte distribution pattern was mostly accounted for
by environmental factors rather than a null model of geometric constraints. However,
laboratory experiments also showed that epiphytes may have a substantial degree of
physiological plasticity in response to the diverse habitat of forest canopies. In addition, a field
experiment indicated an intraspecific genetic adaptation to elevation for a widespread species.
Information on physiological plasticity along with genetic adaptation is essential for assessing
the climate change impacts on epiphyte biodiversity. Lastly, it is concluded that the midelevation Cypress and Picea forests that have a large number of niche-specialized epiphytic
species deserve special attention for conservation purposes.
162 總結
台灣位於東亞(北緯 21°45'–25°56'N，東經 119°18'E–124°34'E)，是一個面積約
3 萬 6 千平方公尺的島嶼，中央山脈由北至南縱貫全島，約有 70%的陸地皆由海拔
1000 公尺以上的山地所覆蓋(最高峰玉山 3952 公尺)，年雨量則由 1000 至 6000 公厘不
等。筆者自 1996 年便在台灣東北部的福山保護區進行附生植物的調查，於碩士論文
研究期間適逢 1997 至 1998 的強烈聖嬰現象，觀察到樣區中附生植物族群有大規模的
乾枯死亡，也因此激發筆者想要了解未來氣候變遷將會對附生植物族群造成的影響，
然而，當時台灣對於附生植物的研究尚少，尤其是本地附生植物的組成、分布以及
生理生態方面的研究報告皆付之闕如，更無從推論附生植物在氣候變遷下，面臨溫
度、雨量甚或二氧化碳濃度變化的反應，遑論相關保育政策的制定了，因此本研究
的目標便是針對台灣附生植物進行如前所述的基礎研究，並進一步了解氣候變遷可
能對附生植物造成的影響。
本論文針對以下幾個假設加以檢視驗證: 1. 台灣的附生植物組成與其他熱帶
地區類似；2. 台灣的附生植物組成受季節風向的影響，並具有鄰近植物區系的複合
特徵；3. 台灣附生植物的多樣性在中海拔呈現高峰，此現象與環境因子的相關性勝
過地形上的中域效應(Mid-domain effect)；4. 潮濕森林中景天代謝(Crassulacean acid
metabolism)的演化起源與二氧化碳的可得性有關；5. 台灣山蘇(Asplenium nidus)吸收較
多日照的葉片，其光合作用潛力也較大；6. 海拔分布廣泛的山蘇(Asplenium antiquum)，
不同族群存在著種內變異，也對氣候變遷產生不同的反應；7. 附生植物的分布與植
被類型相關；8. 某些附生植物及植被類型對氣候變遷較為敏感。
為了測試上述的假設，本論文的研究方法包括敘述統計、野外實驗、實驗室
試驗及電腦模式預測。首先筆者根據田野觀察，編製台灣的附生植物名錄，並整理
全台灣各大標本館採集紀錄，以分析附生植物的組成、多樣性及分布型式(假設 1-3，
163 總結 章節 2 與 3)。田野及試驗室的試驗則包含：量測毬蘭(Hoya carnosa)於森林中時空間變
化下、不同二氧化碳濃度的光合代謝產物，以證實景天代謝的演化動力(假設 4，章
節 4)；量測台灣山蘇(Asplenium nidus)葉片於不同日照環境下的光合作用潛力(假設 5，
章節 5)。針對山蘇(Asplenium antiquum)的小苗進行田野試驗，以模擬氣候變遷對本物種
可能造成的影響(章節 6，假設 6)。最後以電腦模式建立台灣的附生植物分布型式(假
設 3，章節 3)以及模擬附生植物在氣候變遷下的反應(假設 7 與 8，章節 8)。
第 2 章：台灣維管束附生植物的組成及親緣地理
經檢視標本館的標本、文獻紀錄以及野外觀察，筆者編製了台灣的維管束附
生植物名錄，此名錄共包含 336 個物種，分屬於 105 個屬及 24 個科，並得出台灣的
附生植物商數(Epiphyte-Quotient，附生植物占全部植物種數的比例)為 8%，此比例明
顯低於世界平均水準的 10%，推測夏秋兩季頻繁侵襲的颱風是減低附生植物多樣性
的原因之一。與其他熱帶區域的組成類似，台灣的附生植物的組成偏重在少數分類
群，其中蕨類植物占了 171 種，而蘭科植物佔了 120 種，特有種比例偏高(21.3%)，其
中有一半都是蘭科植物。台灣附生植物的親緣地理關係與馬來西亞、東亞及中南半
島等植物區系皆有密切關係，其中蘭科植物與中南半島植物的親緣最為相近，推測
可能與夏季旺盛的西南氣流有關。
第 3 章：台灣附生植物的分布及可能的影響因子
藉由收集標本館藏及訪談植物學家，筆者編製了附生植物資料庫共包含
39084 筆採集紀錄。分析將近 300 個物種的採集紀錄，顯示物種最豐富的區間位於
500-1500 公尺的中海拔，然而此中海拔物種特別豐富的現象，經統計檢測後無法以中
域效應解釋，顯示附生植物的豐富度與環境因子較為相關。此外，筆者使用物種分
布模式工具 MaxEnt 及相關的環境因子來模擬台灣附生植物的分布，模式結果與前述
利用採集紀錄分析的結果相同，也顯示附生植物的多樣性在中海拔到達峰值，且模
164 總結 式結果同時也指出，有 2 個位於台灣北部及中部的中海拔地理區，擁有較高的附生
植物多樣性，推測這 2 個附生植物熱點的形成原因，可能與盛行風向相關。序列分
析(Ordination analysis)結果顯示，年雨量以及與海拔相關的年均溫是影響附生植物分
布最主要的因子，不過不同的分類群對溫度的偏好也有差異，例如半附生植物多生
長在低海拔的溪谷，而附生蕨類則喜愛涼爽的山地氣候。序列分析結果也指出，台
灣的附生植物對溫度敏感(Thermal specialisation)的程度，隨海拔上升而增加，此趨勢
顯然與 Rapoport 法則的假設相反，而在全球暖化的趨勢下，台灣高海拔物種可能將
遭受比較大的威脅。最後，局部序列分析(Partial ordination)顯示颱風對附生植物的分
布造成顯著的影響。
第 4 章：潮濕森林中景天代謝的演化起源與二氧化碳的可得性
本研究選取了位於 2 種不同棲地環境的毬蘭樣本共 20 株，其中 10 株附著在
茂密的森林中，另外 10 株則生長在零星分布於空曠區域的大樹上。量測顯示夜間大
氣中的二氧化碳濃度顯著高於白晝(差距約在 40-60 µmol mol-1)，且茂密森林裡的二氧
化碳濃度也高於空曠區域，推測為周遭樹木夜間的呼吸作用釋放出二氧化碳所致。
然而針對毬蘭葉片檢測碳-13，卻發現樣本內的穩定性同位素濃度並未顯著低於其它
景天代謝植物，顯示毬蘭並未充分利用宿主在夜間所釋出的二葉化碳作為碳源，此
外，控制實驗顯示，毬蘭並非嚴格的景天代謝植物，在白晝期間也會進行 C3 的光合
作用，而因此更進一步降低細胞內的碳-13 濃度。以上種種結果顯示，毬蘭雖然是分
布在潮濕森林裡的附生植物，其景天代謝的演化動力，可能還是為了降低水分散失，
而不是為了利用宿主在夜間所釋放出來多餘的二氧化碳。
第 5 章：台灣山蘇葉片光合作用潛能的可塑性(plasticity)
台灣山蘇是一種葉片呈蓮座狀排列的大型附生蕨類，本研究選取從內圈至外
圈 3 種不同傾斜角度的葉片(垂直、45 度角、水平)，野地量測不同天然光照條件下，
165 總結 葉片兩面的二氧化碳交換率。結果顯示，除了最內圈的垂直葉片以外，接收較多光
照時數的山蘇葉面基本上擁有較高的光合作用速率，由於所有量測樣本皆呈現相似
的氣孔導度(Stomatal conductances)及蒸散率(Transpiration rates)，推測差異歸因於受較
長日照葉面具有較高的生化承載力(Biochemical capacity)，本試驗結果顯示附生植物的
生理可塑性與其微環境的高度變異有極大相關性。
第 6 章：移栽(Reciprocal transplantation)試驗模擬山蘇對氣候變
遷的適應性
山蘇在台灣是一種廣布種植物，本研究以為期兩年的移栽試驗，模擬山蘇的
小苗對氣候變遷的適應性。研究結果顯示，不同海拔種源的山蘇小苗，在低、中、
高海拔(600、1100 與 1950 公尺)樣區呈現顯著不同的生長與存活率，在生長環境最為
極端的高海拔樣區，種源來自於當地的山蘇小苗具有較佳的適應力，顯然比中、低
海拔種源小苗擁有更高的存活率。本研究顯示在評估氣候變遷對物種的潛在衝擊時，
必須一併考慮種內遺傳多樣性對物種的環境適應所造成的影響。
第 7 章：利用物種分布模式(Species distribution model)模擬氣候
變遷對附生植物以及森林植群的影響
本研究改良傳統物種分布模式，採用階層式(hierarchical)的模式建立流程，來
評估在未來氣候變遷條件下，現有的森林植群以及 237 種真附生植物，將在分布上產
生何種變化? 為了增進物種分布模式於氣候變遷預測的準確性及真實性，本研究考量
樹木的傳播距離以及樹木在不適宜環境下的耐受性，此外還將地形、坡向等非氣候
性的環境因子一併納入模式建立。由於附生植物的分布受植被類型影響甚鉅，在模
擬氣候變遷下附生植物的分布時，亦將未來的森林分布作為環境因子納入模式。結
果顯示，氣候條件的平均值與季節性變異對物種分布有同樣重要的影響，而地形等
166 總結 非氣候性因子，則在氣候變遷的條件下，對物種未來的分布維持同樣的影響力。本
研究指出若干植被類型(例如：雲杉、檜木)以及某些對環境溫度及水分變化敏感的附
生植物，在未來氣候變遷的條件下將會受到較大的衝擊。
結論
歸納本研究的發現如下，台灣附生植物的組成具典型熱帶島嶼特徵，相對於大陸塊，
擁有較高的特有種比率及較低的物種數，且為少數分類群(蕨類及蘭科植物)所支配。
序列分析結果顯示，頻繁侵襲的颱風對台灣附生植物的分布有顯著的影響，影響程
度雖然還不明朗，但台灣附生植物佔整體植物誌偏低的現象可能與颱風的干擾有關。
物種分布模式結果顯示，台灣附生植物的海拔分布，主要受環境因子(溫度及雨量)的
影響，而非單純幾何限制所造成空間分布上的中域效應。然而本研究也以實驗結果
證實，附生植物具有卓越的生理可塑性以適應樹冠層多變的棲地環境。此外，移栽
試驗顯示附生植物的廣布種，其種內的基因變異可能會影響物種對氣候變遷的適應
性，應在保育評估中納入考量。最後，本研究指出中海拔的檜木林及雲杉等霧林帶，
擁有許多分布狹隘的特有種附生植物，未來應特別關注氣候變遷對其造成的衝擊。
167 168 SAMENVATTING
Een sterk El Niño effect in de winter van 1997 had een dramatische achteruitgang van epifyten
op Taiwan tot gevolg (pers. observatie). Dit voorval heeft achteraf gezien mijn belangstelling
voor de studie van de invloed van klimaatverandering op epifytische populaties gewekt en de
resultaten van die studie vormen een belangrijk onderdeel van deze dissertatie. Ten tijde van de
El Niño gebeurtenis was er vrijwel nog geen informatie beschikbaar over de floristische
samenstellingen en verbreiding van de epifyten op Taiwan en slechts weinig studies betroffen
de ecofysiologie van epifyten, in het bijzonder in samenhang met klimaatverandering en de
veronderstelde bijhorende veranderingen in CO2 aanbod en zonneschijn. Daardoor was het
niet mogelijk om een goed onderbouwde voorspelling te doen over de invloed van
klimaatverandering op de epifyten van Taiwan.
Taiwan is een eiland in Oost Azië (21°45'–25°56’N en 119°18'–124°34'O) met een oppervlakte
van 36,000 km2. De topografie van het eiland wordt gedomineerd door een longitudinale
centrale bergketen en ongeveer 70% van het eiland is bergachtig (1000-3952 m). De jaarlijkse
neerslag varieert tussen 1000 en 6000 mm, afhankelijk van de overheersende windrichting.
Dit onderzoek beoogt inzicht te verkrijgen in de invloed van klimaatverandering op Taiwanese
epifyten. In het bijzonder worden de volgende hypotheses getest: 1) de samenstelling van de
epifytische flora is vergelijkbaar met die in andere tropische gebieden; 2) de epifytische flora
bestaat uit een mengeling van de soorten in nabijgelegen gebieden en wordt beïnvloed door de
richting van de aanlandige winden; 3) in de bergen wordt de hoogste diversiteit aan epifyten op
een middenhoogte gevonden, hetgeen beter verklaard wordt milieufactoren dan door een
neutraal verbreidingsmodel van soorten langs de helling; 4) de evolutie van Crassulacean Acid
Metabolism (CAM) in epifyten van vochtige bossen werd aangedreven door de
beschikbaarheid van CO2; 5) de fotosynthetische capaciteit in de varen Asplenium nidus is groter
aan de zonzijde van het blad dan aan de schaduwzijde; 6) de wijdverbreide epifytische varen
Asplenium antiquum vertoont intraspecifieke variatie waardoor individuele planten verschillend
reageren op klimaatverandering; 7) de verbreiding van epifyten is gecorreleerd met die van
bosformaties; 8) bepaalde epifyten en bosformaties zijn relatief gevoelig voor
klimaatverandering.
Teneinde de verschillende hypotheses te testen zijn zowel beschrijvende als experimentele
(laboratorium- en veldstudies) en modelmatige onderzoeken uitgevoerd. Een op
herbariumcollecties gebaseerde beschrijvende studie gaf inzicht in de floristische samenstelling
van de epifytische vaatplanten en hun verbreiding en diversiteitpatronen (hypothese 1-3,
hoofdstuk 2,3). Laboratorium- en veldexperimenten gaven inzicht in de veronderstelde
evolutie van CAM naar aanleiding van de dagelijkse schommeling in CO2 (Hoya carnosa) en in
169 SAMENVATTING de fotosynthetische capaciteit van varenbladeren (Asplenium antiquum) onder gesimuleerde
omstandigheden van klimaatverandering (hypothese 4,5, hoofdstuk 4,5). Een veldexperiment
werd ook ingezet om de mogelijke biologische aanpassingen van lokale populaties van de
wijdverbreide varen Asplenium antiquum te onderzoeken onder diverse gesimuleerde
omstandigheden van klimaatverandering (hypothese 6, hoofdstuk 6). Tenslotte is een
modelmatige aanpak gekozen om de verbreidingspatronen van epifyten te analyseren
(hypothese 3, hoofdstuk 3) en om de potentiële impact van klimaatverandering op
bosformaties en hun epifyten te kunnen voorspellen (hypothese 7,8, hoofdstuk 7).
Hoofdstuk 2. Samenstelling en fytogeografie van de epifytische
flora (Composition and phytogeography of the epiphyte flora)
Door middel van raadpleging van herbariumcollecties, literatuuronderzoek en veldobservaties
is een checklist van epifytische vaatplanten opgesteld met in totaal 336 soorten in 105 genera
en 24 families. Het Epifyten- Quotiënt (i.e. het aandeel van het aantal soorten epifyten in de
flora) bedroeg slechts 8%. Vermoedelijk hebben de frequente tropische stormen (tyfoons)
bijgedragen aan de relatief geringe diversiteit aan epifyten op Taiwan. Evenzo als in andere
tropische gebieden wordt de epifytische flora gedomineerd door slechts enkele groepen
planten, in het bijzonder varens (171 soorten) en orchideeën (120 soorten). De bijdrage van het
aantal endemische soorten is hoog (21.3%), waarvan de helft orchideeën. De fytogeografie van
de flora laat een vergelijkbare affiniteit met de flora’s van Maleisië, Oost Azië en de
Indonesische archipel zien, maar de epifytische orchideeën hadden veel soorten gemeen met
de flora van Indo-China wat vermoedelijk een gevolg is van de dominerende windrichting
vanuit die streek.
Hoofdstuk 3. Verklarende factoren voor de verbreidingspatronen
van epifyten (Epiphyte distribution pattern and explanatory factors)
Met behulp van de analyse van 39,084 ongedupliceerde botanische collecties en waarnemingen
is de verbreiding van 300 soorten epifyten in de bergen gedocumenteerd. Een piek in het
aantal soorten op een hoogte van 500 tot 1500 m kon niet verklaard worden door het neutrale
‘mid-domain effect’, hetgeen doet vermoeden dat vooral milieuvariabelen de verbreiding van
soorten bepalen. De potentiële verbreiding van iedere soort is daarnaast gemoduleerd met
behulp van verbreidingsmodellen (MaxEnt). De resultaten van de modellen bevestigden de
piek in diversiteit op de middenhoogte en identificeerde twee gebieden met een uitzonderlijk
hoge diversiteit. Deze epifyten ‘hotspots’ danken hun bestaan waarschijnlijk aan de optredende
windrichting. Een correspondentie-analyse liet zien dat de verbreiding van epifyten vooral
bepaald wordt door de factoren temperatuur (hoogte) en neerslag, maar dat taxonomische
groepen verschillen in hun voorkeur op de thermische hoogtegradiënt. Zo kwamen
170 SAMENVATTING hemiepifyten vooral in het laagland voor terwijl varens een voorkeur voor grotere hoogtes
hadden. In tegenstelling tot de verwachting bij het optreden van het Rapoport effect liet de
correspondentie-analyse ook zien dat de mate van thermische specialisatie van de soorten
toenam met de hoogte waardoor soorten hoog in de bergen mogelijk meer gevoelig zijn voor
opwarming van de aarde dan laaglandsoorten. Een partiële correspondentie-analyse van de
invloed van tyfoons op de verbreiding van epifyten waarbij het effect van alle overige
variabelen werd geëlimineerd liet tenslotte zien dat tyfoons inderdaad de verbreiding van
epifyten significant beïnvloeden.
Hoofdstuk 4. De koppeling tussen de evolutie van CAM in de
epifyt Hoya carnosa en de beschikbaarheid van CO2 (CO2
availability and the evolution of CAM in the epiphyte Hoya carnosa)
De ophoping van zuren en de verhouding van stabiele isotopen in het blad werden
geanalyseerd van twintig Hoya carnosa CAM planten die verzameld werden onder twee
categorieën van veldomstandigheden. Tien planten groeiden op bomen in een ongestoord
dicht bos met een gesloten kronendak en tien planten waren afkomstig uit een meer open bos.
In het gesloten bos was de CO2 concentratie ’s nachts significant hoger (40-60 µmol mol-1) dan
overdag en gedurende de nacht was de concentratie in het gesloten bos ook hoger dan in het
open bos wat toegeschreven kan worden aan de hogere ademhaling van de bomen in het
dichte bos. Desalniettemin, was de verhouding van stabiele isotopen in de Hoya carnosa planten
niet substantieel lager dan in veel andere CAM planten wat suggereert dat de gedurende de
nacht extra hoeveelheid vrijkomende CO2 geen belangrijke CO2 bron was voor deze CAM
planten. Bovendien lieten laboratorium experimenten zien dat Hoya carnosa planten overdag een
aanzienlijke hoeveelheid CO2 opnemen wat de koolstof isotoopwaarden van deze soort zelfs
zou moeten verlagen. Het ziet er dus naar uit, dat de ademhaling van de gastbomen niet
fundamenteel bijdraagt aan de CO2 huishouding van de epifyten in het kronendak en daarmee
wordt de hypothese onderbouwd dat in de evolutie van epifyten CAM is ontstaan om
waterverlies te beperken en niet als reactie op de dagelijkse schommelingen in CO2
concentratie.
Hoofdstuk 5. Plasticiteit in het fotosynthetische vermogen van de
epifytische varen Asplenium nidus (Plasticity of photosynthetic
capacity in the epiphytic fern Asplenium nidus)
Het verschil in fotosynthetische vermogen tussen de zon- en schaduwzijde van epifytische
varenbladeren werd met elkaar vergeleken door in situ de CO2 gasuitwisseling te meten.
Bladeren van Asplenium nidus vormen een rozet en afhankelijk van de positie van de bladeren in
171 SAMENVATTING het rozet werden verticale-, horizontale- en bladeren onder een hoek, met elkaar vergeleken.
De drie bladtypen komen overeen met verschillen in bloostelling aan de zon. De resultaten van
de metingen gaven aan dat de intensiteit van de fotosynthese hoger was indien de kant van het
blad die normaal meer zonlicht ontvangt in het experiment belicht werd, met uitzondering van
verticale bladeren. Omdat de stomatale weerstand en de daaraan gekoppelde
transpiratiesnelheden gelijk waren werd de hogere CO2 opnamesnelheid toegeschreven aan een
groter biochemisch fotosynthetisch vermogen. Het onderzoek liet zien dat onder verschillende
milieuomstandigheden epifyten beschikken over fysiologische plasticiteit.
Hoofdstuk 6. Adaptatie van de wijdverbreide epifytische varen
Asplenium antiquum aan gesimuleerde omstandigheden van
klimaatverandering. (Adaptation of a widespread epiphytic fern,
Asplenium antiquum, to simulated climate change)
Om het vermogen tot adaptatie van de wijdverbreide epifytische varen Asplenium antiquum aan
gesimuleerde omstandigheden van klimaatverandering te bestuderen werd een tweejarig
veldexperiment uitgevoerd waarbij kiemplanten wederkerig uitgeplant werden op drie locaties
langs een hoogtegradiënt. De resultaten lieten zien dat de drie locaties, op 600m, 1100m en
1950 m hoogte, een sterke invloed hadden op zowel de groei als de mortaliteit van de jonge
Asplenium antiquum planten. Op de hoge locatie had de lokale populatie een significant lagere
mortaliteit dan de twee andere populaties waarmee de lokale planten aantoonbaar beter
aangepast waren aan het meer extreme klimaat ter plaatse. Deze resultaten suggereren dat intraspecifieke genetische variatie meegewogen moet worden bij de evaluatie van het potentiële
effect van klimaatverandering op soorten.
Hoofdstuk 7. Een model van het effect van klimaatverandering op
bossen en hun epifyten. (Modelling climate change impacts on forests
and associated epiphyte)
Hiërarchische soort-distributiemodellen (SDM’s) werden ontwikkeld om het effect van
klimaatverandering op bossen en 237 soorten epifytische vaatplanten op Taiwan te kunnen
voorspellen. Om de nauwkeurigheid en het realisme van de het SDM’s te vergroten werd een
nieuwe manier van aanpak toegepast bij de constructie van de modellen waarbij rekening
gehouden werd met 1) de beperkingen aan de ruimtelijke verspreiding van de soorten, de
resistentie van bomen volgend op een verandering van het klimaat, met niet klimaat
gerelateerde factoren, en 2) met de biotische interactie tussen epifyten en hun waardbomen. De
gemodelleerde resultaten deden vermoeden dat de verbreiding van epifyten sterk afhankelijk
was van het type bos. De jaargemiddelden van de klimaatvariabelen en de bijhorende variaties
172 SAMENVATTING hadden evenveel invloed op de gemodelleerde verbreiding van de soorten en niet aan het
klimaat gerelateerde factoren behielden over het algemeen hun invloed bij veranderende
klimaatomstandigheden. Ons model gaf ook aan dat sommige bosformaties (e.g. Cypress and
Picea bossen) en sommige temperatuur- of droogtegevoelige soorten relatief kwetsbaar waren
voor de voorspelde scenario’s van klimaatverandering op het eiland.
Samenvattend heeft het beschrijvend onderzoek aan de epifyten laten zien dat de Taiwanese
epifytische flora niet uitzonderlijk is voor een tropisch eiland. Zo is ook op Taiwan de
diversiteit aan soorten relatief gering in vergelijking tot het vaste land, weliswaar met veel
endemische soorten, en zijn de dominante groepen, zoals varens en orchideeën, vergelijkbaar
met wat we kennen van andere tropische eilanden. De relatief lage diversiteit, zowel absoluut
als in verhouding tot de gehele flora, kan vermoedelijk in ieder geval deels toegeschreven
worden aan de veelvuldig optredende verwoestende tyfoons op Taiwan. De modellen lieten
zien dat de verbreiding van de soorten in de bergen, en de daaraan gekoppelde verschillen in
diversiteit, vooral bepaald werd door milieufactoren en in mindere mate verklaard kon worden
door een ruimtelijk beperkt neutraal model. Het laboratoriumonderzoek liet evenwel ook zien
dat epifyten een aanzienlijk fysiologisch aanpassingsvermogen ten toon kunnen spreiden aan
de verschillende milieuomstandigheden in het kronendak van het bos. Een veldexperiment gaf
bovendien aan dat een wijdverbreide soort intraspecifieke adaptatie vertoont tussen populaties
van verschillende hoogten. Om de invloed van klimaatverandering op de epifytische
biodiversiteit goed te kunnen voorspellen is informatie over de mate van fysiologische
plasticiteit en de mate van genetische adaptatie van de soorten essentieel.
De Cypress en Picea bossen op een middenhoogte in de bergen, tenslotte, herbergen een groot
aantal specialistische epifyten en verdienen daarmee de bijzondere aandacht van
natuurbeschermers.
173 I write because I don’t know what I think until I read what I have to say.
–Flannery O’Connor
174 ACKNOWLEDGEMENTS
It seems unreal to me at this moment when I finish this book filled with gratefulness. For a
time, I even wondered whether being a scientist would be a suitable profession for me.
Therefore, I have to express my sincere gratitude to my promoters, Jan, Gerard, Wil and Geert,
for your valuable advice and great support of my work, and training me to think orderly and
logically. Although most of the time I stayed in Taiwan’s forests doing my work, the spatial
distance has not been a gap between us at all. Especially for Jan, without your guidance and
endless patience, I would not have been able to make it. Thanks! And I do enjoy your
incredible stories about epiphyte research in the field. I also own special thanks to Niels Raes
and Craig Martin, for your inspiration and long term influence (in a good way) from the very
beginning of my academic career.
I appreciate the Taiwan forestry research institute (TFRI) for giving me great freedom to do
my own research. The laboratory support of spore germination provided by Chiou, W.‐L. (邱
文良) and Huang, Y.‐M. (黃耀謀) is gratefully acknowledged. My special thanks go to Yu S.‐
K.(余勝焜), Chung S.‐W. (鐘詩文), Lu P.‐F. (呂碧鳳), Chang Y.‐H. (張藝翰), and Chen, C.‐H.
(陳志輝) for sharing their observations on Taiwanese epiphytes in the field, and to Su, D.‐J. (蘇
德忠) and Yu, S.‐Y. (余偲嫣) for their assistance with the tiring fieldwork. Special thanks are
extended to Lin, T.‐C. (林登秋), Lin, K.‐C. (林國銓), Lin, S.‐H. (林信輝), Zhuang, Z.‐H. (莊志
弘) and Tung, G.‐S. (董景生) for their help with several aspects of this study.
I would like to thank Lin, S.‐H. (林淑華) (Academia Sinica) and Lee, P.‐F. (李培芬) (Spatial
Ecology Laboratory) as well as Chiu, C.‐R. ( 邱祈榮 ) (Lab natural resource investigation &
analysis) at National Taiwan University for providing climatic data layers and forest
occurrences. I am grateful for impressive herbarium collections kindly provided by the
herbarium of the Taiwan forestry research institute (TAIF), the herbarium of the biodiversity
research center, academia sinica, Taipei (HAST), the herbarium of national Taiwan university
(TAI), the national museum of natural science herbarium (TNM) and the Taiwan endemic
species research institute (TESRI).
To Brian, my life partner, thank you for tolerating my capricious mood, sometimes“weirdity”
during this period of time, and being very frank when I need your opinions. I am proud to
present this book to you now.
最後，僅以本書獻給我摯愛的父母與家人，謝謝你們!!
嘉君 exuxvvt
`tçA ECDF
Rebecca C.‐C. Hsu (徐嘉君) was born on Jan. 11th, 1974 in HsinChu, Taiwan. In 1996,
she earned a Bachelor degree in Industrial Design at National Cheng Kung University,
yet also took courses in Biology and was supervised by Prof. Kuo, Chang‐Sheng (郭長
生). The same year she continued studying Botany at National Taiwan University.
Under supervision of Prof. Kuo, Chen‐Meng (郭城孟) and Dr. Hong, Fu‐Wen (洪富文)
from Taiwan Forestry Research Institute (TFRI), she started the study on epiphyte
biomass and nutrient contents in the Fushan experimental forest, and obtained her
MSc. degree in 1998. Since then, she worked in TFRI and at the Taiwan Forestry
Bureau, assisting research projects and forest management affairs. She also worked
with a visiting scholar, Prof. dr. Craig Martin, and together they published several
ecophysiological studies on vascular epiphytes in FuShan. In 2005, she received a
scholarship from the Nation Science Council to study Sustainability and Biodiversity at
Leiden University for a second MSc degree. She was supervised by Dr. Niels Raes and
Dr. Wil Tamis and finished two research projects concerning the genus Cymbidium
(Orchidaceae) and the distribution of Taiwanese vascular epiphytes, using computer
modelling tools. During this period, she began to write a PhD proposal, which was
inspired by an El Niño event during her research in FuShan. After graduating from
Leiden University in 2007, she came back working in TFRI and accomplished the
present doctoral thesis. For her publications, please visit the website:
http://rebecca.ecogarden.tw/.
ISBN: 978-94-91407-12-3
Institute for Biodiversity
and Ecosystem Dynamics
Institute of Environmental Science
KRONENDAK