Variation in Tree Fern Stipe Length with Canopy Height: Tracking Preferred Habitat through Morphological Change

15 Pages·2000·6.9 MB·English

by  Nan Crystal Arens

MISSOURI
BOTANICAL
MM
8
American Fern Journal 90(1]:1-15 (2000) 2000
^A^DEN
LIBRARY
Canopy
Variation Tree Fern Length with
in Stipe
Through
Height; Tracking Preferred Habitat
Change
Morphological
Nan
Crystal Arens and Sanchez Baracaldo
Patricia
CA
Department 94720-3140
of Integrative Biology, University of California, Berkeley,
—
common
Abstract. Cyathea caracasana a open-habitat tree fern in the Andes of Colombia.
is
In full sun, stem growth rates are high (up to 2 cm/month) and individuals regularly produce
woody
spores. However, even the fastest growing ferns are overtopped by angiosperms after 10 to
As
15 years of natural forest regeneration. individuals are overtopped, C. caracasana produces
nearly vertical fronds with long stipes (commonly over 3 m] apparently to place the photosynthetic
We
surface into the canopy. compared stipe length and blade length and width among individuals
growing in open sites and in the understories of two regenerating forests: one with a canopy of
20-25 m, and one with a canopy of 5-8 m. Stipes and blades were shortest in open habitat and
longest in the low-canopy forest. Ferns in the high-canopy forest had intermediate measurements.
number among
Despite the change in frond length, the of primary pinnae per-frond did not differ
the habitats sampled. This suggests that elongation cues are received late in the development of
the frond. This conclusion supported by a positive relationship between stipe length and the
is
distance of the fern meristem below the canopy. Because both understory populations show stipe
elongation relative to open-hapitat ferns, the cue to elongate likely a low red/far-red wavelength
is
ratio of the light received by the apical meristem. Extraordinary elongation is probably made
possible by extra carbon resources available to low-canopy plants, which still have leaves in full
sun. This sense and response mechanism allows individual plants to produce elongated fronds as
their apical meristems are overtopped. Functionally, the long-stiped plants remain in full sun even
open
after they are overtopped, thus they "track" their preferred, habitat.
When
Plant populations respond to environmental variation in several ways.
the environment fluctuates infrequently relative to the span of individuals,
life
may
adaptive segregation produce differential response to environmental cues
among separate populations. Consider variation in light environment. Sun-
showed
adapted populations of Impatiens capensis enhanced growth in
L.
nm) nm)
response changes the of red (600-700 (700-850
to in ratio to far-red
shade-adapted populations (Dudley and Schmitt, Sim-
light relative to 1995).
sun/shade segregation in growth response to red/far-red (R/FR) has heen
ilar
many angiosperm open compared
observed species in understory)
in [as to
when
communities (Morgan and Smith, 1979). In contrast, the environment
span
varies frequently relative to the life of individuals, adaptive segregation
cannot occur and individuals must rely on morphological and physiological
(Thompson, 1991; Sultan, 1993; Ackerly and Bazzaz,
flexibility (plasticity)
may
1995; Arens, 1997). In such cases, plants change form or physiology to
suit new conditions, for example switching from sun-leaf to shade-leaf anat-
omy
(Arens, 1997). Alternatively, individuals *'track" or follow their preferred
been
moves. Conventionally, habitat-tracking in plants has applied
habitat as
it
to intergenerationa! migration (Davis et al., 1986; Webb, 1987; Davis and Sa-
AMERICAN FERN JOURNAL: VOLUME NUMBER
2
90
1 (2000)
W
80
5°N
km
100
Reserva
Natural
La
Planada
Fig. America
1.
Within
(Salzman
may
form
of habitat-tracking also be applied to differential organ growth
in
non-clonal
plants.
When
growing low-canopy secondary some
in
forest, individuals of the
tree
fem
Cyathea caracasana Domin
(Klotzsch) produce unusually
long, vertically
oriented stipes that place the photosynthetic portion of the frond
into the can-
sun. This naoer documents
full
oiu'
an example bv means
oi habitat-tracking wth
dividual plant organs.
Study and Methods
Site
Planada
Andean
mm
77°59'13"W
"N, La Planada
(Fig. receives about 4,500
1).
ARENS & SANCHEZ BARACALDO: VARIATION TREE
IN
3
of rainfall annually (cumulative data from 1982 1997) with season
to a '*dry"
from
June to August. Cloud-harvesting approximately doubles the moisture
available to plants (C. Rios, Instituto von Humboldt, Villa de Leyva, Colombia,
unpublished data). Average daily temperature with no seasonal
is ig^'C, vari-
Vegetation within
ation. the reserve is typical of Pacific slope cloud forest at
m
this elevation, with a natural canopy approximately 25 in height composed
common
of relatively few tree species. The most canopy include ^7-
trees
chornea
(Euphorbiaceae), Clusia (Guttiferae), Inga (Fabaceae], Miconia (Melas-
tomataceae), Myrica and Psidium and Otoba
(Myrtaceae),
(Myristicaceae).
Eleven described tree fern species have been reported from the reserve (Arens
and Sanchez The
Baracaldo, 1998]. majority of the reserve's 3,200 hectares
are
human
covered with primary has been by
forest that little-disturbed
activity.
Within primary forest, canopy gaps are an important part of the habitat dy-
namic. Primary turnover due gap formation and canopy was
forest to closure
3%
estimated at annually (Samper K., 1992). This suggests that at a given point
—
in the forest, a canopy gap will open every 30 years, on average well within
span most The
the life of tree ferns. reserve also contains areas of recently
abandoned pasture and secondary which represent than 50
forest, less years
of natural regeneration. Secondary was by homo-
forest identified relatively
geneous canopy
size class of trees; land-use records establish the time of aban-
donment. abandoned
Regeneration of pasture rapid, with canopies of 6
is 5 to
meters closing within 10 20 years abandonment. Land use
to after history in-
formation available the reserve permits precise estimation of regeneration-
at
time. This makes La Planada a valuable site for studies of successional dynam-
ics in the cloud forest ecosystem.
We
work was 1992 and completed
Field initiated in in 1994. identified co-
whose
occurring individuals of Cyathea caracasana apical meristems
experi-
enced three distinct habitats: open habitat, the understory of 35-year-
(1) [2)
old secondary and the understory of 10-year-old secondary
forest, (3) forest.
Open grew abandoned and
habitat plants in recently pasture cleared areas,
m
commonly Abandoned
within 1,000 of reserve buildings. pastures were in
was by
the earliest stages of forest regeneration. Vegetation characterized grass-
es, herbaceous angiosperms, moss, lycophytes, ferns, and Miconia saplings. In
open grew Understory grew
habitat, tree ferns in full sun. individuals in the
shade below the closed canopy of secondary forest [approximately 35 years of
sample "high canopy" subsequent
regeneration); this referred to as forest in
is
m
The canopy was approximately 25 height with no emergent
discussion. in
m
uncommon
Canopy gaps were and none were observed greater than
trees. 2
in diameter. The younger secondary forest represented approximately 10 years
and had
of regeneration based on land-use history records, a multi-layered
m m
(uncommonly up m)
canopy ranging from 5 to 8 to 10 in height; this
sample referred to as "low canopy" forest in subsequent discussion. Most
is
ferns in low-canopy forest placed their blades in the canopy; however, the
below
apical meristems of these individuals resided canopy. For this study,
we low-canopy had
chose only those individuals in the forest that placed
their
photosynthetic surfaces in the canopy, because they maintain open-habitat
AMERICAN FERN VOLUME NUMBER
JOURNAL:
4 90
(2000)
1
c
Fig. Frond of Cyothea caracasana showing the position measurements made
2. of
in this study.
Length of stipe was measured from A-B; length of frond blade was measured
along
the rachis
from B-C; between
distance pinnae
first-order at three standardized positions (D) along the rachis.
Blade width was measured
at the widest part of the blade E-F. Original drawing by Caroline A.E.
Stromberg.
Sanchez
as habitat-trackers. Co-occurring individuals (commonly with trunk
heights
their blades in the canopy and hmctioned
and
measured
blade number
length, length, blade width, primary pinnae
of [Cyathea
cara-
casana
and
twice-pinnate-pinnatifid), between
is distance pinnae
first-order
at
standardized
range
sam
(p 0.001). Since trunk height is commonly used as a rough proxy for age in
erns (Conant, 1976; Seller. 1981: Bittner and Breckle. 1995: iqqrI
Sfiilpr
this result suggests that each habitat sam
ARENS
& SANCPIEZ BARACALDO: VARIATION TREE
IN
5
Table
Pearson product-moment
1. correlation coefficients (R) for log-transformed frond measurements.
Stem
height, stipe length, blade length, and blade width at widest point measured in centimeters. Log
(distance) is the logarithmic transformation of average distance along the rachis between primary pinnae.
NS ^ < <
not * **
statistically significant; significant at p 0.01; significant at p 0.001.
Log Log Log Log Log
(stem
height) (stipe length) (blade length) (blade width) (distance)
Log
(stem
height)
1
Log -0.03 NS
(stipe length)
1
Log
0.34* 0.44**
(blade length)
1
Log NS NS NS
(blade width) 0.29 0.30 0.19
1
NS
Log (distance) -0.02 0.90** 0.53** 0.30 NS
1
mensurate range of age-size classes. Furthermore, there was no
statistically-
=
significant correlation between trunk height and stipe length (R -0.03, Table
Thus, differences in frond measurements among habitats were not due
1). to
populations sampled,
different age-size or age- or size-related allometric var-
iation.
Light environment experienced by individual ferns was estimated using
photo-paper (Azon non-erasable diazo sepia paper, Preprint) sensors
light
placed at the apical meristem of each tree fern (Friend, 1961; Kitajima and
Augspurger, 1989). Paper sensors were calibrated for a 24 hour interval with
a LI-COR 1000 light meter (Lambda Instruments Corp., Lincoln, NE). Photo-
was
synthetically active radiation (PAR) calculated for each paper photo sensor
increment using the regression derived from LI-COR Although
calibration data.
method
precise than instantaneous meter measurements,
this less light
is
it
does give integrated measures of light environment that can be compared
among From
measurements
individuals. the plant's perspective, integrated
are
more meaningful than instantaneous meter which
ecologically light readings,
cloud cover and time While method
are subject to the vagaries of of day. this
on
does provide useful information light quantity, it is insensitive to light
which important
quality, also ecologically (Perez-Garcia et 1994; Ritchie,
is al.,
van Hinsberg and van Tienderen,
1997; 1997).
SYSTAT
and were performed
Manipulations analyses in 5.2 (Wilkinson,
MS
and Macintosh.
1989) Excel 5.0a for
Results
We
compared frond measurements in pair-wise fashion between habitats
Mean
among
using assuming unequal variances. stipe length differed
a t-test
< < when
This held
three habitats 3A, p 0.001). result 0.001) stipe
all (Fig. (p
lengths were logarithmetically transformed to compensate for unequal vari-
cm
±
ances. Open habitat ferns possessed the shortest stipes (66.4 15.22 cm.
had
Appendix ferns living in the high-canopy forest intermediate stipe
1);
cm ± growing low-canopy produced
lengths (102.9 21.7 cm); ferns in forest
cm
±
longer and more variable stipes (207.9 49.1 cm). Blade
significantly
between low- and high-canopy samples, and low-canopy
forest
length differed
AMERICAN FERN VOLUME NUMBER
6 JOURNAL:
90
(2000)
1
300
300 B
E 250
-
250 o
B
200
o
-
200
150
- o
L^O
o
100
100
c
50
-
50
Low
Open High canopy canopy
Open High canopy Low
canopy
B
200 25
o
20 -
150
15
-
100
Q
10 -
5 TJ
50
o
5
>
< PL,
^ ^^H^^^
^ Open High canopy Low canopy Open High canopy Low canopy
13
35
o
Z 30
1-
25
I-
C
20
-
15
o
I
? 10 -
o
5
E
Open High canopy Low
canopy
Fig. Average frond measurements compared among
3. habits. Data (Appendix are averages of
1]
sampled
20 individuals per habitat; error bars are one standard deviation. A) Length of stipe from
trunk to the first primary pinna; B) length of the frond blade measured along the rachis from the
primary pinna
first to the frond tip; C) width of the frond blade at its widest point; D) average (of
measurements
three taken standard
at positions] distance along between
the rachis primary pinnae
on same
the side of the frond; E) average number of first-order pinnae per frond.
ARENS SANCHEZ
& BARACALDO: VARIATION TREE FERN LENGTH
IN STIPE
7
^ R
Log = <
(blade length) 0.45 0.001
p
Log R =
(average between <
distance first-order pinnae) 0.90 p 0.001
2.5
2
1.5
1
0.5
1.6 1,8 2 2.2 2.4
Log
length)
(stipe
Fig. 4. Logarithmically-transformed data for stipe length, blade length, and the average distance
between between
first-order pinnae. Linear relationships these parameters suggest an allometric
growth relationship between stipe length and both blade length and average distance between
pinnae.
first-order
<
and open However, was no
forest habitat (Fig. 3B, p 0.001). there significant
between open and
difference in blade length habitat high-canopy Sim-
forest.
blade width differed between low- and high-canopy and low-
ilarly, forest,
<
canopy forests and open habitat (Fig. 3C, p 0.001); blade width did not differ
and
among
between pinnae
tance first-order also differed the three habitats
(Fig.
<
number
However, pinnae
3D, p 0.001). the of first-order per frond did not
<
among
any open
differ statistically 0.001) of the three habitats: habitat
(p
mean
mean mean
27.7
27.5
=
=
(Appendix
26.9 4.0
1).
Variation in stipe length might be allometrically related to size or age of the
we
To evaluate hypothesis, performed a logarithmic transformation
plant. this
on measurement and calculated Pearson product-moment
data correlation co-
parameter combinations an
(Table allometric
efficients fR) for all 1). If rela-
would show
was such transformed data high
tionship present, correlations.
Non-significant correlations between trunk height and parameters except
all
<
blade length 0.01, Table 1) indicate that differences in frond size were
(p
not due primarily to an allometric relationship with size-age class. Moderate
correlation between stipe length and blade length (Fig. and between blade
4),
AMERICAN FERN VOLUME NUMBER
JOURNAL:
8 90
1 (2000)
30
T
T T
25
Canopy
height (m)
Fern trunk height (m)
20
B
15
'S
X
10
5
5 10 15 20 25 30 35 40
Age
(yrs)
Canopy
Fig. 5. height in regenerating secondary forest at La Planada and average stem height of
Cyatbea caracasana
plotted through time. Average canopy heights were estimated based on
plots
known
of regenerating
forest of age within the reserve. Tree fern stem heights based on average
growth
rates of plants within open and understory
habitats.
growth
parameters
blade and
length, the distance betv/een pinnae show
first-order (Table
Fig.
1, 4)
a allometric growth between
clear, relationship these parameters.
In contrast to stipe and blade lengths, the number of pinnae per
first-order
frond does not vary with environment The
invariance param-
(Fig. 3E). of this
compared
with
eter, differences in stipe, blade, and distance between
first-
order pinnae,
suggests that variation in frond dimension arise by elongation
of support structures (stipe and rachis) during the elongation phase
of fi-ond
expansion, rather than development when number
early in the pinnae
of are
set.
Discussion
among
the
(Arens
was common
cies also very along roadsides and open
other Within
habitats.
mature
caracasana produced when
forest, C. spores only growing
in forest
many
years, hi
full
an
ARENS & SANCHEZ BARACALDO: VARIATION TREE FERN LENGTH
IN STIPE
9
6
C/D
200 400 600 800 1000
Distance of meristem below canopy (cm)
Untransformed measurements
Fig. of stipe length plotted against the distance of the plant's
6.
meristem below the canopy. The linearity of this relationship suggests a direct gradient response
below low
of stipe length to the light gradient the canopy.
Sanchez Baracaldo, 1998) and reached a height of up to two meters before
woody
species formed a canopy. Individuals of C. caracasana in sunny habitat
new
produced abundant and
spores recruited sporoph3^es. In contrast, C. car-
acasana grew slowly in the understory of the closed canopy forest and seldom
produced Arens and Sanchez These
spores (Arens, 1996; Baracaldo, 1998).
caracasana open, sunny environments such
data suggest that C. prefers as
human-disturbed landscapes or forest gaps. Since mature forest canopy turn-
over high in this montane forest, long-lived ferns like C. caracasana might
is
be expected to show habitat-tracking or plastic responses to such changes in
environment.
light
showed
Growth rate data for forest ferns that individuals placing their pho-
canopy by means have
tosynthetic blades in the of long, vertical stipes statis-
growth compared open
indistinguishable rates to those in habitats (Ar-
tically
low-canopy
ens and Sanchez Baracaldo, 1998). Similarly, ferns in a forest that
do not produce stipes sufficiently long to reach the canopy had low growth
individuals growing below the closed canopy (Arens and San-
similar
rates, to
chez Baracaldo, 1998). These observations support the conclusion that stipe
Cyathea caracasana allows individual plants continue growth
elongation in to
—
and spore production as open-habitat plants to track their preferred habitat
by means morphological change.
of a
10 AMERICAN FERN JOURNAL: VOLUME NUMBER
90
1 (2000)
a
o
N
O
6
o
c
00
1000
Distance of meristem below canopy
(cm)
Untransformed
Fig. measurements
7. of stipe angle plotted against the distance of the plant's
meristem below the canopy. The
Unearity
of this relationship suggests
a direct gradient response
of stipe length to the light gradient below the low canopy.
For
this strategy to work, individual must
plants sense and respond
to their
environment How
in appropriate ways.
does Cyathea
caracasana low
detect a
canopy which Wh' y
into might extend
it frond? extreme
its
is
observed
only in the low-canopy
forest?
In angiosperms, changes
in
light quality (R/FR) can induce and
internode
Low
petiole R/FR
elongation. induced
significant internode
elongation
in Gly-
max
cine (LJ Merr. (Thomas and
Raper, Phaseolus
1985),
vulgaris
L. (Beall et
Meerb
(Morgan and
Smith,
1979), Sinapis alba and Datura
L.,
fi
stim
ulated by low R/FR
(van Hinsberg and van Low
Tienderen, R/FR
1997). pro-
duced
petiole elongation
in Citrullus lanatus (Thunb.) Matsum. &
Nakai
(Graham and
Decoteau, and
1997), Tblaspi arvense Com-
L. (Metzger, 1988).
pared with grown
plants high R/FR
in
elongated
light, structures
result pri-
marily from
lengthening
individual cells with a minor component
of addition-
al cell division in support
structures
(Beall These
et al.. 1996). data are con-
sistent with our
observation
that support structure and
(stipe rachis) length in
Cyathea
caracasana
increases produce
to fronds can
that reach the canopy
of
young
secondary
however,
forest; the
architecture of the frond (number
of
first-
order piimae) does not
chai