|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
FUNCTIONAL ECOLOGY OF THE GAMETOPHYTES AND SPOROPHTYES
OF TROPICAL FERNS
JAMES EDWARD WATKINS, JR.
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
To my parents James E. Watkins and Juanita P. Watkins for encouraging my curiosity in
the natural world, for spending hours looking for material to add to my annual
elementary-school leaf collections, and for writing sick notes to my high school principal
so that I could skip class and spend the early days of Spring in search of the elusive
Many people and organizations contributed to the successful completion of my
work. I thank my supervisory committee members (Stephen Mulkey, Michelle Mack,
Thomas Sinclair, and Pamela Soltis) for their many contributions and their patience
throughout. I also thank members of the University of Florida Ecology Group
(specifically Louis Santiago, Juan Posada, Grace Crummer, Jordan Major and Jason
Vogel) for all of their help. I am especially indebted to Jason Vogel for his countless
hours of statistical discussions and for his ability to raze my radical political ambitions.
Throughout my doctoral studies, I often relied on the sage advice of Dr. Jack Ewel be it
academic, personal, or hunting: I am grateful. I also thank Robbin Moran (New York
Botanical Garden), and Donald Farrar (Iowa State University, Ames) for always lending
an ear or helping hand when it was needed most.
Some of this work took place at La Selva Biological Station in Costa Rica and I am
indebted to the Organization for Tropical Studies for opening the doors to the tropical
world to me. I also thank my wife, Catherine Cardelus, for her immense patience and
support throughout this process. I also thank my son Santiago for teaching me what life is
really all about. Funding was provided by the National Science Foundation, the
Organization for Tropical Studies, the Mellon Foundation, the American Fern Society,
and the University of Florida Graduate School and Department of Botany.
TABLE OF CONTENTS
A C K N O W L E D G M E N T S ................................................................................................. iii
LIST OF TABLES .............. ................................................. ....... vi
L IST O F F IG U R E S .... ......................................................... .. .......... .............. vii
ABSTRACT .............. ......................................... ix
1 IN T R O D U C T IO N ............................................................................. .............. ...
2 GAMETOPHYTE ECOLOGY AND DEMOGRAPHY OF TROPICAL
EPIPHYTIC AND TERRESTRIAL FERNS ............................................................5
In tro du ctio n ....................................................................................................... .... .. 5
M materials and M methods ................................................................. ........................ 6
S tu d y S ite ................................................... 6
G am etophyte Transects ............................................................................. 7
D istu rb an ce P lots ............................................................................... 7
D em ography ....................................... ..............
G am etophyte Survival A analysis ................................... ....................................... 9
R e su lts ...................................... .......................................................10
T ran sects ...................................... ............................... ................ 10
Disturbance Plots ...................................................................... ... ...... 11
D em og rap hy ................................................................12
D isc u ssio n ............................................................................................................. 1 4
Gametophyte Distributions........................................................ 14
D ensity and Species R ichness ........................................ ......... ............... 17
Sporophyte E cology ............................................ .. ........ .............. 18
C o n c lu sio n s ........................................................................................................... 1 9
3 COMPARATIVE DESICCATION TOLERANCE OF TROPICAL FERN
GAMETOPHYTES: ECOLOGICAL AND EVOLUTIONARY
C O N SE Q U E N C E S .......................................................................... .....................2 9
Introdu action ...................................... ................................................. 2 9
M materials and M methods ....................................................................... ..................33
Spore Material and Growth Conditions...........................................................33
Desiccation Experiments ........... ......... .......................................... 33
Chlorophyll-Fluorescence M easurements.................... .............................. ..35
Statistical A analysis .......................... .......... ............... .... ..... .. 35
Results ................ ................................ ...............36
D esiccation Survey ....... .. ..... ..... .. ...... .. .. ......... .... ... ............ ... 36
D esiccation R ates ........................................ .. .. .... ........ ..... .... 37
D esiccation C ycles ........................................ .................. ........ 38
D discussion ......... ..... .... .......................................................... ......38
V aviation in VPD ............ .. ............................... ........ .... .. ............ 39
D esiccation C ycles ........................................ .................. ........ 41
C o n c lu sio n s........................................................................................................... 4 3
4 NITROGEN-15 NATURAL ABUNDANCE AND NITROGEN USE
STRATEGIES OF THE GAMETOPHYTES AND SPOROPHYTES OF
TROPICAL EPIPHYTIC AND TERRESTRIAL FERNS........................................55
Intro du action ...................................... ................... ............................ 5 5
M material and M methods .......................................................... .. ............... 57
Study Site............................................. 57
Isotopic Natural Abundance and 615N Labeled Uptake ....................................59
N utrient U ptake Calculations ........................................ ......................... 60
R e su lts................... ........... ........ ........... .............. ......... ................ 6 0
615N Natural Abundance and N concentration (mg g) ............................... 60
615N L abeled U ptake ............................................ .. .. .. ...... ........... 6 1
D isc u ssio n ...................... .. ............. .. ......................................................6 2
C o n c lu sio n s........................................................................................................... 6 7
5 CON CLU SION S ................................................................75
LIST OF REFERENCES ... ................................ .......... ..............................79
B IO G R A PH ICA L SK ETCH .......................................................................... ... 88
LIST OF TABLES
2-1 Relationship of gametophyte density and richness with three levels of
experimental disturbance and two light levels .............................. ...................21
2-2 Demographic and survival analyses for the gametophytes of 5 fern species using
the Wilcoxon test to compare survival distribution functions for different species.22
3-1 Species and life form from the initial desiccation survey ........................................44
3-2 Fv/Fm recovery results from the repeated measures ANOVA for gametophytes
exposed to three different desiccation intensities................... .................45
3-3 Fv/Fm recovery results from the repeated measures ANOVA for gametophytes
exposed to 1, 2, or 3 desiccation cycles ....................................... ............... 46
4-1 Species, life form and ecology for the natural abundance and uptake experiments 68
LIST OF FIGURES
2-1 Number of gametophytes counted and their relation to disturbance from natural
transects ........... ................................... ......... .................... 23
2-2 The percentage of fern gametophytes as influenced by type of disturbance............24
2-3 The relationship between canopy openness and gametophyte density for A.
canopy and B. terrestrial species ........._. ..... ...... .......... ........... ... ............ 25
2-4 Gametophyte densities as influenced by light and disturbance in experimental
p lo ts. ............................................................. ................ 2 6
2-5 Mean longevity (months) A) and percent gametophytes still alive and un-
recruited B) for the 25-month period of the study .................................................27
2-6 Kaplan-Meier survivorship curves A) and proportion recruiting B) of 5 species
of fern gametophytes over the 25-month study period. No data were collected
during m months 4-7 and 15-24 ............................................................................28
3-1 A) Gametophyte drying curves from 12 tropical fern species of different
habitats. Species were exposed to a VPD to 1.32KPa (-50%RH) for 45 min. B)
Depression of photochemical efficiency in the same gametophytes over a series
of decreasing thallus w ater contents ....................................... ............ ............... 48
3-2 The rate of absolute water loss relative to gametophyte size as indexed by dry
m ass ...............................................................................................4 9
3-3 A) Rate of thallus drying as calculated from 3-1 for 12 tropical fern species of
different habitats. (B) Proportional recovery of the pre-treatment dark adapted
value of Fv/Fm in these same species ........ ................................................. ............... 50
3-4 Proportional recovery of the pre-treatment dark adapted value of Fv/Fm and rate
of thallus water loss expressed as A) relative water content ((g fresh weight- g
dry weight)/g saturated weight g dry weight))* 100 and B) absolute water
content (g w et w eight/g dry w eight) ............................................. ............... 51
3-5 Fv/Fm recovery graphs for gametophytes of A) Diplazium subsilvaticum, B)
Phlebodium pseudoaureum, (c) Polypodium triserale exposed to three different
desiccation intensities: VPD-0.53kPa (20%RH), VPD-1.32kPa (50%RH), and
V P D 2 .12kP a (80% R H ) .............................................................. .....................52
3-6 Proportional Fv/Fm recovery results for gametophytes exposed to 1, 2, or 3
desiccation cycles at VPD- 1.32kPa (50%RH). .............................................. 53
3-7 Morphology in fern gametophytes is diverse and is closely related to species
ecology and phylogeny................................................... .............................. 54
4-1 Sporophtyic and Gametophytic 615N natural abundance signatures of 10 tropical
fern species. ...........................................................................69
4-2 Sporophtyic and Gametophytic (a) 615N natural abundance signatures and (b) N
concentration (mg g-1) of epiphytic, terrestrial, and hemiepiphytic tropical fern
sp e cie s. ........................................................................... 7 0
4-3 615N natural abundance signatures from the hemiepiphytic fern Lomariopsis
v e s tita ............................................................................ 7 1
4-4 Uptake curves from 615N labeled solutions................................... ............... 72
4-5 Uptake saturation values (Vmax) of each N form derived from Michaelis-Menten
functions of the data from Fig. 4-2................................... ...................... .. .......... 73
4-6 12 Uptake saturation values (Km) of each N form derived from Michaelis-
M enten functions of the data from Fig. 4-2 ............. ...............................................74
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
FUNCTIONAL ECOLOGY OF THE GAMETOPHYTES AND SPOROPHTYES OF
James E. Watkins, Jr.
Chair: Stephen S. Mulkey
Cochair: Michelle Mack
Major Department: Botany
Ferns are an important part of both temperate and terrestrial floras, yet their
ecology remains poorly understood. Although ferns are dispersed by tiny wind-blown
spores, most species are limited to specific habitats; on local levels, ferns are no more
widespread than angiosperms. One aspect of fern biology that poses unique ecological
problems is the dependence on a free-living gametophyte. I examined the autecology and
ecophysiology of the fern gametophyte to understand this structure's role in shaping fern
distributions. My study showed that the gametophytes of epiphytic and terrestrial ferns
respond differently to light, disturbance, and desiccation stress, and show unexpected
versatility in nutrient relations. In all cases, such variation is closely linked to species
ecology. Selective pressures acting on the gametophyte generation may be largely
responsible for species distributions.
The ferns, with some 12,000 species, are the third most species group of land
plants following angiosperms and bryophytes. Their intermediate evolutionary position
affords the group a combination of both non-vascular and vascular plant life histories.
Ferns rely on a supposedly delicate, short-lived, and usually haploid independent
gametophyte (one of their connections to non-vascular plants), and in some, recruitment
from the gametophyte often follows into the usually diploid sporophyte stage with
lignified vascular tissue (their most obvious connection to vascular plants). It is the
ecology of the gametophyte in this choreographed alternation of generations that remains
largely unstudied. Such scientific deficit has been commented upon for decades (Pickett,
1914; Holttum, 1938; Cousens, Lacey, and Kelly, 1985; Greer and McCarthy, 1999), yet
there has been little movement to increase our understanding of basic fern ecology.
Richard Eric Holttum (1895-1990) is arguably the founding father of fern ecology.
Professor Holttum was trained at Cambridge and spent much of his time teaching and
studying the ferns of Southeast Asia. By his hand one of the first great treatises of fern
ecology was written (Holttum 1938). His seminal "The ecology of tropical pteridophytes"
first approached the topic in a synthetic way by incorporating intimate knowledge of
ferns and combining it with careful observation and questioning to get to a big-picture
conclusion about the ecology of the group. He also addressed ecology of the gametophyte
We are accustomed to see and to marvel at the great varied form and adaptation of
the sporophtyes, which are the ferns as we know them, but indeed there must be
nearly as much variety of adaptation among the gametophytes. It is true that if the
prothallus ofPlatycerium grew upon the forest floor, the resulting sporophyte, if
produced, would find itself in uncongenial surroundings, and would not develop
very far; but it is also true that the Platycerium prothallus must be able to develop
in relatively exposed position on the tree trunk in which prothalli of many ferns
would be unable to exist (Holttum 1938, pp. 421-422).
One of Holttum's key observations was the recognition of gametophyte-mediated
controls on fern recruitment. Platycerium, commonly know as staghorn ferns, are Old
World epiphytes with diverse ecology; many species grow on highly exposed emergent
canopy tree trunks, but never on the forest floor. This observation combined with the
copious production of wind-dispersed spores showed that dispersal is perhaps of only
modest importance in the distribution of ferns: the gametophyte played the critical role in
recruitment. Unfortunately, Holttum's call to arms was largely ignored and we have
progressed little since the publication of his work.
Many factors have limited the study of fern gametophyte ecology. Avoidance of
the gametophyte generation may have been driven by some of the comments made by
Frederick Orpen Bower (Bower, 1923) in "The Ferns". Bower saw little taxonomic value
in fern gametophytes and doubted their utility in advancing pteridology. Bower was such
a recognized figure (and the depauperate literature of the time basically supported his
ideas) that many botanists took his words to heart. Fortunately, many studies have since
incorporated gametophytic characters into phylogenies, and we have learned that
gametophytes have systematic value and have many taxonomic characters that allow for
species or morpho-type identification (Atkinson and Stokey, 1964; Nayar and Kaur,
1971; Chiou and Farrar, 1997; Watkins Jr. and Farrar, 2005).
There has been limited application of modem ecological methodology and
statistics to the study of fern gametophyte ecology. In a series of observational field and
elegant lab experiments, Pickett (1913; 1914) showed that the gametophytes of
Asplenium rhizophyllus and A. platyneuron could be long-lived and survive winter
temperatures and extreme drought. Pickett's work helped usher in a new way of thinking
about fern gametophyte ecology and later reports by Mottier (1927) and Walp (1951)
showed that the gametophytes of some temperate species were essentially indeterminate
and could grow in vitro for decades, if reproduction were prevented. This notion was
further supported by Donald Farrar who developed the now classic story of tropical
gametophytes growing independently of sporophytes in the temperate Appalachian
Mountains (Farrar, 1967, 1971; Farrar, 1998). In many cases, these species form
populations of asexually reproducing gametophytes that number in the tens of thousands
and appear to survive winter freezes and summer droughts.
In a series of papers, Michael Cousens developed the concept of gametophyte safe
sites and showed that multiple factors act at the level of the gametophyte to shape their
distribution and recruitment (Cousens, 1979, 1981; Cousens, Lacey, and Kelly, 1985;
Cousens, 1988; Cousens, Lacey, and Scheller, 1988). Cousens' work largely developed in
the backdrop of earlier studies showing that fern gametophytes have a distinct ecology
relative to sporophytes.
From the turn of the century with Pickett's work, to more recent work by Cousens
and Farrar, we have learned that gametophytes in temperate forests have a distinct
ecology relative to sporophytes, that gametophytes can be long-lived in natural and
especially in vitro settings, and robust when dealing with abiotic stresses. Yet, the
ecology of tropical gametophytes remains essentially unstudied. The goal of this
dissertation is to examine multiple aspects of fern ecology, focusing on the gametophyte
GAMETOPHYTE ECOLOGY AND DEMOGRAPHY OF TROPICAL EPIPHYTIC
AND TERRESTRIAL FERNS
Ferns are conspicuous components of temperate and especially tropical wet
forests. Yet, general fern ecology is poorly understood. Much early work was anecdotal
or derived from studies and observations made from sporophytes or comments obscured
in floristic inventories. A flurry of recent studies attempted to describe both the patterns
of fern sporophyte diversity and the causal relations behind such patterns (Tuomisto and
Ruokolainen, 1994b; Tuomisto and Dalberg, 1996; Tuomisto, Poulsen, and Moran, 1998;
Tuomisto and Poulsen, 2000; Jones et al., 2006; Watkins et al., 2006). These studies were
critical in developing ecological models to better understand the biology of the fern
sporophyte. Yet, focusing on sporophyte ecology, only told us a small part of fern
ecology. Missing are studies on the ecology of the free-living gametophyte.
Ferns alternate between two independent generations: the haploid gametophyte
and the diploid sporophyte. The gametophyte is a fundamentally different organism than
the sporophyte. It lacks vascular tissue, produces rhizoids instead of true roots, has poorly
developed to non-existent cuticles, and is comparatively small. We know from early work
that gametophytes can be more widespread and can grow in areas that are uninhabitable
to sporophtyes. Yet, recruitment happens in the gametophyte generation, and the resulting
sporophyte distributions depend on gametophyte ecology.
Gametophyte biology is complex, and ontogeny and morphology vary
tremendously among species (Atkinson and Stokey, 1964; Nayar and Kaur, 1971). A
common observation is that there are the apparent fundamental differences in
morphology and potentially longevity between epiphytic and terrestrial species (Dassler
and Farrar, 1997, 2001). Epiphytic species often produce gametophytes with diverse
morphologies that are frequently capable of asexual reproduction and are potentially
long-lived. Most terrestrial species are thought to produce the short-lived, textbook
cordate thallus and exhibit little ability to reproduce asexually (but see Watkins and
Farrar 2005). Little quantitative data have been generated to back up such longevity
claims, and I have been unable to find a single paper that describes factors that influence
the distribution and mortality of tropical gametophtyes.
The goal of our study was to examine the causal mechanisms of the distribution of
fern gametophytes and the demography of several tropical epiphytic, hemiepiphytic, and
terrestrial species. We examined the distribution of epiphytic and terrestrial gametophytes
and ask what in situ factors control gametophyte establishment. Then we examined the
gametophyte demography of 5 species from different habitat types, to assess gametophyte
survival and recruitment rates and relate these to life history.
Materials and Methods
This study was conducted at La Selva Biological Station (Heredia Province) in the
Atlantic lowlands of northeastern in, Costa Rica. La Selva is a 1400 ha tropical wet forest
having a mean annual rainfall of about 4,300 mm, with peaks of precipitation in June-
July and November-December, and a drier period in March. Mean monthly rainfall
nevertheless never falls below 150 mm in any month during the dry season based on
long-term meteorological records.
In order to describe the occurrence of gametophytes in nature, 425 plots of
25 cm x 25 cm were placed along 50 randomly chosen terrestrial 50m transects, and 425
25 cm x 25 cm canopy plots were placed in 9 canopy trees. The total number of
gametophytes (irrespective of identification) was counted and recorded. Each quadrat
was coded for level of disturbance: 0=undisturbed (<5cm2 bare substrate); 1 = low
disturbance (i.e. >5cm2 bare substrate); 2 = medium disturbance (>5cm2 bare substrate
and substrate disturbed); 3= high disturbance (100% bare substrate and substrate turned
over). Additionally, each quadrat with a disturbance rating of Level 1 and above was
coded for the type of damage when possible. To assess light environment, a digital
hemispherical photograph was taken with a Nikon Coolpix 950 digital camera (Melville,
NY, U.S.A.) with a fisheye lens attachment, then analyzed using Gap Light Analyzer
software (Frazer et al., 1999) to estimate the percentage of total light transmittance.
Photos were taken 25 cm above each quadrat. To determine the influence of light
environment on density, both number of gametophytes and percent canopy transmittance
were log transformed and analyzed by regression analysis. We used ANOVA to
determine the influence of level of disturbance on gametophyte density. Unless otherwise
stated, all analyses were performed with the computer program JMP version 5.01 (SAS-
To better understand the influence of disturbance and light on terrestrial fern
establishment, 20 disturbance plots were established and monitored for gametophyte
density at 5 months post establishment. All plots were established on the same soil type
in primary forest, with 10 plots placed in low-light understory habitats and 10 placed in
high-light canopy gaps of similar age. Each plot measured 1m2 and was divided into for
0.5 m x 0.5 m subplots of increasing disturbance that were similar in degree to those
found in nature. The undisturbed treatment subplot acted as the control, and no leaf litter
was removed. For low disturbance level, we removed all leaf litter with no mechanical
damage to the soil. The medium disturbance level was raked with a metal sand rake to
disturb the first 5 cm of soil. The high disturbance level was physically turned over with a
shovel, to a depth of approximately 20 cm. Gametophyte density and diversity were
recorded in the center 25 cm2 area. Litter-fall was removed from the disturbed plots
weekly, and after 5 months, plots were assessed for density (and when possible, diversity
of gametophytes). Light environment was determined with digital photography as
discussed above. Determination of gametophyte identity was difficult, and individuals
were thus lumped into "types" that in actuality may represent multiple species. Types
were assigned based on morphological characters that were identifiable by the use of a
10X or 20X hand lens and were identified and organized based on: trichome presence and
type, rhizoid color, gametophyte shape, and the presence and morphology of gemmae.
Identification of fern species from gametophytic characters was complicated and should
be taken as a conservative estimate of actual species richness. Only those gametophytes
that were mature were counted. A 2X4 full factorial ANOVA was used to determine the
effects of both light and disturbance intensity on gametophyte density and diversity.
In June of 2003, gametophytes from three populations of each of 5 species (See
table 2-2) were located and marked in the field. Marked individuals were checked once
each month and followed for the next 15 months with one final census made at 25
months. No data were recorded during the 5-7th month. At each census, individuals were
recorded as present, dead or missing, or as recruits into the sporophyte generation. When
possible, individuals were coded for their cause of mortality.
In the case of terrestrial species, gametophytes were marked with a numbered
aluminum nail; whereas, the epiphytic species were either marked with a nail or with a
numbered tag attached to the substrate with copper wire. It was not possible in all cases
to determine precisely the initial age of marked gametophytes. Therefore, individuals
were chosen according to their initial size. Initial sizes were held constant within a
species but differed among the species. Longevities were calculated as the time between
the initial mark (treated as birth) and death of each gametophyte. Species were chosen to
represent different functional types as discussed below. A major flood event took place in
month 11; individuals were sampled three days before the flood (for the regularly
scheduled 11 month survey) and then three days after the flood to serve as an extra
survey to determine the influence of flooding. The next sample period took place on the
next corresponding survey day and was recorded as the month 12 survey period. This
allowed for more precise determination of mortality due to flooding rather than
categorizing these individuals in the unknown category.
Gametophyte Survival Analysis
As with many demographic studies, individuals can leave the study by different
avenues. Such absent samples were coded as right-censored data points (Hollander and
Wolf, 1999). In this study, only those individuals that recruited into the sporophyte
generation and those still alive at the end of the experiment (25 months) were recorded as
censors. Censoring individuals reduces the sample size of individuals at risk after the
time of censorship. Censoring, therefore, reduces the number of individuals contributing
to the curve, and each death after a censored point represents a higher proportion of the
remaining population. Subsequent deaths will result in greater decreases in overall
survivorship. Censorships that occur early in the study have a greater effect on
survivorship curves than those removed at later periods. Thus, the data from the
survivorship curve after the first censor represent an estimate and not the actual
survivorship of the population. In order to clarify the survival curves, we also plotted the
cumulative proportion of individuals that recruited at each time interval.
Gametophyte survival functions were estimated using non-parametric Kaplan-
Meier product-limit survival functions (Collett, 2003). These analyses were also used to
estimate mean life span for each species. Log-rank X2 statistics were computed to test for
homogeneity of the survival functions for all species. Weibull distributions were used to
model survivorship functions and to calculate the parameters a and P. The scale
parameter a is a measure of the degree of hazard for the species; whereas, the shape
parameter 0 determines the degree of change in the hazard function over time. Large
values of a correspond to low hazard levels (i.e. greater survivorship) where low values
equate to rapidly decaying survivorship. Large values of 0 (i.e. >1) correspond to an
increasing hazard rate that affects older individuals over younger individuals. With a P <1
younger individuals are more likely to die within the period of the experiment.
A combined total of 2096 gametophytes were sampled with 329 recorded from
canopy, 538 from low-trunk, and 1229 from terrestrial habitats. Level of disturbance had
a highly significant effect on the number of gametophytes in terrestrial habitats (Fig. 2-
2b, r2=0.65, F=53.58, p<0.001) with greater percentages of gametophytes occurring in
more disturbed habitats. Less than 1% of terrestrial gametophytes were found in areas
without disturbance. The opposite trend was apparent in canopy habitats where level of
disturbance had less influence on numbers of gametophytes (Fig. 2-2a, r2=0.11, F=2.40,
p=0.08) and greater percentages of gametophytes occurred in less disturbed habitats (58%
of canopy gametophytes were found in areas with no disturbance).
In terrestrial transects, seven causes of disturbance were identified: leaf-litter
removed, new and old root tip-ups from fallen trees, rotten logs, erosion, branch falls, and
animal causes (Fig. 2-2b). A total of six causes of disturbance were identified in canopy
habitat: insect, branch falls, epi-slides, physical damage, animal, and unknown causes
(Fig. 2-2a). Identification of causes in the low-trunk habitat was difficult and thus was
excluded from all analyses. In the case of terrestrial species, recent root tip-ups harbored
the greatest number of gametophytes (>50%). The disturbance category with the greatest
percentage of gametophytes in canopy habitats was animal disturbance with -20%. In
addition to disturbance, canopy openness (as a surrogate for light level) exhibited a
positive effect on the number of terrestrial gametophytes (Fig. 2-3a, r2=0.423,
p=<0.0001), but exhibited little influence on canopy gametophyte density (Fig. 2-3b,
A total of 1247 gametophytes from 16 morpho-types were counted in the
experimentally disturbed plots. There were 6 non-unique morpho-types found in the low
light treatment. There were a total of 16 types with 10 unique types in the high light
treatments. There was a significant and positive effect of both increasing light and
disturbance and the interaction of light and disturbance on gametophyte density among
the plots (Fig. 2-4, Table. 2-1). Likewise, morpho-type richness was significantly
influenced by both light and disturbance but exhibited no significant interaction (Fig. 2-4,
All combined, 809 gametophytes from the five species were marked and followed
throughout the demography study. A total of 263 gametophytes were marked from the
understory terrestrial species Danaea wendlandii. The three populations of this species
were all recorded in the understory of primary forests from sites that were at least 50m
from trail sides. We marked 275 gametophytes of Pityogramma ebenea, an abundant
species often found in full to partial sun in disturbed sites such as road and trail sides. All
populations of this species were recorded from disturbed sites within the forests or in
open areas away from trail sides. Sixty-seven gametophytes of the understory
hemiepiphyte Lomariopsis vestita were marked from small diameter trees in primary
forests. Two canopy epiphytes were also marked: 98 from the high-light epiphyte Vittaria
lineata, and 106 from the medium-light understory epiphyte Campyloneurum
Survival distribution functions varied significantly among species for the entire
survey period (Table 2-2, log-rank X2 = 386.2, d.f. = 4, p < 0.0001). Campyloneurum
brevifolium had the highest mean longevity and was 5 times that of the lowest:
Pityrogramma ebenea (Fig. 2-5a, 2-6a). When combined, the epiphytic species: C.
brevifolium, Lomariopsis vestita, and Vittaria lineata had higher mean longevities than
the terrestrial species (log-rank 2 = 212.3, d.f. = 1, P < 0.0001). In all cases 0 >1,
indicating an increasing hazard rate suggesting that older individuals are more likely to
die than younger individuals over the study period. Percent recruitment varied among the
species, with C. brevifolium < V. lineata < D. wendlandii < L. vestita ~ P. ebenea. The
cumulative proportion of gametophytes recruiting varied for all species over the sampling
time of the study (Fig. 2-6b). Initial recruitment was highest for both terrestrial species.
More than 30% of gametophytes of P. ebenea had recruited between plot establishment
and the first census. No additional individuals recruited beyond month eight. Initial
recruitment was lower for D. wendlandii, but increased throughout the study period.
Recruitment was lowest for V lineata and C. brevifolium with essentially no recruitment
occurring after the third census up until the 25th month (Fig. 2-6b). The percent of
gametophytes still alive at the end of the study also varied from 0% in P. ebenea to just
over 70% in C. brevifolium (Fig. 2-5b).
A total of 7 causes of mortality, including an unknown category, were surveyed in
the field. Catastrophic habitat failure occurred when habitats were over 95% of the
individuals were destroyed. This happened when entire trees fell in the case of epiphytes
or when hill sides collapsed with some terrestrial species. Flooding also resulted in
catastrophic failure, but was separated as unique disturbance type because we were able
to directly assess its influence (Fig 2-2a). Minor erosion also resulted in the loss of some
individuals as did a massive flood in month eight of the study. Fungal attack, herbivory,
and physical damage (as would occur from a branch or rock fall that physically removed
individuals from the population) were also identifiable causes of mortality. The unknown
category likely consisted of a contribution of all of these plus unidentifiable novel
disturbances. Each cause resulted in different magnitudes of mortality, with some
categories completely absent from some species and/or populations (Fig. 2-2). Most %
mortality was attributable to unknown causes such as an individual simply missing from
the population without any sign of disturbance, etc.
The present study clearly demonstrates the importance of disturbance for
gametophyte establishment. Even minor disturbances that remove leaf litter and turn up
the soil can produce sites for gametophyte establishment. Disturbance that physically
turns up soil not only produces an exposed and competition free habitat, is can also
exposes the underlying spore bank and provide additional propagules that may further
contribute to density and richness. Surprisingly, the maximum number of terrestrial
gametophytes found in the lowest level of natural disturbance was three and the majority
of the undisturbed sites had no established gametophytes.
To my knowledge, disturbance has never been reported to be an important factor
influencing gametophyte density or shaping species distributions. However, studies on
the gametophytes of temperate species have highlighted the importance of nutritional and
edaphic safe sites for the gametophytes ofLorinseria (Woodwardia) areolata (Cousens,
Lacey, and Scheller, 1988) and Blechnum spicant (Cousens, 1981). Little is known of
other factors influencing gametophyte distributions and a few temperate studies have
produced mixed results suggesting that gametophyte gender expression may influence
gametophyte distribution (Klekowski, 1969; Crist and Farrar, 1983) while others have
found no relationship (Holbrook-Walker and Lloyd., 1973; Greer and McCarthy, 1999).
Apart from these studies, little is known of the influence of these characters on tropical
fern gametophyte distributions. Numerous studies have however, examined factors
behind sporophyte distributions and have fingered important roles of microclimate
(Nobel, 1978) and water availability in dry sites (Marquez et al. 1997) and edaphic
factors (Tuomisto and Ruokolainen, 1994a; Tuomisto and Poulsen, 1996; Tuomisto,
Poulsen, and Moran, 1998) in wet tropical sites. None of these studies have directly
addressed the role of disturbance.
The nature of disturbed habitats creates a positive feedback for species that prefer
disturbed sites. By their very nature such sites are unstable and result in increased
mortality due to continued habitat erosion. Pityrogramma ebenea is perhaps the most
common species of disturbed habitats at La Selva. The species produces large numbers of
spores with high fecundity, and the gametophytes can be found in virtually any habitat
where disturbance is present, i.e. exposed road/trail cuts and the relatively dark
understory (pers. obs.). The gametophytes of this species also germinate, grow, and
recruit rapidly (Fig. 2-6). Rapid development is necessary in species that occupy highly
disturbed habitats. Catastrophic events are common in the habitat of this species, and in
two populations such disturbance resulted in the near-complete habitat destruction and in
continued habitat instability exacerbated by wet season rains. Epiphytic species are also
subject to catastrophic disturbances; a single population of Vittaria lineata experienced
this sort of disturbance following a large tree fall which resulted in 100% mortality of one
of the study populations. However, these events seem relatively rare and epiphytic
habitats tend to be more stable when compared to terrestrial habitats in this forest.
Based on these data, terrestrial gametophytes simply do not establish in sites that
are disturbance-free. However, there are varying levels of tolerance and clearly different
life histories in terrestrial species. Danaea wendlandii is a eusporangiate fern. The
eusporangiates have many unique characters, but of particular interest for this study is the
production of liverwort-like gametophytes that are several cell layers thick. Individual
gametophytes are often large and more resistant to physical damage relative to the single
layered leptosporangiate species (pers. obs.). Such tough gametophytes that occur in sites
with minor disturbance confer longevity. Indeed, the gametophytes of Danaea
wendlandii exhibited 3 times the mean longevity and significantly less recruitment than
those of P. ebenea. Two populations ofDanaea wendlandii fell in the flood zone, and
while such disturbances are clearly part of the biology of this species, this event resulted
in lower mean longevities in this study. The eusporangiate biology of this species places
it nearly opposite of P. ebenea in terms of life history strategy.
There were also surprising differences between epiphytic and terrestrial species.
One emergent difference between these two groups is the percent of gametophytes that
survived but did not recruit. Over 70% of the gametophytes of C. brevifolium, and over
50% of both V lineata and L. vestita were alive and un-recruited by the 26th month. This,
compared to the less than 5% in Danaea wendlandii and 0% in Pityrogramma ebenea.
This observation highlights fundamentally different gametophytic life history strategies
that have evolved in the two life forms. In fact, recent phylogenetic analysis of the ferns
has revealed a recent split between terrestrial and epiphytic clades in the Eu-Polypodiales
One of the defining gametophytic characters of the epiphytic clade is indeterminate and
asexually reproducing gametophytes. Dassler and Farrar (1997) have argued that such
longevity and asexual reproduction is a mechanism to encourage outcrossing in epiphytic
species that may carry significant genetic load. Long-lived thalli and thus genotypes can
produce numerous archegonia over space and time to ensure fertilization of newly
dispersed genotypes. Such differences in life history are significant and may have been
critical in the radiation from terrestrial species into canopy habitats.
Density and Species Richness
Canopy gametophyte density is clearly more sensitive to disturbance, with nearly
58% occurring in undisturbed sites. Additionally, there was not a detectable relationship
between gametophyte density and light environment as was shown for terrestrial density.
In general, canopy light environments are significantly higher than terrestrial sites and the
lack of response to light in the former is not surprising in a habitat where this light is not
limiting. However, the canopy does experience temperature and humidity extremes and it
is plausible that microclimate and water availability play larger roles in these habitats
relative to terrestrial sites. Such an observation would be in line with reports by Hietz and
Briones (1998a) who demonstrated that within canopy distribution of fern sporophytes
was largely a function of species water relations.
Quantification of gametophyte species/morphotype richness in the natural
transects was abandoned largely due to time constraints. However, there was a clear
light-disturbance-density relationship for terrestrial species, and for this reason we
examined these variables more completely in the terrestrial disturbance experiment. In
one high disturbance plot, we counted six different morphotypes with 65% of the density
dominated by Pityrogramma ebenea. There are numerous life history strategies in the
ferns, and P. ebenea is a species with high germination and recruitment rates. Indeed, it is
possible that all of the species that were encountered in the higher disturbance plots
exhibited similar life histories. Clearly there are specific differences in gametophyte
ecology that influence densities at a given site. However, the near complete absence of
gametophytes in undisturbed habitats suggests that disturbance may be critical to the
majority, if not all terrestrial species regardless of life history. Such natural observations
combined with experimental manipulations offer strong evidence that light and
disturbance both act to structure terrestrial gametophyte density and richness.
Unlike comparisons of seedling-adult distributions, fern gametophytes are
completely and fundamentally different organisms from sporophytes. For this reason,
gametophytes would not necessarily be expected to behave like sporophytes. Firstly,
gametophytes are often, if not always more widespread than sporophytes (Peck, 1980;
Peck, Peck, and Farrar, 1990). Such plasticity and simplicity in function may reduce the
role that nutrients or other edaphic factors play in gametophyte distribution. Secondly,
gametophytes can be long lived with some individuals living decades in culture (Mottier,
1927; Walp, 1951) and in years field settings (pers. obs.). Additionally, gametophytes
may be relatively sensitive to desiccation and temperature changes; however, many
studies have shown that gametophytes are relatively robust to environmental stress (Sato
and Sakai, 1980; Cousens, 1981; Sato and Sakai, 1981; Cousens, Lacey, and Scheller,
1988; Ong and Ng, 1998). While survival from stress and edaphic requirements may be
important, for all species, emergence in litter free sites seems critical.
Gametophyte and sporophyte distributions are related; however, the point of
gametophyte establishment is not necessarily the point of mature sporophyte distribution.
Epiphytic species are often associated with creeping rhizomatous growth. Such growth
may allow a perfectly healthy gametophyte to produce sporophytes in less than optimal
conditions. Such sporophytes may have the ability to then grow into more favorable
habitats where they can reproduce. The resultant mature sporophyte distributions
produced by such a strategy may obscure much of the species biology. Epiphytic species
may also form gametophyte banks that are long lived and stress tolerant and like many
seed banks, have the ability to wait for appropriate conditions to appear before recruiting
in to the sporophyte generation. This may be especially true for species whose
gametophytes also have a means of vegetative reproduction. This ability is common in
epiphytes (Atkinson and Stokey, 1964; Farrar, 1990; Farrar, 1998) and has been reported
in some terrestrial species (Watkins and Farrar, 2005). Such abilities question the
common assumption that gametophyte 'safe sites' limit the establishment of sporophytes
in epiphytic species. This hypothesis may hold more merit with terrestrial species but is
unlikely as important for epiphytic species.
Such apparent and fundamental differences in life history and the way that
epiphytic and terrestrial life forms respond to disturbance and light provides evidence for
adaptively meaningful variation in life histories that has evolved in the two groups.
Epiphytic species have evolved in a high light, highly competitive, yet relatively stable
matrix. Such environments reduce the light limitations encountered by terrestrial species,
yet they incorporate closer contact with bryophytes. Dassler and Farrar (1997) have
argued that differences in gametophyte morphology and asexual reproduction between
epiphytic and terrestrial species have largely evolved due to pressures form bryophyte
competition. Such changes may have only been possible in canopy habitats where
disturbance is less intense. Radiation into canopy habitats required a suite of adaptive
characters in both the gametophyte and sporophyte generation. One major advantage of
the canopy habitat is reduction in litter that can cover developing gametophyte. While
canopy habitats accumulate enormous amounts of organic matter (Cardelus and Chazdon,
2005), wind often removes a significant proportion of leaves that land on internal and
especially outer branches (Nadkarni and Matelson, 1991). The presence of leaf litter may
be the most important limiting factor to terrestrial species establishment as there are
relatively few morphological or physiological pathways that would allow a species to
survive under leaf litter. The mechanisms behind sporophyte distributions remain
complicated as they clearly also rely on gametophyte ecology. This is further complicated
by spore dispersal which may obscure the relationship between patterns of distribution
and habitat heterogeneity. Regardless of dispersal limitations or lack thereof, terrestrial
sporophytes are largely elements of disturbance past. Additional work on gametophyte
ecology will need to take into account edaphic and microclimatic factors to better
understand variables shaping the distribution of this magnificent group of plants.
Table 2-1. Relationship of gametophyte density and richness with three levels of
experimental disturbance and two light levels
Source DF F P>F
Light 1 52.643 0.000
Disturbance 3 18.567 0.000
Light*disturbance 3 6.152 0.001
Number of species
Source DF F P>F
Light 1 46.522 0.000
Disturbance 3 6.662 0.000
Light*disturbance 3 0.364 0.779
Table 2-2. Demographic and survival analyses for the gametophytes of 5 fern species using the Wilcoxon test to compare survival
distribution functions for different species
Number of Gametoohvtes
Species Dead Censored Total Months Std X2 P > X2
Campyloneurum brevifolium (Lodd. ex
Link) Link 19 87 106 22.9 0.57 386.2 <0.0001
Danaea wendlandii Rchb. f. 168 95 263 12.3 0.56
Lomariopsis vestita E. Fmyn. 22 45 67 20.8 0.82
Pityrogramma ebenea (L.) Proctor 165 110 275 4.1 0.28
Vittaria lineata (L.) Sm. 40 58 98 16.6 1.08
0 1 2 3
Level of Disturbance
25- B r2=0.65, F=53.58, p<0.001 a
Level of Disturbance
Figure 2-1. Number of gametophytes counted and their relation to disturbance from
natural transects in both A) Canopy habitats and B) Terrestrial habitats. Here
0 represents no disturbance (< 5cm2 bare soil), 1 = low disturbance (i.e. >
5cm2 bare soil), 2 = medium disturbance (>5cm2 bare soil and soil disturbed),
3= high disturbance (100% bare soil and soil turned over).
A r2=0.11, F=2.40, p=0.08
b b b
IN BF ES PD UN AT
Type of Disturbance
B Terrestrial Habitats
NL OT RT ER BF
Type of Disturbance
Figure 2-2. The percentage of fern gametophytes as influenced by type of disturbance
identified. A) In canopy habitats (IN: insect damage, BF: branch fall, ES: epi-
slide, PD: physical damage, UN: Unknown, AT: animal trail) and B) In
terrestrial habitats (NL: no leaf litter, OT: old tip-up mound, RT: rotting
tree/wood, ER: erosion, BF: branch fall, AT: animal trail, TU: recent tip-up
A Canopy Habitats
-- -- -
A Canopy Habitats r2=0.000, p=0.085
n *O ** *
0.5 o0 n e
0.0 m m m
I I .
1.4 1.5 1.6 1.7 1.8
Log (Canopy Openess)
(D 1.0 -
E 0.5 -
B Terrestrial Habitats
*** e *
e m r2=0.327, p=0.000
.0 0.5 1.0 1.5 2.0 2
Log (Canopy Openess)
Figure 2-3. The relationship between canopy openness and gametophyte density for A.
canopy and B. terrestrial species
c'4 800 I I Low Light
0 i 7-- I T
Control Low Med Hi
Level of Disturbance
Figure 2-4. Gametophyte densities as influenced by light and disturbance in experimental
plots. Control plots were those that had <5cm2 of bare soil exposed. Low
disturbance was created by only removing litter, medium plots were
established by removing litter and raking soil to a depth of 5cm, high plots
were created by removing litter and turning soil over with a spade to a depth
Campyloneurum Lomariopsis Vittaria Danaea Pityrogramma
brevifolium vestita lineata wendlandii ebenea
Campyloneurum Lomariopsis Vittaria Danaea
brevifolium vestita lineata wendlandii
Figure 2-5. Mean longevity (months) A) and percent gametophytes still alive and un-
recruited B) for the 25-month period of the study
\ .... ......
4 r ~- .'-.go. ---
0 .. -
__, Lomariopsis vestita
S Vittaria lineata
15 24 25 26
15 24 25 26
S Pityrogramma ebenea
-- Campyloneurum brevifolium
0 5 10 15 24 25 26
Figure 2-6. Kaplan-Meier survivorship curves A) and proportion recruiting B) of 5 species
of fern gametophytes over the 25-month study period. No data were collected
during months 4-7 and 15-24
---a Pityrogramma ebenea
COMPARATIVE DESICCATION TOLERANCE OF TROPICAL FERN
GAMETOPHYTES: ECOLOGICAL AND EVOLUTIONARY CONSEQUENCES
Overwhelming evidence indicates that land plants evolved from simple aquatic
algal ancestors (Bold, 1957; Niklas, 1997). The radiation of once-aquatic plants onto dry
land required the evolution of adaptive character suites that permitted life in what was
and remains deadly dry air. To survive this environment, plants have evolved two
mechanisms of surviving desiccating conditions. One mechanism is the avoidance of
desiccation as demonstrated in most modern terrestrial plants and has been accomplished
by the development of characters such as highly organized cuticles with effective
stomatal control. Cacti of the dry deserts perhaps represent the pinnacle of avoidance
with their succulent water storing stems, reduced leaves, and thick, well-developed
cuticle. These characters and many other associated with desiccation avoidance are all
thought to be highly derived and early land plants most certainly did not have such
structures. Early plants relied on the unique mechanism of survival from desiccation,
referred to as desiccation tolerance (DT). Desiccation tolerance has been differently
defined; the most commonly accepted definition is survival of drying to equilibrium with
surrounding air (Bewley, 1979). Such drying is more than sufficient to kill most any plant
that relies on desiccation avoidance: at least 99% of the world's vascular flora (Alpert
and Oliver, 2002).
Many of the characters that facilitated DT in ancestral land plants are still found
in modern algae and bryophytes (Alpert, 2000; Oliver, Tuba, and Mishler, 2000; Alpert,
2005). Much as the cacti are to avoidance, the bryophytes are to tolerance. Fantastic
stories exist in the literature demonstrating that some bryophytes can recover from 23
years of desiccation in herbaria (Alpert 2000 and references herein). Such remarkable
abilities clearly have ecological consequences and many studies have shown that more
desiccation tolerant bryophytes are often associated with more xeric habitats.
Consequences exist even within bryophyte species with one example being the
overrepresentation of female gametophytes in dioecious bryophytes of xeric habitats.
Such sex-based disparity is often related to greater DT in females relative to males (Stark
2005). Desiccation tolerance holds tremendous potential to influence the ecology of
species (Deltoro et al., 1998; Csintalan, Proctor, and Tuba, 1999; Robinson et al., 2000;
Apart from the linkage of DT to phylogeny, there have also been demonstrated
differences in desiccation tolerance of different generations, such as larval-adult in some
invertebrates or gametophyte-sporophyte in some bryophytes. The ability and degree of
desiccation tolerance in the different stages can be radically different. In some cases a
high degree of tolerance may exist in one stage but be absent in the other. For example,
the gametophytes of some bryophytes exhibit a much greater degree of tolerance than
sporophtyes (Proctor, 2000; Proctor, 2001). Such variation also occurs invertebrates
where the larvae of the fly Polypedilum vanderplanki exhibits greater desiccation
tolerance relative to the adult stage (Watanabe et al., 2002). Within the vascular plants,
desiccation tolerance of spores and seeds is well known but much less is known about
tolerance in lineages with two separate free-living stages.
The only lineage of vascular plants to exhibit two separate free-living generations
is the pteridophytes. Of particular interest to desiccation tolerance are the morphological
and physiological differences between these two stages in the ferns. The gametophyte is
the point of gamete formation and fertilization and is small, lacks vascular tissue, and has
a poorly developed to non-existent cuticle. The sporophyte produces spores and is thus
the primary stage for dispersal; it has a well developed vascular system and a waxy
cuticle complete with stomata. These differences alone result in unique life-cycle-
mediated ecological strategies, especially as they relate to water relations. True
desiccation tolerance in the sporophyte stage is known from and likely only exists in
relatively few species (Gaff, 1987; Porembski and Barthlott, 2000). In a recent review on
the subject, Proctor and Pence (2002) recorded that <1% (64 species) of the ferns studied
exhibited DT and of those, 40 were Cheilanthoid taxa that are commonly associated with
desert-like habitats. Much less is known of species from tropical habitats, but DT has
been recorded in genera as phylogenetically disparate as Asplenium and Polypodium
(Kappen, 1964; Gaff, 1987; Proctor and Pence, 2002). The species in which it does occur
are often extreme xerophytes living in deserts or other highly exposed and dry
environments. As with bryophytes, the apparent degree of desiccation tolerance in the
sporophyte stage has been linked with species ecological distribution (Harten and
Eickmeier, 1987; Hietz and Briones, 1998b). Studies are still too sparse to determine the
extent of this character in structuring populations.
Studies on desiccation tolerance of the gametophyte generation of the ferns are
fewer in number. Some of the earliest comments were made by Goebel (1900) regarding
the ability of the buried tubercles of Annogramme chaerophylla to resume growth
following dry spells. A similar observation was made by Cambell (1904) on the unburied
gametophytes of Gymnogramme triangularis that appeared to have survived a dry
California summer. The first evidence was generated by Pickett (1913; 1914; 1931) who,
through a series of desiccation experiments was the first to clearly show that the
gametophytes ofAsplenium rhizophyllum and A. platyneuron could recover growth
following extreme desiccation. He also discovered that there was a greater degree of
tolerance in A. Rhizophyllum, a species of more exposed and drier habitats, relative to A.
platyneuron, which is often confined to more mesic sites. This was the first link of
desiccation tolerance in the gametophyte generation with sporophyte distributions and
species ecology. Pickett's work has largely been the last of its kind, and apart from
anecdotal reports (Gilbert, 1970) and observations on the ability of the gametophytes of
Pyrossiapilosellodes to recover from drought (Ong and Ng, 1998) nothing is known of
the ability of fern gametophytes to tolerate desiccation and of their rates of recovery.
The goal of this paper is to survey a broad range of tropical fern gametophytes to
determine the extent of desiccation tolerance in this phase of the fern life cycle. We first
examine the ability of several species to recover from a single desiccation event. We then
subject a select number of several species of varying life histories to a more extensive
repeated dry down cycles and drying intensities and relate the results to the ecological
distributions of the species.
Materials and Methods
Spore Material and Growth Conditions
Spore material from 12 species of varying ecology was collected from La Selva
Biological Station in the Atlantic lowlands of northeastern Costa Rica at 37-100 masl.
Fertile fronds were gathered in the field and put into glassine envelopes with tape-sealed
seams. Envelopes with plant material inside were stored in an air-conditioned lab and
allowed to dry under these conditions. Spores were brought back to the University of
Florida where they were cultured. The growth of a broad sampling of species with
different ecologies required extensive experimentation with culture techniques, it was
discovered that species grew best on a combination of organic soil collected from canopy
trees at La Selva mixed with a small amount of vermiculite. Spores were sown on this
medium into 60mm x 15mm Fisherbrand Petri plates. These plates were stored in sealed
clear plastic containers (Pioneer Plastics Model 395-c, Dixon, KY). Cultures were
exposed to 20tmol m-2 sec-1 for 10hrs day- from GE fluorescent plant and aquarium
40watt growbulbs and watered with deionized water every 10-12 days.
For the initial survey experiment, 5-10 mature gametophytes (one gametophyte
per Petri plate) of all 12 species (see Table 3-1) were allowed to desiccate at a vapor
pressure deficit (VPD) =1.3kPa (50% relative humidity) in a VPD controlled chamber
that was constructed using one of the plastic growth boxes connected via Bevline tubing
to a Licor dew point generator (Model 610, Lincoln, NE) set to a flow rate of 0.5 L min'.
Samples were allowed to dry for 45min in a constant vapor pressure deficit and were
removed from the box every 5min and placed in a Sartorious microbalance (Gottingen,
Germany) where weight and a measurement of Fv/Fm was taken (see methods below). The
samples were then placed back into the box. The volume of the box was relatively small,
and a Hobo Pro RH/Temp Data Logger (Bourne, MA) was used to verify that the
container typically regained the 1.3kPa VPD within one min of the top being replaced. To
evaluate recovery, upon completion of the desiccation treatment, samples were
consecutively rehydrated by adding 5-10 drops of deionized water to the thallus. After 1
hr of rehydration, measurements of Fv/Fm were again made at 5min, 24hrs, and 48hrs post
rehydration. Samples were then dried for 72hrs in a drying oven at 700C and weighed to
determine gametophyte dry weight. Relative water content was plotted against time and
A second desiccation experiment was designed to test the effect of drying
intensity on recovery of Fv/Fm. The gametophytes ofDiplazium subsilvaticum and
Phlebodium pseudoaureum, and Polypodium triserale were chosen to represent the
extremes of tolerance from the initial desiccation experiment and they were dried at three
different intensities VPD=0.5kPa (20% RH), VPD=1.3kPa (50% RH) and VPD=2.lkPa
(80%RH) following the methods above. Gametophytes were kept at these VPD for 72hrs
after which time they were rehydrated with deionized water and measurements of Fv/Fm
were taken at 24, 48, and 72hr post rehydration. These values were related to the dark
adapted value of Fv/Fm to determine the mean percent recovery.
A third experiment was run to examine the influence of consecutive desiccation
cycles on photochemical efficiency. Plant material from six species was chosen from the
survey experiment to represent the different recovery abilities. Thirty gametophytes were
selected and ten were dehydrated for one, two, or three cycles at VPD=1.3kPa and kept at
this level for 72 hrs using the methods described above. Material was then rehydrated
with deionized water and measurements of Fv/Fm were again made at 24hrs, 48hrs, and
72hrs post rehydration. These values were related to the dark adapted value of Fv/Fm to
determine the mean percent recovery.
Variation in photochemical efficiency (Fv/Fm) was measured as the desiccation
dependent change in ratio of variable and maximal fluorescence Fv/Fm where Fv is the
difference between the maximum (Fm) and the minimum (Fo) fluorescence emissions
(Mulkey and Pearcy, 1992; Horton, Ruban, and Walters, 1996) measured using an Opti-
Sciences pulse modulated fluorometer (Model OS-500, Hudson NH). Minimal
fluorescence was measured under a weak pulse of modulating light over 0.8 s, and
maximal fluorescence was induced by a saturating pulse of light (8000 imol m-2 -1)
applied over 0.8 s. The parameter Fv/Fm was first measured after 20mins dark adaptation,
and this measurement was taken as the index of recovery. Dark-adapted FV/FM provides
an estimate of the maximal quantum efficiency of Photosystem II, which in unstressed
material is generally around 0.76-0.83 (Demmig-Adams and Adams, 1992).
For the initial desiccation survey, a series of regressions was run on arcsin square root
transformed relative water content (RWC) data against time for each individual within a
species to determine the rate of drying of gametophytes exposed to a VPD of 1.3kPa
(50% RH) over the 45min time interval. The slopes of these regression lines were
calculated to generate a species mean drying rate. These rates were then analyzed by a
one way ANOVA followed by a post hoc Tukey's test to determine differences among
species. Linear regression analysis was used to asses the influence both final relative and
absolute water content at 45min and the slope of the individual drying curves on species
recovery ability at 48h post rehydration. Percent species recovery data were also arcsin
square root transformed for these analyses. The mean species drying rates, RWC, and
absolute water content (AWC) at 45min were then plotted against each species' mean
percent recovery at 48h. Depression in photochemical efficiency was also graphed as a
function of thallus RWC and AWC.
To assess the influence of consecutive desiccation cycles (experiment 3) and VPD
(experiment 2) on photochemical efficiency, a repeated measures ANOVA was
performed with number of desiccation cycles or VPD and recovery time as the fixed main
effects. Data were first examined for sphericity following the Mauchly criterion. Pairwise
comparisons were made across recovery times with Bonferroni-adjusted multiple t-tests.
Klockars and Sax (Klockars and Sax, 1986 p. 38-39) recommend using the more
stringent Bonferroni-adjusted multiple t-test when the number of planned comparisons is
greater than the number of degrees of freedom for between-groups. In cases where data
did not meet the sphericity criterion, p-values were adjusted using both Greenhouse-
Geisser and Huynh-Feldt methods based on the respective epsilons (Scheiner and
For the initial desiccation survey, a series of regressions were run on each species
to determine the change in relative water content (RWC) of gametophytes exposed to
1.32kPa (50% RH) over the 45min time interval. All species exhibited rapid rates of
thallus relative water content loss (Fig. 3-1A) and absolute water content loss (data not
shown). In all species regressions on the arcsin square root transformed data were linear
(Table 3-1). In all cases, linear regressions were used to calculate the slopes of species'
RWC and AWC drying curves to determine desiccation rates. The absolute size (mass
based) of individual gametophytes had little influence on the absolute rate of water loss
(Fig 3-3a, r2=0.05, p=0.06). These rates varied significantly among species with the
fastest dry down in the terrestrial Thelypteris balbisii and slowest rates in the terrestrial
Cyclopeltis semicordata and the epiphyte Polypodium triserale (Fig. 3-3A). Depression
in photochemical efficiency (Fv/Fm) as gametophytes desiccated was non-linear with
respect to RWC and varied among species (Fig. 3-1B).
Species exhibited differential abilities to recover following desiccation (Fig 3-
3B). This recovery ability was more closely related to the decay rate of RWC (r2=0.288,
p<0.0001) when compared to the final RWC reached (r2=0.193, p=0.0008), the decay
rate of AWC (r2=0.0001, p=0.81), or final AWC reached (r2=0.097, p=0.008) after
The VPD of the different desiccation treatments significantly influenced the
recovery abilities of both Diplazium subsilvaticum and Phlebodium pseudoaureum, but
had little influence on Polypodium triseriale. For the understory terrestrial D.
subsilvaticum, the ability to recover following the 2.12 and 1.3kPa VPD treatments was
essentially non-existent. Additionally, the Fv/Fm values reached at these VPDs are
suggestive of significant photoinhibition and photodamage. The 0.53kPa treatment also
depressed Fv/Fm but to a lesser degree and gametophytes exposed to this treatment
exhibited clear recovery following rehydration. Phlebodium pseudoaureum exhibited
relatively less depression in Fv/Fm and exhibited greater recovery than Diplazium
subsilvaticum. In all three VPD treatments, gametophytes exhibited recovery albeit with
lower rates from the 2.12 and 1.32kPa treatments. Polypodium triserale exhibited
remarkable Fv/Fm fidelity at all three rates with no significant degree of Fv/Fm
depression at any desiccation intensity (Fig. 3-5, Table 3-2).
Six species representing different life histories and desiccation tolerance from the
initial survey were chosen and exposed to multiple desiccation cycles (1, 2, or 3). In all
cases excluding Polypodium triserale, percent recovery was greater following one versus
two or three desiccation cycles (Fig. 3-6 f, Table 3-3). Recovery ability was closely
linked to species ecology with slow to no recovery following >1 desiccation cycle in the
understory species that occur in more mesic habitats: Diplazium subsilvaticum, Adiantum
latifolium, and Cyclopeltis semicordata. Within this group, Cyclopeltis semicordata
exhibited less inhibition and the greatest recovery following a single desiccation cycle.
With multiple cycles, all species were significantly inhibited and there was little evidence
of recovery following 2 and 3 cycles. There was a much greater degree of recovery
following 2 and 3 cycles in the species from more xeric habitats: the terrestrial
Pityrogramma ebenea, and the epiphytes: Phlebodium pseudoaureum, Polypodium
triserale. Unlike the other species in this xeric category, Polypodium triserale exhibited
similar degrees of recovery from 1 and 2 cycles but experienced a much slower recovery
following the third desiccation cycle.
In all species, fern gametophytes exhibited rapid rates of water loss (Fig la.).
With poor control of transpiration and inefficient absorptive organs, gametophytes likely
rely on water derived directly from the atmosphere and/or that which flows over them. As
such, gametophytes face considerable variations in water content throughout the day and
must be able to withstand long periods of desiccation. This is especially true of epiphytic
gametophytes that can live for years (Watkins et al. Chapter 2). The only recourse that
gametophytes have is to tolerate desiccation or perish.
The initial survey produced a surprising degree of desiccation tolerance across
many species in a gametophyte that is known to require water for fertilization and is
thought to require humid conditions for survival. After exposure to a VPD of 1.3kPa
(rH=50%) for 24 hours, all species exhibited greater than 50% recovery of the pre-
treatment Fv/Fm values and the majority had recovered more than 70% of this value (Fig.
3-3). Gametophytes were dried to near constant state and thus match the definitions of
desiccation tolerant by Bewley (1979) and Alpert and Oliver (2002). The species in this
study are all tropical in origin and while they experience various degrees of humidity in
nature, some of the more exposed canopy trees for this same forest average VPD-3.0kPa
at mid-day during the dry season (Cardelus and Chazdon, 2005). The gametophytes in
this study experience VPD levels of 1.3kPa, but the value remains on the extreme end of
what they typically experience. There was no significant difference in the absolute water
loss rates based on gametophytes mass (Fig. 3-2). In natural settings, individuals will be
exposed to daily variation in RWC and their ability to recover from low relative water
content is crucial. The only apparent mechanism that gametophytes have to control water
loss is an increase in size and perhaps alteration of thallus morphology (see below).
Variation in VPD
To further examine the influence of desiccation intensity on recovery of
pretreatment photochemical efficiency, three species were exposed to three different
desiccation intensities: roughly the everyday vapor pressure deficient in nature (0.53kPa),
that which is likely to reflect a typical drought event (1.32kPa) and an extreme value
representing a VPD that species in this site rarely if ever experience (2.12kPa). The
results from this experiment demonstrated remarkable tolerance to desiccation intensity
that is tightly linked to species ecology. Diplazium subsilvaticum is a low-light creek-side
species and has little tolerance of desiccation intensities below 0.53kPa (80% RH) (Fig 3-
5A). The biggest decrease in Fv/Fm occurred in this species between 0.53 and 1.32kPa.
The two epiphytes occur in similar habitats; and, Phebodium pseudoaureum is typical of
open and exposed habitats such as roadsides and open clearings, Polypodium triseriale is
often found in more highly exposed areas such as fence posts and tree trunks. The level
of tolerance to desiccation intensity was linked to the habitats with the most highly
exposed Polypodium triseriale exhibiting essentially no sensitivity to desiccation
intensity (Fig. 3-5C).
These patterns enforce the notion that desiccation rates can influence the survival
of certain species. In some bryophytes (Gaff, 1997; Alpert and Oliver, 2002; Alpert,
2005) and at least in the sporophtyes of the Hymenophyllaceae (Proctor, 2003),
intermediate desiccation intensities that maintain intermediate RWC's are more likely to
result in mortality compared to rapid dry downs. Alpert and Oliver (2002) have argued
that one reason for this pattern is need for rapid yet organized metabolic shutdown that
may not occur at slower desiccation rates. Additional studies that incorporate different
drying intensities and longer time spent in these conditions are clearly needed.
The link between DT and species distributions corresponds closely with those
reported for bryophyte species from xeric habitats exhibiting greater desiccation tolerance
than those from mesic habitats (Oliver, Mishler, and Quisenberry, 1993; Deltoro et al.,
1998; Proctor, 2001; Cleavitt, 2002; Alpert, 2005). It also suggests that there may be
some connection between gametophyte and sporophyte physiologies.
Not only do species experience different intensities of desiccation in nature, they
also experience multiple desiccation cycles throughout the day and/or growing season.
The species in this study clearly exhibited different abilities to cope with consecutive
desiccation cycles at a VPD of 1.3kPa with species of more mesic habitat exhibiting little
ability to cope with more than one cycle of desiccation (Fig. 3-6). The more mesic
creekside Diplazium subsilvaticum was the most desiccation-sensitive after one cycle and
had the worst recovery whereas the more xeric Adiantum latifolium and Cyclopeltis
semicordata had higher recoveries following one cycle (Fig. 3-6 c&b). Polypodium
triseriale, the species of more xeric habitats also exhibited depression in Fv/Fm, but
recovery was largely independent of two desiccation cycles. One aspect that was
common for all terrestrial species is the relative tolerance to a single desiccation cycle
and the extreme depression caused by repeated cycles. Repeated desiccation cycles
likely induce radical biological damage and recovery depends more on actual DNA repair
and new protein synthesis than simple recovery of PSII function related to the release of
excess excitation energy. Light and desiccation combined are often a deadly combination;
and that gametophytes were returned to the original culture conditions upon rehydration
could have resulted in increased photodamage. This was however similar to what species
experience in nature and is likely reflective of species biology.
The ability of species to recover from desiccation was more closely related to the
rate of drying rather than the final RWC (Fig 3-4). While gametophytes initially seem to
have relatively few options to control the rate of thallus desiccation there was variation in
rates among species exposed to identical drying conditions (Fig 3-3a). For example, the
terrestrial Thelypteris nicaraguensis a RWC decay rate that was more than twice as fast
as the canopy epiphyte Polypodium triseriale. The variation in desiccation rate was
linked to gametophyte morphology (Fig. 3-4). Species with complex three-dimensional
morphologies or those that tended to produce prothallia hairs exhibited significantly
slower dry down rates than gametophytes that are one dimensional and glabrous. This
observation suggests a novel mechanism that gametophytes may employ to control water
loss. There is limited internal capacitance in fern gametophytes, and in a manner similar
to many bryophytes, fern gametophytes with even a minor degree of morphological
complexity can hold external water. Such exohydric abilities may function in a natural
setting to help slow the rate of water loss. The decrease in rate was more likely a
combination of water vapor becoming trapped in the folds and overlapping wings of
more complex thalli and the increase of external boundary layer produced by
gametophyte proliferations and hairs.
Variation in gametophyte morphology especially in complexity has long been
commented upon. Dassler and Farrar (1997, 2001) have speculated that complex
morphologies found in many long-lived gametophytes of epiphytic species have arisen
from competition with bryophytes and as a mechanism to ensure outcrossing. One
obvious trend across the taxa is that gametophytes of species from drought-prone habitats
such as those in epiphytic habitats, deserts, rock outcroppings, etc., tend to produce thalli
that often exhibit complex branching, overlapping wings, proliferation, and hairs. Apart
from this link between gametophyte morphology and ecology there is also a link between
gametophyte morphology and phylogeny. Recent phylogenetic analysis of the ferns has
revealed a split between terrestrial and epiphytic clades in the Eu-Polypodiales (Pryer,
Smith, and Skog, 1995). The athyrioids, thelypteroids, onocleoids, woodsioids and
blechnoids are almost entirely terrestrial; perhaps less than 1% of the known species of
this clade are true epiphytes. On the other hand, the dryopteroids, lomariopsoids,
elaphoglossoids, oleandroids, davallioids, and polypodioids have many epiphytic species;
perhaps as many as 60-70% of the species in this group as a whole are epiphytic (Fig 3-6)
(R. Moran pers. comm.). Most epiphytic species exhibit complex morphologies whereas
terrestrial species often less complex ones. Increases in thallus size and complex three
dimensional morphologies may provide the only water conservation mechanisms
available to fern gametophytes. Complex water conserving morphology may have been
critical in the radiation of ferns into canopy and more exposed habitats.
The data presented in this paper show remarkable desiccation tolerance in fern
gametophytes. While all species exhibited recovery following an extreme desiccation
event, the extent of recovery differed among species and was closely linked to the
ecology of the species. The role of the fern gametophyte in controlling recruitment
remains unclear, but these data suggest that gametophytes are relatively robust in dealing
with desiccation especially in limited cycles. While desiccation intensity clearly
influenced recovery, repeated cycles of desiccation were more likely to limit recovery
and in the case of the most mesic species likely resulted in significant photodamage. The
degree of recovery following desiccation and its relation to species ecology suggest that
fern gametophytes exhibit adaptively meaningful variation in this character. It is likely
that selective pressures acting on the gametophyte are largely responsible for the
distribution of ferns and have played a major role in the evolution of ecological diversity
within the group.
Table 3-1. Species and life form from the initial desiccation survey, a series of regressions were run on each species to determine the
change in relative water content (slope) of gametophytes exposed to 1.32kPa (50% rH) over the 45min time interval. All
species exhibited rapid and uncontrolled rates of thallus water loss. In all species regressions on the arcsin square root
transformed data were linear (STE are standard errors, RWC is relative water content expressed as ((g fresh weight- g dry
weight)/g saturated weight g dry weight))* 100, AWC is absolute water content expressed as mg water / mg dry mass, %
rec corresponds to percent recovery of initial pre-treatment dark adapted Fv/Fm values
Adiantum latifolium Lam.
Cyclopeltis semicordata (Sw.) J. Sm.
Dennstaedtia bipinnata (Cav.)
Diplazium subsilvaticum H. Christ
Nephrolepis biserrata (Sw.) Schott
Phlebodium pseudoaureum (Cav.)
Pityrogramma ebenea (L.) Proctor
Polypodium triseriale Sw.
Pteris altissima Poir.
Thelypteris balbisii (Spreng.) Ching
Thelypteris curta (H. Christ) C. F.
Thelypteris nicaraguensis (E. Fourn.)
C. V. Morton
Life Form r2 p Slope STE RWC STE AWC STE %rec STE
Terrestrial 0.711 <0.0001
Terrestrial 0.762 <0.0001
-1.463 0.116 18.875 5.829 3.380 0.794 0.600 0.127
-2.199 0.061 19.254 3.512 3.690 0.370 0.555 0.087
Table 3-2. Fv/Fm recovery results from the repeated measures ANOVA for gametophytes exposed to three different desiccation
intensities: 20%RH (-0.53kPa), 50%RH (-1.32kPa) and 80%RH (-2.12kPa). The gametophytes of Diplazium
subsilvaticum are often found in the understory, whereas those of Phlebodium pseudoaureum, and Polypodium triserale
were collected in the mid and exposed canopy respectively. Gametophytes were kept at the VPD levels for 48hrs after
which time they were rehydrated with deionized water and measurements of Fv/Fm were taken at 24, 48, and 72hr post
rehydration. These values were related to the dark adapted value of Fv/Fm to determine the mean percent recovery
Diplazium striatastrum Lellinger df F p Mauchly x p
VDP 2 38.41 <0.0001 0.739 3.24 0.662
ID(VPD) 12 1.86 0.0741
Recovery Time 2 7.32 0.0033
Recovery Time*VPD 4 6.39 0.0012
Phlebodium pseudoaureum (Cav.) Lellinger df F p Mauchly x p
VDP 2 42.19 <0.0001 0.716 3.59 0.61
ID(VPD) 12 2.2 0.0482
Recovery Time 2 24.31 <0.0001
Recovery Time*VPD 4 7.81 0.0003
Polypodium triseriale Sw. df F p Mauchly x p
VDP 2 1.46 0.2716 0.779 2.67 0.75
ID(VPD) 12 4.53 0.3531
Recovery Time 2 0.6 0.6352
Recovery Time*VPD 4 5.09 0.0041
Table 3-3. Fv/Fm recovery results from the repeated measures ANOVA for gametophytes exposed to 1, 2, or 3 desiccation cycles at
50% RH (VPD- 1.32kPa). Gametophytes were kept at this level for 48 hrs. Material was then rehydrated with deionized
water and measurements of Fv/Fm were again made at 24hrs, 48hrs, and 72hrs post rehydration. These values were related
to the dark adapted value of Fv/Fm to determine the mean percent recovery. Adjusted p-vaules are G-G Greenhouse-Geisser
and H-F Huynh-Feldt adjusted probabilities
Diplazium striatastrum Adjusted
Lellinger df F p Mauchly X p G-G H-F
Rate 2 3.07 0.0649 0.6301 5.081 0.0788 0.0043 0.0016
ID(Trt) 12 3.58 0.3931
Recovery Time 2 98.7 <0.0001 GG E = 0.73
Recovery Time*Trt 4 4.06 0.0118 HF E = 0.944
Adiantum latifolium Lam. df F p Mauchly X p G-G H-F
Rate 2 1.53 0.2378 0.691 4.06 0.1312 0.0073 0.0029
ID(Trt) 12 2.09 0.4982
Recovery Time 2 324.53 <0.0001 GG E = 0.764
Recovery Time*Trt 4 5.44 0.0029 HF E = 0.999
Cyclopeltis semicordata (Sw.)
J. Sm. df F p Mauchly X p G-G H-F
Table 3-3. Continued
Pityrogramma ebenea (L.)
Proctor df F p Mauchly 2 p G-G H-F
Rate 2 57.65 <0.0001 0.857 1.688 0.4298 <0.0001 <0.0001
ID(Trt) 12 0.56 0.792
Recovery Time 2 80.09 <0.0001 GG E = 0.875
Recovery Time*Trt 4 11.67 <0.0001 HF E = 1
(Cav.) Lellinger df F p Mauchly 2 p G-G H-F
Rate 2 98.81 <0.0001 0.462 8.48 0.0143 <0.0001 <0.0001
ID(Trt) 12 0.38 0.8699
Recovery Time 2 169.52 <0.0001 GG E = 0.65
Recovery Time*Trt 4 20.49 <0.0001 HF E = 0.818
Polypodium triseriale Sw. df F p Mauchly 2 p G-G H-F
Rate 2 24.82 <0.0001 0.492 7.78 0.02 0.0046 0.0019
ID(Trt) 12 1.65 0.548
Recovery Time 2 87.4 <0.0001 GG E = 0.664
Recovery Time*Trt 4 6.78 0.0008 HF E = 0.839
0 10 20 30 40 5C
100 80 60 40 20 0
Relative Water Content
-o- Thelypteris nicaraguensis
-o- Thelypteris curta
-- Pityrogramma ebenea
-- Adiantum latifolium
-A- Dennstaedtia bipinnata
-- Pteris altissima
--- Diplazium subsilvaticum
-*- Thelypteris balbisii
-*- Phlebodium pseudoaureum
-*- Nephrolepis biserrata
-- Cyclopeltic semicordata
-*- Polypodium triseriale
Figure 3-1. (a) Gametophyte drying curves from 12 tropical fern species of different
habitats. Species were exposed to a VPD to 1.32KPa (-50%RH) for 45 min.
(b) Depression of photochemical efficiency in the same gametophytes over a
series of decreasing thallus water contents
-* -0.32 % *
ao e* e
0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0.0040
Dry Mass (mg)
Figure 3-2. The rate of absolute water loss relative to gametophyte size as indexed by dry
. x ^ ..
0 .0 i i I i I I Ii Ii I
Figure 3-3. (a) Rate of thallus drying as calculated from Figure 3-1 for 12 tropical fern
species of different habitats. (b) Proportional recovery of the pre-treatment
dark adapted value of Fv/Fm in these same species
A F=10.23, p<0.0001
t b b b b
-2.2 -2.0 -1.8 -1.6 -1.4 -1.2 -1.0
Change in RWC min-
Figure 3-4. Proportional recovery of the pre-treatment dark adapted value of Fv/Fm and
rate of thallus water loss expressed as (a) relative water content ((g fresh
weight- g dry weight)/g saturated weight g dry weight))* 100 and (b)
absolute water content (g wet weight/g dry weight)
-0.4 -0.3 -0.2 -0.1
Change in AWC min1
0.6 a a a
LL 0.4 b
0.2 b b
Diplazlum subsilvaticum (understory)
Dark 24h 48h 72h
b b b
Phlebodium pseudoaureum (Mid-Canopy)
Dark 24h 48h 72h
Dark 24h 48h 72h
Figure 3-5. Fv/Fm recovery graphs for gametophytes of (a) Diplazium subsilvaticum, (b)
Phlebodium pseudoaureum, (c) Polypodium triserale exposed to three
different desiccation intensities: VPD-0.53kPa (20%RH), VPD-1.32kPa
(50%RH), and VPD -2.12kPa (80%RH). The gametophytes of Diplazium
subsilvaticum are often found in the understory, whereas those of Phlebodium
pseudoaureum, and Polypodium triserale were collected in the mid and
exposed canopy respectively. Gametophytes were kept at the VPD levels for
48hrs after which time they were rehydrated with deionized water and
measurements of Fv/Fm were taken at 24, 48, and 72hr post rehydration.
Pairwise comparisons were made across recovery times with Bonferroni-
adjusted multiple t-tests
-- 2 12 Kpa(20% Rh)
-0- 1 32 Kpa (50%r Rh)
-V- 0 53 Kpa (80% Rh)
Polypodium trlserale (Exposed Canopy)
100oo A Diplazium subsilvaticum loo D Pityrogramma ebenea
80 Cycle 1 80 a
-0- Cycle 2
8 -0- Cycle 3
S40 a a 40
20 b b b 20-
0 Vb b 0
24 48 72 24 48 72
100 B Adiantum latifolium 100 E Phlebodium pseudoaureum
80 80 -
8 a b
I 60 a --- 60
8 40 40 b
0 b__ ____________ b 20
24 48 72 24 48 72
C Cyclopeltis semicordata 100 F Polypodium triseriale a
80 a a a a
8 60 -
40 b b 40
b 4 0
20 c 20
24 48 72 24 48 72
Recovery Time (hrs) Recovery Time (hrs)
Figure 3-6. Proportional Fv/Fm recovery results for gametophytes exposed to 1, 2, or 3
desiccation cycles at VPD- 1.32kPa (50%RH). Gametophytes were kept at
this level for 48 hrs. Material was then rehydrated with deionized water and
measurements of Fv/Fm were again made at 24hrs, 48hrs, and 72hrs post
rehydration. These values were related to the dark adapted value of Fv/Fm to
determine the mean percent recovery. Pairwise comparisons were made within
recovery times with Bonferroni-adjusted multiple t-tests. (a) Diplazium
subsilvaticum, (b) Adiantum latifolium, (c) Cyclopeltis semicordata, (d)
Pityrogramma ebenea, (e) Phlebodium pseudoaureum, (f) Polypodium
triseriale (see Table 3-1 for species habitat information)
i iHe AHemidictyoid
$ : Nephrolepis
Figure 3-7. Morphology in fern gametophytes is diverse and is closely related to species
ecology and phylogeny. Gametophytes of species from drought prone habitats
such as those in epiphytic habitats, deserts, rock outcroppings etc. tend to
produce thalli that often exhibit complex branching, overlapping wings,
proliferation and hairs; whereas, species from more buffered habitats have less
ornamented and simple morphologies. The athyrioids, thelypteroids,
onocleoids, woodsioids and blechnoids are almost entirely terrestrial, perhaps
less than 1% of the known species are true epiphytes. On the other hand, the
dryopteroids, lomariopsoids, elaphoglossoids, oleandroid, davallioid and
polypodioids have many epiphytic species, perhaps as many as 60-70% of the
species in this group as a whole are epiphytic (R. Moran pers. comm.). Most
epiphytic species exhibit complex morphologies whereas terrestrial species
often less complex ones. Such differences in morphology are significant and
may have been critical in the radiation from terrestrial species into canopy
habitats. A) Adiantum latifolium, B) Thelypteris sp. 1, C) Thelypteris sp.2, D)
Vittaria, E) Campyloneurum
NITROGEN-15 NATURAL ABUNDANCE AND NITROGEN USE STRATEGIES OF
THE GAMETOPHYTES AND SPOROPHYTES OF TROPICAL EPIPHYTIC AND
A central goal of plant ecology is to develop a mechanistic understanding of
species distributions in both space and time. One important abiotic factor that is known to
influence both plant performance and distribution is nitrogen. In N-limited systems or in
systems where N availability is heterogeneous, plants can compete for N in many ways,
one of which is by partitioning this resource in both space and time or by uptake of
different chemical forms of N. Such partitioning may result in species coexistence
through sorting along resource gradients which ultimately structures species distributions.
The partitioning and uptake of different chemical forms of N has been receiving
increased attention largely due to the revelation that plants can circumvent the N cycle
and directly uptake organic N, in the form of amino acids from the soil solution (Chapin,
Moilanen, and Kielland, 1993a; Kielland, 1994, 1997; Lipson and Nasholm, 2001; Finzi
and Berthrong, 2005). While it has been known for some time that plants can acquire
organic N, early work in tundra and boreal ecosystems demonstrated that a large
component of an individual's N budget could be supported by direct uptake of organic
relative to inorganic N (Chapin, Moilanen, and Kielland, 1993b; Kielland, 1994, 1997).
Subsequent studies have shown that plants from a wide range of ecosystems can directly
access organic N as an important part of their N nutrition (Lipson and Nasholm, 2001).
Few studies, however, have examined the ability of plants from tropical wet forests to
take up organic nitrogen and to our knowledge no studies have examined this ability in
The ferns pose a unique set of ecological constraints due to both the dispersal of
tiny wind-blown spores and the occurance of an independent and free-living haploid
gametophyte. The mineral nutrition of fern sporophytes is not well studied; however,
evidence indicates that fern sporophytes behave as seed plant sporophytes when
confronted with increased inorganic N (Prange and Ormrod, 1982; Walker and Aplet,
1994; Pillai and Ong, 1999). Less is known of the mineral nutrition of gametophytes and
unlike sporophytes (which rely on well developed root systems), fern gametophytes are
thought to rely primarily on rhizoid uptake of nutrients with possible uptake of water and
nutrients across the thallus (Racusen, 2002). The ability of fern gametophytes to grow on
different N forms was reviewed by Miller (1968), and there is limited evidence to show
that the gametophytes of at least one species can grow well on specific mixtures of amino
acids in the absence of inorganic forms.
For plants other than ferns, changes in 815N values have been shown to occur
through ontogeny and with increases in plant size (Zotz, 1997; Schmidt, Stuntz, and Zotz,
2001; Hietz and Wanek, 2003; Reich et al., 2003; Zotz et al., 2004; Casper, Forseth, and
Wait, 2005). In the case of hemiepiphytic plants, changes occur as a direct result of the
connection of once aerial roots to terrestrial water and nutrient pools (Putz and Holbrook,
1989; Field, Lawton, and Dawson, 1996; Wanek et al., 2002). As ferns continue their
life-cycle from gametophyte to sporophyte, radical changes in their ecophysiology,
especially nutrient relations are likely to occur.
The goal of this paper is to determine if species differ in their ability to take up
different forms of nitrogen in both the gametophyte and sporophyte generations and to tie
this to the natural abundance of 615N to determine if species access different nitrogen
source pools. We then compare these data across epiphytic and terrestrial species and a
developmental series of a hemiepiphytic species to better understand the dynamics of
nitrogen nutrition of different life forms.
Material and Methods
This study was conducted at La Selva Biological Station of the Organization for
Tropical Studies in Heredia Province, Costa Rica (10026' N, 84o00' W). La Selva is a
1400 ha tropical wet-forest positioned in the Caribbean lowlands with an average
monthly temperature of 25.8 C and annual rainfall of 4000 mm per year (Sanford et al.,
1994). The site boasts a diversity of ferns with multiple species from epiphytic,
hemiepiphytic and epiphytic life forms (Grayum and Churchill, 1987).
The gametophytes and sporophytes of 10 species were field collected from
100X100m grids to control for differences in soil type in the terrestrial species and from
the trees in the case of epiphytic species (Table 4-1). The following species were
Adiantum latifolium Lam. is a terrestrial species that is common in disturbed areas
in both primary and secondary forests. The species typically grows along trail sides and
can be encountered under a wide rage of light and soil regimes.
Danaea nodosa (L.) Sm. and Danaea wendlandii Rchb. F. are both eusporangiate
ferns and as such have gametophytes that are several cell layers thick. Danaea
wendlandii is perhaps the most common fern at La Selva and often grows on upland well
drained sites and in disturbed understory habitats. Danaea nodosa has similar habitat, but
is more often associated with wetter sites becoming most abundant along creek sides.
Diplazium subsilvaticum H. Christ is a terrestrial arborescent species that is
common along creek banks and wetter areas in both primary and secondary forests.
Lomariopsisjapurensis (Mart.) J. Sm. and Lomariopsis vestita E. Foum are both
understory hemiepiphtyes. In both species, the gametophytes develop on the trunks of
small trees and remain epiphytic throughout their life. Young sporophytes produce roots
that grow down and contact the soil and rely on the host tree for support. Adult plants
never loose contact with the forest floor.
Olfersia cervina (L.) Junze has been classified as both a terrestrial and
hemiepiphyte species. It is restricted to grow on soils with high organic content and is
most commonly found growing on rotting logs, but can also grow on large tree trunks
where sufficient detritus has accumulated.
Elaphoglossum latifolium (Sw.) J. Sm. and Campyloneurum brevifolium (Lodd.
Ex Link) Link are both canopy epiphytes with the former more often found in highly
exposed portions of the canopy and often on bare bark. The latter species also occurs in
exposed sites, but is most common in the inner canopy rooted in canopy soil organic
Antrophyum lineatum (Sw.) Kaulf: An understory epiphyte that grows on the
trunks of living trees. Both gametophytes and sporophytes are common in primary and
secondary forests at La Selva.
Isotopic Natural Abundance and 61 N Labeled Uptake
Foliar and gametophytic samples for natural abundance of all nine species were
field collected within a 3-wk period during June 2005. Collections were brought back to
the lab where they were washed in deionized water to remove all soil and then dried at
60C for 48h. Samples were analyzed for nitrogen concentration and isotope ratio using a
Costech elemental analyzer coupled with a Finnigan Delta XL PlusTM continuous flow
mass spectrometer at the University of Florida, Gainesville. Based on repeat analyses of
NIST peach leaves standard (SRM 1547; 615N 1.91%o), average Is precision was 0.07%o
In order to compare uptake of both organic and inorganic N forms we used
excised root techniques (Treseder and Vitousek, 2001) in sporophtyes and whole plant
uptake in gametophytes of Danaea wendlandii, Lomariopsis vestita, and Campyloneurum
brevifolium. Immediately prior to the uptake trials, plants were field collected and
brought back to the lab where roots or gametophytes were rinsed with deionized water to
remove soil and other debris. For sporophytes, fine roots were selected, excised, and
placed into a series of solutions containing increasing concentrations of 615N labeled
organic and inorganic N forms (see below). For gametophyte trials, 2-5 individual
gametophytes of similar size and maturity were placed directly into a separate set of
solutions. All samples were allowed to incubate for 60min in solutions containing 0
deionizedd water only), 10, 50, 100, 300, and 500 imol concentrations containing only
labeled NH4+, NO3-, (99 atom%), a cocktail of equal proportions of the amino acids:
Aspartic and Glutamic Acids, and Glycine (98 atom%), and a cocktail of all solutions
(NH4 + NO3- + the 3 amino acids). All solutions were amended with 0.01 mol/L sucrose
as an energy source and 0.5 mmol/L CaC12 to maintain membrane integrity (Kielland
1997). Immediately following the incubation trial, roots and gametophytes were removed
and rinsed for 2min in a solution containing Immol/L KC1 to remove any excess 615N
from the external surfaces. The material was then dried at 60C for 73h, weighed, and
ground for analysis of 615N using the same Finnigan Delta XL PlusTM continuous flow
Nutrient Uptake Calculations
815N enrichment was calculated as F=[T(As-AB)]/AF, where F is the weight of N
derived from the 615N tracer, Tis the total weight of N in the sample, As is atom% excess
615N in the labeled sample, AB is atom% excess 615N in the natural abundance sample,
and AF is atom% excess in the 615N tracer (Knowles and Blackburn, 1993). To calculate
the kinetic uptake parameters of maximum uptake (Vmax) and the saturation constant
(Km), we fitted the data to a Michaelis-Menten function. We also quantified the .mols of
N per g root dry mass against the solution concentration using V= Vmax S/ Km + S, where
V is the velocity of uptake, and S is the concentration. The parameter Vmax is an estimate
of the maximum uptake rate for a given ion and is controlled by the activity of membrane
bound proteins specific to that ion. The value of experimental calculations of Vmax is that
it gives an estimate of the root's/gametophyte's total capacity of ion uptake. The value
Km is estimated to describe the affinity of specific membrane-bound proteins to a given
ion and is related to the capacity to utilize low concentrations of this ion. Roots with
lower Km values have higher affinities at low concentrations for the ion in question.
61sN Natural Abundance and N concentration (mg g-)
Species differed significantly in 615N and N concentration (mg g-l) values (F=3.06,
p=0.0053 and F=3.69, p=0.0009 respectively). There were also significant differences
within and between life forms (615N F=3.74, p=0.029; N concentration (mg g-') F=10.66,
p<0.0001), generations (615N F=15.52, p=0.0002; N concentration (mg g-1) F=57.65,
p<0.0001), and a significant life form by generation interaction (615N F=16.95, p<0.0001;
N concentration (mg g-l) F=8.85, 0.0004) (Fig. 4-1, 4-2a). Within both hemiepiphytic and
terrestrial species, the 615N values of gametophytes were more enriched F=64.93,
p<0.0001 and F=3.84, p=0.0061) than that of sporophtyes; whereas, the opposite was the
case for the epiphytes (F=3.99. p=0.059). Across life forms, the gametophytes of
terrestrial species were slightly depleted yet not significantly so when compared to
epiphytic and hemiepiphytic species (F=2.63, p=0.086). Greatest differentials were
observed among the sporophtyes with hemiepiphytes significantly more depleted than
terrestrial species than epiphytic species (F=17.12, p<0.0001). Gametophytes exhibited
significantly higher N concentration (mg g-1) relative to sporophytes within each life form
and varied between life forms with epiphytes and hemiepiphytes exhibiting higher N
concentration (mg g-1) in gametophytes and sporophtyes than terrestrial species (Fig. 4-
2b). To better understand ontogenetic shifts that occur in hemiepiphytic ferns, we
measured variation in 15N natural abundance of different stages of Lomariopsis vestita.
There was a clear series of increasing 615N enrichment from epiphytic gametophytes and
sporophtyes to terrestrially-rooted adult sporophtyes (Fig. 4-3).
61sN Labeled Uptake
The gametophytes and sporophytes of both Danaea wendlandii and
Campyloneurum brevifolium exhibited uptake capacity of both inorganic and organic
forms of N. For both species uptake of NO3- in both gametophytes and sporophytes was
limited. The gametophytes of both species had higher Vmax for all N forms than
sporophytes. The gametophytes and sporophytes of C. brevifolium had higher Vmax
values for the amino acid mix followed by NH4 and the all solution cocktail (Fig. 4-
4a&b and Fig. 4-5a&b) and all solution cocktails. The gametophytes ofD. wendlandii
had highest Vmax values for the all solution cocktail where the Vmax for NH4+ and amino
acid was similar (Fig. 4-5c). Km values exhibited considerable variation both within and
between species and generations. In all cases, Km was lowest and therefore affinity
highest for NO3- relative to all other N forms (Fig. 4-6a-d).
The results from the uptake studies indicate that ferns show preference for
specific N forms and that they do so differently in sporophyte and gametophyte
generations. In the case of the epiphytic C. brevifolium, both gametophytes and
sporophtyes exhibited high potential for uptake of amino acid N followed by inorganic
NH4+ (Fig. 4-5a-b). Uptake potentials shifted slightly in the gametophytes of the
terrestrial D. wendlandii with uptake of amino acids and NH4+ essentially equal. The
gametophytes and sporophytes of the epiphytic species exhibited higher uptake capacities
for N-derived from amino acids relative to the terrestrial Danaea wendlandii.
Campyloneurum brevifolium is a mid- to low-canopy species that is frequently rooted in
canopy soil; whereas, Danaea wendlandii is a species that is always rooted on mineral
soil. The occurrence of species on such different soil types is likely to produce radically
different nitrogen nutrition and produce different nutrient use strategies between such
Canopy soil is fundamentally different from terrestrial soil in that the former is
almost entirely organic. In a study on soil nutrients at La Selva, Cardelus and Mack (pers.
com.) have shown that canopy soil organic matter had significantly greater bulk nitrogen,
NH4 and dissolved organic nitrogen than terrestrial forest floor soils and that nitrogen
mineralization rates were significantly lower in the canopy. They were also able to show
that within canopy soils, the concentration of NO3 was low and NH4+ and dissolved
organic nitrogen dominate this matrix. For this reason, it is not surprising that epiphytic
species would exhibit preferential uptake ofNH4+ and organic nitrogen.
Both the uptake rate (Vmax) and half saturation constants (Km) were higher for
NH4+ than either amino acids or NO3s for the sporophytes ofDanaea wendlandii (Fig 4-5,
4-6). In the gametophytes of this species, the values of Km for amino acids were several
fold higher than those from the sporophtyes and from the gametophytes and sporophtyes
of Campyloneurum brevifolium. Uptake rates of amino acids and NH4+ from these same
gametophytes were not significantly different. These patterns indicate that NH4+ is an
important N source for both, but especially, terrestrial sporophtyes in this study. NO3s
concentration within terrestrial soils at this site is high (Cardelus and Mack pers. com.).
Nitrate is highly mobile and ferns must compete with microbes and other plants for this
resource and may have partitioned uptake to NH4 and amino acids to avoid or lessen this
competition. Such plants would not be highly invested in NO3s carriers and would be
expected to exhibit higher affinities for this ion at lower concentrations; a result
demonstrated by this study (Fig. 4-6).
The importance of amino acids as components of species' N budgets is receiving
increased attention and has been shown in species from tundra (Kielland, 1994), to
temperate (Finzi and Berthrong, 2005) and subtropical (Schmidt and Stewart, 1999)
ecosystems. We believe that our data are some of the first to demonstrate this ability in
species from tropical lowland forests and clearly the first to do so in the ferns. The
significance of amino acids varies but clearly makes up a critical component (>50% by
some estimates) of species' N budgets from tundra ecosystems (Kielland 1994). In many
ways, canopy soil is functionally equivalent to tundra and boreal soils as it is organic in
origin and potentially has high concentrations of free amino acids. As such, there may be
convergence to greater investment in amino acid uptake in organic soils.
The gametophytes of both species had higher uptake rates of all N forms
excluding NO3- relative to sporophytes. There are fundamental differences in anatomy
and morphology of these two stages of the life cycle and comparisons across stage must
be made with care. Gametophytes produce primitive yet functional rhizoids that likely aid
in nutrient uptake (Smith, 1972a, 1972b), yet they may also take in nutrients via diffusion
across cells of the thallus (Racusen, 2002). This could result in much greater
gametophyte surface area and transporter density compared to root surface area and result
in greater uptake per unit mass in gametophytes. Expression of uptake rates on an area
basis is made difficult as the gametophytes of both species develop complex three
dimensional morphologies that make accurate determination of area difficult.
There are a number of mechanisms that control the natural abundance 615N
signatures in plant tissues and for this reason, 615N signatures reflect a series of integrated
fractionation events (Evans, 2001; Robinson, 2001; Dawson et al., 2002). In spite of the
complexities of interpreting natural abundance 615N signatures, such data provide
evidence of a process when confounding variables that influence fractionation can be
identified or eliminated (Robinson, 2001). Differences in plant 615N signatures have been
shown to be primarily related to 1) uptake of different N sources with distinct signatures
(Robinson, 2001), 2) N availability and plant demand (Kolb and Evans, 2003), 3)
mycorrhizal associations (Hobbie and Hobbie inpress), 4) rooting depth (Nadelhoffer
and Fry, 1994) or a combination of these events (Dawson et al., 2002).
Both N availability and plant demand can have impacts on tissue 615N values
through kinetic fractionations Kolb and Evans (2003) have shown that when N supply is
greater that a plant's assimilatory ability, plant tissues can become highly depleted in 15N
relative to the source. This, however, only occurs in ecosystems where external N
concentrations are high and such kinetic fraction seems unlikely to be responsible for the
differences observed in our study. It is possible that mechanisms other than supply and
demand drive kinetic fractionations differently in gametophytes and sporophtyes.
Unfortunately, little is known of how root vs. rhizoid/thallus uptake differentially
Ectomycorrhizal associations can result in large fractionation events (8-10 /oo)
whereas fractionation caused by arbuscular mycorrhizal symbioses seems to have little
effect on 615N values of host plants (Schmidt and Stewart, 2003). Both terrestrial and
epiphytic leptosporangiate fern species have arbuscular mycorrhizal symbioses in the
sporophtyes, but never in the gametophytes (Gemma, Koske, and Flynn, 1992). The
differences in natural abundance 615N signatures between gametophytes and sporophtyes
are unlikely due to such associations.
Plant uptake from source pools with different 615N signatures is a major
contributor influencing tissue values. Plants can take up both organic and inorganic
nitrogen and each of these sources has different signatures: NO3 is highly depleted
relative to NH4+ which can be more depleted that amino acids. In our data, gametophyte
natural abundance 615N signatures are either significantly depleted or equal to sporophyte
values. The data from the uptake experiment indicate that NO3 is not a major source of
either gametophyte or sporophtyes N nutrition and that source alone is not responsible for
the differences between gametophyte and sporophtyes. Sources can however be derived
from N species from enriched soil or highly depleted atmospheric sources. Evidence
indicates that epiphytes rely heavily on depleted atmospheric N sources (Hietz et al.,
2002) and the result is major offset of 15N values when compared to terrestrially rooted
plants (Watkins, unpublished data). This may also help explain the differences between
gametophytes and sporophtyes. As throughfall leaches through the canopy to the forest
floor, it may contain significant concentrations of depleted N (Cardelus and Mack pers.
com.) which may then be taken up more directly by gametophytes. Canopy epiphytic
gametophytes also rely on depleted N species and may exhibit greater direct atmospheric
uptake than sporophytes. Our uptake data indicate that overall uptake rate in
gametophytes is much greater than that in sporophtyes. This may provide gametophytes
greater opportunity for uptake of depleted pools that may be rapidly diluted by rainfall.
Several studies have now shown that tissue 615N values from deep rooted
individuals are more enriched relative to shallow rooted individuals (Nadelhoffer and
Fry, 1988; Nadelhoffer and Fry, 1994; Handley and Scrimgeour, 1997). Rooting depth is
a potentially major contributor to the observed differences in 615N values between the
surface rooted gametophytes and deeper sporophyte roots.
One way to observe shifts in N sources is to follow ontogenetic series and track
changes in 615N values. The gametophytes of Lomariopsis vestita are epiphytes on
understory trees and produce epiphytic sporophtyes with roots that grow down the trunk
into the soil. The difference between gametophytes and young sporophtyes that were not
attached to the soil were large with the latter being more enriched (Fig. 4-3). This is
possibly due to a combination of atmospheric uptake in gametophytes and quickly shifts
to direct root uptake in sporophtyes that may be more efficient in tapping nutrients from
more enriched soil pools. Terrestrially rooted young and adult individuals would be
predicted to exhibit enriched 815N signatures relative to epiphytic individuals. Such
plastic abilities are critical in the life of hemiepiphytes that face frequent water stress and
a temporally heterogeneous nutrient environment.
The observed differences in foliar 815N values and evidence of differential uptake
of N forms indicate that ferns can partition N by form. The preference of N form varied
with greater preference on organic N in the epiphytic vs. terrestrial species and between
gametophytes and sporophtyes. These data indicate that there are different nutrient use
strategies between the two life forms and between generations. Individuals that can
circumvent mineralization in the N cycle by direct amino acid uptake may have a
tremendous competitive advantage relative to those relying on inorganic forms. What is
critically needed are studies that incorporate dual labeled C and N isotopes to precisely
determine if amino acids are hydrolyzed at the cell membrane or if they are taken directly
Adiantum latifolium Lam.
Danaea nodosa (L.) Sm.
Danaea wendlandii Rchb. F.
Diplazium subsilvaticum H. Christ
Lomariopsisjapurensis (Mart.) J. Sm.
Lomariopsis vestita E. Foum
Olfersia cervina (L.) Junze
Elaphoglossum latifolium (Sw.) J. Sm.
Campyloneurum brevifolium (Lodd. Ex Link)
Antrophyum lineatum (Sw.) Kaulf
Disturbed areas, high medium light
Understory, low light
Understory, disturbed low light areas
Understory exposed mesic areas
Understory secondary and primary forests
Understory secondary and primary forests
Understory, on mounds of organic matter or decaying trees
Highly exposed, on bare bark in canopy
Moderate light, rooted in soil, inner canopy
Understory secondary and primary forests
Table 4-1. Species, life form and ecology for the natural abundance and uptake experiments
Figure 4-1. Sporophtyic and Gametophytic 615N natural abundance signatures of 10
tropical fern species. Tissue was field collected from 100X100m grids to
control for differences in soil type in the terrestrial species and from the same
trees in the case of epiphytic species
7, T r
Epiphyte Hemi-Epiphyte Terrestrial
Epiphyte Hemi-Epiphyte Terrestrial
Figure 4-2. Sporophtyic and Gametophytic (a) 615N natural abundance signatures and (b)
N concentration (mg g-) of epiphytic, terrestrial, and hemiepiphytic tropical
fern species. Post hoc test were generated using Tukey tests. Capitol letters
refer to within life-form whereas lower case letter refer to across life form
S Gametophyte Bb
T Ba Aa
-1 i i
Gametophytes Young Young Adult
Sporophytes Sporophytes Sporophytes
Not Attached Attached Attached
Figure 4-3. 615N natural abundance signatures from the hemiepiphytic fern Lomariopsis
vestita. Gametophytes of this species are completely epiphytic on understory
trees. Young sporophtyes are initially produced that have true roots, but no
connection to the forest floor at very early stages. Young sporophtyes
eventually attach to the soil and rooting depth increases with adult plants
0 100 200 300 400 500
N Concentration (umol/L)
C DW Gametophyte
0 100 200 300 400 500
N Concentration (umol/L)
B CB Sporophyte
0 100 200 300 400 500
N Concentration (umol/L)
D DW Sporophyte
0 100 200 300 400 500
N Concentration (umol/L)
-- Amino Acid Mix
-- All Solutions
Figure 4-4. Uptake curves from 615N labeled solutions. Fine roots were selected, excised,
and placed into a series of solutions containing increasing concentrations of
815N labeled organic and inorganic N forms. For the gametophyte trials
individual gametophytes of similar size and maturity were placed directly into
a separate set of solutions. All samples were allowed to incubate for 60min in
solutions containing only 15N labeled NH4 NO3-, (99 atom%), a cocktail of
equal proportions of the amino acids: Aspartic and Glutamic Acids, and
Glycine (98 atom%), and a cocktail of all solutions (NH4+ + NO3- + the 3
A CB Gametophyte
NH4 N03 AA
NH4 N03 AA
D F=61.55, p<0.0001
NH4 N03 AA All
Figure 4-5. Uptake saturation values (Vmax) of each N form derived from Michaelis-
Menten functions of the data from Fig. 4-2, error bars are standard errors.
Campyloneurum brevifolium is a mid-canopy epiphyte of exposed habitats;
Danaea wendlandii is a low-light understory terrestrial species
NH4 N03 AA All
C F=13.0, p=0.0019
.__L -- -
NH4 N03 AA All NH4 N03 AA All
Nitrogen Form Nitrogen Form
NH4 N03 AA
NH4 N03 AA
Figure 4-6. V2 Uptake saturation values (Km) of each N form derived from Michaelis-
Menten functions of the data from Fig. 4-2, error bars are standard errors.
Campyloneurum brevifolium is a mid-canopy epiphyte of exposed habitats;
Danaea wendlandii is a low-light understory terrestrial species.
This dissertation is one of the first attempts to examine and apply modern
ecological and ecophysiological techniques to the study of fern gametophyte ecology.
This work has demonstrated that fern gametophytes can be extremely long-lived in situ
and that there are differences in factors that influence the distribution and demography of
epiphytic and terrestrial ferns. Differences in life history and the way that epiphytic and
terrestrial life-forms respond to disturbance and light provide evidence for adaptively
meaningful variation in life histories that has evolved in the two groups. Epiphytic
species have evolved in a high light, highly competitive, yet relatively stable matrix. Such
environments reduce the light limitations encountered by terrestrial species, yet they
incorporate closer contact with bryophytes. These habitat mediated conditions may be
largely responsible for the observed variation in longevity.
Dassler and Farrar (1997) have argued that differences in gametophyte longevity
between epiphytic and terrestrial species have largely evolved due to pressures from the
genetic consequences of intergametophytic selling. Asexually reproducing indeterminate
gametophytes of many epiphytes can produce large and long-lived clones. Such clones
greatly increase the longevity of individual genotypes which is hypothesized to increase
the chance of outcrossing. The data generated from chapter one clearly show that
epiphytic gametophytes are significantly longer lived than terrestrial species. One
remarkable discovery is that epiphytic gametophytes can live for years where even
terrestrial species on relatively stable substrates rarely lived beyond 6 months. I have
observed gametophytes of some understory epiphytes that are over 6 years old.
Are the differences in longevity between epiphytes and terrestrial species related to
some adaptively meaningful variation between the life-forms as has been hypothesized,
or is it simply a result of some intrinsic instability of epiphytic vs. terrestrial habitats?
The answer to this question remains elusive as there is no clear understanding of the
differences in disturbance between epiphytic and terrestrial habitats. The data in my study
would indicate that epiphytic habitats are far more stable than terrestrial habitats. This
clearly needs to be established and a much greater survey of demography needs to be
undertaken to better understand the extent of the differences that I have reported.
Gametophytes that can live for months and especially those that live for years have
to cope with stress associated with extreme abiotic variation. The second part of my
dissertation has revealed a surprising and completely unexpected degree of extreme
gametophyte stress tolerance to desiccation across several species. All species surveyed
exhibited more desiccation tolerance than current pteridological dogma would suggest. In
addition, such tolerance was clearly linked to species sporophyte ecology with those from
drought prone habitats, such as the epiphytes in the study, exhibiting greater degrees of
tolerance compared to those in mesic habitats. Epiphytic species were also robust in
dealing single dry down events and exhibited significantly greater recovery following
extreme desiccation intensities and multiple desiccation cycles compared to more mesic
terrestrial species. Not only are epiphytic species longer-lived, they are also considerably
more desiccation tolerant: two characters that are likely connected.
Stress tolerance has been show to structure some bryophyte communities (Cleavitt,
2002) but I have been unable to find any additional work to suggest that ferns are sorting
along dines of stress tolerance. Much additional work needs to be completed on
gametophyte physiology to truly understand the role that gametophytic longevity and
desiccation tolerance plays in sorting species. Additional studies need to include
combinations of temperature and light stress with desiccation to develop a better
understanding of the interaction of these characters and the relative tolerance of more
species. This work also has basic science applications beyond ferns and can be applied to
aspects directly related to genetic engineering of desiccation tolerance in crop plants.
One factor that seems closely tied to sorting of terrestrial fern sporophtyes are
edaphic factors (Tuomisto, 1998, Tuomisto, 1994, Tuomisto, 2002). The role that
nutrients play in shaping fern gametophyte distributions is virtually unknown. My work
on nitrogen relations of tropical fern gametophytes has revealed unexpected versatility in
nitrogen acquisition between both gametophytes and sporophtyes and between epiphytic
and terrestrial species. Of great significance was the discovery that ferns can partition
nitrogen by form and have the ability to take up amino acids and use them as an
important component of their nitrogen budgets. The importance of amino acid uptake as
critical components of species' N budgets is currently receiving increased attention and
has been shown in species from tundra (Kielland, 1994), to temperate (Finzi and
Berthrong, 2005) and subtropical (Schmidt and Stewart, 1999) ecosystems. Nitrogen
associated with amino acids is clearly important in both the N cycle of tropical fern
gametophytes and sporophtyes and such flexibility in accessing different nitrogen forms
may provide species with differential competitive abilities result in one mechanism by
which species are sorted along nutrient gradients. What is critically needed are studies
that incorporate dual labeled C and N isotopes to precisely determine if amino acids are
hydrolyzed at the cell membrane or if they are taken directly into plan.
LIST OF REFERENCES
ALPERT, P. 2000. The discovery, scope, and puzzle of desiccation tolerance in plants.
Plant Ecology 151: 5-17.
2005. The limits and frontiers of desiccation-tolerant life. Integrative and
Comparative Biology 45: 685-695.
ALPERT, P., and M. J. OLIVER. 2002. Drying without dying. In M. Black and H. W.
Pritchard [eds.], Desiccation and Survival in Plants, 3-43. CABI Publishing,
ATKINSON, L. R.,and A. G. STOKEY. 1964. Comparative morphology of the gametophytes
of homosporous ferns. Phytomorphology 14: 51-70.
BEWLEY, J. D. 1979. Physiological-aspects of desiccation tolerance. Annual Review of
Plant Physiology and Plant Molecular Biology 30: 195-238.
BOLD, H. C. 1957. Morphology of Plants. Harper and Row, New York.
BOWER, F. O. 1923. The Ferns. Cambridge University Press, Cambridge.
CARDELUS, C. L., and R. CHAZDON. 2005. Inner-crown microenvironments of two
emergent tree species in a lowland wet forest. Biotropica 37: 238-244.
CASPER, B. B., I. N. FORSETH, and D. A. WAIT. 2005. Variation in carbon isotope
discrimination in relation to plant performance in a natural population of
Cryptantha flava. Oecologia 145: 541-548.
CHAPIN, F. S., L. MOILANEN, and K. KIELLAND. 1993a. Preferential use Of organic
nitrogen for growth by a nonmycorrhizal arctic sedge. Nature 361: 150-153.
1993b. Preferential Use of Organic Nitrogen for Growth by a Non-Mycorrhizal
Arctic Sedge. Nature 361: 150-153.
CHIOU, W.-L., and D. R. FARRAR. 1997. Comparative gametophyte morphology of
selected species of the family Polypodiaceae. American Fern Journal 87: 77-86.
CLEAVITT, N. L. 2002. Stress tolerance of rare and common moss species in relation to
their occupied environments and asexual dispersal potential. Journal of Ecology
COUSENS, M. I. 1979. Gametophyte ontogeny, sex expression, and genetic load as
measures of population divergence in Blechnum spicant. American Journal of
1981. Blechnum spicant: habitat and vigor of optimal, marginal, and disjunct
populations and field observations of gametophytes. Botanical Gazette 142: 251-
1988. Reproductive strategies of pteridophytes. In J. L. Doust and L. L. Doust
[eds.], Plant Reproductive Ecology: Patterns and Strategies, 307-328. Oxford
University Press, New Youk.
COUSENS, M. I., D. G. LACEY, and E. M. KELLY. 1985. Life-History Studies Of Ferns A
Consideration Of Perspective. Proceedings Of The Royal Society Of Edinburgh
Section B-Biological Sciences 86: 371-380.
COUSENS, M. I., D. G. LACEY, and J. M. SCHELLER. 1988. Safe sites and the ecological
life history ofLorinseria areolata. American Journal ofBotany 76: 797-807.
CRIST, K. C., and D. R. FARRAR. 1983. Genetic load and long-distance dispersal in
Aspleniumplatyneuron. Can. J Bot. 61: 1809-1814.
CSINTALAN, Z., M. C. F. PROCTOR, and Z. TUBA. 1999. Chlorophyll fluorescence during
drying and rehydration in the mosses Rhytidiadelphus loreus (Hedw.) Warnst.,
Anomodon viticulosus (Hedw.) Hook. & Tayl. and Grimmia pulvinata (Hedw.)
Sm. Annals ofBotany 84: 235-244.
DASSLER, C. L., and D. R. FARRAR. 1997. Significance of form in fern gametophytes:
clonal, gemmiferous gametophytes of Callistopteris baueriana
(Hymenophyllaceae). International journal of plant sciences 158: 622-639.
2001. Significance of gametophyte form in long-distance colonization by
tropical, epiphytic ferns. Brittonia 53: 352-369.
DAWSON, T. E., S. MAMBELLI, A. H. PLAMBOECK, P. H. TEMPLER, and K. P. Tu. 2002.
Stable isotopes in plant ecology. AnnualReview of Ecology and Systematics 33:
DELTORO, V. I., A. CALATAYUND, G. GIMENO, and E. BARRENO. 1998. Water relations,
chlorophyll fluorescence, and membrane permeability during desiccation in
bryophytes from xeric, mesic and hydric environments. Canadian Journal of
Botany 76: 1923-1929.
DEMMIG-ADAMS, B., and W. W. ADAMS. 1992. Photoprotection and Other Responses of
Plants to High Light Stress. Annual Review ofPlant Physiology and Plant
Molecular Biology 43: 599-626.
DEMMIG-ADAMS, B., and W. W. ADAMS III. 1992. Photoprotection and other responses of
plants to high light stress. Annu. Rev. Plant Physiol. Plant Mol. Biol. 43: 599-626.
EVANS, R. D. 2001. Physiological mechanisms influencing plant nitrogen isotope
composition. Trends in Plant Science 6: 121-126.
FARRAR, D. R. 1967. Gametophytes of four tropical fern genera reproducing
independently of their sporophytes in the southern Appalachians. Science 155:
1971. The biology of ferns with asexually reproducing gametophytes in the
eastern United States. Ph.D. Dissertation. Ph.D. Dissertation, University of
Michigan, Ann Arbor, MI.
1990. Species and evoluiton in asexually reproducing independent fern
gametophytes. Systematic Botany 15: 98-111.
1998. The tropical flora of rockhouse cliff formations in the eastern United
States. Journal of the Torrey Botanical Club 125: 91-108.
FIELD, T. S., R. O. LAWTON, and T. E. DAWSON. 1996. Comparative nutrient relations in
canopy-rooted and ground-rooted Didymopanax pittieri Marchal (Araliaceae)
hemiepiphytes in a wind-exposed tropical montane rain forest. Biotropica 28:
FINZI, A. C., and S. T. BERTHRONG. 2005. The uptake of amino acids by microbes and
trees in three cold-temperate forests. Ecology 86: 3345-3353.
FRAZER, G. W., C. D. CANHAM, P. SALLAWAY, and D. MARINAKIS. 1999. Gap Light
Analyzer. Simon Fraser University & Institute for Ecosystem Studies, Burnaby,
British Columbia, Canada & Millbrook, New York.
GAFF, D. 1997. Mechanisms of desiccation tolerance in resurrection plants. In A. S. Basra
and R. K. Basra [eds.], Mechanisms of environmental stress resistance in plants,
43-58. Harwood Academic Publishers, London.
GAFF, D. F. 1987. Desiccation tolerant plants in South America. Oecologia (Berlin) 74:
GEMMA, J. N., R. E. KOSKE, and T. FLYNN. 1992. Mycorrhizae in Hawaiian
pteridophytes: occurrence and evolutionary significance. American Journal of
Botany 79: 843-852.
GILBERT, O. L. 1970. Biological flora of the British Isles. Dryopteris villarii (Bellardi)
Woynar (Aspidium rigidum.; Lastrea rigidia (Sw.) C. Presl).
GRAYUM, M. H., and H. W. CHURCHILL. 1987. An introduction to the Pteridophyte flora
of Finca La Selva, Costa Rica. American Fern Journal 77: 73-89.
GREER, G. K., and B. C. MCCARTHY. 1999. Gametophytic plasticity among four species
of ferns with contrasting ecological distributions. International Journal of Plant
Sciences 160: 879-886.
HANDLEY, L. L., and C. M. SCRIMGEOUR. 1997. Terrestrial plant ecology and N-15
natural abundance: The present limits to interpretation for uncultivated systems
with original data from a Scottish old field, Advances in Ecological Research, Vol
HARTEN, J. B., and W. G. EICKMEIER. 1987. Comparative Desiccation Tolerance of 3
Desert Pteridophytes Response to Long-Term Desiccation. American Midland
Naturalist 118: 337-347.
HIETZ, P., and 0. BRIONES. 1998a. Correlation between water relations and within-
canopy distribution of epipytic ferns in a Mexican cloud forest. Oecologia 114:
1998b. Correlation between water relations and within-canopy distribution of
epiphytic ferns in a Mexican cloud forest. Oecologia 114: 305-316.
HIETZ, P., and W. WANEK. 2003. Size-dependent variation of carbon and nitrogen isotope
abundances in epiphytic bromeliads. Plant Biology 5: 137-142.
HIETZ, P., W. WANEK, R. WANIA, and N. M. NADKARNI. 2002. Nitrogen-15 natural
abundance in a montane cloud forest canopy as an indicator of nitrogen cycling
and epiphyte nutrition. Oecologia 131: 350-355.
HOLBROOK-WALKER, S. G., and R. M. LLOYD. 1973. Reproductive biology and
gametophyte morphology of the Hawaiian fern genus Sadleria (Blechnaceae)
relative to habitat diversity and propensity for colonization. Botanical Journal of
the Linnean Society 67: 157-174.
HOLLANDER, M., and D. A. WOLF. 1999. Nonparametric statistical methods. John Wiley
and Sons Inc., New York.
HOLTTUM, R. E. 1938. The ecology of tropical pteridophytes. In F. Verdoorn [ed.],
Manual of pteridology., 433. M. Nijhoff, The Hague.
HORTON, P., A. V. RUBAN, and R. G. WALTERS. 1996. Regulation of light harvesting in
green plants. Annual Review OfPlant Physiology And Plant Molecular Biology
JONES, M. M., H. TUOMISTO, D. B. CLARK, and P. OLIVAS. 2006. Effects of mesoscale
environmental heterogeneity and dispersal limitation on floristic variation in rain
forest ferns. Journal of Ecology 94: 181-195.
KAPPEN, L. 1964. Untersuchungen uber die Jahreslauf der Frost-, Hitze- und
Austrocknungsresistenz vin sporophyten einheimischer Polypodiaceen. Flora
KIELLAND, K. 1994. Amino acid absorption by artic plants: implications for plant
nutrition and nitrogen cycling. Ecology 75: 2373-2383.
1997. Role of free amino acids in the nitrogen economy of artic cryptogams.
Ecoscience 4: 75-79.
KLEKOWSKI, E. J. J. 1969. Reproductive biology of the pteridophyta. II. Theoretical
considerations. Botanical Journal of the Linnean Society 63: 347-359.
KLOCKARS, A. J., and G. SAX. 1986. Multiple comparisons. Sage Publications., Thousand
KNOWLES, R., and T. H. BLACKBURN. 1993. Nitrogen isotope techniques. Academic
Press, New York.
KOLB, K. J., and R. D. EVANS. 2003. Influence of nitrogen source and concentration on
nitrogen isotopic discrimination in two barley genotypes (Hordeum vulgare L.).
Plant Cell And Environment 26: 143 1-1440.
LIPSON, D., and T. NASHOLM. 2001. The unexpected versatility of plants: organic
nitrogen use and availability in terrestrial ecosystems. Oecologia 128: 305-316.
MILLER, J. H. 1968. Fern gametophytes as experimental material. Botanical Review 34:
MOTTIER, D. M. 1927. The behavior of certain fern prothallia under prolonged
cultivation. Botanical Gazette 83: 244-266.
MULKEY, S. S., and R. W. PEARCY. 1992. Interactions between acclimation and
photoinhibition of photosynthesis of a tropical forest understory herb, Alocasia
macrorrhiza, during simulated canopy gap formation. Funct. Ecol. 6: 719-729.
NADELHOFFER, K. F., AND B. FRY. 1988. Controls On Natural N-15 And C-13
Abundances In Forest Soil Organic-Matter. Soil Science Society Of America
Journal 52: 1633-1640.
NADELHOFFER, K. J., and B. FRY. 1994. Nitrogen isotope studies in forest ecosystems. In
K. Lajtha and R. H. Michener [eds.], Stable Isotopes in ecology and
environmental science, 22-44. Blackwell scientific publications, Oxford.
NADKARNI, N. M., and T. MATELSON. 1991. Litter dynamics within the canopy of a
neotropical cloud forest, Monteverde, Costa Rica. Ecology 72: 849-860.
NAYAR, B. K., and S. KAUR. 1971. Gametophytes of homosporous ferns. The Botanical
Review 37: 295-396.
NIKLAS, K. 1997. The Evolutionary Biology of Plants. University of Chicago, Chicage.
NOBEL, P. S. 1978. Microhabitat, Water Relations, and Photosynthesis of a Desert Fern,
Notholaena-Parryi. Oecologia 31: 293-309.
OLIVER, M. J., B. D. MISHLER, and J. E. QUISENBERRY. 1993. Comparative Measures of
Desiccation-Tolerance in the Tortula-Ruralis Complex. 1. Variation in Damage
Control and Repair. American Journal ofBotany 80: 127-136.
OLIVER, M. J., Z. TUBA, and B. D. MISHLER. 2000. The evolution of vegetative
desiccation tolerance in land plants. Plant Ecology 151: 85-100.
ONG, B. L., and M. L. NG. 1998. Regeneration of drought-stressed gametophytes of the
epiphytic fern, Pyrrosia pilosellodes (L.) Price. Plant cell reports 18: 225-228.
PECK, J. H. 1980. Life history and reproductive biology of the ferns of Woodman Hollow,
Webster County, Iowa. Ph.D. Dissertation, Iowa State University, Ames.
PECK, J. H., C. J. PECK, and D. R. FARRAR. 1990. Influences of life history attributes on
formation of local and distant fern populations. American Fern Journal 80: 126-
PICKETT, F. L. 1913. Resistance of the prothallia of Camptosorus rhizophyllus to
desiccation. Bul. Torrey Bot. Club 40: 641-645.
1914. Some ecological adaptations of certain fern prothallia Camptosorus
rhizophyllus Link., Asplenium platyneuron Oakes. American Journal ofBotany 1:
1931. Notes on xerophytic ferns. America Fern Journal 21: 49-57.
PILLAI, R. S., and B. L. ONG. 1999. Effects of inorganic nitrogen availability on the
sporophytes of Acrostichum aureum L. Ph1,niy,,1nthetiiI 36: 259-266.
POREMBSKI, S., and W. BARTHLOTT. 2000. Granitic and gneissic outcrops (inselbergs) as
centers of diversity for desiccation-tolerant vascular plants. Plant Ecology 151:
PRANGE, R. K., and D. P. ORMROD. 1982. Effects Of Ammonium And Nitrate Nutrition
On The Ostrich Fern (Matteuccia-Struthiopteris). Canadian Journal OfPlant
Science 62: 195-201.
PROCTOR, M. 2001. Patterns of desiccation tolerance and recovery in bryophytes. Plant
G1 ,fi I i Regulation 35: 147-156.
PROCTOR, M. C. F. 2000. The bryophyte paradox: tolerance of desiccation, evasion of
drought. Plant Ecology 151: 41-49.
2003. Comparative ecophysiological measurements on the light responses, water
relations and desiccation tolerance of the filmy ferns Hymenophyllum wilsonii
Hook. and H-tunbrigense (L.) Smith. Annals ofBotany 91: 717-727.
PROCTOR, M. C. F., and V. PENCE. 2002. Vegetative tissues: Bryophytes, vascular
resurrection plants and negative propagules. In M. Black and H. W. Prichard
[eds.], Desiccation and survival in plants: drying without dying, 207-237. CAB
International, Wallingford, UK.
PRYER, K. M., A. R. SMITH, and J. E. SKOG. 1995. Phylogenetic relationships of extant
ferns based on evidence from morphology and rbcL sequences. American Fern
Journal 85: 205-282.
PUTZ, F. E., and N. M. HOLBROOK. 1989. Strangler fig rooting habits and nutrient
relations in the llanos of Venezuela. American Journal ofBotany 76: 781-788.
RACUSEN, R. H. 2002. Early development in fern gametophytes: Interpreting the
transition to prothallial architecture in terms of coordinated photosynthate
production and osmotic ion uptake. Annals ofBotany 89: 227-240.
REICH, A., J. J. EWEL, N. M. NADKARNI, T. DAWSON, and R. D. EVANS. 2003. Nitrogen
isotope ratios shift with plant size in tropical bromeliads. Oecologia 137: 587-
ROBINSON, D. 2001. delta N-15 as an integrator of the nitrogen cycle. Trends In Ecology
& Evolution 16: 153-162.
ROBINSON, S. A., J. WASLEY, M. POPP, and C. E. LOVELOCK. 2000. Desiccation tolerance
of three moss species from continental Antarctica. Australian Journal of Plant
Physiology 27: 379-388.
SANFORD, R. L., P. PAABY, J. C. LUVALL, and E. PHILLIPS. 1994. Climate,
geomorphology, and aquatic systems. In L. A. McDade, K. S. Bawa, H. A.
Hespenheide, and G. S. Hartshorn [eds.], La Selva: Ecology and Natural History
of a Neotropical Rain Forest, 19-33. University of Chicago Press, Chicago.
SAS-INSTITUTE. 2005. JMP User's Guide. SAS Instite, Inc., Cary.
SATO, T., and A. SAKAI. 1980. Freezing resistance of gametophytes of the temperate fern,
Polystichum retroso-paleaceum. Canadian journal of botany 58: 1144-1148 ill.
1981. Cold tolerance of gametophytes of some cool temperare ferns native to
Hokkaido. Can. J. Bot. 59: 604-608.
SCHEINER, S. M., and J. GUREVITCH. 2001. Design and analysis of ecological
experiments. Oxford University Press, Oxford.
SCHMIDT, A., and G. R. STEWART. 2003. Delta 15N values of tropical savanna and
monsoon forest species reflect root specilizations and soil nitrogen status.
Oecolgia 134: 569-577.
SCHMIDT, G., S. STUNTZ, and G. ZOTZ. 2001. Plant size: an ignored parameter in epiphyte
ecophysiology? Plant Ecology 153: 65-72.
SCHMIDT, S., and G. R. STEWART. 1999. Glycine metabolism by plant roots and its
occurrence in Australian plant communities. Australian Journal OfPlant
Physiology 26: 253-264.
SMITH, D. L. 1972a. Localization of phosphatases in young gametophytes of Polypodium
vulgare L. Protoplasma 74: 133-148.
1972b. Staining and osmotic properties of young gametophytes of Polypodium
vulgare L. and their bearing on rhizoid function. Protoplasma 74: 465-479.
TRESEDER, K. K., and P. M. VITOUSEK. 2001. Effects of soil nutrient availability on
investment in acquisition of N and P in Hawaiian rain forests. Ecology 82: 946-
TUOMISTO, H., and K. RUOKOLAINEN. 1994a. Distribution of Pteridophyta and
Melastomataceae Along an Edaphic Gradient in an Amazonian Rain Forest.
Journal of Vegetation Science 5: 25-34.
1994b. Distribution of Pteridophyta and Melastomataceae Along an Edaphic
Gradient in an Amazonian Rain-Forest. Journal of Vegetation Science 5: 25-34.
TUOMISTO, H., and A. D. POULSEN. 1996. Influence of edaphic specialization on
pteridophyte distribution in neotropical rain forests. Journal of biogeography 23:
TUOMISTO, H., and A. DALBERG. 1996. Influence of edaphic specialization on
pteridophyte distributions in neotropical rain forests. Journal ofBiogeography 23:
TUOMISTO, H., and A. D. POULSEN. 2000. Pteridophyte diversity and species composition
in four Amazonian rain forests. Journal of vegetation science: official organ of
the International Association for Vegetation Science 11: 3 83-396.
TUOMISTO, H., A. D. POULSEN, and R. C. MORAN. 1998. Edaphic distribution of some
species of the fern genus Adiantum in western Amazonia. Biotropica 30: 392-
WALKER, L. R., and G. H. APLET. 1994. Growth And Fertilization Responses Of
Hawaiian Tree Ferns. Biotropica 26: 378-383.
WALP, R. L. 1951. Fern Prothallia under Cultivation for 12 Years. Science 113: 128-129.
WANEK, W., S. K. ARNDT, W. HUBER, and M. POPP. 2002. Nitrogen nutrition during
ontogeny of hemiepiphytic Clusia species. Functional Plant Biology 29: 733-740.
WATANABE, M., T. KIKAWADA, N. MINAGAWA, F. YUKUHIRO, and T. OKUDA. 2002.
Mechanism allowing an insect to survive complete dehydration and extreme
temperatures. Journal OfExperimental Biology 205: 2799-2802.
WATKINS, J. E., and D. R. FARRAR. 2005. Origin and taxonomic affinities of Thelypteris
(subgen. Stegnogramma) burksiorum (Thelypteridaceae). Brittonia 57: 183-201.
WATKINS, J. E., C. CARDELUS, R. K. COLWELL, AND R. C. MORAN. 2006. Species
richness and distribution of ferns along an elevational gradient in Costa Rica.
American Journal ofBotany 93: 73-83.
WATKINS JR., J. E., and D. R. FARRAR. 2005. Origin and taxonomic affinities Thelypteris
(subgen. Stegnogramma) burksiorum (Thelypteridaceae). Brittonia 57: 183-201.
ZOTZ, G. 1997. Photosynthetic capacity increases with plant size. Botanica Acta 110:
ZOTZ, G., A. ENSLIN, W. HARTUNG, and H. ZIEGLER. 2004. Physiological and anatomical
changes during the early ontogeny of the heteroblastic bromeliad, Vriesea
sanguinolenta, do not concur with the morphological change from atmospheric to
tank form. Plant Cell And Environment 27: 1341-1350.
James Edward (Eddie) Watkins, Jr. was born on 12 March, 1974 in Ozark, Dale
County, Alabama. He attended the now-condemned Flowers Elementary School where
his first introduction to science was a project on turtles by his first grade teacher Mrs.
Hopper. He graduated to attend D.A. Smith Middle School. It was there, under direction
of his eighth grade teacher, Dena Byers he first began to develop an understanding of
scientific experiment. He competed in several regional science fairs, making it as far at
the Alabama State Science Fair for his work on factors controlling the rate at which mice
could exit a complicated maze.
During these years, he spent much of his time fishing, hunting, building forts, and
observing nature from his daily hikes in to the forests surrounding his home. During these
early years he developed a true connection with the natural world. He gained his early
understanding of how this world was put together by his first biological mentor Ms.
Linda Dees: science teacher at Carroll High School. During these formative high school
years he began to put together what he would do for the rest of his life. One of the most
important developments came when Ms. Dees required his freshman biology class to
complete a plant collection. This Eddie did by only collecting live ferns that were then
transplanted into the school's nature preserve. In the course of this collection, he
discovered two of the rarest ferns in Alabama and went on to publish some of this work
in a peer-reviewed journal. During his early fern forays, he made the acquaintance of
Professor Warren Herb Wagner, Jr.; the world's leading fern authority at the time, and
continues to inspire Eddie's work today. After graduation, he attended Auburn
University, where he worked under the direction of Dr. John D. Freeman and Dr. Robert
S. Boyd. In the lab and courses of Dr. Boyd that Eddie was finally able to put nature and
experimental sciences together and begin to develop an understanding of the complexities
of ecology. After graduation, he attended Iowa State University under Dr. Donald Farrar
where he attained an MS degree with his thesis on Thelypteris burksiorum. These years
were paramount to his Pteridological development, as Dr. Farrar is one of the last old-
school fern biologists and was as smitten with ferns as Eddie. Eddie graduated in 2000
and spent a year living with his wife Catherine in Costa Rica, studying the magnificent
array of ferns at La Selva Biological Station and beyond. He then returned to the South
where he began his doctoral studies with Dr. Stephen Mulkey and Dr. Michelle C. Mack.
After completing of his doctoral studies, Eddie will begin a post doctoral fellowship in
the lab of Michelle Holbrook at Harvard University.