HYDROPHILY: PHYLOGENETIC AND EVOLUTIONARY CONSIDERATIONS

15 Pages·1991·8.3 MB·English

by  C Thomas Philbrick

RHODORA,
Vol. 93, No. 873, pp. 36-50, 1991
AND
HYDROPHILY: PHYLOGENETIC
EVOLUTIONARY
CONSIDERATIONS
Thomas
Philbrick
C.
ABSTRACT
The two forms of abiotic pollination, hydrophily and anemophily, exhibit con-
trasting taxonomic, ecological and phylogenetic Anemophily wide-
patterns.
is
among
spread angiosperms whereas hydrophily occurs in only one dicot and seven
monocot
families. Ecological limitations of hydrophily likely parallel those of
anemophily,
yet the processes involved in the former are not well understood.
Although hydrophily
in probability polyphyletic, the phylogenetic
is all relation-
among
ships hydrophilous groups arc unresolved. Specialization reproductive
in
accompany
structures that hydrophilous pollination makes the recognition of
homology
difficult, thus phylogenetic hypotheses in groups in which hydrophily
occurs are tentative. Hydrophily and anemophily are both geographically wide-
spread, yet the general trend of decreasing incidence of anemophily with decreasing
may
latitude is lacking in the distribution of hydrophily. This contrast be asso-
ciated with markedly different geographic patterns of species richness aquatic
in
versus
terrestrial habitats.
Key
Words:
hydrophily,
species richness, geographic distribution, phylogcny,
evolution
INTRODUCTION
Hydrophily, water- mediated cross-pollination, entails dramatic
modifications of the systems of angiosperms. These
floral terrestrial
changes from
adaptation provide
arise to for the release, transport
and
capture of water-borne, often wet, pollen. Hydrophilous taxa
are infrequent in angiosperms; only 140 of the of 225,000
total ca.
angiosperms Thorne, comm.)
(R. hydrophilous
pers. are
The
(.00062%). infrequency of hydrophily has no doubt
contrib-
uted to the perception that this pollination system "unimpor-
is
angiosperm
tant" in evolution. In an understanding of hy-
fact,
may
drophily provide unique on
a perspective the evolution of
angiosperm
reproductive systems. Hydrophilous systems
are per-
haps examples
the best of the evolutionary "plasticity'' of the
angiosperm
floral biology; in hydrophily the "aerial"
i.e., floral
system that dominates angiosperms has been aban-
essentially
doned. Furthermore, the occurrence of hydrophily makes
restricted
possible to gain an overall understanding of evolution, an
it
its
36
—
Hy
Philbrick drophily 3 7
1991]
understanding which perhaps unattainable in virtually other,
all
is
much more
widespread pollination systems.
Two
sometimes forms of hy-
but rather arbitrary,
general,
epihydrophily; pollination via pollen
drophily are recognized:
1)
and hypohy-
transport at the water surface (two-dimensional), 2)
below water
drophily; pollination via pollen transport the surface
dimensional)
may some
operate
in
combined
are in this
most works on pollination biology, the two forms
anem
media
similar systems that operate in different
anem
between
similarities
systems
evolutionary patterns
have
Whitehead, 1969, 1986; Regal, 1983) ad-
anemo
been
graphic context, but the evolution of hydrophily has largely
may
be informative
it
hope
systems with the
I
them and
place these
to
lutionary
Many
concerning
questions
basic
remain
manne
angiosperms
whether was a prerequisite for, or a consequence of, the invasion
it
Why
hydrophily
of the marine environment (Philbrick, 1988). is
What
common
most monocots, and virtually absent in dicots?
in
might discrepancy suggest about the evolutionary histories
this
monocots Are monocots, or certain groups
of the versus dicots?
somehow
monocots,
hydrophily?
may
among
pattern
similarities
The
purpose
mophily.
taxonomic
le
summarized.
of hydrophily will be con-
ondly, the ecogeographic distribution
anem
Rhodora
38
[Vol. 93
Table
Les
1.
The
number
genera are followed by the oT species/type of hydrophily: epihy-
E,
H, Taxonomy
drophily; hypohydrophily. of monocots follows Dahlgren and Ras-
mussen
(1983).
Dicot:
Nymphaeales
Ceratophyllaceae—
cosmopolitan.
Ceratophyllum 6/H
Monocot:
Hydrocharitales
—
Hydrocharitaceae warm
cosmopolitan, mainly
regions
Appertiella
1/E
Elodea
5/E
Enhalus
1/H
Halophila 8/H
Lagarosiphon
9/E
Nechamandra
1/E
Thalassia 2/H
Vallisneria 2/E
Zosterales
Najadaceae—
cosmopolitan.
Najas
30-50/H
—
Posidoniaceae Mediterranean, W.
Asia, Australia.
S.
Posidonia 3/H
Potamogetonaceae—
Ruppia) temperate and
(re: subtropical regions
Ruppia
1-7/E
Zosteraceae—
temperate coasts, excluding South America and
S.W.
Africa,
Heterozostera 1/H
Phyllospadix 5/H
Zostera 12/H
Zannichelliaceae—
cosmopolitan.
Alikenia 2/E
Lepilaena
4/E
Zannichellia 1-5/H
Cymodoceaceae—
and
tropical subtropical
Amp
bo 2/H
hi
lis
Cymodocea
4/H
Halodule 6/H
Syringodium 2/H
Thalassodcndron 2/H
may
or biological factors that influence distribution. Through
its
a consideration of these two general topics seek to focus attention
I
on some
of the important issues that remain to be adequately
addressed
regarding the evolution of hydrophily.
Hydrophily 39
Philbrick—
1991]
TAXONOMIC AND PHYLOGENETIC CONSIDERATIONS
taxonomic
monocots
which hydrophily has been documented. Al-
family
dicot for
hydroph-
containing
though the Callitrichaceae often cited as
is
evidence suggests otherwise (Philbrick
the available
ilous species,
hydrophily consid-
and Anderson, unpubl. data). In contrast, is
where seven
more widespread monocots, occurs in
erably in it
and 134 (.0026% of monocots).
(Table species If
families ca.
1)
among
angiosperms
we of hydrophily aquatic
consider the extent
monocots
of
ilous (Philbrick, 1990).
been hydrophily has evolved several
has long recognized that
It
taxonomic
times. Yet, the
some
taxonomic
autogamy have been misinterpreted as hy-
manifestations of that
In
being Philbrick, 1984, 1988).
drophily, clarified
is still (cf.
among
where hydrophily occurs tentative.
is still
mo
among monocots
phylogenetic history
Ceratophyllaceae.
the Its
much more complex. The seven monocot families in which
is
and Zos-
two Hydrocharitales
hydrophily occurs are in orders:
much
There has been speculation concerning
(Table
terales
1).
among study
relationships the families in these orders, yet is
masks
plagued by the extreme modification of floral structure that
A
and Rasmussen
by Dahlgren
homology. recent cladistic analysis
of
provided perhaps the best phylogenetic hypothesis re-
(1983)
among
the families
from
most
recent ancestor, the
aerial-flowered
mussen phylogeny suggested
(Figure th;
1)
minimum monocots
a of four times in
times
from
systems
may
synapomorph
be
Pota-
Figure with a subsequent reversal to aerial pollination in
1
,
mogeton (Potamogetonaceae) (Figure This latter "reversal-
If)-
40 Rhodora
93
[Vol.
Zosterales
Hydrocharitales
1
f r
o
Q
Z
CO
CO
< < O o 1 O
O
O N N o. n r 1
Q_ DC
Q. 05 e
h
C
f
D
B
A
E
A
Figure cladogram modified from Dahlgren and Rasmussen The
1. (1983).
synapomorphies
that support each node are not included but are listed in Dahlgren
and Rasmusscn's
Figure 359. Families which hydrophily
10, p. in occurs are in
upper case those which
letters; in aerial pollination occurs are in lower case
letters.
The
arbitrary division of the Hydrocharitaceae into hydrophilous and ento-
(h)
mophilous Upper
taxa mine.
(e) is case letters adjacent to arrows designate where
may
hydrophily have
lower
arisen; case adjacent arrows
letters to designate re-
from
versals hydrophily to aerial pollination systems. NAJ., Najadaccae; CYM.,
Cymodoceaceae; ZAN.,
Zanichelliaceae; ZOS., Zosteraccae; POS., Posidoniaceae;
R., Ruppia\ Potamogeton; jun., Juncaginaceae; Scheuchzeriaccae;
p., sch., apo.,
Aponogetonaceae; but., Butomaceae; HYD., Hydrocharitaceae; POT, Potamo-
getonaceae.
scenario" requires one less step than does the former. However,
the reversal-scenario requires the evolution of
aerial pollination
from
a hydrophilous precursor, which contrary
is to the tradi-
tional belief that aerial pollination systems are primitive. Yet,
in
no
principle there reason from
to refrain reversing
is this polarity.
Could some
aerial floral systems in groups be derived from hy-
drophilous ones? At the species level there no a priori reason
is
from
may
natives, with the hope of gaining a new perspective on the prob-
lem. noteworthv that the resolution invnlvp^; the fpwpQt
It is th;?t
Philbrick— Hydrophily
41
1991]
Q
O
a.
o CO CO
a o < o <
o
o N
1
CL
jQ 03 cr Q.
h
E g
D
B
F
I
H c
J
A
An
Figure 2. intuitive phylogeny of Robert Thorne (pers. comm., 1990)
il-
proposed between same shown
lustrating the relationships the families in Figure
1, Families in which hydrophily occurs are in upper case letters, those in which
hydrophily does not occur are in lower case letters. For the sake of illustration
the Hydrocharitaceae are arbitrarily divided into taxa that exhibit hydrophily (h)
and those that are entomophilous Upper case letters adjacent to arrows indicate
(e).
may
where hydrophily have originated; lower case letters represent reversals from
hydrophily to aerial pollination systems, apo., Aponogctonaceae; sch., Scheuchze-
Juncaginaceae; POT., Potamogetonaceac; Ruppia; Potamoge-
riaceae; jun., R.,
p.,
ton\ POS., Posidoniaceae; CYM., Cymodoceaceae; ZAN., Zanichelliaccae; ZOS.,
HYD.,
Zosteraceae; NAJ., Najadaceae; Hydrocharitaceae; Butomaceae.
but,
from
Steps (the reversal-scenario) includes a single reversal hy-
drophily
to aerial pollination.
may
It be useful here to compare the cladogram from Dahlgren
and Rasmussen
with another phylogenetic pattern that has been
proposed
for the Zosterales, Figure 2 the intuitive phylogenetic
is
Thome
Thome
of Robert comm.).
tree (pers. Several features of the
phylogenetic tree contrast with that of Dahlgren and Rasmussen,
Thome's
e.g., placement of the Najadaceae relative to the Zos-
Using same Thorne
teraceae. the reasoning as above, the phy-
logenetic tree suggests that hydrophily has evolved
at least five
monocots
times 2A, D,
in (Figures B, C, E). If reversals to aerial
most
pollination systems are included, the parsimonious
reso-
would
lution require five steps. For instance, hydrophily
still if
A
and
arose at B, to account for occurrence in the Najadaceae
its
Rhodora
42
[Vol. 93
minimum
and and of two
Hydrocharitaceae, respectively, at a
J,
reversals would be necessary. Thus, the Dahlgren and Ras-
i)
(g,
more
mussen topology (Figure provides a parsimonious (fewer
1)
resolution of the distribution of hydrophily than does the
steps)
Thorne
phylogenetic of (Figure
tree
2).
The complexity of interpretation of structure throughout
floral
the Zosterales has been an obstacle in establishing confidence in
and
phylogenetic hypotheses for this order (Dahlgren ClilTord,
Dahlgren and Rasmussen, Posluszny and Tomlinson,
982; 1983;
1
1977; Tomlinson, 1982). This difficulty has been attributed to
accompany
modification of reproductive structures that the evo-
lution of hydrophily. However, the presence of non-hydrophilous
Taxa
such
taxa in this order also leads to phylogenetic puzzles.
Lilaea and Triglochin (Juncaginaceae) and Scheuchzeria
as
(Scheuchzeriaceae) display floral features that are difficult to assess
homologous with those other monocot groups (Tomlinson,
as in
1982). Thus, appears that there are factors other than hydrophily
it
make
that interpretation of phylogenetic relationship
itself diffi-
A
some
cult in the Zosterales. reassessment of of the basic ideas
we
have concerning system evolution, such the evolu-
as
floral
may
tionary polarity of aerial versus hydrophilous pollination,
new
provide
a perspective.
important note that neither the Dahlgren and Rasmussen
to
It is
Thome
(Figure nor (Figure phylogenies provide resolution
1) 2)
among
for relationships taxa in the Hydrocharitaceae (nor are
meant
they Yet, this family deserves special consideration
to).
The
here, given diversity of pollination types. Hydrocharita-
its
ceae the only family in which entomophily, epihydrophily and
is
hypohydrophily occur Cook, Unfortunately,
(cf 1982).
all little
work
phylogenetic has been carried out in the Hydrocharitaceae
D. K. Cook, comm.) and unclear whether hydrophily
(C. pcrs.
it is
mono-
Kaul
or polyphyletic in the family. 968, 970) included
is (1
1
a diagram of floral evolution derived from his studies of floral
He
showed
and inflorescence anatomy. that modifications of
floral
may
and accommodate hydrophily
inflorescence structures that
have occurred via several developmental pathways. This evidence
A
suggests that hydrophily polyphyletic within the family. phy-
is
logenetic framework against which to assess the evolution of pol-
lination types would allow us to address whether one form of
hydrophily has led to another epihydrophily to hypo-
(e.g.,
whether have
hydrophily) or they arisen independently.
Philbrick- Hydrophily 43
1991]
TAXONOMIC, GEOGRAPHIC, LIMNOLOGICAL AND
BIOLOGICAL CONSIDERATIONS
Anemophily
taxonomic
versus hydrophily: distribution
The
differences in taxonomic distribution of anemophily and
Anemophily
hydrophily taxonomically widespread
are
striking.
is
and occurs in virtually every major angiosperm group. Ane-
mophilous groups Quercus
are often species-rich; for instance,
(Fagaceae) encompasses 600 Salix 400
ca. species, (Salicaceae)
ca.
and Poa 500
species, (Poaceae) ca. species. In addition, the Po-
aceae, a largely anemophilous family, has perhaps 6000
species.
Thus, reasonable to say that speciation in these groups does
it is
not seem have been hindered by an anemophilous
to pollination
system.
Hydrophily, angiosperm fam-
in contrast, restricted to eight
is
Groups
ilies (Table that include hydrophily are generally spe-
1).
cies-poor Najas with 50
(Les, 1988); in fact, its ca. species is the
most The
species-rich hydrophilous genus. next hydroph-
largest
The
ilous genus Zostera which has only 12 species. remaining
is
Whether
genera contain 10 or fewer hydrophily
species.
it is itself
or the associated attributes of in the aquatic milieu which
life
limit speciation an issue which has been addressed to some
is
degree by Les although remains unresolved.
(1988), largely
it
Geographic
considerations
Although there are relatively few hydrophilous species, hy-
drophily itself as geographically widespread as anemophily, yet
is
two show
the abiotic pollination types different geographic pat-
Anemophily
terns. increases in incidence with increasing latitude
and/or altitude (Whitehead, 1969; Regal, 1983; Berry and Calvo,
mode
Anemophily
primary of
1989). the pollination across
is
large geographic areas, northern boreal forests and grasslands.
e.g.,
uncommon
However, anemophily
relatively in tropical regions.
is
common
In contrast, hydrophilous species seem to be equally in
make
temperate and tropical regions, but up a small component
of the flora throughout. All eight families in which hydrophily
and
occurs are geographically widespread, five of the eight families
many
exhibit broad geographic ranges across temperate and trop-
regions (Table
ical
1).
The num-
broad distribution of hydrophily by the
illustrated
is
Rhodora
44
93
[Vol.
Table Summary of the total number of species (A-C) and gcnera(D) that
2.
occur per 10 degrees of latitude. The taxa in each latitudinal increment are not
necessarily mutually exclusive. A. The number of species ofNajas in Central and
North America (modified from Haynes, 1979). B. The number of species ofNajas
in the Neotropics (modified from Lowden, 1986). C. The number of species of
Najas Malaysia (modified from dc Wilde, 1972). D. The number of genera of
in
seagrasses worldwide (modified from den Hartog, 1970).
NaJas Seagrasses
Degrees
.
D
A C
B
Latitude 4-6
4-8
70
2
60
2
2
-53
50
-578
4
5
40
-3-8
7
30
5
20 -2-8
-4-11
4
8
5 5
ION
8
lOS
20
30
40
- - -
2
50
ber of representative taxa that occur in temperate and tropical
Lowden
Haynes and
For pro-
latitudes. instance, (1979) (1986)
of Najas throughout Central
vided distribution data for species
The num-
and North America, and the Neotropics, respectively.
bers of species of Najas per 10 degrees of latitude (Tables 2A, B)
The
shows geographic uniformity. similar
a relative overall in-
cidence of najad species in tropical and temperate regions con-
markedly with what one would predict the distribution
trasts if
of anemophily were used as a model: increasing incidence with
increasing The distribution of Najas in Malaysia (de
latitude.
Wilde, 1972) also illustrates the extent of this genus in tropical
The of hy-
regions (Table 2C). temperate-tropical distribution
by worldwide of
drophily further illustrated the distribution
is
number
of genera
the genera of sea-grasses (Table 2D). In fact, the
somewhat than
per 10 degrees of latitude higher in tropical
is
temperate
regions.
The nature of the comparisons that are being attempted here
com-
For meaningful
lead to several questions. instance, this a
is
parison given the vastly different sample sizes? That there are
is,
Philbrick— Hydrophily 45
1991]
only 140 species of hydrophilous angiosperms, compared with
the thousands of anemophilous taxa from which
to assess distri-
No
bution patterns. doubt our perception of the distribution of
hydrophily colored by the distribution of relatively few species.
is
However, problem
this inherent in the nature of the pattern
is
being discussed, that comparing a ubiquitous to a rare polli-
is,
nation system. Secondly, although anemophily infrequent
in
is
how many
tropical regions relative to biotic pollination, tropical
taxa are in fact anemophilous? Could be that on a taxon-by-
it
taxon anemophily more
basis as frequent, or frequent, in trop-
is
ical regions as hydrophily? Although, this frequency comparison
may may
or not be the significance of the comparison being
true,
made
in the incidence of the pollination system throughout
lies
broad
geographic areas, not in proportion relative to other
its
pollination systems within a particular
flora.
Limnological
factors
Our
understanding of hydrophily might from
benefit a discus-
sion of the factors that limit distribution, an approach taken
its
regarding anemophily (Whitehead, 1969; Regal, 1983; Dauben-
The
mire, 1972). ecogeographic distribution of hydrophily no
is
doubt a result of a complex array of interactions of limnological
and biological factors. Although we run the risk of over-simpli-
may
fication, be useful nonetheless to attempt to construct
it
generalized patterns.
Aquatic habitats are considered uniform regardless of
fairly
den
latitude (Arber, 1920; Sculthorpe, 1967; Hartog, 1970;
Tiff-
ney, 1981; Les, 1988). Although differences in water chemistry,
growth season duration and temperature do aquatic plant
affect
communities, the aquatic habitat seems be more uniform than
to
due thermal of
terrestrial habitats, in part to the properties water.
Environmental cues that initiate flowering in hydrophilous
may
The
species be insensitive to latitudinal change. seagrasses
(Cymodoceaceae, Hydrocharitaceae
[Enhalus, Thalassia], Posi-
doniaceae, Zosteraceae) provide a good example, where tidal cy-
have been
cles implicated in stimulation of flowering
Pettitt,
(cf.
1984). Although the amplitude of the tidal cycle varies with lat-
may
itude, the cycle ubiquitous. Similar ubiquitous cues
itself
is
also be instrumental in the flowering of freshwater hydrophilous
However,
taxa. pollination systems in freshwater hydrophilous