Dioecism in Elingamita johnsonii (Myrsinaceae)

Abstract

Examination of the flowers of Elingamita johnsonii confirms that the species is functionally dioecious. Female flowers have a well‐developed ovary, capitate stigma, and anthers which produce pollen with occluded apertures. Male flowers produce abundant aperturate pollen and have a rudimentary ovary that lacks a normally developed stigma and placenta. The sizes of the floral parts differ in each sex, with the female having smaller sepals, petals, and anthers, and shorter filaments. Male plants have lager inflorescences, more flowers per inflorescence, and more pollen grains per anther.

Full text

New Zealand Journal
of
Botany,
2000, Vol.
38:
569-574
0028-825W00/3804-0569 $7.00
0
The Royal Society of New Zealand 2000
5
69
Dioecism
in
Elingamita johnsonii
(Myrsinaceae)
P. B. HEENAN
Landcare Research
P.
0.
Box 69
Lincoln 8 152, New Zealand
different
reports
of the floral biology
of
E.
johnsonii
prompted the present study, the aim of which is to
provide empirical data on the breeding system of the
species.
Abstract
Examination of the flowers of
MATERIALS
AND
METHODS
Elingamita johnsonii confirms that the species is
functionally dioecious. Female flowers have a well- Flowering plants of E. johnsonii in cultivation at
developed ovary, capitate stigma, and anthers which Landcare Research sites at Lincoln (near
produce pollen with occluded apertures. Male flow- Christchurch
south
Island) and Mt Albert (Auck-
ers produce abundant aperturate pollen and have a land,
North
Island) and at several gardens
in
Awk-
rudimentary ovary that lacks a normally developed land city, and herbarium material at
AK
and
CHR,
stigma and placenta. The sizes of the floral parts were available for Study. The sample included Plants
differ in each sex, with the female having smaller grown from seeds collected on the Three Kings
1s-
sepals, petals, and anthers, and shorter filaments. hds and other Plants grown from seeds collected
Male plants have lager inflorescences, more flow- from mainland-cultivated plants. General floral
ers per inflorescence, and more pollen grains per morphology was examined using a binocular micro-
anther. scope, and floral measurements were made with
Mitutoyo digital calipers. Pollen stainability was
Keywords
Myrsinaceae; Elingamita; Elingamita asessed
bY
staining the cytoplasm With Alexander's
johnsonii dioecious; Three
Kings
Islands
flora
N~~
differential stain (Alexander 1969). Examination of
Zealand flora stained pollen was made with a Zeiss light micro-
scope. The number of pollen grains per anther were
counted by subsampling with a haemocytometer.
Pollen grains from herbarium specimens were exam-
ined with a Leica
S440
scanning electron microscope
the Three Kings Islands (,-.
58
km
NW
of
cape
Hand
pollinations Were conducted on
two
female
Reinga, New Zealand, 3409's, 17208'E) and
was
plants at Lincoln to aSSeSS the SUCCeSS Of Self- and
described from a small amount
of plant
material
cross-pollination. Six inflorescences on
two
female
collected in January 195
1
(Baylis 195
1).
In the Plants at Mt Albert were also bagged
to
assess self-
original description the species was considered to Pollination.
have hermaphrodite flowers, but this has Analysis of variance (ANOVA) was conducted
subsequently been emended to dioecious (Scott
On
inflorescence
length,
flower
number
Per
198 1; Baylis 1995). Salmon (1982) indicated that inflorescence, and pollen grains per anther for male
cultivated plants have either male flowers only or
and
female
plants.
both hermaphrodite and female flowers. These
INTRODUCTION
The monotypic Elingamita johnsonii is endemic to Using
an
Oxford CTl500 cold stage at
-
180°C.
RESULTS
The male inflorescence of E. johnsonii is larger and
has more flowers per inflorescence than those of
female plants. The male flowers are characterised
by
BO0009
Received
19
April
2000; accepted
27hfy 2000

5
70
long filaments that arch outward, strongly dehiscent
anthers, a high number of pollen grains per anther,
aperturate pollen, and a poorly developed placenta
with shrivelled ovules, and they lack a prominent and
well-developed stigma (Table 1; Fig. 1,
3,
4). In
comparison, the female flowers have a large and
well-developed ovary with a prominent capitate
stigma, a well-developed placenta, and short
filaments, the anthers appear normally developed but
are not always dehiscent, the pollen is inaperturate,
and there are fewer pollen grains per anther (Table
1
;
Fig. 1,2,4).
Elingamita johnsonii
clearly has
two
different forms of flower and is considered to be
functionally dioecious.
The differences in inflorescence length, the
number of flowers per inflorescence, and the number
of pollen
grains
per anther for male and female plants
are significant at the 99.9% confidence level (Table
1).
The male to female sex ratio of the
Elingamita
samples is close to
1
:
1
(
12 male
:
14 female).
Pollen grains stained with Alexander’s differen-
tial stain were between 76.0% and 98.2% stainable
in the male plants, and (4.8-)71.9-98.9% stainable
in the female plants (Table
1).
The pollen grains of
male plants have three apertures (Moar 1993), and
those of female plants have occluded apertures (Fig.
2, 3). Sometimes, when pollen grains from female
plants are examined under a light microscope, they
appear
to
be aperturate. Examination by SEM has
shown this to be due to slightly collapsed walls (Fig.
2).
The two abundantly fruiting female plants at Mt
Albert had six inflorescences with 164 well-
developed flowers bagged to exclude foriegn pollen
and any pollinators, and each of these flowers failed
to produce fruit (Table 2). In contrast, three control
and open-pollinated female inflorescences with 32
flowers produced
1
I
developing
hit
and, as there
were flowering male plants in the vicinity,
it
is
most
likely
that
these males acted
as
the pollen source. The
pollen vector
is
not
known.
Hand female-to-female
pollinations on
two
cultivated plants at Lincoln also
failed
to
produce any fruit (Table 2). After hand
pollination the ovaries of self- and cross-pollinated
female flowers did not noticeably swell, and the
flowers abscised after about 34 weeks, abscising at
the same time
as
control and unbagged flowers that
were not pollinated.
Elingamita johnsonii
male
flowers
(Fig. 1,3,4)
Inflorescences 40-77 mm long, with
1
1&2 13
flowers; rachis and pedicels with extrafloral nectaries
(Heenan 1999). Pedicels 2.34.2
x
c. 1.2
mm.
Sepals
New Zealand Journal of Botany, 2000, Vol. 38

Heenan-Dioecism
in
Elingamita
57
1
Fig.
1
Flowers of
E.
johnsonii.
Male
(a,
b,
c)
and female
(d,
e,
f)
flowers. a, d: flower habit;
b,
e:
androecium; c,
f
gynoecium. Scale
bars
=
1
mm.
Fig.
2,
3
Pollen grains of
E.
johnsonii.
Fig.
2
(left) Pollen from
female plants with occluded aper-
tures.
Fig.
3
(right) Pollen from
male plants with normally devel-
oped apertures.
Table
2
Specimen
Pollination
type
Date of pollination
Number
of
flowers Flower
abscission Fruit formed
Lincoln
I
self
25
Mar
1998
45
c.
3 weeks
0
Lincoln
I
self
5
Mar
1999
70
c.
4
weeks
0
Lincoln
I
cross (Lincoln
2)
5
Mar
I999
38
c.
4
weeks
0
Lincoln
2
self
5
Mar
1999
50
c.
4
weeks
0
Lincoln
2
cross (Lincoln
1
)
5
Mar
19%
84
c.
4
weeks
0
Lincoln
2
control (unbagged)
5
Mar
1999
105
c.
4
weeks
0
Mt Albert
1
self (bagged)
9
Mar
1998
I05
c.
3 weeks
0
Mt Albert
2
self (bagged)
9
Mar
1998
59
c.
3 weeks
0
Pollinations of female plants
of
€.johnsonii.
-
II
Mt
Albert
1
control (unbagged)
9
Mar
1998
32

572 New Zealand Journal of Botany, 2000, Vol. 38
Fig.
4
of
E.
johnsonii.
Relationship
of
flower number
per
inflorescence and inflorescence length
for
male
(0)
and female
(A)
plants
2.2-2.6
x
c. 1.2 mm, usually flushed red when
mature, triangular, glandular; apex acute to subacute.
Petals 2.1-2.2
mm
long, connate. Anthers 1 &1.7
x
c. 1.2 mm. Filaments 3.0-3.2 mm long. Ovary
c.
!
.C;
x
c.
0.75
mm;
placenta
c.
0
2
x
c
0
!
5
mm,
malformed. Style c. 1.2 mm long. Stigma absent.
Pollen with 3 apertures.
VOUCHERS:
Three
Kings 1s1ands,A.
E.
wright
52687
Hunter,
Feb
19657
AK
105825; cu1tivated,
Mt
A1bert,
A.
E.
Esler,
4
Feb
19709
CHR
205237';
356450; cultivated, Mt Albert, Auckland,
A.
E.
Esler,
2
I
Feb 1978, AK 2 19494B; cultivated, Mangonui
Co.,
R. H. L.
Ferguson,
24 Jan 1983, AK 167232;
cultivated, Lincoln,
p.
B.
Heenan,
26 Jan 1998, CHR
514783; cultivated, Auckland University,
p.
Heenan,
9 Mar 1999, CHR
5
14877-5 14880;
cultivated, Grey Lynn, Auckland,
P.
Heenan,
9 Mar
1999, CHR
5
1488 1; cultivated, Mount Albert
Research Centre, Auckland,
P.
Heenan,
9
Mar 1999,
CHR 514873; cultivated, Mount Albert Research
Centre
Auckland,
P.
Heenan;
9
Mar
1999;
CHR
514870.
ILLUSTRATIONS:
Baylis (1951, fig. 1; 1995, fig. 1);
Salmon (1982, fig. 4,5). Baylis (1995) reproduced
suggested that these are female (fig. la) and male
(fig. lb, c). The illustration of the putative female
stigma, and is almost certainly a
male
flower
prior
to
anthesis,
NOTES:
Each branch
of
the inflorescence terminates
in 1-5 aborted buds. A 1-m-high male plant in
cultivation at Lincoln flowered prolifically with
terminal inflorescences in late January and early
February 1998, and then in early March
it
produced
about 25 inflorescences from the axils of the leaves
17 Dec 1982, AK 162641; cultivated, Mt Albert,
J.
his 1951 figure of the flowers
Of E.
johnsonii,
and
cultivated, Mt Albert,
A.
E.
Esler,
2
1
Feb 1978, CHR lacks a swollen ovary and normally developed
*Voucher specimen
for
the pollen grains described by
Moar
(
1993).
(CHR 514933).

Heenan-Dioecism
in
Elingamita
573
Elingamita johnsonii
female flowers (Fig. 1,2,4)
Inflorescences 1546 mm long, with 9-72 flowers;
rachis and pedicels with extrafloral nectaries
(Heenan 1999). Pedicels 1.4-3.3
x
1.6-2.1 mm.
Sepals 1.1-2.8
x
1.0-2.2 mm, usually green,
triangular, glandular; apex acute to subacute. Petals
1.0-1.8
mm
long, connate. Anthers 1.0-1.4
x
1.0-
1.3 mm, Filaments
up
to
0.2
mm
long.
ovary
1.7-
2.6
x
1.5-2.3 mm; placenta 0.9-1.2
x
0.8-1.2 mm,
ovoid. Style 0.75-0.95 mm long. Stigma capitate,
prominent. Anthers appear normally developed, but
usually indehiscent. Pollen with occluded apertures,
cytoplasm usually present.
VoUCHERS:
Cultivated
Mt
A1bert3
Auckland'
A.
E'
pollinators
as
the flowers lack floral nectar, although
they do have extrafloral nectaries (Heenan 1999).
The
female
flowers
maY therefore mimic hemaPh-
rodite
flowers
(Gilbert 1975).
A
number of the floral characteristics of
Elingamita
also Occur in many anemophilous sPe-
cies. Characters of wind-pollinated species include
terminal inflorescences that
are
held well beyond the
leaves, unisexual flowers,
an
insignificant perianth,
a
low
Ovule number,
the
anthers
of
male flowers
being exserted well beyond the corolla tube, and the
Pollen
dry
and abundant (Bullock 1994).
Elingamita
also lacks some attributes of wind-pollinated species,
such as
an
exserted stigma. Flower morphology
is
clearly not an adequate basis to confidently propose
Esler
21
Feb
1978?
AK
219494A; cultivated,
Lin-
coln,
"
B.
Heenan,
26
Jan
1998,
CHR
5
14784;
cul-
the pollen
vector
type,
and field
studies
should be
undetaken to determine the mode of
pollen
disper-
tivated'
Mount
Albert
Research
Centre'
Auckland'
sal.
Furthermore, flower morphology should not be
p.
Heenan,
9
Mar
1999,
CHR
5
14867; cultivated,
Mount
Albert
Research
Centre,
Auckland,
p.
used to determine the type of
gender
dimorphism
as
many
genera
may
be critically dioecious (Bawa
&
Heenan,
9 Mar 1999, CHR
5
14869; cultivated,
Opler
1975).
The
gender dimorphism of E.
johnsonii
Mount
Albert Research
Centre?
Auckland,
p.
Heenan,
9
Mar
1999,
CHR
5
14872; cultivated,
Mount
Albert Research
Centre,
Auckland,
p.
is cryptic,
as
based
on
qualitative
flower
morphol-
ogy
it could be interpreted
as
androdioecious,
a
par-
ticularly
rare
breeding system (Charlesworth 1984;
Heenan,
9 Mar 1999, CHR 514871; cultivated, Grey
Lynn, Auckland,
P.
Heenan,
9 Mar 1999, CHR
5
14882; cultivated, Government House, Mount
Eden, Auckland,
P.
Heenan,
9 Mar 1999, CHR
5
14876.
ILLUSTRATIONS:
Salmon (1982, fig. 2, 3); Hunter
(1964, p. 195).
NOTES:
Most female inflorescences have a
ring
of
5 -
7
aborted
buds
at
the terminal
pm
Of
each
branch
Of
the
inflorescence.
The
Ovary
Of
the
female
flower
sometimes Pushes the corolla tube and anthers off
early
in
the development of the flower (Fig. Id), and
if this happens the flower appears to be
morpho-
logically strctly
female.
this
may
account
for
the
record of some plants having both hermaphrodite
and female flowers (Salmon 1982).
Webb
1999).
ACKNOWLEDGMENTS
I
thank
Ross
Beever for providing plant material, assist-
ante
at Mt Albert, and comments on the draft manuscript;
Colin Webb for discussion and comments on the draft
manuscript; Peter de Lange for assistance with collect-
ing plant material; Matt McGlone for discussion on
pol-
len; Tim Galloway for Fig.
l;
Neil Andrews
for
assistance
with Fig.
2
and
3;
and Brent Maxwell, Tony Palmer,
Helen Preston-Jones, Peter de Lange, and
J.
E.
&
1.
J.
Bellingham for plant material. Funds for the research
were provided by the Foundation for Research, Science
and
Technology
(contract co9405),
DISCUSSION
REFERENCES
The flowers of
Elingamita
are typical of many New
Zealand genera in
being
inconspicuous, of small
size
Alexander,
M.
P. 1969: Differential staining of aborted and
and simple structure, and with white or pale colour-
nonaborted pollen.
Stain
Technology
44:
1
17-1
22.
ing (Godley 1979; Lloyd 1985; Webb
C%
Kelly
Bawa,
K.
s.;
Opler P.
A.
1975: Dioecism in tropical
1993). Such flowers are usually suited to
forest trees.
Evolution 219:
167-179.
unspecialised
insect
pollinators, and hence
Baylis,
G.
T.
S.
1951:
Elingamita
(Myrsinaceae), a new
Elingamita
is
most likely to be entomophilous
monotypic genus from West Island, Three Kings.
(Webb et al. 1999). The presence of pollen-bearing
Records of
the
Auckland Institute and Museum
4:
anthers
in
the female flower could be a reward for
99-102.

5
74
Baylis,
G.
T.
S.
1995:
Sex and
Elingamita:
a cautionary
tale.
New Zealand Botanical Society Newsletter
41:
20-21.
Bullock,
S.
H.
1994:
Wind pollination
of
neotropical
dioecious
trees.
Biotropica 26:
172-1 79.
Charlesworth,
D.
1984:
Androdioecy and the evolution
of
dioecy,
Biological Journal
of
the
Linnean
so-
ciety 22:
333-348.
Gilbert,
L.
E.
1975:
Ecological consequences
of
a
coevolved mutualism between butterflies and
plants.
In:
Gilbert, L. E,; Raven, p.
H.
ed,
Coevolution
of
animals and plants. Austin, Uni-
versity
of
Texas Press. Pp.
2 10-240.
Godley, E.
J.
1979:
Flower biology in New Zealand.
New
Zealand Journal of Botany
I
7
44 1466.
Heenan, P.
B.
1999:
Extrafloral nectaries in
Elingamita
johnsonii
(Myrsincaeae).
New Zealand Journal
of
Botany 37
183-185.
Hunter,
J.
A.
1964:
Recently discovered native tree
has
horticultural merit.
New Zealand Gardener 21:
1
94-
I
95. 130.
New Zealand Journal
of
Botany,
2000,
Vol.
38
Lloyd,
D.
G.
1985:
Progress
in
understanding the natural
history
of
New Zealand plants.
New Zealand
Jour-
nal of Botany 23:
707-722.
Moa, N.
T.
1993:
Pollen grains
of
New Zealand
dicotyledonous plants. Lincoln, Manaaki Whenua
Press.
Salmon,
J.
T.
1982:
The native
trees
of
New Zealand.
Wellington, A.
H.
&
A. W. Reed.
Scott, L.
198
1
:
Elingamita. Botany Division Newsletter
Webb, C.
J.
1999:
Empirical studies: evolution and main-
tenance
of
dimorphic breeding systems.
In:
Geber,
M. A.; Dawson,
T.
E.;
Delph, L.
F.
ed.
Gender
and sexual dimorphism in flowering plants. Ber-
lin, Springer-Verlag.
4.61-95.
Webb, C.
J.;
Kelly, D.
1993:
The reproductive biology
of
the New Zealand
flora.
Trends in Ecologv and
Evolution 12:
142-147.
Webb, C.
J.;
Lloyd,
D.
G.;
Delph,
L.
F.
1999:
Gender
dimorphism in indigenous New Zealand
seed
plants.
New Zealand Journal of Botany
37
I
19-
67:
7-8.