Directory UMM :Data Elmu:jurnal:T:Tree Physiology:Vol15.1995:
Tree Physiology 15, 477--484
© 1995 Heron Publishing----Victoria, Canada
Constraints to seed production: temperate and tropical forest trees
J. N. OWENS
Graduate Center for Forest Biology, University of Victoria, Victoria, British Columbia V8W 2Y2, Canada
Received May 31, 1994
Summary Reforestation requires a constant supply of high
quality seed. Different methods have been used to measure seed
and fruit production as a proportion of that which potentially
could be produced. These estimates coupled with developmental studies permit determination of when and to what extent
different biological constraints reduce seed and fruit production. These vary among species, years and sites. Overcoming
biological constraints may be possible in seed orchards and
seed production areas. Biological constraints include: (1) periodic or inadequate floral initiation, (2) asynchronous development and flowering, (3) floral abortion, (4) ovule abortion,
(5) embryo abortion, and (6) failure of seeds and fruits to
mature and our inability to determine maturity. Solutions to
these constraints are varied but all must begin with an understanding of reproductive biology of each species.
Keywords: embryo, floral initiation, flowering, fruit production, maturation, ovule, reforestation.
Introduction
Reforestation requires a constant supply of high quality tree
seed or vegetative propagules. However, seed is often in short
supply, as a result of much seed being of low quality with
variable maturity and a limited storage life. Seed production is
variable among species and regions, but the causes of these
variations are often uncertain. Even in seed orchards many
constraints exist and mating is usually not random. The seed
produced may represent only a portion of the genetic diversity
intended by the phenotypic selection and orchard design (ElKassaby 1992).
Seed production research must first identify species where
there are biological constraints to seed or fruit production and
then determine the biological constraints for each species in a
particular location. Most seed production research has been
done on temperate conifers (Owens and Blake 1985, Owens
1991). We know little about the reproductive biology and seed
production of tropical species, particularly the tropical hardwoods about which there is now increased interest.
Even the basic phenologies of the reproductive cycles are
poorly understood for tropical conifers and hardwoods. Most
temperate and many tropical forest trees exhibit indirect flowering in which there is a period of dormancy between floral
initiation and pollination. However, there are at least four other
groups of tropical flowering periodicities that are recognized
(Janzen 1978): (1) everflowering species, such as Ficus, that
initiate flowers throughout the year and, therefore flowers,
seeds and fruits at various stages of development are always
present; (2) nonseasonal-flowering species that exhibit flowering periodicity among plants and from branch to branch; (3)
gregarious-flowering species that initiate floral buds continuously, but these remain closed for weeks or months until
environmental conditions are favorable for anthesis, sometimes resulting in anthesis at the same time over a wide geographical area, as in Coffea; and (4) seasonal-flowering
species that flower in response to rainy or dry seasons or subtle
seasonal variations in day length or temperature, as has been
proposed for certain dipterocarps (Ashton et al. 1988).
Frequency of flowering (Table 1) (Owens et al. 1991a),
which is usually measured by frequency of seed or fruit production at the end of the reproductive cycle, is variable. Seed
production is measured as seed per hectare or per tree with
little concern for the amount of seed produced relative to the
potential seed that could be produced. More accurate, though
more laborious methods are available. For example, in conifers, seed efficiency (SEF), or seed to ovule ratio (S/O), measures the filled seed produced per cone as a percentage of the
seed potential (number of fertile ovules per cone) (Owens et al.
1990, 1991a). In flowering plants, reproductive success (RS),
which is a measure of the fruit to flower ratio (Fr/Fl) multiplied
by the S/O ratio, has been used (Weins et al. 1987). Both of
these approaches make it possible to rank species, clones or
individual trees with respect to RS and identify quickly major
constraints to seed or fruit production (Table 2).
However, use of SEF or RS does not give a complete picture
of the constraints or when they occur in the reproductive cycle.
Careful dissections or detailed anatomical techniques (Owens
et al. 1991b) are needed to determine the times and possible
causes of losses (Figure 1).
Identifying biological constraints
The reproductive phenology, from floral initiation through
pollination and seed or fruit maturity, of temperate conifers is
well known, but less is known about temperate hardwoods.
Biological constraints and solutions
Constraints and their relative importance vary among species,
sites and years. Constraints result from developmental irregu-
478
OWENS
Table 1. Bud types, age and frequency of flowering in selected hardwood and conifer trees.1
Family/
Botanical name
Aceraceae
Acer platanoides
Platanus occidentalis
Betulaceae
Alnus glutinosa
Betula pendula
Dipterocarpaceae
Dipterocarpus spp.
Fagaceae
Fagus sylvatica
Quercus petraea
Salicaceae
Populus tremuloides
Salix nigra
Oleaceae
Fraxinus excelsior
Verbenaceae
Tectona grandis
Cupressaceae
Chamaecyparis lawsoniana
Thuja plicata
Pinaceae
Abies grandis
Larix decidua
Picea abies
Pinus contorta
Pinus caribaea
Pseudotsuga menziesii
Tsuga heterophylla
1
Common name
Buds simple (S)
or mixed (M)
Buds terminal (T)
or axillary (A)
First good
crops (years)
Intervals between
seed crops (years)
Norway maple
American sycamore
M
M
T, A
A
25--30
25--30
1--3
1--3
Common alder
Silver birch
S
S
A
T, A
15--20
15
2--3
1--3
M
A
20--30
1--6
Common beech
Sessile oak
M
M
A
A
50--60
40--50
5--15
2--5
Quaking aspen
Black willow
S
S
A
A
50--702
25--752
4--5
1--3
Common ash
M
A
25--30
3--5
Teak
M
T
20--25
1--3
Lawson cypress
Western red cedar
S
S
T
T
20--25
20--25
4--5
2--3
Grand fir
European larch
Norway spruce
Lodgepole pine
Caribean pine
Douglas-fir
Western hemlock
S
S
S
M
M
S
S
A
T
T, A
A
A
A
A, T
40--45
25--30
30--35
15--20
10--15
30--35
25--30
3--5
3--5
3--5
2--3
1--3
5--7
3--4
Data compiled from various sources.
larities causing abortion, or abnormal or delayed development
of a reproductive structure. An anomaly may not become
manifest until long after its cause has occurred. The longer the
reproductive cycle, the more opportunity there is for climatic
and other factors to affect seed and fruit production adversely
(Owens 1991).
Failure to initiate floral structures
A common cause of low seed production is a failure to initiate
cones or flowers. Seasonal flowering trees do not flower frequently. They have a particular developmental stage and season at which floral initiation occurs that is mediated through
the interaction of developmental, physiological and environmental events, many of which are poorly understood. In trees
with distinct vegetative buds (Table 1), floral initiation usually
only occurs in some of these buds long before floral structures
are clearly visible. In many north-temperate conifers (Figure 2) and a few hardwoods, the time and method of floral
initiation have been determined by anatomical study (Owens
and Blake 1985). In most cases, floral initiation and early
development of floral structures occur before winter dormancy. In the few tropical forest trees that have been studied,
Pinus caribaea Morelet. (Harrison and Slee 1991), Eucalyptus
(Moncur and Boland 1989) and some dipterocarps (Ng 1981,
Ngamkhajorniway and Wasuwanich 1986, Ashton et al. 1988),
as in temperate conifers and hardwoods, floral initiation appears to be correlated with lateral shoot elongation and stages
of vegetative bud development. Such correlations are useful
because they allow us to predict the time of floral initiation
based on external morphology.
Floral enhancement is usually achieved by manipulating the
environmental factors known to promote flowering or by application of fertilizers or plant growth regulators (PGR). These
methods have been quite successful in north-temperate conifers (Owens and Blake 1985, Ross and Pharis 1985) but have
had little success in tropical hardwoods. Notable exceptions
are some fruit trees and Eucalyptus (Griffin et al. 1993, Moncur 1994), where flowering has been promoted by paclobutrazol. These flower enhancing methods may only be applicable
to potted trees, seed production areas and seed orchards where
trees are small and environmental conditions can be controlled.
Environmental factors known to affect flowering include
light intensity, temperature, water stress and nutrients. Photoperiod may influence flowering indirectly through its control
of shoot growth and bud development (Owens and Blake
1985), but it is unlikely to affect flowering in tropical forest
trees. Increasing light is a common practice in north-temperate
CONSTRAINTS TO SEED PRODUCTION
479
Table 2. Reproductive success (RS) within Acacia auriculiformis A. Cunn. ex Benth. × A. mangium hybrid.1
Spike
Flowers per spike
% Flower
survival
Fruits per
spike
Fruit per
flower
Ovules per
ovary
Seed per
pod
Seed per
ovule
RS2
1
2
3
1
3
4
2
2
5
1
8
6
2
2
1
2
6
3
10
4
0.01
0.02
0.03
0.01
0.03
0.04
0.02
0.02
0.05
0.01
0.08
0.05
0.02
0.03
0.01
0.02
0.05
0.04
0.11
0.04
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
12
14
5
7
8
8
7
12
11
8
8
6
14
16
13
7
4
5
10
9
0.75
0.87
0.31
0.43
0.50
0.50
0.44
0.75
0.69
0.50
0.50
0.35
0.35
0.31
0.81
0.44
0.25
0.31
0.62
0.56
0.0075
0.0174
0.0093
0.0043
0.0150
0.0200
0.0088
0.0150
0.0340
0.0045
0.0410
0.0175
0.0055
0.0083
0.0730
0.0080
0.0120
0.0110
0.0678
0.0224
0.034
16
0.58
0.0196
Prepollination
Postpollination
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
100
100
95
94
78
97
98
98
100
110
96
101
107
73
105
89
113
81
92
96
1
6
9
15
14
5
21
18
5
6
6
14
19
14
12
3
17
9
4
11
1
6
9
16
18
5
21
18
5
5
6
14
18
19
11
3
15
11
4
11
Mean
96.7
10.2
10.8
1
2
3.3
9.2
Unpublished data collected from trees at the ASEAN Forest Tree Seed Center, Muak Lek, Saraburi, Thailand.
RS is the fruit to flower ratio multiplied by the seed to ovule ratio.
seed production areas in which irradiation to the crown is
increased by thinning, proper spacing and crown management.
High solar irradiance also enhances flowering in branches of
teak (Tectona grandis L.F.) (Nanda 1962). High temperature at
the time of floral initiation may enhance flowering in temperate trees but timing is important (Ross 1989). Temperature-enhanced flowering has been accomplished by tenting (see
Owens and Blake 1985), proper location of seed orchards and
moving potted trees into greenhouses (Philipson 1983, Ross
1988). However, this technique may be of limited value in
tropical hardwoods. In contrast, low winter temperatures have
enhanced flowering in Eucalyptus nitens (Deane & Maiden)
Maiden grafts in conjunction with paclobutrazol application
(Moncur and Hasan 1994). Water stress and root factors have
also been used to enhance flowering in temperate hardwoods
(Holmsgaard and Olson 1966) and conifers (Ross 1991). Water
stress may be crudely controlled by proper drainage, site selection and root pruning in soil-grown trees and by withholding
water in potted trees. Results with tropical trees have not been
reported. Nutrient availability may be manipulated through
fertilizer application in seed production areas and seed orchards (see Owens and Blake 1985).
In general, gibberellins (GA) promote flowering in conifers.
Cones have been induced in several species of Cupressaceae
and Taxodiaceae with GA3, often without adjunct treatments,
and in the Pinaceae, with the less polar GA4, GA7 and GA9, but
usually with a mixture of GA4/7 and adjunct treatments (see
Owens and Blake 1985). However, GAs often inhibit flower-
ing in hardwoods (Jackson and Sweet 1972, Ross and Pharis
1985, Pharis and Ross 1986, Singh and Kumar 1986). In
E. nitens, application of paclobutrazol reduced the concentration of endogenous GAs in apical tissue and enhanced flowering in trees maintained outside over winter (Moncur and Hasan
1994).
The timing of floral induction treatments is important and
must precede reproductive bud determination (Figure 2). Because biochemical changes begin before anatomical differences are visible, accurate phenological studies that correlate
floral initiation with vegetative shoot growth are essential so
that floral induction or enhancement treatments can be applied
precisely and cost effectively.
Pollination
Most temperate forest trees are wind pollinated, and the period
of anthesis is short and cued by specific environmental events
such as temperature, whereas for many tropical forest trees,
distinct environmental cues are lacking. In tropical Agathis and
Pinus, pollen shed and seed-cone receptivity are not well
synchronized resulting in long periods of pollination with only
a small proportion of cones shedding pollen and being receptive at one time. A pollen cloud, which is necessary for good
seed set in many wind-pollinated species, is absent. This major
constraint (Figure 1) may be overcome in seed orchards by
supplemental mass pollination (SMP) (Webber 1991); however, this technique cannot be applied in most seed production
areas or in natural stands where trees are large.
480
OWENS
In tropical hardwoods that are mostly pollinated by insects,
it is essential to know the phenology of the trees and pollinators
(Bawa et al. 1985). Pollinators and host trees may be very
specific and synchronized, or there may be a variety of pollinators visiting a variety of plants over long periods of time.
Because pollinators may rely on other plant species as nesting
sites and for food at other times of the year, the weed-free
understory of north-temperate seed orchards may be undesirable in tropical hardwood seed orchards. Supplementing pollinators, especially if they are colonial bees, is a viable
alternative (Moncur 1993). In many tropical hardwood seed
orchards, it may be advantageous to have widely spaced trees
with open, managed crowns to enhance pollination and an
understory of selected herbaceous or shrubby species to enhance the pollinator population.
Incompatibility responses
Figure 1. The seed efficiencies and causes for seed loss in three
conifers (adapted from Colangeli and Owens 1990, and Owens et al.
1990, 1991a).
It has generally been thought that incompatibility does not
occur in conifers (Sedgley and Griffin 1989). Rather, postzygotic self-inviability has been the only proposed mechanism.
Multiple fertilization takes place and simple polyembryony
occurs, but most embryos resulting from selfing abort during
early development, whereas embryos resulting from cross fertilization usually survive. In most conifers, if a high degree of
selfing has occurred, there will be a significant reduction in
seed production. However, we have found that, in Douglas-fir,
incompatibility may occur through reduced pollen germination and pollen tube growth or by prevention of fertilization
(Takaso and Owens 1994). The incompatibility mechanism is
less precise than in angiosperms and results from secretions
Figure 2. The time of cone initiation
in some north-temperate conifers
(Owens and Blake 1985).
CONSTRAINTS TO SEED PRODUCTION
from the integument, nucellus, megagametophyte or egg.
These secretions may only decrease the possibility of fertilization from self and certain out-cross pollen rather than completely blocking it. Therefore, incompatibility in conifers may
only reduce seed production.
In hardwoods, incompatibility is genetically controlled by
the gametophyte or sporophyte and, consequently, may occur
at different times. Sporophytic incompatibility reactions result
because the coating materials in the pollen exine are derived
from the male sporophyte parent during pollen development.
The interaction of the coating materials and the female sporophyte tissues may prevent pollen from adhering to the stigma,
prevent pollen hydration or germination, or inhibit pollen tube
penetration of the stigma (Figure 3). Gametophytic incompatibility reactions result because the pollen intine, formed entirely by the gametophyte, reacts with the female sporophyte
stigma, style or ovule tissues inhibiting pollen-tube growth
(Sedgley and Griffin 1989). Any of these reactions may prevent fertilization by incompatible pollen thus reducing seed
production, but they may also increase seed quality. These are
prezygotic (prefertilization) events.
Selfing, which is an important cause of reduced seed production in most conifer seed orchards, can be reduced by SMP
(Webber 1991) and proper seed orchard design and management. In tropical hardwood seed orchards, selfing resulting in
flower, young fruit or ovule abortion may be reduced and seed
production increased by proper spacing of trees and increasing
the number of pollinators.
481
Floral abortion
Young pine seed cones that are inadequately pollinated are
usually shed (cone drop) during the first year (Sweet 1973). In
other conifers, unpollinated seed cones usually continue to
mature, but they may be small and produce no filled seed
(Owens et al. 1991a). In hardwoods, there are many more
flowers than fruits produced. Floral abortion may occur at
many stages but is most frequent at or shortly after pollination.
In small-flowered hardwoods, abundant flowers serve as attractants for pollinators. Unpollinated flowers usually abscise
immediately, greatly reducing the Fr/Fl ratio (Table 2). Incompatibility may also result in flower abortion in a manner and
time similar to unpollinated flowers. Adverse climatic conditions, commonly low temperatures in temperate regions and
heavy rainfall and winds in tropical regions, and predators also
cause flower loss.
Ovule abortion
The S/O ratio is extremely variable among species and years.
In conifers, the number of ovules per cone may vary from a few
in the Cupressaceae to several hundred in some Pinaceae and
Araucariaceae. In temperate regions, early ovule abortion may
result from low temperatures during spring pollination (Owens
et al. 1990). Seed insects that lay eggs in conelets at pollination
(Hedlin 1974) and diseases (Sutherland et al. 1987) destroy
many ovules and young seeds. These losses can be minimized
through good seed orchard management.
Figure 3. The pistil of a flowering plant showing the five stages at which incompatibilitycan occur (Owens et al. 1991b).
482
OWENS
Most conifers have variable numbers of poorly developed
ovules at the base and tip of cones. This is a developmental
phenomenon with no apparent solution. Lack of pollination is
the major cause of ovule abortion in most conifers (Figure 1).
Unpollinated ovules often develop normally until the time of
fertilization in many Pinaceae, Cupressaceae and Araucariaceae. Seed coats may partially develop and seeds, which
may externally appear normal and filled, are actually empty or
contain only the degenerated megagametophyte (Owens et al.
1990, 1991a). In Pinus, unpollinated ovules abort before fertilization leaving small, rudimentary ‘‘seeds’’ (Owens et al.
1982). It has been estimated that if more than 20% of ovules in
a pine cone abort, the cone will abort (Sarvas 1962). In conifers, the major solution to ovule abortion is to ensure adequate
cross-pollination through seed orchard design and management.
In tropical hardwoods, the number of ovules per flower
varies (Tectona (Hedegart 1976) and Pterocarpus (Troup
1921) have four, Acacia has 16 (Table 2), and dipterocarps
generally have six (Ashton 1988)), whereas seed per fruit is
generally much less (Tectona and Pterocarpus have one to two,
Acacia has two to 16 (Table 2), and dipterocarps generally
have one (Ashton 1985)). In some temperate hardwoods such
as species of Quercus and Betula (Mogensen 1975) in which
the number of ovules is four and six, respectively, the number
of surviving ovules appears to be fixed at one; however, in most
species it is variable. Weins et al. (1987) concluded that the
number of surviving ovules in herbaceous species is fixed in
many inbreeders but random in many outcrossing species. This
has not been tested in hardwoods. Ovule abortion decreases
seed production, but it is also an important means of selection
to ensure the survival of remaining ovules (Shaanker et al.
1988). Abortion has been interpreted as a trade-off between
seed number and seed quality (Stephenson and Winsor 1986)
that permits the parent plant to match seed production with
available resources.
Ovule abortion in hardwoods is commonly caused by a lack
of pollen, but other more subtle factors are also involved
(Shaanker et al. 1988). For example, ovules closest to the
source of nutrients have the best chance of survival. Biochemically, the more fertilized ovules, the greater the plant growth
regulator production, which may result in a stronger sink for
nutrients resulting in higher ovule survival in that fruit (Sweet
1973). Developmentally, ovule abortion often results from lack
of fertilization due to incompatibility reactions (Dumas and
Knox 1983). Generally, the mechanisms by which ovule, embryo and seed ontogeny are controlled are poorly understood
in forest trees. Only recently has gene regulation of these
events begun to be studied in a few herbaceous angiosperms in
which the development and genetics are well understood
(Reiser and Fischer 1993, Thomas 1993). Such studies are not
yet possible in forest trees.
Embryo, seed and fruit development and maturation
Despite numerous studies of embryogenesis in conifers (Singh
1978), there have been few studies designed specifically to
determine the times and possible causes of embryo abortion
(Dogra 1967, 1984, Colangeli and Owens 1990, Owens et al.
1990, 1991a). Our knowledge of flowering plant embryogeny
is much more complete in herbaceous species than in hardwoods (Wardlaw 1955, Davis 1966, West and Harada 1993).
But as in conifers, there are still relatively few studies to
determine times and causes of embryo abortion. We know
almost nothing about tropical hardwood embryogeny. The
interaction of the embryo and megagametophyte in conifers is
complex and poorly understood. Embryos may abort followed
by degeneration of the megagametophyte and vice versa (Colangeli and Owens 1990). Polyembryony occurs in most conifers and increases the number of embryos per ovule and hence
the possibility of one embryo surviving. Therefore, although
early embryo abortion is a common occurrence in most ovules,
seed production only decreases when all embryos abort. In
contrast, true polyembryony is rare in hardwoods (Gifford and
Foster 1989), therefore, any embryo abnormality is likely to
cause seed loss. In conifers and hardwoods, the frequency of
embryo abortion tends to decrease as embryos develop.
Our understanding of the molecular biology underlying the
control of embryo development and subsequently abnormalities and abortion in herbaceous species is rapidly expanding
(Goldberg et al. 1989, Thomas 1993). Seed proteins are the
area of focus because they provide distinct markers of the DNA
regulated embryogenic events. Results from these studies may
provide insight into embryo development in forest trees.
Once embryos, seeds, cones and fruits are fully enlarged,
most must undergo a period of maturation. Maturation may
require certain developmental, physiological and biochemical
changes before seeds or fruits can be harvested. For example,
Tectona requires about 120 days from pollination until fruit
maturity. Fruit growth and development occur for the first 50
days and fruit maturation occurs during the final 70 days
(Hedegart 1976). However, it is difficult to distinguish immature fruits from mature fruits. Easily distinguishable external
morphological features, such as size, shape, texture and color
of seeds, cones or fruits, must be correlated with internal
embryo, endosperm, megagametophyte, seed coat or fruit wall
ontogeny and water content. Increases in abscisic acid (ABA)
occur as fruits and seeds mature. Although ABA content is a
good indicator of maturity, it is not easily identified and measured. The accumulation of storage products, most commonly
proteins but in some seeds, carbohydrates or lipids, is correlated with increases in ABA. These can easily be identified by
simple histochemical tests (Owens et al. 1991b) on tissue
slices. Such studies can assist in developing field criteria for
identifying mature seeds and fruits for harvest.
Conclusions
Constraints to seed reproduction are many and varied and act
at all stages of the reproductive cycle. The number and relative
importance of these constraints vary with species. Before these
constraints can be fully understood and possibly controlled,
the reproductive cycle of the species must be understood.
Reproductive success has usually been measured as fruit or
seed produced at the end of long reproductive cycles; however,
CONSTRAINTS TO SEED PRODUCTION
it could be more informative to determine reproductive success
in relation to reproductive potential using fruit to flower and
seed to ovule ratios. This information, coupled with more
detailed studies of ontogeny, would allow constraints to be
identified and solutions sought.
Acknowledgments
The research that formed the basis of most of this paper was supported
by a research grant from the Natural Sciences and Engineering Research Council of Canada. Some unpublished information on tropical
forest trees was provided by Mr. Prasert Sornsathapornkul and Mr.
Suwan Tangmitcharoen, graduate students under Canadian International Development Agency scholarships.
References
Ashton, P.S. 1988. Dipterocarp biology as a window to the understanding of tropical forest structure. Annu. Rev. Ecol. Syst. 19:347-370.
Ashton, P.S., T.J. Givnish and S. Appanah. 1988. Staggered flowering
in the Dipterocarpaceae: new insights into floral induction and the
evolution of mast fruiting in the aseasonal tropics. Am. Nat.
132:44--66.
Bawa, K.S., D.R. Perry and J.H. Beach. 1985. Reproductive biology
of tropical lowland rain forest trees. I. Sexual systems and incompatibility mechanisms. Am. J. Bot. 72:331--345.
Colangeli, A.M. and J.N. Owens. 1990. Cone and seed development
in a wind-pollinated, western hemlock (Tsuga heterophylla) clone
bank. Can. J. For. Res. 20:1432--1437.
Davis, G.L. 1966. Systematic embryology of the angiosperms. Wiley,
New York, 528 p.
Dogra, P.D. 1967. Seed sterility and disturbances in embryogeny in
conifers with particular reference to seed testing and tree breeding
in Pinaceae. Stud. For. Suec. 45:1--97.
Dogra, P.D. 1984. The embryology, breeding systems and seed sterility in Cupressaceae----a monograph. In Glimpses in Plant Research.
Aspects of Reproductive Biology. Ed. P.K.K. Nair. Vikas Publishing
House PVT Ltd., New Delhi, India, pp 1--127.
Dumas, C. and R.B. Knox. 1983. Callose and determination of pistil
viability and incompatibility. Theor. Appl. Genet. 67:1--10.
El-Kassaby, Y.A. 1992. Domestication and genetic diversity----should
we be concerned? For. Chron. 68:687--700.
Gifford, E.M. and A.S. Foster. 1989. Morphology and evolution of
vascular plants, 3rd Edn. W.H. Freeman and Co., New York, 626 p.
Goldberg, D.J., S.J. Barker and L. Parez-Grav. 1989. Regulation of
gene expression during plant embryogenesis. Cell 56:149--160.
Griffin, A.R., P. Whiteman, T. Rudge, I.P. Burgess and M. Moncur.
1993. Effect of paclobutrazol on flower-bud production and vegetative growth in two species of Eucalyptus. Can. J. For. Res. 23:640-647.
Harrison, D.L.S. and M.U. Slee. 1991. Gibberellin A4/7 enhanced
flowering in Pinus caribaea var. hondurensis. Can. J. For. Res.
21:788--793.
Hedegart, T. 1976. Breeding systems, variation and genetic improvement of teak (Tectona grandis L.f). In Tropical Trees. Eds. J. Burley
and B.T. Styles. Academic Press, London, pp 109--123.
Hedlin, A.F. 1974. Cone and seed insects of British Columbia. Environ. Can., Can. For. Serv. Inf. Rep. BC-X- 90, 61 p.
Holmsgaard, E. and H.C. Olsen. 1966. Experimental induction of
flowering in beech. Forst. Forsogsraes. Dan. 30:3--17.
483
Jackson, D.I. and G.B. Sweet. 1972. Flower initiation in temperate
woody plants. A review based largely on the literature of conifers
and deciduous fruit trees. N.Z. For. Serv. Reprint No. 508, pp 9--24.
Janzen, D.H. 1978. Seeding patterns of tropical trees. In Tropical Trees
as Living Systems. Eds. P.B. Tomlinson and M.H. Zimmermann.
Cambridge University Press, Cambridge, U.K., pp 83--128.
Moncur, M.W. 1994. Flower induction and enhancement in tropical
species. In Proceedings: International Symposium on Genetic Conservation and Production of Tropical Forest Tree Seed. Eds. M.
Drysdale, S.E.T. John and A.C. Yap. Asean-Canada Forest Tree
Seed Center, Chaingmai, Thailand.
Moncur, M.W. and D.J. Boland. 1989. Floral morphology of Eucalyptus melliodoro A. Cunn. ex. Schau. and comparison with other
eucalypt species. Aust. J. Bot. 37:125--135.
Moncur, M.W. and O. Hasan. 1994. Floral induction in Eucalyptus
nitens. Tree Physiol. 14:1303--1312.
Nanda, K.K. 1962. Some observations on growth, branching behaviour and flowering of teak (Tectona grandis L.F.) in relation to light.
Indian For. 88:207--218.
Ng, F.S.P. 1981. Vegetation and reproductive phenology of dipterocarps. Malays. For. 44:197--221.
Ngamkhajomawat, S. and P. Wasuwanich. 1986. Preliminary studies
on flower development in Dipterocarpus alatus roxb. Embryon
2:56--61.
Owens, J.N. 1991. Flowering and seed set. In Physiology of Trees. Ed.
A.S. Raghavendra. John Wiley and Sons, pp 247--271.
Owens, J.N. and M.D. Blake. 1985. Forest tree seed production. A
review of literature and recommendations for future research. Environ. Can., Can. For. Serv. Inf. Rep. PI-X-53, 161 p.
Owens, J.N., A.M. Colangeli and S.J. Morris. 1990. The effect of self,
cross and no pollination on ovule, embryo, seed and cone development in western red cedar (Thuja plicata Donn). Can. J. For. Res.
20:66--75.
Owens, J.N., A.M. Colangeli, and S.J. Morris. 1991a. Factors affecting seed set in Douglas-fir (Pseudotsuga menziesii (Mirb) Franco).
Can. J. Bot. 69:229--238.
Owens, J.N., S.J. Simpson and M. Molder. 1982. Sexual reproduction
of Pinus contorta. II. Postdormancy ovule, embryo, and seed development. Can. J. Bot. 60:2071--2083.
Owens, J.N., P. Sornsathapornkul and S. Tangmitchareon. 1991b.
Manual: studying flowering and seed ontogeny in tropical forest
trees. ASEAN-Canada Forest Tree Seed Center Project, Muak-Lek,
Saraburi, Thailand, 134 p.
Pharis, R.P. and S.D. Ross. 1986. Pinaceae. Hormonal promotion of
flowering. In Handbook of Flowering, Vol. 5. Ed. A. Halevy. C.R.C.
Press, Boca Raton, FL, pp 2--41.
Philipson, J.J. 1983. The role of gibberellin A4/7, heat and drought in
the induction of flowering in Sitka spruce. J. Exp. Bot. 34:291--302.
Reiser, L. and R.L. Fischer. 1993. The ovule and the embryo sac. Plant
Cell 5:1291--1301.
Ross, S.D. 1988. Effects of temperature, drought, gibberellin A4/7, and
their timing on flowering in potted Picea engelmannii and P. glauca
grafts. Can. J. For. Res. 18:163--171.
Ross, S.D. 1989. Temperature influences on reproductive processes in
conifers. In Climate Applications in Forest Renewal and Forest
Production. Eds. D.C. MacIver, R.B. Street and A.N. Auclair. Can.
Gov. Print. Center, Ottawa, pp 40--43.
Ross, S.D. 1991. Promotion of flowering in potted white spruce grafts
by root pruning: its relationship to drought and shoot elongation.
Can. J. For. Res. 21:680--685.
484
OWENS
Ross, S.D. and R.P. Pharis. 1985. Flower induction in crop trees:
different mechanisms and techniques with special reference to
conifers. In Attributes of Trees as Crop Plants. Eds. M.G.R. Cannell, J.E. Jackson and J.C. Gordon. Institute of Terrestrial Ecology,
Monkswood Experimental Station, Abbots Ripton, Huntingdon,
U.K., pp 383--397.
Sarvas, R. 1962. Investigations on the flowering and seed crop of
Pinus sylvestris. Commun. Inst. For. Fenn. 53:1--198.
Sedgley, M. and A.R. Griffin. 1989. Sexual reproduction of tree crops.
Academic Press Ltd., London, 378 p.
Shaanker, R.V., K.N. Ganeshaiah and K.S. Bawa. 1988. Parent--offspring conflict, sibling rivalry, and brood size patterns in plants.
Annu. Rev. Ecol. Syst. 19:177--205.
Singh, K. and S. Kumar. 1986. Manipulation of flowering in wood
plants by the use of gibberellin. Indian J. For. 9:331--336.
Stephenson, A.G. and J.A. Winsor. 1986. Lotus corniculatus regulates
offspring quality through selective fruit abortion. Evolution
40:453--458.
Sutherland, J.R., T. Miller and R.S. Quinard. 1987. Cone and seed
diseases of North American conifers. North American Forestry
Commission Publication No. 1, Victoria, B.C., Canada, 77 p.
Sweet, G.B. 1973. Shedding of reproductive structures in forest trees.
In Shedding of Plant Parts. Ed. T.T. Kozlowski. Academic Press,
New York, pp 341--382.
Takaso, T. and J.N. Owens. 1994. Effects of ovular secretions on
pollen in Pseudotsuga menziesii (Pinaceae). Am. J. Bot. 81:504-513.
Thomas, T.L. 1993. Gene expression during plant embryogenesis and
germination: an overview. Plant Cell 5:1401--1410.
Troup, R.S. 1921. The silviculture of Indian trees, Vol. 1. Clarendon
Press, Oxford, U.K., pp 287--292.
Wardlaw, C.W. 1955. Embryogenesis in plants. Methuen, London,
381 p.
Webber, J.E. 1991. Interior spruce pollen management manual. Land
Management Report. No. 70., B.C. Ministry of Forests, 25 p.
West, M.A.L. and J.J. Harada. 1993. Embryogenesis in higher plants:
an overview. Plant Cell 5:1361--1369.
Wiens, D., C.L. Calvin, C.A. Wilson, C.I. Davern, D. Frank and S.R.
Seavey. 1987. Reproductive success, spontaneous embryo abortion,
and genetic load in flowering plants. Oecologia 71:501--509.
© 1995 Heron Publishing----Victoria, Canada
Constraints to seed production: temperate and tropical forest trees
J. N. OWENS
Graduate Center for Forest Biology, University of Victoria, Victoria, British Columbia V8W 2Y2, Canada
Received May 31, 1994
Summary Reforestation requires a constant supply of high
quality seed. Different methods have been used to measure seed
and fruit production as a proportion of that which potentially
could be produced. These estimates coupled with developmental studies permit determination of when and to what extent
different biological constraints reduce seed and fruit production. These vary among species, years and sites. Overcoming
biological constraints may be possible in seed orchards and
seed production areas. Biological constraints include: (1) periodic or inadequate floral initiation, (2) asynchronous development and flowering, (3) floral abortion, (4) ovule abortion,
(5) embryo abortion, and (6) failure of seeds and fruits to
mature and our inability to determine maturity. Solutions to
these constraints are varied but all must begin with an understanding of reproductive biology of each species.
Keywords: embryo, floral initiation, flowering, fruit production, maturation, ovule, reforestation.
Introduction
Reforestation requires a constant supply of high quality tree
seed or vegetative propagules. However, seed is often in short
supply, as a result of much seed being of low quality with
variable maturity and a limited storage life. Seed production is
variable among species and regions, but the causes of these
variations are often uncertain. Even in seed orchards many
constraints exist and mating is usually not random. The seed
produced may represent only a portion of the genetic diversity
intended by the phenotypic selection and orchard design (ElKassaby 1992).
Seed production research must first identify species where
there are biological constraints to seed or fruit production and
then determine the biological constraints for each species in a
particular location. Most seed production research has been
done on temperate conifers (Owens and Blake 1985, Owens
1991). We know little about the reproductive biology and seed
production of tropical species, particularly the tropical hardwoods about which there is now increased interest.
Even the basic phenologies of the reproductive cycles are
poorly understood for tropical conifers and hardwoods. Most
temperate and many tropical forest trees exhibit indirect flowering in which there is a period of dormancy between floral
initiation and pollination. However, there are at least four other
groups of tropical flowering periodicities that are recognized
(Janzen 1978): (1) everflowering species, such as Ficus, that
initiate flowers throughout the year and, therefore flowers,
seeds and fruits at various stages of development are always
present; (2) nonseasonal-flowering species that exhibit flowering periodicity among plants and from branch to branch; (3)
gregarious-flowering species that initiate floral buds continuously, but these remain closed for weeks or months until
environmental conditions are favorable for anthesis, sometimes resulting in anthesis at the same time over a wide geographical area, as in Coffea; and (4) seasonal-flowering
species that flower in response to rainy or dry seasons or subtle
seasonal variations in day length or temperature, as has been
proposed for certain dipterocarps (Ashton et al. 1988).
Frequency of flowering (Table 1) (Owens et al. 1991a),
which is usually measured by frequency of seed or fruit production at the end of the reproductive cycle, is variable. Seed
production is measured as seed per hectare or per tree with
little concern for the amount of seed produced relative to the
potential seed that could be produced. More accurate, though
more laborious methods are available. For example, in conifers, seed efficiency (SEF), or seed to ovule ratio (S/O), measures the filled seed produced per cone as a percentage of the
seed potential (number of fertile ovules per cone) (Owens et al.
1990, 1991a). In flowering plants, reproductive success (RS),
which is a measure of the fruit to flower ratio (Fr/Fl) multiplied
by the S/O ratio, has been used (Weins et al. 1987). Both of
these approaches make it possible to rank species, clones or
individual trees with respect to RS and identify quickly major
constraints to seed or fruit production (Table 2).
However, use of SEF or RS does not give a complete picture
of the constraints or when they occur in the reproductive cycle.
Careful dissections or detailed anatomical techniques (Owens
et al. 1991b) are needed to determine the times and possible
causes of losses (Figure 1).
Identifying biological constraints
The reproductive phenology, from floral initiation through
pollination and seed or fruit maturity, of temperate conifers is
well known, but less is known about temperate hardwoods.
Biological constraints and solutions
Constraints and their relative importance vary among species,
sites and years. Constraints result from developmental irregu-
478
OWENS
Table 1. Bud types, age and frequency of flowering in selected hardwood and conifer trees.1
Family/
Botanical name
Aceraceae
Acer platanoides
Platanus occidentalis
Betulaceae
Alnus glutinosa
Betula pendula
Dipterocarpaceae
Dipterocarpus spp.
Fagaceae
Fagus sylvatica
Quercus petraea
Salicaceae
Populus tremuloides
Salix nigra
Oleaceae
Fraxinus excelsior
Verbenaceae
Tectona grandis
Cupressaceae
Chamaecyparis lawsoniana
Thuja plicata
Pinaceae
Abies grandis
Larix decidua
Picea abies
Pinus contorta
Pinus caribaea
Pseudotsuga menziesii
Tsuga heterophylla
1
Common name
Buds simple (S)
or mixed (M)
Buds terminal (T)
or axillary (A)
First good
crops (years)
Intervals between
seed crops (years)
Norway maple
American sycamore
M
M
T, A
A
25--30
25--30
1--3
1--3
Common alder
Silver birch
S
S
A
T, A
15--20
15
2--3
1--3
M
A
20--30
1--6
Common beech
Sessile oak
M
M
A
A
50--60
40--50
5--15
2--5
Quaking aspen
Black willow
S
S
A
A
50--702
25--752
4--5
1--3
Common ash
M
A
25--30
3--5
Teak
M
T
20--25
1--3
Lawson cypress
Western red cedar
S
S
T
T
20--25
20--25
4--5
2--3
Grand fir
European larch
Norway spruce
Lodgepole pine
Caribean pine
Douglas-fir
Western hemlock
S
S
S
M
M
S
S
A
T
T, A
A
A
A
A, T
40--45
25--30
30--35
15--20
10--15
30--35
25--30
3--5
3--5
3--5
2--3
1--3
5--7
3--4
Data compiled from various sources.
larities causing abortion, or abnormal or delayed development
of a reproductive structure. An anomaly may not become
manifest until long after its cause has occurred. The longer the
reproductive cycle, the more opportunity there is for climatic
and other factors to affect seed and fruit production adversely
(Owens 1991).
Failure to initiate floral structures
A common cause of low seed production is a failure to initiate
cones or flowers. Seasonal flowering trees do not flower frequently. They have a particular developmental stage and season at which floral initiation occurs that is mediated through
the interaction of developmental, physiological and environmental events, many of which are poorly understood. In trees
with distinct vegetative buds (Table 1), floral initiation usually
only occurs in some of these buds long before floral structures
are clearly visible. In many north-temperate conifers (Figure 2) and a few hardwoods, the time and method of floral
initiation have been determined by anatomical study (Owens
and Blake 1985). In most cases, floral initiation and early
development of floral structures occur before winter dormancy. In the few tropical forest trees that have been studied,
Pinus caribaea Morelet. (Harrison and Slee 1991), Eucalyptus
(Moncur and Boland 1989) and some dipterocarps (Ng 1981,
Ngamkhajorniway and Wasuwanich 1986, Ashton et al. 1988),
as in temperate conifers and hardwoods, floral initiation appears to be correlated with lateral shoot elongation and stages
of vegetative bud development. Such correlations are useful
because they allow us to predict the time of floral initiation
based on external morphology.
Floral enhancement is usually achieved by manipulating the
environmental factors known to promote flowering or by application of fertilizers or plant growth regulators (PGR). These
methods have been quite successful in north-temperate conifers (Owens and Blake 1985, Ross and Pharis 1985) but have
had little success in tropical hardwoods. Notable exceptions
are some fruit trees and Eucalyptus (Griffin et al. 1993, Moncur 1994), where flowering has been promoted by paclobutrazol. These flower enhancing methods may only be applicable
to potted trees, seed production areas and seed orchards where
trees are small and environmental conditions can be controlled.
Environmental factors known to affect flowering include
light intensity, temperature, water stress and nutrients. Photoperiod may influence flowering indirectly through its control
of shoot growth and bud development (Owens and Blake
1985), but it is unlikely to affect flowering in tropical forest
trees. Increasing light is a common practice in north-temperate
CONSTRAINTS TO SEED PRODUCTION
479
Table 2. Reproductive success (RS) within Acacia auriculiformis A. Cunn. ex Benth. × A. mangium hybrid.1
Spike
Flowers per spike
% Flower
survival
Fruits per
spike
Fruit per
flower
Ovules per
ovary
Seed per
pod
Seed per
ovule
RS2
1
2
3
1
3
4
2
2
5
1
8
6
2
2
1
2
6
3
10
4
0.01
0.02
0.03
0.01
0.03
0.04
0.02
0.02
0.05
0.01
0.08
0.05
0.02
0.03
0.01
0.02
0.05
0.04
0.11
0.04
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
16
12
14
5
7
8
8
7
12
11
8
8
6
14
16
13
7
4
5
10
9
0.75
0.87
0.31
0.43
0.50
0.50
0.44
0.75
0.69
0.50
0.50
0.35
0.35
0.31
0.81
0.44
0.25
0.31
0.62
0.56
0.0075
0.0174
0.0093
0.0043
0.0150
0.0200
0.0088
0.0150
0.0340
0.0045
0.0410
0.0175
0.0055
0.0083
0.0730
0.0080
0.0120
0.0110
0.0678
0.0224
0.034
16
0.58
0.0196
Prepollination
Postpollination
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
100
100
95
94
78
97
98
98
100
110
96
101
107
73
105
89
113
81
92
96
1
6
9
15
14
5
21
18
5
6
6
14
19
14
12
3
17
9
4
11
1
6
9
16
18
5
21
18
5
5
6
14
18
19
11
3
15
11
4
11
Mean
96.7
10.2
10.8
1
2
3.3
9.2
Unpublished data collected from trees at the ASEAN Forest Tree Seed Center, Muak Lek, Saraburi, Thailand.
RS is the fruit to flower ratio multiplied by the seed to ovule ratio.
seed production areas in which irradiation to the crown is
increased by thinning, proper spacing and crown management.
High solar irradiance also enhances flowering in branches of
teak (Tectona grandis L.F.) (Nanda 1962). High temperature at
the time of floral initiation may enhance flowering in temperate trees but timing is important (Ross 1989). Temperature-enhanced flowering has been accomplished by tenting (see
Owens and Blake 1985), proper location of seed orchards and
moving potted trees into greenhouses (Philipson 1983, Ross
1988). However, this technique may be of limited value in
tropical hardwoods. In contrast, low winter temperatures have
enhanced flowering in Eucalyptus nitens (Deane & Maiden)
Maiden grafts in conjunction with paclobutrazol application
(Moncur and Hasan 1994). Water stress and root factors have
also been used to enhance flowering in temperate hardwoods
(Holmsgaard and Olson 1966) and conifers (Ross 1991). Water
stress may be crudely controlled by proper drainage, site selection and root pruning in soil-grown trees and by withholding
water in potted trees. Results with tropical trees have not been
reported. Nutrient availability may be manipulated through
fertilizer application in seed production areas and seed orchards (see Owens and Blake 1985).
In general, gibberellins (GA) promote flowering in conifers.
Cones have been induced in several species of Cupressaceae
and Taxodiaceae with GA3, often without adjunct treatments,
and in the Pinaceae, with the less polar GA4, GA7 and GA9, but
usually with a mixture of GA4/7 and adjunct treatments (see
Owens and Blake 1985). However, GAs often inhibit flower-
ing in hardwoods (Jackson and Sweet 1972, Ross and Pharis
1985, Pharis and Ross 1986, Singh and Kumar 1986). In
E. nitens, application of paclobutrazol reduced the concentration of endogenous GAs in apical tissue and enhanced flowering in trees maintained outside over winter (Moncur and Hasan
1994).
The timing of floral induction treatments is important and
must precede reproductive bud determination (Figure 2). Because biochemical changes begin before anatomical differences are visible, accurate phenological studies that correlate
floral initiation with vegetative shoot growth are essential so
that floral induction or enhancement treatments can be applied
precisely and cost effectively.
Pollination
Most temperate forest trees are wind pollinated, and the period
of anthesis is short and cued by specific environmental events
such as temperature, whereas for many tropical forest trees,
distinct environmental cues are lacking. In tropical Agathis and
Pinus, pollen shed and seed-cone receptivity are not well
synchronized resulting in long periods of pollination with only
a small proportion of cones shedding pollen and being receptive at one time. A pollen cloud, which is necessary for good
seed set in many wind-pollinated species, is absent. This major
constraint (Figure 1) may be overcome in seed orchards by
supplemental mass pollination (SMP) (Webber 1991); however, this technique cannot be applied in most seed production
areas or in natural stands where trees are large.
480
OWENS
In tropical hardwoods that are mostly pollinated by insects,
it is essential to know the phenology of the trees and pollinators
(Bawa et al. 1985). Pollinators and host trees may be very
specific and synchronized, or there may be a variety of pollinators visiting a variety of plants over long periods of time.
Because pollinators may rely on other plant species as nesting
sites and for food at other times of the year, the weed-free
understory of north-temperate seed orchards may be undesirable in tropical hardwood seed orchards. Supplementing pollinators, especially if they are colonial bees, is a viable
alternative (Moncur 1993). In many tropical hardwood seed
orchards, it may be advantageous to have widely spaced trees
with open, managed crowns to enhance pollination and an
understory of selected herbaceous or shrubby species to enhance the pollinator population.
Incompatibility responses
Figure 1. The seed efficiencies and causes for seed loss in three
conifers (adapted from Colangeli and Owens 1990, and Owens et al.
1990, 1991a).
It has generally been thought that incompatibility does not
occur in conifers (Sedgley and Griffin 1989). Rather, postzygotic self-inviability has been the only proposed mechanism.
Multiple fertilization takes place and simple polyembryony
occurs, but most embryos resulting from selfing abort during
early development, whereas embryos resulting from cross fertilization usually survive. In most conifers, if a high degree of
selfing has occurred, there will be a significant reduction in
seed production. However, we have found that, in Douglas-fir,
incompatibility may occur through reduced pollen germination and pollen tube growth or by prevention of fertilization
(Takaso and Owens 1994). The incompatibility mechanism is
less precise than in angiosperms and results from secretions
Figure 2. The time of cone initiation
in some north-temperate conifers
(Owens and Blake 1985).
CONSTRAINTS TO SEED PRODUCTION
from the integument, nucellus, megagametophyte or egg.
These secretions may only decrease the possibility of fertilization from self and certain out-cross pollen rather than completely blocking it. Therefore, incompatibility in conifers may
only reduce seed production.
In hardwoods, incompatibility is genetically controlled by
the gametophyte or sporophyte and, consequently, may occur
at different times. Sporophytic incompatibility reactions result
because the coating materials in the pollen exine are derived
from the male sporophyte parent during pollen development.
The interaction of the coating materials and the female sporophyte tissues may prevent pollen from adhering to the stigma,
prevent pollen hydration or germination, or inhibit pollen tube
penetration of the stigma (Figure 3). Gametophytic incompatibility reactions result because the pollen intine, formed entirely by the gametophyte, reacts with the female sporophyte
stigma, style or ovule tissues inhibiting pollen-tube growth
(Sedgley and Griffin 1989). Any of these reactions may prevent fertilization by incompatible pollen thus reducing seed
production, but they may also increase seed quality. These are
prezygotic (prefertilization) events.
Selfing, which is an important cause of reduced seed production in most conifer seed orchards, can be reduced by SMP
(Webber 1991) and proper seed orchard design and management. In tropical hardwood seed orchards, selfing resulting in
flower, young fruit or ovule abortion may be reduced and seed
production increased by proper spacing of trees and increasing
the number of pollinators.
481
Floral abortion
Young pine seed cones that are inadequately pollinated are
usually shed (cone drop) during the first year (Sweet 1973). In
other conifers, unpollinated seed cones usually continue to
mature, but they may be small and produce no filled seed
(Owens et al. 1991a). In hardwoods, there are many more
flowers than fruits produced. Floral abortion may occur at
many stages but is most frequent at or shortly after pollination.
In small-flowered hardwoods, abundant flowers serve as attractants for pollinators. Unpollinated flowers usually abscise
immediately, greatly reducing the Fr/Fl ratio (Table 2). Incompatibility may also result in flower abortion in a manner and
time similar to unpollinated flowers. Adverse climatic conditions, commonly low temperatures in temperate regions and
heavy rainfall and winds in tropical regions, and predators also
cause flower loss.
Ovule abortion
The S/O ratio is extremely variable among species and years.
In conifers, the number of ovules per cone may vary from a few
in the Cupressaceae to several hundred in some Pinaceae and
Araucariaceae. In temperate regions, early ovule abortion may
result from low temperatures during spring pollination (Owens
et al. 1990). Seed insects that lay eggs in conelets at pollination
(Hedlin 1974) and diseases (Sutherland et al. 1987) destroy
many ovules and young seeds. These losses can be minimized
through good seed orchard management.
Figure 3. The pistil of a flowering plant showing the five stages at which incompatibilitycan occur (Owens et al. 1991b).
482
OWENS
Most conifers have variable numbers of poorly developed
ovules at the base and tip of cones. This is a developmental
phenomenon with no apparent solution. Lack of pollination is
the major cause of ovule abortion in most conifers (Figure 1).
Unpollinated ovules often develop normally until the time of
fertilization in many Pinaceae, Cupressaceae and Araucariaceae. Seed coats may partially develop and seeds, which
may externally appear normal and filled, are actually empty or
contain only the degenerated megagametophyte (Owens et al.
1990, 1991a). In Pinus, unpollinated ovules abort before fertilization leaving small, rudimentary ‘‘seeds’’ (Owens et al.
1982). It has been estimated that if more than 20% of ovules in
a pine cone abort, the cone will abort (Sarvas 1962). In conifers, the major solution to ovule abortion is to ensure adequate
cross-pollination through seed orchard design and management.
In tropical hardwoods, the number of ovules per flower
varies (Tectona (Hedegart 1976) and Pterocarpus (Troup
1921) have four, Acacia has 16 (Table 2), and dipterocarps
generally have six (Ashton 1988)), whereas seed per fruit is
generally much less (Tectona and Pterocarpus have one to two,
Acacia has two to 16 (Table 2), and dipterocarps generally
have one (Ashton 1985)). In some temperate hardwoods such
as species of Quercus and Betula (Mogensen 1975) in which
the number of ovules is four and six, respectively, the number
of surviving ovules appears to be fixed at one; however, in most
species it is variable. Weins et al. (1987) concluded that the
number of surviving ovules in herbaceous species is fixed in
many inbreeders but random in many outcrossing species. This
has not been tested in hardwoods. Ovule abortion decreases
seed production, but it is also an important means of selection
to ensure the survival of remaining ovules (Shaanker et al.
1988). Abortion has been interpreted as a trade-off between
seed number and seed quality (Stephenson and Winsor 1986)
that permits the parent plant to match seed production with
available resources.
Ovule abortion in hardwoods is commonly caused by a lack
of pollen, but other more subtle factors are also involved
(Shaanker et al. 1988). For example, ovules closest to the
source of nutrients have the best chance of survival. Biochemically, the more fertilized ovules, the greater the plant growth
regulator production, which may result in a stronger sink for
nutrients resulting in higher ovule survival in that fruit (Sweet
1973). Developmentally, ovule abortion often results from lack
of fertilization due to incompatibility reactions (Dumas and
Knox 1983). Generally, the mechanisms by which ovule, embryo and seed ontogeny are controlled are poorly understood
in forest trees. Only recently has gene regulation of these
events begun to be studied in a few herbaceous angiosperms in
which the development and genetics are well understood
(Reiser and Fischer 1993, Thomas 1993). Such studies are not
yet possible in forest trees.
Embryo, seed and fruit development and maturation
Despite numerous studies of embryogenesis in conifers (Singh
1978), there have been few studies designed specifically to
determine the times and possible causes of embryo abortion
(Dogra 1967, 1984, Colangeli and Owens 1990, Owens et al.
1990, 1991a). Our knowledge of flowering plant embryogeny
is much more complete in herbaceous species than in hardwoods (Wardlaw 1955, Davis 1966, West and Harada 1993).
But as in conifers, there are still relatively few studies to
determine times and causes of embryo abortion. We know
almost nothing about tropical hardwood embryogeny. The
interaction of the embryo and megagametophyte in conifers is
complex and poorly understood. Embryos may abort followed
by degeneration of the megagametophyte and vice versa (Colangeli and Owens 1990). Polyembryony occurs in most conifers and increases the number of embryos per ovule and hence
the possibility of one embryo surviving. Therefore, although
early embryo abortion is a common occurrence in most ovules,
seed production only decreases when all embryos abort. In
contrast, true polyembryony is rare in hardwoods (Gifford and
Foster 1989), therefore, any embryo abnormality is likely to
cause seed loss. In conifers and hardwoods, the frequency of
embryo abortion tends to decrease as embryos develop.
Our understanding of the molecular biology underlying the
control of embryo development and subsequently abnormalities and abortion in herbaceous species is rapidly expanding
(Goldberg et al. 1989, Thomas 1993). Seed proteins are the
area of focus because they provide distinct markers of the DNA
regulated embryogenic events. Results from these studies may
provide insight into embryo development in forest trees.
Once embryos, seeds, cones and fruits are fully enlarged,
most must undergo a period of maturation. Maturation may
require certain developmental, physiological and biochemical
changes before seeds or fruits can be harvested. For example,
Tectona requires about 120 days from pollination until fruit
maturity. Fruit growth and development occur for the first 50
days and fruit maturation occurs during the final 70 days
(Hedegart 1976). However, it is difficult to distinguish immature fruits from mature fruits. Easily distinguishable external
morphological features, such as size, shape, texture and color
of seeds, cones or fruits, must be correlated with internal
embryo, endosperm, megagametophyte, seed coat or fruit wall
ontogeny and water content. Increases in abscisic acid (ABA)
occur as fruits and seeds mature. Although ABA content is a
good indicator of maturity, it is not easily identified and measured. The accumulation of storage products, most commonly
proteins but in some seeds, carbohydrates or lipids, is correlated with increases in ABA. These can easily be identified by
simple histochemical tests (Owens et al. 1991b) on tissue
slices. Such studies can assist in developing field criteria for
identifying mature seeds and fruits for harvest.
Conclusions
Constraints to seed reproduction are many and varied and act
at all stages of the reproductive cycle. The number and relative
importance of these constraints vary with species. Before these
constraints can be fully understood and possibly controlled,
the reproductive cycle of the species must be understood.
Reproductive success has usually been measured as fruit or
seed produced at the end of long reproductive cycles; however,
CONSTRAINTS TO SEED PRODUCTION
it could be more informative to determine reproductive success
in relation to reproductive potential using fruit to flower and
seed to ovule ratios. This information, coupled with more
detailed studies of ontogeny, would allow constraints to be
identified and solutions sought.
Acknowledgments
The research that formed the basis of most of this paper was supported
by a research grant from the Natural Sciences and Engineering Research Council of Canada. Some unpublished information on tropical
forest trees was provided by Mr. Prasert Sornsathapornkul and Mr.
Suwan Tangmitcharoen, graduate students under Canadian International Development Agency scholarships.
References
Ashton, P.S. 1988. Dipterocarp biology as a window to the understanding of tropical forest structure. Annu. Rev. Ecol. Syst. 19:347-370.
Ashton, P.S., T.J. Givnish and S. Appanah. 1988. Staggered flowering
in the Dipterocarpaceae: new insights into floral induction and the
evolution of mast fruiting in the aseasonal tropics. Am. Nat.
132:44--66.
Bawa, K.S., D.R. Perry and J.H. Beach. 1985. Reproductive biology
of tropical lowland rain forest trees. I. Sexual systems and incompatibility mechanisms. Am. J. Bot. 72:331--345.
Colangeli, A.M. and J.N. Owens. 1990. Cone and seed development
in a wind-pollinated, western hemlock (Tsuga heterophylla) clone
bank. Can. J. For. Res. 20:1432--1437.
Davis, G.L. 1966. Systematic embryology of the angiosperms. Wiley,
New York, 528 p.
Dogra, P.D. 1967. Seed sterility and disturbances in embryogeny in
conifers with particular reference to seed testing and tree breeding
in Pinaceae. Stud. For. Suec. 45:1--97.
Dogra, P.D. 1984. The embryology, breeding systems and seed sterility in Cupressaceae----a monograph. In Glimpses in Plant Research.
Aspects of Reproductive Biology. Ed. P.K.K. Nair. Vikas Publishing
House PVT Ltd., New Delhi, India, pp 1--127.
Dumas, C. and R.B. Knox. 1983. Callose and determination of pistil
viability and incompatibility. Theor. Appl. Genet. 67:1--10.
El-Kassaby, Y.A. 1992. Domestication and genetic diversity----should
we be concerned? For. Chron. 68:687--700.
Gifford, E.M. and A.S. Foster. 1989. Morphology and evolution of
vascular plants, 3rd Edn. W.H. Freeman and Co., New York, 626 p.
Goldberg, D.J., S.J. Barker and L. Parez-Grav. 1989. Regulation of
gene expression during plant embryogenesis. Cell 56:149--160.
Griffin, A.R., P. Whiteman, T. Rudge, I.P. Burgess and M. Moncur.
1993. Effect of paclobutrazol on flower-bud production and vegetative growth in two species of Eucalyptus. Can. J. For. Res. 23:640-647.
Harrison, D.L.S. and M.U. Slee. 1991. Gibberellin A4/7 enhanced
flowering in Pinus caribaea var. hondurensis. Can. J. For. Res.
21:788--793.
Hedegart, T. 1976. Breeding systems, variation and genetic improvement of teak (Tectona grandis L.f). In Tropical Trees. Eds. J. Burley
and B.T. Styles. Academic Press, London, pp 109--123.
Hedlin, A.F. 1974. Cone and seed insects of British Columbia. Environ. Can., Can. For. Serv. Inf. Rep. BC-X- 90, 61 p.
Holmsgaard, E. and H.C. Olsen. 1966. Experimental induction of
flowering in beech. Forst. Forsogsraes. Dan. 30:3--17.
483
Jackson, D.I. and G.B. Sweet. 1972. Flower initiation in temperate
woody plants. A review based largely on the literature of conifers
and deciduous fruit trees. N.Z. For. Serv. Reprint No. 508, pp 9--24.
Janzen, D.H. 1978. Seeding patterns of tropical trees. In Tropical Trees
as Living Systems. Eds. P.B. Tomlinson and M.H. Zimmermann.
Cambridge University Press, Cambridge, U.K., pp 83--128.
Moncur, M.W. 1994. Flower induction and enhancement in tropical
species. In Proceedings: International Symposium on Genetic Conservation and Production of Tropical Forest Tree Seed. Eds. M.
Drysdale, S.E.T. John and A.C. Yap. Asean-Canada Forest Tree
Seed Center, Chaingmai, Thailand.
Moncur, M.W. and D.J. Boland. 1989. Floral morphology of Eucalyptus melliodoro A. Cunn. ex. Schau. and comparison with other
eucalypt species. Aust. J. Bot. 37:125--135.
Moncur, M.W. and O. Hasan. 1994. Floral induction in Eucalyptus
nitens. Tree Physiol. 14:1303--1312.
Nanda, K.K. 1962. Some observations on growth, branching behaviour and flowering of teak (Tectona grandis L.F.) in relation to light.
Indian For. 88:207--218.
Ng, F.S.P. 1981. Vegetation and reproductive phenology of dipterocarps. Malays. For. 44:197--221.
Ngamkhajomawat, S. and P. Wasuwanich. 1986. Preliminary studies
on flower development in Dipterocarpus alatus roxb. Embryon
2:56--61.
Owens, J.N. 1991. Flowering and seed set. In Physiology of Trees. Ed.
A.S. Raghavendra. John Wiley and Sons, pp 247--271.
Owens, J.N. and M.D. Blake. 1985. Forest tree seed production. A
review of literature and recommendations for future research. Environ. Can., Can. For. Serv. Inf. Rep. PI-X-53, 161 p.
Owens, J.N., A.M. Colangeli and S.J. Morris. 1990. The effect of self,
cross and no pollination on ovule, embryo, seed and cone development in western red cedar (Thuja plicata Donn). Can. J. For. Res.
20:66--75.
Owens, J.N., A.M. Colangeli, and S.J. Morris. 1991a. Factors affecting seed set in Douglas-fir (Pseudotsuga menziesii (Mirb) Franco).
Can. J. Bot. 69:229--238.
Owens, J.N., S.J. Simpson and M. Molder. 1982. Sexual reproduction
of Pinus contorta. II. Postdormancy ovule, embryo, and seed development. Can. J. Bot. 60:2071--2083.
Owens, J.N., P. Sornsathapornkul and S. Tangmitchareon. 1991b.
Manual: studying flowering and seed ontogeny in tropical forest
trees. ASEAN-Canada Forest Tree Seed Center Project, Muak-Lek,
Saraburi, Thailand, 134 p.
Pharis, R.P. and S.D. Ross. 1986. Pinaceae. Hormonal promotion of
flowering. In Handbook of Flowering, Vol. 5. Ed. A. Halevy. C.R.C.
Press, Boca Raton, FL, pp 2--41.
Philipson, J.J. 1983. The role of gibberellin A4/7, heat and drought in
the induction of flowering in Sitka spruce. J. Exp. Bot. 34:291--302.
Reiser, L. and R.L. Fischer. 1993. The ovule and the embryo sac. Plant
Cell 5:1291--1301.
Ross, S.D. 1988. Effects of temperature, drought, gibberellin A4/7, and
their timing on flowering in potted Picea engelmannii and P. glauca
grafts. Can. J. For. Res. 18:163--171.
Ross, S.D. 1989. Temperature influences on reproductive processes in
conifers. In Climate Applications in Forest Renewal and Forest
Production. Eds. D.C. MacIver, R.B. Street and A.N. Auclair. Can.
Gov. Print. Center, Ottawa, pp 40--43.
Ross, S.D. 1991. Promotion of flowering in potted white spruce grafts
by root pruning: its relationship to drought and shoot elongation.
Can. J. For. Res. 21:680--685.
484
OWENS
Ross, S.D. and R.P. Pharis. 1985. Flower induction in crop trees:
different mechanisms and techniques with special reference to
conifers. In Attributes of Trees as Crop Plants. Eds. M.G.R. Cannell, J.E. Jackson and J.C. Gordon. Institute of Terrestrial Ecology,
Monkswood Experimental Station, Abbots Ripton, Huntingdon,
U.K., pp 383--397.
Sarvas, R. 1962. Investigations on the flowering and seed crop of
Pinus sylvestris. Commun. Inst. For. Fenn. 53:1--198.
Sedgley, M. and A.R. Griffin. 1989. Sexual reproduction of tree crops.
Academic Press Ltd., London, 378 p.
Shaanker, R.V., K.N. Ganeshaiah and K.S. Bawa. 1988. Parent--offspring conflict, sibling rivalry, and brood size patterns in plants.
Annu. Rev. Ecol. Syst. 19:177--205.
Singh, K. and S. Kumar. 1986. Manipulation of flowering in wood
plants by the use of gibberellin. Indian J. For. 9:331--336.
Stephenson, A.G. and J.A. Winsor. 1986. Lotus corniculatus regulates
offspring quality through selective fruit abortion. Evolution
40:453--458.
Sutherland, J.R., T. Miller and R.S. Quinard. 1987. Cone and seed
diseases of North American conifers. North American Forestry
Commission Publication No. 1, Victoria, B.C., Canada, 77 p.
Sweet, G.B. 1973. Shedding of reproductive structures in forest trees.
In Shedding of Plant Parts. Ed. T.T. Kozlowski. Academic Press,
New York, pp 341--382.
Takaso, T. and J.N. Owens. 1994. Effects of ovular secretions on
pollen in Pseudotsuga menziesii (Pinaceae). Am. J. Bot. 81:504-513.
Thomas, T.L. 1993. Gene expression during plant embryogenesis and
germination: an overview. Plant Cell 5:1401--1410.
Troup, R.S. 1921. The silviculture of Indian trees, Vol. 1. Clarendon
Press, Oxford, U.K., pp 287--292.
Wardlaw, C.W. 1955. Embryogenesis in plants. Methuen, London,
381 p.
Webber, J.E. 1991. Interior spruce pollen management manual. Land
Management Report. No. 70., B.C. Ministry of Forests, 25 p.
West, M.A.L. and J.J. Harada. 1993. Embryogenesis in higher plants:
an overview. Plant Cell 5:1361--1369.
Wiens, D., C.L. Calvin, C.A. Wilson, C.I. Davern, D. Frank and S.R.
Seavey. 1987. Reproductive success, spontaneous embryo abortion,
and genetic load in flowering plants. Oecologia 71:501--509.