Table Of Content云 南 植 物 研 究 2008, 30 (3): 279～294
Acta Botanica Yunnanica DOI: 10.3724 SP.J.1143.2008.08017
Advances in Understanding Seed Dormancy at the Whole-seed Level:
An Ecological, Biogeographical and Phylogenetic Perspective
1,2 1
BASKIN Carol C , BASKIN Jerry M
(1 Department of Biology, University of Kentucky, Lexington, KY 40506, USA;
2 Department of Plant and Soil Sciences, University of Kentucky, Lexington, KY40546, USA)
Abstract: Following a brief account of the early foundations of seedgermination ecology sensu lato, some historical and re-
cent developments pertaining to the ecology, biogeography and phylogeny of seed dormancy are discussed .
Key words: Biogeography of seed dormancy; Ecology of seed dormancy; Phylogeny of seed dormancy; Seed dormancy;
Seed germination
CLC number: Q948 Document Code: A Article ID: 0253 -2700 (2008) 03- 279- 16
A celebration publication (festschrift) is an oppor- complexities of seed germination in the natural environ-
tunity to (1) think about the early work that helped to ment .This excitement was further enhanced when ecol-
establish a particular discipline, (2) reflect on its cur- ogists began to pay attention to data from studies on wa-
rent focal points of research activity and (3) contemplate ter-permeable seeds with physiological dormancy in the
what the future directions might be as the field continues temperate climate of the United Kingdom ( Ratcliffe,
to develop . Here, we wish to celebrate the ecology of 1961; Newman, 1963) and those on water-imperme-
seed dormancy and germination in its broadest sense able (physically-dormant) seeds done in the arid region
( sensu lato) . of Australia by Quinlivan (1961) , all of which showed
that timing of dormancy-break played an important role
Foundations in timing of germination in the field .
Control of seed germination in nature With recognition that dormancy-break was corre-
lated with certain seasons of the year, e. g ., cold,
The date when people first began to make obser-
wet conditions of winter (cold stratification), seed ger-
vations on seed germination phenology and to wonder
mination ecologists paid increased attention to the work
why species might have different germination seasons
has not been recorded . However, the Greek philoso- of seed physiologists who had been investigating the
pher Theophrastus ( ca . 372 - 287 B . C .) had a keen specific dormancy-breaking requirements of seeds in the
interest in seeds and commented on the fact that germi- laboratory greenhouse for several decades . For exam-
nation was influenced by climatic factors ( Evenari, ple , Barton (1930) was studying the effects of cold
1980 - 81) . Indeed, one of the major focal points of stratification on dormancy break in the early 1930s .
early research on germination ecology was trying to ex- The subsequent“marriage”of ecology and physiology
plain what controlled the timing of seed germination in in studies of seed dormancy break and germination
the field, i. e ., under natural habitat conditions . The turned out to be a very productive approach .
work of Went (1949) in the deserts of southwestern One aspect of the control of timing of seed germi-
USA and that of Koller (1955) , Evenari (1980 - 81) nation in the field that received considerable early at-
and Gutterman ( 1981 ) in the deserts of Israel did tention was germination of seeds in the soil, especially
much to excite ecologists about trying to understand the weed seeds . It had long been recognized that arable
Dedicated to the30th Anniversary of Acta Botanica Yunnanica
Received date: 2008- 02-15, Accepted date: 2008-02-22
2 8 0 云 南 植 物 研 究 30卷
soils may contain large numbers of seeds (Darwin, the Boyce Thompson Institute in Yonkers [ now at
1859 [ 1892 ]) . Further, studies by Brenchley and Ithaca], NewYork (USA), investigated the dormancy-
Warington (1936) showed that buried seeds of many breaking requirements of seeds of many economically-
weeds were not in a constant state of readiness to germ- important species in the 1920s and 1930s (Crocker and
inate . Thus, if seed-containing soil was disturbed in Barton, 1957 ) . Much research was done in various
one season ( e. g ., spring and or autumn) seeds of a laboratories to determine (1) the best techniques for
particular species might germinate; however, if it was making seed coats permeable in species with physical
disturbed in another season no seedlings of that species dormancy, (2) the minimum length of the cold stratifi-
appeared but those of another one might . The first di- cation period require to break dormancy and (3) the
rect evidence that buried seeds of some species undergo optimal period of dry storage at room temperatures (af-
seasonal changes in their dormancy state was obtained terripening) required to break dormancy (see Crocker
by Courtney (1968) for seeds of Polygonum aviculare, and Barton, 1957) . Crocker (1916) developed one of
and it was confirmed by Schafer and Chilcote (1970) the first dormancy classification systems, in which he
and Taylorson (1970) shortly thereafter .Concern about distinguished seven kinds of dormancy with regard to
buried seeds, especially from a weed-control perspec- their cause: (1) underdevelopment of the embryo, (2)
tive, resulted in studies on longevity of seeds in soil water-impermeable seed or fruit coat, (3) mechanical
(Telewski and Zeevaart, 2002), development of sim- resistance of seed covering layers, (4) low gas perme-
ple models of the fate dynamics of seeds in the soil ability of seed covers, (5) metabolic (physiological)
(Schafer and Chilcote, 1969, 1970; Roberts, 1972)
block in the embryo, (6) combined dormancy and (7)
and descriptions of two types of transient and two types secondary dormancy .
of persistent seed banks (Thompson and Grime, 1979; When Nikolaeva began her Ph . D . dissertation
Grime 1981); the list of persistent seed bank types research on the“physiology of deep dormancy in seeds”
subsequently was expanded to seven ( Roberts, 1981;
(translated into English in 1969) at the Komarov Bo-
Baskin and Baskin, 1998) .
tanical Institute in St . Petersburg, Russia, she knew
One of the products of early studies on control of about Crocker′s classification system and the work done
timing of germination in the field was the development
on seeds in various laboratories around the world (see
of theoretical models, in particular those of Cohen
references cited in Nikolaeva, 1969 ) . Using her
(1968) . He worked in the randomly varying ( unpre-
knowledge of the literature and her vast experience with
dictable rainfall) environment of the Negev Desert in
seed dormancy, Nikolaeva devised the most logical and
Israel . His models identified various ecological situ-
detailed seed dormancy classification system yet devel-
ations that may have lead to the selection of seed
oped (Nikolaeva, 1969) , which she subsequently re-
dormancy, and they have served as the foundation for
vised slightly (Nikolaeva, 1977, 2001) . Her system is
much additional theoretical thinking about seeds (Ven-
based on seed internal structure ( e. g ., fully-devel-
able and Lawlor, 1980; Bulmer, 1984; Philippi and
oped vs . underdeveloped embryo), role of seed cover-
Seger, 1989; Brown and Venable, 1991) .
ing layers ( i. e ., impermeability to water, mechanical
Development of a dormancy classification system
restriction to radicle emergence, presence of chemical
Dormancy in seeds is recognized as the failure of inhibitors); physiological requirements for dormancy-
seeds to germinate, being particularly troublesome to break (cold stratification and or warm stratification);
people who may sow seeds that subsequently fail to ger- and effects of plant growth regulators on breaking phys-
minate . On the other hand, nondormant seeds are iological dormancy .
those that germinate readily when placed under suitable In her system of dormancy classification, Nikolae-
temperature, light, moisture and oxygen conditions . As va gave both a name and a letter number formula to the
mentioned above, plant physiologists, e. g ., those at seven types and various subtypes of dormancy she rec-
3 期 BASKIN and BASKIN: Advances in Understanding Seed Dormancy at the Whole-seed Level: An ... 2 8 1
ognized: three types of exogenous (physical, chemical terman and Porath, 1975), light quality (McCullough
and mechanical) , three types of endogenous ( physio- and Shropshire, 1970 ), mineral nutrition (Thurston,
logical, morphological and morphophysiological) and 1951), position on mother plant (Negbi and Tamari,
many types of combinational (exogenous plus endoge- 1963), soil moisture (Pallas et al., 1977) and tem-
nous) dormancy . For example, deep simple epicotyl perature (Junttila (1971) . Plants of the same species
e
dormancy was represented by the formula B-C , where growing in sites varying with respect to elevation (Mey-
3
B refers an underdeveloped embryo, C to deep physio- er, 1990) , latitude (Wilcox, 1968), salinity (Sands,
3
1981) and soil moisture (Youngberg, 1952) also vary
logical dormancy and superscript e to epicotyl (Niko-
laeva, 1977 ) . After surveying available evidence, in seed dormancy and germination characteristics . One
Baskin and Baskin (1998) proposed that chemical and of the challenges has been to decide if the observed
variation is due to genetics and or preconditioning .
mechanical dormancy be recognized only as a part of
physiological dormancy, and Nikolaeva (2004) agreed Thus, a number of common garden studies have been
with this conclusion .Later, Baskin and Baskin (2004) done, and genetic (Hacker and Ratcliffe, 1989), pre-
published a modified version of Nikolaeva′s system that conditioning environmental (Nelson et al., 1970) and
included three hierarchical layers of classification: genetic x environmental (Quinn and Colosi, 1977) ef-
class, level and type . In the modified version, Niko- fects have been found .
laeva′s physical, physiological, morphological, mor- Phylogenetic relationships
phophysiological and one type of combinational (physi- Martin′s (1946 ) family tree of seed phylogeny
cal plus physiological ) dormancy are “ classes of
and the accompanying body of work on embryo mor-
dormancy”(Baskin and Baskin, 2004) .
phology of 1287 genera of plants was an important start-
Variation in dormancy and germination ing point for considerations of the phylogenetic relation-
Within-in species variation in seed dormancy and ships of seed dormancy . He placed seeds with small,
germination characteristics has great economic as well underdeveloped embryos ( i. e ., they must grow inside
as ecological importance .Thus, it is not surprising that the seed before germination can occur) at the base of
inheritance of seed dormancy has been under almost the tree . Subsequent work in Russia by Grushvitzky
continuous investigation since the early 1900s, e. g ., (1967), analyses of taxonomic position [using Takhta-
studies on peanut by Stokes and Hull (1930) . Fur- jan′s (1986, 1987) plant taxonomic classification sys-
ther, various aspects of seed dormancy and germina- tem] and seed characters by Nikolaeva (1999) and an
tion, including degree of dormancy (Adkins et al., evaluation of evolution of embryo size in seed plants by
1986), rate of dormancy loss (Jana et al., 1979) and Forbis et al. (2002) supported the idea that underde-
germination requirements (Whittington et al., 1970), veloped embryos are primitive . However, it is well rec-
are at least partly controlled by genetics . Seed poly- ognized that underdeveloped embryos also are present in
morphism has a genetic component (Antonovics and seeds of some advanced groups of plants (Grushvitzky,
Schmitt, 1986), and the consequences of seed size on 1967; Nikolaeva, 1999 ) . The first attempt to map
germination (Harper and Obeid, 1967) and on subse- kinds of dormancy onto phylogenetic charts was by
quent size of seedlings (Wulff and Bazzaz, 1992) have Baskin and Baskin ( 1998 ), who placed classes of
been studied in a number of species . dormancyonto Takhtajan′s (1980) diagram for showing
Another source of variation in seed dormancy and the evolutionary relationships within the angiosperms .
germination characteristics is the environment experi-
enced by the mother plant while seeds are maturing . Recent (1980 - 2008) and on-going research
This type of variation is called preconditioning or ma- Research on the ecology of seed dormancy and
th
ternal effects, and some of the many factors causing it germination in the last part of the 20 century and be-
st
include age of plant (Olson, 1932), day length (Gut- ginning of the 21 century has added significantly to the
2 8 2 云 南 植 物 研 究 30卷
strong early foundation of knowledge . In some cases, Myrtaceae, Rhizophoraceae, Rubiaceae, Rutaceae,
work on previously-studied topics such as classes of Symplocaceae and Verbenaceae that did not germinate
dormancy and soil seed banks has continued, but some until after 90 days in Ng′s (1991, 1992) germination
new fields of investigation such as the world biogeogra- phenology studies of Malaysian forest trees also have
phy of seed dormancy and evolution of seed dormancy deep PD .
in closely-related taxa has emerged . We will briefly Physical dormancy (PY) . Since 1980, several
survey achievements in these various research areas . significant advances have been made with regard to
Classes of seed dormancy seeds with PY . First, the list of families with one or
more species with water-impermeable seed or fruit coats
Physiological dormancy (PD) . For seeds with
has been refined and now includes Anacardiaceae,
nondeep PD, it has been known for many years that,
Bixaceae, Cannaceae, Cistaceae, Cochlospermaceae,
depending on the species, cold stratification ( Carpita
Convolvulaceae (including Cuscutaceae) , Curcurbita-
et al., 1983), GA (Baskin and Baskin 1971), in-
3
ceae ( Sicyos), Dipterocarpaceae ( subfamilies Mon-
cubation temperatures (Junttila, 1973), light (Scheibe
toideae and Pakaraimoideae, but not Diptero-
and Lang, 1965) or darkness (Chen, 1968) may in-
carpoideae), Fabaceae (subfamilies Caesalpinioideae,
crease growth potential of the embryo enough for the
Mimosoideae and Papilionoideae) , Geraniaceae, Mal-
radicle to emerge through the seed coat, and any other
vaceae (including Bombacacaceae, Sterculiaceae, and
covering layer ( s) that might be present, and thus
Tiliaceae, sensu APG, 2003), Nelumbonaceae, Rham-
germinate . However, for endospermous seeds it has
naceae, Sapindaceae, Sarcolaenaceae (Baskin et al.,
been found that the endosperm becomes less restrictive
2000) and Surianaceae (Baskin et al., 2006) . Also,
for growth, just prior to emergence of the radicle .
the specialized structure for entry of water (water gap)
Breakdown of the endosperm cap over the tip of the
has been identified in five additional families: Anacar-
radicle is correlated with production of GA (Groot and
diaceae, carpellary micropyle (Li et al., 1999); Can-
Karssen, 1987) and with activities of various enzymes
naceae, imbibition lid (Grootjen and Bouman, 1988);
( e. g ., Toorop et al., 2000) and expansins (Chen et
Cistaceae, bixoid chalazal plug ( Thanos and
al., 2001) . This is an on-going area of research, and
Georghiou, 1988; Nandi, 1998 ) ; Convolvulaceae,
the genes responsible for the various aspects of endo-
bulges near hilum in most genera (Jayasuriya et al.,
sperm cap weakening are being determined (Mella et
al., 2004) . 2007) but hilum in Cuscuta (Jayasuriya et al., un-
In temperate regions, seeds with deep PD require publ .) and Dipterocarpaceae, bixoid chalazal plug
several weeks to a few months of cold stratification to (Nandi, 1998) . Although fungi and or soil abrasion
become nondormant (ND) , at which time seeds usually have received some attention as the agents that made
germinate best at low temperatures; sometimes germi- seeds water permeable; still, little, or no, experimen-
nating at the cold-stratifying temperature (Baskin and tal evidence has been produced to support these ideas
Baskin, 1998) .Deep PD that is broken only by sever- (Baskin and Baskin, 2000) . However, there is an in-
al months of warm (25 15℃) stratification has been creasing body of data indicating that the water gap
found in seeds (actually drupelets) of Leptecophylla opens in response to changes in environmental condi-
tameiameiae ( Ericaceae) from the tropical montane tions, in particular temperature (see below); thus,
zone of Hawaii (Baskin et al., 2005) . After dormancy these structures serve as environmental sensors .
break, 25 15℃ was the optimum temperature regime One of the most exciting discoveries about seeds
for germination . Future research may prove that mem- with PY is that, at least in some species, dormancy-
bers of the Burseraceae, Clusiaceae, Combretaceae, break in nature involves two steps: (1) seeds are made
Euphorbiaceae, Fagaceae, Flacourtiaceae, Hernandi- sensitive but remain impermeable, and (2) sensitive
aceae, Lecythidaceae, Meliaceae, Menispermaceae, seeds respond to certain environmental conditions and
3 期 BASKIN and BASKIN: Advances in Understanding Seed Dormancy at the Whole-seed Level: An ... 2 8 3
thus become water-permeable and germinate . Depend- Hawaiian lobelioid shrubs (Baskin et al., 2005 ) .
ing on the species, seeds may become sensitive if However, although seeds of Drosera anglica, which
stored moist (or at high RH) at high, e. g ., Trifoli- Martin considered to be dwarf, have small embryos in
um subterraneum (Taylor, 1981, 2005 ), Ornithopus relation to size of endosperm, embryos did not grow
compressus (Taylor and Revell, 1999) and Ipomoea la- prior to radicle emergence ( Baskin and Baskin,
cunosa (Jayasuriya et al., 2008), or at low, e. g ., 2005) .
Melilotus albus, Medicago lupulina, Lotus cornicula- With regard to Martin′s (1946) classification sys-
tus and Trifolium repens (Van Assche et al., 2003), tem of seeds, there are 10 types based on embryo and
temperatures . Although seeds are exposed to the first endosperm characteristics and two additional types
step and become sensitive to dormancy break, they do based on seed size: dwarf, 0.3 - 2.0 mm and micro,
not germinate unless step 2 occurs . In fact, seeds of ≤0.2 mm . However, his“dwarf seed families”could
I. lacunosa can cycle between sensitive and insensitive not be distinguished from his other small-seeded non-
states .Thus, if sensitive seeds of I. lacunosa are dried dwarf families on the basis of endosperm texture, seed
at high temperatures (instead of being on moist sand) coat anatomy, embryo morphology, class of dormancy
they become insensitive and will not respond to the or phylogenetic position . Consequently, Martin′s key
treatment (35℃ on wet sand for ≥2 h) that breaks has been revised to place all seeds into categories based
dormancy, i. e ., causes water gap to open (Jayasuriya on embryo and endosperm characteristics (Baskin and
et al., 2008) . Seeds made insensitive can again be Baskin, 2007) .
made sensitive . Seed banks
Morphophysiological dormancy ( MPD ) . To
Worldwide, the one topic related to the ecology of
break MPD, seeds must be exposed to appropriate en-
seed dormancy and germination that has received the
vironmental conditions to break PD and to promote
most attention from 1980 to present is seed banks .
growth of the embryo ( break MD) . Seven levels of
Much of the work is related to soil seed banks, but
MPD were distinguished by Nikolaeva (1977 ), and
considerable effort has been devoted to aerial seed
they can be divided into two general categories: sim-
banks, especially serotiny (Cowling et al., 1987;
ple, embryo grows at relatively high ( > 10℃, i. e . Lamont et al., 1991) . The suggested minimum period
temperatures too high for cold stratification); and com- of time for seeds to live in soil and be called a persis-
plex, embryo grows during cold stratification . Nondeep tent seed bank has been expanded from ≥ 1 year
simple MPD was not mentioned by Nikolaeva (1977), (Thompson and Grime, 1979) to the second germina-
and it was first described in seeds of the winter annual tion season ( Walck et al., 2005 ) to accommodate
Chaerophyllum tainturieri ( Baskin and Baskin, seeds that are in the dormant phase of the annual
1990) .Thus, eight levels of MPD now are known . In dormancy nondormancy cycle at the end of 1 year and
seeds of C. tainturieri, PD is broken during summer, those that require more than 1 year to become nondor-
and embryo growth occurs in autumn if seeds are ex- mant . It has been made emphatically clear that seeds
posed to light . In some seeds with nondeep simple do not have to be dormant to remain ungerminated and
MPD, however, PD is broken during cold stratifica- viable in the soil for long periods of time (Thompson et
tion, and embryogrowth and germination occur in spri- al., 2003; Fenner and Thompson, 2005) .
ng (Walck et al., 1999) . Studies to determine the species represented in
Another interesting discovery related to MPD is persistent seed banks and number of seeds of these spe-
that some of Martin′s (1946)“dwarf seeds”, e. g ., cies have been (and continue to be) done in various
Campanulaceae and Gentianaceae (Baskin and Baskin, habitats around the globe . Many aspects of soil seed
2005), have small underdeveloped embryos .Underde- banks are being studied, and all we can do in a brief
veloped embryos also occurred in seeds of six endemic review is to list some of them: (1) shapes and sizes of
2 8 4 云 南 植 物 研 究 30卷
seeds with a high probably of forming a persistent seed break dormancy in seeds of winter annuals (Baskin and
bank (Thompson et al., 1993; Funes et al., 1999); Baskin, 1998) . Further, as many seeds with PD come
(2) reasons for spatial heterogeneity (Fernandez et out of dormancy the range of temperatures over which
al., 2002); (3) causes of seed death including preda- they germinate increases, and as they enter dormancy
tion ( Honek et al., 2003), solarization (Mas and the range of temperatures over which theygerminate de-
Verdu, 2002 ) and flooding (Voesenek and Blom, creases; this is called the dormancy continuum (Baskin
1992); (4 ) comparison of seed rain and seed bank and Baskin, 1985b) . Models of the seasonal dynamics
(Pakeman and Small, 2005) ; (5) dynamics of buried of buried seeds include seasonal changes in temperature
seeds of individual species ( Wilson and Witkowski, range for germination (Vleeshouwers and Bouwmeester,
2003; Shimono and Washitani, 2004); (6) longevity 2001), as well as changes in sensitivity to light (Batlla
of seeds in soil (Bekker et al., 1998; Arroyo et al., and Benech-Arnold, 2005 ) and hydrothermal time
2004) ; (7) maximum number of seeds in arable soils (Ekeleme et al., 2005) . Seeds also change in sensi-
(D′Angela et al., 1988) ; ( 8) maximum depth of tivity to other factors, e. g ., plant growth regulators
seedling emergence (Benvenuti et al., 2001; Grundy and substrate moisture during the dormancy cycle
et al., 2003); (9) weed seed emergence models (Col- (Baskin and Baskin, 1998) .
bach et al., 2005; Wang et al., 2006); (10) seed
Germination requirements of nondormant seeds
bank management to facilitate restoration of plant com-
After seeds, especially those with PD, become
munities (Hoelzel and Otte, 2003; Mengistu et al.,
nondormant various environmental factors may play a
2005); and (11) potential of seed bank for restora-
role in determining if they germinate (Vleeshouwers et
tion, including good species representation (Kalamees
al., 1995) . It has long been known that seeds may
and Zobel, 2002; Sakai et al., 2005), some species
require light for germination, and this is an important
not in seed bank (Senbeta and Teketay, 2002; Ro-
reason why nondormant seeds in the soil do not germi-
drigues and Matos, 2006), weed seeds present ( Bid-
nate (Fenner and Thompson, 2005) . For nondormant
well et al., 2006; Roovers et al., 2006) , seeds of
seeds on the soil surface, there are various factors that
aliens present (Shen et al., 2004; Yakimowski et al.,
may play a role in controlling germination, i. e ., pro-
2005) and seeds of rare species present (Auld and
vide a signal that conditions are suitable for seedling
Scott, 2004) .
establishment . For example, light filtered through
Dormancy cycling
green leaves (Teketay, 1998) or constant temperatures
Annual dormancy nondormancy or conditional (Pons and Schroder, 1986) may reduce or prevent ger-
dormancy nondormancy cycles have been documented mination of nondormant seeds, depending on the spe-
in buried seeds of various winter annuals, summer an- cies . For some species, there are chemical cues from
nuals and perennials exposed to the temperate-zone an- the environment that promote germination of nondor-
nual temperature cycle (Khan and Karssen, 1980; mant seeds . One example is seeds of Schoenoplectus
Baskin and Baskin 1998) . However, some seeds come hallii which germinate in shallow pools of water in spri-
out of dormancy and remain nondormant during burial ng . Seeds must be cold stratified (under moist but not
for many years (Baskin and Baskin, 1985a) . Under- flooded conditions) to break PD, after which they
standing why buried seeds can undergo changes in their require flooding, exposure to high (30 15℃) tempera-
dormancy state has been enhanced by experimental data tures, light and ethylene to germinate (Baskin et al.,
showing that lowwinter temperatures break PD in seeds 2003) .
of many summer annuals and can induce nondormant The many observations of appearance of thousands
seeds of winter annuals into dormancy . On the other of seedlings following fire, especially in matorral vege-
hand, high summer temperatures can induce nondor- tation, have prompted research to determine the effects
mant seeds of summer annuals into dormancy and can of heat and chemicals from fire on germination . de-
3 期 BASKIN and BASKIN: Advances in Understanding Seed Dormancy at the Whole-seed Level: An ... 2 8 5
Lange and Boucher (1990) were the first to find that tain germination inhibitors ( Traveset and Verdu,
smoke from burning plant material would promote germ- 2002) .
ination, and subsequently smoke and smoke-water have
Desiccation sensitive seeds
been used to promote seed germination of many species
All desiccation sensitive (recalcitrant) seeds have
(Stevens et al., 2007) . However, smoke or smoke-
a high moisture content (30 - 70%) at time of dispers-
water often is more effective in promoting germination
al, and they die if moisture content declines to < 15 -
after seeds have been given a treatment to break PD,
40% , depending on the species . Also, seeds are sen-
e. g ., burial in soil and thus exposure to summer field
sitive to low temperatures and die if frozen . In cont-
temperatures (Baker et al., 2005) , indicating that at
rast, orthodox seeds have a moisture content of 15 -
least in some species smoke promotes germination of
20% at maturity, can be dried to 6 - 12% and most
nondormant seeds and is a cue that the area has been
tolerate freezing (Baskin and Baskin, 1998) .Desicca-
burned .The compound in smoke that promotes germi-
tion sensitive seeds are more likely to be found in spe-
nation has been identified ( Flematti et al., 2004),
cies living in moist, aseasonal habitats than in those
and it is a butenolide which more specifically is known
living in arid, seasonal habitats, but some arid-zone
as karrikinolide (Flematti et al., 2007) .
species do have desiccation sensitive seeds (Tweddle et
Effects of animals
al., 2003) . Overall, seeds of only a relatively small
In addition to studies on seed predation and seed proportion ( ca .10% ) of seed plants are recalcitrant .
dispersal by animals, much attention has been given to Somewhat surprising is that more than 90% of aquatic
effects of animals on germination of seeds that have and marsh species have orthodox seeds ( Dickie and
been regurgitated or passed through the digestive sys- Pritchard, 2002) . However, due to the big problem of
tem . The assumption often is made that seeds with PY trying to store recalcitrant seeds of economically-valu-
must pass through the digestive system of an animal to able species, especially trees, much ecological, bio-
germinate in nature . However, careful studies show chemical and ultrastructure research has been done on
that depending on plant species and on kind of animal, them in the last 30 years .
seeds with PY may (1) be digested, i. e ., destroyed Metabolism is continuous in recalcitrant seeds,
(Smit and Rethman, 1996), (2) become permeable which helps to explain why they need to be continuous-
and germinate in the digestive system and thus die ly hydrated to prevent DNA damage and possible death
(Gardener et al., 1993), (3) germinate to higher per- (Boubriak et al., 2000 ) . In fact, some recalcitrant
centages after they have been defecated than the non- seeds are dormant when they are dispersed and thus
ingested controls (Shiferaw et al., 2004) and (4) ex- must be kept moist during the time required for dorman-
hibit no increase in germination after defecation (Orte- cy break and for germination (Flores, 1996) . In Pana-
ga-Baes et al., 2001) . The story for water-permeable ma, desiccation-tolerant seeds are more likely to be dis-
seeds that are eaten by animals is about the same .That persed in the dry season, and desiccation sensitive seeds
is, passage through the digestive system may increase are more likely to be dispersed in the wet season (Daws
( Serio-Silva and Rico-Gray, 2002 ), decrease et al., 2005) . Attempts to store recalcitrant seeds for
( Charalambidou et al., 2005 ) or have no effect extended periods of time at high moisture levels have not
(Figueroa and Castro, 2002) on germination percent- been very successful due to high susceptibility of seeds
ages, compared to those of control seeds . For many to attack by fungi (Berjak et al., 2004), ineffective mi-
species, however, removal of fruit material from around crobial defense systems in the seeds (Anguelova-Merhar
seeds significantly increases germination ( Burrows, et al., 2003) and high possibility of seeds germinating
1993) .Thus, in addition to seed dispersal, one of the (Corbineau and Come, 1986) . At present, cryopreser-
big effects of animals, especially birds, on seed germi- vation ( < -130℃) of seeds and zygotic embryos of re-
nation is that they remove fruit materials that may con- calcitrant species is receiving much attention . Methods
2 8 6 云 南 植 物 研 究 30卷
to dehydrate tissues to the glassy state using vitrification sion (with respect to germination) have been found
solutions (Fahy et al., 2004) or to rapidly super-cool (Galloway and Etterson, 2007), often there is no evi-
them (Wesley-Smith et al., 2004) are being developed dence of it (Molina-Freaner et al., 2003; Pico and
as ways to safely cryopreserve recalcitrant seeds . Koubek, 2003) .
An important discovery with regard to seed storage Research on seed polymorphism continues and
is that some seeds are intermediate between orthodox studies can be divided into two general categories: het-
and recalcitrant in terms of tolerance to low moisture erocarpy and size mass . The dimorphic seeds of Aster-
levels and low temperatures (Ellis et al., 1991) . A aceae ( Gibson and Tomlinson, 2002; El-Keblawy,
number of Citrus (Hong and Ellis, 1995) and Coffea 2003; Brandel, 2007) and Chenopodiaceae (Khan et
(Dussert et al., 1999) species have seeds with inter- al., 2004; Mandak and Pysek, 2005) and to a certain
mediate storage behavior . Berjak and Pammenter extent Brassicaceae (Cordazzo, 2006) and a few other
(2001) emphasize that there is a continuum of sensitiv- families ( Bandera and Traveset, 2006 ) have been
ities to dehydration between recalcitrant and orthodox studied .The focus of these studies has been to evaluate
seeds, and even within seeds with intermediate storage the effects of polymorphism on germination, as well as
behavior there is a continuum . the adaptive significance and fitness .
Seed size mass may have effects not only on germ-
Genetics
ination but also on early growth and successful (or not)
The inheritance of seed dormancy has received
establishment of seedlings (Khurana and Singh, 2004;
much attention, especially with regard to rice ( Oryza
Moles and Westoby, 2004) and on evolutionary stable
sativa) (Oard et al., 2000; Gu et al., 2004) but also
strategies (Rees and Venable, 2007) . The interest in
many other species (Leon et al., 2006a) . Quantitative
seed size continues to increase, and now divergence
trait loci are the subject of intense study in various spe-
analysis, using information for thousands of species, is
cies (Gu et al., 2006; Bettey et al., 2000), while
being used to ask what factors are most closely associat-
genes that control various aspects of seed dormancy and
ed with changes in seed mass . This analysis showed
germination (Fukuhara and Bohnert, 2000), especially
that growth form and seed mass are closely related
for Arabidopsis thaliana ( Chiwocha et al., 2005;
(Moles et al., 2005a, b), and there is a close corre-
Gonzalez et al., 2004; Salaita et al., 2005; Cadman
lation between seed mass and size of the genome
et al., 2006; Finch-Savage et al., 2007) are being
(Beauleu et al., 2007) . Divergence in seed mass also
identified in other species . On the other hand, some
is related to other factors, such as temperature, precip-
work is being done on the effects of the mother plant
itation and latitude (Moles et al., 2005b) . On a glo-
( preconditioning ) on dormancy and germination of
bal scale, seed mass generally decreases with latitude,
seeds, with particular focus on effects of precondition-
but at 20 - 25°(at the edge of the tropics) there is a 7-
ing on characteristics of the resulting seeds such as col-
fold decrease in mass, seemingly correlated with differ-
or (Luzuriaga et al., 2006), mass (Luzuriaga et al.,
ences in vegetation type and growth form of species
2006; McPeek and Wang, 2007), thickness of fruit
(Moles et al., 2007) .
seed (Qaderi et al., 2003) and concentration of UV-
World biogeography of seed dormancy
absorbing compounds (Griffen et al., 2004) .
Inbreeding depression in plants is a topic of much A vast amount of information is available on pres-
interest, and several hundred papers have been pub- ence vs . absence of dormancy and on class of seed
lished on the subject in the last 30 years . From a seed- dormancy of plant species from throughout the world .
germination perspective, if seeds resulting from selfed Further, new data from each of the major vegetation
flowers have lower germination than those from crossed zones recognized by Walter (1979) continue to be pub-
flowers this is interpreted to be evidence of inbreeding lished: evergreen rainforests ( Ferraz and Varla,
depression . Although many cases of inbreeding depres- 2003 ), semievergreen rainforests ( Sautu et al.,
3 期 BASKIN and BASKIN: Advances in Understanding Seed Dormancy at the Whole-seed Level: An ... 2 8 7
2007) , tropical montane (Tigabu and Oden, 2001), means the opportunities to do seed dormancy studies in
tropical deciduous forests ( Zamith and Scarano, a phylogenetic context are increasing . That is, germi-
2004), savannas (Danthu et al., 2003), hot deserts nation studies done using members of a lineage could
(Flores et al., 2006 ), matorral (Cochrane et al., provide new insight into the evolutionary relationships
2002), broadleaved evergreen forests ( Chien et al., of classes of dormancy (or nondormancy) . Although
2006), deciduous forests (Kondo and Sato, 2007), much research currently is being done on estimation of
steppes (Wesche et al., 2006 ), cold deserts ( Ren divergence times and biogeography of various taxonomic
and Tao, 2004 ), boreal forests ( Rosner and Har- groups (Xiang et al., 2004; Nie et al., 2005), infor-
rington, 2004), tundra (Cummins and Miller, 2000) mation on phylogeny has not yet been widely used to
and mountains (Phartyal et al., 2003) . Thus, efforts help gain an understanding of evolutionary relationships
to compile these data and to begin to study general pat- of seed dormancy . In one study, using four species in
terns of the biogeography of dormancy are underway a subclade of Aristolochia subgenus Siphisia, it was
(Baskin and Baskin, 1998, unpubl .) . found that both trait stasis and divergence (adaptation)
Data on dormancy vs . nondormancy, and if have occurred in dormancy-breaking and germination
dormant the class of dormancy, for 7351 species have requirements in this lineage (Adams et al., 2005) .
been organized by vegetation zone . For all vegetation Further insights can be gained by combining informa-
zones except tropical rainforests and semievergreen tion on phylogeny, classes of dormancy (or nondorman-
rainforests, more species have dormant than nondor- cy) and fossil history . For example, in the Dipsacales,
mant seeds . In rainforests, 52% and 48% of the spe- all the clades except the most advanced Valerina clade
cies have nondormant and dormant seeds, respectively, have underdeveloped embryos and thus MD or MPD .
and in semievergreen rainforests 50% and 50% of the Members of the Valerina clade have seeds with fully
species have nondormant and dormant seeds, respec- developed embryos and thus either PD or are nondor-
tively (Baskin and Baskin, unpubl .) . Generally, in mant . The fossil record suggests that MPD or MD was
tropical subtropical areas the portion of species with present in Dipsacales earlier than PD or nondormancy
dormant seeds increases across the gradient of vegeta- (Baskin et al., 2006) .
tion zones as temperature and precipitation decrease .In
temperate arctic areas, the portion of species with
The future
dormant seeds increases across the gradient of vegeta-
It is clear that research on the ecology of seed
tion zones as temperature and precipitation decrease
dormancy and germination already spans a broad spec-
until one comes to the relatively cool vegetation zones
trum of topics, but this spectrum will, no doubt, con-
(woodland, montane, boreal and tundra), where the
tinue to expand . Some topics, e. g ., the genetic as-
proportion of species with dormant seeds decreases
pects of the ecology of seed dormancy and germination,
(Baskin and Baskin, 1998, unpubl .) . PD ( 29.8 -
already shows signs of much diversification, including
79.7% of the species) is the most important class of
effects of breeding system (Verdu et al., 2004), pop-
dormancy in all vegetation zones on earth, with PY
ulation size (Paschke et al., 2002) , habitat fragment-
(3.2 - 30.0% ) and MPD (0 - 16.3% ) being second
ation ( Cascante et al., 2002 ), herbicide resistance
and third in importance, respectively . MPD ( 0 -
( Recasens et al., 2007 ) and phenotypic plasticity
3.1% ) and (PY + PD) (0 - 2.1%) are not very im-
(Donohue, 2005) on germination and fitness; pro-
portant in any vegetation zone .
teomics (Leon et al., 2005b); and spread of trans-
Phylogeny
genes into wild populations ( Garnier and Lecomte,
The on-going process of updating and revising the 2006) ; invasive vs . non-invasive species (Mandak,
phylogeny of angiosperms (APG, 2003), as well as 2003); and rare vs . common species (Osunkoya and
improving phylogenies of various orders and families Swanborough, 2001) .
2 8 8 云 南 植 物 研 究 30卷
We foresee that the determination of class of response to smoke water or heat [J] . Sci Sci Res, 15: 339—348
Bandera MC, Traveset A, 2006 . Reproductive ecology of Thymelaea ve-
dormancy (or nondormancy) , as well as the dormancy-
lutina (Thymelaeaceae )-factors contributing to the maintenance of
breaking and germination requirements of seeds from all
heterocarpy [J] . Plant Syst Evol, 256: 97—112
over the world will continue at a rapid rate .The stimu-
Barton LV, 1930 . Hastening the germination of some coniferous seeds
lus for this work comes from the desire of many scienti- [J] . Amer J Bot, 17: 88—115
sts to know how things work in nature and from the Baskin CC, Baskin JM, 1998 .Seeds: Ecology, biogeography, and evo-
lution of dormancy and germination [M] . San Diego: Academic
need for information on how to propagate species for
Press
economic and conservation reasons . As information on
Baskin CC, Baskin JM, 2005 . Underdeveloped embryos in dwarf seeds
(1 ) seed dormancy class ( or nondormancy ) and
and implications for assignment todormancy class [J] . Seed Sci Res,
dormancy-breaking and germination requirements in-
15: 357—360
creases for plant taxa from all vegetation zones of the Baskin CC, Baskin JM, 2007 . A revision of Martin′s seed classification
world, and (2) phylogenetic relationships within orders system, with particular reference to his dwarf-seed type [J] . Seed
Sci Res, 17: 11—20
and families are increasingly refined (using both DNA
Baskin CC, Baskin JM, Chester EW et al., 2003 .Ethylene asa possible
and fossil evidence), good tools with which to under-
cue for seed germination of Schoenoplectus hallii (Cyperaceae) , a
take studies of evolutionary relationships of seed
rare summer annual of occasionally flooded sites [J] . Amer J Bot,
dormancy and germination within plant lineages will be 90: 620—627
available . Thus, major advancements in our knowledge Baskin CC, Baskin JM, Yoshinaga A, 2005 . Morphophysiological
dormancy in seeds of six endemic lobelioid shrubs (Campanulaceae)
of the world biogeography of seed dormancy and of evo-
from the montane zone in Hawaii [J] . Can J Bot, 83: 1630—1637
lutionary origins and relationships of the various classes
Baskin CC, Baskin JM, Yoshinaga A et al., 2005 . Germination of
of dormancy (and nondormancy) can be expected .
drupelets in multi-seeded drupes of the shrub Leptecophylla tame-
iameiae ( Ericaceae) from Hawaii: A case for deep physiological
References: dormancy broken by high temperatures [J] . Seed Sci Res, 15:
349—356
Adams CA, Baskin JM, Baskin CC, 2005 .Trait stasis versus adaptation ?
Baskin JM, Baskin CC, 1971 . Effect of chilling and gibberellic acid on
in disjunct relict species: Evolutionary changes in seed dormancy-
growth potential of excised embryos of Ruellia humilis [J] . Planta,
breaking and germination requirements in a subclade of Aristolochia
100: 365—369
subgenus Siphisia (Piperales) [J] . Seed Sci Res, 15: 161—173
Adkins SW, Loewen M, Symons SJ, 1986 .Variation within pure lines of Baskin JM, Baskin CC, 1985a .Does seed dormancyplay a role ingermi- #
wild oats ( Avena fatua) in relation to degree of primary dormancy nation ecologyof Rumex crispus? [J] . Weed Sci, 33: 340—343
[J] . Weed Sci, 34: 859—864 Baskin JM, Baskin CC, 1985b . The annual dormancy cycle in buried
AngiospermPhylogeny Group (APG), 2003 . An update of the Angio- weed seeds: A continuum [J] . BioScience, 35: 492—498
sperm Phylogeny Group classification for the orders and families of Baskin JM, Baskin CC, 1990 .Germination ecophysiologyof seeds of the
flowering plants: APG II [J] . Bot J Linn Soc, 141: 399—436 winter annual Chaerophyllum tainturieri: Anewtype of morphophysi-
Anguelova-Merhar VS, Calistru C, Berjak P, 2003 .A studyofsome bio- ological dormancy [J] . J Ecol, 78: 993—1004 /
chemical and histopathological responses of wet-stored recalcitrant Baskin JM, Baskin CC, 2000 . Evolutionary considerations of claims for
seeds of Avicennia marina infected by Fusarium moniliforme [J] . physical dormancy-break bymicrobial action and abrasionby soil par-
Ann Bot, 92: 401—408 ticles [J] . Seed Sci Res, 10: 409—413
Antonovics J, Schmitt J, 1986 . Paternal and maternal effects on Baskin JM, Baskin CC, 2004 . A classification system for seed dormancy
propagule size in Anthoxanthum odoratum [ J] . Oecologia, 69: [J] . Seed Sci Res, 14: 1—16
277—282 Baskin JM, Baskin CC, Dixon KW, 2006 .Physical dormancy in the en-
Arroyo MTK, Cavieres LA, Humana AM, 2004 . Experimental evidence demic Australiangenus Stylobasium, a firstreport for the familySuri-
of potential for persistent seed bank formation at a subantarctic alpine anaceae (Fabales) [J] . Seed Sci Res, 16: 229—232
site in Tierra del Fuego, Chile [J] . Ann Missouri Bot Gard, 91: Baskin JM, Baskin CC, Li X, 2000 . Taxonomy, anatomy and evolution
357—365 of physical dormancy in seeds [J] . Plant Species Biol, 15: 139—
AuldTD, Scott J, 2004 . Estimating population abundance in plant spe- 152 7
cies with dormant life-stages: Fire and the endangered plant Gre- Baskin JM, Hidayati SN, Baskin CC et al., 2006 . Evolutionary consid-
villea caleyi R .Br [J] . Ecol Manage Restor, 5: 125—129 erations of the presence of both morphological and physiological seed
Baker KS, SteadmanKJ, Plummer JA et al., 2005 .Dormancy release in dormancy in the highly advanced euasterids II order Dipsacales [J] . '
Australian fire ephemeral seeds during burial increases germination Seed Sci Res, 16: 233—242
