This article comments on:Fernández Farnocchia RB, Benech-Arnold RL, Mantese A, Batlla D. 2021. Optimization of timing of next-generation emergence in Amaranthus Weed Seed Dormancy In this unit we will learn about seed dormancy as seen by many respected seed biologists, the way most scientists in the seed discipline view this complex area of plant
Seed dormancy and weed emergence: from simulating environmental change to understanding trait plasticity, adaptive evolution, and population fitness
Kazumi Nakabayashi, Gerhard Leubner-Metzger, Seed dormancy and weed emergence: from simulating environmental change to understanding trait plasticity, adaptive evolution, and population fitness, Journal of Experimental Botany, Volume 72, Issue 12, 28 May 2021, Pages 4181–4185, https://doi.org/10.1093/jxb/erab150
Predicting weed emergence in crop production systems is a global challenge that requires understanding mechanisms of weed ecology and trait evolution in response to climate change and altered agricultural practices. Seed dormancy is a highly adaptive trait that controls this by defining the environmental conditions in which the seed is able to germinate ( Finch-Savage and Leubner-Metzger, 2006). Weed soil seed bank persistence and the timing of seedling emergence depend on dormancy ( Baskin and Baskin, 2006; Walck et al., 2011; Batlla et al., 2020). Integrating mechanisms of seed dormancy dynamics in variable field environments and across generations with population-based models and realistic ecophysiological simulations ( Fernández Farnocchia et al., 2021) are essential for more sustainable weed management strategies.
Charles Darwin wrote in his letter to Joseph Hooker (12 April 1857) ‘I have been interested in my ‘weed garden’ of 3×2 feet square: I mark each seedling as it appears, and I am astonished at number that come up.’ The timing of weed seedling emergence, often as seasonal flushes, has critical and agronomical implications as weeds produce the highest potential yield loss (30–40%) in the major crop production systems ( Oerke, 2006). Weed soil seed bank dynamics depend on seed dormancy, a trait with high plasticity in weed species and with enormous adaptive value to adjust the population to a cropping system ( Baskin and Baskin, 2006; Westwood et al., 2018; Schwartz-Lazaro and Copes, 2019; Batlla et al., 2020). The control of germination timing is achieved by seed dormancy, which can be considered as block(s) to the completion of germination of an intact viable seed under otherwise favourable conditions, namely after the seed becomes non-dormant ( Finch-Savage and Leubner-Metzger, 2006). Primary dormancy is established during seed maturation prior to dispersal, whereas secondary dormancy refers to the acquisition of dormancy in a mature seed after dispersal and after the loss of primary dormancy ( Graeber et al., 2012; Finch-Savage and Footitt, 2017; Penfield and MacGregor, 2017). The molecular mechanisms underpinning the seasonal seed dormancy cycling to time germination in variable field environments have been investigated with Arabidopsis thaliana ecotypes adapted to different climates ( Finch-Savage and Footitt, 2017). Seeds continually adjust their dormancy status by sensing a range of environmental signals. Temperature is related to slow seasonal change and used for temporal sensing to determine the time of year and adjust the depth of dormancy accordingly. This alters seed sensitivity to signals related to the spatial environment, including light and soil moisture. The sensing of these signals is more ultimate as they indicate if conditions are suitable for germination and therefore trigger dormancy release. Molecular mechanisms and large-scale molecular datasets of A. thaliana seed dormancy states (see references in Finch-Savage and Footitt, 2017) and of weed trait plasticity ( Maroli et al., 2018) require integration by using threshold population-based models and realistic ecophysiological simulations ( Batlla and Benech-Arnold, 2014; Donohue et al., 2015; Finch-Savage and Footitt, 2017). These models and simulations provide an ecophysiological framework and are especially complex if seed dormancy regulation is investigated across weed generations to capture how it maximizes weed population fitness.
Using the summer annual weed Amaranthus hybridus (smooth pigweed), Fernández Farnocchia et al. (2021) provide a sophisticated and well-integrated analysis of how the primary dormancy level at dispersal established during maturation in different maternal environments synchronizes next-generation seedling emergence timing to maximize weed population fitness. Nine Amaranthus (pigweed) species, including A. hybridus, A. retroflexus, and A. palmeri, are listed as invasive or noxious weeds, with Palmer amaranth being the most troublesome herbicide-resistant weed in south-eastern USA ( Trucco et al., 2009; Ward et al., 2013; Assad et al., 2017). Fernández Farnocchia et al. (2021) found that primary seed dormancy depth was lower when harvested from late season mother plants where seed maturation occurred in a short photoperiod maternal environment ( Box 1). However, these observed variations in dormancy depth in the laboratory experiments did not affect seedling emergence timing in the field experiments. To interpret these results, Fernández Farnocchia et al. (2021) developed threshold population-based models and performed realistic simulations which generated a better ecophysiological framework for predicting seedling emergence patterns under natural conditions. Their major conclusion is that it is crucial to consider the effects of distinct maternal environments leading to variations in the depth of primary dormancy for correctly predicting weed soil seed bank dynamics, and how these contribute to the synchronization of next-generation emergence timing to maximize population fitness. Other well-investigated examples where regulation of seed dormancy by the maternal environment, in particular photoperiod and temperature during maturation ( Box 2), was instrumental for maximizing population fitness in the field are A. thaliana ( MacGregor et al., 2015; Huang et al., 2018; Footitt et al., 2020) and the weed Polygonum aviculare ( Batlla and Benech-Arnold, 2014; Fernández Farnocchia et al., 2019; Batlla et al., 2020).
Amaranthus seed structure with peripheral embryo and perisperm, and maternal effects on seed cost thickness
The typical seed of the Amaranthaceae and of many other core Caryophyllales families evolved in the early Cretaceous and is characterized by a peripheral embryo curved around a central starchy perisperm (dead storage tissue) (Baskin and Baskin, 2019). Most Amaranthus species, including the weeds A. hybridus, A. retroflexus, and A. palmeri, and the amaranth food crops A. caudatus and A. cruentus, disperse seeds from one-seeded dehiscent fruits which open at maturity ( Irving et al., 1981; Trucco et al., 2009; Ward et al., 2013; Assad et al., 2017; Ninfali et al., 2020; Fernández Farnocchia et al., 2021). The inner seed coat consists of a sclerified parenchyma layer with osteosclereids on either side. The outer seed coat layer of palisade sclereids can vary considerably in thickness. Fernández Farnocchia et al. (2021) found that the maternal environment during seed maturation on the mother plant determined seed coat thickness and depth of primary physiological dormancy. The seed coat morphological and physicochemical properties are most important for mediating the interactions between the embryo and the ambient environment. Other core Caryophyllales species with perispermic seeds disperse one-seeded indehiscent fruits ( Sukhorukov et al., 2015) in which the fruit coat (pericarp) properties serve this role; an example for this from the Amaranthaceae family is sugar beet ( Hermann et al., 2007). The maternal environment during reproduction also affects the primary dormancy depth of the dispersed fruits of the Caryophyllales (Polygonaceae) weed P. aviculare ( Fernández Farnocchia et al., 2019), but the possible effects on pericarp properties have not been investigated. The figure shows a drawing of A. cruentus seed structure modified from Irving et al. (1981), with permission from the publisher John Wiley and Sons; seed coat thickness of A. hybridus from Fernández Farnocchia et al. (2021).
Of particular interest from a mechanistic point of view is the finding that the maternal environment, photoperiod ( Fernández Farnocchia et al., 2021) and temperature ( MacGregor et al., 2015), affected primary dormancy depths, at least in part, by altering seed coat morphological (Box 1) and physicochemical ( Box 2) properties. In species with coat-imposed dormancy, the seed and fruit coat properties are a decisive component of this trait ( Finch-Savage and Leubner-Metzger, 2006; Lepiniec et al., 2006; Steinbrecher and Leubner-Metzger, 2017; Francoz et al., 2018). In cereal grains and in A. thaliana, proanthocyanidins (tannins, brownish pigments) accumulate during seed coat development on the mother plant. The extent of this and thereby primary dormancy depths varies with temperature during seed production ( MacGregor et al., 2015), and transparent testa (tt) mutants ( Debeaujon et al., 2000) have reduced dormancy and altered permeability properties ( Box 2). The typical seed of the Amaranthaceae ( Box 1) and of many other Caryophyllales species has a peripheral embryo curved around a central starchy perisperm (dead storage tissue) evolved in the early Cretaceous ( Baskin and Baskin, 2019). Amaranthus hybridus seed coat thickness and primary dormancy depths were affected by the reproduction environment on the mother plant ( Fernández Farnocchia et al., 2021). Dormancy is, however, not the only trait affected by seed coat thickness: a comparison of several weed species demonstrated that seed mortality in the soil seed bank is related to seed coat thickness ( Gardarin et al., 2010). In this work, the estimated annual seed mortality rates in the soil seed bank and the associated seed coat thicknesses of A. hybridus and A. thaliana were very similar, ranking in the middle tier of 18 species. Seed coats are indeed more than a protective shield formed of dead cell layers ( Francoz et al., 2018). They play important roles in seed germination, dormancy, longevity, and the persistence of the soil seed bank. As maternal tissues, the interaction between the mother plant’s genotype and the maternal environment during reproduction is decisive in maximizing population fitness across generations. This knowledge is required not only for developing more sustainable weed management strategies ( Westwood et al., 2018), but also for better understanding of the underpinning mechanisms of trait plasticity and adaptive evolution upon environmental change.
Physicochemical seed coat properties determine the flux of compounds required for the control of germination
Physicochemical properties of seed and fruit coats have been shown to play important roles in the control of seed germination by providing permeability and/or mechanical restraints on germination processes ( Steinbrecher and Leubner-Metzger, 2017). The outer seed coverings consist mostly of dead tissues and represent the seed’s interphase with the external environment. In addition to providing mechanical restraint, coat-associated mechanisms control or even prevent water uptake, leaching of inhibitors for embryo elongation such as abscisic acid (ABA), or gaseous exchange which may cause oxygen deficiency within the embryo. An excellent example to illustrate this are the transparent testa (tt) mutants of Arabidopsis thaliana which exhibit lighter testa (seed coat) colour (see figure) due to defects in flavonoid metabolism and in turn reduced proanthocyanidin biosynthesis ( Lepiniec et al., 2006). During A. thaliana seed coat development, proanthocyanidins accumulate in the endothelium, the innermost cell layer of the inner integument, while the outermost cell layer of the outer integument differentiates into mucilage-producing cells; and at seed maturity the testa consists entirely of dead tissue with oxidized proanthocyanidins as brownish pigments. In many mutants, reduced pigmentation often led to thinner testa and increased permeability for hormones or other compounds (see figure), and this was associated with reduced dormancy phenotypes of the tt mutants ( Debeaujon et al., 2000). Flavonoid biosynthesis during seed coat development was shown to be higher when seeds were matured in cool conditions (see figure), which was associated with a less permeable testa and increased primary dormancy ( MacGregor et al., 2015). Furthermore, the seed coat of many tt mutants remained permeable even when matured under low temperature. These results clearly indicate that temperature regulation for increased primary dormancy involves altering testa properties by accumulation of flavonoids. Increased permeability not only permits a greater influx of water and oxygen, but also allows leaching out of endogenous compounds which are inhibitory to germination or embryo growth (see figure). Similarly to Arabidopsis where temperature has been demonstrated to be a major factor, the maternal environmental signalling and dormancy control in Amaranthus seem to be an interaction between embryo and seed coat, with photoperiod during reproduction as the major factor ( Fernández Farnocchia et al., 2021).
Weed Seed Dormancy
In this unit we will learn about seed dormancy as seen by many respected seed biologists, the way most scientists in the seed discipline view this complex area of plant biology. The concepts and organization of this unit follow this traditional way of looking at seed dormancy. It is essential that you understand these concepts, and see them from this perspective, if you are to have an understanding of weed seed biology and the scientific literature on seed dormancy.
Having said that, it is important that you realize that I am not in complete agreement with this established model of seed dormancy. Will will evaluate these different views in our classroom discussions. For starters, I view the term dormancy as the “biology of what isn’t“:
Dormancy: a state in which viable seeds, spores or buds fail to germinate under conditions favorable for germination and vegetative growth.
What this definition of dormancy obscures is what important phenomena are hidden from our view, it tells us nothing about what is happening in that seed, or its potential to germinate, its a “black box” definition. Many different seed phenomena, potentially caused by a multitude of different mechanisms, all fall under this vague term What we call dormant is a range of germinability states, from those right on the edge of germination and those profoundly dormant..
Germinability: the capacity of an seed, bud or spore to germinate under some set of conditions.
The dormant seed requires after-ripening for it to become capable of germination. After-ripening of weed seed usually occurs in the soil from the time it is shed in the growing season until it germinates, often over one or more winters (in the north temperate regions like Iowa). In the soil, physiological, chemical and physical changes occur and after-ripening proceeds.
After-ripening: period after dispersal when the seed, spores and bud cannot germinate, even under favorable conditions, and during which changes occur allowing it to germinate.
Imagine a seed at the end of this after-ripening period and process, it is still dormant but it is almost ready to germinate. If we were able to place this single seed in many slightly different but “favorable” germination conditions (parallel universes?), it might germinate in some and not others. This dormancy term just doesn’t tell us much. The ability to germinate is often a “window” of sensitivity to environmental conditions. As germination conditions in the soil become more favorable, the “window” widens and more weed seed germinate.
Weed seed, and their germination requirements, are very diverse. They range from the deeply and profoundly dormant (imagine a velvetleaf seed buried 18 inches in some cold Minnesota soil, its hard coat protecting it, low oxygen is preventing oxidative stress and germination: happy as a clam and able to last a hundred years) to viviparous (imagine a foxtail seed germinating right on the panicle in a warm, moist, foggy field along the edge of the Chesapeake Bay in Maryland).
Vivipary: germinating while still attached to the parent plant.
Viviparous: producing offspring from within the body of the parent.
In this unit we will learn the established model of seed germination and dormancy. We will also discuss how a better model might be developed. Warning signs that you are headed in the wrong direction in your thinking can be signaled by the following terms, often used. BEWARE!:
Block to Dormancy
Switch for Germination
Finally, it is important to realize there is very good reason for this confusion, why seed biologists like nothing better to do than invent new terms: dormancy is an extremely complicated area of biology. No one has figured out the mechanism of embryo dormancy. Most of the literature doesn’t differentiate between dormancy imposed by the embryo and surrounding (often inhibiting) envelopes when they discuss experimental seed germination results. And then there is the problem of figuring out how these seed behave in the soil: agricultural soil seed bank dynamics.
The Traditional Seed Dormancy Model
Reading assignment: Harper: Ch. 3: pp. 61-82; summary p. xiv-xv
- Introductory concepts
- Innate dormancy
- Enforced dormancy
- Induced dormancy
- Dormancy: 1. A state in which viable seeds (or buds; spores) fail to germinate under conditions of moisture, temperature and oxygen favorable for vegetative growth (Amen, 1968); 2. A state of relative metabolic quienscence
- Seasonal dormancy: in an environment in which favorable growth conditions are seasonal, dormancy is usually clocked by solar rhythm; consequences to the population: predictive
- Opportunistic dormancy: when there is only a small seasonal element in the occurrence of favorable conditions dormancy tends to be both imposed and released by the direct experience of the unfavorable or favorable conditions; consequences to the population: responsive
- Innate dormancy: the condition of seeds as they leave the parent plant, and is a viable state but prevented from germinating when exposed to warm, moist aerated conditions by some property of the embryo or the associated endosperm or maternal structures
- Induced dormancy: is an acquired condition of inability to germinate caused by some experience after ripening; in the opportunistic dormancy category
- Enforced dormancy: is an inability to germinate due to an environmental restraint: shortage of water, low temperature, poor aeration, etc.; in the opportunistic dormancy category
- Dormancy occurs during periods of unfavorable conditions; is more resistant to environmental hazards
- Dormancy can be seen as “dispersal in time”
- Dispersal phase usually dormant; dehydrated seeds weight less (esp. wind dispersal) and are metabolically slower
a. rhythmic adaptation of weeds to the temporal rhythms in the environment
b. strategic alternative to dispersal
- Seed-Environment interactions: Seed dormancy a product of the interaction of the seed (embryo, envelopes like a seed coat, seed food reserves like the endosperm or cotyledons) and the environment (temperature, gases, water, light, soil, temperature)
Evolutionary and environmental context of weed seed dormancy as an adaptive strategy in the struggle for existence
1. Dormancy weak solution to problem of adaptation in changing environment: time lost in capturing resources, in struggle for existence with neighbors
- Probability of suffering greater hardship by continuing growth; evolutionary solution: annual habit, dormancy
- The effort or cost of a seasonally dimorphic phenotype; evolutionary solution: e.g. desert shrubs with somatic polymorphism, large leaves in wet season, small leaves and scales in dry season
- The cost of producing a homeostatic growth form that is tolerant of the whole range of environmental conditions: cost of wide tolerance is great, obtained at cost of reduced efficiency at optimum conditions; evolutionary solution: e.g. evergreen trees
- a. Grain: the way in which an individual plant experiences the heterogeneity of the environment
- Fine-grained environment: individual exposed to environmental factors in small doses, short-termed flucuations; each individual in each year experiences the same range of environments in the same frequencies and there is no uncertainty; example: long-lived perennial tree experiences entire range of environmental factors in yearly changes
- Coarse-grained environment: each individual spends its whole life, or a long period at least, in a single environmental alternative; example: annual weed gets a dry year, or a wet year
- Optimal adaptive strategies in a heterogeneous environment:
-if environments experienced are not too different, the optimal adaptive strategy is a single type of best suited to some intermediate environment
-if the environments experienced are very different, and acts in a fine-grained way, the optimal strategy is a single type of dormancy adapted to the more common environment
-if the environments experienced are very different, and act in a coarse-grained way, the optimal strategy is polymorphism; seed dormancy in weeds in temperate climates good example here
- Innate dormancy conferred by the fact that the process of growth of an embryo to a stage fit for germination has not been completed while the the embryo was still borne on the parent plant, it is shed morphologically incomplete
- Example: Heracleum sphondylium: development of embryo continues at the expense of extra-embryonic food reserves for several months after seed is shed
- This process imposes a necessary time lag between dispersal and germination
Control by a biochemical trigger
1. A biochemical process may need to be stimulated before the germination process can begin
2. Often this trigger is a seasonally related stimulus which can switch on germination at an adaptively appropriate time of year
- Cues and triggers involved in breaking innate dormancy do not produce a clear “all or nothing” effect: only a portion of the seed germinate at one time; a spectrum of requirements by seeds in a single sample which may reflect:
-different maternal influences
-different ages and ripening conditions (influence of different environmental conditions at different times during reproductive phase, in same plant: Cavers)
- Light and phytochrome
1. Example Betula pubescens (UK)
a. require light and long days for germination
b. length of dark period critical: germination declines with increasing dark period length
c. temperature dependence complicated light dependence
-at 20C light dependence lost
-with chilling treatment light dependence lost
2. Several other species follow variations on this same theme: e.g. many dicot weeds (Isikawa, 1954)
- Temperature: chilling or flucuating temperature-
example: Papaver spp.: diurnal flucuation between 10 and 30C breaks dormancy; occurs in upper layers of UK soils in April and May, fixes time of germination
- Nitrate ion: NO3-
-nitrate concentration of the soil solution often rises quite sharply as the soil temperature increases in the spring (Russell, 1962)
-stimulation of Chenopodium album seed germination in the field, and several other species, stimulated by nitrate
- Germination stimulants
-e.g. ethanol, anesthetics, etc.
-[ecological, agricultural relevance?]
- Triggering of biochemical process which destroys a germination inhibitor: breakdown process of inhibitor which occurs within the tissues of the seed
- Physical leaching, or removal of the source of, and inhibitor: leaching or destruction of inhibitor by an external agent
Physical restriction of gas exchange and growth
- Impermeable (or relatively impermeable) seed or fruit coat may prevent water or gas uptake by seed and prevent germination until physical damage occurs to this barrier
-example: Avena fatua (wild oat) seed dormancy broken easily by pricking pericarp
- Common innate dormancy in species inhabiting sand dunes; abrasion by sand
- Scarification: seeds that require abrasion tend to break dormancy at different times rather than in a sudden flush
-example: Avena fatua (wild oat) seed dormancy broken easily by pricking pericarp
-example: Abutilon theophrasti hard seed coat: germination broken readily with treatments cracking hard seed coat (50C for 15 minutes; 10-15 minutes in sulfuric acid); hard seed coat confers very long dormancy and viability in soil seed bank
2. Dormancy caused by mechanical restriction of growth by embryo coverings (pericarp, testa, perisperm, endosperm)
-example: cocklebur: upper seed (of two in capsule) radicle is restricted, insufficient thrust to rupture testa and germinate
- Innate dormancy appears to be under strict and simple genetic control; often modified by maternal effects (i.e. endosperm effects from mother; maternal origin of ovary)
- Commercial crop seed have lost dormancy present in wild relatives in process of domestication; some dormancy left as protection from precocious germination of crop seed while still in inflorescence (?) in wet weather near harvest time
- Genetically controlled polymorphism: distinctly different dormancy genotypes
-example: Spergula arvensis: 3 different seed coats, each control different levels of seed dormancy
-example: Nicandra physaloides: presence or absence of isochromosomes determines whether the seed is non- dormant (2n = 20) or dormant (2n = 19)
Somatic polymorphism and innate dormancy
Somatic polymorphism: Production of seeds of different morphologies or behavior (phenotypes) on different parts of the same plant; not a genetic segregation but a somatic differentiation
- Adaptive advantage to producing seed on one plant with different qualities
- Common adaptation limited to weedy species usually
- Seed dormancy somatic polymorphism is common in weedy species of Gramineae, Compositae, Chenopodiaceae and Cruciferae families
- A quality lacking in genetic polymorphisms: continuum of responses, not just a few genetic alternatives
- Proportion of morphs can be subtly and directly altered by environmental conditions experienced by the parent plant
- Water stress in mature leaves plus short days may induce abscissic acid production
- ABA may have an effect on developing seeds as they differentiate histologically in developing seed
- Differences in dormancy in seed may be a function of water stress at time of seed development
- Germinability of seeds as a function of maternal environment (Gutterman, Y. 1978. Acta Horticulurae 83:41-55)
- 1. example: Rumex crispus (curled dock)
-progeny of individual plants vary enormously in ability to germinate in darkness or at common temperature
-variation is greater between plants than between habitats, no one germination response
- Example: Xanthium spp. (Cocklebur);
-seed borne in pairs in capsule: large and small seed dispersed together
-upper seed in capsule usually dormant, lower germinates first due to differences in testa permeability to water entry, leaching of endogenous germination inhibitors
-dormancy breaking requirements different for 2: 12 month difference insurance second will become established if first year unfavorable
- Example: Avena fatua (wild oat), and Avena ludoviciana
-grains borne on different parts of the spikelet have different germination requirements
-first grain of spikelet lacks dormancy, remainder have deep dormancy
- Example: Compositae germination behavior differentiated by seed size, seed formed in ray versus disc flowers
- Example: Chenopodium album (common lambsquarters) may produce 4 different kinds of seed on same plant
-two color categories: brown and black; two seed coat categories: reticulate and smooth
-brown: thin-walled, larger, germinate quicker than black, even at low temperatures; killed by winter, but if they survive have the capacity to produce very large plants with high reproductive output; only 3% of seed on a plant; among the first to be produced by a plant
-black: require cold treatment, nitrate to break dormancy
-ratio of brown:black govered by environmental conditions
- Example: common purslane seed varied from nondormant to dormant on same plant (Egley, 1974)
- Imposed dormancy: state of seed dormancy maintained by the absence of necessary conditions for germination
-E.g. shortage of water, temperature, unfavorable soil atmosphere, etc.
-E.g. seed buried deeply in soil by tillage, etc., commonly has enforced dormancy
- Carbon dioxide narcosis in soils common factor in enforced dormancy; e.g. high respiration sites in soil elevate CO2 (seed respiration, soil microorganisms)
- Lowered oxygen tension in the soil also important here; e.g. severe oxygen starvation in waterlogged or compacted soils
- Temperate agricultural regions: low temperature enforces dormancy
A seed has acquired dormancy which is not innate and which does not require continued enforcement
CO2 narcosis: example: Brassica alba dormancy induced by high CO2 treatment
- ilum acts as a hygoscopically activated valve
- when air is dry the hilum valve opens and allows water loss from seed
- in wet air it closes,
- embryo progressively dries to a value equal to that of the driest environment it experienced
- hard seed dormancy only broken by seed coat scarification
- white and red clover seed
- collected seed from soil in dark after burial treatment
- later seed of many species (buckhorn plantain, corn spurry, field poppy) tested had light requirement for germination which was not needed when freshly harvested
Cold treatment induced light requirement for germination of Stellaria media (Chickweed); one way autumn shed seed acquire light requirement by spring
High temperature exposure of imbibed seeds coupled with restriction of oxygen availability induced dormancy (Villiers, 1972).