Summer Dormancy in Perennial Temperate Grasses

Abstract

• Background and Aims Dormancy has been extensively studied in plants which experience severe winter conditions but much less so in perennial herbaceous plants that must survive summer drought. This paper reviews the current knowledge on summer dormancy in both native and cultivated perennial temperate grasses originating from the Mediterranean Basin, and presents a unified terminology to describe this trait.

• Scope Under severe drought, it is difficult to separate the responses by which plants avoid and tolerate dehydration from those associated with the expression of summer dormancy. Consequently, this type of endogenous (endo-) dormancy can be tested only in plants that are not subjected to moisture deficit. Summer dormancy can be defined by four criteria, one of which is considered optional: (1) reduction or cessation of leaf production and expansion; (2) senescence of mature foliage; (3) dehydration of surviving organs; and (4, optional) formation of resting organs. The proposed terminology recognizes two levels of summer dormancy: (a) complete dormancy, when cessation of growth is associated with full senescence of foliage and induced dehydration of leaf bases; and (b) incomplete dormancy, when leaf growth is partially inhibited and is associated with moderate levels of foliage senescence. Summer dormancy is expressed under increasing photoperiod and temperature. It is under hormonal control and usually associated with flowering and a reduction in metabolic activity in meristematic tissues. Dehydration tolerance and dormancy are independent phenomena and differ from the adaptations of resurrection plants.

• Conclusions Summer dormancy has been correlated with superior survival after severe and repeated summer drought in a large range of perennial grasses. In the face of increasing aridity, this trait could be used in the development of cultivars that are able to meet agronomic and environmental goals. It is therefore important to have a better understanding of the genetic and environmental control of summer dormancy.

INTRODUCTION

In his often-quoted review (Salazs and Levinsh, 2004), Vegis (1964) observed that dormancy is an adaptive response, which has evolved in the environment of origin of the species, enabling survival during seasons when environmental conditions are most threatening. He noted that dormancy over the winter months was common among species from high latitudes, where winters are severe, but also acknowledged the existence of a similar developmental stage in plants from Mediterranean climates, which, in contrast, occurs during the summer when drought and heat threaten survival. Summer-dormant plants include many geophytes, with species representating the Alliaceae, Orchidaceae, Poaceae and Liliaceae. As a group, they have received little attention (Vegis, 1964), although, more recently, it has been acknowledged that the physiology of summer dormancy warrants further research (Parsons, 2000; Vaughton and Ramsey, 2001; Kamenetsky et al., 2003).

One of the earlier studies of summer dormancy in the Poaceae showed that, under the arid summer climate of southern California, a few species of perennial grasses (in particular Poa scabrella and Poa bulbosa) became dormant even though supplied with water throughout the dry summer (Laude, 1953). For perennial herbaceous plants originating in Mediterranean environments, where summer drought is long and severe, growth cessation and the onset of dormancy before the unfavourable season ensures survival by maintaining the viability of plant meristems (Koller, 1969). By retarding or suspending most metabolic processes, dormancy prevents regrowth in the event of occasional summer rains, as this might be detrimental to summer survival (Hoen, 1968), especially when typical dry, hot conditions return. In the same way, winter dormancy prevents new shoot growth, should unseasonally warm conditions occur (Anderson et al., 2005).

It is notoriously difficult to define dormancy (Rees, 1981), a term loosely used to describe any state of suspended activity in organisms (Villiers, 1975). However, the definition: ‘Dormancy is a temporary suspension of visible growth of any plant structure containing a meristem’ (Lang et al., 1987) has gained some acceptance, and is the basis for the definition of three types of dormancy: (a) endo-dormancy, when the regulation is mediated through the physiology of the affected structure; (b) para-dormancy, when the regulation involves a specific biochemical signal originating outside the affected structure (such as apical dominance); and (c) eco-dormancy, which implies regulation by environmental factors without internal control.

An alternative approach to conceptualizing dormancy has been provided by Rees (1981) and Van der Schoot et al. (1995), who view the phenomenon as the stationary phase of plant development. This approach has advantages, as the dormant state can be characterized separately from its mode of induction (Van der Schoot, 1996). Endo-dormancy can then be defined as the most stable state of the apical meristem because it is trapped in this state by an alteration of the symplasm interconnections, and the only difference between the types of dormancy is the ‘control’ over their stability.

The classification proposed by Lang et al. (1987) has also been criticized by Junttila (1988), who observed that it would widen the historical meaning of dormancy from the earlier, commonly accepted, meaning proposed by Amen (1968), Salisbury and Ross (1985) and Black et al. (1987), in which true or obligate dormancy was restricted to cases where the suspension of growth occurred under favourable environmental conditions. For example, Junttila (1988) noted that under the new classification, dry seed, without any specific requirement for germination other than addition of water, would be classified as dormant (in this case eco-dormant), although, under the earlier definitions it was not. Similarly, it is argued that the buds of over-summering grasses cannot be assessed as dormant unless they are provided with water during the alleged summer dormancy period. Thus ‘eco-dormancy’, particularly of summer-dormant grasses growing in a summer-dry environment, is a concept of questionable value since no plant is able to grow without water; and it can be argued that categories of eco-dormancy such as quiescence, relative dormancy (Hoen, 1968), facultative dormancy (Vaughton and Ramsey, 2001) or conditional dormancy (Vegis, 1964) are merely the prevention of growth by an environmental constraint such as shortage of water, which is one of the basic requirements for normal growth (Villiers, 1975).

Furthermore, it is difficult to demonstrate unequivocally the occurrence of summer dormancy as it is frequently confounded with responses by which plants avoid and tolerate dehydration. Indeed, when subjected to severe water deficit at any period of the year, temperate perennial grasses respond similarly, with a gradual decrease in leaf elongation, leading eventually to cessation, and the progressive senescence of all mature leaves, followed by a stage when only the meristems survive for varying times depending on the dehydration tolerance of the population or species (Volaire et al., 1998; Volaire and Lelievre, 2001). It would be incorrect to describe these responses as autumn or spring dormancy if drought was imposed during those seasons and, when these responses are exhibited under a summer drought, they are not necessarily indicative of the potential of the plant to exhibit summer endo-dormancy.

It is, therefore, necessary to propose a clear definition of summer dormancy that is consistent with a method to assess the level of dormancy when the constraining effect of water deficit is removed. A review of the literature relating to summer dormancy reveals a varied terminology, often imprecisely defined, that originates from a wide diversity of interests and research approaches (e.g. physiological, ecological, genetic, etc.). A unified terminology and associated scoring methods are essential to improve mutual understanding in the research community and introduce relevant objectives and traits, with benefits for both basic research and applied breeding programmes.

CHARACTERISTICS OF SUMMER DORMANCY

In arid areas with low irregular rainfall, the adaptative responses of native tussock grasses such as Stipa tenacissima are based on arrested development, which cannot be ascribed to summer dormancy since the plants can regrow opportunistically within days after water is applied (Pugnaire et al., 1996). Conversely, under temperate or continental climates, where rainfall is spread over the year, low temperature is normally the main limiting factor for the growth of forage plants, and winter dormancy and/or cold tolerance are the main adaptative responses (Eagles, 1989).

Therefore, summer dormancy appears to be an adaptation of perennial species subjected to the predictable long and dry summers of Mediterranean climates. This strategy of extreme drought escape mimics the response of annual plants that avoid the deleterious effects of drought as seed. However, whereas the seed embryo is a whole plant, buds consist of a shoot only (Crabbé, 1994). Thus, in herbaceous plants, seed and bud dormancy appear to be similar phenomena but the mechanisms involved may be different (Okagami, 1986; Dennis, 1996). Species exhibiting summer dormancy include wild grasses such as P. scabrella (Laude, 1953), P. bulbosa (Ofir and Kigel, 1999; Volaire et al., 2001), Hordeum bulbosum (Ofir et al., 1967) and some populations of forage grasses such as Dactylis glomerata ssp. hispanica ‘Kasbah’ (Volaire, 2002) and ecotypes of cocksfoot from Morocco (Al Faïz, INRA Morocco, pers. comm.) and from Israël (Malinowski, Texas University, pers. comm.) that are all currently under evaluation. All of these genotypes exhibit four physiological and developmental processes after flowering in late spring, which are characteristic of the summer dormancy trait: (a) cessation of leaf growth; (b) senescence of most aboveground herbage; (c) dehydration of the bases of the youngest leaves at the base of vegetative tillers, which contain the meristematic tissues; and (d) formation of swollen basal internodes (corm/tuber) or swollen leaf bases (such as bulbs) that can be regarded as resting organs with associated regeneration buds (Burns, 1946). It is characteristic of species with summer dormancy that these processes occur, irrespective of soil water availability.

Physiological and biochemical analyses to extend characterization of summer dormancy in D. glomerata have shown that the more dormant the population, the lower the metabolic activity during the summer, this being associated with an earlier decline in the content of monosaccharides in the leaf bases in spring (Volaire et al., 2005). A comparative study of cocksfoot and P. bulbosa showed an accumulation of dehydrin proteins, originally found in dormant seeds (Close, 1997), in water-stressed leaf bases of three populations of cocksfoot that could dehydrate down to 40 % water content, although a greater number of dehydrins were expressed in bulbs of P. bulbosa which could dehydrate to 10 % water content (Volaire et al., 2001; Volaire, 2002). Dehydrin expression did not correlate with the drought survival of the tested genotypes but was merely a function of the water status of the tissues. Indeed, these proteins were present in all dormant, and therefore dehydrated, plants under irrigation, reflecting the endogenous control of dehydration in these tissues, irrespective of the soil moisture (Volaire et al., 2005). Later research demonstrated that non-dormant populations of cocksfoot and tall fescue (Festuca arundinacea) responded to a summer rainstorm with a decline in dehydrin expression in leaf bases, whereas no such decline occurred in the dormant cultivars. Indeed, in the dormant cocksfoot cultivar which did not regrow and had a lower rate of water use, the leaf bases remained drier than in the non-dormant cultivar after the storm, whereas, in tall fescue, no difference in moisture content was seen between dormant and non-dormant populations (Norton et al., 2006a, b).

Studies of the hydration of surviving meristematic tissues of common perennial forage grasses have shown that none tested so far could be considered to be desiccation tolerant (Volaire et al., 1998; Volaire, 2002; Norton et al., 2006b). In the case of cocksfoot, the lethal water content of meristems at the bases of the last enclosed leaves was found to be around 0·5–0·8 g H₂O g⁻¹ dry weight, irrespective of the dormancy of the plants (Volaire et al., 2001). Other research with phalaris (Phalaris aquatica) and H. bulbosum showed that severing their roots from a source of sub-soil moisture resulted in plant death. This confirmed that a continuous water supply, however small, is indispensable for survival of these species during the summer resting period (Laude, 1953; Ofir et al., 1967; McWilliam and Kramer, 1968). Therefore, summer dormancy appears to confer superior drought avoidance but not necessarily higher dehydration tolerance as was initially hypothesized (Volaire, 2002). Moreover, it has been shown that dehydration tolerance and summer dormancy, although usually associated, are independent phenomena, at least in P. bulbosa and D. glomerata, since dehydration tolerance is exhibited when plants are subjected to drought at any time of the year, whereas dormancy is exhibited only in summer when induced by daylength and temperature (Volaire et al., 2001).

The case of P. bulbosa is noteworthy, as it is a rare example of a desiccation-tolerant grass, with bulbs that can dry to lower than 10 % moisture and yet maintain their regrowth potential over extended periods. In general, summer dormancy differs from desiccation tolerance, which is possessed by a rare group of angiosperms termed resurrection or poikilohydric plants (Gaff, 1971), which can survive the loss of most of their tissue water for long periods, but revive rapidly upon rewatering (Bartels et al., 1996; Scott, 2000). The ability of the air-dried foliage to restore photosynthetic function as soon as water is available is an adaptation to very xeric sites (Lazarides, 1992). However, desiccation tolerance differs from summer dormancy since (a) it involves mature tissues rather than meristems [although P. bulbosa is sometimes and questionably referred to as a resurrection plant (Lazarides 1992)]; and (b) it involves rapid recovery of tissues whenever water is not limiting and therefore is imposed by a restriction of water. In contrast to dormancy, therefore, it is not a periodical endogenous regulation of metabolism controlled by photoperiod and temperature.

INDUCTION AND RELEASE OF SUMMER DORMANCY

In temperate plant species, the environmental cues, such as photoperiod and temperature, for induction of processes, such as flowering or the onset and release of dormancy, are sensed by the meristematic regions of buds (Metzger, 1996). Endo-dormancy of meristems of Mediterranean temperate geophytic grasses such as P. bulbosa, H. bulbosum and P. aquatica (Ofir et al., 1967; McWilliam, 1968; Ofir and Kerem, 1982; Ofir and Kigel, 1999) develops under increasing daylength and temperature. Induction apparently starts during early winter since it is enhanced by pre-exposure to short days and low temperatures (Ofir and Dorenfield, 1992). Conversely, the gradual release of dormancy in these species is mediated by relatively high temperatures (decreased inhibition by these high temperatures) at the end of the summer, while sprouting of the dormant buds is accelerated once the lower temperatures characteristic of early autumn recommence (Ofir, 1986). Moreover, the formation of organs associated with dormancy (e.g. bulbs, corms and tubers) and of those associated with stem elongation and flowering may be linked processes (Ofir et al., 1967; Hoen, 1968). However, in P. bulbosa, dormancy development and flowering are independent, so that, in some populations, flowering is rare or absent although the plants still become dormant (Ofir and Kerem, 1982; Ofir and Kigel, 2003, 2006). On the other hand, in H. bulbosum, the formation of corms and their adjoining resting regeneration buds occurs at an early stage of reproduction. (Ofir et al., 1967; Ofir and Koller, 1974). In addition, the more arid the site of plant origin, the earlier populations of D. glomerata and P. bulbosa flowered (Volaire and Lelievre, 1997; Ofir and Kigel, 2003) and this was correlated with an earlier onset of dormancy and superior drought survival. Hoen (1968) showed that only reproductive plants of P. aquatica carried dormant buds, whereas Sankary et al. (1969) observed that populations originating from the more arid sites had longer dormancy.

Conditions required for induction and release of summer dormancy in perennial forage grasses have received little attention. In cultivars of cocksfoot and tall fescue, growth reduction and senescence of mature foliage under summer irrigation were much less when plants were sown in spring than in the preceding autumn (Norton et al., 2006a, b). These results highlighted the effects of early inductive factors characteristic of previous winter conditions, that could be mediated through gibberellin (Ofir, 1976) and/or abscisic acid, as shown in H. bulbosum and P. bulbosa (Ofir and Koller, 1974; Ofir, 1976; Ofir and Kigel, 1998). Indeed, similar results, which emphasized the role of early induction in activation of these hormones, were found in studies with seed embryos (Le Page-Degivry et al., 1996) and tubers (Suttle, 2000). In contrast, as shown for P. bulbosa and D. glomerata, a change in soil water content is unlikely to influence or control induction of dormancy, since induced dehydration of the meristems associated with full leaf senescence occurs before the onset of water deficiency, and is maintained even under irrigation (Volaire, 2002). Instead, as shown for cold resistance of winter-dormant buds (Erez, 2000), low water potential in the meristems can even enhance drought avoidance of the surviving organs (Norton et al., 2006a).

HOW TO DEFINE SUMMER DORMANCY

Until now, the intensity of summer dormancy has been assessed by a range of methods that reflect the diversity of definitions applied. When the plant material is assessed under summer drought, dormancy has been scored as the percentage of herbage senescence in summer (Silsbury, 1961, 1964; Lorenzetti et al., 1981; Ceccarelli and Somaroo, 1983) and related to summer survival. This technique does not permit discrimination between plant responses associated with dormancy and those involved in the typical response to severe drought, which can, as previously discussed, occur at any season. Therefore, this approach cannot be used to test the hypothesis that summer survival can be attributed unequivocally to summer dormancy rather than to other adaptations involving high dehydration tolerance. For example, a Mediterranean cultivar of cocksfoot ‘Medly’, that was claimed to be summer dormant because of its high persistence under typical summer droughts of the south of France, was shown to be active under summer irrigation; its survival under extreme drought was lower than that of ‘Kasbah’, whose resistance was known to be associated with summer dormancy (Norton et al., 2006a).

Conversely, most studies in P. aquatica have measured dormancy in terms of the regrowth of the buds after a midsummer rainstorm in the field (Oram, 1983, 1984; Oram and Freebairn, 1984; Culvenor and Boschma, 2005), or after transfer into controlled conditions, at lower temperature and higher soil moisture (Hoen, 1968; Biddiscombe et al., 1977). Dormancy is expressed as the relative number of shoots activated by watering, in the field method, or the number of days for 50 % of the final number of new tillers to appear after the relocation of plants from the field to controlled conditions. These techniques are analogous to those used for testing the winter dormancy of tree buds (Williams et al., 1979). However, rewatering plants in the middle of the summer is difficult to reproduce and standardize, since the duration of the dormancy period can vary greatly, for example from 30 to 127 d within ecotypes of P. aquatica in California (Sankary et al., 1969). The actual environmental conditions caused by a rainstorm depend on the intensity and duration of the storm and the subsequent (lower) temperatures. The response of the plants will be affected by the amount of available water in the field, which will be a function of the local soil and climate, and other environmental factors that vary seasonally (temperature and photoperiod). Where plants are transferred into a phytotron, the condition of the root system of the uprooted plants may also affect the shoot regrowth, as suspected by Volaire et al. (2005).

To avoid these problems, it is proposed to assess summer dormancy in the field using a protocol based on the following principles: (a) autumn sowing to ensure optimal induction conditions; (b) irrigation sufficient to compensate for evapotranspiration in all seasons and in particular during summer; (c) measurement of seasonal biomass to assess the potential for herbage production in summer; and (d) scoring of foliage senescence in summer to complete the characterization of the sward under summer irrigation. This protocol has been used in field experiments on cocksfoot and tall fescue (Norton et al., 2006a, b), providing information for the formulation of terminology suitable for the range of summer behaviour in temperate perennial grasses. The proposed definitions are based on three sets of responses, plus an additional and optional anatomical criterion (4): (1) leaf growth in summer (cessation, reduction or continuation); (2) senescence of mature herbage (from none to total); (3) dehydration (from none to desiccation) of enclosed bases of the youngest leaves which represent the last surviving aerial organs (Volaire et al., 1998); and (4) formation of resting organs such as bulbs or corms.

According to the combination of responses, three major groups of populations of temperate perennial grasses can be distinguished.

1. Populations that maintain active growth under irrigation. The foliage of these plants remains photosynthetically active during the summer, with rates of foliage senescence not higher than normal turnover. Consequently, all aboveground organs of the plants remain well hydrated and no resting organs are formed. This pattern of response, characteristic of continental populations of cocksfoot (Volaire, 1995) and tall fescue (Norton et al., 2006b), indicates that these populations do not have the summer dormancy trait.

2. Populations that cease growth completely for a minimum of 4 weeks during the summer (the drought period may last much longer in the site of origin of these populations), and in which most mature aerial tissues senesce. It is proposed to define this pattern of response as complete dormancy. In this group, the endogenously induced dehydration of surviving tissues can vary from moderate (e.g. D. glomerata ‘Kasbah’) to full, air-dry desiccation (P. bulbosa). Species may possess resting organs such as bulbs which exhibit varying degrees of enlargement, e.g. P. bulbosa, H. bulbosum and D. glomerata ‘Kasbah’ which exhibits swollen tiller bases (Norton et al., 2006a).

3. Populations that exhibit markedly reduced growth, associated with partial senescence of foliage, but no dehydration of leaf bases. It is proposed that this pattern of response should be classified as incomplete dormancy. It is commonly found within forage grasses developed from Mediterranean germplasm of D. glomerata and F. arundinacea such as ‘Flecha’ (Lumaret, 1988; Volaire, 1995; Volaire and Lelievre, 1997; Reed et al., 2004a, b; Piano et al., 2005; Norton et al., 2006b). This pattern is also exhibited by P. aquatica, a species that produces mainly reproductive tillers under summer irrigation, and whose buds regrow after a summer rain. Bud regrowth occurs even in the most dormant cultivars such as ‘Sirocco’ and ‘Atlas PG’ (Culvenor and Boschma, 2005). Phalaris aquatica also complies with the fourth criterion of dormancy: the internode at the base of the reproductive tiller swells to form a tuber, albeit remaining highly hydrated over the summer (McWilliam and Kramer, 1968), with axillary dormant buds located directly below and in association with the tuber (Hoen, 1968).

The term ‘incomplete dormancy’ implies a quantitative expression of the trait. The proposed classification is consistent with the model for dormancy cycling, as developed by Rinne et al. (2001), which depicts the meristem as passing through a series of states of cellular communication, with characteristic sensitivities to distinct environmental cues. Moreover, as dormancy induction is mediated through hormonal changes in response to environmental factors, it is likely that these changes might be gradual and vary amongst genotypes according to their place of origin, since photoperiod responses can vary among different ecotypes of perennials that can grow over a broad range of latitudes (Salisbury and Ross, 1985).

The proposed approach parallels the quantification of autumn dormancy in lucerne (Medicago sativa), which is based on the growth potential of lucerne populations in autumn (Teuber et al., 1998), although whether dormancy of lucerne is true endo-dormancy or just reduced growth under decreasing temperatures is unclear. The association between autumn dormancy and winter hardiness was positive (Cunningham et al., 2001), but the two traits were shown to be improvable simultaneously in this species (Brummer et al., 2000; Weishaar et al., 2005). In line with the approach for lucerne, the establishment of summer dormancy would be more reliable after repeated measurements in several locations and over more than 1 year (Teuber et al., 1998). Similarly to the need to measure autumn dormancy and winter survival in regions experiencing harsh winters (Castonguay et al., 2006), the assessment of summer dormancy should be conducted only in typical Mediterranean climates (between a 30° and 45° latitude) experiencing cool season rainfall and typical hot summer droughts (Daget, 1977), and therefore subjected to the range of temperatures and daylengths likely to initiate the endo-dormancy adaptative responses described herein.

CONCLUSION: AGRONOMIC AND ENVIRONMENTAL IMPORTANCE OF THE SUMMER DORMANCY TRAIT

In cultivated perennial grasses, most of the research to date on abiotic stresses has focused on low temperatures and freezing, as these occur widely in temperate grasslands (Renault et al., 2006). However, in a changing global climate with the prospect of increasing aridity, progress should be made in the development of grass cultivars with improved resilience to summer drought, where the over-riding requirement is that the pasture should survive and regrow rapidly when autumn rains start (Humphreys et al., 2006). As found previously in a range of species (Biddiscombe et al., 1977), it is confirmed that summer dormancy, whether complete, as in cocksfoot (Norton et al., 2006a), or incomplete, as in tall fescue (Norton et al., 2006b) and phalaris (Culvenor and Boschma, 2005), is correlated with superior survival and autumn regrowth after severe summer drought. Clearly, summer dormancy confers superior persistence and can be useful in achieving a range of agronomic and environmental goals. The first of these is the development of germplasm for forage production in areas with regular summer droughts. As complete dormancy has been associated mostly with populations of low yield potential, it will be necessary to determine the genetic control of the dormancy trait to provide tools for the development of cultivars with a range of dormancy associated with a higher level of productivity. A second, increasingly important, objective is to develop perennial grasses that are able to fulfil a range of environmental functions, e.g. perennial grasses can improve the structure of soils through biological action of the roots (McCallum et al., 2004) and, as perennial canopies, they can reduce the need for chemical fertilizer and reduce nitrate leaching (King and Berry, 2005). Indeed, dormant grasses have an increasing role in Mediterranean vineyards and orchards where competition for water from the grass must be minimized in summer, while their vigorous growth during the cool season (Knight, 1965, 1968; Culvenor et al., 2002; Reed et al., 2004a) is desirable to minimize weeds and nitrate leaching.

Further study to strengthen understanding of the genetic and environmental control of summer dormancy is necessary for exploitation of the trait in breeding programmes.