Discussion

Proximate Causes of Flowering Promotion in Water-Stressed Loquats

Previous experiments have demonstrated that posharvest DI is a useful strategy to advance bloom and harvest dates in loquat, making the crop more profitable (Hueso and Cuevas, 2007). However, the causes determining bloom earliness could not be explored in that experience. Shoot growth involvement in early flowering has been now specifically addressed. The correlation found among the advancements in terminal bud dormancy, panicle development and bloom dates informs that loquat earliness in response to DI is due to a complete displacement of the reproductive cycle, from flower initiation to full bloom and harvest.

This phenological displacement can be explained by DI effects on shoot growth and rest. Our results show that DI treatments progressively reduced soil water content, effect that in turn was translated into the plant. The reduction of water availability determined an earlier suspension of shoot growth in water-stressed trees, allowing terminal buds to differentiate into panicles before. SEM and conventional microscopy images confirm that the first anatomical changes in the apical meristem leading to the formation of the panicle occurred three weeks earlier in W0% trees. This transition to flowering occurred days before the full establishment of dormancy in the apical bud, when the formation of new leaves slowed down (i.e. increasing plastochron length). Panicle initiation preceding bud dormancy has been observed in avocado indicating that rest is not a prerequisite for the transition to flowering (Salazar-Garcia et al., 1998). Certain growth of buds during the exposure to cool inductive temperatures also appears to be necessary for panicle initiation in mango (Nunez-Elisea and Davenport, 1995). In apple, hand-defoliation soon after harvest makes possible a second annual crop in Indonesia, by preventing dormancy entrance in a time in which flowers are already initiated (Edwards, 1985).

After panicle initiation, flower development will continue if water deficit does not prevent it. At this regard, bud break and panicle development phenophases also proceed first in water-stressed trees, because the timely resumption of full irrigation. On the contrary, a prolongation of DI during August has been proven detrimental for bloom earliness because water stress delays the last steps of panicle elongation (Cuevas et al., 2007). In contrast to the success of moderate and severe DI treatments, pulses of water shortages as those caused by W50% advanced flowering date in a minor extent because they were not able to make buds to enter in dormancy long before controls. Light water-stress caused by a 20% reduction in watering along the season has also failed to substantially modify flowering date in previous experiments (Hueso and Cuevas, 2007). Work in progress is trying to determine the optimum levels of water stress that more rapidly activates the flowering process in loquat. The hypothesis is that a moderate water stress switches on the flowering program without restraining panicle development. A severe stress may serve, however, to extend the positive effect to the whole bud population making blooming season more compact and uniform. The gradual modification of loquat phenology from panicle initiation to bloom in response to increasing water deficits demonstrates that bloom earliness under DI is due to an advancement of the flower induction process and not due to a higher developmental rate in water-stressed trees, as it has been argued in mango (Nunez-Elisea and Davenport, 1994). Bringing first discernible response to DI to summer can be useful to farmers since provides an early indication of success.

Flowering advancement in water-stressed loquats coincides with observations carried out in citrus and mango. Out of season flowering in response to water withholding is a well-known strategy for citrus producers (Barbera et al., 1985). In several species of genus Citrus, a severe water stress imposed in summer provokes, after rewatering, a second bloom that sets a more valuable crop next summer (Maranto and Hake, 1985). In mango, water stress also advance bloom date (Nunez-Elisea and Davenport, 1994; Lu and Chacko, 2000). Because fully irrigated mangoes bloom profusely (as our control trees), Nunez-Elisea and Davenport (1994) conclude that water stress is not essential for induction of floral morphogenesis in mango grown the subtropics, where cool temperatures have been identified as the main flowering stimulus. In the tropics, however, night temperatures remain too high for induction and a dry period is proposed as the environmental cue for flower induction (Lu and Chacko, 2000).

It is worthwhile to mention that although the onset of summer rest took place before in DI treatments, the number of phytomers, structural segments composed of a leaf, bud, node and internode, was not modified by water deficit. From our results is deduced that the last phytomers were initiated and its founder cells recruited at the time water stress was imposed. At this moment, a number of leaves may still expand, but others remain as undifferentiated foliar primordia below the differentiating terminal bud, until the recovery of plant water status. In lychee, a tropical species with terminal panicles, the same levels of DI here applied (50%, 25% and 0%) greatly reduce postharvest shoot growth and increase flowering and yield. This results made to the authors conclude that resource competition among vegetative and reproductive growth operate in lychee. In loquat, same shoot leaf number and similar length in fully irrigated and water-stressed loquats suggests that the amount of resources allocated to vegetative versus reproductive growth scarcely changed in response to DI.

Ultimate Reasons behind Flowering Promotion in Water-Stressed Fruit Trees

A model for explaining flowering promotion in water-stressed trees is then alternatively proposed. From Arabidopsis studies, we have learned that the transition to flowering in annuals may be regulated by multiple signals and multiple pathways. In Arabidopsis, flowering is controlled by four pathways. All these pathways converge to regulate the meristem identity gene LEAFY (Soltis et al., 2002). A LEAFY-like gene has been recently isolated in six species of Maloideae including loquat, where the highest levels of transcription are expressed at bud break (Esumi et al., 2005; Liu et al., 2007). One of the floral pathways identified in Arabidopsis is GA dependent, but whereas GA is a floral promoter in long-day annuals, it inhibits flowering in fruit trees (Sedgley and Griffin, 1989). This fundamental difference in the role of GA has to be taking into account when proposing a model for flowering in fruit trees. In annuals, an increase in GA activates floral pathways integrators that regulate the formation of flowers (Ausin et al., 2005; Percy, 2005). The situation must be reverse in fruit trees where fast shoot growth as that provided by GA cancel the chance of bud dormancy and flower initiation in fruit trees. For this reason, flower initiation in Angiosperm woody plants is not only compatible but it may require bud dormancy.

In our model, drought promotes abscisic acid (ABA) synthesis and transport to the leaves to induce rapid stomata closure (Beardsell and Cohen, 1975). ABA antagonism with growth promoters hormones, noticeably gibberellins (GA), may eventually lead to a hormonal balance favourable to the onset of terminal bud dormancy (Wareing, 1978; Michalczuk, 2005), allowing flowering program to be expressed. This is not to say that ABA plays a role of floral promoter but that its antagonism with the floral inhibitor (GA) indirectly promotes flowering in water-stressed trees. Goldschmidt and Samach (2004) argue that woody perennials may be constantly induced, but the flowering is repressed by a floral inhibitor (GA in our model) which would act in a similar manner as FLOWERING LOCUS C gene represses flowering transition in Arabidopsis. No need of floral promoter in fruit trees is deduced from this approximation. GA involvement in shoot growth is well documented in annuals and woody plants, where low GA content reduces growth, especially internode elongation. ABA, formerly known as "dormin" (Eagles and Wareing, 1963), specifically inhibits GA biosynthesis and blocks the formation of enzymes as a-amylase that are stimulated by GA to obtain energy for maintenance and growth. Furthermore, GA and ABA are both terpenoids that share the same promoter, the mevalonate, allowing competition for substrate to take place. Common tree flowering promotion in response to triazoles application and ABA may respond to the antagonism of both molecules to GA biosynthesis.

Note that in this model, new leaves as a source of flower inhibitors are not required for explaining lack of flowering in active growing shoots. Rather on the contrary, the presence of new leaves is the negation of the conditions required for flowering transition to occur. Reduced activity in the apical meristem (i.e. increased plastochron) due to water stress is the only switch needed for the activation of the flowering program that includes bud competence and flower initiation during summer. Some kind of citokinins (CK) involvement in bud dormancy onset and release seems likely since its synthesis in roots and delivery to the leaves is usually decreased in water-stressed plants (Pospisilova, 2003). Low CK levels during dormancy and an increase during dormancy release suggest to CK may reinforce the role of GA in bud break. This theory is coherent with the promotion of flowering in response to growth inhibitors and with the negative effects that gibberellins have on tree blooming and fits a series of field observations including tree response to pruning, watering and nitrogen overfertilization.

Many others tropical and subtropical fruit crops seem to require reduced vegetative activity for flowering transition. In lychee, a period of vegetative dormancy is needed to initiate floral buds. This dormancy can be induced by low temperatures, water stress, withholding fertilizers, cincturing and auxin sprays (Menzel, 1993). Whiley and Schaffer (1994) have also noted that flower induction in woody subtropical and tropical evergreen species usually follows a period of quiescence in the canopy caused by environmental conditions (temperature and drought). Bower et al. (1990) propose a simple model to explain flowering in avocado in response to low temperatures that also relays on vegetative growth stops and low GA content. In this model lack of shoot growth (low GA) would conduce to a reduction in available carbohydrates and to an increase in CK and ABA synthesis by new roots. In avocado, CK would increase the number of sprouting buds while ABA would regulate the transition to flowering in apical buds. This scenario coincides with Ben-Tal (1986) inhibitory theory, in which vegetative growth end is marked as a necessary step for flowering to take place in fruit trees. Ben-Tal (1986) emphasizes that vegetative growth disturbance is the general rule that explains tree flowering in response to many different inductive factors such as day length, temperature, hormones, plant size and, as in water-stressed loquats, drought.

Although the previous model for flowering promotion in water-stressed loquats seems plausible, an identification of the environmental factors stimulating terminal bud dormancy and flower initiation in well-irrigated loquats is still needed. This floral stimulus must be naturally produced and must be responsive to water-stress. Two different possibilities arise: correlative inhibition (paradormancy) exerted by competing organs and an environmentally induced dormancy (ecodormancy). In temperate-zone trees, paradormancy develops as days shorten in late summer (Faust et al., 1997). During this period, ABA content increases, although dormancy is still relatively shallow. In late fall, dormancy becomes more intense as dehydrins accumulate in the bud triggered by ABA and decreasing air temperatures (Faust et al., 1997). Arora et al. (2003) acknowledge the difficulty delinking the functions of ABA in cold hardiness versus dormancy in the buds of temperate-zone fruit trees. Different results make the authors yet to assign a major role of ABA in cold acclimation. However, this function does not fit into loquat characteristics since its terminal buds are formed in summer and do not exhibit cold resistance (loquats bloom on November). Although the literature commonly infers that the short day length of late summer is responsible of the cessation of shoot growth, many temperate-zone woody plants, as well as loquat, form terminal buds in early summer with long day periods (Powell, 1987). Shoot growth cessation at this time may be due to the competition of numerous metabolic sinks for essential metabolites (Powell, 1987). Whatever the final reason and exact time may be, shot elongation ceases and apical bud dormancy is established. Correlative inhibition of the apex by mature leaves operate in apple as it has been shown in hand-defoliation experiments (Faust et al., 1997). Interestingly, an ABA decrease, a GA increase and small changes in CK take place in these apple floral buds forced to break paradormancy (Edwards, 1985). Mango apical buds also exhibit foliar paradormancy (Nunez-Elisea and Davenport, 1995). Apical bud paradormancy triggered by an early increase of ABA in mature leaves is compatible with the flowering advancements achieved under DI.

On the other hand, the analysis of loquat phenology has shown that in control trees terminal bud enters in dormancy in August when the evapotranspirative demand and temperatures are high (Cuevas et al., 1997). Water soil content and plant water status in fully irrigated trees negate, however, a situation of water stress, leaving high temperature as the only relevant environmental factor. In this situation the great similarities between seed and bud dormancy (Powell, 1987) may be of help. Seed dormancy induced by high temperature is known as thermodormancy. Thermodormancy inhibits seed germination in late summer in many important crops. ABA and other GA inhibitors cause seed thermodormancy, while chilling and GA release seeds from dormancy. Same situation may apply to loquat buds. This mechanism for inducing ecodormancy is compatible too with the flowering advancement observed under DI. At this regard, it is well known that drought increases leaf temperature which indirectly may advance bud thermodormancy. Rewatering, on the contrary, reduces leaf temperature. Environmental effects on bud dormancy maintenance can be deduced by comparison of loquat behaviour in contrasting climates. In the tropics (San Juan del Obispo, Guatemala), with constant moderate temperatures (25/15°C) along the year, loquat has modified its annual cycle at high altitudes and forms panicles during a dry period in spring

(the usual time for harvest), reaching bloom during the wet summer. This shift suggests a favourable effect of water shortages on flowering transition, but disregards temperatures as main dormancy inducers (B. Sercu, com. pers.). The effects of defoliation and high temperatures on potted loquats are now in study trying to elucidate the factors causing the onset and release of terminal bud dormancy in this species.

Was this article helpful?

0 0

Post a comment