Pisum sativum cv. Guido grown under controlled environment conditions was exposed to either low or high UV-B radiation (2·2 or 9·9 kJ m–2 d–1 plant-weighted UV-B, respectively). Low or high UV-B was maintained throughout growth (LL and HH treatments, respectively) or plants were transferred between treatments when 22 d old (giving LH and HL treatments). High UV-B significantly reduced plant dry weight and significantly altered plant morphology. The growth and morphology of plants transferred from low to high UV-B were little affected, when compared with those of LL plants. By contrast, plants moved from high to low UV-B showed marked increases in growth when compared with HH plants. This contrast between HL and LH appeared to be related to the effect of UV-B on plant development. Exposure to high UV-B throughout development consistently reduced leaf areas. In fully expanded leaves there was no significant UV-B effect on cell area and reduced leaf area could be attributed to reduced cell number, suggesting effects on leaf primordia. Further reductions in the leaf area of younger leaves were the result of the slower development rate of plants grown at high UV-B, which also resulted in significant reductions in leaf number.
Recent studies have shown that the effects of UV-B radiation (280–315 nm) on growth in the field cannot be attributed to reduced CO2 assimilation rate per unit leaf area (e.g. Beyschlag et al. 1988; Caldwell, Flint & Searles 1994; Mepsted et al. 1996). The effects of UV-B on biomass may result from reductions in leaf area rather than reductions in assimilation per unit area. Morphological changes such as reduction in leaf area and plant height are a consistent response of plants exposed to UV-B radiation (Barnes, Flint & Caldwell 1990; Ziska et al. 1993; Sullivan et al. 1996). However, the mechanisms underlying such morphological responses to UV-B remain poorly defined. Where UV-B effects on photosynthesis have been reported, the underlying mechanisms are better understood, as UV-B can affect stomata, photosystem II reaction centres, photosynthetic enzymes and pigments (Teramura & Sullivan 1994; Tevini 1994). These latter effects of UV-B on photosynthesis are well characterized in cultivated pea (He et al. 1993; He et al. 1994; Strid, Chow & Anderson 1990). However, such studies have typically used high UV-B radiation in combination with low photosynthetically active radiation, and plants have been transferred from zero UV-B to a very high-UV-B environment. The marked photosynthetic responses observed in these experiments may be a consequence of the shock caused by this sudden UV-B exposure of unadapted plants, as we observed no effect of UV-B on photosynthesis in pea grown under moderate UV-B doses throughout development (González et al. 1996).
To determine whether a sudden exposure to UV-B can inhibit photosynthesis in pea, we measured photosynthesis in plants grown at high and low UV-B doses and then exchanged between treatments. At the same time, as UV-B can reduce the growth of pea even in the absence of any photosynthetic response (González et al. 1996), we examined whether UV-B acted through changes in leaf expansion and, ultimately, cell expansion or division in the epidermis. In addition, other UV-B responses, such as protective mechanisms, were assessed.
MATERIALS AND METHODS
Pisum sativum L. (cv. Guido) plants were grown from seed in 12 cm pots in John Innes no. 2 compost in two replicate controlled-environment chambers. The yield of this cultivar was particularly sensitive to the effects of elevated UV-B in a previous field experiment (Mepsted et al. 1996).
Photosynthetically active radiation (PAR, 400–700 nm) was provided by a combination of two high-pressure sodium lamps (Philips SON-T 1000 W; Starna Ltd, Romford, Essex, UK) and two sets of four metal-halide lamps (Osram HQI-T 400 W; Starna Ltd, Romford, Essex, UK). Lamps were switched on and off in three steps so that maximum irradiances (850–950 μmol m–2s–1 PAR at soil level) were maintained for 12 h of the 16 h photoperiod. The metal-halide lamps also provided some UV-A radiation (315–400 nm) and this was supplemented using four ‘black-light’ fluorescent tubes (Philips TLD-30/12) to give a UV-A irradiance at plant height of 0·98 ± 0·4 W m–2 (equivalent to an integrated UV-A dose of 43 kJ m–2 d–1).
Each controlled-environment chamber was divided longitudinally into two compartments by acrylic sheets opaque to UV-B, so that low-UV-B and high UV-B treatments were provided within each chamber. UV-B radiation was supplied to each subcompartment of each chamber by two or four UV-B fluorescence tubes (Philips TL-40/12), held 0·75 m above the top of the plants. To filter out ultraviolet radiation of wavelength below 290 nm, sheets of cellulose diacetate (100 μm thick Clarifoil, Courtaulds Ltd, Derby, UK) were supported ≈17 cm from the UV-B tubes. At this distance from the lamps, the rate of cellulose diacetate photoageing was slow and was countered by twice-weekly measurement and adjustment of UV-B irradiances, as described previously (González et al. 1996). UV-B plant-weighted according to Caldwell's generalized plant action spectrum (normalized to 300 nm, PAS300; Caldwell 1971; Caldwell et al. 1986) was supplied for 10 h during the middle of the photoperiod to give 2·2 and 9·9 kJ m–2 d–1, respectively, in each subcompartment of each chamber. All measurements of PAR, UV-A and UV-B radiation were measured with a double monochromator spectroradiometer (model SR991-PC; Macam Photometrics, Livingston, West Lothian, UK) calibrated for spectral irradiance against deuterium (240–350 nm) and tungsten (350–800 nm) standards traceable to the National Physical Laboratory (Teddington, Middlesex, UK), and for wavelength alignment checked against a mercury–argon lamp (LOT Oriel, Leatherhead, Surrey, UK).
Twenty-two days after germination, when all plants had developed at least three fully expanded leaves, a subsample (14 and 12 plants from the low-UV-B and high-UV-B treatments, respectively) was taken to assess physiological and growth parameters (see below). At the end of the 22nd day, half of the remaining plants growing under low-UV-B were transferred to high-UV-B treatment, and half of the remaining high-UV-B-treatment plants moved into low UV-B. This crossover of treatments resulted in four combinations of ‘primary’ and ‘secondary’ UV-B treatments (germination–22 d and 22 d–final harvest, respectively). To summarize, the four treatments were: (1) low UV-B radiation throughout the experiment (LL); (2) low primary UV-B followed by transfer to high secondary UV-B (LH); (3) high UV-B radiation throughout the experiment (HH), and (4) high primary UV-B followed by transfer to low secondary UV-B (HL).
Physiology and UV-absorbing pigments
Immediately before crossover, chlorophyll fluorescence and gas exchanges were measured in situ in the youngest fully expanded leaf. UV-absorbing compounds were measured in the youngest fully expanded leaf of destructively harvested plants. The youngest fully expanded leaf of each remaining plants was then marked, and gas exchange and chlorophyll fluorescence were measured on these leaves at 15, 36, 65, 85 and 95 h after the crossover. Gas exchange and chlorophyll fluorescence were also recorded in the youngest fully expanded leaf at the end of the experiment (≈11 d of UV-B radiation after the crossover). To facilitate comparison, these leaf measurements were always made at the same time of day. At the final destructive harvest (12 d after crossover), the absorbance of UV-absorbing compounds was measured in the marked leaf, in the youngest fully expanded leaf and in unexpanded tissue taken from the apex of the main stem. Measurements of chlorophyll fluorescence, leaf gas exchanges and UV-absorbing compounds were made as previously described (González et al. 1996).
In plants harvested immediately before crossover, plant height (length of the main shoot to the base of the uppermost stipule), the number of lateral shoots on the main stem and the number of pairs of stipules on the main stem and laterals were quantified. The area of each pair of leaves and stipules was determined separately at each internode of the main stem using a Li-Cor 3100 leaf-area meter (Li-Cor, Lincoln, NB, USA). Remaining plants were destructively harvested 12 d after the UV-B crossover. At this time, plant height was measured and each plant was separated into main shoot, lateral shoots, leaves and stipules. The area of each pair of leaf and stipules, from both main stem and lateral shoots, was measured separately. All component parts were then oven-dried at 60 °C to constant weight for dry-mass determinations.
At the final harvest, the area of a single leaflet of the second and fifth node in every plant was measured, and a film of transparent varnish was applied on the widest adaxial surface of the leaves, avoiding the central vein, to obtain replicas of the epidermal surface. When dry, the leaf imprint was carefully removed from the leaf, transferred to a microscope slide, and viewed under a microscope. The image was projected on a television monitor using a video camera and a microimage analyser (Quick Image image capture system, software version 2·0, Mass Microsystems Inc., Sunnyvale, CA, USA) was used to count the epidermal cells. The number of cells per unit area was calculated from the number of cells displayed on the monitor (area represented 0·0628 mm2). Cells crossing the bottom and right-hand edges of the screen were included, and those crossing the top and left-hand edges excluded. The average number of cells was calculated for each leaf from observations of 10 randomly chosen fields of view. Epidermal cell area was then calculated from the unit area divided by average number of cells. The number of cells per leaf was obtained from the leaf area divided by average cell area.
The number of experimental plants in each subcompartment of each chamber was 26. To prevent any positional effect in the chambers, plants were randomized completely between chambers every second day. To prevent edge effects, additional plants were placed on either edge of each compartment. All data obtained before crossover were analysed using one-way analysis of variance (ANOVA). Data obtained after the crossover were analysed using a two-way factorial ANOVA that tested the main effects of primary and secondary treatments, and their interaction. All analyses were made using the SPSS 6·0 software package (SPSS Inc., Chicago, IL, USA). Significantly different means were separated by Tukey's HSD multiple-range test with a significance level set at P < 0·05.
Primary UV-B treatments
The primary UV-B treatments had no significant effects on UV-absorbing pigments (per unit leaf weight or area), chlorophyll fluorescence or gas exchange of the youngest fully expanded leaves (data not presented).
The high UV-B dose, however, significantly (P < 0·001) reduced the height and number of fully expanded leaves of the main stem (Figs 1a & b). Total plant foliage area (i.e. leaves plus stipules) was reduced by ≈30% under the high UV-B dose (P < 0·001, Fig. 1c). Most of this reduction was attributable to a 60% reduction in the foliage area of lateral shoots (Fig. 1d), even though the number of lateral shoots was not affected (2·0 ± 0·005 and 1·9 ± 0·008, respectively, in low and high UV-B). On the main axis, elevated UV-B caused reduction in leaf and stipule area at each node; but the reduction was significant (P < 0·01) only from the third and fourth node, respectively (Figs 1e & f).
Secondary UV-B treatments
Secondary UV-B treatments had no significant effect on absorbance at 300 nm (A300) in mature leaves (i.e. youngest fully expanded leaf at the time of crossover) or unexpanded tissues at the end of the study (data not presented). However, in the youngest fully expanded leaves at the final harvest, A300 was significantly higher (21%, P = 0·04) in plants always grown under high UV-B (HH) than it was in plants that received only low UV-B (LL) (Table 1). The interaction between primary and secondary treatment was also significant (P = 0·01). In plants grown at high UV-B and transferred to low UV-B (HL), A300 decreased by 20% (P < 0·01). However, absorbance of these pigments was not significantly altered when plants grown under low UV-B were moved to high UV-B radiation (LH).
Table 1. . Absorbance at 300 nm (g–1 fresh weight) in youngest expanded leaves in pea (cv. Guido) plants exposed to four UV-B treatments
There were no significant treatment effects on transpiration rate, stomatal conductance, net photosynthetic rate or internal CO2 concentration in either mature or youngest fully expanded leaves (data not presented). Time-course measurements of chlorophyll fluorescence made over the first 95 h after crossover revealed significant changes in Fv/Fm, but these were transient, small (< 5%) and all measurements were within the range (0·81–0·86) and unlikely to be biologically significant (Table 2).
Table 2. . The Fv/Fm ratio in marked leaves of pea (cv. Guido) at different times after the cross-over of plants between different UV-B treatments
Growth and morphology
The marked effects of UV-B observed before crossover were reflected in the substantial and significant (consistently P < 0·01) effects of the primary treatments on plant biomass and morphology at the final harvest (Figs 2 & 3). By contrast, while usually significant (P < 0·05), the effects of high UV-B during the secondary treatment were consistently smaller than expected from responses to primary treatments.
Above-ground biomass was significantly reduced in plants exposed to high UV-B throughout the experiment (38% between HH and LL, P < 0·01) but not in plants moved from low to high UV-B (6% reduction between LH and LL: Fig. 2a). By contrast, compared with HH plants above-ground biomass in plants transferred to low UV-B (HL) increased significantly (by 33%: Fig. 2a). Similar patterns of response to primary and secondary UV-B treatments were observed in the dry weights of total foliage (Fig. 2b), the main stem and its foliage (Figs 2c & d) and of lateral shoots (Fig. 2e). The most marked contrast between primary and secondary UV-B treatments was apparent in the dry weight of foliage on lateral shoots (Fig. 2f), where high UV-B throughout growth caused a significant (40%) reduction but a change from low to high UV-B had no effect and a change from high to low resulted in a significant (50%) increase compared with HH. This was the only growth parameter where the interaction between primary and secondary UV-B treatments was statistically significant (P < 0·05).
Plant height (Fig. 3a) was significantly reduced by high UV-B during both primary and secondary treatments (P < 0·001 and P < 0·05, respectively). The foliage areas of the main stem (Fig. 3b) and lateral shoots (Fig. 3c) were particularly responsive to a shift from high to low UV-B, increasing significantly (by 22% and 47%, respectively) in HL plants compared with HH plants. However, plants were relatively insensitive to a change from low to high UV-B (reductions of 6–7% in LH compared with LL, both non-significant). The number of expanded nodes of the main stem was significantly reduced (P < 0·001) under the primary but not secondary UV-B treatment (Fig. 3d).
In plants exposed to high UV-B throughout development, the areas of fully expanded leaves and stipules were invariably, although not always significantly, reduced (5–15% comparing HH with LL: Fig. 4). However, as at the first harvest, the effects of UV-B on the area of leaves and stipules on the main axis were greatest on younger leaves (Fig. 4). Thus, high UV-B, in both primary and secondary treatments, caused significant reductions in the areas of leaves at node 4 and above (Fig. 4a). In plants exposed to high UV-B during primary treatment, there was no expansion of either leaves or stipules at node 9.
The area of leaf 2 was significantly reduced by elevated UV-B (13% comparing HH with LL: Table 3). This reduction in leaf area was a function of a marked (13%), but not quite significant (P = 0·055), reduction in the number of epidermal cells per leaf, while epidermal cell area was unaffected by UV-B (< 1% comparing HH and LL, P = 0·61).
Table 3. . Leaf area and epidermal characteristics of a) the second leaf and b) the fifth leaf of pea (cv. Guido) under different UV-B treatments
Exposure to high UV-B throughout the experiment (HH) induced a significant reduction in the area of the fifth leaflet compared with that observed with the LL treatment (17%, P = 0·01), which corresponded to a 12% reduction in cell area (P = 0·14) and a 6% reduction in cell number (P < 0·05, Table 3). Plants transferred to low UV-B (HL) partly offset this reduction in leaf area through a significant increase in cell area, while there was no change in cell number (Table 3). Overall, factorial ANOVA showed that high UV-B during the primary treatment significantly (P < 0·05) reduced the number of cells but did not affect cell area, while high UV-B during secondary treatment significantly (P < 0·001) reduced cell area, but did not affect cell number. There were no significant interactions between primary and secondary treatments on cell area or number.
One aim of this experiment was to test whether plants grown under low UV-B from germination would respond to transfer to high UV-B conditions with a readily detectable inhibition of photosynthesis. However, changes in gas exchange and chlorophyll fluorescence after crossover were not significant or too small to have any biological significance, or both. On this basis, the large photosynthetic response of pea observed in some other studies (e.g. He et al. 1993, 1994) cannot be attributed simply to the shock effect of sudden transfer to high UV-B but rather to UV-B doses far higher (≈fivefold), and PAR substantially lower, than used here.
Another interesting observation from the crossover, which would not have been apparent had plants been grown under constant UV-B conditions, was the change of UV-absorbing pigments. Two lines of evidence indicate that induction of UV-absorbing pigments in cultivar Guido was related to a direct exposure of UV-B radiation. First, the UV-B treatments had no effect on UV-pigments in unexpanded apical tissue that was enclosed within several stipules and hence not directly exposed to incident UV-B. The lack of UV-B induction in folded tissue resembled that found earlier in another pea line (JI1389) but not in other lines (see González et al. 1996). Secondly, there was a great decrease in the absorbance of these pigments in young expanded leaves of HL plants compared with HH plants — that is, pigments changed in response to current UV-B rather than to past exposure. Conversely, plants transferred from low to high UV-B did not show the increase in absorbance of UV-pigments in young expanded leaves after the crossover that would be expected if UV-absorbing pigments had been induced by direct exposure. At present, the mechanisms underlying these contradictory responses in HL and LH plants are not understood. However, it is clear that the mechanisms by which UV-B induces accumulation of UV-absorbing pigments in both expanded leaves and tissue still enclosed in the bud differ between lines of cultivated pea.
High UV-B treatment altered plant growth and morphology. Plants were shorter and the dry weight and the areas of leaves and stipules were reduced from the first node, resulting in a cumulative reduction in overall leaf area and biomass. In the absence of any change in photosynthesis per unit area, reductions in growth appear to be a function of two inter-related processes: a delay in development and a reduction in leaf expansion.
A delay in plant development at elevated UV-B radiation was apparent at both destructive harvests, particularly in the delayed expansion of younger leaves. At the time of crossover, the large reduction in area of the leaves and stipules of the fourth node was attributable to their incomplete expansion at high UV-B (Fig. 1e,f). Similarly, at the final harvest, in HH plants the areas of leaves and stipules at nodes 7 and above were severely reduced, and there was no expansion at the ninth node (Fig. 4). At any of these nodes, the area in HH was broadly comparable to that of leaves or stipules at the subsequent node in LL, that is development was delayed by approximately one node. There are few reports that UV-B delays whole-plant development, but increased UV-B extended the duration of leaf expansion in barley (Liu, Gitz & McClure 1995), delayed leaf expansion in sweetgum (Dillenburg, Sullivan & Teramura 1995) and delayed flowering in several species (e.g. Musil 1995; Saile-Mark & Tevini 1997).
The delay in development of plants grown at high UV-B under the primary treatment resulted in smaller plants at the time of crossover. This in turn was important in determining the responses of plants to secondary UV-B. Plants initially grown under high UV-B and transferred to low UV-B showed an increase in growth in almost all organs, and especially in dry weight of lateral shoots and leaf area of these shoots. The mean relative growth rates for both height and total foliage area after crossover were significantly increased (P < 0·001 and P < 0·01, respectively) by high UV-B during primary treatment (Table 4). This indirect effect of previous UV-B exposure is consistent with the expected decrease in relative growth rate during continued growth and development. The increase in relative growth rate after high primary UV-B is particularly marked in plants transferred from high to low UV-B. These HL plants had a significantly higher relative growth rate than plants receiving any other treatment, including LL (Table 4); this could be attributed to: (1) their smaller size at the time of transfer to low UV-B combined with (2) the absence of direct growth inhibition by UV-B during secondary treatment. The delay in development observed in plants under high primary treatment did not persist after transfer to low UV-B, and the new UV-B environment not only allowed an increase in leaf expansion but also increased the rate of leaf emergence (Fig. 4). The capacity of plants to recover from early UV-B exposure could have environmental implications given the episodic nature of stratospheric ozone depletions; those occurring during late winter and early spring, for example, result in unseasonably high UV-B doses during early developmental stages (e.g. Jokela et al. 1995).
Table 4. . Mean relative growth rate (RGR) for height and foliage area during the post-cross over period of growth in pea plants
The response of plants grown at low UV-B and transferred to high UV-B (LH) is less easy to interpret. The increase in UV-B at cross-over reduced growth less than was expected from the reductions found in HH plants. This was particularly marked in the development of lateral shoots, which was essentially unaffected by high UV-B during the secondary treatment (Fig. 2f). This indication that UV-B response is a function of the developmental stage at which the stress occurs is consistent with the findings of previous studies. For example, Teramura et al. (1984) reported that soybean plants were more sensitive to enhanced UV-B during the vegetative phase of growth than they were later in development. In pea cultivar Guido, the large effect of exposure to high UV-B early in development appeared to be related to the inhibition of lateral shoot growth (Fig. 1), which was far less marked if high UV-B was imposed later in development (Figs 3 & 4).
Although delayed development appeared to be the major factor resulting in reduced plant biomass at harvest, there was also a small but consistent reduction (5–15%) in the final area of fully expanded leaves and stipules (Figs 1 & 4). Leaf expansion is the consequence of cell division and cell enlargement (Cosgrove 1994), with the epidermis playing an important limiting role (Kutschera 1992). In dicotyledonous plants, a period of rapid cell division and slow expansion is followed by a period of rapid cell expansion during which leaf growth rate is maximum (Digby & Firn 1985). Both cell division and cell expansion may be affected by environmental factors. For example, drought imposed early in leaf development reduced leaf expansion in pea through reduced cell division but when imposed later inhibited cell expansion (Lecoeur et al. 1995). In pea cultivar Guido, it was clear that reduced expansion of leaf 2 under elevated UV-B was not caused by any reduction in cell expansion because the area of epidermal cells was not affected by UV-B treatments. Thus, reduced leaf area could be attributed wholly to reduced epidermal cell number (Table 3a). Similarly, in leaf 5, the number of epidermal cells per leaf area was significantly reduced by high UV-B during primary (P < 0·05) but not secondary treatment (Table 3b). The additional effect of high secondary UV-B treatment in reducing epidermal cell area in leaf 5 (P < 0·001) seems likely to be a function of UV-B effects on leaf development. This would result from incomplete expansion of leaf 5 under high UV-B at the time of measurement, which resulted in smaller epidermal cell area. Incomplete leaf expansion at the time of cross-over was also apparent in the response to secondary UV-B treatments, in which low UV-B allowed a significant increase in cell area (Table 3b). Overall, the present study indicates that the reduced area of fully expanded leaves following exposure to high UV-B is the result of a reduction in cell number. While cell division in leaf primordia was not directly quantified, we infer from our measurements of cell number that division within the leaf primordium was inhibited under elevated UV-B. The hypothesis that UV-B acts on cell division requires confirmation through more detailed microscopic analysis. However, it is notable that Dickson & Caldwell (1978) showed that UV-B altered the rate of cell division in Rumex patientia, resulting in smaller leaves and slower early growth rates, while in Petunia hybrida, increased leaf area under elevated UV-B was attributed to increased cell division (Staxén & Bornman 1994). As cell division ceases earlier in epidermal tissues than in the mesophyll, the number of epidermal cells is determined at an early stage of development of the leaf primordium (Korner & Pelaez Menendez-Riedl 1989). A direct UV-B effect on the early development of leaf primordia is unlikely in pea, where unexpanded leaves and stipules protect the shoot apex from incident radiation. Therefore, UV-B exposure of mature leaves appears to have indirect effects on meristematic tissues that are not directly exposed to incident radiation.
In conclusion, in pea cultivar Guido, growth inhibition caused by elevated UV-B appears to be partly the result of an irreversible reduction in leaf expansion but mainly attributable to a reversible inhibition of the development of individual organs (both leaves and lateral shoots) and of the whole plant. Under field conditions of fluctuating UV-B, the importance of developmental stage in determining whole-plant UV-B responses may result in marked interactions with other environmental factors that influence plant development.
We thank Jason Queally and Richard Taylor for technical assistance and Professor W. J. Davies for comments on the manuscript. This work was supported by the Government of Galicia (R.G.), and the UK Department of the Environment (Contract PECD 7/12/21) and Ministry of Agriculture, Fisheries and Food (Contract OC9218).