The effects of solar ultraviolet-B radiation on the growth and yield of barley are accompanied by increased DNA damage and antioxidant responses

Authors

  • C. A. MAZZA,

    1. IFEVA, Departamento de Ecología, Facultad de Agronomía, Universidad de Buenos Aires, Av. San Martín 4453 (1417) Buenos Aires, Argentina, and,
    Search for more papers by this author
  • D. BATTISTA,

    1. IFEVA, Departamento de Ecología, Facultad de Agronomía, Universidad de Buenos Aires, Av. San Martín 4453 (1417) Buenos Aires, Argentina, and,
    Search for more papers by this author
  • A. M. ZIMA,

    1. IFEVA, Departamento de Ecología, Facultad de Agronomía, Universidad de Buenos Aires, Av. San Martín 4453 (1417) Buenos Aires, Argentina, and,
    Search for more papers by this author
  • M. SZWARCBERG-BRACCHITTA,

    1. IFEVA, Departamento de Ecología, Facultad de Agronomía, Universidad de Buenos Aires, Av. San Martín 4453 (1417) Buenos Aires, Argentina, and,
    Search for more papers by this author
  • C. V. GIORDANO,

    1. IFEVA, Departamento de Ecología, Facultad de Agronomía, Universidad de Buenos Aires, Av. San Martín 4453 (1417) Buenos Aires, Argentina, and,
    Search for more papers by this author
  • A. ACEVEDO,

    1. IGEAF CICA-INTA (1712) Castelar, and Departamento de Ciencia y Tecnología, Universidad Nacional de Quilmes (1876) Bernal, Argentina
    Search for more papers by this author
  • A. L. SCOPEL,

    1. IFEVA, Departamento de Ecología, Facultad de Agronomía, Universidad de Buenos Aires, Av. San Martín 4453 (1417) Buenos Aires, Argentina, and,
    Search for more papers by this author
  • C. L. BALLARÉ

    1. IFEVA, Departamento de Ecología, Facultad de Agronomía, Universidad de Buenos Aires, Av. San Martín 4453 (1417) Buenos Aires, Argentina, and,
    Search for more papers by this author

C. L. BallaréFax: +54-1514-8730; e-mail: ballare@ifeva.edu.ar

Abstract

There is limited information on the impacts of present-day solar ultraviolet-B radiation (UV-B) on biomass and grain yield of field crops and on the mechanisms that confer tolerance to UV-B radiation under field conditions. We investigated the effects of solar UV-B on aspects of the biochemistry, growth and yield of barley crops using replicated field plots and two barley strains, a catalase (CAT)-deficient mutant (RPr 79/4) and its wild-type mother line (Maris Mink). Solar UV-B reduced biomass accumulation and grain yield in both strains. The effects on crop biomass accumulation tended to be more severe in RPr 79/4 (≈ 32% reduction) than in the mother line (≈ 20% reduction). Solar UV-B caused measurable DNA damage in leaf tissue, in spite of inducing a significant increase in UV-absorbing sunscreens in the two lines. Maris Mink responded to solar UV-B with increased CAT and ascorbate peroxidase (APx) activity. No effects of UV-B on total superoxide dismutase (SOD) activity were detected. Compared with the wild type, RPr 79/4 had lower CAT activity, as expected, but higher APx activity. Neither of these activities increased in response to UV-B in RPr 79/4. These results suggest that growth inhibition by solar UV-B involves DNA damage and oxidative stress, and that constitutive and UV-B-induced antioxidant capacity may play an important role in UV-B tolerance.

INTRODUCTION

The effects of ultraviolet-B (UV-B) radiation on terrestrial plants have been studied in considerable detail over the last two decades. Experiments carried out by physiological ecologists have produced information on the morphological and growth responses to UV-B of several agronomically important species under realistic field conditions [references in Caldwell et al. (1995) and Rozema et al. (1997)]. In parallel with these efforts, laboratory studies have addressed the effects of UV-B on damage to cellular components (e.g. in the form of lipid peroxidation, protein degradation, and DNA damage), and investigated adaptive biochemical responses to UV-B, such as the accumulation of UV-absorbing pigments, the increase in antioxidant enzymes, and the effects of UV on photoenzymatic DNA repair [see reviews by Bornman & Teramura (1993) and Strid, Chow & Anderson (1994)].

Several generalizations have emerged from these studies. From field experiments it has been established that ambient levels of solar UV-B received at low (Searles, Caldwell & Winter 1995), intermediate (Ballaréet al. 1996), and high latitudes (Rousseaux et al. 1998) can inhibit stem elongation and cause reductions in leaf area expansion in various plant species. Photosynthesis per unit leaf area was found to be rather insensitive to solar UV-B (or to realistic supplementation treatments) under field conditions (Beyschlag et al. 1988; Searles et al. 1995). Nonetheless, plant growth (biomass accumulation) can be inhibited by ambient UV-B as a result of UV-B-induced reductions in leaf area expansion (Ballaréet al. 1996). Growth inhibition appears to be transient in some systems, and limited to the early phases of development (Rozema et al. 1997). However, the published information on growth responses of mature plants and on the impacts of solar UV-B on crop yield is still very limited. In the reports available at present, morphological responses have been observed frequently, but conflicting data have been published regarding the effects of solar UV-B on crop yield [references in Bornman & Teramura (1993)].

Laboratory and greenhouse studies have shed some light on the mechanisms of UV-B-induced growth inhibition and UV-B protection. Evidence has been obtained for the idea that some of the morphological responses to UV-B (e.g. inhibition of hypocotyl elongation) are mediated by specific photoreceptors (Ensminger 1993; Ballaré, Barnes & Flint 1995a), and may play adaptive roles reducing UV-B exposure (Ballaréet al. 1995b). The accumulation of phenylpropanoid derivatives is a common response to UV-B; biochemical studies have shown that several key enzymes of the pathway are coordinately induced in response to UV irradiation (Beggs & Wellmann 1994), and the importance of phenylpropanoid compounds in UV screening and protection has been convincingly demonstrated using flavonoid-deficient mutants [see Landry, Chapple & Last (1995) and Reuber, Bornman & Weissenböck (1996)]. Laboratory studies also suggest that UV-B can cause oxidative damage to lipids and other cellular components (Kramer et al. 1991; Malanga & Puntarulo 1995; Takeuchi et al. 1995), and activate antioxidant defences (Willekens et al. 1994; Malanga & Puntarulo 1995; Kim et al. 1996a; Rao, Paliyath & Ormrod 1996; Takeuchi et al. 1996). As an additional damaging effect, UV-B radiation can induce lesions in DNA in the form of cyclobutane pyrimidine dimers (CPDs) and pyrimidine (6,4) pyrimidinone dimers (6,4 photoproducts) (Britt 1995).

There is a significant knowledge gap between field and laboratory studies, which has two components. First, the occurrence of certain effects of UV-B under field conditions has not yet been demonstrated. There are virtually no published studies on oxidative damage and antioxidant responses induced by solar UV-B under field conditions. Kim et al. (1996a) have shown an increase in antioxidant enzymes in rice and cucumber in response to UV-B irradiation in laboratory experiments, but UV-B supplementation studies with rice by the same group failed to detect significant antioxidant responses in the field (Kim et al. 1996b). Second, although some responses are known to occur in the field, their functional implications are still unclear. For example, although it has been demonstrated that solar UV-B increases the steady-state level of CPDs in field-grown plants (Ballaréet al. 1996), the physiological consequences of variations in the total CPD load have not been established (Taylor et al. 1996).

The field experiments that we report in this paper were designed to assess the impact of ambient UV-B on the growth and yield of barley and to narrow the gap between biochemical and physiological studies. Barley is an interesting model system for two reasons. First, commercial varieties have short stature, which facilitates studies that extend until crop maturity. Second, it provides an opportunity to study the effects of solar UV-B on a cool-season crop. This is significant because most field studies on growth and biochemical responses to solar UV-B reported thus far have been carried out in the summer. Specifically, we addressed the following questions. (1) Does springtime solar UV-B affect the morphology, biomass accumulation and yield of barley crops? (2) Does springtime solar UV-B affect DNA integrity? (3) Do barley plants respond to solar UV-B radiation with changes in antioxidant enzyme levels? (4) Is antioxidant capacity an important factor in UV-B tolerance? As a first approach to define the role of antioxidant capacity in UV-B tolerance we carried out comparative experiments that included a catalase (CAT)-deficient mutant and its wild-type mother line.

MATERIALS AND METHODS

Plant material and experimental design

All the experiments were carried out in the experimental fields of IFEVA (34°35′ S; 58°29′ W), Buenos Aires, Argentina. Seeds of the CAT-deficient barley (Hordeum vulgare L.) mutant RPr 79/4 and its normal mother line cv. Maris Mink (Kendall et al. 1983) were sown in individual 300 cm3 plastic cones filled with standard topsoil. There were two sowing dates: 10 June 1997 and 17 June 1997. Seedlings were allowed to emerge in the field under either clear polyester films (0·1 mm thick, Mylar-D, Dupont, UK), which virtually cut off all UV radiation below 310 nm [– UV-B treatment, see spectrum in Ballaréet al. (1996)], or Aclar films (0·04 mm thick, Allied Signal, Pottsville, PA, USA), which have very high transmittance over the whole UV waveband (+ UV-B treatment). Sowing was carried out in pots (rather than directly in the soil) in order to ensure uniform initial size; on 4 July 1997 [24 or 17 d after sowing (DAS)] the seedlings were transplanted to their final field location. The seedlings were planted in rows in 1 × 1 m plots; the space between the rows was 18 cm with the two genotypes planted in alternate rows; planting density was ≈ 50 m–2. For each sowing date there were three replicates of each UV-B treatment, following a randomized complete block design (total number of field plots = 12). On the north and south side of each plot there was a border row of wheat; the east and west sides were covered with curtains of clear polyester or Aclar, depending on the UV-B treatment. The filters were raised periodically to maintain them ≈ 5 cm above the barley canopy. The level of UV-B attenuation was checked with a broad-band UV-B detector (SUD/240/W attached to a IL-1700 research radiometer; International Light, Newburyport, MA, USA; peak spectral response at 290 nm; half-band width = 20 nm), and it was found to be consistently greater than 95% at the centre of the – UV-B plots. The plots were irrigated as needed to maintain the soil water content near field capacity, and weeds were controlled manually. The crops were harvested on 15 December 1997, when the plants in both treatments were yellowish and ripe.

Growth and morphology

All morphological measurements were taken on plants located at the centre of each plot. Leaf length and tiller numbers were measured several times during the growing season. On the final harvest the number of tillers, spikes, and seeds produced per plant were counted. Dry mass was obtained after oven-drying the plants at 70 °C for at least 48 h.

DNA damage analysis

For DNA damage analysis we collected the middle third of the youngest fully expanded leaf available at the time of sampling (three leaves per plot) at 1300 h. The samples were immediately frozen in liquid nitrogen. DNA extraction was carried out under dim orange light, essentially as described by Doyle & Doyle (1987) using a cetyltrimethylammonium bromide (CTAB)-based procedure modified by the use of polyvinyl polypyrrolidone (PVP) to eliminate polyphenols during DNA purification. DNA was quantitated with ethidium bromide (Gallagher 1994) using a Peltier-cooled CCD camera/imager system (Fluor-S MultiImager, Bio-Rad, Hercules, CA, USA) for fluorescence detection. DNA damage was assayed by determination of CPDs using a method adapted from Stapleton, Mori & Walbot (1993). In brief, DNA samples (≈ 3000 ng) in TE buffer were denaturated and immobilized on a charged Nylon blotting membrane (Zeta-Probe, Bio-Rad); CPDs were detected using the TDM-2 monoclonal antibody (gift from Dr Toshio Mori, Nara Medical University, Japan). The method is based on the detection of primary-bound antibody by alkaline phosphatase-conjugated secondary antibody (Bio-Rad) using a chemiluminescent substrate (CSPD®, Tropix, Bedford, MA, USA). Chemiluminescence was detected with the CCD/imager system.

Antioxidant enzymes and pigments

Samples for enzyme and pigment determinations were always collected at solar noon on sunny days. For enzyme analysis we collected the middle third of the youngest fully expanded leaf available at the time of sampling (three leaves per plot). The tissue was wrapped in aluminium foil, placed on ice and processed within 15 min of the sampling time. For determinations of total CAT and ascorbate peroxidase (APx) activities, the tissue (100 mg) was homogenized in 1 cm3 of 50 mol m–3 HEPES (pH 7·5) containing 0·1 mol m–3 ethylene diamine tetraacetic acid (EDTA), PVP, and centrifuged at 9500g for 20 min at 4 °C. Enzymes were assayed in the supernatant. CAT activity was measured spectrophotometrically by following the consumption of H2O2 at 240 nm in a reaction mixture containing 50 mol m–3 potassium phosphate buffer (pH 7·0), 24 mol m–3 H2O2 and 20 mm3 of the sample in a final volume of 600 mm3 (Aebi 1984). The reaction was recorded for 60 s after the addition of the sample. APx activity was measured spectrophotometrically by the decrease in absorbance of ascorbate at 290 nm in a reaction mixture containing 50 mol m–3 potassium phosphate buffer (pH 7·0), 0·1 mol m–3 EDTA, 0·5 mol m–3 ascorbate and 20 mm3 of the sample in a final volume of 620 mm3. The reaction was started by adding 3 mm3 of 9·8 mol m–3 H2O2 and the change in absorbance was followed for 30 s. No correction for the oxidation of ascorbate in the absence of sample was necessary (Nakano & Asada 1981). For determinations of superoxide dismutase (SOD) activity, the tissue (100 mg) was homogenized in 5 cm3 of 90% (v/v) acetone and centrifuged at 565g for 20 min at 4 °C. The pellet was resuspended in 5 cm3 of 90% (v/v) acetone and centrifuged at 4000g for 10 min at 4 °C. This step was repeated and the pellet was resuspended in 50 mol m–3 potassium phosphate buffer (pH 7·8), 0·5 mol m–3 EDTA and centrifuged at 10 000g for 15 min at 4 °C. SOD activity was assayed spectrophotometrically (550 nm) in the supernatant based on its capacity to inhibit the reduction of cytochrome c by superoxide radicals generated by xanthine–xanthine oxidase (Beuchamp & Fridovich 1971). The reaction mixture contained 50 mol m–3 potassium phosphate buffer (pH 7·8), 0·1 mol m–3 EDTA, 20 mmol m–3 cytochrome c, 53 mmol m–3 xanthine, and the amount of xanthine oxidase required to detect a change of 0·04 absorbance units min–1. One unit of SOD activity was defined as the amount of sample necessary to produce 50% inhibition of the cytochrome c reduction rate. Protein determination was carried out according to Bradford (1976).

For pigment analysis we collected four leaf discs per plot (0·45 cm diameter, youngest fully expanded leaf). Each disc was placed in 1·4 cm3 of 99:1 methanol:HCl and allowed to extract for 48 h at – 4 °C. Absorbance of the extracts was read at 305 nm or 546 nm for determinations of total UV-absorbing compounds or anthocyanins, respectively. The dry mass of the discs was determined after oven-drying at 70 °C for 4 h.

Statistical analyses were performed using PROCGLM in the SAS v 6·12 package (SAS Institute, Cary, NC, USA); appropriate transformations of the primary data were used when needed to meet the assumptions of the analysis of variance.

RESULTS

Solar UV-B affects morphological development and increases the concentration of UV-absorbing compounds

Morphological measurements taken during early vegetative growth showed clear phenotypic differences between the RPr 79/4 mutant and its normal mother line cv. Maris Mink. The leaves of the CAT-deficient mutant elongated more slowly (Fig. 1) and developed translucent lesions that became necrotic with time (not shown). Tiller production was also less profuse in the CAT mutant than in Maris Mink. Solar UV-B reduced tiller production in both lines, although the early season effects were more apparent in Maris Mink than in the RPr 79/4 mutant (Fig. 2). Specific leaf mass was ≈ 10% higher in Maris Mink than in the CAT-deficient mutant (P≤ 0·02), but it was not significantly affected by solar UV-B (data not shown).

Figure 1.

. Effects of solar ultraviolet-B (UV-B) on the elongation of the primary leaf in two lines of barley.

Figure 2.

. Effects of solar ultraviolet-B (UV-B) on tiller production in two lines of barley. Factorial analysis: T, main effect of UV-B treatment; G, main effect of genotype (i.e. Maris Mink versus RPr 79/4); Τ×G, UV-B by genotype interaction; the figures above the columns indicate the significance of the UV-B effect in each genotype, which is reported only when the relevant interaction term was significant. The final harvest data were analysed together for the two sowings.

Solar UV-B increased the content of UV-absorbing compounds per unit of leaf dry mass in both barley lines (Fig. 3, upper panel). The increase in UV-absorbing compounds did not appear to result from a general upregulation of flavonoid synthesis, as the concentration of anthocyanins per unit dry mass fell (in Maris Mink) in response to solar UV-B (Fig. 3, lower panel).

Figure 3.

. Effects of solar ultraviolet-B (UV-B) on the accumulation of UV-absorbing compounds (A305) and anthocyanins (A546) in two lines of barley. The results are for samples collected on 20 October 1997; qualitatively similar results were obtained earlier in the spring. Factorial analysis: see details in legend to Fig. 2.

Solar UV increases the CPD burden in field-grown plants

Leaf tissue harvested around noon on sunny days in late winter/early spring had measurable levels of CPDs (Fig. 4). A significant fraction of the total DNA damage was caused by the UV-B component of sunlight, as indicated by the significant difference in damage density between + UV-B and – UV-B plots. Additional experiments suggest that part of the ‘residual’ damage in – UV-B plots is caused by the UV-A component of sunlight (Szwarcberg-Bracchitta et al. in preparation). The two barley lines had similar steady-state levels of CPDs at noon, although damage density tended to be higher in RPr 79/4 (Fig. 4).

Figure 4.

. Effects of solar ultraviolet (UV) radiation on DNA damage density at midday in the youngest fully expanded leaf of two lines of barley (second sowing). Each bar represents the average of three true replicates (blocks); each replicate consisted of a pooled sample of three leaves, each from a different plant. The assay was run in triplicate. The results are for samples collected on 9 October 1997 and are adjusted for DNA mass. In Maris Mink, the + UV-B/– UV-B damage density ratio varied between 1·5 and 3·3 in early spring (data from five independent sampling dates during late September–early October). Factorial analysis: see details in legend to Fig. 2.

Solar UV-B increases CAT and APx activity in Maris Mink plants

The activities of several antioxidant enzymes were measured on fresh samples collected at midday from the youngest fully expanded leaf available at the time of sampling. Levels of SOD activity were similar in the two barley stocks, and tended to increase during the course of the growing season. There were no significant differences between UV-B treatments in SOD activity (Fig. 5). CAT activity at noon was much higher in Maris Mink than in the RPr 79/4 mutant, as expected. Exposure to solar UV-B increased the specific activity of CAT by ≈ 25% in Maris Mink, whereas in the mutant no significant response to UV-B was detected (P = 0.62) (Fig. 6). Interestingly, the differences in APx activity between lines mirrored those measured for CAT, with APx activity much higher in the RPr 79/4 mutant than in the mother line (Fig. 7). There was a significant effect of solar UV-B on APx activity, which was detectable only in the Maris Mink mother line (Fig. 7).

Figure 5.

. Effects of solar ultraviolet-B (UV-B) on total superoxide dismutase (SOD) activity at midday in the youngest fully expanded leaf of two lines of barley (second sowing). Each bar represents the average of three true replicates (blocks); each replicate consisted of a pooled sample of three leaves, each from a different plant. Factorial analysis: see details in legend to Fig. 2.

Figure 6.

. Effects of solar ultraviolet-B (UV-B) on total catalase (CAT) activity at midday in the youngest fully expanded leaf of two lines of barley. Each bar represents the average of three true replicates (blocks); each replicate consisted of a pooled sample of three leaves, each from a different plant. The results are for samples collected at noon on 27 September 1997; similar results were obtained in the winter (late July). Factorial analysis: see details in legend to Fig. 2.

Figure 7.

. Effects of solar ultraviolet-B (UV-B) on ascorbate peroxidase (APx) activity at midday in the youngest fully expanded leaf of two lines of barley. Each bar represents the average of three true replicates (blocks); each replicate consisted of a pooled sample of three leaves, each from a different plant. The results are for samples collected at noon on 27 September 1997. Factorial analysis: see details in legend to Fig. 2.

Solar UV-B reduces biomass accumulation and yield

Crop biomass at harvest was significantly larger in Maris Mink than in the RPr 79/4 mutant line (Fig. 8), and it was larger in the first sowing than in the second sowing. Solar UV-B reduced final crop biomass, and this reduction tended to be larger in the RPr 79/4 mutant (≈ 32%) than in Maris Mink (≈ 20%; Fig. 8). The changes in crop biomass were accompanied by variations in spike and grain production (Fig. 9). The reduction in grain yield caused by solar UV-B in Maris Mink varied between 17 and 31%. The yield of the RPr 79/4 mutant was very poor in the second sowing and no effects of UV-B were apparent; in the first sowing, when yields were higher, solar UV-B caused a 39% reduction in grain production in the CAT-deficient mutant line (Fig. 9, lower panels).

Figure 8.

. Effects of solar ultraviolet-B (UV-B) radiation on biomass accumulation in two lines of barley. Factorial analysis: see details in legend to Fig. 2. The primary data (x) were transformed by ln (x + 1) before analysis, and the means were back transformed for presentation. The digital photographs at the bottom of the figure illustrate the impact of solar UV-B on the growth of Maris Mink and the RPr 79/4 mutant. The plants are from the second sowing and, for each genotype, each tuft corresponds to a different block.

Figure 9.

. Effects of solar ultraviolet-B (UV-B) on spike production and grain yield in two lines of barley. Factorial analysis: see details in legend to Fig. 2.

DISCUSSION

The results of this study indicate that ambient UV-B levels can reduce biomass accumulation and yield of barley in Buenos Aires. This result is important for two reasons. First, only a limited number of UV-B-exclusion experiments with cultivated species covered the whole growing season producing data on crop yield, and few general results have emerged thus far. Only some of these studies had true replicates of the UV-B treatments; plants were generally grown in pots and the sample size was frequently small. The study of Bartholic, Halsey & Garrard (1975) was conducted on large, replicated field plots, but factors such as the apparent interaction between treatments and pest attack were not controlled, and certainly complicate the interpretation of their experiments. Saile-Mark & Tevini (1997) recently reported large reductions in the yield of bush-bean plants in response to solar UV-B; unfortunately, the UV-B-exclusion treatment had no replicates and, therefore, no statistical inference can be made from the results. However, if the results of Saile-Mark & Tevini (1997) are taken together with those reported here (Fig. 9) and the ones from the well-replicated and realistic supplementation experiments of Mepsted et al. (1996) with pea, the general picture that emerges is that ambient and moderately increased UV-B levels frequently reduce crop yield. Second, this study is the first to show that ambient UV-B can limit growth and grain production in a cool-season crop. In Buenos Aires, the daily UV-B (305 nm) integral (measured with a GUV-511 multichannel radiometer; Biospherical Instruments, San Diego, CA, USA) is about two times lower in early spring (October) than in early summer (end of December–beginning of January). In addition, the UV-B:UV-A ratio (obtained by dividing the readings of the 305 nm and 340 nm channels) is substantially lower (≈ 56%) in early spring than in the summer (Orce, Rae & Helbling 1995). In general, the differences between UV-B treatments tended to increase during the growing season (see, e.g. tiller numbers, Fig. 2) and, particularly in the first sowing of the RPr 79/4 mutant, the UV-B effects were larger in terms of seed yield than in terms of morphological variables measured earlier in the growing season, such as leaf elongation (Fig. 1) and tiller numbers at 27 DAS (data not shown). This increase in the UV-B effect may reflect the trend towards higher UV-B levels with the progress of the season, an increase in sensitivity with ontogenic development, and/or the accumulation of UV-B-induced damage. The lack of effect of UV-B on leaf expansion early in the season (Fig. 1) is consistent with the results of previous greenhouse studies with barley (Liu, Gitz & McClure 1995), but it contrasts with the results of several field studies carried out with dicots (Searles et al. 1995; Ballaréet al. 1996; Rousseaux et al. 1998), which show that leaf area expansion is frequently reduced by solar UV-B. Regarding the effects of UV-B on tillering, we found a clear inhibitory effect in barley (Fig. 2), which might be a mere consequence of the reduced biomass accumulation under ambient UV-B treatments (see Fig. 8). Previous greenhouse studies with other monocots have shown increased tillering in response to supplemental UV-B treatments that did not affect biomass accumulation (Barnes, Flint & Caldwell 1990).

Our results provide evidence for both UV-B-induced DNA damage and oxidative stress as potential causes of growth reduction. Regarding DNA damage, it has been shown previously that plants grown outdoors in the summertime contain measurable quantities of CPDs (Ballaréet al. 1996; Stapleton, Thornber & Walbot 1997). Selective filtration experiments showed that the UV-B component of sunlight was responsible for a sizable fraction of the CPD load found in leaves of the annual weed Datura ferox (Ballaréet al. 1996). Our results with barley (Fig. 4) similarly indicate that ambient levels of solar UV-B in early spring do increase DNA damage over the levels detected in plants grown under Mylar filters (Fig. 4), and that this increased DNA damage is accompanied by reductions in growth and yield (Figs 8 & 9). The damaging effect of solar UV-B on DNA occurred in spite of the significant UV-B-induced increase in UV-absorbing compounds (Fig. 3), which are known to greatly reduce UV penetration into the leaf in barley (Reuber et al. 1996).

Regarding oxidative stress, previous laboratory experiments that measured active oxygen intermediates and products of oxidative damage (e.g. carbonyl groups, lipid peroxidation) have suggested that the mechanisms of UV toxicity involve oxidative damage (Strid et al. 1994; Malanga & Puntarulo 1995). This hypothesis is consistent with the results of a recent growth chamber study that reported an apparent increase in sensitivity to high UV-B doses in an ascorbic acid-deficient mutant of Arabidopsis (Conklin, Williams & Last 1996). Furthermore, there is evidence to support the idea that reactive oxygen species are involved in the signalling mechanisms that activate some UV-induced defence genes (Green & Fluhr 1995).

Although we did not measure levels of reactive oxygen species or oxidative damage in our field experiments, we did detect a significant antioxidant response to solar UV-B in barley. Other studies, carried out under laboratory or greenhouse conditions, have also detected changes in antioxidant enzymes in UV-B-treated plants at the mRNA, protein, and activity levels, but there is virtually no information on the effects of ambient UV-B on antioxidant compounds for plants grown in the natural environment. SOD, the activity of which did not vary consistently in our experiments (Fig. 5), catalyzes the dismutation of superoxide anion into H2O2. Strid et al. (1994) reported reduced chloroplastic SOD mRNA levels in peas exposed to UV-B, whereas Willekens et al. (1994) found no effects of supplemental UV-B on four SOD transcripts in Nicotiana plumbaginifolia. Malanga & Puntarulo (1995) and Rao et al. (1996) found that artificial UV-B treatments significantly promoted total SOD activity in Chlorella vulgaris cultures and Arabidopsis thaliana (Ler ecotype) seedlings, respectively. Similar results were reported for rice and cucumber by Kim et al. (1996a), whereas in other studies the effects of artificial UV-B treatments on total SOD activity were found to vary (from promotion to inhibition) with temperature, seedling age (Takeuchi et al. 1996), and duration of UV-B treatment (Dai et al. 1997).

CAT (in the peroxisomes and mitochondria) and APx (in the cytosol and chloroplasts) catalyse the reduction of H2O2 to H2O. In our experiments both total CAT and APx activity increased in Maris Mink in response to solar UV-B; no significant effects of UV-B were found in the RPr 79/4 mutant (Figs 6 & 7). In a growth chamber study with N. plumbaginifolia, Willekens et al. (1994) found that UV-B affected the expression of three CAT transcripts (Cat1 was repressed and Cat2 and Cat3 were induced); Malanga & Puntarulo (1995) and Dai et al. (1997) reported an increase in CAT activity in response to UV-B in Ch. vulgaris and rice; in contrast Rao et al. (1996) failed to detect significant effects of UV-B on total CAT activity in A. thaliana (Ler ecotype). APx activity was found to increase with UV-B irradiation in growth chamber studies with Arabidopsis (Landry et al. 1995; Rao et al. 1996), rice (Kim et al. 1996a), and cucumber cotyledons (Takeuchi et al. 1996) and mature leaves (Kim et al. 1996a); conversely, Dai et al. (1997) reported a decrease in APx activity in rice leaves exposed to very high UV-B doses in a greenhouse study. An interesting finding in our experiments was the increased APx activity in the CAT-deficient mutant RPr 79/4 compared with the mother line (Fig. 7). Sen Gupta et al. (1993) reported increased APx activity in transgenic tobacco plants that overexpressed a Cu–Zn SOD gene. This and other observations have lent support to the idea that the expression of H2O2-scavenging enzymes (APx and CAT) is upregulated by increased H2O2 levels in the tissue (Sen Gupta et al. 1993; Prasad et al. 1994). This hypothesis is consistent with our observations (Fig. 7), because the H2O2 level is likely to be much higher in the CAT mutant RPr 79/4 than in Maris Mink.

In conclusion, our results show that solar UV-B can affect biomass accumulation and yield of barley at temperate latitudes (Figs 8 & 9). This effect on growth is accompanied by increased DNA damage (Fig. 4), and by putatively adaptive responses such as the accumulation of UV-absorbing pigments (Fig. 3) and the increase in H2O2-scavenging enzymes (Figs 6 & 7). The larger effect of solar UV-B on the CAT-deficient mutant, which was apparent in the final biomass data (Fig. 8) and in the grain yields of the first sowing (Fig. 9), suggests that constitutive and UV-B-induced antioxidant capacity (Figs 6 & 7) may be important components of UV-B tolerance in the field.

Acknowledgements

This research was supported by grants from the University of Buenos Aires (no. AG-023) and the Secretariat of Science and Technology (BID 802 OC-AR, PID no. 394) to C. L. B. and A. L. S., and also in part by a grant from the E. U. (INCO-DC: IC18-CT96–0124) to A. A. We thank Dr Toshio Mori for the antibodies used for the detection of CPDs, Dr Ann Stapleton for her advice on DNA damage analysis, and Gabriela Malanga and Dr Susana Puntarulo for their advice on the determination of antioxidant enzymes. Andrés Arakelian, M. Laura Federico, María Irianni, and Sebastián Munilla provided excellent technical assistance.

Ancillary