Effects of a UV-absorbing greenhouse covering film on tomato yield and quality

The effect of blocking the ultraviolet (UV) solar radiation using a UV-absorbing low density polyethylene (PE) film on tomato crop yield and fruit quality was evaluated in a two-year study in two arched roof greenhouses located in Central Greece. The UV-A and UV-B radiation transmission values of the greenhouse covered by the UVabsorbing PE film during the first year were 0.4% and 1.2%, respectively and increased to 0.8% and 1.3% in the 2nd year, while the respective values in the greenhouse covered by a traditional PE film were 20.7% and 12.5% during the 1st year and 28.7% and 26.7% during the 2nd year. Under the UV-absorbing film the number of insect injured fruit was reduced and the marketable yield was similar or higher than that under the common PE film, while fruit quality characteristics (size, shape), nutritional value (ascorbic acid and lycopene) and organoleptic quality (pH, titratable acidity and total soluble solids) were similar under both covering materials. Moreover, the reduction of incoming UV radiation had an appreciable effect on fruit skin color, indicating an effect on pigments other than lycopene. Additional key words: fruit quality; insects; polyethylene; transmission; ultraviolet radiation; UV-stabilizers.

Lycopene content and ascorbic acid are crucial quality parameters (Giuntini et al., 2005) but while ascorbic acid is present in all vegetables (Davey et al., 2006); lycopene is found only in red tomato and watermelon [Citrullus lanatus (Thunb.)]fruit (Bramley, 2000).Lycopene is considered one of the most important carotenoids in European and North American diet (Giuntini et al., 2005) and has attracted considerable interest as a health promoting phytochemical (Tadmor et al., 2005).
The aim of this work was to study the effects of UV-absorbing greenhouse covering films on fruit yield and quality characteristics of a common tomato cultivar (Solanum lycopersicum L., cv.Belladonna) in Greece.

Greenhouse facilities and plant material
The experiments were conducted during spring and summer 2003 and 2004 (transplanting at the end of February until July) in two similar arched roof greenhouses, N-S oriented, located at the University of Thessaly farm (39°22'N 22°44'E, altitude 85 m), on the continental area of Eastern Greece.The geometrical characteristics of each greenhouse were as follows: eaves height of 2.9 m; ridge height of 4.1 m; total width of 8 m; total length of 20 m; ground area of 160 m 2 , and volume of 524 m 3 .The greenhouses were equipped with two continuous side roll-up windows located at a height of 0.6 m above the ground with a maximum opening area of 27 m 2 (two vents of 15 m length × 0.9 m opening height) for both vents.A flap roof window was also located longitudinally on the whole greenhouse roof (20 m long) with 0.9 m maximum opening height (18 m 2 opening area).The vents were controlled automatically via a controller (Macqu, Geometions SA, Athens, Greece) and opened in steps; they began to open when greenhouse air temperature exceeded 22°C, and reached their maximum aperture when temperature reached 25°C.

Introduction
Greenhouses create an ideal environment not only for crop but also for pest and disease development.Chemicals pest and disease control are common practice for fresh vegetable protection.As an alternative, many growers use insect proof screen use, placing them in the ventilators.Thus, screening reduces ventilation (Katsoulas et al., 2006;Teitel, 2007) making the high internal temperature of Mediterranean greenhouses during summer even worse.
A complement to using insect screens is the use of a UV-absorbing film for greenhouse covering, which creates a light environment unfavourable to harmful insects.UV-absorbing films do not only block insect pests but also reduce spread of insect-borne viruses (Raviv & Antignus, 2004).Furthermore, UV-absorbing films can reduce crop diseases caused by a range of fungi that use UV as an environmental cue for sporulation (Nigel et al., 2005).
It is well documented that UV-absorbing films suppress several foliar diseases (Raviv & Antignus, 2004), but their impact on crop yield and quality still needs investigation.In Israel, no significant differences were found on growth, yield, maturation time and fresh and dry weight values of tomatoes (Solanum lycopersicum L.) grown under UV-absorbing films (Raviv & Antignus, 2004).In studies carried out in Spain, an increase in tomato yield was reported when UV-absorbing films were used (González et al., 2004;Monci et al., 2004).Kittas et al. (2006) compared the effect of UV-absorbing films on eggplant (Solanum melongena L.) crop behaviour and production.In the absence of UV radiation eggplants were taller (21%), with a larger leaf area (17%) and produced more marketable fruit yield.The adoption of UV-absorbing films by growers as an alternative technique to chemical pest and disease control seems promising as long as fruit yield and quality and concomitantly the anticipated economic return is not negatively affected.
In terms of nutritional quality, it is well established that light affects lycopene content (Cox et al., 2003) as well as ascorbic acid (Giuntini et al., 2005) and other compounds that contribute to fruit composition.Some authors report that UV radiation affects plant secondary metabolism by restricting the production of UV-absorbing compounds including flavonoids and other phenolics (Allen et al., 1998;Caldwell et al., 2003).These compounds significantly affect fruit composition and therefore fruit nutritional quality.Effects of UV-absorbing covers on tomato quality or 18°C during the day and interrupted it at 16°C during the night or at 20°C during the day.Greenhouse soil was covered by a white on black plastic sheet.
One greenhouse was covered by a common low density PE film (C-PE, Plastika Kritis S.A., Heraclion, Crete, Greece) and the second one by a UV-absorbing PE film (UV-PE, Plastika Kritis S.A., Heraclion, Crete, Greece).Both films had 180 μm thickness and equal amounts of infrared, diffusion and ethyl-vinyl-acetate additives but had different UV-absorbers content in the main PE master batch formation, which resulted in differences in UV radiation transmissivity.
The tomato crop (Lycopersicon esculentum, cv.Belladonna) was transplanted on 22/02/2003 and 25/02/2004, at the stage of 5-6 true leaves in bags filled with perlite at a density of 2.4 plants m -2 .Fertigation was automatically controlled by a computer with set points for electrical conductivity of 2.1 dS m -1 and pH of 5.6.Plants were pruned to one stem, topped after the 7 th truss and treated equally with fungicides [chlorothalonil (72-75% wettable powder, wettable granules, concentrating liquid suspension)] and miticides [clotentezine (50% concentrating liquid suspension) and propargite (30% wettable powder) for Tetranychus eggs and adults respectively] application when deemed necessary.A beehive of bumblebees Bombus terrestris (L.), was installed in each greenhouse to facilitate pollination.Plants in each greenhouse were divided into four blocks having 90 plants each.

Measurements
The following climatic data were recorded outside (on a meteorological mast) and inside (centre) of each greenhouse using sensors calibrated before the experimental period: air temperature, relative humidity, global solar radiation (model Middleton EP08-E, Brunswick Victoria, AUS), photosynthetically active radiation (model LI-190SA; Lincoln, NE, USA), UV-B radiation (290-315 nm, model SKU 430, Sky instruments LTD, UK) and UV-A radiation (315-380 nm, model SKU 420, Sky instruments LTD, UK).Each sensor was scanned every 30 s and averaged every 10 min using a Delta-T data logger system (model DL3000, Delta-T devices, Cambridge, UK).
PE film spectral transmittance measurements were made in the laboratory on three samples per PE film taken before their installation in the greenhouses using a LI-COR portable spectroradiometer (model LI-1800, LI-COR, Lincoln, NE, USA) equipped with a 10 W glass halogen lamp and an external integrating sphere (model LI-1800-12S, LI-COR, Lincoln, NE, USA) (Kittas & Baille, 1998).
Fruit harvesting took place twice a week, at the light red stage of maturity, according to the classification of Grierson & Kader (1986).Harvested fruit from 24 selected plants in each greenhouse were weighed, their dimensions were measured and the total production was calculated.Fruit were sorted into marketable and rejected production.The rejected production was due to (i) physiological problems, including small fruit (weight < 100 g), fruit with defects, breaks, scars, blossomend rot and other physiological disorders; and (ii) insects damage (was assessed in the scale 1 -5: 1 = none, 2 = slight, 3 = moderate, 4 = severe, 5 = extreme).The most prevalent insect causing fruit injuries was thrips (Frankliniella occidentalis).Thrips number was monitored every week by means of 12 blue sticky HORIVER ® traps (25 × 10 cm) and measurements on ten tomato plants (top, middle, bottom leaves) in each greenhouse (Vatsanidou et al., 2009).
Two measuring procedures were used for fruit skin colour measurements.The first one was applied both years (2003,2004) recording colour of all harvested mature fruit.The second one, applied only during 2004 experimental period, included measurement of fruit on the vein for colour evolution evaluation through the six maturity stages (green, breaker, turning, pink, light red and red).Forty two non-shaded, randomly selected fruit of the same truss with uniform shape and size were labelled in each greenhouse and their colour was measured every 1 to 3 days, according to severity of colour change, from green maturity stage until fully ripe fruit.Colour was measured by a Miniscan TM XE Plus (Hun-terLab, Hunter Associates Lab, Inc., Reston, VA, USA) tristimulus colour analyzer and colour around fruit equatorial region was recorded (6 measurements per fruit).Measurements were reported in the L*, a*, b* system [CIELAB, L* varies between light (L* = 100) and dark (L* = 0), a* varies between green (a* = -50) and red (a* = 50), and b* varies between yellow (b* = 50) and blue (b* = -50) colour].Chroma and hue (h°) values were also calculated (McGuire, 1992).
Fresh fruit were macerated in a blender for titratable acidity, total soluble solids, ascorbic acid and lycopene determinations.Total soluble solids (TSS) content was measured using a refractometer (model PR-1, Atago, Tokyo, Japan).Titratable acidity was measured by titra-tion with 0.1 N NaOH to pH 8.2 and results were expressed as % citric acid.Ascorbic acid was extracted in 1% oxalic acid and measured with reflectoquant ascorbic acid test strips (RQflex 10 Reflectoquant, Merck KGaA, Darmstadt, Germany) in a reflectometer (model 116970 Reflectometer Merck KGaA, Darmstadt, Germany).Lycopene was extracted by homogenizing 1 g sample with 25 mL of acetone in a centrifuge tube shaken in the dark.Absorbance at 503 nm was measured by means of a spectrophotometer and lycopene content was calculated using the molecular extinction coefficient of 17.2 • 10 4 L mol -1 cm -1 and expressed as mg per 100 g fruit weight (Mencarelli & Saltveit, 1988).

Data analysis
Comparison of means was performed by one-way analysis of variance at a significance level of 0.05 using the SPSS statistical package (SPSS-16, Chicago, IL, USA).

Greenhouse microclimate
The average transmission coefficients for PAR, UV-A and UV-B radiation, as calculated by the ratio of inside to outside measured value of each parameter during 2003 experimental period were 74.3%, 20.7% and 12.5% in the C-PE greenhouse and 75.5%, 0.4% and 1.2%, in the UV-PE covered greenhouse, respectively.The values changed during the replicate experimental period of 2004 to 73.0%, 28.7% and 26.7% in the C-PE greenhouse and to 72.7%, 0.8% and 1.3% in the UV-PE greenhouse.These results indicate that the UV radiation absorbing properties of the UV-PE film were maintained for two summer seasons after its placement in the greenhouse.In addition, measurements conducted in the laboratory concerning films' spectral transmittance revealed that both films had similar radiation properties in blue (400-500 nm) and green to red (500-700 nm) spectrum regions during both years (data not shown).Thus, any differences that may have emerged on crop yield and fruit quality characteristics could be ascribed to differences of the covering materials in UV-radiation transmittance.According to Krizek et al. (2005) plant productivity of greenhouse crops is greatly influenced by the amount of UV radiation, photosynthetically active radiation (PAR) and IR transmitted through the covering material of these structures which is not the case in our tested films since their transmittance in blue, green to red and PAR radiation regions was the same.
Air temperature and vapour pressure deficit (VPD) data during 2003 experimental period are presented in Table 1.Temperature and VPD values followed a similar pattern during the experimental period in both greenhouses.Similar values were also found during 2004 replicate experimental period (data not shown).
The UV-radiation energy flux corresponds to about 3.8% of the total solar radiation energy flux outside the greenhouse (Robaa, 2004).Considering that the films tested differ in UV transmission by about 28%, the difference in the incoming solar energy flux is expected to be about 1% (= 28% × 3.8%), which,

VPD i (kPa)
UV-A (kJ m -2 day -1 ) Effects of UV-absorbing covers on tomato quality for an average temperature of 20°C, may induce a 0.2°C temperature difference between the two greenhouses.
Consequently, the only difference between the environments of the two greenhouses could be considered the quality of light.

Fruit yield
No significant differences (p ≥ 0.05) were found between the two greenhouses in total yield, fruit number, marketable mean fruit weight and non-marketable fruit production ought to physiological problems (meaning zipper, blossom end rot, cutface, sunscald, small size).Under the UV-PE greenhouse, a significant decrease of fruit with insect injuries was observed during both years (Table 2).This could be explained by the significantly lower number of thrips observed in the UV-PE greenhouse, as indicated by the significantly lower number of thrips caught in the sticky traps in the UV-PE greenhouse (Table 3).It has also been observed that under UV-absorbing films some crops such as tomato (González et al., 2004;Monci et al., 2004), eggplant (Kittas et al., 2006) and lettuce (Nigel et al., 2005;Diaz et al., 2006) show better performance and yield.
The majority of the reports that concern the effect of UV-absorbing films on crop, focus on pest and disease control.They measure the efficiency of UVabsorbing films on insect population and its impact on marketable yield, although Antignus et al. (1996) mention that the effect of UV-absorbing films on crop yield due to a decrease of virus transmission may be higher than the effect on yield due to reduced insect population.Park et al. (2007) found that the number of marketable fruit decreased as thrips density increased and they also suggest that thrips fruit damage should not be evaluated in terms of yield loss, but in terms of the percentage of damaged fruit.They stated that thrips caused direct damage to fruit by feeding on and laying eggs in the developing fruit causing darkness and malformation.Similar impacts on crop yield were also observed in the present study.Rheinländer et al. (2006) stated that between thirteen species of insects and mites found on tamarillo leaves and fruit, most of these were not likely to cause any damage on fruit (e.g.spiders, lacewings ladybirds and beetles) and declared that thrips was the only insect which could cause such scarring.
The cosmetic scars that were observed in the harvested fruits of the present study were similar to those presented by Rheinländer et al. (2006), according to which thrips was the major cause for fruit damages.It was found that, compared to the C-PE greenhouse, thrips' population was 70% lower in the UV-PE than in the C-PE greenhouse, during the total experimental period and that the percentage of harvested fruit affected by thrips (injury score > 2), varied from 5% to 25% during both experimental periods.

Fruit quality
Tomato specific fruit weight (g mL -1 ) was similar in the two greenhouses ranging from 1.08 to 1.16 and from 0.93 to 0.95 for 2003 and 2004 experimental periods, respectively.Fruit shape parameters, which are components of fruit quality according to the European Commission (1983), were similar in both greenhouses (Table 4).
Table 2. Marketable yield, fruit number per greenhouse area, fruit weight and rejected production due to physiological problems such as zipper, blossom end rot, cutface, sunscald, small size, and insect damage in the tomato crop grown under the greenhouse covered by the UV-absorbing (UV-PE) and the common polyethylene (C-PE) film during the two experimental periods.Values [mean ± SE (standard error)] followed by different letters within year are significantly different (p ≤ 0.05)

Year Greenhouse
Total yield (kg m -2 ) In fruit shape formation, pollinators play a major role (Al-Attal et al., 2003).As fruit shape parameters did not differ among UV environments of the present work it could be assumed that bumblebees' activity was not affected by the UV-absorbing film.
Total soluble solids (TSS), ascorbic acid, lycopene content, pH and titratable acidity values were similar under both PE-films (Table 4).TSS were increased with time as truss number also increased under both greenhouses (Fig. 1).Similar results have been also reported by Bertin et al. (2000)The influence of low UV-B radiation levels on ascorbic acid, lycopene (Giuntini et al., 2005) and TSS content (Krizek et al., 2006) depends on tomato genotype and accordingly, the results of the present study may be valid only for the tomato cultivar studied.The non significant differences in fruit ascorbic acid and TSS content that was found in the present study may indicate a neutral reaction to UV radiation reduction of the specific cultivar tested.
Colour values (L*, hue, chroma) of harvested fruit were similar under both greenhouses and experimental periods (Fig. 2A, B and C, data for 2004 are not shown).The fruit colour is mainly attributed to the β-carotene and lycopene content (Grierson & Kader, 1986).Since lycopene content was similar under both greenhouses (Table 4) it could be assumed that the observed colour differences are attributed to the differences in β-carotene content.Hue colour parameter could be used as an objective colour index for vine ripened tomato fruit (Lopez-Camelo & Gomez, 2004).In the present work, hue values did not show any difference (Figs.2B, 2E) between the two greenhouses and accordingly, it could be concluded that tomatoes produced under the UV-PE greenhouse are not inferior of those produced under the C-PE covered greenhouse.Lastly, on the vein colour evolution evaluation through the six maturity stages (green, breaker, turning, pink, light red and red) showed that fruit under the UV-PE greenhouse had significantly higher chroma values than under the C-PE during light red and red maturity stages (Fig. 2F).These differences could be attributed to differences in synthesis rate of others than lycopene pigments.
It could be concluded that use of UV-absorbing greenhouse covering films that modify the UV radiation light environment leads to reduction of number of insect injured  Effects of UV-absorbing covers on tomato quality fruit and to similar or higher marketable yield, while fruit quality characteristics (size, shape), nutritional value (ascorbic acid and lycopene) and organoleptic quality (pH, titratable acidity, total soluble solids) are not affected.Therefore, UV-absorbing films can be considered as an effective practice in tomato crop protection from insect, without negative effects on fruit quality attributes.

Table 1 .
Monthly average values of solar radiation, air temperature and vapour pressure deficit measured outside and under the greenhouse covered by the UV-absorbing (UV-PE) and the common polyethylene (C-PE) film during 2003, for the period of day between 08:00 to 20:00.Standard deviation of the values is given in parenthesis

Figure 1 .
Figure 1.Evolution, along the crop cycle, of tomato fruit soluble solids content measured in the two greenhouses.Triangles: UV-PE greenhouse; circles: C-PE greenhouse.The vertical bars represent the standard error of means.

Figure 2 .
Figure 2. Skin colour values of harvested fruit (A, B, C) (experimental period 2003) and skin colour change on fruit on the vein (D, E, F) (experimental period 2004).Triangles: UV-PE greenhouse; circles: C-PE greenhouse.The vertical bars represent the standard error of means.

Table 3 .
Mean number of thrips caught on sticky traps per week under the greenhouse covered by the UV-absorbing (UV-PE) and the common polyethylene (C-PE) film during the two experimental periods.Values [mean ± SE (standard error)] followed by different letters within year are significantly different (p ≤ 0.05)