Evaluation of the Performance of an Ozonization System for the Disinfection of the Nutrient Solution of a Greenhouse Tomato Crop

An ozonization system (Ozomax OZO 6VTT, IGT limited, Ontario, Canada) was installed in a commercial plastic-covered greenhouse in Danville, Quebec, Canada (www.savoura.com) to study the effects of the disinfection and recirculation of the nutrient solution in a tomato crop. The ozonization system was compared to a conventional growing system where the nutrient solution was not treated, neither recirculated. Samples of both treated and non-treated nutrient solutions were analysed for the presence of 38 potential pathogens using DNA Multiscan Technology. Detailed measurements of the presence of Pythium spp. and Fusarium spp. in the nutrient solutions, the growing media and the root system were achieved at various periods. We also determined the evolution of bacteria in the treated and non-treated nutrient solutions. A weekly complete mineral analysis, including pH and EC of both treated and non-treated nutrient solutions was done. For both treatments, plant growth and development were measured on a weekly basis while fruit yield and quality were determined for every harvest. INTRODUCTION Protection of the underground water and of the environment requires that nutrient solutions used for greenhouse tomato production are recuperated and recirculated. Various systems using heat or UV radiation are commonly employed mainly in northern Europe to treat nutrient solutions (Runia, 1994, 2001). These systems were shown to be very efficient to eliminate pathogens in the nutrient solutions, but relatively expensive to operate (0.15-0.25 Can$/m) (Le Quillec, 2002). Other systems, using strong oxidants such as ozone, peroxide or chloride to destroy pathogens were experimented (Rey, 2001; Poncet et al., 2001; Runia, 1994, Vanacher, 1988). The addition of chloride to the nutrient solutions caused symptoms of phytotoxicity to tomato plants when concentration exceeded 2-4 ppm (Le Quillec, 2002). A French company (Trailigaz, Garges-lès-Gonesse) developed an ozonization system to disinfect nutrient solutions using ozone and peroxide. In Canada, IGT Ltd developed and commercialized an ozonization system coupled with a bubbling system to increase oxygen concentration in the nutrient solution and potentially improve its efficiency. This project aims at measuring the efficiency and the agronomic performance of the latter system. MATERIALS AND METHODS The experiment was conducted in a 27 000 m double polyethylene greenhouse equipped with a HPS supplemental lighting system supplying 120 μmol.m.s. Tomato plants Cv Trust were grown at a plant density of 2,4 plants/m in rockwool slabs (Grodan master type) placed on gutters located 1 m. high. Standard cultural practices were used for Proc. IC on Greensys Eds.: G. van Straten et al. Acta Hort. 691, ISHS 2005 390 nutrition, irrigation and plant training. Biological control of insects was primarily achieved by introducing predators such as Eretmocerus spp. and Dicyphus spp. Before being reused, the nutrient solution from 50% of the growing area was recuperated and treated with the ozonization system (Ozomax OZO 6VTT, IGT limited, Ontario, Canada) initially able to produce 40 g and later upgraded to produce 60 g of O3 per hour. The rate of nutrient solution treated was adjusted between 3 to 10 m per hour. It was consequently possible to compare the effects of 3.0 , 5.3 , 7.7 and 14.1 g of ozone per mof nutrient solution. The capacity of the ozonization system related to each treatment rate is described in Table 1. The nutrient solution of the other half of the growing area was neither recuperated, nor recirculated. In each section of the greenhouse, three bays of 337 m were selected to measure growth, development, yields and fruit quality (Dorais, 2001). Samples of the nutrient solutions were taken before and after ozonization and analysed for their mineral contents. Four additional experiments were conducted. First, we treated nutrient solutions for 30, 60, 90 or 120 minutes with ozone. Second, we sampled treated and non-treated nutrient solutions with ozone at 14.1 ppm and we later treated both of them with 0, 1, 2, 3, 4 or 5 ppm of peroxide. Thirdly, recirculated nutrient solutions were treated with both ozone (7.7 and 14.1 ppm) and peroxide at 5 ppm. Finally, nutrient solutions were treated with ozone at 7,7 ppm, peroxide at 5 ppm and chloride at 1, 2 or 5 ppm. For all four experiments, we counted bacteria on petri dishes as described further. Microbial analysis were achieved by Relab den Haan in The Netherlands, the Laboratoire de diagnostic en phytoprotection of MAPAQ and by the Centre de recherche en horticulture (CRH) at université Laval. Relab den Haan uses DNA technology to detect for the presence of 38 pathogens in the nutrient solutions. Nutrients solutions and root samples were analysed for the presence of fungi by the Laboratoire de diagnostic en phytoprotection of MAPAQ. Root segments or 100 μL of nutrient solutions were spread on three different media: Synthetic nutrient agar with antibiotics, P5ARP specific for Pythium and P5ARPH specific for Phytophthora. Bacterial counts were made at Université Laval using 50 μL of nutrient solution spread in petri dishes containing a tryptic soy agar. Nutrient solutions were sampled before the ozonization treatment, immediately after and 18 hours after in the storage reservoir. RESULTS AND DISCUSSION Data measured by Relab Den Haan from DNA analysis (Table 2) indicate that the treatment of ozonization did not eliminate all pathogens from the nutrient solutions, even at the highest concentrations. However, ozonization reduced the level of infection in the nutrient solutions for many pathogens, especially Pythium spp., Pythium dissocutum and Fusarium oxysporum. In some cases, the level of infection was not reduced at all and in a few cases, it was even increased for unknown reasons. An experiment will be conducted shortly to verify if the sand filter located before the ozonization system was the source of sporadic contamination. Analysis of fungi (data not presented) in the nutrient solutions and in roots by the Laboratoire de diagnostic en phytoprotection of MAPAQ indicated the presence of Plectosporium spp., Fusarium oxysporum, and Cladosporium spp. in similar amounts in both the treated and non-treated nutrient solutions. Fungi were also found in root segments. The increase in the duration of ozonization from 30 to 120 minutes improved the efficiency of the disinfection treatment by reducing the number of bacteria present in the nutrient solutions (Table 3). These results may be explained by either a longer period of treatment and/or an increase in the oxydo-reduction potential caused by higher concentration of ozone in the nutrient solutions. Data presented in table 4 clearly indicate the effectiveness of peroxide to reduce the number of bacteria in the nutrient solutions. The best results were obtained at the highest peroxide concentrations (4-5 ppm). Peroxide appears to be more efficient when added after the ozonization treatments (Table 5) when we sampled the treated nutrient