Final Activity Report Summary - NAGREF+PPO (Sustainable fertigation techniques for vegetable production in Greece) Vegetable production constitutes one of the most important farming activities in Greece and has the potential to develop further. However optimal levels of vegetable production require high water inputs, especially under Mediterranean conditions. Fertigation (which means the combination of fertilisation and irrigation) is a tool for analysing and achieving optimal levels of water and fertiliser use.To optimise nutrient and water inputs, a program was developed in the Netherlands by Praktijk Onderzoek Plant & Omgeving (PPO; applied plant research - division glass; Wageningen-University & Research) in the 1960s. This program, which has evolved over time, uses a 1:2 v/v soil-water extract as a base for fertigation recommendations for a variety of vegetable crops.In total, about 5 000 growers, both in the Netherlands and abroad, are using the system. New elements have to be introduced into this system to make it appropriate for the Mediterranean situation. Due to the higher risk of soil salinisation and different soil characteristics of the Mediterranean region, modifications to this program are probably warranted.This project was undertaken to determine if the Dutch glasshouse system was affected by:1) irrigation water quality and quantity; 2) soil solution dilution in the 1:2 volume water extract; and3) soil nitrate concentrations.Problems with the extraction procedure and interpretation of results may occur due to filtration problems and gypsum in the soil which leads to overestimation of the European Commission of the soil solution. Quantity of fertilisers, and costs and benefits for the grower were also determined. The work plan consisted of evaluation of the Greek situation, workshops, contacts with laboratories agricultural universities, agricultural extension services, growers association, on-farm demonstrations, trials and introducing the system into laboratories and extension services.An experienced researcher (> 10 years of research experience) of NAGREF was trained at PPO during 3 months. A more experienced researcher (with >10 years of research experience) of PPO was hosted by NAGREF during 10 months and provided training to partner organisations. The project management structure consisted of a project committee and a project management team. The project committee included these two researchers together with: the director of NAGREF, the director of 'olive and horticultural crops' at the institute of Kalamata, a representative of an agricultural educational institute, of private laboratories, extension services and growers association. The project management team (experienced researcher of NAGREF and more experienced researcher of PPO) was responsible for project planning, monitoring progress of the project and monitoring the budget.The investigation was conducted in the Kiparisia area in the southwest part of Peloponnese, Greece (latitude 36 deg 5 N, longitude 20 deg 8 E). The study was conducted in four greenhouses; two with cucumber (Cucumis sativus L.) and two with tomato (Lycopersicon esculentum Mill.) from August 2004 to December 2004 (first season) and from January 2005 up to July 2005 (main crop season). All crops were grown on a soil with low and high lime content under plasticulture. Occasionally other greenhouse soils were analysed for comparison. Greenhouses were heated and ventilated, depending upon the season, and covered from 0.2 to 1.2 ha. Water and fertigation solutions typically used by producers were placed in glass bottles. Samples were non-acidified during transport to the laboratory. Electrical conductivity, pH, K, Ca, Mg, NH4, Na, Si, NO3, SO4, P, HCO3, Cl, Fe, Mn, Zn, B, Cu and Mo were measured using colorimetry, potentiometry, inductively coupled plasma atomic emission spectrometry, atomic absorption / emission and titration. In-line drip irrigation systems were used with one emitter per plant with a discharge rate of 4 L*hr-1. Irrigation schedules were based on grower practices, and related to solar radiation and plant size. Fertilisers were provided as a water soluble, greenhouse grade, N-P2O5-K2O-MgO (+trace elements) or as individual salts, mainly KNO3, Mg(NO3)2 or urea. Fertiliser use was recorded. Soil was plowed and rotavated. In each greenhouse 20 random soil samples, taken to a depth of 25 cm, were obtained. Fertiliser was applied and the soil again rotavated. Additional soil samples were obtained. After planting, during cultivation, soil samples were taken using an Edelman auger (dia 7 cm) under the emitters and between plant rows (25 cm from emitters) down to 20 cm below the soil surface from each greenhouse.Ten samples were taken from under the emitters and 25 cm from the emitters within each greenhouse. It was later decided to standardise the procedure by taking joint samples of equal amount under and 25 cm distant from emitters, which were mixed into one composite sample per greenhouse. Samples were taken every 2 - 3 weeks during the growing period. Soil texture was determined, and the soil analysed for pH, organic matter, lime content, P-Olsen and exchangeable cations according to the 1:2 volume water extract method, and saturated paste extract methods. Some samples were taken shortly after irrigation and were above field capacity. These samples were spread flat for air-drying for several hours, prior to analysis. Soil samples were lightly ground by hand, non-sieved, and the soil put into a transparent container filled with 200 mL of demineralised water until the total volume reached 300 mL to obtain the 1:2 (v/v) soil:water mix. Samples were left to equilibrate for 2 hrs and the resultant slurry shaken on a horizontal shaker for 20 min at 120 horizontal movements*min-1, or stirred with a mixer for 2 min at 17 rounds*sec-1. Soil extracts having a low filtration rate (i.e. low EC and high clay contents) were left overnight to facilitate filtration with suction. In the filtrate EC, pH, K, Ca, Mg, NH4, Na, NO3, SO4, P, HCO3, Cl, Mn and B were measured.Quality control of the 1:2 volume extraction methods was assessed by: a) comparison of the sum of the cations NH4, K, Ca, Mg, Na with the sum of the anions NO3, Cl, SO4, HCO3, H2PO4 in equivalents; b) comparison of the calculated EC according to the linear segment method of McNeal et al. (1970) with the measured EC; and c) comparison of the sum of the cations and anions in equivalents with the measured EC. Sub samples were dried to zero % moisture to measure mass water content. Aliquots of samples were air-dried, moistened until saturation and the EC, Na, K, Ca and Mg in the saturated paste extracts analysed. Base N fertiliser requirements were calculated so that on average it was assumed that, 1 L of extract in the 1:2 volume water extract originated from 0.625 L of bulk soil in the greenhouse. The required fertigation concentrations were calculated according to Dutch fertigation system. The N-concentration was calculated with reference to the NO3-concentration in the 1:2 volume water extract. Concentrations of K, Mg, Ca and SO4 were calculated with reference to the ratio to NO3 and the optimal ratios of these elements to NO3.The mean values and coefficients of variation were calculated for duplicates of dilution, under water density and ECs in extracts. Correlation analysis was used to test relationships between EC-measured, EC-calculated from cation and anion contents, and EC-calculated from sum of cations and anions. Results Composition of irrigation water: Well water contained Na, Ca, Mg, Si, Cl, SO4, HCO3, B and sometimes NO3. With an EC31 mg*kg-1), likely due to past fertilisation practices. Exchangeable Ca and K appeared to be different for the two soil types with apparently higher K and lower Ca in the low lime soils than in the high lime soils. In low lime soil, exchangeable K was higher (1.2-1.4 meq/100 g) than in the high lime soil (0.3-0.5 meq/100 g).Soil sampling during cultivation: Under emitters in the greenhouses of the experimental sites, elemental contents appeared to be lower than between emitters. Since the difference was large, standardisation of the sampling procedure was considered to be important, which was why the samples under the emitters and 25 cm from emitters were combined. Quality assessment of soil analysis: Coefficient of variation of the duplicates was in most cases less than 8 and 4 % for underwater density and EC, respectively. The EC calculated by the segment linear method of McNeal was 93 % of the measured EC.The sum of cations in meq*L-1 divided by 20.4 was equal to the measured EC in dS*m-1 Comparison of soil extraction methods: The relationship between EC, Ca-and Na-contents in the saturated paste and the 1:2 volume extract were significant. However, the relationship for K was not significant. The K-values in the 1:2 volume extracts were less than 0.5 mmol*L-1 in soils with high lime content, and more than 0.5 mmol*L-1 in soils with low lime content. The K-values in the saturated paste extracts varied in both soil types. Dilution of soil solution in the 1:2 volume extract: Low water contents (at field capacity) resulted, as expected, in higher dilution of soil solution than high water contents. This was seen as an advantage, because in this way the interpretation of the EC and element contents in the 1:2 volume extract could be equal for soils with variable water contents.Recommendation of base fertiliser: Most greenhouses had a higher EC than was considered optimal for tomato or cucumber growth. The high values were due to high Na, Ca, Mg, NO3, Cl and SO4 contents. At Na > 5.0 mmol*L-1, soil leaching is recommended. However, with trickle irrigation salts are leached from the root zone under emitters. Base N, K and Mg fertiliser input was calculated using data from the soil analysis before planting and fertilisation, and the target values recommended by Dutch model. For example, the average recommendation for N was 241 kg*ha-1 of N to a 25 cm depth. This roughly corresponded with the experience of the growers. The K recommendations for the low lime soils were considered high, i.e. was 401 kg*ha-1 of K2O at a 25 cm depth. High K fertilisation was needed since the high lime soils had K-fixing properties and the cation adsorption complex had to be filled with K. A recommendation for P could not be estimated from the 1:2 volume water extracts since water extracts are considered as a non-reliable estimate of plant availability of P. Since the P-Olsen values were high, there was no need for P fertilisation. Fertigation recommendation: Soil nitrogen concentration was higher before planting, and after base fertilisation, than during the growing cycle. This was caused by differences in sampling procedure, where before planting, the whole bulk soil had been sampled and during the growing cycle, the sample was a compound sample from under and between emitters. Under emitters, leaching lowered NO3 content, but these were often in the optimal range during the growing cycle. Only a few greenhouses had low or high NO3 contents. When all N, K, Ca, Mg and SO4 contents were in the optimal range the fertigation recommendation was the so-called 'standard', The K-, Ca-, Mg-, and SO4-contents in the 1:2 volume extract, were compared with the NO3-content in the 1:2 volume extract and when the ratios were in the optimal range the recommendation was not changed. When ratios in the 1:2 volume extract were not in the correct optimal range, the 'standard' was changed. For NO3 this was according to a model, and the recommendation for the EC was changed from the standard in the way that the recommended EC was higher when NO3 in the 1:2 volume extract was too low and vice versa. Since all well water contained enough B, more Ca, and in some wells more SO4, than the standard composition of the fertigation solution, no B, Ca or SO4 containing fertilisers were recommended.In practice difficulties were found in that the growers did not fertilise continuously, but used intermittent fertilisation, i.e. in some irrigation treatments, unamended well water, and in others fertiliser diluted in water. The standard recommendation could be changed so that the fertiliser dose application is alternated with unamended water (1:1). In this situation the optimal fertiliser concentration should be two times higher than the standard recommendation for continuous fertiliser dose application.Fertiliser use: Fertiliser use appeared to be greater for tomato than for cucumber crops. Fertilisation of tomato with high nutrient needs, in a preceding crop in the rotation, likely reduced the need of fertiliser in the succeeding cucumber crop with lower nutrient requirements. The input of K2O in the tomato greenhouse with soil high in lime was lower than optimal. Costs and benefits of soil analysis: Transport of samples to a laboratory, analysis and recommendation cost were considered excessive by producers. However, by reducing 12-15 kg*ha-1 of N-NO3 in a growing cycle, the cost of one analysis would be returned. It appears that K fertilisation had to be increased on K-fixing soil. Higher fertiliser is recommended than that currently used by growers to increase fruit quality and production.
