This paper reports osmotic effects from a series of experiments where greenhouse crops were grown in substrates with different EC values of the nutrient solution. Results of yield characteristics were published elsewhere: in the present paper, the nutrient uptake in relation to the EC will be discussed in the light of growth of lily, lettuce and kohlrabi. The crops were grown in substrates within a circulation system and thus, the water absorption of the crops could be precisely determined. The nutrient absorption was determined by tissue analysis. For lettuce whole heads were sampled and analysed; for lily, bulbs and leaves and for kohlrabi, tubers and leaves were sampled and analysed separately. The results showed a strong increase of plant nutrient concentrations in the EC-domain until the optimum growth response of the crops. Higher EC values did not affect the plant nutrient dry matter concentrations seriously, except for K concentrations of the lily crop. Calculations showed that the K concentration in the plant sap played an important role in the osmotic adjustment of the different crops to high external concentrations. The uptake concentrations of the different crops were presented and discussed in relation to the EC value in the external solution. External nutrient concentrations and plant nutrient concentrations in relation to optimum yield are presented. Calculations are made of the nutrient efficiency of a system with drainage to waste compared with a system with reuse of drainage water. INTRODUCTION Results of a study of osmotic effects on greenhouse crops grown in substrate (Sonneveld et al., 2004) outlined the relationships between the yields of a series of crops and osmotic potentials in the external solution. In the experiments, differences in the osmotic potential of the external solution were achieved by addition of nutrients. In the present paper results of nutrient uptake in three crops, in relation to the osmotic potential of the external solution, will be discussed. These include a tuber crop, a leaf crop and a bulb crop, represented by kohlrabi, lettuce and lily, respectively. METHODS AND MATERIALS The crops were grown in sand and as well in granulated rock wool substrates placed in a 0.15 m thick layer in basins with a size of 0.8 m x 1.6 m. Nutrient solutions of different concentrations were prepared and added with the aid of a sprinkler system. The irrigation time, the quantity of water and the concentration of nutrients in the water were varied according to the crop, the growing conditions and the osmotic potential aimed at. The leaching fraction in the growing system was focused on 0.25 and 0.50 and the drainage water was reused in the system. Six treatments were laid out in four parallel-randomized blocks in which the EC in the substrate solution roughly varied between 1 and 8 dS m. The composition of the nutrient solutions used was tuned to the need of the crop according to the recommendations for growers (Sonneveld and Straver, 1994). The different EC values were achieved by addition of various amounts of major elements. NH4 and P, however, were kept constant, since NH4 may influence the pH of the nutrient solution (Sonneveld, 1991) Proc. IS on Growing Media Ed.: J.-C. Michel Acta Hort. 779, ISHS 2008 314 and P can become toxic to some crops (Howell and Bernhard, 1961), when added with too high concentrations. The rough composition of the nutrient solution as used related to an EC value of 1.8 dS m was following: in mmol L NH4 1.0–1.3, K 6.0–7.5, Ca 2.8– 3.3, Mg 0.8–1.5, NO3 12.6–13.5, SO4 0.8–1.5, H2PO4 1.0–1.3 and in μmol L Fe 10–60, Mn 0–20, B 15–30, Cu 0.25–0.75, Mo 0.5–0.75. Sufficient Zn was available in the primary water, either rain water or desalinated water. The average EC of this water was 0.15 dS m and the concentrations of Na and Cl were about 0.6 mmol L. The kohlrabi crop cv. Quickstar was grown between Feb.-Apr., 1991; the lettuce between cv. Cortina Jun.-Jul., 1992; and lily cvs. Connecticut King and Star Gazer between Feb.-Jun., 1995. Both cultivars of lily were grown in the same experimental plots; thus part of the calculations could not be made for the cultivars separately. In such cases average values of both cultivars are given. The planting density was 20.3, 18.8 and 56.2 plants per m for kohlrabi, lettuce and lily respectively, in agreement with horticulture practices in The Netherlands. The concentrations of nutrients and the EC in the root environment were estimated by measurement of these parameters in the solution supplied and in the drainage water. The quantities of water supplied and drained out were recorded and used for calculation of the water absorption of the crop. Tissue samples were gathered at the end of the cropping periods and analysed for major nutrients, as well as Na and Cl. For kohlrabi tubers and tops, and for lilies bulbs and tops were gathered separately. With lettuce, representative tissue samples were taken from whole heads. The samples of the lily bulbs grown in the rock wool could not be cleaned sufficiently from these fibres and thus were not analysed. All other samples were dried at 80°C, ground and extracted by total destruction (acid digestion) to determine the element concentrations. At the beginning of the experiments, small kohlrabi plants in rock wool plugs were set out, and with lettuce small plants in peat cubes were used. The quantities of nutrients brought in with these materials were very small and considered neglected for the purpose of calculations. For lily young bulbs were planted out and tissue analysis showed that 5 to 45% of the nutrients absorbed by the crop were already available in the bulbs used at the start. These quantities were taken into account in the calculations of the nutrient uptake. RESULTS AND DISCUSSION The relationships between the yields of the crops and EC values in the root environment are shown in Fig. 1. A function of an exponential model showed the best relationship between the EC and the yield characteristics (Sonneveld et al., 2004). The maximum yield was obtained at EC values of 2.3, 2.7, 0.5 and 1.1 for kohlrabi, lettuce, lily cv. Connecticut King and lily cv. Star Gazer respectively. The water use in the experiments, covering the absorption by the crop and the evaporation by the substrate surface, was not different for the EC values of the kohlrabi and the lettuce and amounted on average 109 and 100 L m respectively. The water use of the lily decreased from 325 till 229 L m with increasing EC in the root environment from 1.2 to 6.0. This reduction in water use of 30% is in agreement with the previously reported growth reduction in this range, which was also about 30% (Sonneveld et al., 2004). The fact that kohlrabi and lettuce did not show differences in water use with increasing EC can be explained by the type of canopy formed by these crops. Both crops quickly form a closed canopy over the total area, even when there is a moderate growth reduction. In such cases the transpiration is related to the area covered. Lily, however, has an open canopy in which the transpiration will be related to the total leaf area. These results contradict the (still) generally held believe that the principal factor associated with growth reduction of crops by salinity is water availability, and support conclusions of previous workers such as Lagerwerff and Eagle (1961), and Maas and Nieman (1978), who concluded that the plants adjust for the water availability. Tissue analysis between the crops grown in sand and those grown in rock wool