LETTUCE RESPONSE TO LEACHATE RECYCLING IN AN ARID ZONE GREENHOUSE

Lettuce response to salinity has been studied in open irrigation systems but not under solution recycling conditions. The objective of this study was to determine effects of threshold EC values for solution replacement on stalk weight and quality, solution chemistry and quantities of water and nitrogen removed from the system. Results showed that lettuce yield response to threshold EC in recycled solutions was similar to response to steady salinity in open irrigation systems reported in the literature. The salinity induced yield decline was stronger in the spring than in the autumn growing season. Accumulated NaCl in solution reduced Mg, Ca and K concentration in leaves in the spring experiment, but only Mg was significantly correlated with untrimmed and trimmed stalk yield. Recirculation reduced water and nutrients discharge and hence their input into the system. Increasing the threshold EC from 2 to 4-5 dS/m reduced water discharges by 60-to-100%, while the yield decline was ~10% in autumn and ~20% in spring. INTRODUCTION Leachate recycling causes salt accumulation in recirculated solutions depending on crop transpiration and fresh water electrical conductivity (EC). Salt accumulation follows a transient state that continues until a predetermined threshold EC is obtained, and this EC is maintained constant with time (steady salinity state) due to frequent solution replacement by fresh water and nutrients. Lettuce yield response to salinity was investigated under constant EC conditions in soil (van den Ende et al., 1975; Mass and Hoffman, 1977; Sonneveld, 1988), growth substrate (Bar-Yosef et al., 1993) and flowing solution (Feigin et al., 1993) but so far not under closed loop irrigation, low quality water conditions. The salinity induced yield reduction reported in the above mentioned studies varied between 3.1 and 13% per unit increase in EC (dS/m), depending on system. In addition to osmotic effects, elevated solution Na concentration was found to stimulate tip burn-rot incidence (Sonneveld, 1988) and inhibit K uptake (Feigin et al., 1991). The objectives of this work were to (i) study lettuce yield response to salt accumulation in closed loop irrigation systems, and (ii) evaluate recycling effects on nutrients uptake, transpiration and water and N discharge to the environment. MATERIALS AND METHODS The leachate recycling system (Fig. 1) was part of a climate and fertigation controlled polycarbonate greenhouse. Water losses by evapotranspiration and discharge (Vout) were replenished daily by fresh water (Vad) of known salt (Cad) and nutrients concentration. The solution was boosted from an outside solution reservoir (Vr=1500 L) by pump (P) into the greenhouse trickler irrigation system. Water treatment (TRT) comprised a 100 mesh filter. Suspended particles were removed by circulating the reservoir solution 4 times a day through a 120 mesh filter. Water meters on pumps and on fresh water inlet pipe allowed daily evapotranspiration (ET) estimation. Nutrients uptake was determined by analyzing entire plants 3-4 times along the experiment. Two experiments, both with lettuce (Lactuca sativa L.) grown in rockwool (RW) were conducted. In the first one, cultivar 9283 was planted on Oct. 31 and harvested on Dec 19 (autumn experiment). In the second (spring experiment) cultivar 9273 was planted in the used RW on Apr. 23 and harvested on May 30. Plants stand was 4.8 m grounds. Each plot consisted of seven polystyrene boxes (1.2x1x0.2 m), placed on elevated rigid plastic sheets provided with drainage canals on both sides of the bed. The boxes had five 1.5 cm openings along the walls to allow leachate drainage to the canals. Each box contained two RW slabs (1.2x0.2x0.08 m) placed 20 cm apart yielding a total RW volume of 1.34 m/plot. Number of Proc. IS on Soilless Cult. and Hydroponics Ed: M. Urrestarazu Gavilan Acta Hort. 697 ISHS 2005 244 plants and emitters (2 L h) per plot were 350 and 420, respectively. Treatments differed in threshold EC for solution replacement (Table 1). The drainage solution of 5 replicates per treatment were combined and circulated as depicted above. Fertigation frequency was 12 per day, 7 minutes per pulse. Samples from emitter and drainage solutions were collected for 24 h then analyzed for EC, pH, and pertinent ions. The target N, P, K concentrations were 10, 1, and 5 mM and those of Fe, Zn, and Mn (added as EDTA) 1.0, 0.5 and 0.5 mg L; Fe was applied as EDDHA to give an extra 1 mg Fe L too. All concentrations were elevated gradually from 50% at planting to 100% 14 d later. The deviation between target and actual concentrations was < 10%. The concentrations of Cl, Na, Ca, Mg and SO4 in fresh water were 6.2, 5.0, 1.6, 1.4, and 0.8 mM; EC, pH and HCO3 were 1.1 dS m, 7.5 and 3.5 mM, respectively. In some cases NaCl was added to the solution in order to obtain the planned threshold EC (Table 1). The replenishment solution NH4:NO3 ratio was 1:3, but due to fast nitrification, preferential NH4 uptake and large N storage (3.4 m solution/treatment) the time averaged NH4:NO3 ratio in recycled solutions was 1:10. The solution pH in the two experiments fluctuated between 7.5 and 6.0. Weather conditions are summarized in Table 2. Plants were grown according to best horticultural practices. Sampled plants were dried and element contents were determined in inner and outer leaves. Organic-N, P and K in H2SO4-H2O2 digest employing autoanalyzer and flame photometer; cations in HNO3-HClO4 digest using atomic absorption; leaf nitrate in water extract using the Merck RQflex kit. Nitrate, ammonium, P, and SO4 in recycled solution were determined by autoanalyzer; Na and K by flame photometry, and all other ions by ICP. Oxygen concentration and COD (chemical oxygen demand) were determined by YSI polarograph and HACH reactor (150C)/DR/2000 spectrophotometer, respectively. RESULTS AND DISCUSSION Chemistry of Recycled Solutions The solution EC (Fig. 2) was determined primarily by Cl and Na accumulation in recycled solutions (data not presented). The solution pH increased with increasing threshold EC because the frequency of solution replacement decreased and the time averaged NH4/NO3 ratio in solution declined (data not shown). When solution pH exceeded 6.5 (usually just before solution replacement) Mn precipitated and its concentration decreased to 0.02 mg/L or less (data not shown). Oxygen concentration in solution varied between 7.4 and 8.7 mg/L, depending on temperature only; the COD at the end of the autumn experiment was 22 mg O2 L in Tr 1 and 40 mg O2 L in treatments 2 and 3. In the spring experiment the COD in Tr 1to-4 was 37, 55, 73 and 72 mg O2 L (detailed data not presented). This indicates that dissolved organic C accumulation in solution increased as solution discharge decreased. Treatments and autumn vs. spring weather conditions had a negligible effect on total bacteria concentration in the recycled solution (0.7x10 to 4x10 CFU mL). Pythium was found in solution only in the 3 week of the spring experiment in treatments 1, 2 and 3. Plants infected by root-disease were not observed in the reported experiments. Yield and Fruit Quality Increasing the threshold EC from 2 to 4 dS/m caused no fresh or dry yield decline in the autumn experiment (Table 3). In the spring, fresh yield (Y) decreased as the threshold EC was raised from 2.5 to 5.3 dS/m (Table 3). The correlations between Yuntrimmed and Ytrimmed (g/stalk) at harvest and the steady solution EC (dS/m), represented by the time averaged EC during the last week of growth (ECe) were: Yuntrimmed = -60.26 ECe + 751.1 (R = 0.94) [1] Ytrimmed = -30.02 ECe + 503.3 (R = 0.98) [2] When the entire growing season averaged EC was the independent variable the model R values were 0.93 and 0.97, respectively (correlations not presented), indicating that the steady EC had a greater impact on yield than the transient EC. The relative decrease in trimmed stalk yield (slope divided by the control yield, 6.7% per 1 dS/m) was similar to that reported by Feigin et al (1991) for lettuce grown in flowing solution, greater than the value found by Sonneveld (1988) for greenhouse lettuce (3.1-to-