Equipment Performance for Determining Water Needs of Tomato Plants Grown in Sawdust Based Substrates and Rockwool

Transcription

1 Equipment Performance for Determining Water Needs of Tomato Plants Grown in Sawdust Based Substrates and Rockwool Martine Dorais 1, Jean Caron 2, Geneviève Bégin 1,2, André Gosselin 2, Linda Gaudreau 3 and Claudine Ménard 1 1 Agriculture and Agri-Food Canada, Horticultural Research Centre, Laval University, Quebec, Canada, G1K 7P4 2 Horticultural Research Centre, Laval University, Québec, Canada, G1K 7P4 3 Les Serres du St-Laurent inc., 700 rue Thibodeau, Portneuf, Québec, Canada G0A 2Y0 Keywords: Chl a fluorescence, load cell, photosynthesis, reflectometer, TDR, tensiometer Abstract There might be a benefit in using sawdust for replacing rockwool in terms of reducing substrate costs while maintaining an adequate productivity and root growth. However, recent research showed that water availability and transfers in sawdust were limited, thereby increasing water stress during active plant growth. To lessen this risk, the objective of this study was to characterize the relationship of substrate matric potential with volumetric water content and dielectric (TDR) sensor in a range of soilless substrates moisture that could be observed in greenhouses. In addition, we investigated whether TDR, load cell, tensiometer and canopy temperature could accurately sense plant water stress and monitor and control irrigation scheduling for tomato plants grown in sawdust based substrates. Thus, equipments were installed in three substrates (1- rockwool, 2- pure sawdust, and 3-70% sawdust + 30% wood fibers) for the production of greenhouse tomato grown on raised-gutters with supplemental lighting. Under different solar radiation in winter and spring, and different water contents of the growing media, we measured gas exchanges, Chl a fluorescence and leaf water potential. We also measured plant growth and productivity during the whole growing seasons. For sawdust + wood fiber substrate, results showed significant correlations between TDR, load cell, and tensiometer. Matric potentials were generally correlated to TDR and load cell of sawdust and rockwool substrates. The canopy temperature was a poor indicator of the matric potential but a good indicator of the substrate water content during spring. Regardless of the growing season, TDR and load cell measurements were correlated to CO 2 assimilation rate and F v /F m ratio of plant grown in sawdust based substrates and rockwool. However, relationships between water content measurements, matric potential, and physiological parameters varied with the growing media and season. No significant differences were observed for the plant development, leaf area, plant dry weight, carbohydrate partitioning and yield of tomato plants grown under different growing media. INTRODUCTION Rockwool is actually used by over 80 % of Canadian greenhouse producers, but is not recyclable and its disposal is expensive, contrary to wood waste products. Recent results (Allaire et al., 2004a,b) showed the benefits of using sawdust compared to rockwool in terms of reducing costs with equivalent yield and root growth. Hence, favouring the latter would diminish production costs and reduce environmental impacts associated with greenhouse culture. However, our results showed that easily-available water (EAW) and transfers in sawdust were limiting, which increased the risk of water stress during active plant growth (Allaire et al., 2004b). This therefore implies designing an appropriate irrigation strategy with such substrate. The irrigation strategy must be based on appropriate sensors. The irrigation management of greenhouse tomato crops is generally based on time and solar radiation with approximately, 1 ml per joule cm -2. Even though close relationship occurs between Proc. IC on Greensys Eds.: G. van Straten et al. Acta Hort. 691, ISHS

2 light and the plant evapotranspiration and consequently their water need, a time-lag between the root water supply and uptake and the physiological water need may be observed under some growing conditions. Also, using time and solar radiation remains an approximate tool as it does not take into account the substrate characteristics and hence does not ensure optimum water use. In the industry, direct and indirect approaches have been developed for an integration of plant water supply and plant water needs (Baille, 1992; Giacomelli, 1998). Substrate based measurements have been proposed as indices of EAW. Widely used in greenhouses, tensiometers measure the substrate/soil matric potential (ϕ m ) and constitute a good indicator of the substrate water availability to the plant when the substrate-water characteristic curve is known (Murray et al., 2004; da Silva et al., 1998). Sensors based on volumetric water content have also been developed like time domain reflectometry probes (TDR) and load cells (Campbell and Campbell, 1982; Noborio, 2001; Topp, 1985), and should be indirectly linked to substrate water potential, as plant activity is linked to substrate water potential in the substrate. The objective of this study was thus to characterize the relationship of substrate matric potential with volumetric water content and dielectric (TDR) sensor in a range of soilless substrates moisture that could be observed in greenhouses. In addition, we investigated whether TDR, load cell and tensiometer could accurately sense plant water stress and monitor and control irrigation scheduling for tomato plants grown in sawdust based substrates. MATERIALS AND METHODS Plant Material The study was conducted into a commercial area of 2500 m 2 at Les Serres du St- Laurent Inc. (Savoura), Portneuf, Quebec, Canada (longitude 71 W, latitude 46 N). Two tomato crops, Lycopersicon esculentum Mill. cv Trust ( plants m -2 ; plantation in August 2003 and interplanting in February 2004; 6 weeks of cohabitation) have been cultivated on raised-gutters. Supplemental lighting was provided with HPS lamps (400 W, P.L. Lighting Systems Canada Inc., 120 µmol m -2 s -1, 17-h photoperiod) when the natural photosynthetic photon flux (PPF) light level was less than 550 µmol m -2 s -1. Nine randomized experimental units measuring 278 m 2 each (complete block design with three treatments and three replicates), were used for the comparison of the water content and matric potential of three types of substrate: 1) sawdust of fresh white spruce Picea glauca (Moench) Voss. (Allaire et al., 2004a), 2) sawdust of fresh white spruce added to 30% wood fibers that favours water transfer, and 3) rockwool slabs (Grodan Master Dry type, 10 cm x 24 cm x 50 cm). Table 1 gives the substrates different characteristics. Sawdust substrates were put into 18 L bucket (45 cm x 23 cm x 18 cm) with four plants per bucket (4.5 L of substrate per plant). Drip irrigation supplied a standard nutrient solution to the plants. The solution contained in mm: 16.2 N, 1.6 P, 8.6 K, 4.9 Ca, 1.8 Mg and in µm: 28.6 Fe, 1.19 Cu, 12.2 Mn, 5.8 Zn, B and 0.7 Mo. The electrical conductivity of the solution was maintained between 2.5 to 3.5 ms cm -1 and adjusted daily according to solar radiation. The ph was also adjusted daily to For all substrates, the volume of nutrient solution given per irrigation was 100 ml per plant, and the irrigation frequency was based on intercepted solar radiation (1 ml joule -1 cm -2 ) measured on site. Carbon dioxide enrichment at 1000 µmol CO 2 mol -1 was supplied during daylight hours. Average 24 h air temperature in the greenhouse was maintained at 18.0 to 19.3 C. Under different solar radiation in winter (December 9-10; January 20-22) and spring (April 7-8) times (Table 2), different water contents of the growing media were obtained by withholding nutrient solution supply and by adding nutrient solution as shown in Figure 1. Load cells (Groscale, Priva) were installed and calibrated for each substrate treatment as recommended by Priva and connected to the Priva control system. Twelve plants per load cell system were studied. Two TDR probes (Tektronix 1502C) and four tensiometers (Omega instrument PX26) were installed in each load cell system of 294

3 each treatment and connected to CR10 dataloggers (Campbell scientific, Logan, Utah) for continuous monitoring. An infra red canopy temperature sensor (Priva) was installed within one experimental unit of the sawdust + wood fiber treatment. Measured Parameters During the two growing seasons, growth parameters were weekly measured, while plant biomass and leaf area were measured every 8 weeks. Fruits were harvested three times a week and classified for their quality every three weeks (Allaire et al., 2004b). For the experimental periods, gas exchange measurements (Licor 6400, Li-Cor, Nebraska, USA), F v /F m ratio (Handy PEA, Hansatech, UK), and leaf water potential (pressure chamber) were hourly measured between 9h00 to 15h00 (December and January) or 16h00 (April) on the 5 th leaves from the apex of each treatment with 3 to 6 replicates. Substrate water content and matric potential were simultaneously collected. ANOVA and the mean separations were performed using the protected LSD test available in SAS stat software v.8.2. (SAS Institute, Cary, NC). Correlations between equipments and physiological parameters were also performed. RESULTS AND DISCUSSION Equipment Performance Relationship between investigated methods (tensiometer, load cell and TDR) varied with the media (Table 3), most likely because of differences of the desorption curves of the different substrates (Fig. 2), and of an important hysteretic phenomena (the water desorption curves differ for drainage and rewetting). For example, in some cases, the lack of correlation between the water content and the matric potential may be due to the fact that over the large range of matric potential observed, the curve of sawdust was very flat (Fig. 2), resulting in very little changes in volumetric water content despite important changes in matric potential. The load cell of the sawdust wood fiber substrate was strongly correlated to the TDR measurements (both are linked to the substrate volumetric water content) and in pure sawdust and rockwool for the winter period. Matric potential of all substrates was significantly correlated to the water content expressed by TDR or load cell, but the correlation was not significant for the year round load cell, likely due the highly non linear shape of the water desorption curve and the hysteretic phenomenon. The canopy temperature was closely related to the TDR and load cell (substrate water content), particularly during the spring. However, no significant correlation was observed during the winter period when the solar radiation is the limiting growing factor. Correlation with Physiological Parameters Our results suggest that in general, TDR and load cell measurements could be a good indicator of the CO 2 assimilation rate and F v /F m ratio of plant grown in sawdust based substrates and rockwool (Table 4). In winter, TDR, load cell, and matric potential measurements of all substrates were correlated to leaf stomatal conductivity, except for load cell and matric potential in rockwool. In spring, the water content of all substrates expressed by load cell was correlated with CO 2 assimilation rate, while the leaf stomatal conductivity was correlated to the water content of sawdust and rockwool substrates. For the sawdust + wood fiber substrate, leaf water potential was significantly correlated to the water content and matric potential (Fig. 3). In general, water content indicated by TDR was significantly correlated to physiological parameters most of the time and more often than the load cell. Tensiometers data appeared the worst predictor, a fact most likely due to the very sharp decrease of the water desorption curve in a very narrow range of matric potential (Fig. 2). This then results in a steep decrease in plant physiological parameters resulting in the observation of partial wilting without any further decrease in substrate water potential. This non linear response was observed for most parameters for sawdust and rockwool, consistent with their flat water desorption curve (Fig. 2), Thus, water 295

4 content measurements could be used as a predictive tool for water management of crops grown in all substrates investigated, but the growing period should be considered to improve their accuracy. In agreement with previous studies (Fracheboud and Leipner, 2003), water content of the media did not induce a strong decrease in the PSII photochemistry judged by Fv/Fm of dark adapted leaves, even though slight wilting became apparent under matric potential of -12 to -15 cm for sawdust as well as rockwool, and -25 cm for the sawdust wood fiber medium. The Chl a fluorescence curves of dark adapted leaves (Kautsky curves) showed that under moderate water stress conditions (-22 cm for the sawdust wood fiber medium; -8 cm for sawdust and -12 cm for rockwool), F 0 and F m of plants grown in sawdust and sawdust wood fiber substrates were similar or slightly lower than plants grown on rockwool (data not shown), which may be related to a higher nonphotochemical quenching. Growth and Yield For both crops, no significant difference was observed for the plant development expressed by weekly stem growth and diameter, leaf length and fruit set (data not shown), as well as leaf area, plant dry weight and carbohydrate partitioning of tomato plants grown under different growing media (Table 5). However, fruit dry weight and the electrical conductivity of the fruit juice of tomato grown on sawdust + wood fiber substrate were significantly higher than fruits from plants cultivated in rockwool (data not shown). Even though we observed under high light or CO 2 levels lower photosynthetic capacity of plants cultivated into pure sawdust as compared to sawdust wood fibers and rockwool (Fig. 4), no significant difference (p 0.05) was observed during the winter and spring times for tomato yield, fruit size and fruit external quality (Table 6). In conclusion, relationships between water content measurements, performed either with load cell or TDR probes, and physiological parameters vary with the growing media and season. However, the relationship between physiological parameters and substrate water content is not unique because of different causes: 1) differences in the desorption curve of the different media, along with the hysteretic phenomena of the curves themselves and 2) the physical evolution of the organic substrates after the first three months of growth (settling, compaction and biological degradation), which lead to a 30% decrease of the substrate volume. Hence, this suggests that frequent calibration of the water content-plant parameters relationship may be needed for a good accuracy. Additional experiments are ongoing to investigate further this aspect. Literature Cited Allaire, S.E., Caron, J., Ménard, C. and Dorais, M. 2004a. Growing media varying in particle size and shape for greenhouse tomato. Acta Hort. 644: Allaire, S.E., Caron, J., Ménard, C. and Dorais, M. 2004b. Growing media as a replacement of rockwool for greenhouse tomato production. Can J. Soil Sci. 85. Baille, A Water status monitoring in greenhouse crops. Acta Hort. 304: Campbell, G.S. and Campbell M.D Irrigation scheduling using moisture measurements: theory and practice. Adv. Irr. 1: Fracheboud, Y. and Leipner, J The application of chlorophyll fluorescence to study light, temperature, and drought stress. In: Practical applications of chlorophyll fluorescence in plant biology. J.R. DeEll and P.M.A. Toivonen Ed., Kluwer Academic Publishers. Giacomelli, G.A Monitoring plant water requirements within integrated crop production systems. Acta Hort. 458: Murray, J.D., Lea-Cox, J.D. and Ross, D.S Time domain reflectometry accurately monitors and controls irrigation water applications in soilless substrates. Acta Hort. 633:

5 Noborio, K Measurement of soil water content and electrical conductivity by time domain reflectometry: a review. Computers and Electronics in Agriculture 31: Silva da, F.F., Wallach, R., Polak, A. and Chen, Y Measuring water content of soil substitutes with time-domain reflectometry (TDR). J. Amer. Soc. Hort. Sci. 123: Topp, G.C Time domain reflectometry (TDR) and its application to irrigation scheduling. Adv. Irr. 3: ACKNOWLEDGEMENTS The authors thank Agriculture and Agri-Food Canada and Les Serre du St-Laurent Inc. (Savoura) for their financial support as well as the technical staff, Jocelyne Moreau and Audrey Boulianne, at Savoura. We are grateful to Priva for their technical support and collaboration. 297

6 Tables Table 1 Substrate physical properties at the beginning of both crops. θ s (% vol) AFP (% vol) EAW (% vol) Sawdust + wood fibers Sawdust Rockwool θ s = water content at ϕ=-0.1 cm; AFP (Air Filled Porosity): air content at ϕ = -10 cm EAW (Easily Available Water): difference between water content at ϕ=-10 cm and water content at ϕ=-50 cm Table 2. Climate conditions in the greenhouse during the 3 experimental periods. Solar radiation PPF* (µmol m -2 s -1 ) Temperature ( C) HR CO 2 ** (µl l -1 ) (J cm -2 day -1 ) min max average min max average average December January April * Data are means of PPF measured outside the greenhouse at 9:00 to 15:00. **Data are means of CO 2 concentrations measured inside the greenhouse at 9:00 to 15:

7 Table 3. Correlation coefficients between TDR, load cell, tensiometer and infra red canopy temperature during the whole year, winter and spring. Year-round Winter Spring T C Load cell ϕ m T C Load cell ϕ m T C Load cell ϕ m Sawdust + wood fibers n=113 n=42 n=71 TDR 0.21* 0.81*** 0.55*** *** 0.48** 0.55** 0.81*** 0.55*** Load cell 0.20* ** *** 0.61** * ϕ m Sawdust n=105 n=34 n=71 TDR *** *** 0.78*** *** Load cell *** Rockwool n=100 n=29 n=71 TDR ** *** 0.41* * Load cell ** *** * P 0.10; *P 0.05; ** P 0.01; *** P

8 300 Table 4. Correlation coefficients between TDR, load cell, tensiometer and physiological parameters during the whole year, winter and spring. CO2 assimilation rate (µmol m -2 s -1 ) Year-round Winter Spring gs (cm s -1 ) Fv/Fm CO2 assimilation rate (µmol m -2 s -1 ) gs (cm s -1 ) Fv/Fm CO2 assimilation rate (µmol m -2 s -1 ) gs (cm s -1 ) Fv/Fm Sawdust + wood fibers n=113 n=42 n=71 TDR ** 0.71*** 0.46** * Load cell * ** *** ϕm * Sawdust n=105 n=34 n=71 TDR 0.20* 0.27** 0.29** 0.76*** 0.62*** 0.43** * 0.15 Load cell 0.52*** * * ϕm * 0.29** 0.32* 0.39* 0.45** Rockwool n=100 n=29 n=71 TDR 0.33*** 0.35*** 0.29* *** 0.48** *** 0.17 Load cell 0.39*** * *** 0.24* 0.25* ϕm *** P 0.10; *P 0.05; ** P 0.01; *** P

9 Table 5. Leaf area, plant dry weight and carbohydrate partitioning of tomato plants grown under different growing media. Samples were collected in December (first crop) and April (second crop) (n=9). Leaf area Plant dry weight Carbohydrate partitioning (cm 2 /plant) (g/plant) Stem Leaves Fruits December Sawdust + wood fibers Sawdust Rockwool P 0.05 April Sawdust + wood fibers Sawdust Rockwool P 0.05 Table 6. Yield of tomato plant (kg plant -1 ) cultivated in different growing media (week 42 to week 17). Data are means of three replicates. Total (kg plant -1 ) #1 size (g) #2 #3 #4 BER CC Sawdust + wood fibers Sawdust Rockwool ANOVA (P 0.05) (p=0.68) (p=0.32) (p=0.34) (p=0.42) (p=0.27) (p=0.48) (p=0.23) (p=0.70)

10 Figurese Fig. 1. Relationship between matric potential of each substrate and time during the experiment period of April. Similar patterns were observed for December and January experimental periods. 1 Volumetric water content (cm 3 cm -3 ) Sawdust Sawdust + Wood fibers Rockwool Peat 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0, Matric potential (cm) 0,1 Fig. 2. Typical desorption curves of soilless substrates measured by Caron (unpublished). 302

11 Leaf water potential (-bars) A) r=0.61* TDR (v/v) B) r=0.66** C) r=0.53* Load cell (kg) ,5 3 2, Matric potential (mbar) Fig. 3. Relationships between leaf water potential of tomato plant grown in sawdust + wood fiber ( ), sawdust (Ο) and rockwool substrate ( ), and their water content expressed by A) TDR (v/v), B) load cell (kg) and C) matric potential (mbar) during the whole growing season. Dash lines represent the correlation of the sawdust + wood fiber substrate. Leaf water potential (-bars) 303

12 Photosynthetic rate (µmol CO 2 m -2 s -1 ) A) B) Sawdust + Wood fiber o Sawdust X Rockwool PPF (µmol m -2 s -1 ) CO 2 (µl L -1 ) Fig. 4. Light and CO 2 saturation curves of tomato leaves cultivated in different growing media. 304