minimized oxygen stress in closed nutrient systems
TRANSCRIPT
M P R A K T I J K O N D E R Z O E K P L A N T & O M G E V I N G
Minimized oxygen stress in closed nutrient systems enabling high flux nutrient supply
FINAL report of project no. 97/8 funded by the Dutch-Israeli Agricultural Research Program
(DIARP)
Rob Baas, Dick van den Berg
Applied Plant Research (PPO), Division Glasshouse Horticulture,
Linnaeuslaan 2A, 1431 JV Aalsmeer The Netherlands
Michael Raviv, Shlomit Medina and Arkady Krasnovsky
Agricultural Research Organization, Newe Ya'ar Research Center, POB 1021, Ramat Yishay,
30095, Israel
PPO 551
€ 64,50
Applied Plant Research Sector Glasshouse Crops November 2001
W A G E N I N G E N
Table of contents page
SUMMARY 5 The Netherlands 5 Israel 6
1 OBJECTIVES OF THE ORIGINAL RESEARCH PROPOSAL 6
2 THE NETHERLANDS 7 2.1 MATERIALS AND METHODS 7
2.1.1 Cultivation system 7 2.1.2 Experimental designs 7 2.1.3 General information. 8 2.1.4 Determinations root environment 8 2.1.5 Determinations of plant parameters 9
2.2 RESULTS 9
2.2.1 Experiment 1 (week 40 1998 - week 24 1999) 9 2.2.2 Experiment 2 (week 25 1999 - week 4 2000) 15 2.2.3 Experiment 3 (week 5 - week 32 2000) 17 2.2.4 Combined experiments (week 40 1998-week 32 2000) 20
3 ISRAEL 24 3.1 MATERIALS AND METHODS 24
3.1.1 Hydroponic experiments 24 3.1.2 Growing medium experiment 27
3.2 RESULTS 28
3.2.2 Hydroponic experiments 28 3.2.3 Growing media experiment 41
4 DISCUSSION AND CONCLUSIONS 43 4.1 THE NETHERLANDS 43
4.1.1 Growing media effects 43 4.1.2 EC and NaCI effects 43 4.1.3 Discharge reduction possibilities 44 4.1.4 Transpiration, Water use efficiency 44
4.2 ISRAEL 45
4.2.2 Physical characteristics of growing media 45 4.2.3 Salinity effects on plant performance 45 4.2.3 Flow rate effects on plant performance 46 4.2.4 Discharge reduction possibilities 46
5 DESCRIPTION OF COOPERATION 47
6 EVALUATION OF THE RESEARCH ACHIEVEMENTS WITH RESPECT TO THE AIMS OF THE ORIGINAL RESEARCH PROPOSAL 48
6.1 NETHERLANDS 48
6.2 ISRAEL 48
7 REFERENCES 50
8 APPENDIX 51
Summary The project was aimed at high-flux irrigation in order to allow higher concentrations of non-nutritional elements such as Na and CI.
The Netherlands In the first experiment with cut rose 'Frisco' (week 40 1998- week 24 1999) the extent of oxygen deficiency under high-flux irrigation conditions was investigated. Experimental variables were growing media and irrigation frequencies. Growing media used in the 3 liter containers (height 15 cm) were coir, and three perlite fractions. Physical determinations were determined before and during the experiment, showing the air-filled porosity (AFP) to range from on average 31% (fine perlite), 33% (coir), to 53% (perlite mid), and 61% (perlite coarse). Irrigation amounts were very high compared to normal horticultural practice, i.e. approximately 1, 1.9 and 3.8 I per plant/container per day. In addition to the time-based irrigation control, F(requency) D(omain) sensors monitored permittivity (related to water content) in the coir treatments, and were used to trigger irrigation in case a setpoint was reached. This resulted in increased irrigation frequencies at higher transpiration in the low-frequency treatments. Since transpiration (determined on weekly basis) was between ca. 0.1 and 0.5 I per plant per day, drain percentages were on average 80, 87 and 94 % in the irrigation treatments. EC was decreased from 1.9 to 1 mS/cm in week 6. Despite the low concentrations, no visual nutrient deficiencies occurred, although average nutrient uptake concentrations - as determined by fertilizer addition - were close to concentrations in the nutrient solution. No effects of the irrigation frequencies, or interaction of irrigation frequency with growing media were
found on production. Total FW production was lowest in the coarse perlite treatment, due to a lower number of stems produced per plant. The lower above-ground production coincided with reduced root growth. It was suggested that the decreased root growth is the result of higher penetration resistance in coarse perlite. Highest root and stem production but lowest stem weight occurred in the coir and fine perlite treatment. In vitro root ADH measurements performed during the experiment showed large differences between harvest dates, but generally activity was higher at lower AFP. The results indicated that oxygen deficiency was not a problem under the given circumstances (irrigation rates/growing media).
A refinement of the experiment was conducted from week 25 1999 until week 4 2000. In this experiment the effect of irrigation frequencies, growing media and EC (0.9 and 1.9 mS/cm) of the nutrient solution on production, transpiration and nutrient uptake were investigated. The underlying hypothesis was that by decreasing nutrient concentrations, nutrient discharge of elements could be decreased, both in open and in closed nutrient systems. The average drainage percentages varied between 62% for the low (8x/day), up to 95% for the high (32x/day) irrigation frequency. Permittivity (water content) increased slightly during the experiment. No effects of irrigation frequency on production, transpiration or mineral nutrient concentrations were found. Production was also not significantly influenced by EC 0.9 compared to EC 2, although leaf concentrations of K, N and P were lower in this treatment. The result offered perspectives for lowering the EC under commercial growing conditions.
In the final experiment which lasted from week 5 to week 32 2000, one irrigation frequency (16x/day) was used. Gas analysis of air samples taken from the rhizosphere confirmed that aeration was sufficient in all the growing media. Experimental variables were EC (0.9 and 1.9 ) and NaCI addition (0 and 10 mM) in order to simulate Na accumulating conditions. The lower production in coarse perlite persisted in the last year of the project. There was a significant (negative) effect of NaCI on stem weight and total weight production. However, this was only apparent in the low EC treatment. Without NaCI in the nutrient solution, stem weight was highest at low EC. The apparent negative effect of NaCI at low nutrient-EC was particularly related to higher CI concentrations in the leaves. The negative effect of NaCI accumulation on production was not found at EC 1.9, which agreed with previous results. The project shows that NaCI concentrations up to 10 mM can be tolerated without problems provided the major nutritional elements are not below certain limiting concentrations. It was calculated that an increase in Na concentration tolerated to 10 mM reduces emission
from closed systems to 60 % compared to the currently used 4 mM.
Israel In the first year (Experiment A), rose plants (CV. Frisco) were grown in closed, recirculated systems, in three size fractions of tuff and two size fractions of perlite at high flow rates (86 l/plant.day). Physical characteristics of the media, including particle size distribution, water release curve, saturated and unsaturated and hydraulic conductivities were determined. Perlite-grown plants suffered slightly from short events of low pH and low calcium availability (as manifested in Ca content of the collected plant material). The performance of tuff-grown plants was closely related to their physical traits. Very low AFP and low hydraulic conductivity characterize fine tuff while very low available water content and low unsaturated hydraulic conductivity characterizes coarse tuff. Tuff of a medium particle size has adequate AFP, available water content and hydraulic conductivity. Consequently, it yielded the highest yields and its photon quantum efficiency was the highest. The results of the first year (experiment A) suggest that a priori physical characterization of growing media, especially as related to AFP and unsaturated hydraulic conductivity, can be used to predict potential productivity of substrates.
Coarse tuff and perlite and medium tuff were compared under high flow rate (Experiment B). Low nutrient concentration was applied, as compared to the first experiment, assuming that the high flow rate (86 l/plant.day) will enable sufficient nutrient uptake. This was clearly not the case, suggesting that lowering salinity through reduced nutrient concentration is not feasible under the experimental conditions. Perlite-grown plants had slightly better gas exchange rate and water relations than tuff-grown plants. These advantages were not expressed, however, in yield terms due, probably to the untimely interruption of this experiment and/or due to the more severe effect of nutrient deficiency on perlite vs. tuff. In another hydroponic trial (Experiment C) two levels of salinity (1.1 and 4.4 mS/cm) were compared under two flow rates (corresponding to 6 and 144 l/plant.day). The experiment was done in coarse perlite, in order to avoid nutritional effects, typical to tuff. High salinity did not affect carbon exchange rate, while high flow rate increased it, as well as oxygen concentration. High flow rate also increased stomatal conductance and specific transpiration rate under both low and high salinity levels. Nutrient uptake was also promoted by high flow rate, on the background of low (but not of high) salinity. Under high salinity conditions high flow rate reduced CI uptake. In spite of the marginal effects of high salinity on gas exchange parameters of rose plants, it inhibited expansive growth, leaf area and yield. High flow rate, although positively affecting most nutritional and gas exchange parameters, could not alleviate the negative salinity effects on yield of rcses. It is possible that a longer period is required for rose plants, in order to adapt to high salinity conditions. This can be interpreted by the minimal effect of salinity on plants grown in the semi-commercial experiment (see below). In hydroponic experiment D, coarse perlite was tested under 3 levels of salinity and 3 flow rates (3-72 l/plant.day). Similarly to the results obtained in experiment C, high salinity did not affect carbon exchange rate, while high flow rate increased it. High flow rate also increased stomatal conductance and specific transpiration rate. Nutrient uptake was also promoted by high flow rate, on the background of low (but not of high) salinity. High salinity, however, inhibited expansive growth, resulted with lower leaf area and yield.
Besides the hydroponic experiments, a semi-commercial medium experiment was conducted during two years. It was hypothesized that increased flow rate of recycled solution, combined with increased amount of applied water (to compensate for the assumed increase in water transpiration that accompanies higher water availability) should allow normal yield production similar to non-recycled control. This assumption was verified. A clear reduction in water and fertilizers input was achieved. It was found that a decrease of -30-35% of water and fertilizer consumption is possible, with resulted savings of ~$1 m2 year1. The use of water recirculation is now steadily growing in Israel.
1 Objectives of the original research proposal
In order to decrease emission of polluting nutrient elements such as N and P from glasshouse horticulture, the following sequence of events may be envisaged to occur.
PROBLEM
Nutrient + water loss =>
=>Na (and CI) accumulation
=> ion-toxicity
=> salinity
=> nutrient deficiency =>
=> hypoxic conditions =>
SOLUTION
=> Recirculate nutrient solution =>
=>determine threshold concentrations
=>decrease nutrient concentrations =>
=> increase irrigation frequency =>
=> optimize oxygen supply medium
According to this scheme, prerequisites for minimizing the nutrient discharge of polluting elements such as N and P are that oxygen and nutrient supply should be optimized in closed nutrient systems.
The original hypotheses of the research proposal were as follows: Oxygen availability in growing media can be optimized to meet demand by plant roots. At high irrigation rates and non-limiting oxygen supply, lower nutrient concentrations and higher sodium and/or chloride concentrations can be tolerated than at low-irrigation rates.
The specific objectives that were derived from the above hypotheses were:
1. To determine crop-specific optimal physical characteristics (particularly air-filled porosity) of a growing medium to be used under high-flux saline fertigation conditions by using parameters indicating oxygen deficiency.
2. To determine threshold conditions which limit (physiological) plant processes related to growing conditions in closed nutrient systems.
As a model crop cut rose was used during the three-year project.
2 The Netherlands
2.1 Materials and methods
2.1.1 Cultivation system
The experimental site consisted of 6 rolling benches (length 12 meter, width 1.2 m) in a 150-m2 greenhouse. On each bench two gutters were fixed on which containers (diameter 15-20 cm, height 15 cm, volume 31) filled with growing medium were placed. Nutrient solution was supplied with one dripper per container. The drained nutrient solution was collected in a 500-liter circulation tank. The containers were covered with plastic to avoid evaporation. Each experimental unit had a separate circulation tank and consisted of two rows of 6 m and 60 plants at the start of the experiment. Set-point temperature in the greenhouse was 18°C day/night during the experiment. C02 was supplied at 1000 ppm in case no
ventilation took place. Assimilation lighting was used until May Kweek 18) if the outside radiation was less than 100 W/m2 (=100 J/s.m2). No assimilation light was supplied from 8 p.m. until 4 a.m. Through holes in the plastic, rose cuttings of c.v. Frisco propagated in coir dust cylinders were planted in the growing media in week 40 1998 (5 plants per m).
2.1.2 Experimental designs
During the project, the same plants were used in three different experiments.
2.1.2.1 Experiment 1 The experiment lasted from week 40 1998 until week 24 1999 in order to determine the extent oxygen deficiency in the roots may occur under high-flux irrigation conditions. Experimental variables were 4 growing media (coir dust, perlite fine, mid and coarse) * 3 irrigation frequencies (8, 16, 32x/day, corresponding with 960, 1920 and 3840 ml per plant per day). There were 4 replicates per treatment. Some physical characteristics of the media are given below.
Table 1. Physical characteristics of Growing medium
coir perlite fine (0-1 mm) perlite mid (0.6-2.5 mm) perlite coarse (1-7.5 mm)
bulk density (kg/m3) 95 89
75
157
porosity (% v/v)
94 • 97
97
94
the tested g water -3.2 cm (% v/v) 86 79
56
40
rowing media. water -10 cm (% v/v) 80 78
40
30
water -32 cm (% v/v) 49 51
31
24
water -50 cm (% v/v) 45 42
27
21
Water -100 cm (% v/v) 40 36
21
18
org. matter (% w/w)
83 0
0
1
2.1.2.2 Experiment 2 The experiment was conducted from week 25 1999 until week 4 2000. Experimental variables were 3 irrigation frequencies * 4 growing media * 2 EC's of the nutrient solution (0.9 and 1.9 mS/cm). Irrigation frequencies and growing media were as in experiment 1. There were two replicates per treatment.
2.1.2.3 Experiments The experiment lasted from week 5 2000 until week 32 2000. Experimental variables were 4 growing media * 2 EC' s of the nutrient solution (0.9 and 1.9 mS/cm). * 2 NaCI concentrations (0 and 10 mM). The addition of NaCI resulted in increased EC until at most 2.9 mS/cm. There were 3 replicates per treatment. Irrigation frequency was 16x/day.
2.1.3 General information.
In Appendix 1 the layout of the greenhouse and the experimental plots in experiment 3 is given. The EC and pH of nutrient solution in the circulation tanks was adjusted every week, and nutrient analysis was performed every two weeks. The amount of water and nutrients added to adjust to the target EC was recorded on a weekly basis. Target concentrations of major nutrients at 1.9 mS/cm were: (mM) NH4 0.1; K 5; Ca 5; Mg 3; N0312.5; S04 3; P 0.9. At 0.9 mS/cm the ratio between all elements was kept the same as at 1.9 mS/cm. Target concentrations of minor element were always (uM): Fe 25; Mn 3; Zn 3.5; B 10; Cu 1; Mo 0.5. Growing media used were commercially available coir dust and perlite fractions as described in Table 1. Irrigation frequencies were maintained primarily by a time-based control as follows: lx/180 min, lx/90 min, lx/45 min with 35-40ml/min drippers. Duration of each irrigation was 1 minute per irrigation up to week 51. Since comparison of EC and pH between tank and medium solution showed unwanted deviations, irrigation was increased to 3 minutes per irrigation throughout the rest of the experiment. Total irrigation per container/plant therefore was 960 ml, 1920 ml and 3840 ml per day from week 51. In addition to the time-based control, Frequency Domain (FD) sensors (electrode-length 6.5 cm, IMAG-DLO, The Netherlands) were positioned vertically at half-height in the containers of the coir dust treatments (see Appendix 1). In case a threshold permittivitity was reached, irrigation started. This system therefore ensured irrigation on top of the time-based control.
2.1.4 Determinations root environment
2.1.2.1 Permittivity and EC Permittivity and EC readings with FD sensors (IMAG-DLO) were taken every 10 minutes in 18 pots filled with coir dust (6 replicates per irrigation frequency) during the experiment. Permittivity was related to water content after calibration in 15 cm high cylinders filled with coir.
EC and pH was also recorded weekly in the recirculation tanks.
2.1.2.2 Gas composition rhizosphere In order to obtain information on the aeration status of the growing media during the last year of the project, gas samples were taken. Perforated PVC tubes (volume 3.8 ml) were brought horizontally into the growing medium at 5 cm from the bottom of the containers. The tubes were closed with a septum at the end, which allowed to take gas samples with a syringe. The samples (5 replicates per growing medium) were analyzed for C02 and 02 using gas-chromatography.
2.1.2.3 Root weight Root weight was determined during experiment 1. Since the root samples were polluted with perlite particles, root dry weight of week 21 was determined by determining the organic matter content of the dried samples (at 70°C for 48 h) by heating at 450°C.
2.1.2.4 ADH The activity of in vitro alcohol dehydrogenase activity in root samples taken at different harvests in experiment 1 was determined as described by Baas et al. (1995, modified 1997).
8
Leaf analysis was performed a number of times during the project. Only young fully expanded 5-leaflet leaves were harvested and analyzed for Ca, K, Mg, P and total-N. Occasionally N03l Na and CI were analyzed.
2.1.5 Determinations of plant parameters
2.1.2.1 Transpiration Transpiration was determined on a weekly basis by determining the amount of nutrient solution needed to fill up the recirculation tanks to a predetermined level. Since the different growing media were combined per recirculation tank, no distinction could be made in transpiration between the growing media.
2.1.2.2 Production Production was determined up to three times per week by weighing and counting the harvested stems per plot.
2.2 Results
2.2.1 Experiment 1 (week 40 1998 - week 24 1999)
2.2.1.1 Root environment
2.2.1.1.1 water content The results in Table 2 show that large differences in water content between the growing media existed; no significant differences between the frequencies were apparent however. The results fit - in comparison with the laboratory results - in the -10 to -32 cm range. From the results and the porosity data (Table 1) the air-filled porosity (AFP) was calculated (Fig. 1). Since it is known that perlite contains closed pores, which apparently do not contribute to oxygen transport to the roots, AFP of perlite can be 3 (fine perlite) to 13 % (coarse perlite) lower.
Table 2. Volumetric water conl Growing medium coir
Perlite fine
Perlite mid
Perlite coarse
Frequency
lx/45 min lx/90 min lx/180 min lx/45 min lx/90 min lx/180 min lx/45 min lx/90 min lx/180 min lx/45 min lx/90 min lx/180min
ent as determined by weight of the pots during week 37
67
57
38
27
week 44
60
66
36
26
week 50
59 61 60 68 74 68 46 47 44 35 32 37
the experiment week 12
61 60 61 72 72 73 49 48 47 37 34 38
week 21
60 52 55 67 56 67 53 46 52 40 35 42
-coir
perl fine
- perl mid
-perl coarse
35 45 55 65 75
time (weeks)
Fig. 1. Air-filled porosity during experiment 1
Permittivity (short term measurements)
permittivity day 118 (w 17)
§ 70 n
à 6 5 ' £ "5 1 60
£ 55
1 50
—•—m
1
N° 'b0 <•? A° °>° N 0 ^ 0 <b° A° °>° •$ <b°
tirne (rrinutes)
permittivity day 147 (w 21)
O O
•
O
O
•HL
t Ë a> a
5 0 liiiiiMiiuiiiiiiiiiiiiiirriiiiiiiiiiiMiiiiiiiriniiiinrTTïiiiiiiiiiiiiiiiiinii m um imiiiiiii
N° <b° <o° A ° <*° N° ^ ° <b° A ° < ° N° <i?
time (minutes)
Fig 2 Average permittivity (each line represents average of six sensors) on two days with different irradiation levels (day 118, irradiation sum 2224 J/cm2 ;day 147, irradation sum 2683 J/cm )
Two examples of the irrigation feed-back control are shown in Fig. 2. In case the setpoints were not reached, as on day 118, the normal time-based control worked. In case of high irradiation levels, e.g. at day 147', setpoints were reached in the I treatment, and irrigation frequency increased.
2.2.11.2 Nutrient concentrations and uptake Average realized concentration as determined by analysis of the recirculation tanks is given in Table 2. Note that the EC and concentrations were decreased during the experiment as is illustrated by the N03 and K concentrations (Fig. 3,4). From the ratio of added fertilizers in order to maintain the target concentrations and the realized transpiration the so-called removal concentrations were calculated (Table 3). As can be seen concentrations of N03 and K in the recirculation tanks at the end of the experiment were lower than the average removal concentrations. Visual deficiencies did not occur however. Note the higher uptake of NH4. B and Mn than the realized concentration, suggesting either active uptake of these elements, or loss by other mechanims (nitrification, precipitation, absorption).
10
Fig. 3. Nitrate concentration in recirculation tank. Fig. 4. Potassium concentration in recirculation tank during experiment 1.
Table 3. Average realized concentrations in recirculation tanks and average calculated removal concentrations
K
Ca
Mg
NH4
Na
N03
H2P04
S04
CI
Fe
Mn
Zn
B
Cu
mmol/1
mmol/1
mmol/1
mmol/1
mmol/1
mmol/1
mmol/1
mmol/1
mmol/1
umol/l
umol/l
umol/l
umol/l
umol/l
Removal concentration
1.4
1.0
0.2
1.2
0.0
3.3
0.3
0.2
0.0
11.2
2.8
0.0
13.0
0.1
Nutrient solution concentration
2.6
2.4
1.6
0.3
1.2
7.2
0.5
1.8
0.6
24
2.6
4.5
7.1
1.3
2.2.1.2 Plant parameters
2.2.1.2.1 RootADH'activity Root ADH activity as determined at several occasions showed large differences during the experiment (Table 4). One reason for the low activities in week 49 is probably the long storage of the samples from week 44 and week 49 prior to analysis in week 12. The roots had a dessicated appearance after being stored at -20°C. The difference between week 12 and week 21 may have been due to differences in root activity, or to the larger amount of roots in the substrate, thereby decreasing effective air content. Although the trend was similar at all harvest dates only in week 12 a significant difference in ADH activity was found. The higher activity in fine perlite correlated with the lower air content of fine perlite at the same harvest date (Fig. 4).
11
Table 4. Root ADH activity (pmol NADH/g prot.min) at different harvest dates growing medium coir perlite fine perlite mid perlite coarse LSD
week 49 517 543 397 438 -
Week 12 1149 1515 947 858 325
week 21 2217 2465 2001 1869 -
Fig 5. Relation between air content and root ADH activity experiment 1.
2.2.1.2.2 Transpiration and irrigation Transpiration (weekly average) steadily increased during the experiment due to increased irradiation (Fig. 6). Since irrigation amounts exceeded transpiration, drainage percentages were very high compared to standard horticultural drainage percentages of 30%, namely on average 80, 87 and 94% in the frequency treatments. In week 51 the duration of the dripping period was increased from 1 to 3 minutes in all frequencies, since differences in concentrations between drain and in the recirculation tank (drip solution) differed (data not shown). The sharp increase in the low frequency treatment from week 6 to 9 1999 was the result of failure of the feed-back control: permittivity was not increased above the setpoint by irrigation, causing the irrigation to be triggered. As could be expected, feed-back control was only realized in the low frequency treatment.
— 4
.1 1
•—8x/day
A 16x/day
-X—32x/day
-transp
0.5
'j^yy^jTi v * ti*yvtw? • v ^ ^ * «^vi w 0.1
0
41 45 49 53 4 8 12 16 20 24
time (week no.)
Fig. 6. Realized transpiration (l/plant.day) as average of all growing media and irrigation amounts during experiment 1.
12
2.2.1.2.3 Production No effects of the irrigation frequencies were found on production (Table 5). There were also no significant irrigation * growing medium interaction effects. However, the growing media differed in production parameters. Less stems per plant were produced in the coarse perlite compared to the perlite fine and coir dust treatments (Table 6). Average stem weight however was higher in perlite coarse compared to coir dust, thereby nearly but not fully compensating total weight production. Fig. 7,8 show that the difference already was established in the beginning of the experiment, which corresponded with the realized root weight production at that moment (Table 7). Apparently, the roots had difficulty to grow from the coir dust plugs, in which the cuttings had been propagated, into particularly perlite mid and perlite coarse, which had a negative effect on above-ground production in that period, despite the abundant supply of water and nutrients. The difference in rooting was maintained during the experiment, since at the end the root weight in the perlite mid and perlite coarse was still significantly less than in coir dust.
Table 5. Analysis of variance of production parameters experiment 1. n.s. = not significant; * = P<0.05; ** = P<0.01.
Irrigation growing medium Irrigation'growing medium
no. stems/plant w40-w24 n.s. * * n.s.
stem weight w40w24 n.s. * n.s.
total weighl/plot w40-w24 n.s. * n.s.
Table 6. Production parameters of experiment 1 Treatment freq 32x/day freq 16x/day freq 8x/day LSD
Coir perlite fine perlite mid perlite coarse LSD
no. stems/plant 12.4 12.5 12.6 -
13.2 13.0 12.3 11.5 1.0
g FW/plant 522 511 522 -
532 532 513 486 35
stem FW 42.1 40.9 41.4 -
40.3 40.9 41.7 42.3 1.3
Table 7. Root wei Growing medium
Coir Perlite fine Perlite mid Perlite coarse LSD
ght at harvests during the experiment. Week 44 feFW) 1.9 1.5 0.7 0.4 0.5
week 50 te FW) 13.1 8.9 4.8 7.6 2.2
Week 21 (ß FW) 163 140 91 74 -
gDW) 16.7 10.2 6.2 9.1 6.5
13
60 -,
5 50 -du O)
!> 40 0 S
o; 30 -co
20
—#—coir
O—perl fine
—£—perl mid
—M—perl coarse
^ K « ^ ^ K ^ tfS^r i
s M
/ r
fc. ^
^
12 13 1 2 3 4 5 6
period ('98-'99)
Fig. 7 . Production per plant per period of 4 weeks in the different growing media in experiment 1.
CO E -> CO
0
—«—coir
—a—perl fine
—*—perl coarse
J ^
12 13 1 2 3 4 5 6
period ('98-'99)
Fig. 8. Average stem weight per period.
2.2.1.2.4 Nutrient analysis Only minor differences in nutrient concentrations were found in leaves as a result of growing medium and/or frequency (Table 8). Potassium was higher in coir. In coarse perlite lower total N and K but higher P was found. The low frequency resulted in higher Ca and lower P concentrations in leaves. The lower EC after week 6 resulted only in lower K and to a lesser extent P concentrations in week 17 compared to week 44. Although the values were below the guide values for cut rose (de Kreij et al 1992), the concentrations were still higher than considered in the deficiency range (Appendix 2)
14
Table 8. Nutrient concentrations (mmol/1
Growing medium coir perlite fine perlite mid perlite coarse LSD
coir Perlite fine Perlite mid Perlite coarse LSD
Frequency Low Medium High LSD
Week 44 K 871 880 881 859 -
Week 17 606 562 574 565 30
572 581 577
Na 10 11 11 10 -
<g DM) in 1
Ca 257 255 263 258 -
364 343 339 330 -
359 343 330 21
ully-grown
Mg 120 129 136 131 -
147 167 169 168 13
167 165 156 -
five leaflet
P 98 102 105 89 -
90 86 83 80 -
81 83 90 4
eaves in experiment 1
N03 41 54 46 25 -
CI 20 31 31 33 -
S04 45 39 39 35 -
N total 2576 2641 2591 2358 161
2532 2590 2502 -
2.2.2 Experiment 2 (week 25 1999 - week 4 2000)
2.2.2.1 Root environment
2.2.2.1.1 Transpiration and irrigation As in the previous experiment, irrigation rates were very high in relation to the measured transpiration (Fig. 9); on average, drainage rates were 62, 81 and 95% for the low, medium and high irrigation frequency, respectively. The dip in the irrigation at week 52 is the result of a temporary technical failure. Highest transpiration rates were found in weeks 28 and 31 at 0.62 l/plant.day (weekly average).
Fig. 9. Realized transpiration (average all growing media) and irrigation during experiment 1.
15
2.2.2.1.2 EC measurements with FD sensors he time-coarse of EC measurements with the FD sensors in coir shows that the target EC's of 0.9 (Fig. 10) and 1.9 mS/cm (Fig. 11) were reached relatively well.
2.0
1.5
1.0 I U u 0.5
0.0
*fffo\tf**+?9f**
i i i i i i i i i i M M i i M i i i i M i i i i i i i
25 29 33 37 41 45 49 1
time (weeks)
Fig. 10. Time-course of EC as measured with FD sensors in coir at different frequencies at EC 0.9
1.0 I I M I I I I I I I I I I I I I M I I I I I l i l i l i i
25 29 33 37 41 45 49 1
time (weeks)
8x/üay
16x/fciay
32x/t)ay
Fig. 11. Time-course of EC as measured with FD sensors in coir at different frequencies at EC 1.9
2.2.2.2 Plant parameters
2.22.2.1 Production No significant differences were found for the EC, the frequency and the growing media treatments during the experimental period (Table 9). Apparently, nutrition did not become limiting at 0.9 mS/cm.
16
Table 9. Production parameters in the experiment week 25 1999-w 4 2000 Treatment
EC 0.9 EC 1.9 LSD Freq 32x/day Freq 16x/day Freq 8x/day LSD Coir Perlite fine Perlite mid Perlite coarse LSD
no. stems/plant
35.6 37.1 -36.6 36.1 36.2 -37.4 37.5 35.8 34.5 -
g FW/plant
805 808 -806 806 807 -820 831 799 775 -
stem FW
22.6 21.8 -22.0 22.4 22.3 -22.0 22.2 22.3 22.5 -
Transpiration L/plant 70.4 69.9 -70.3 69.5 70.6 -
2.2.2.2.2 Nutrient analysis
Table 10. Nutrient concentrations (mmolAg DM) in fully-grown
Treatment EC 0.9 EC 2 LSD freq 32x/day freq 16x/day freq 8x/day LSD Coir perlite fine perlite mid perlite coarse LSD
week 40 1999 K 538 618 32 588 585 561 -585 --571 -
Ca 354 378 -375 383 341 -381 --352 -
Mg 161 151 6 153 161 152 -148 --163 6
P 66 77 4 73 72 71 -73 --71 -
Total N 2089 2312 108 2264 2210 2128 -2193 --2208 -
eaves in week 40 1999 and week 4 2000 week 4 2000 K 688 722 26 710 703 701 -714 716 690 699 -
Ca 309 295 -307 319 279 -302 306 294 307 -
Mg 124 128 -127 130 121 -119 128 128 129 -
P 82 91 5 89 86 85 -89 89 83 85 -
Total N 2296 2364 -2367 2350 2273 -2293 2400 2351 2277 -
There were only minor effects of the decreased nutrient concentrations in the nutrient solution on chemical composition as measured in week 40 1999 and week 4 2000 (Table 10). K and P concentrations and total N in week 40 were lower, whereas Mg concentration was higher probably as a result of K/Mg antagonism in uptake. In week 4 only significant effects on K and P were found. The irrigation frequencies had no effect on chemical composition. With regards to the growing media, only a smaller Mg concentration in coir in week 40 was found. In week 4 there were no differences in nutrient concentrations.
2.2.3 Experiment 3 (week 5 - week 32 2000)
2.2.3.1 Root environment
2.2.3.1.1 EC realized Fig. 12 shows that the 10 mM NaCI addition resulted in an EC increase of 1 mS/cm, which makes a comparison comparable between the EC 1.9 and EC 0.9+10 Na treatments, and resulted in a EC span of 1-3 mS/cm in the experiment.
17
£ o <
%
O LU
3
2
1
M ^ V / I / ^
o
**V* •^^^•^^*
M M I M I ! I I I I I M I I I I I I
• - EC 0.9+ 0 Na
— EC 0.9+ 10 Na
A EC 1.9+ 0 Na
-K—EC 1.9+ 10 Na
5 10 15 20 25 30
week
Fig. 12. Realized EC in recirculation tanks during experiment 3.
2.2.3.1.2 Gas composition rhizosphere The gas composition of samples taken 5 cm from the bottom of the containers showed that significant differences in concentration oxygen and carbon dioxide existed between the growing media (Table 11). The media with the lowest AFP (cf Fig. 1) showed the highest carbon dioxide and lowest oxygen concentrations. However, the oxygen concentrations hardly differed from ambient conditions (greenhouse air), which confirmed the conclusion from the first year that aeration was no physiological problem in the growing media used.
Table 11. Oxygen and carbon dioxide concentrations in the media as determined in april 2000.
Coir Perlite fine Perlite mid Perlite coarse Control (greenhouse air) LSD
0, 21.9 21.8 22.2 22.1 22.1 0.1
CO, 0.16 0.25 0.10 0.09 0.04 0.08
2.2.3.2 Plant parameters
2.2.3.2.1 Production Production was significantly affected by the treatments (Table 12). Production (number stems/plant and FW/plant ) was lower in the perlite coarse treatment. With regards to the EC and NaCI treatments there was a negative effect of NaCI addition. However, this negative effect became only significant at the low EC treatment. In the absence of NaCI, the low EC treatment produced heavier stems than the EC 1.9 treatment. Total transpiration over the experimental period was - although not significantly different - related to total FW production.
18
Table 12. Production parameters and transpiration in the experiment week 5-32 2000 Treatment
EC 0.9 NaCI 0 EC 0.9 NaCI 10 EC 1.9 NaCI 0 EC 1.9 NaCI 10 LSD
EC 0.9 EC 1.9 LSD NaCIO NaCI 10 LSD Coir Perlite fine Perlite mid Perlite coarse LSD
NaCI*EC EC*medium NaCrmedium NaCI*EC*medium
No. stems/plant
41.9 38.5 43.5 42.0 -
40.2 42.7 -42.7 40.3 -43.1 44.1 40.2 38.5 3.9
----
g FW/plant
976 855 961 928 93
917 944 -969 895 62 970 983 909 855 88
----
stem FW
23.3 22.2 22.1 22.1 0.7
22.8 22.1 0.5 22.7 22.2 0.5 22.5 22.3 22.6 22.2 -
* ---
Transpiration L/plant 80 69 76 78 -
74 77 -78 74 -
-
2.2.3.2.2 Nutrient analysis
Table 13. Nutrient concentrations (mmolAg DM) in young fully developed leaves sampled in weeks 12 and 31 2000
Treatment EC 0.9 NaCI 0 EC 0.9 NaCI 10 EC 1.9 NaCIO EC 1.9 NaCI 10 LSD
EC 0.9 NaCI 0 EC 0.9 NaCI 10 EC 1.9 NaCIO EC 1.9 NaCI 10 LSD
Week 122000 Na 3 5 4 3 -
CI 15 39 8 26 9
K 634 627 670 661 33
Ca 270 263 300 266 -
Mg 122 117 128 123 -
P 72 74 81 74 7
Total N 2183 2221 2251 2197 -
N03
Week 31 2000 1 1 1 1 -
29 53 11 35 8
663 679 728 754 32
229 207 224 210 -
151 133 143 141 7
66 68 73 72 3
1870 1925 2050 2040 97
7 9 12 10 -
A clear effect of NaCI addition was visible in the CI concentrations in the leaf (Table 13). Particularly in the EC 0.9 treatment CI concentrations were high, which coincided with the lower production in the EC 0.9 NaCI 10 treatment. Na concentrations were very low, and did not show significant changes, confirming the effective exclusion of Na by rose roots. With respect to the other major elements K, P and Mg were higher at higher EC. Total N was significantly higher in week 31 in the higher EC treatment. No significant differences in the Ca concentration were found.
2.2.3.2.3 Nutrient removal concentrations From transpiration and fertilizer addition in order to maintain nutrient concentrations in the nutrient solution nutrient removal concentrations were calculated (Table 14). Except NH4, removal concentrations generality were higher in the EC 1.9 treatments.
19
Table 14. Nutrient removal concentrations as calculated by fertilizer addition divided by transpiration. Between brackets the concentrations in the nutrient solution are given.
K Ca Mg NH4 Na N03 H2P04 S04 Cl Fe Mn Zn B Cu
mmol/1 mmol/l mmol/1 mmol/l mmol/1 mmol/l mmol/l mmol/1 mmol/l umol/l umol/l umol/l umol/l umol/l
EC 0.9 + 0 NaCI 2.0(2.5) 0.9(2.5) 0.5(1.5) 0.5(0.1) -3.5(6.3) 0.4(0.5) 0.4(1.5) -19 (25) 3(3) 2 (3.5) 9(10) 0.6(1)
EC 0.9 + 10 NaCI 2.1 0.9 0.6 0.8 -3.7 0.4 0.4 -20 3 2 10 0.6
EC 1.9+0 NaCI 2.6 (5) 1.1(5) 0.7(3) 0.4(0.1) -4.5(12.5) 0.5 (0.9) 0.5 (3) -21 (25) 4(3) 2(35) 12(10) 0.7(1)
EC 1.9+10 NaCI 2.4 1 0.6 0.6 -5.0 0.6 0.6 -24 5 2 14 0.8
2.2.4 Combined experiments (week 40 1998 - week 32 2000)
2.2.4.1 Permittivity/water content
Fig. 13. Water content in the coir as determined by permittivity readings
Water content as calculated from permittivity readings remained relativey constant during the cultivation period in coir (Fig. 13). During experiment 3 only one frequency of irrigation (16x/day) was used. If porosity did not change, a water content of 70% would still mean that air-filled porosity was at least 20%, which - as concluded before - did not lead to aeration problems.
2.2.4.2 Transpiration Transpiration during the experiment more or less coincided with irradiation sum as determined outside the greenhouse (Fig. 14). It should be realized that assimilation lighting was available in the period from september until may (week 1-18 and 35-week 18) which -at least partly explains the relatively higher transpiration to irradiation ratio in winter time. When plotted against each other (Fig. 15), the maximum
20
transpiration can be estimated for summer conditions, which can serve as minimum daily irrigation requirement.
2400
CO T3 c\i E -y
1800
3 1200 c o
'XS CO
TJ to
6 0 0
irradiation
assim. light
transpiration
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0 r- CO Lf>
week
Fig. 14. Realized transpiration and radiation during the experimental period. Data based on weekly average.
0.9
_ 0.8
I 0.7 1 0.6 | 0.5
0 0 .4
1 0.3 a g 0.2 CO
~ 0.1
0.0
Transpiration < 3E"1,lrrJ + 2E"08lrH + 8E06lrr+ 0.2
* ^
1 1 1 1 • 1 1 1 1 1
0 250 500 750 1000 1250 1500 1750 2000 2250 2500
irradiation (J/cm2.day)
Fig. 15. Relation between outside irradiation sum and transpiration as determined from weekly average. Fitted curve for maximum transpiration is given. Note that data without assimilation lighting are above ca. 1200J/cm2.day.
2.2.4.3 Production The differences in production as found in the first year (experiment 1) persisted throughout the total experimental period (Table 15). The coarse perlite treatment showed a yield reduction of 10% compared to the coir and fine perlite treatments. The middle fraction perlite (which is commercially sold for rose production) yielded intermediately. Stem weight did not differ significantly at the end of the project (in contrast to the results of experiment 1, Table 6).
21
Table 15. Production during the total experimental period week 40 1998 - week 32 2000) Medium Coir Perlite fine Perlite mid Perlite coarse LSD
no. stems/plant 95 96 90 86 6
tot.weight/plant 2380 2406 2278 2167 154
stem weight 25.0 25.1 25.4 25.3 -
In Fig. 16-18 the time course of the production is given. During the first year not so many, but very heavy stems were produced per plant (Fig. 18). Later, as more stems per plant were produced per harvest cycle, stem weight (and length) decreased and more or less stabilized between 25-30 g FW.
3 0 0
- •—co i r dust
-û—perl mid
——irrad. sum
- •—per l f ine
-X—perl coarse
«- co in
period ( '98-'OO)
Fig. 16. Fresh weight production in the growing media and light sum per period of 4 weeks during the experimental period.
10
coir
perl f ine
perl mid
perl coarse
period ( ' 98 - ' 00 )
Fig. 17. Stems per plant produced per period of 4 weeks in the growing media during the project.
22
Fig. 18. Average stem weight produced in the growing media during the project.
Fig. 19. Water and light use efficiency during the experimental period calculated from 2-period average harvest, transpiration and light sum data.
From the data as given in Fig 14 and Fig. 16 water use efficiency (WUE in g FWAg H20) and light use efficiency (LUE in g FW/1000J.cm2) was calculated (Fig. 19). WUE increased steadily in the first year to reach its peak during the low transpiration period 11-12, whereafter it declined again. Under summer conditions without assimilation light WUE was between 12-15 g FWAg H20. LUE also showed a peak under low-light conditions, although it's peak was some 2 periods later than the the peak in WUE. Under summer conditions LUE was around 3-4 g FW/1000J.cm2.
23
3 Israel
3.1 Materials and methods
3.1.1 Hydroponic experiments
3.1.1.1 Experiment A Nine closed, fully recirculated systems were used. Each system accommodated 9 containers. One plant (cut rose, cv. "Frisco" grafted on Rosa indica major cv. Sharon) per container was planted at 3.9.98. Medium volume per plant is 2.5 liter. The containers were mulched using black aluminized film in order to minimize evaporation. Two systems were not planted yet operated in a similar manner to the planted systems. They served as blanks (one for tuff and one for perlite), in order to account for evaporation from the media and thus enable transpiration calculation. The nutrient solution flow rate was very high - 3.6 liter/hour and the pumps were operated continuously throughout the experiments. At this flow rate, the concentration of dissolved oxygen, EC and pH of the solution are completely uniform at all parts of the system, including the collecting tank.
Immediately after planting, all the buds were removed and the stems were bent in order to promote bottom break formation. The bottom breaks were left for further growth. Flowers were harvested at commercial opening stage and their length was measured. They were then divided to leaves, flower buds and stems. All the removed plant material was collected, weighed, dried, weighed again and kept for chemical analysis. At the termination of the experiment (19.1.99) the plants were removed carefully from the medium, separated to different organs and were treated in a similar manner.
The nutrient solutions consisted of collected rainwater and (in mM) 5 N (75% N03, 25% NH4), 1 P, 2.5K, 2 Ca, 1 Mg, + Hoagland level micronutrients. The EC of this solution is 0.8 mS cm1 and its pH is 6.2. The solution volume is 17 liters per system. The volume was monitored frequently and solution was added so as to keep it close to the maximal level. The added volume was recorded. Frequent EC, pH and nutrient analyses ensured that the solutions will be at all times less than ± 25% of the original levels. It was difficult, however, to control the pH levels, especially in the perlite and this resulted in some plant disorders. Once this became clear, the ammonium/nitrate ratio was lowered in the perlite stock solution. This facilitated pH control. The whole solution volume was replaced whenever the nutrient concentration deviated from the preset values. The volume and chemical content of discarded solutions were recorded.
Media Our intention was to work with uniform, graded materials. Tuff and perlite - two commercially important media were screened using 4, 10 and 20 mesh screens, to generate fractions of >4.75 mm (only tuff), 2.0-4.75 mm and 0.85-2.0 mm. Perlite particles larger than 4.75 mm are not commercially available. Fractions finer than 0.85 mm were also collected but were not used at this time. The fractions were defined, for the sake of convenience, as tuff 4, 10, 20 etc.
Measurements The various size fractions were further subjected to particle size distribution analysis. Water release curves of the media were measured using a sintered glass funnel, as described by Raviv and Medina (1997). Saturated hydraulic conductivity of the various media was measured using glass columns having an inner diameter of 33.1 mm. The columns were filled with 180 cm3 medium and saturated with water overnight. After draining excess water, water were added at the same rate of percolation from the lower end of the column, while measuring both filling rate of a collecting tank and flow rate from a supply burette. The measurements were replicated 10 times per size fraction. Saturated hydraulic conductivity (Ks) was
24
calculated as the flow rate (cm3!!1) divided by the column surface area (cm2) and is expressed as cm h1. Unsaturated hydraulic conductivity of the various media was calculated based on the above values and based on the measured water release curves, according to the Mualem & van Genuchten equation. The calculations were performed by Prof. R. Wallach of the Hebrew University of Jerusalem, Faculty of Agriculture, Rehovot.
Measured plant parameters included periodical measurements of net assimilation rates (NAR) that were measured using Li-Cor 6200 Photosynthesis meter. Leaf water potential was measured early in the morning, using a pressure chamber (Kfar Haruv, Israel). The collected plant material was chemically analyzed at the end of the experiment.
Water analyses included content of nutritional and major non-nutritional ions, EC and pH. Chemical composition of the solutions was analyzed bi-weekly and at the end of the experiment. Whenever an "old" solution was replaced with a fresh one, its chemical content was analyzed. Dissolved oxygen in the water reservoir was measured sporadically, until it became clear that D02 levels of all media are at all times very close to saturation (results are not presented). Unless otherwise stated, all growing procedures and mesurements followed the same pattern in the other hydroponic experiments.
3.1.1.2 Experiment B The same hydroponic systems served for this experiment. One plant per container was planted at 28.2.99. Three of the media were tested again: coarse (>4.75 mm) and medium (24.75mm) tuff and coarse (2-4.75mm) perlite. Each medium was used in 3 systems. Two of the systems were planted and one served as a blank. Flowers were harvested at commercial opening stage, starting 11.4.99. The experiment was terminated at 31.5.99 and the plants were treated as described above. This experiment was terminated abruptly, due to malfunction of some of the pumps. This prevented us from doing complete water and mineral balance, as planned. Measured plant parameters included periodical measurements of NAR and leaf water potential. The nutrient solutions consisted of collected rainwater and their content appears in Table 16. We tried, in this case, to take advantage of the high flux, so we reduced considerably the ionic concentration.
Table 16: content (in mM L ') of the nutrient solutions, and their maximal EC levels. Compound Tuff Perlite
1.25 1.0 0.5 0.8 0.4 0.5 6.5
Microelements were added using chelates, to the concentration of Hoagland solution. The pH of the solutions was adjusted to 6.2. However, the pH tended to drop and frequent additions of K0H were required to maintain the pH in the range 5-6. Water analyses were conducted as decribed above.
3.1.1.3 Experimente
Four systems served for this experiment. Plants were planted at 29.2.2000 in coarse perlite (24.75 mm). Two nutrient solution irrigation capacities were compared: 0.24 l/hour (low) and 6.0 l/hour (high). At these irrigation capacities the EC and pH of the solution are completely uniform at all parts of the system, including the collecting tank. It is not the case, however, for the concentration of dissolved oxygen in the low irrigation capacity, which was highest at the first irrigation outlet and decreases as the solution passed through the containers. Flowers were harvested beginning 3.4.2000. The experiment was terminated at 6.6.2000 and the plants were treated as described above. Fresh roots were subjected to analysis of ADH
25
KN03
NH4N03
NH,H2P04
Ca(N03)2
MgS04
EC (mS cm1) pH
2.0 2.0 1.0 1.0 0.5 0.8 6.5
activity, as described by Baas et al. (1995, modified 1997). The nutrient solutions consisted of collected rainwater and their contents appear in Table 18.
Table 18: content (in mM) of the 3 nutrient solutions compared, and their maximal EC levels-Compound Low salinity High salinity
8 2 2 11 4 2 4.5 4.4
KN03
NH4N03
NH4H2P04
NaCI Ca(N03)2
MgS04
Maximal EC (planned, mS/cm) Actual average EC (mS/cm)
2 0.5 0.5 0 1 0.5 1.5 1.1
Measured plant parameters included periodical measurements of NAR that were measured using Li-Cor 6200 Photosynthesis meter. Stomatal resistance and specific transpiration rate were measured early in the morning, using a Li-Cor 1600 Diffusive Porometer. The collected plant material was chemically analyzed at the end of the experiment.
3.1.1.4 Experiment D Nine systems served for this experiment. Plants were planted at 12.11.2000, in coarse (2-4.75 mm) perlite. Three nutrient solution irrigation flow rates were tested: 0.12, 0.6 and 3.0 l/h. The pumps were operated continuously throughout the experiments. At each flow rate, 3 salinity levels were used. The highest salinity level aimed to mimic the composition of a closed, recirculated irrigation system while in the lower nutrient level we tried to compensate for the low nutritional content by applying high flow (equivalent to 40 times the normal application rate in commercial greenhouse). The nutrient solutions consisted of collected rainwater and their contents appear in Table 17.
Table 17: Actual concentrations of several ions (in mM) of the nutrient solutions, and their EC levels (Average of 58 observations).
NO, NH/ PO.3
K+
Ca++
Na+
CI' Average EC (mS cm1)
Low salinity 2.18 0.15 0.13 1.74 1.20 0.79 1.17 0.52
Medium salinity 3.15 0.17 0.33 3.79 1.65 1.18 1.63 0.87
High salinity 10.34 0.32 0.95 8.78 3.79 13.01 12.47 3.06
Magnesium concentrations in the stock solutions were 0.5, 1.0 and 2.0 mM in the low, medium and high salinity treatments, respectively. Microelements were added using chelates, to the concentration of Hoagland solution. The pH of the solutions was adjusted to 6.0. However, the pH tended to drop and frequent additions of KOH were required to maintain the pH in the range 5-6. This resulted with negligible concentrations of bicarbonates. Water analyses were conducted as described above.
Flowers were harvested starting 19.12.00. The experiment was terminated at 27.3.01 and the plants were treated as described above. Measured plant parameters included periodical measurements of NAR, stomatal resistance and specific transpiration rate.
26
3.1.2 Growing medium experiment
Tuff (0-8 mm) was used as a growing medium. It was placed in 1.4X0.8X0.2m polystyrene boxes, situated, 2 per bed, on stainless still troughs. The troughs have one drainage outlet, leading to a collecting tank, where submersible pump is located. Twenty-eight rose plants (CV. Golden Gate, grafted on Rosa indica Major) were planted at mid-April 2000 per bed (identical with commercial density, at 5.8 plants per m2, including paths). Twenty beds served for the experiment at 5 replicates per treatment, arranged in a randomized block design. Harvest started at 23/5/2000. Water recirculation started at 17/7/2000. The treatments were:
1. Control, flow-through solution, at a leaching fraction that enables maintaining EC of the effluent < 3.5 dS m1. Irrigation is scheduled with a tensiometer, located at a depth of 8-10 cm. The setpoint for irrigation is not uniform. It is high (1.7 kPa) during the night, falling to low (1.1 kPa) between 09:00-12:00 and gradually increasing again towards night. Typically, this regime results with 4-6 pulses per day, at rather regular hours. The EC of the tap water is 1.1-1.2 mS cm1 and with fertilizers (see in the following), 2.1-2.2 mS cm1.
2. Closed system, where the effluents are pumped back to an operative tank, where fresh water is added to replenish transpired or discharged water. EC of the mixture is set to 2.3 mS cm * and the mixture is brought to 3.0 mS cm"1 with fertilizers. Irrigation scheduling mimics that of treatment 1. This is facilitated by the regularity of the watering demand of treatment 1.
3. Same as treatment 2, but with double the number of irrigation pulses per day, with the same watering distribution during the day. Based on previous experience it was assumed that increasing irrigation frequency will result with increased transpiration, therefore the cumulative amount of applied water was increased by 10%.
4. Same as treatment 3, but with triple the number of pulses per day, with the same watering distribution during the day. Based on previous experience it was assumed that increasing irrigation frequency will result with increased transpiration, therefore the cumulative amount of applied water was increased by 20%.
Excess drainage water that was not used for creating the mixture described above was discharged from the collection tanks. Amount of discharged water was recorded. The target values for nutritional ions in the control treatment were (in mM) K , 4-5; P04 1; N03i 7-8; NH4 2-3. Concentrations were, of course, higher, in the recirculated treatments. Moisture content in the media was measured using frequency domain (FD) sensors (Silora, Kfar Masaryk, Israel), at a 15-minute interval, over the entire experimental period.
The plants were treated as common in commercial greenhouses. Weak flowers were bent and pinched. Flowers were harvested daily, sorted for length, marketability and weighed. Harvest results were analyzed using analysis of variance. Differences among treatments were determined using Duncan's Multiple Range test.
Assuming that the large difference in osmotic potentials among the treatments (ca. 36 kPa between the control and the recycled treatments) may affect flower longevity, we tested their vase life, under controlled conditions. The flowers were transferred immediately after harvest to a cold room and stored at 4+1 °C for 24 hours. Subsequently, they were transferred to simulated room conditions: 25±1°C, RH = 55±5%, 16 h photoperiod. The diameter of the flower was measured every other day and its visual appearance was assessed visually, on a 1-5 scale, 1 being commercial opening stage and 5 - wilted. The test was replicated twice. Fifteen flowers per treatment were sampled at each sampling date.
27
3.2 Results
3.2.2 Hydroponic experiments
3.2.2.1 Experiment A: Physical and hydraulic characteristics of the media
After screening the various fractions of the media, a particle size distribution was conducted, to validate the screening efficiency. The results appear in Table 19.
Table 19: Particle size distribution of tuff and per ite fractions (%, W/W) Fraction size (mm) >4.75 2.004.75 0.85-1.99 0.43-0.84 0.25-0.42 0.10-0.24 <0.10
Tuff 4
83.1 15.8 0.0 0.0 0.1 0.3 0.8
Tuff 10
3.6 84.6 9.0 0.6 0.2 0.5 1.5
Tuff 20
0.0 0.1 52.6 23.1 8.6 9.6 6.1
Perlite 10
2.9 80.6 14.4 0.9 0.8 0.3 0.1
Perlite 20
0.0 0.8 87.8 6.1 3.1 1.8 0.4
It can be seen that the screening was efficient in removing particles larger than the screened range. However, some smaller particles may be found in each specific fraction, mainly due to adhesive forces (especially in tuff). The water release curves (Fig. 20) demonstrate the specific physical characteristics of some of the fractions. Perlite 20 has the highest water content at most of the measured matric potentials but relatively low AFP at 1 kPa, which is considered in the literature as the matric potential at container capacity, based on the average depth of standard containers. Tuff 20, on the other hand, has virtually zero AFP at this matric potential and relatively high water content at all the measured range above this point. Tuff 4 and 10 have similar water release curves - very high AFP at 1 kPa and very low water content at most of the measured range above it. Perlite 10 has high AFP and intermediate water content. Only perlite 20 and tuff 20 can release significant amount of water above matric potential of 1 kPa. It should be mentioned that under our experimental conditions, AFP at 1 kPa is not the relevant parameter to predict oxygen diffusion rate in the pore space as the plants were under near-saturation conditions at all times. At these conditions, however, the relativem capacities among the substrates were similar to those reported for 1 kPa. It can be concluded that only tuff 4, tuff 10 and perlite 10 have simultaneously acceptable physical conditions at near-saturation. Except for fine tuff (tuff 20), all size fractions have very high hydraulic conductivities (HC) near saturation (Figure 21). However, HC of tuff 4 drops much sharper than those of the other media with increasing matric potential. This leaves tuff 10 and perlite 10 as the best media in terms of physical characteristics, among the tested media.
28
W*er release curves of perlite and ti if size fractions
0 10 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 100
Pressire head (cm(
Fig. 20. Water release curves from different media and fractions.
Hydraulic conductivities of perlite and tuff size fractions
• • . ' I K . .
• Peflrte 10
• Tutu
• Tuft 10
«Turf 20
Mairie potential (kPa x 10)
Fig. 21. Combined results of the measurement of saturated hydraulic conductivity and calculation of unsaturated hydraulic conductivity.
3.2.2.2 Experiment A: Harvested plant material and nutrient uptake
Number of flowers per plant and other harvest-related data are presented in table 20.
29
Table 20: Total and harvested dry weight, flower parameters and water use parameters of Frisco roses, grown in several
Flowers/ Plant Flower length (cm) Flower diameter (cm) Total DW/plant (g) Yield DW/plant (g) Water transpired/plant (L) BiomassWUE* (L/g) Yield WUE (L/g)
size fractions o Tuff 4 4.33 ±0.97 44.0 ± 0.94
0.44 ±0.01
46.9
20.6
13.378
3.51
1.54
f tuff and perlite. Tuff 10 4.56 ±0.81 48.3 ±0.79
0.51 ±0.01
66.2
29.2
18.711
3.54
1.56
Tuff 20 4.11 ±0.41 47.7 ±1.64
0.48 ± 0.02
48.9
22.6
13.111
3.73
1.73
Perlite 10 3.67 ±0.66 43.5 ±1.29
0.47 ±0.02
55.3
16.7
15.289
3.62
1.10
Perlite 20 5.00 ±0.66 41.5 ±1.30
0.44 ± 0.01
54.6
22.0
16.589
3.29
1.33 * WUE - Water use efficiency
Perlite 20-grown plants were most prolific in terms of number of flowers per plant but the flowers were of a low quality, as reflected in their short stem length and small diameter. Perlite 10-grown plants were the least productive although their total dry matter production was relatively high. This unusual behavior was due to a large proportion of "blind" shoots, harvested from these plants. This fact also resulted with low yield WUE. As mentioned above, the poor showing of the perlite fractions may result from events of low pH values and probably are not necessarily related to their physical characteristics. Tuff 10-grown plants yielded the best, as could be expected by its adequate physical properties. Tuff 10-grown plants also transpired the highest amount of water, as could be expected according to their leaf fresh and dry weights that were higher than those of the other media (data not shown).
The input of various ions was calculated as the difference between the amount introduced to the systems, minus the amount withdrawn with the replaced solutions. This input was further divided between the plant uptake and the amount of ions retained by the media. A summary of the results is shown in Table 21.
Table 21: Macronutrients consumption by rose plants, grown in several size fractions of perlite and t Medium
Tuff 4
Tuff 10
Tuff 20
Perlite 10
Perlite 20
Amount (mg/plant) Input Plant % Input Plant % Input Plant % Input Plant % Input Plant %
N
1458 1053 72 1978 1674 85 1213 1089 90 1440 1300 90 1463 1262 86
P
551 201 36 731 297 41 747 214 29 271 260 96 280 259 92
K
1622 851 52 2272 1388 61 2013 866 43 1403 1139 81 1616 1112 69
Ca
459 436 95 643 635 99 230 440 191 526 413 78 586 469 80
tuff.
30
Tuff 10-grown plants accumulated the highest amount of all four macronutrients, in accordance with their produced biomass. The contents of N, P and K were correlated to the dry matter production per plant. However, Ca deviated considerably from this trend. Based on the presumed contribution of the tuff itself (see >100% uptake efficiency of Ca in tuff 20-grown plants) to Ca uptake by the plants, it is conceivable that Ca availability was a limiting factor to the flower production of perlite-grown (especially the coarser fraction) plants. Higher Ca concentrations will be supplied to perlite-grown plants in the next experiments.
3.2.2.3 Experiment A: Water relations and net assimilation rates Several measurements of leaf turgor pressure and dry matter content of representative (fully expanded, 5-leaflet leaves) leaves were done during the experiment. No statistical differences were found among the media and it seems that at this high water availability, all media can satisfy the atmospheric demand for transpiration, under the aerial conditions that prevailed in the greenhouse.
Net assimilation rate (NAR) of representative leaves was measured during 2 days. The results are presented in Figure 22, as a function of photosynthetic photon flux density (PPFD).
Net assimilation rate of rose leaves, grown in various size fractions of tuff and perlite, as a function of PPFD
•
•
• X
Perlite 20
Perlite 10
Tuff 20
Tuff 10
Tuff 4
•Logaritmisch (Perlite 10)
•Logaritmisch (Tuff 10)
0 100 200 300 400 500 600 700 800 900
PPFD (micromol/m2/sec)
Fig. 22. Net assimilation rate as function of light intensity in the different media used.
At the time the measurements were conducted, light saturation was never achieved and it seems that most
31
data points are not distinguishable from each other. Still, trend lines were added to the two extreme media: perlite 10 and tuff 10 and they clearly show different response to PPFD. At light intensities as low as 400 nmol nT2sec\ tuff 10-grown plants maintained higher NAR than that of perlite 10-grown plants and the difference became even greater at higher light intensities.
3.2.2.4 Experiment B Dissolved oxygen (DO) concentrations in the systems were close to saturation in all cases (Table 22). No statistical differences were found among the results, yet the oxygen consumption of the roots is evident, when comparing the planted vs. blank systems. All values are in the range that enables uninterrupted root function.
Table 22: Average values of DO (percent of saturation) in the systems. Medium Coarse tuff (tuff 4) Coarse tuff (tuff 4) Medium tuff (tuff 10) Medium tuff (tuff 10) Coarse perlite (perlite 10) Coarse perlite (perlite 10)
Plant + -+ -+ -
DO (% of saturation) 87 93 88 91 89 92
Photosynthesis was measured only once in this experiment, due to its untimely termination. The large variability, typical to this kind of measurements, coupled with the relative small number of data points does not enable drawing clear-cut conclusions (data not shown).
Leaf water potential was measured between 07:00 and 08:00 (when low PPFDs prevail) on 3 different days. No difference was found between the average results of the two tuff types (3.07 and 3.06 atmospheres for coarse and medium tuff, respectively). Water potential of perlite-grown leaves was much lower: 2.25 atmospheres. T-test indicates a significant difference at p<0.05. This is in agreement with the apparent superiority of perlite-grown plants in terms of NAR at low PPFDs, as presented in Figure 1. The apparent advantage of perlite in terms of water relations under low light levels may be related to its higher unsaturated hydraulic conductivity, as compared to the two tuff types (see Figure 21). Yield parameters are shown in Table 23.
Table 23: Yield and weight parameters of Frisco roses, grown in different media. Different letters in the same row denote statistical difference at p<0.05.
Flowers/plant Ave. Flower length (cm) Flower diameter (cm) Total DW/plant (g) Yield DW/plant (g) Root DW/plant (g)
Coarse tuff 3.11 a 48.4 a 0.55 a 47.5 20.7 4.8
Medium tuff 3.38 a
48.4 a 0.54 a 53.1 22.4 4.3
Coarse perlite 3.28 a 46.5 a 0.52 b 42.4 20.2 4.2
Productivity of the three media was similar, in terms of number of flowers. However, the perlite flowers were somewhat shorter and thinner. The accumulated dry matter, produced by the perlite-grown plants was also lower than that of the two tuff media. Yield DW of medium tuff was the highest. The efficiency of ionic uptake by the plants, as affected by the various media, is expressed as amount of accumulated elements in the plant tissues. For the sake of simplicity, the elemental content of all plant organs was combined and presented in Table 24.
32
Table 23: Amount (mg plant1) of several nutritional and non-nutritional elements, accumulated over the experimental period by rose plants (CV. Frisco) in 3 growing media. Element N P K Ca Mg Na CI
Coarse tuff 1257
203 1036
401 273 63
80
Medium tuff 1409
213 1128
530 254 59 89
Coarse perlite 933 151 787 320 197
41 98
When taking into account the accumulated dry weight, the relative content of the various elements suggests somewhat clearer picture (Table 25).
Table 25: Content (% in DM) of several nutritional and non-nutritional elements, accumulated over the experimental period by rose plants (CV. Frisco) in 3 growing media. Element N P K Ca Mg Na Cl
Coarse tuff 2.40 0.39 1.98 0.77 0.52 0.12 0.15
Medium tuff 2.45 0.37 1.97 0.92 0.44 0.10 0.16
Coarse perlite 2.00 0.32 1.69 0.69 0.42 0.09 0.21
Based on the above results, as well as on our previous experience, it can be concluded that concentration of nutritional elements in the flowing solutions (see Table 15 for details) were somewhat low for tuff and markedly low for perlite. It is therefore suggested that at the current irrigation capacity (that was rather high) lowering the concentration of nutrients cannot be pushed so much. This conclusion is corroborated by the results of analyses of the media where plants were growing, as shown in Table 26.
Table 26: Content (milimol in 1:10 extraction solution)) of several ions in the media, served for growing plants (CV. Frisco) in recirculating systems. Ion N-NCv N-NH/ P-PO,3
K+
EC (mS cm1) PH
Coarse tuff 4.04 0.24 0.21 1.47 0.78 6.05
Medium tuff 3.13 0.08 0.17 1.06 0.70 6.00
rose
Coarse perlite 0.51 0.08 0.27 0.50 0.25 6.25
It can be clearly seen that except for phosphate, which is strongly adsorbed to tuff surfaces (Silber and Raviv, 1996), other macronutrients are deficient in perlite, as compared to the two tuff fractions. It is therefore suggested that the higher potential productivity of perlite was not expressed in this experiment due to nutritional deficiency. In spite of this and with the intention to avoid chemical conflicts with non-inert material such as tuff, perlite 10 was chosen for our next experiments, aimed at understanding the relative importance of salinity and water irrigation capacity.
3.2.2.5 Experiment C Dissolved oxygen concentrations in the collecting tanks are presented in Table 27.
33
Table 27: DO concentrations (average of 6 measurements, % of saturation) in the collection tanks of closed systems, planted with roses (CV. Frisco), as affected by salinity and irrigation capacity. Different letters denote significant difference at p<0.05 Irrigation capacity
Salinity Low High
Low
80 b 81 b
High
91 a 90 a
Irrigation capacity had a clear (and expected) effect on the DO concentration in the collection tanks. As explained in the Materials and Methods, we measured also the gradient of DO along the system; The DO in the collection tank represented the situation in the first growing container. As solution moved from container to container, its DO level decreased. It was found that the actual DO to which the last plants were exposed was somewhat lower than those measured in the tanks (data not shown). Salinity, on the other hand, had no consistent effect on DO.
Alcohol dehydrogenease (ADH) is an inducible enzyme, serves in the plant to detoxify compounds (specifically ethanol) that are formed during periods of anaerobic respiration. Its occurrence in plant tissues can thus indicate exposure to anaerobic conditions. Due to technical difficulties, ADH content in the roots was determined only for low salt roots. Plant roots grown under low irrigation capacity yielded 18% more ADH than those of high irrigation capacity (10908 vs. 9273 Limol NADH g1 protein min1).
Net assimilation rate by the rose roots (in (imol C02m~2s1) growing conditions is described in Figure 23.
as a function of PPFD, as affected by the
• Flow 2
• Flow 10
A Flow 50
^^^^Logaritmisch (Flow 2)
Logaritmisch (Flow 10)
^^^" logar i tmisch (Row 50)
C Q.
y
y = 8.2512Ln(x)- 39.116
R2 = 0.7989
\
. • s£§S^f*
= 7.7386Ln(x) - 34.97 R2 =0.8189 A
A ^ * ^ 2 A ^ ^ ^ i " * ' - ^
^ • •
X j ^ y /ab3Ln(X) - 36.902
R2 = 0.8698
•
400 500 600 700 800
PPFD(LLmol m-2 s1)
2,-1 Figure 23: NAR of rose (CV. Frisco) leaves (in i mol C02m2s capacities and salinity levels, as a function of PPFD.
grown under different solution irrigation
In general, high irrigation capacity had a beneficial effect, especially at high PPFD levels, irrespective of the salinity level. This is consistent with our previous findings that net carbon exchange rate of roses is not sensitive to salinity in this range (Raviv and Blom, 2001). In this previous research we found, however, that salinity did affect expansive growth, thus markedly affecting leaf area. Leaf area data are presented, together with yield data, in Table 28.
34
Table 28: The effects of salinity and irrigation capacity on leaf area yield and weight parameters of Frisco roses. Different letters in the same row denote statistical difference at p<0.05.
Irrigation capacity/salinity Flowers/plant Ave. flower length (cm) Flower diameter (cm) Total DW/plant (g) Yield DW/plant (g) Leaf area (cm2 plant1)
Low/low 5.1a 42.3 ab
0.52 a 40.11 a 30.39 a
655 a
Low/high 3.7 a
35.7 b 0.44 b
22.96 b 14.53 b
413 b
High/low 5.3 a
45.4 a 0.53 a
41.90 a 30.07 a
632 a
High/high 3.6 a
36.1 b 0.43 b
22.72 b 15.36 b
438 b
Contrary to the absence of salinity effect on net photosynthesis of rose leaves, salinity did affect most yield-related tested parameters. It is plausible that the cause of this phenomenon is salinity effect on expansive growth, as stated above. High salinity had a negative effect on accumulated leaf area per plant, on flower length, weight and diameter and therefore on the total accumulated yield and biomass produced. The effect of the treatments on water relation parameters is shown in Table 29.
able 29: The effects of salinity and Irrigation capacity/salinity Stomatal conductance (cm s1) Specific transpiration rate (nmol H^OmV) Water consumption (L plant1) Specific leaf water consumption (ml cm2) W U E ^ t g r a m L 1 )
rrigation capacity Low/low 596 4.73
13.00 19.8
3.09
on water relation parameters of Low/high 315 3.95
7.78 18.8
2.95
High/low 798 4.78
15.00 23.7
2.79
Frisco roses. High/high 560 4.74
10.14 23.2
2.24
Water relations are affected by both salinity and irrigation capacity. Stomatal conductance was highest at low salinity and high irrigation capacity. Both increased salinity and reduced irrigation capacity lowered stomatal conductance while their combination further reduced stomatal conductance. STR behaved somewhat differently, probably closely reflecting atmospheric demand for water. This could not be matched only by low flow/high salt treatment, due to its low stomatal conductance. STR generally determined the specific water consumption, but periods of very high atmospheric demand that could not be supplied at low irrigation capacity resulted with relatively low values of both low flow treatments. Specific water consumption, coupled with the leaf area determined total water consumption that was, as expected, the lowest for low flow/high salt treatment. It is interesting to note that high flow treatments consumed more water than their low flow counterparts, suggesting higher potential yield (Raviv and Blom, 2001). We have no explanation for the low WUEbiomassof the high flow/high salt treatment. All other treatments exhibited normal values for closed systems (Raviv and Blom, 2001). Both irrigation capacity and salinity affected the concentration of elements in the plant tissues (Table 30).
Table 30: Content (% in DM) of several nutritional and non-nutritional elements, accumulated over the experimental period by rose plants (CV. Frisco) as affected by salinity and irrigation capacity. Irrigation capacity/salinity Element N P K Ca Mg Na Cl
Low/low
2.54 0.28 2.64 0.88 0.24 0.05 0.20
Low/high
2.99 0.44 2.97 0.81 0.22 0.14 0.53
High/low
2.59 0.38 2.60 0.88 0.24 0.04 0.16
High/high
3.14 0.47 2.98 0.88 0.25 0.13 0.41
35
The higher concentrations of the high-salt solutions reflected in higher tissue concentrations of all elements except calcium and magnesium. Higher irrigation capacity increased tissue content of phosphor and caused lower chlorine concentration. Total elemental uptake by the plants was somewhat affected by both salinity and irrigation capacity (Table 31).
Table 31: Amount (mg plant1) of several nutritional and non-nutritional elements, accumulated over the experimental period by rose plants (CV. Frisco), as affected by sa inity and irrigation capacity. Irrigation capacity/salinity Element N P K Ca Mg Na Cl
Low/low
914 100 951 317 88 18 71
Low/high
704 104 700 190 52 34 124
High/low
1086 159 1091 369 100 17 68
High/high
743 111 705 208 59 31 97
Contrary to the impression received by tissue concentrations, when calculated as total elemental uptake, the lower biomass production of the high salt piants resulted with lower elemental yield. A clear exception for this rule is the case of sodium and chlorine, where exclusion mechanisms could not cope with the high ambient concentrations. Still, it is interesting to note that in high flow the increase in these elements was lower than in low flow. This strengthens the case of beneficial effect of high flow. Another exception is the case of phosphor at low irrigation capacity, for which we have no explanation.
Another approach to understanding the effects of various factors on uptake processes is by calculation of the average element concentration in the solution taken up by the plants by dividing the data of Table 30 by their respective transpiration data. Results of such calculation are presented in Table 32.
Table 32: Concentration (mg I"1) of several nutritional and non-nutritional elements in the solution taken rose plants (CV. Frisco) as affected by salinity and irrigation capacity. Irrigation capacity/salinity Element N P K Ca Mg Na CI
Low/low
70.3 7.7 73.2 24.4 6.8 1.4 5.5
Low/high
90.5 13.4 90.0 24.4 6.7 4.4 15.9
High/low
72.4 10.6 72.7 24.6 6.7 1.1 4.5
up by
High/high
73.3 10.9 69.5 20.5 5.8 3.1 9.6
The ionic uptake flux is only marginally affected by the ion concentration in the nutrient solution. It is also quite indifferent to the irrigation capacity, suggesting that it is tightly controlled by physiological factors, mainly growth rate-related. The only notable exceptions are, again, sodium and chlorine. Taking into account their concentration in the saline solution, their uptake rate was extremely low. Their exclusion rate, however, is positively affected by irrigation capacity.
3.2.2.6 Experiment D Average DO concentrations in the systems are presented in table 33. As can be expected, reducing the flow rate, significantly affected DO concentrations. Increased salinity also somewhat reduced DO concentration. However, under our experimental conditions its effect was much less pronounced.
36
Table 33: Average values of DO (percent of saturation, average of weekly measurements throughout the experiment) in the systems, as affected by flow rate and salinity Flow\salinity 0.12 L/H 0.6 L/H 3.0 L/H Average
Low salinity 65.3 a* 79.9 b 91.4 c 78.9
Medium salinity 59.8 a 76.2 b 89.9 c 75.3
High salinity 60.2 a 71.1 b 86.0 c 72.4
Average 61.8 75.7 89.1
Different letters in the same column denote statistical difference at p<0.05.
Net photosynthesis was measured several times during this experiment. As we found previously (Raviv & Blom, 2001), salinity (within the tested range) had no effect on net carbon exchange rate (NCER) (data not shown). On the other hand, flow rate significantly affected NCER (Fig 1). On the average, increasing the flow rate from 0.12 to 3.0 L/H resulted with an increase of NCER of ca. 1.5 |umol C02m2 , -2 s i
• Flow 2
• Row 10
A Row 50
^^^"Logaritmisch (Row 2) 1 Logaritmisch (Row 10)
^^^"Logaritmisch (Row 50)
C
y = 7.7386Ln(x) - 34.97 R> = 0.8189 À
: 8.2512Ln(x)- 39.116 R2 = 0.7989
200 300 400 500 600 700 SOO 900 1000
PPFD^molm^s1)
Figure 23: NCER of rose (CV. Frisco) leaves (in nmol C02 m2 s \ ) , grown in perlite under various flow
rates, as a function of photosynthetic photon flux density (PPFD).
The observed increase in NCER of high-flux plants can result from both the negative effect of oxygen deficiency on NCER of low-flux plants and, indirectly, through the effect of low flow rate on nutrient availability. Unlike the lack of salinity effect on NCER, we found both in the above cited research (Raviv and Blom, 2001) and in this research (previous annual report) that salinity did affect expansive growth, thus markedly affecting leaf area. Total amount of leaf area per plant, developed over the entire experimental period is presented, together with yield data, in Table 33.
37
Table 34: Yield, weight and developed leaf area of Frisco roses, grown under different salinity levels and flow rates. Different letters in the same row denote statistical difference at p<0.05. Treatment
Flow rate (U/H) 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00
Salinity
Low Low Low Low Medium Medium Medium Medium High High High High Average Average Average
Flowers/plant
3.78 be* 4.22 b 5.67 a 4.56 4.11 b 4.33 b 5.00 ab 4.48 4.44 b 3.00 c 4.33 b 3.92 4.11 3.85 5.00
Ave. length (cm)
56 53 54 54 51 59 54 55 50 47 52 50 52 53 53
DW/flower (gr)
9.04 7.59 8.77 8.47 8.49 9.86 7.81 8.72 7.23 8.05 8.56 7.95 8.25 8.50 8.38
Flowers DW/plant (gr)
34.17 32.03 49.73 38.62 34.89 42.69 39.05 39.07 32.10 24.15 37.06 31.16 33.91 32.72 41.90
Developed leaf area (cm2)**
4552 4193 5122 4622 3999 5047 4031 4359 3365 3389 4683 3812 3972 4210 4612
Different letters in the same column denote statistical difference at p<0.05. * * Developed leaf area was calculated by adding the removed leaf area throughout the experiment to the final leaf
area and subtracting the initial leaf area from the above value.
Contrary to the absence of salinity effect on net photosynthesis of rose leaves, salinity did affect most yield-related tested parameters. It is plausible that the cause of this phenomenon is salinity effect on expansive growth, as stated above. High salinity had a negative effect on number of flowers per plant, accumulated leaf area per plant and therefore on the total accumulated biomass produced. In general, productivity of the plants exposed to high salinity was lower than that of plants exposed to either medium or low salinity by an average of 21%.
Increasing the flow rate from 0.60 to 3.0 L/H increased productivity by an average of ca. 26%. The difference resulted from a change in the number of flowers/plant, while flower length and DW of the various treatments were not statistically different from each other. The effect of the flow rate on productivity can be attributed to two factors: the above-mentioned increase in NCER (Fig. 23) and, as mentioned above, to the effect of flow rate on availability of nutrients to the plant roots.
For the sake of simplicity, elemental content of leaves, as a representative organ, are presented in Table 35. The homeostatic nature of plants does not allow major variations among elemental contents of different treatments. Even a difference between no NaCI addition to the recirculating solutions of the low-and medium-salinity media vs. a concentration of ca. 12 mmol NaCI in the high-salinity treatment, resulted with only a relatively modest change in the content of foliar Na and CI. The main phenomenon that worth mentioning is the effect of flow rate on CI content. High flow rate sharply decreased foliar CI content, even under high salinity conditions. Ca uptake was dramatically affected by salinity. Although Ca concentration in the saline solution was twice that of the low- and medium-salinity solutions, much lower Ca content was found in leaves of plants grown under saline conditions. Somewhat different picture can be obtained by looking at the total amount of elements, taken up by the plants. This method of presentation overcomes the masking effect, created by homeostasis, while allowing for the beneficial effects of increased availability, to be expressed, as a result of the higher accumulated DW. Results of "elemental yield" are presented in table 36.
38
Table 35. Content (% in DM) of several nutritional and non-nutritional elements, accumulated over the experimental period by rose leaves (CV. Frisco), as affected by sa inity and flow rate. Treatment Flow rate (L/H) 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00
Salinity
Low Low Low Low Medium Medium Medium Medium High High High High Average Average Average
N
3.20 3.35 3.41 3.32 3.34 3.35 3.59 3.43 3.17 3.58 3.65 3.47 3.24 3.43 3.55
P
0.414 0.505 0.466 0.462 0.501 0.587 0.614 0.567 0.437 0.800 0.442 0.560 0.451 0.631 0.507
K
3.18 3.46 3.08 3.24 3.74 3.46 3.36 3.52 4.11 4.02 3.83 3.99 3.68 3.65 3.42
Table 36. Amount (g plant1) of several nutritional and non-nutritionc experimental period by rose plants (CV. Frisco), as affected by sa Treatment Flow rate (L/H) 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00
Salinity
Low Low Low Low Medium Medium Medium Medium High High High High Average Average Average
N
1.12 1.19 1.53 1.28 1.12 1.38 1.23 1.25 0.91 0.92 1.31 1.05 1.05 1.16 1.36
P
0.17 0.20 0.24 0.20 0.19 0.26 0.24 0.23 0.18 0.20 0.22 0.20 0.18 0.22 0.23
K
1.14 1.26 1.41 1.27 1.22 1.43 1.15 1.27 1.12 1.01 1.34 1.16 1.16 1.23 1.30
Ca
1.63 1.60 2.74 1.99 1.25 1.61 1.44 1.43 1.03 1.11 1.13 1.09 1.30 1.44 1.77
Na
0.05 0.07 0.05 0.06 0.05 0.07 0.05 0.06 0.10 0.09 0.08 0.09 0.07 0.08 0.06
Cl
0.20 0.21 0.19 0.20 0.19 0.17 0.17 0.18 0.53 0.32 0.28 0.38 0.31 0.23 0.21
al elements, accumulated over the nity and flow rate. Ca
0.54 0.56 0.93 0.68 0.53 0.57 0.47 0.52 0.28 0.30 0.37 0.32 0.45 0.48 0.59
Na
0.02 0.03 0.03 0.03 0.03 0.03 0.03 0.03 0.05 0.06 0.06 0.06 0.03 0.04 0.04
Cl
0.06 0.07 0.09 0.07 0.07 0.07 0.07 0.07 0.07 0.09 0.10 0.09 0.07 0.08 0.09
When taking into account the accumulated dry weight, the relative contents of the various elements suggest a clear negative effect of salinity on N and Ca uptake, which is somewhat offset by high flow rate. Chloride accumulation is not greatly affected by salinity, probably due to the osmotic effect that resulted with smaller plants. Sodium accumulation, however, overcame this effect as well as part of the exclusion mechanism. In addition to its effect on N and Ca uptake, high flow rate also improved P and K uptake. The effect of the various treatments on water relation parameters is shown in Table 37.
39
Table 37. The effects of salinity and flow rate on Treatment
Flow rate (L/H) 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00
Salinity
Low Low Low Low Medium Medium Medium Medium High High High High Average Average Average
Stomatal conductance (cm s1)*
242 237 310 263 250 252 250 251 236 203 226 222 243 231 262
water relation parameters of Frisco roses. STR (umol H20 m2s')*
4.77 4.90 6.71 5.46 5.05 5.13 4.63 4.94 5.30 4.34 4.95 4.86 5.04 4.79 5.43
Water consumption (L plant1)
12.31 13.25 13.82 13.13 11.48 14.80 13.22 13.17 10.37 11.07 12.27 11.24 11.39 13.04 13.10
WUE^JgramL1)
3.76 3.59 4.29 3.88 3.77 3.61 3.54 3.64 3.60 3.06 3.83 3.50 3.71 3.42 3.89
* Average of 20 observations
Water relations are affected by both salinity and irrigation capacity. Stomatal conductance was highest at low salinity and high irrigation capacity. Both increased salinity and reduced irrigation capacity lowered stomatal conductance. In this case, STR did not reflect stomatal conductance or water consumption, probably due to its high variability, necessitating a much larger number of observations; to compensate for changing atmospheric demand during the course of measurements and biological variability among sampled leaves. Both low flow rate and high salinity significantly reduced water consumption. It is interesting to note that high flow treatments consumed more water than their low flow counterparts, suggesting higher potential yield (Raviv and Blom, 2001).
Another approach to understanding the effects of various factors on uptake processes is by calculation the average element concentration in the solution taken up by the plants by dividing the data of Table 36 by their respective water consumption data (table 35). Results of such calculation are presented in Table 38.
Table 38. Concentration (mmol I1) of several nutritional and non-nutritional elements in the solution taken up bv rose plants (CV. Frisco) as affected by salinity and irrigation capacity. Treatment Flow rate (L/H) 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00 Average 0.12 0.60 3.00
Salinity
Low Low Low Low Medium Medium Medium Medium High High High High Average Average Average
N
6.51 6.42 7.93 6.95 6.98 6.68 6.65 6.77 6.30 5.94 7.64 6.62 6.60 6.34 7.41
P
0.44 0.48 0.56 0.49 0.55 0.58 0.58 0.57 0.57 0.59 0.57 0.58 0.52 0.55 0.57
K
2.37 2.44 2.62 2.48 2.73 2.49 2.23 2.48 2.78 2.34 2.81 2.64 2.62 2.42 2.55
Ca
2.18 2.13 3.36 2.55 2.30 1.92 1.79 2.00 1.34 1.35 1.51 1.40 1.94 1.80 2.22
Na
0.09 0.10 0.09 0.09 0.10 0.09 0.08 0.09 0.16 0.25 0.23 0.21 0.12 0.15 0.13
Cl
0.15 0.15 0.18 0.16 0.17 0.14 0.14 0.15 0.20 0.22 0.23 0.22 0.17 0.17 0.18
The ionic uptake flux is only marginally affected by the ionic concentration of the nutrient solution,
40
suggesting that it is tightly controlled by physiological factors, rather than concentration in the surrounding solution. For example, an increase in sodium concentration from 0.79 to 13 mM (low- and high-salinity, respectively) resulted with a modest average increase from 0.09 to 0.21 mM in the uptake solution. This remarkable sodium exclusion of rose roots enabled a relatively uninterrupted leaf area development and biomass production (decrease of 18-20%, see table 3) in the high-salt, comparing to the low-salt treatment. This modest growth inhibition almost disappeared under high-flux conditions. The apparent main effect of salinity on ion uptake was expressed in the levels of calcium. In spite of the fact that high-salt solution contained ca. 3 times more ca than the low-salt solutions, the later provided for 82% more Ca++ in the influx solution. In addition to the osmotic effect of high salinity, this may be one of the main causes for reduction in biomass production. Unfortunately, this effect cannot be negated by high flux and this subject calls for additional research, aiming at improving Ca uptake under saline conditions.
3.2.3 Growing media experiment
The presented results cover the period 17/7/2000 - 16/3/2001. Details of water usage are shown in Table 39.
Table 39: Irrigation quantities, amount discharged and consumed (1 m2 day1) and EC (mS cm1) of the various treatments
Treatment 1 2 3 4
Irrigation 7.579 4.914 5.123 6.402
Discharge 4.066 1.076 1.342 1.748
Transpiration 3.513 3.838 3.781 4.654
EC - dripper 2.28 3.03 3.10 3.00
EC- effluent 3.24 4.11 4.14 4.02
Water saving by recirculation was 35% for the regular treatment (# 2) and 16% when 20-30% more water were used for irrigation (treatment 4). 2.3-3.0 I m2 day1 of drainage water were contained. The increased transpiration of treatment 4 can be explained by the greater water availability. Amount of the various ions contained and discharged by the various treatments is shown in Table 40.
Table 40. Amount of selected ions (mg m2
during the period 17/7 - 1/12/2000. Treatment 1 contained 1 discharged 2 contained 2 discharged 3 contained 3 discharged 4 contained 4 discharged
N(NO, + NH/) 519 684 560 291 518 348 468 .470
day1) contained and discharged by the various treatments
P - PO„3
199 77 170 25 170 28 178 37
K+
275 1283 545 558 485 637 348 867
Na+
218 1553 395 698 336 776 207 997
CI 336 2705 977 1176 857 1333 704 1668
The amount of nitrogen contained in the systems did not differ substantially among the treatments. It represents quite closely the amount of biomass produced (see table 41). Effluent recirculation reduced the amount of discharged N by 57%. The reduction in discharged N (393 mg m2 day1) could bring about 11m3
pure water to above the European standards for potable water. This figure dramatically illustrates one of the main environmental benefits of drainage water recirculation. The use of higher water amounts and increased irrigation frequency necessarily resulted with increased discharge but had no effect on containment of N. Similar picture was obtained with P. Water recirculation considerably reduced the discharge of K, Na and CI. K and Na probably replaced H+ and other cations on adsorbing sites. Moreover, organic acids, secreted by rose roots, gradually increasing tuff's CEC, thus enabling continued adsorption of cations. Another containment mechanism results from increased plant uptake from concentrated solutions. Similarly to our previous findings (Raviv et al., 1998 and last year's report), yield parameters show that the performance of the rose plants was not markedly damaged by the use of recycled effluents
41
at EC of up to 3 mS cm1 and that the slight damage was offset by increasing the recirculation rate (Table 41).
Table 41. The effect of water recirculation and irrigation frequency on yield parameters of rose pants (CV. Golden Gate) over the period 17/7/2000 - 16/3/2001. Different letters in the same column denote statistical difference at p<0.05.
Treatment
1 2 3 4
Flowers m2 day1
0.75 a 0.71a 0.72 a 0.79 a
Marketable flowers m2 day1
0.55 a 0.55 a 0.54 a 0.59 a
Flower length (cm)
53.8 a 51.6 b 52.0 b 52.5 b
Flower weight (gram) 28.7 a 27.2 a 27.0 a 28.1 a
Fresh weight (gram m^day1)
21.5a 19.3 b 19.4 b 22.2 a
Although number of flowers and their weight was not affected by the increased salinity of the recirculated solution, the produced fresh biomass was lower in treatments 2 and 3. This effect was negated by increasing the flow rate (treatment 4). Average flower length was also affected by salinity, resulted with a significant shortening of the flowers by 1.3-2.2 cm.
No statistical differences were found among the tested vase life parameters of the flowers of the various treatments. In both sampling dates the vase life was 14.5-15.5 days in the test room with no uniform trend. The maximal flower diameter was 9.1-9.7 cm., again, without significant difference, although a slight trend towards larger flowers of treatment #4 could be observed.
42
4 Discussion and conclusions
4.1 The Netherlands
One of the prerequisites of the project was that the oxygen supply would not be limiting under high-flux irrigation conditions since water content would be close to container capacity. From the results in Fig. 1 compared to the physical characteristics, it can be concluded that pressure head was between -10 cm and -32 cm in the 15 cm high containers. However, under these conditions air-filled porosity (AFP) was still at a minimum of 25% (Fig. 1). Although the induction of the enzyme ADH in the roots was somewhat higher in the growing media fine perlite and coir (Table 4, Fig. 5) the differences between the growing media were small. Furthermore, gas composition in the bottom of the containers did hardly differ from ambient conditions (Table 11), and root growth (Table 7) and production (Table 15) was best in the 'wet' growing media. It is therefore concluded that the pants did not suffer from oxygen-deficient conditions under the given circumstances. The results are in accordance with other experimental and simulation data on oxygen supply and stress in growing media. For instance, rose cuttings showed reduced root and shoot growth in rockwool blocks at AFP only below 25% around the roots (Baas et al. 1997). For chrysanthemum it was concluded that oxygen stress could be found at air-filled porosities already below 35%, although the effects very much depended on climatic conditions (Baas and Warmenhoven 1995). In cucumber in rockwool, no effects of high water contents on growth were found; gas composition in the rhizosphere only started to differ at AFP below 25% (Baas 2001). Finally, using respiration data of cucumber in combination with oxygen diffusion coefficients determined in fine perlite and rockwool, it was calculated that at an AFP of 30% or higher, no oxygen depletion in rockwool or perlite would occur (Baas et al 2001).
4.1.1 Growing media effects
Differences in production were found between the growing media. Particularly the coarse perlite lagged behind in number of stems and total weight production per plant (Table 15). This reduction was already noticed in the first year of the project, and was related to the inferior root growth as monitored during this year (later on it was impossible to establish root weight). Since there were no irrigation frequency * growing media interactions found, and also no or hardly any effects on leaf nutrient concentrations, the growth could not be related to decreased nutrient availblility in coarse perlite. It is suggested that the penetration resistance in coarse perlite is higher than in the other growing media. The middle fraction perlite possibly with a presumed intermediate resistance showed intermediate root growth and production. It may be speculated that the decreased root growth may have an influence on the production and export of cytokinins, which can have an influence on shoot formation/bud break (Dieleman 1998).
4.1.2 EC and NaCI effects
The project has shown that in the absence of NaCI, EC in the nutrient solution could be decreased to values of 0.9-1.0 mS/cm without adverse effects on production. Indeed, in the final experiment, stem weight was even highest in the EC 0.9 treatment. Although cut rose is not particularly salinity-sensitive (Sonneveld et al, Raviv 2001) salinity tests usually were at EC levels starting from ca. 2 mS/cm. Cultivar Sonia grown in rockwool at low supply rates (de Kreij and van der berg 1990) showed decreased production at EC 1.0 probably due to nutrient deficiency. Under our high-flux irrigation conditions, nutrient supply apparently was sufficient, although total N, P and K concentrations in the leaf were lower at EC0.9 compared to EC 1.9 (Tables 10 and 13). In contrast, Mg concentration were higher at low EC, possibly due to less K/Mg
43
antagonism. More research on production under high-flux irrigation conditions at low nutrient concentrations is needed to extend the results to other cultivars and conditions. The advantage of using ECO.9 in open systems compared to e.g. EC 1.6 with respect to nutrient emission is obvious. In closed systems, the advantage of decreasing the EC of the nutrient solution is only relevant in case drainage solution is discarded in case of inacceptable high concentrations of e.g. Na and CI. In the presence of NaCI, the positive effect of a low EC was counteracted. At EC 0.9 the addition of 10 mM NaCI negatively affected production, whereas this was not the case at EC1.9 (Table 11). The effect of NaCI on production was related to higher CI concentrations in the leaf, but could - with the exception of slightly lower Mg concentrations - not be related to deficiencies. Na concentrations in the leaves were - as found before (Baas en van den Berg 2000) - very low (Table 12). Since particularly the number of stems appeared to be reduced by the presence of NaCI (Table 11), it might be speculated that root formation/root death may have been affected by NaCI at low EC.
4.1.3 Discharge reduction possibilities
For the Netherlands legislation, Na concentration in the drainage water has to be at least 4 mM to allow run-off of the drainage solution in cut rose cultivation; the outcome of this research shows that the concentration of Na can be increased to 10 mM without any yield or quality decrease. Which savings in terms of fertilizer reduction can be realized by increasing this Na concentration up to 10 mM? The leaching fraction needed to maintain the maximum Na concentration (Fd) in the drainage solution is given by (Sonneveld and Van der Burg 1991): Fd=(Naw + NarNac)/(Nad-Nac) in which Naw is the Na concentration of the irrigation water, Naf the increase of the Na concentration by impurities of fertilizer addition, and Nad the Na concentration of the drainage water. Nac is the Na removal (crop uptake and loss in growing medium due to precipatition and adsorption). Nac has been shown to be linearly related to (Naw + Naf) in e.g. Madelon as: Nac = 0.005*(Naw+Naf) + 0.037 (Baas and van der Berg 2000). Table 42 shows some that, nutrient emission can be reduced by 60 % by increasing Nad from 4 to 10 mM.
Table 42. Calculated drainage fractions needed to maintain constant Na concentration in drainage water (Nad)as dependent on the water quality and fertilizer input (Naw + Naf) and removal concentration (Nac). Nac
Based on data obtained with cv. Madelon Baas and van der Berg 2000). Na, 0.1 0.1 2.0 2.0 3.5 3.5
Na, 0.5 0.5 0.5 0.5 0.5 0.5
Na 0.057 0.087 0.057 0.087 0.057 0.087
Na
10
10
10
0.14 0.06 0.62 0.24
1.00 0.39
4.1.4 Transpiration, Water use efficiency
Transpiration was followed weekly during the experiment (Fig. 13) and correlated with irradiation (Fig. 14). Maximum transpiration was 0.75 l/plant.day. At the planting density of 10 plants/m2 this would mean 7.5 mm/day at an irradiation sum of 2200 J/cm2.day. This transpiration is 50% higher than the ca. 5 mm/day found for cultivar First Red at this irradiation sum (de Hoog 1998). These differences may be the result of differences in cultivar, planting density, crop stage and greenhouse conditions (light transmittance, C02
concentrations, VPD, ventilation rate). The scatter in Fig. 14 shows that 50% difference in transpiration rate can already be caused by crop stage and greenhouse conditions. Irrigation quantity should be sufficient to cope with the maximum transpiration rates as given in Fig. 14: T = 3 E11 * I3 + 2 E"8 * I2 + 8 E5 * I + 0.2 with T in l/plant.day and I in J/cm2.day
44
Water Use Efficiency (WUE) as determined from transpiration and production data (Fig. 18) showed a remarkable peak during the winter period, when assimilation lighting (on at light intensities less than 100 J/s.m2) was the predominant light source. Under summer conditions WUE was between 12-15 g FWAg H20. At a dry matter percentage of 25% WUE would be 34 g DMAg H20 which is considerably higher than the 2.3-2.8 g DMAg H20 as given from data collected in recirculation systems (Raviv 2001).
4.2 Israel
4.2.2 Physical characteristics of growing media
The results of the present study suggest that physical characteristics of the media, specifically water release curve, saturated and unsaturated hydraulic conductivities largely determine plant performance. This occurs only when nutrition and chemical status of the root zone are not limiting factors, as found in perlite-grown plants, that suffered from short events of low pH and low calcium availability (as manifested in Ca content of the collected plant material, Table 21). The performance of tuff-grown plants was closely related to their physical traits. Very low air filled porosity (AFP) and low hydraulic conductivity characterizes fine tuff while very low available water content and low unsaturated hydraulic conductivity characterize coarse tuff. Tuff of a medium particle size has adequate AFP, available water content and hydraulic conductivity (Figures 20 and 21). Consequently, it yielded the highest yields (Table 20) and its photon quantum efficiency was the highest (Figure 22). The results of the first year suggest that a priori physical characterization of growing media, especially as related to AFP and unsaturated hydraulic conductivity, can be used to predict potential productivity of substrates.
Nutrient solutions of all tested media contain, under normal working conditions, enough DO to support root function. At regular irrigation capacity (3.6 I h"1, experiment A) it can be assumed that the media are at all times at or near container capacity. It is interesting to note that slow irrigation capacity (below 0.61 h1), which should enable more air into large pores, is accompanied by lower rather than higher DO levels (Tables 27 and 33). This suggests that DO delivery using mass flow is faster than its replenishment through dissolution of gaseous oxygen or that oxygen enrichment of the nutrient solution by a higher circulation rate plays a significant role. This coincides with one of the main hypotheses of this research.
4.2.3 Salinity effects on plant performance
As expected, salinity resulted with lower yields in the hydroponic experiments (Tables 28 and 34). This, however, was not the case in the semi-commercial experiment (Table 41). The increased salinity under recirculation conditions resulted with a slight negative effect on flower length (Table 41) while number of flowers or of marketable flowers was not affected. Total productivity, as reflected by production of FW was lowered by the increased salinity by -10%. This negative effect was completely negated by increasing irrigation frequency. These contradictory results cannot stem only from the difference in salinity levels in the two cases (4.4 and 3.06 mS cm * in the first and second hydroponic experiments, respectively, and -3.0 mS cm1 in the medium experiment). Our current working hypothesis is that rose plants can cope with high salinity after a certain period of adaptation. The hydroponic experiments were probably too short to enable such adaptation.
Salinity effect may be attributed to 3 factors: the negative effect of osmotic potential on water uptake, specific detrimental effect of toxic ions such as Na+ and the negative effect of high salinity on Ca uptake. The first two factors were only marginally affected by high salinity (Tables 29, 36). Calcium uptake, on the other hand, showed clear reduced uptake under saline conditions (Table 36). Increasing the flow rate only partly overcame this effect. Under practical conditions of high-salinity closed systems foliar Ca content must be carefully monitored in order to prevent any possible deficiency, caused by high osmotic potential in the rhizosphere.
45
4.2.3 Flow rate effects on plant performance
One of the hypotheses of this research was that high flow rate of the nutrient solution can compensate for low level of nutrients, by increasing their availability. We found that this is, indeed, the case, as can be seen in Table 35. This suggests that closed growing systems, operated under irrigation capacities of >20 I m2 h 1 can support normal plant production using nutrient concentrations far below common values. Increasing the flow rate within the root zone increased the concentration of DO (Tables 27 and 33). It is assumed that it also increased water and nutrient availability. As a result, increased flow rate exerted profound effects on plant's physiological responses. Examples for such effects are increased stomatal conductance and STR (Table 29), increased photosynthetic efficiency (Figures 23 and 24), increased transpiration (Table 39), decrease in root ADH activity, enhanced P (Table 30) and Ca (Table 35) uptake, decrease in CI uptake under saline conditions (Tables 30 and 35). These potentially beneficial responses were not always reflected in yield increase, probably due to other, yet to be identified, factors. In some cases, however, increased flow rate led to higher number of flowers and leaf area under saline conditions (Table 34) and to increased harvested FW (Table 41). These results are in agreement with the research's hypothesis. It should be mentioned here that the limitations of our semi-commercial systems prevented us from further increasing the flow rate, to values similar to those exerted positive effect in the experimental closed systems. In the hydroponic experiment, high flow rate had a positive effect on plant performance. Moreover, high flow rate enabled a decrease in nutrient's concentration. In our medium experimental system such a flow rate is not possible and the maximal flow rate (3 x as high as in the commercial system) only marginally improved plant performance. This is in accoradance with the absence of any frequency effects on production parameters in the Dutch experiments. The system's limitations are not inherent to any commercial system and, if required, a commercial system can be devised to accommodate the recommended irrigation frequency. Based on these results, it is strongly suggested to increase the irrigation capacity and frequency in commercial greenhouses, provided that the hydraulic conductivity of the media is large enough to allow uninterrupted drainage. This can only be practiced in closed systems.
4.2.4 Discharge reduction possibilities
The economical savings of both water and fertilizers, as found in this study, is around 1$ m2 year1
(treatment 2, Table 39). This calculation is based on an average water consumption of 3.5 m3 m2 year1 in open systems at a cost (including fertilizers) of about $1 per m3 and savings of 30-35% of the water consumption by recirculation. For the Israeli rose growers this represents annual savings of $2.5 million. For the national water economy this is reflected in savings of 2.5 million m3. The environmental impact of such practice is highly desirable, preventing some 1600 tons of nitrates from entering to water reservoirs. At higher flow rates (e.g. treatment 4), lower savings can be expected (around 0.5 $ m2 year1) but the assumed (although not statistically proven) yield increase of $2.2 m2 year1 should justify this policy. In reality, most Israeli growers that recirculate drainage water (currently about 40% of all growers, and growing) widely adopted this policy, claiming their yields were actually increased after switching to recycled systems.
46
5 Description of cooperation
During the project the project leader R. Baas visited M. Raviv in Israel in february 1999 and may 2000 to discuss results and ongoing experiments. By e-mail the two groups frequently exchanged data and methods used. The senior Israeli technician spent some time in the Dutch lab to study ADH analysis.
47
6 Evaluation of the research achievements with respect to the aims of the original research proposal
6.1 Netherlands
Objective 1) To determine optimal physical characteristics of a growing medium to be used under high-flux saline fertigation by using parameters indicating oxygen deficiency.
With respect to the first objective several indications were found that oxygen deficiency was not a major problem under the given circumstances. Although oxygen supply was slightly hampered in the growing media fine perlite and coir as determined by root ADH activity and gas composition of the rhizosphere, yield and root growth clearly was not affected. The media containing an air-filled porosity of at least 30% therefore proved to be adequate for high-flux fertigation.
Objective 2) To determine threshold conditions (refreshment rate, nutrient concentrations, osmotic potential) which limit physiological plant processes related to growing conditions in closed nutrient systems.
With respect to the second objective no effects of irrigation frequency on production were found. With respect to leaf nutrient concentrations only on one occasion lower -but still adequate- P concentrations were found. Apparently, the lower limiting threshold irrigation was not reached in the project, which is not surprising considering the irrigation amounts to be well above irrigation rates normally used in horticultural practice. With respect to nutrient and NaCI concentration an interesting interaction was found. Without NaCI in the nutrient solution, EC could be decreased succesfully to less than 1 mS/cm. However, in the presence of NaCI, yield decreased compared to the control treatment of 1.9 mS/cm without NaCI. In this respect the threshold conditions became evident.
6.2 Israel
Objective 1) to determine optimal physical characteristics of a growing medium to be used under high-flux saline fertigation by using parameters indicating oxygen deficiency.
Determination of physical properties, suitable for high-flux irrigation was conducted during the first year by growing rose plants in fractionated tuff and perlite. It was found that high-flux irrigation is compatible with a wide range of physical properties. Two tuff fractions (4 and 20) resulted with lower biomass production. Tuff 4 (particle size >2 mm, no fine particles) has low unsaturated hydraulic conductivity (K) in the relevant matric potential (about 0.5 kPa). Within this range its K is an order of magnitude lower than that of tuff 10 (most particles in the range of 24.75 mm). Tuff 20 (particle size <2 mm) has low AFP (<3% between 0 and 10 cm pressure head)). In spite of the fact that constant irrigation means also constant fresh supply of dissolved oxygen, it appears that a somewhat larger AFP is required to satisfy the rhizosphere need for oxygen.
Objective 2) To determine threshold conditions (refreshment rate, nutrient concentrations, osmotic potential) which limit physiological plant processes related to growing conditions in closed nutrient systems.
In relation to the second specific objective, we tried to define threshold levels for salinity and flow rate in closed systems. Up to the level of 3 mS cm ] it is possible to counteract negative salinity effects by increasing the flow rate. This was found both in mini systems with constant flow and in a semi commercial system with intermittent, high-frequency irrigation. In completely closed mini systems, a flow rate of 3 L h_1
was sufficient to reverse salinity-mediated effects such as reduction in N uptake rate. This corresponds to
48
about >100 L m2 day ~\ far above the current values (about 18 L m2 day_1 in warm summer days). A good measure for flow rate is the resulting dissolved oxygen content. In addition to the flow rate, it is also affected by root respiration but as we assume high respiration rate as a desirable phenomenon DO as a setpoint must be high enough to prevent any hidden oxygen deficiency. Our results suggest that DO has to be >80% of saturation. At this level we found decreases in NCER, stomatal conductivity, STR and nitrogen uptake. In respect to nutrient concentration, no exact threshold were determined. It was shown, however, that using high flow rate, plant can tolerate low nutrient concentrations (in mM: 2.33 N, 0.13 P and 1.74 K), as compared to normal values. This was evidenced by both the nutritional status of the plants (Tables 35 and 38) and yield (Table 34).
These results suggest that recirculation of drainage water is a practical alternative, even at relatively high EC values. High flow rate can compensate for decreased nutritional concentration. This can further decrease the discharge rate, while lowering the cost of fertilizers and the environmental hazards.
49
7 References
Baas R 2001. Verminderde zuurstofvoorziening in opkweekblokken komkommer heeft weinig gevolgen. Groente en Fruit 2 maart 2001: 18-19.
Baas R, Warmenhoven MG 1995. Alcoholdehydrogenase indicating oxygen deficiency in chrysanthemum grown in mineral media. Acta Hort. 401: 273-282.
Baas R, Gisler0d HR, Van den Berg TJM 1997. Do roots of rose cuttings suffer from oxygen deficiency during propagation in rockwool? Acta Hort. 450: 123-131.
Baas R, Van den Berg 2000. Sodium accumulation and nutrient discharge in recirculation systems: a case study with roses. Acta Hort. 507: 157-164.
Baas R, Wever G, Kooien AJ, Tariku E, Stol KJ 2001. Oxygen supply and consumption in soilless culture: evaluation of an oxygen simulation model for cucumber. Acta Hort, (in press)
Dieleman JA 1998. Cytokinins and bud break in rose combination plants. Thesis Wageningen University ISBN 90-5485-850-8.
De Kreij C, Van den Berg T.J.M., 1990. Nutrient uptake, production and quality of Rosa hybrida'm rockwool as affected by EC of the nutrient solution. Proc. Xlth Int. Plant Nutrition Colloq. 1989, Kluwer, Dordrecht: 519-523.
Kreij C de, Sonneveld C, Warmenhoven MG, Straver NA 1990. Guide values for nutrient content of vegetables and flowers under glass. Research station for Floriculture and Greenhouse Vegetables report no. 15.
Raviv, M., A. Krasnovsky, Sh. Medina and R. Reuveni (1998). Assessment of various control strategies for recirculation of greenhouse effluents under semi-arid conditions. The Journal of Horticultural Science and Biotechnology1'3:485491.
Raviv M, Blom TJ 2001. The effect of water availability and quality on photosynthesis and productivity of soilless-grown cut roses. A review. Scientia Hort. 88: 257-276.
Sonneveld C, Van der Burg, 1991. Sodium chloride salinity in fruit vegetable crops in soilless culture. Neth. J. Agric. Sei. 39:115-122.
Sonneveld C, Baas R, Nijssen HMC, De Hoog J 1999. Salt tolerance of flower crops grown in soilless culture. J. of PI. Nutr. 22(6): 1033-1048.
Silber, A. and M. Raviv (1996). Effects on chemical surface properties of tuff by growing rose plants. Plant and Soil186: 353-360.
Publications
Baas R, Berg TJM van den 2002. Limiting nutrient emission from a cut rose closed system by high-flux irrigation and low nutrient concentrations? Acta Hort, (submitted)
50
8 Appendix
Appendix 1 . Layout of the experimental plots of experiment 2 in the greenhouses in the Netherlands, and position of the FD-sensors.
Freq. l x / 90 min EC 0.9
SI
S2
S3
SI
S3
S4 FD FD FD
S4
S2
Freq. l x /45 min EC 1.9
Freq. l x /45 min EC 0.9
S4 FD FD FD
S2
S4
SI
S3
SI
S3
S2
Freq. l x /180 min EC 1.9
Freq. l x /180 min EC 0.9
S4 FD FD FD
S2
S2
S3
SI
S3
SI
S4
Freq. l x / 90 min EC 1.9
Freq. l x / 45 min EC 1.9
SI
S4 FD FD FD
S2
S4
S2
S3
S3
SI
Freq. l x / 180 min EC 0.9
Freq. l x / 90 min EC 1.9
S4 FD FD FD
S2
S4
S2
S3
SI
SI
S3
Freq. l x / 45 min EC 0.9
Freq. l x /180 min EC 1.9
S4 FD FD FD
SI
S3
S2
S2
S3
SI
S4
Freq. l x / 90 min EC 0.9
S4: coir S3: fine perliteO-lmm S2: mid perlite 0.6-2.5 mm S I : coarse perlite 1-7.5 mm
51
Appendix 2. Guide values and deficiency concentrations (mmolAg DM) of nutrients in young fully-grown leaves of cut rose. Source de Kreij et al 1992.
element
K Ca Mg Total IM P CI Na
Lower guide value
800 250 90 1700 100 20 2
Upper guide value
900 450 160 2800 200 50 15
Deficiency Concnetrations <460 <250 <80 <1430 <65
52