Page 28 - Volume 1, Issue 2, July 2008
Basic HTML Version
Troyanos & Hipps
80
cm) filled with medium grade quartz sand.
The sand had been washed with 3% v/v of
concentrated HCl and leached with de-
ionized water (Conductivity
≤
3 μS cm
-1
)
(10) until the pH of nutrient solution drain-
ing from the pots was approx. 5.5. The
analyses of sand samples that taken from
the pots and extracted with 1 M ammoni-
um acetate at pH = 7 showed that the sand
had a residual concentration of Mg of ap-
prox. 30 μmol l
-1
. When the sand leached,
the plants were weighted and planted.
During their establishment, they were wa-
tered on alternate days with 600 ml of half
strength of the nutrient solution present-
ed in Table 1. After two weeks, when the
plants started growing, 600 ml of the full
strength nutrient solution (Table 1) were
applied each day. Three different concen-
trations of Mg: 0, 150 and 1500 μmol l
-1
us-
ing MgCl
2
.6H
2
0 were applied to the plants.
Growth and mineral analyses
During the experiment, four destructive
harvests were done at 30, 50, 71 and 92 days
after transplanting (DAP) by removing the
plants from the pots. The plants were divid-
ed into leaves, stems, new shoots and roots
and were rinsed three times with de-ion-
ized water. The fresh weights of the leaves,
stems and roots were measured. After-
wards, the plant material was dried to con-
stant weight in a forced air oven at 85
0
C for
determination of dry weight and ground for
determination of the nutrient content. The
leaf area (L
A
) was measured with a Delta-T
leaf area meter (Burwell, Cambridge, Eng-
land) and the diameter of the stems was
measured at 10 cm above the root collar.
One hundred milligrams of the dried
ground plant material was digested with
2 ml 18 M H
2
SO
4
containing 1 g l
-1
Se and 1
ml hydrogen peroxide. After cooling, the
digests were made up to 20 ml with dis-
tilled water. Ca and Mg were determined
by atomic absorption spectrophotometer
on the digests after dilution with 20 vol-
umes of a 1 g l
-1
solution of LaCl
3
as a releas-
ing agent. K was determined on the same
solution by atomic emission photometer.
Total nitrogen was determined using an
auto- analyzer by indophenol blue meth-
od (23) and phosphorus determined by
the molybdenum blue method (6).
Experimental design
A complete randomized block design
was used with three [Mg] : 0, 150 and 1500
μmol l
-1
as the treatments allocated to four
blocks and four harvest dates. To take ac-
count of the variation due to initial plant
weight, the plants were put into 4 groups
of 12 plants depending on their initial
fresh weight as follows: the largest being
in the 1
st
group, and the smallest in the 4
th
group. The 12 plants in each group were
then assigned at random to each block
and each plant was assigned at random
to each Mg and harvest combination. By
using this randomization the difference in
the initial plant weight was included into
the block effect (2
)
. At each harvest, one
plant was removed from each pot thus, 12
plants were removed per harvest.
Statistical analyses
The statistical analyses were performed
with Genstat (9) (Genstat 5 Committee,
1987). Plant growth data were analysed us-
ing polynomial regression models fitted to
the data which were logarithmically trans-
Table 1.
Composition of the nutrient solution.
Macronutrients
μmol l
-1
Ca(NO
3
)
2
.4H
2
O
4000
KNO
3
4000
NaH
2
PO
4
.2H
2
O
1330
MgCl
2
.2H
2
O
0, 150, 1500
CaSO
4
.H
2
O
500
Micronutrients
As in Long-Ashton
nutrient solution
FeEDTA
100
pH
5.5
