The effect
of natural rainfall and N, P and K nutrients on the yield of maize was
investigated in 16 years of a long-term fertilization experiment set up at the
Experimental Station of the Institute in Nagyhörcsök. The soil was a calcareous
chernozem, having the following characteristics: pH (KCl): 7.3, CaCO
3
:
5%, humus: 3%, clay: 20-22%, AL-soluble P
2
O
5
: 60-80,
AL-soluble K
2
O: 180-200, KCl-soluble Mg: 150-180; KCl+ EDTA-soluble
Mn, Cu and Zn content: 80-150, 2-3 and 1-2 mg·kg
-1
. The experiment
had a split-split-plot design with 20 treatments in 4 replications, giving a
total of 80 plots. The treatments involved three levels each of N and P and two
levels of K in all possible combinations (3×3×2=18), together with an untreated
control and one treatment with a higher rate of NPK, not included in the
factorial system. The main results can be summarized as follows: An analysis of
the weather in the 16 experimental years revealed that there were no average
years, as two years were moderately dry (1981, 1982), eight were very dry
(1973, 1978, 1986, 1989, 1990, 1993, 1997, 2002) and six were very wet (1969,
1974, 1977, 1994, 1998, 2001). In dry years the N, NP and NK treatments led to
a yield increment of over 3.0 t·ha
-1
(3.2 t·ha
-1
) (81%) compared with the unfertilized control, while the
full NPK treatment caused hardly any increase in the maize yield (7.2 t·ha
-1
).
In the case of drought there was a 4.0% yield loss in the N, NP and NK
treatments compared to the same treatments in the dry years. This loss was only
1.0% in the NPK treatment. In very wet years the positive effects of a
favourable water supply could be seen even in the N, NP and NK treatments (with
yields of around 7.4 t·ha
-1
). The yield increment in these
treatments compared with the droughty years averaged 8%, while balanced NPK
fertilization led to a further 2% increase (10%). Significant quadratic
correlations were found between the rainfall quantity during the vegetation
period and the yield, depending on the nutrient supplies (Ø: R =
0.7787***, N: R = 0.8997***, NP: R = 0.9338***, NK: R = 0.9574***, NPK: R =
0.8906***). The optimum rainfall quantity and the corresponding grain yield
ranged from 328-349 mm and 5.0-7.7 t·ha
-1
, respectively, depending
on the fertilizer rate. The grain yield increment obtained per mm rainfall in
the case of optimum rainfall supplies was found to be 14.3-23.2 kg·ha
-1
,
while the quantity of rainfall utilized during the vegetation period for the
production of 1 kg air-dry matter in the case of maximum yield amounted to 698,
449, 480, 466 and 431 litres in the control, N, NP, NK and NPK treatments,
respectively. It was clear from the 43-year meteorological database for the
experimental station (1961-2003) that over the last 23 years (1981-2003) the
weather has become substantially drier. Compared with the data for the previous
20 years (1961-1980) there was an increase of 20, 500 and 50% in the number of
average, dry and droughty years, no change in the number of wet years and a 71% drop in the number of
very wet years.