Agriculture Reference
In-Depth Information
experiments for fertilizer recommendations is 3 years. However, if it is conducted at several loca-
tions, a 2 year period can also serve as a reasonably good time period to make recommendations.
When conducting field experiments for fertilizer recommendations, care should be taken to
select appropriate rates or levels of a given nutrient. The levels selected should create a wide range
of nutrients in the soil to satisfy the crop needs for a given nutrient from a deficiency level to a
sufficiency level. Grain yield is the best parameter to determine nutrient requirements or fertilizer
recommendations for grain crops. Hence, grain yield should be determined at physiological matu-
rity and the relationship between grain yield and nutrient rates should be determined. Analysis of
variance should be used in data analysis, and the quadratic regression model is used to describe the
yield response to the applied fertilizer rate. The quadratic response function is the most common
functional form to evaluate the yield response to fertilizer rates. The quadratic model is a second-
order polynomial function written as
Y = a + bx + cx
2
where Y is the estimated yield, X is the application rate of the nutrient, and a, b, and c are coef-
ficients estimated by fitting the model to the data. The quadratic function assumes that crop yield
will increase at a decreasing rate as the nutrient application rate increases until the maximum yield
is achieved at
N(Y
max
) = b/2c
where N(Y
max
) is the level of the applied nutrient that achieves the maximum yield. Yield decreases
past this point.
The author conducted field experiments for upland and lowland rice grown under Brazilian con-
ditions using promising genotypes. Results related to lowland rice response to N rates are presented
in Figure 8.17. The response of lowland rice to N fertilization was highly significant and quadratic
when N was applied in the range of 0-200 kg ha
−1
. Based on a quadratic regression equation, a
maximum grain yield of about 6250 kg ha
−1
was obtained with the application of 171 kg N ha
−1
.
Half of the N was applied at the time of sowing and half at 45 days after sowing corresponding to
the active tillering growth stage. Similarly, the author also studied the response of upland rice to N
8000
6000
4000
Y = 3620.7830 + 33.3111X - 0.0974X
2
R
2
= 0.8511**
2000
0
0
40
80
120
160
200
Nitrogen application rate (kg N ha
-1
)
FIGURE 8.17
Response of lowland rice to nitrogen application.
