231 
1983 
the effect 
int, posi- 
rhe sowing 
up to the 
overruled 
-eflectance 
Lbution of 
; described 
he growing 
vegetative 
irared re- 
Lfference 
bsequently 
used for 
)f a special 
his latter 
dated, which 
;rsion prob- 
the vege- 
ression of 
5 described 
this func- 
ind 9 (see 
e, the re- 
frared re- 
g a training 
h both re- 
sequently, 
for estima- 
rs, 1986b). 
e analysed 
ed signif- 
ing season 
d infrared 
influence 
ng of the 
red and 
CORRECTED 
INFRARED REFL. «> 
FIELD TRIAL 116 ir 
d infrared 
stage, 1982. 
Figure 10: Seasonal change in corrected infrared 
reflectance for field trial 116 in 1983 (cf. 
figure 3). 
ESTIMATED LAI 
FIELD TRIAL 116 in 1983 
Figure 11: Seasonal change in estimated LAI for 
field trial 116 in 1983. 
FIELO TRIAL 116 in 1983 
Figure 12: Seasonal change in measured LAI (field 
measurements) for field trial 116 in 1983. 
infrared reflectances in figure 3). The results for 
the estimated LAI are illustrated in figure 11. The 
same remarks about treatment effects are valid as 
for the infrared reflectance in field trial 116 in 
1983. These results are also in agreement with the 
results of the LAI as measured in the field by har 
vesting plants (figure 12). 
3.3 Analysis of variance 
A very important analysis of field measurements from 
agricultural field trials is an analysis of variance 
in order to investigate whether treatment effects 
are significant and whether interactions (meaning 
that the effects of one factor are dependent on the 
levels of another factor, so that they are not mere 
ly additive) occur. Analogously with field measure 
ments, an anlysis of variance was carried out on 
the estimated LAI from reflectance values. 
Table 1 lists the means, estimated coefficients 
of variation (CV) of residuals and the critical 
levels in testing for treatment effects, for the 
LAI (field measurement) in field trial 116 in 1983. 
The variance was determined by means of appropriate 
averaging of inter-plot and intra-plot variances. 
The CV is the ratio of the square root of the mean 
squares of residuals and the mean value of the de 
pendent variable. The critical level (often called 
P-value) is the smallest level of significance at 
which the observed result would just lead to rejec 
tion of the null hypothesis (that a treatment has 
no effects). On 31 May and 6 June the interaction 
effect between sowing date and nitrogen nutrition 
was significant at the 5% level. The sowing date 
effect and the nitrogen nutrition effect were sig 
nificant on most dates. The large critical level 
for sowing date effect on 6 June was caused by the 
fact that the late-sown crop had a smaller LAI than 
the early-sown crop at the beginning of the season 
and caught up with the early-sown crop at the begin 
ning of June, resulting in a larger LAI for the late- 
sown crop at the end of the season. Coefficients of 
variation were quite large. 
Results of the analysis of variance for the esti 
mated LAI from reflectance measurements in field 
trial 116 in 1983 are listed in table 2. On most 
dates critical levels in testing for treatment ef 
fects by means of the estimated LAI were smaller 
than those obtained from field measurements of LAI 
(table 1). This applied in particular to the nitro 
gen nutrition effect. The large critical level for 
sowing date effect on 5 July was caused by similar 
reflectances per band for both sowing dates at the 
end of June and at the beginning of July, resulting 
in similar LAI estimates and no effect of sowing 
date. Also the variability (CV values) of estimated 
LAI was much smaller than variability of measured 
LAI. Results for April are not listed, because bare 
soil predominated on that date. 
Table 1: LAI: means, CVs and critical levels in 
testing for treatment effects, obtained on conse 
cutive sampling dates. Field trial 116 in 1983. 
date 
mean 
CV critical level in 
interaction sowing 
dates 
testing: 
nitrogen 
nutrition 
10 May 
0.54 
0.420 
0.283 
0.003 
0.208 
31 May 
1.91 
0.353 
0.010 
0.030 
0.001 
6 June 
1.96 
0.288 
0.008 
0.356 
0.002 
28 June 
2.18 
0.272 
0.980 
0.018 
0.007 
12 July 
1.09 
0.338 
0.788 
0.001 
0.007 
Table 2: 
LAI estimated by reflectance measurements: 
means, CVs and 
critical levels in testing for treat- 
ment effects, obtained 
on consecutive missions. Fieli 
trial 116 in 1983. 
date 
mean 
CV 
critical 
level in 
testing: 
interaction sowing 
nitrogen 
dates 
nutrition 
6 May 
0.47 
0.056 
0.000 
0.000 
0.000 
30 May 
1.43 
0.120 
0.110 
0.002 
0.001 
7 June 
2.54 
0.086 
0.231 
0.000 
0.002 
21 June 
2.13 
0.090 
0.030 
0.001 
0.000 
5 July 
1.46 
0.109 
0.001 
0.944 
0.000 
12 July 
1.21 
0.089 
0.284 
0.010 
0.000 
22 July 
1.01 
0.080 
0.156 
0.000 
0.000 
8 August 0.16 
0.113 
0.051 
0.022 
0.000 
d infrared 
stage, 1982.