The evaluation of leaf diseases of coconuts, Cocos nucifera L., known as la
rge verrucose and small verrucose, caused by Sphaerodothis acrocomiae and P
hyllachora torrendiella, respectively, come upon the problem of sampling me
thod. Since there is not a consensual method for measuring coconut verrucos
e incidence, the present work was done with the objective of comparing two
sampling methods of common use: method A: sampling six leaflets/plant in a
single leaf, method B: sampling six leaflets/plant, from different leaves.
The study was developed starting from three samples of 300 leaflets (ten pl
ants x five leaves x six leaflets) collected in three coconut genotypes, in
which the stromata number of large verrucose and of small verrucose were c
ounted. Throughout these data the variance was calculated from the leaf in
the plant and the leaflet in the leaf and plant, necessary for the estimate
d sample average in the two compared methods, as well in the other formulat
ed alternatives. For the two diseases, in all the genotypes, the estimate o
f variance of the sample mean calculated by the A method was higher than th
e one calculated by the B method, showing that the last method is more; app
ropriate than the first one. Other sample sizes were also studied and compa
red by the B method, resulting that sample of two or three leaflets collect
ed in six leaves per plant decreased sample variance on respectively 20% or
30%, probably being of more accurate applicability.