I would like to know the standard number of leaves can be selected for measuring the plant leaf morphological traits. I want to know how this measurement can be precisely done.
You can measure the leaf area of second fully expanded leaf of four individual leaves and pool them for leaf area for a particular plant. Elaheh Daghighi
It would have to depend on the variability of the leaf morphology. Are you looking at wild-type plants/ecological studies or mono-cultural cropping plants? There may be quite a difference between sampling a "uniform" crop and studying the variation inherent in a "natural" population.
Perhaps a statistician could answer your question with a specific protocol. You could also look up procedures in Biostatistical Analysis by Zar (currently 6th edn), that consider the effect size, sample size and so on. This might be useful if the leaves are from a uniform grass species, rather than a highly variable multi-form tree species.
My thoughts suggest collecting N leaves (where N is a much larger number than you think you need), measuring the morphological traits of interest, then calculating the appropriate means/range/variance/standard deviations etc for leaves 1 to N. If the measures of variations do not approach an asymptote, then collect another N leaves and continue. When they do approach an asymptote, then pick a number that is close enough and use that.