Was one rice plant analyzed for each genotype? - Then it's impossible to get biological replicates. If this is the case, I don't agree with M. Cooper. You can describe what you measured and highlight plants (genotypes) with the best response to drought stress. Don't calculate p-values. Take this as an hypothesis-generating approach (not a testing approach). Testing differences on technical replicates will only tell you is you have enough data to see that one plant is different from another. There is no information if this difference is attributable to the genotypes. It can be attributable to the position of the plant in the field (all kinds of differences between environmental factors) or to differences in epigenome of the plants, the plants could have acquired different mutations during germination that affect the response, and it can also be attributed to differences in the processing of the biomaterial of the plants (up to the different measurements). Having many different plants in a sample does not tell you what the reason is that a sample of genotype X differs from a sample of genotype Y, but since the genotype is the only factor systematically different between the samples, one can make statistcal inferences about the genotype effect. If you have only technical replicates, there are really many factors perfectly confounded with the genotype, and this is the problem.

Or were there several plants of each genotype analyzed? - Then you can measure biol. replicates, even if these plants are all grown in the same experiment, in the same year and in the same soil. Then you could statistically infer differences between the genotypes, with the limitation that this still refers to the particular year and the particular soil (it might not translate to other years or soils).

Good information Jochen Wilhelm . I analyzed several plants (3 plants) of each genotype in the same year in the same soil. According to your statement, this could be used as biological replicate while it is a technical replicate. Well, if I proceed with it considering biological replicate, wouldn't it statistically wrong and affect my analysis of variance?

If you have individual data from 3 individual plants (per genotype) then you have n=3. These are biological replicates. The interpretation has to acknowledge that the systematic differences you might not be generalize to other years or other soils. But at least for the given year and soil, you can (possibly) show differences that can abe attributed to the genotype.

You have biological replicates n=3) as JW has already suggested.

You have biological replicates n=3) as JW has already suggested. However, be careful with the interpretation of the results: might not be generalize to other years or other soils. Moreover, results can be strongly affectet by the row of cultivation!