I have a plant RNA seq data of control and treated conditions obtained from 4 different days. This SRA dataset obtained from NCBI doesn't have replicates but have mentioned in their article that they have merged the biological replicates drawn during sequencing process in the sample preparation section. When I try to run DESeq, I am getting error because of the absence of replicate sample read counts. How can obtain the differential expressed genes between control and treated of each day separately?

Can anyone please help me how to solve this.

DESeq2 requires replicates to perform statistical inference. It cannot do a 1 vs 1 comparison.

With just one sample of each kind, there is no use for fancy statistics. You can use the library normalization techniques described in DESeq, and then just compare the values directly. You won't be able to generate a p-value.