With htseq version 0.9.1, by default it counts "stranded" reads, that is, specifically aligned to one of the DNA strains. This would not be correct for metagenomics, you'd need to specify the option "--stranded=no"
Important: The default for strandedness is yes. If your RNA-Seq data has not been made with a strand-specific protocol, this causes half of the reads to be lost. Hence, make sure to set the option --stranded=no unless you have strand-specific data!
With htseq version 0.9.1, by default it counts "stranded" reads, that is, specifically aligned to one of the DNA strains. This would not be correct for metagenomics, you'd need to specify the option "--stranded=no"
It is metioned in htseq manual:
Important: The default for strandedness is yes. If your RNA-Seq data has not been made with a strand-specific protocol, this causes half of the reads to be lost. Hence, make sure to set the option --stranded=no unless you have strand-specific data!