Running MLST tool on individual contigs in fasta vs whole genome assembly fasta

This is more of a kind of observation.

When MLST tool was used to assign species and ST on assembled contigs as a whole file the assignment was ecoli species and ST 448. 

But when assembled file is separated into multiple contigs, only one contig was assigned with species "senterica", without ST assignment. The rest of the contigs were not assigned by any species and ST. 

It appears, running all the contigs together in a file gives MLST tool more confidence than running the tool on each single contig.