I only detected a few bacteria

Hi I use humann3 but find only 38 species in metaphaln_bugs_list, but the results from kraken and motus have 6199 and 91 each, humann2 have 52 ?
why humann3 less than huamnn2?

# humann3
$ grep "s__" 10WK-D54_humann_temp/10WK-D54_metaphlan_bugs_list.tsv |wc -l
38

# kraken2
[$ grep "s__" 10WK-D54.report |awk -F"       " '$2 !=0 {print $2}'|wc -l
6199

# motus2
grep -v "#" 10WK-D54_taxonomy.txt |awk -F "     " '$2!=0 {print $2}'|wc -l
91

# humann2
cat 10WK-D54_metaphlan_bugs_list.tsv |grep "s__"|wc -l
52

I can think of a few reasons - MetaPhlAn 3 makes more effort to group together very very closely related species as “species groups.” In addition, MetaPhlAn 2 will list “Genus_X_unclassified” and subpecies-level (t__) clades which are not found in MetaPhlAn 3. If you look at which rows are different between the v2 and v3 output it may provide some additional clues.