The alignment here is between quite close related species.
Just by having a look to the sequences it is quite difficult to determine if one belongs from the opposite DNA strand, but by default it is better to let this option off. One improvement if we expect those homologues to be distant, could be to check the Use translated anchoring option.
The input file you gave to Phylemon has more than 2 sequences. Lagan is only able to compute pairwise alignment, if you want to add more sequences you can use M-Lagan, the same CHAOS algorithm will be used for the computation.