16S ribosomal RNA does not encode a protein, so attempting to align it in any type of codon mode may cause problems.  The 16S ribosomal RNA is a structural RNA that forms many critical stems and loops to create the structure that will be part of a scaffold holding many ribosomal proteins in position so that the ribosome can function.   Within a small range of diversity such as one family of crabs the alignment may be so easy that any method will work well, but if there is a lot of diversity then a method which attempts to consider codons is very counterproductive for any noncoding sequences such as 16S rRNA.

Good day Sir Brian Foley. Thank you for your time.

I am working with only one family of crabs. So using either Clustal W or Clustal W (Codon) would work? 

So it means that even if I am aligning forward and reverse sequences, I may use Clustal W and not Clustal W (codon) given the fact that I am working with 16S rRNA (which does not code for a protein). Please correct me if I am wrong. 

Thank you very much.

I am not sure if CLASTAL W or other alignment programs will automatically reverse complement the reverse sequences for you.  You may need to do that step yourself before asking an alignment program to align the sequences.

Sorry for not clearing that up Sir. I reverse complement the reverse sequences before using an alignment program. When I align in MEGA, I get the option of choosing between the two: Clustal W or Clustal W (Codons). 

Although as I align sequences, I've realized that I may use Clustal W (not with the codons option) because I am working with 16S rRNA, which does not encode for a protein. Is this right? 

Yes, CLUSTAL, MUSCLE, MAFFT and other programs should all do a fine job of aligning 16S rRNA (or rDNA, the rRNA gene) especially within one family or genus of of organisms.   The more distantly related the sequences the harder the job.  The other thing to keep in mind when studying mitochondrial genes such as 16S rRNA, is that for most animals the mitochondria are passed primarily by eggs and rarely by sperm, but there are some exceptions.  So mitochondrial evolution can give somewhat different pictures of populations and evolution than nuclear genes do.  It can be somewhat similar to studying differences in X and Y chromosome evolution vs the autosomes in mammals. 

And yes, you should not use the (codon) option in CLUSTAL W nor use other methods which attempt to do a "codon alignment" when the DNA being aligned is not protein-coding DNA.

Got it. Thank you very much for the enlightenment, Sir Foley. I really appreciate your time for sharing your knowledge on mitochondrial and nuclear gene regions.

I prefer using the Clustal X, because it has a feature that normally does not appear in Clustal w which is the iteration. This feature significantly improves overall alignment. Also the Clustal X has a friendly graphical interface. Another option is the omega Clustal that can be downloaded or used online (http://www.ebi.ac.uk/Tools/msa/clustalo/).

If you prefer to download any version may get the same in:

http://www.clustal.org/

The Clustal w is the older version of Clustal and has no graphical interface.

Thank you for your suggestions Sir Rodrigo Peraira. I appreciate it.

I am currently using Clustal W because it is already built-in in MEGA 6. I shall download Clustal X and try it out. It is my first time to hear about Clustal Omega, I may try it out soon too.

Thank you for you help!