I am not a specialist in Polychaeta, but usually you do need these parts in order to positively identify to the species level and count individuals from that species. This is necessary since it is usual to find fragmented parts of the body (segments, for example) in samples, but naturally, it is difficult to be precise unless you have the anterior parts of an individual. An exception, of course, happens when you do have entire individuals of the same species, in which case you can compare the fragmented individuals with the complete ones. Ideally you should contact a Polychaeta specialist in your geographic region in order to help you with this. Warm regards from Rio!

The pygidium and prostomium really are the easiest paths to identifying maldanid polychaetes.  If you know the other maldanids in the region you might be able to align intermediate segments with those to find genus/species by comparing presence or absence of a collar (around 4th segment), plus number, location and type of neurosetae (neurochaetae).  That being said, the pygidium and prostomium really do hold the best definitive characters.

The text is OCR-treated and the plates are of good quality in this pdf-copy of Arwidssons book (300 dpi, grayscale).

No, You need both 

http://www.scamit.org/tools/toolbox-new/ANNELIDA/Class%20Polychaeta/Subclass%20Scolecida/Family%20Maldanidae/-OTHER%20USEFUL%20TOOLS/Maldanidae%20by%20PRGarwood.pdf

Thank You for the answers! I am working on material from Gulf of Guinea, but I think, that Yours advices will be very useful. I will try to identify as many of them as I can. However only few Maldanids are complete. Thank you again.

It is very difficult to identify maldanids without one of the extremes, in most cases you need to know the total number of segments to identify your material at the generic level. However, if you have a good knowledge of the fauna in your sampling area, you can use anal plaque papillation, the shape of the cephalic plaque and the shape of the uncines to compare. Regards

For northern Europe, there is an NMBAQC Scheme key that allows identification from front ends only. For other areas, it would be more difficult.

I agree that generally disconnected ends are to be considered only identifiable at the family level, except for where there is local knowledge and limited and known faunas, as for the Bogueinae for Florida and the Caribbean that are easily fragmented but found inside characteristic rocky cases, and for which diagnostic characters are seen on the first few setigers. Cheers,  Doug

Yes, it is possible.  Ideally you will have seen at least one whole specimen so that you can match heads and tails.  However, it is still possible with all fragmented.  Use methyl green staining to show up the glands on the segments.  The staining patterns along the body can be very helpful. With different fragment lengths you may well be able to build up complete knowledge of different species.  Hopefully you can then recognise what species are then present.