Even though segmentation appears to be an important evolutionary trait in phylas like

Annelida, Arthropoda, and some members of Chordata, these phylas are not grouped together
and diverge after the common protosome development mode. (Knox et al. 2010, p934)
Annelids have been characterised under the lineage Lophotrochozoa, and the Arthropodes are
on a different lineage,Cuticulata. The main classification of Annelids with Molluscs and
several other phyla are due to their eggs developing by spiral cleavage and their trochophore
larva. This defers from the classification of the Arthropods in the Cuticulata which have 2layered cuticles as body covering (Knox et al. 2010, p955) or Ecdysozoans which undergo
moulting (Adouette et al. 1999).
Even though development of segmentation in these two phyla are similar, meaning the
number of repetitions are fixed as compared to strobilisation in tapeworms and these
repetitions are function units, it is apparently not as important a feature as other
characteristics when it comes to classifying these members into their phyla.
This is because DNA and morphological evidence of phylogeny has shown that on the DNA
level, molluscs and annelids have more relatedness on the same branch than does annelids
and arthropods except for an occurrence of segmentation (Knox et al. 2010, p955). Also,
segmentation in annelids and arthropods differ, suggesting that it is due to functional reasons
why segmentation has occurred and as such are hard to consider in the classification
prioritising segmentation only instead of other synapomorphies between the currently
grouped phyla. The bodies of annelids and arthropods require flexibility, mobility and
increased functional efficiency of organs, which is probably why segmentation occurred.
Locomotion in annelids require segmented muscles and nerves working in conjunction with
hydrostatic pressure in the coelom (Knox et al. 2010, p934), and arthropods have a rigid
exoskeleton that requires a form which facilitates movement for their jointed appendages
(Knox et al. 2010, p970).
However, because segmentation does not occur in other phylas in the same group, it is highly
likely that this trait has evolved independently. There is not enough evidence to support that
segmentation occurred in an early common ancestor (Adouette et al. 1999), because the other
phylas do not show evidence of segmentation. It would be likely that were segmentation to be
from a common ancestor, that these traits would also occur in the other phyla.

As the occurence of segmentation has its reason in the evolution of more complex and
diverse body forms and functions (Knox et al. 2010, p955), it is unlikely that the early
primitive common ancestors were developed or complex enough to require segmentation as a
function for specialised tasks. Another consideration would be that tagmatisation seems to be
a more advanced form of segmentation which joins segments together to perform more
specialised functions.
A point of interest would be that some members at an early stage of phyla chordata are also
segmented, even though this phyla are a member of the deuterostomes. As such, it would be
not be feasible to group phyla together just because of one unique trait. If classification by
segmentation were to be used, protostomes and deuterostomes would be grouped together by
this trait instead of developmental characterisations or DNA.
Therefore, the hypothesis that segmentation evolved independently as a result of function and
not synapomorphy can be argued for with the above considerations. It would otherwise be
hard to explain why the members of phyla did not develop according to segmentation and
have this trait occur in all the phylas.

Knox, B. et al. , S 2010, Biology: An Australian Focus, 4th Edn. McGraw-Hill, Australia.
Adouette et al., S 1999, Animal Evolution, TIG Volume 15 No.3.