Abstract:
Sea spiders, or pycnogonids, are a unique group of exclusively marine invertebrates that are found
worldwide. A scarcity of pycnogonid research is reflected in the unclear position of this group with
regards to the phylum Arthropoda and lack of certainty in their family-level phylogeny. Traditionally,
the pycnogonid phylogeny has relied on the external morphological characters of temperate, shallow
water species. The Antarctic sea spider fauna displays a high degree of endemism and a number of
species have the potential to address several long-standing questions regarding the pycnogonid
evolution. This research uses new sequence data from Antarctic species to provide the most complete
molecular phylogenetic reconstructions of the Pycnogonida, and is the first study to formally test a
number of alternative hypotheses on the interfamilial relationships of this group of organisms.
The BioRoss 2004 pycnogonid collection was classified into 18 different OTUs (5 families & 10
genera) and used, in combination with publicly accessible sequences, to provide samples for this study.
Partial regions of the nuclear 18S and 28S rDNA, mitochondrial 12S and 16S rDNA and protein coding
COI loci were sequenced for each dataset, and the concatenated data tested for incongruence using the
Partition of Homogeneity test. The distance based Neighbour Joining and character based Maximum
Likelihood tree-building algorithms were used to reconstruct the pycnogonid phylogeny for each locus
independently and as a concatenated dataset. A series of alternative evolutionary hypotheses based on
previous studies were examined via the Shimodaira-Hasegawa test. The primary hypothesis examined
was the cephalic appendage reductive trend, which implies that ancestral sea spider taxa possess the
greatest complexity of anterior appendages.
On all the individual locus trees the family Nymphonidae were the earliest diverged lineage of
pycnogonids, although low resolution at the roots of the trees implies that the data are not strong
enough to reject an alternative hypothesis of a basal Ammotheidae group. Pycnogonidae is not the
most recently derived sea spider family and the cephalic appendage loss hypothesis is thus rejected.
None of the phylogenies supported a close relationship between the Colossendeidae and Nymphonidae
families and doubt is raised over the true identification of several GenBank sequences. Polymerous
species do not form a combined, ancestral group but are instead more likely to represent recent
divergences from three separate families. Strong evidence supports the placement of the transient
Austropallene genus (Callipallenidae) at the base of the Nymphonidae family.
This study, and ongoing work, has generated large amounts of new sequence data. This can be used in
future pycnogonid phylogenetic research and/or in investigations on the highly contentious position of
the Pycnogonida with regards to the phylum Arthropoda. A DNA Surveillance website has been
created to assist in the molecular identification of pycnogonids from future benthic bio-discovery
expeditions (http://www.dna-surveillance.auckland.ac.nz).