Mello-Leitão’s (1918) description of Litoporus iguassuensis was based on a single female specimen and a juvenile male from Rio de Janeiro, Brazil. No figure accompanied the original description, and the species did not reappear in the literature until 2000 when I provided a brief redescription and figures of the female genitalia based on the female type specimen (Huber 2000).
When males were finally discovered and collected during the last years, this tiny and at first sight inconspicuous species turned out to be surprisingly exceptional among pholcids in various ways. First, females and juveniles were found to live in camouflaged retreats (below, right photo), a behavior that has not been documented in any other pholcid spider. Second, males were found to live outside the retreats in the domed sheets (below, left photo), and this behavioral difference probably explains the strong sexual size and color dimorphism that is also rare in pholcids. Third, females were found to be color polymorphic, a phenomenon that is equally uncommon in pholcids.
In addition to being thus exceptional, Mello-Leitão’s species continues to defy easy classification, both at the levels of genus and species. Our current conclusion is that Mello-Leitão’s original assignment to the genus Litoporus is not only the best hypothesis but it also generates a number of exciting testable predictions.
The first prediction relates to the fact that females remain unknown in all three closest relatives of the type species L. aerius (i.e. in L. dimona, L. saul, L. secoya; all described in Huber, 2000). We predict that, just as in L. iguassuensis, their females are so different from the males that they were separated from them when the material was initially sorted to species.
The second prediction concerns the phylogenetic relationship between Litoporus and Mesabolivar. If the placement of L. iguassuensis in Litoporus is correct, then this means that Litoporus is nested within a paraphyletic Mesabolivar, eventually requiring splitting or synonymization of Mesabolivar.
At the species level,
we initially assumed to be dealing with up to five species.
differences weakly pointed in that direction (see figure below).
However, molecular markers failed
to provide congruent grouping information, so we decided that with the
evidence it is more parsimonious to argue for a single variable species
for two or more species.