Hi,
is there any modification made in numDescendants in species/?name response?
For example, http://api.gbif.org/v1/species/?name=Lembus returened one record with 2 descendants (I remember it as http://api.gbif.org/v1/species/102321492 ) and two records with 16 descendatns arournt 2015-07-29 04:00 UTC (ish) but now only one record of GBIF Backborn Taxonomy with 16 descendants and others have no descendants.
http://api.gbif.org/v1/species/?name=Lembus%20infusionum was shered lower name in there three records. http://api.gbif.org/v1/species/102321493 is one of Lembus infusionum which has parentKey 102321492. The parent key is the record had two descandants.
The modified implementation requires to try /species/{int}/children even if numDescendants is zero, because there may be descendants. I prefer to the previous implementation to navigate classifications.
James