The taxonomy of the New Zealand wrens has been a subject of considerable debate since their discovery, although they have long been known to be an unusual family. In the 1880s,
Forbes assigned the New Zealand wrens to the suboscines related to the
cotingas and the
pittas (and gave the family the name Xenicidae). Later, they were thought to be closer to the
ovenbirds and
antbirds. Sibley's 1970 study comparing egg-white
proteins moved them to the oscines, but later studies, including the 1982 DNA-DNA hybridisation study, suggested the family was a sister taxon to the suboscines and the oscines. This theory has proven most robust since then and the New Zealand wrens might be the survivors of a lineage of passerines that was isolated when New Zealand broke away from
Gondwana 82–85
million years ago (Mya), though a pre-
Paleogene origin of passerines is highly disputed and tends to be rejected in more recent studies. As no evidence indicates passerines were flightless when they arrived on New Zealand (that
apomorphy is extremely rare and unevenly distributed in Passeriformes), they are not required by present theories to have been distinct in the
Mesozoic. As unequivocal Passeriformes are known from
Australia some 55 million years ago, the acanthisittids' ancestors likely arrived in the
Late Paleocene from Australia or the then-
temperate Antarctic coasts.
Plate tectonics indicate that the shortest distance between New Zealand and those two continents was roughly at that time. New Zealand's minimum distance from Australia is a bit more today – some – whereas it is now at least from Antarctica. The extant species are closely related and thought to be descendants of birds that survived a genetic bottleneck caused by the
marine transgression during the
Oligocene, when most of New Zealand was under water. The earliest known fossil is
Kuiornis indicator from the
Early Miocene St Bathans fauna.
Kuiornis is thought to be more closely related to
Acanthisitta than to other Acanthisittids. The relationships between genera and species were formerly poorly understood. The extant genus
Acanthisitta has one species, the
rifleman, and the other surviving genus,
Xenicus, includes the New Zealand rock wren and the recently extinct bushwren. Some authorities have retained Lyall's wren in
Xenicus as well, but it is often afforded its own monotypic genus,
Traversia. The stout-legged wren (genus
Pachyplichas) was originally split into two species, but more recent research disputes this. The final genus was
Dendroscansor, which had one species, the long-billed wren. A mitochondrial DNA study in 2016 resolved much of the phylogeny, though the placement of
Dendroscansor was unresolved due to lack of DNA testing due to the rarity of its remains. It was found that
Xenicus was
paraphyletic with respect to
Pachyplichas, and that the stout legged wrens must have evolved from a gracile-legged ancestor, and the paper suggested placing the
Pachyplichas species within
Xenicus. It was also found that the split between Lyall's wren and other acanthisittids probably took place during the Oligocene, over 30 million years ago, so acanthisittids must have survived the
Oligocene drowning. A
cladogram is given below: • †
Lyall's wren,
Traversia lyalli • Genus
Acanthisitta •
Rifleman,
Acanthisitta chloris • Genus
Xenicus • †
Bushwren,
Xenicus longipes •
New Zealand rock wren,
Xenicus gilviventris • †
North Island stout-legged wren,
Xenicus jagmi (possibly conspecific with
Xenicus yaldwyni) • †
South Island stout-legged wren,
Xenicus yaldwyni • Genus †
Dendroscansor • †
Long-billed wren (New Zealand),
Dendroscansor decurvirostris Fossil genus and species †
Kuiornis indicator from the
Early Miocene St Bathans fauna ==Description==