Fossil Orchid shows Orchids are ancient.

[Tindomul]

Hi,
The following is a direct quote from the article. Apparently it is all the morphological features of the pollinium and pollen combined to narrow this specimen down to a specific orchid species.

Horizon and locality. Specimen was excavated in the year 2000 from
a mine located east of Santiago, Cordillera Septentrional, Dominican
Republic. Lignite and sandy clay beds, Early to Middle Miocene (15–20 Myr old; ref. 3).

Diagnosis. The species is separated from other members of
Goodyerinae by the bent anther, large angular massulae (,100 per
pollinarium), and tightly packed pollen units (20320 mm). The
amber piece (2031435mm) contains a single inclusion of
Meliorchis caribea. Two complete pollinia (each ,1,0003500 mm),
belonging to a single pollinarium, are firmly attached to the mesoscutellum
of a worker bee, Proplebeia dominicana7 (Fig. 1a). The
tapering pollinia consist of .100 loosely packed angular massulae
(,2003100 mm, Fig. 1b), each of which encapsulates several tetrads;
obovoid pollen units are tightly packed.
These pollinarium features are found only in the Orchidoideae8. A
survey of herbarium specimens of all Neotropical genera within this pollinarium of Meliorchis is attached to the mesoscutellum (dorsal
surface of the thorax) of worker bees of P. dominicana. This indicates
that the flower of M. caribea was gullet-shaped, and, rather than the
bee probing the lip of the flower with its tongue as in modern
Goodyerinae (Fig. 2a), the anterior part of the bee would have had
to enter the flower completely (Fig. 2b).