PhilMarkey

Frank "had to have been" is my conclusion. I'll list my reasons, and perhaps you could share your own conclusions?

1. Habit - Musa itinerans var. formosana and Musa itinerans var. kavalanensis both grow in large clumps, and take advantage of all the suitable places for growing in. Musa insularimontana is rare (exceptionally rare) in the few sites on Lanyu. If it had been there for any length of time you'd expect it to be in large clumps, and have occupied some of the many suitable sites there are to grow in. The pattern I see is a recently introduced species with only a few stems for each clump, and growing (or their descendants) only where they were once cultivated - by the Tian-Chi pond and an old yam paddy on Red Head Mountain. True! I have found a few individual plants self-seeding in a couple of ravines, but this only goes to prove a very recent dispersal.
   
2. I have been told by the Yami people the two main sites (Tian-Chi and Red Head) were where the bananas were cultivated for banana cloth.
   
3. The flora of Lanyu is Philippine. The island is separated from Taiwan by a very deep oceanic trench. There are no banana species native to Lanyu, the Batanes of the Philippines, or the Ryukyu islands of Japan. Yes! It could have arrived on the island by birds, but that must be recent.
   
4. The similarity between Musa itinerans var. formosana and Musa insularimontana is great.
   
5. Musa insularimontana is not suited to Lanyu island. The banana can olny grow at elevation (200 - 500 m), and when you consider Musa itinerans var. formosana on the mainland of Taiwan it too only grows at elevation, in the south of Taiwan I've yet to find Musa itinerans growing lower than 500 m.
   
6. The Japanese "Tashiroi" first described the banana, but he didn't regard it as wild endemic - he said it was cultivated.

There are some arguments for though!
The Chinese have only just occupied the island since the Japanese before them. The Yami people were tribal and very territorial (still today - people from one tribal village on the island can't access to forest of another tribal village). But, being sea-farers the Yami almost certainly visited Okinawa, and Taiwan, where they probably learnt to use banana fibre.
The Yami needed the bananas to produce their only clothing (loin-cloth).
But the Yami arrived on Lanyu from the Batane islands of the Philippines, and they don't have bananas there. There, they made clothing (leaf-skirts) from Pheonix palm leaves. There is only one individual Pheonix palm on Lanyu - you'd expect them to have brought these palms with them if they needed them for clothing!

My conclusion is that the Yami learnt to make banana cloth either from the Ryukyu islands of Japan or from Taiwan, or from the Philippines. They then needed a banana. So they went to Taiwan and found one.
The bananas are only found in a semi-cultivated state in yam paddies or by Tian-Chi pond. And only now are they starting to become naturalized in high ravines on Lanyu.
Musa insularimontana is therefore not endemic to Lanyu island. I just need to find one growing on the Taiwan mainland.

So what's your opinion?

__________________
I'm the seed dealer at www.trebrown.com