There is some evidence that female are fairly faithful to a
nest loction, so at least some of the arrivals to islands are
probably on purpose.

In the last three years of watching, hummingbirds have arrived
on Vancouver Island at least a week before reaching the same
latitude on the mainland. It would seem that first arrivals track
more or less straight north over the strait. The islands in
the Sound seem to see birds much later, more in line with the
peak for arrivals west of the Cascades.

All of these speculations refer to population behaviors, not
behaviors of individuals. There is enough randomness in any
biological phenomenon to produce plenty of puzzling outliers.

"Francis G. Smith" wrote:
>
> Dear Tweeters:
> A rufous hummer showed up this last snowy, cold weekend at my feeder on the
> southern tip of San Juan Island. I often speculate about the over-water
> route the hummers take to get to our island. The Straits of Juan de Fuca
> presents approximately 20 miles of open, often windy salt water for the
> little guys to negotiate before landfall. I know that they make significant
> over-water flights as a part of their migration patterns but wonder if more
> localized crossings are made under the same set of biological factors. It
> is easy to imagine a rufous hummingbird meandering from neighborhood to
> neighborhood (on land) before settling down to a territory, but what about
> an island like San Juan? Do the San Juan Island hummers head here
> deliberately or do they end up here by happenstance?
>
> Best regards,
> Francis Smith
> San Juan Island

--
Mike Patterson
Astoria, OR
celata at pacifier.com

http://www.pacifier.com/~mpatters/bird/bird.html