Yes, as Bumber said, the "tropicality" is dependent on latitude primarily if there are any poles (no pole => no latitudes). If the location is in the tropical region the biome will become tropical (for those biomes that do have tropical/temperate versions: a glacier will still be a glacier at the equator if the temperature is low enough). A band around the equator is the "maybe" zone where other parameters can influence whether the biome becomes tropical or temperate (temperature and rainfall, if I remember correctly), with different biomes having different criteria and different ranges.
Also, as Bumber mentioned as well, the actual temperature DF generates is modified from the PSV ones based on latitude (if present, of course) and elevation.
Thus, if you want a tropical forest in a 17 tile high PSV world with at least one pole you ought to set forest resulting parameters for the equatorial band to guarantee it. I don't remember if the bands immediately to the sides of the equator are in the definitely tropical or the maybe zone, but in the tropical zone you can get your tropical forest to have any non freezing temperature (below that it would be taiga and then glacier below that).
Also note that a single pole has half as many degrees of latitude per tile than a dual pole one, as the world spans only 90 degrees rather than 180, roughly doubling the size of each band (roughly, because the underlying math is tied to latitudes, not tiles, so the boundaries may not align exactly on tile boundaries).