That is a point I should have made. The source shape file indicates an occurrence with a circular polygon having a radius of about 100m. In some instances a record represents a group of trees so the polygon representing the area is much larger.
For ease of plotting (read lazyiness) I calculated the centeroid of each polygon and fixed a placemark at that location. For clarity I perhaps should have plotted the full polygon.
Give that many of the first observations predate wide availability of highly portable GPS equipment and certainly predate the May 1st 2000 removal of ‘selective availability' I wouldn't consider the location information to be highly accurate.