Many of the fossils on which evolution is based on are reconstructed from
the flimsiest evidence, even from a single tooth or bone. The conclusions
based on such fossils are mere speculation.
Evolution is not based on fragmentary fossils. The theory would still
be extremely robust with no fossils at all, based on evidence from
modern life. Furthermore, there are more than enough substantially
complete skeletons to support evolution. The whale
transitional sequence, for example, is based on several excellent
skeletons.
A single bone, even in isolation, can give a surprising amount of
information. A tooth, for example, can show generally what kind of
food an animal ate and give an idea of its size. These conclusions, in
turn, tell how the animal fits into the ecology.
Bones are never considered in isolation; rather, they are compared with
other bones from more complete skeletons. If you have a bone that
looks like an Iguanodon femur but smaller, to give a simple example,
the reconstruction would look a lot like a smaller Iguanodon. A
complete reconstruction, however, is possible only if you can match the
single bone to an animal for which there is a complete skeleton already.
The ability to deduce much about a fossil from a single tooth or bone
was made famous by anatomist and paleontologist Georges Cuvier. In
1804, for example, he confidently announced that a French fossil was an
opossum (then unknown from France) on the basis of only its teeth
(Zimmer 1998, 135-137). Cuvier was a creationist.
References:
Zimmer, Carl, 1998. At the Water's Edge. New York: Touchstone.