Paleobiodiversity: we need new data