The evolution of land animals is one of those iconic geological/paleontological images, even percolating into the popular culture as a symbol of progress. What is interesting, of course, is that popular representations of this seminal event are usually of the fish-to-lungfish-to-vague amphibian sort; what people seem to forget is that the first animals to CONQUER LAND (insert echo here) were invertebrates, bravely going were no metazoan had gone before!
In the past, the first known tracks were thought to belong to an Arthropleura like myriapod (pretty much, a centipede). These tracks are from the Joggins, in Nova Scotia, and are early Silurian in age. Horribly, these tracks seem to suggest that the myriapod that made them was enormous; the picture below, lifted from here, shows a model of one:
Now, however, the earliest terrestrial land animal tracks might belong to a Cambrian arthropod that used a discarded shell, a la Hermit Crabs, to prevent drying out on subaerial tidal sand flat. A recent paper in the April issue of Geology (here's the abstract) by Hagadorn and Seilacher (2009) shows trackways with a peculiar, segmented, shingled-to-the-left tailmarkings. The picture below is their Figure 1 (pg. 295):
The interpreted ethology (that is, behavior) of the critter is labelled in this picture below, Hagadorn and Seilacher (2009) Figure 2 (pg. 296):
These traces show a marked similarity to the traces of modern Hermit Crabs, whose borrowed shells also bump along behind them as they wander the beach. The picture below is of a modern Hermit Crab trace from the Bahamas, and was seized from the Data Repository Items for the paper:
Anyway, the interesting part of this paper is WHY the critter might have carried a shell around behind it. These tracks are found in the Cambrian Elk Mound Group of Wisconsin, and are commonly associated with microbial mats, elephant-skin textures, microbal sand-balls, and other sedimentary structures that suggest extremely shallow to subaerially exposed conditions. One of the reasons that Hermit Crabs lug their shell around is that it serves as a reverse-SCUBA suit; in other words, the Crabs can bring a damp, humid shell along with them to keep their gills in proper working order.
Hagadorn and Seilacher (2009) suggest a similar strategy for this Cambrian tracemaker. The tracks themselves show that the shell was far to small to house the entire critter. Rather, they interpret this as an early behavior that allowed these Cambrian arthropods to exploit the subaerially exposed sand-rich, microbial tidal flats along the Paleo-Wisconsin shorelines. If that's the case, then this is one of the very earliest strategies employed by terrestrial (or at least, amphibious) animals, and it's a pretty derived behavior to boot!
Hagadorn, J.W., Seilacher, A., 2009, Hermit arthropods 500 million years ago?: Geology, v. 37, p. 295-298