Amphipods locomote by walking, jumping, and swimming.  Many semiterrestrial beachhopper species burrow.  Navigation of beachhoppers is done by orientation to the sun and moon. 

Amphipods crawl using the large walking legs or periopods.  When crawling out of water and without normally buoyancy, intertidal species are clumsy and tend to fall over.  However, semiterrestrial species, such as sandhoppers Megalorchestia californiana, are able to walk upright using sideward facing legs. Here, the 3rd, 4th, and 5th periopods are modified so that they stick out as props. They are endowed with large extensor muscles for jumping. Photograph courtesy Ingrid Taylar, Seattle, Washington and FREE QUARK.

A talitrid amphipod Megalorchestia californiana checks
out another amphipod in a burrow. The large red-tinted
2nd antennae are used for stability, righting after a
jump, chemotactile perception, and fighting 10X

Jumping in sandhoppers is a combination of balance and flexing of the abdomen.  The animal first balances on its 3rd periopods, then turns its abdomen under and presses its 3rd uropods and telson into the sand.  At this point the 4th and 5th periopods are parallel to, but not touching, the ground.  The abdomen suddenly flexes and straightens, and the amphipod is propelled into the air.  The 1st and 2nd uropods, and 4th and 5th periopods, together act as shock absorbers to cushion the landing.  Small-sized Megalorchestia can jump half a meter or more in distance, but larger animals are less agile.  Hurley 1959 Pac Sci 13: 107; Hurley 1968 Am Zool 8: 327. Original photograph courtesy Ingrid Taylar, Seattle, Washington and FREE QUARK.

A simulation of a jump based on the above description:

On sand beaches around the Scripps Institution of Oceanography, California tiny (3-6mm) gammarid amphipods Americhelidium (Synchelidium) sp. inhabit the high wave-surge zone.  During low-tide periods they are quiescent in temporary burrows in the sand but, when the tide comes in, they actively swim about in the water column.  As a consequence, the amphipods moved up and down the beach with the changing tide. 

If collected and kept in an aquarium tank in the laboratory, the amphipods exhibit the same tidal-rhythm of activity (see graphs on Left). However, the pattern persists in the laboratory for only a short time and mostly attenuates within 72h.  Interestingly, if animals are placed in sand-filled trays anchored subtidally to the substratum or attached subtidally on a pier, the rhythm is either not present or has a cycle different from the natural tide cycle.  The same is true for specimens in sand-filled trays suspended 1m below the water surface, and for ones contained in soft, compressible) plastic bottles either anchored to the sea bottom or floating at the surface.  The results of these and other experiments cause the author to reject the hypothesis that tidal changes in hydrostatic pressure represent the proximal stimulus for the activity rhythm. 

Other experiments likewise show that factors such as temperature change and photoperiod are not the stimuli for inducing the rhythm.  However, if a sand covering is not present, the rhythm is disrupted, suggesting that the sand's presence or, rather, the mechanical stimulation of the particles on the amphipods’ bodies, may be an important factor in entraining a tidally based rhythm of activity. Enright 1963 Zeitschrift für vergleichende Physiologie 46: 276.

NOTE  in earlier papers the author shows that Americhelidium responds to even small changes in hydrostatic pressure (0.01atm) in a laboratory flask by rapid scrambling and darting about.  The accompanying graph shows an example of responses to a pressure change of 37mb, equivalent to about 0.04atm.  Note that the activity response falls off within a few seconds of application of the continuous pressure stimulus.  Enright 1961 Science 133: 758; Enright 1962 Comp Biochem Physiol 7: 131.