Consider two elephants with 30 Hz (30 cycles/second) tooters. This is below human hearing but not below the African elephants’. One toots from a fine hard pan and the other from a thick, forest humus surface.
Price et al., 1998 tells us that the sound emitted from over the thick humus surface will be 6 dB lower in volume at 10 km than from the elephant on the hard surface. 6 dB is a factor of a 2-fold difference. At 15 Hz even the softest surface an almost perfect acoustic reflector.
If you are a good, deep-voiced elephant, your toots carry as far with the same sound energy over any kind of surface. Tooting for your mate on an uphill slope results in farther, louder sounds. Consider the dumb Dumbo who toots for his mate from hill tops.
The best arrangement is for a 10 Hz bull is to be downhill and the cow in the uphill position (+5 dB gained). The worst configuration is for the Bull to be at the base of the hill and cow to be in the next valley past the hill (-5 dB). An elephant with a falsetto 20 Hz has a +10 dB at hill top and a -10 dB in the next valley. The higher the frequency the more topography matters and preferential courting vs. topography should be considered.
There is not much value in downhill mate calling except in the movies where you find a fine silhouette against the horizon. Now if the wind is upon your back when you howl uphill, the advantages are even greater.
Lamancusa and Doroux, 1993 provide the skinny on topography and gender.