I'd say that it could have been two things (disregarding the possibility that the precip naturally fell in that pattern) -- bright banding and evaporation. As snow / frozen precip approaches the -5C level (or even colder than that), the snow becomes sticky, leading to the possibility of very large snowflakes. The outer parts of the snowflakes begin to melt, resulting in liquid-covered, large volume particles (ice has lower density than water, so a particle of a given mass with have a higher volume in the ice stage than in the liquid stage). Since liquid water liquid reflects radar energy much better than ice, these relatively large, liquid-coated particles usually result in anomalously high reflectivities. This results in the typical "bright-band" effect seen near the freezing level.
The second possibility is that the very dry low-level air led to significant evaporation as the precip fell though the lowest few kilometers. If the precip wasn't very deep, then the higher beam heights (and thus locations farther from the radar given a fixed elevation angle) would have relatively low reflectivity. Meanwhile, locations near the radar on a set elevation would be sampled (by the radar) near the surface, where the dry air is evaporating precip and leading to reduced reflectivity. So, low reflecitivites would characterize the locations far from the radar (high beam height) and very near the radar (low beam height).
The shape of the storm-total precip looks like the result of classic bright-banding, but the second possibility may have aided in this as well. I know there was a lot of evaporation occurring from dry low-level air for precip in much of OK last night, though I haven't looked at an Arkansas surface map today.
