This thread proposes that bioballs are not only less effective than a properly built and maintained sand bed, but can be harmful, keeping unwanted Nitrates at elevated levels. Bioballs can even be dangerous. I am not saying they cannot or should not be used, but because of some LFS that continue to recommend their use, this needs to be settled before anyone else gets hurt.
You may have heard of the Nitrogen cycle. That's one of the processes of decomposing waste. It involves the conversion of uneaten fish food and fish and coral poop into harmless substances. The 3 step process converts it from Ammonia -> Nitrite -> Nitrate and finally -> Nitrogen gas and Water. Ammonia changes to Nitrite very quickly due to the easy action of bacteria in the water and on surfaces of rock and glass in an oxygen rich environment. That's the first step of the cycle. Nitrite then waits to be changed/eaten by another species of bacteria.
Bioballs are used in a technique called a wet-dry filter or trickle filter. These filters have been called Nitrate factories because they are so effecient at changing Nitrite to Nitrate. The trickle filter creates in the tankwater an over-abundance of Nitrate. The bioball accomplishes only the second step of the three part process. Nitrate then needs to be converted into the final non toxic products of Nitrogen gas and water.
The final step of the process can only be accomplished in a low oxygen environment. Today, the best way we know to provide that is in the depths of a deep sand bed (DSB) or Plenum. Down in the bottom sand or in the Plenum away from the flow of oxygenated water, Nitrate can be converted to simple Nitrogen gas and water. The depth of sandbed needed to effectively reach that final step depends upon the speed that water is flowing over the sand and the amount of surface area provided by the different particle sizes of sand commonly used today.
Water flow, sand bed depth and partcle size are discussed below but first, there is an interesting thing about sand beds. In fact, the upper layers of the sand bed accomplish the same step in the process as a trickle filter. Hobbyists who dismantled their trickle filter and made certain the sand bed was of sufficient depth for its particle size have reported a decrease in Nitrates to an acceptable level. An advantage of dismantling the trickle filter also leaves room for an RDP Refugium. The RDP Refugium provides a large host of benefits discussed in this thread:
http://www.utahreefs.com/forum/forum_posts.asp?TID=36396
Faster water flow will require a deeper sand bed to keep some sand away from the oxygenated water. Slower water flow allows a shallower sand bed, but since water flow directly influences the health of a reef aquarium, it's usually best to have faster flow and a deeper sand bed. Some very successful reef aquariums have from 4-8 inches of sand clearly visible at the bottom of the tank. Fortunately, there is a solution to the unsightly sand against too much of the viewable front and sides of an aquarium.
Smaller particle size offers tighter spaces between the sand particles, thus restricting water flow. Smaller particle size also offers more surface area to grow a larger population of bacteria in the same volume area as larger particle sand. Smaller particle size typically allows a shallower sand bed than larger particle sand and still performs as well.
If a larger particle sand is very porous, it's effectiveness increases, but in time all sand beds get filled with detritus which essentially clogs the spaces between particles and the many pores of each sand particle. This is why occasional stirring of the sand bed and even periodic partial vacuuming is helpful. Stirring portions of a sand bed can be done frequently and can be helpful for a reef aquarium because it releases detritus and live larvae into the water column. Coral and small invertebrates love to eat that stuff.

Though stirring the sand with a stick or even by hand is preferred, the historically common practice of sand vacuuming is useful perhaps once a year. The extent of vacuuming is much, much less than commonly understood and is best done only at the top 1 inch of the sand bed or restricted to small portions of the deeper sand bed.
If a system has a trickle filter full of bioballs, in the long run it is best to plan the removal of the bioballs. It's important to remove them slowly. There is detritus loosely lodged in the spaces between the plastic needles of the bioballs. To be safe, remove no more than 20% of them in a single day and take care not to jostle the remainder. Moving the bioballs dislodges the detritus. If too much of it lets go all at once, it can create a very toxic situation. This has been known to kill lots of animals.
What about sulfur? Not as commonly known or recognized but just as important as the process of decomposing Nitrogen compounds is the process of binding and decomposing Sulfur compounds. Without any thought or need for action by the hobbyist, the sand bed accomplishes this task alongside the Nitrogen cycle. Some frightening discussions about rotten egg gas (SO2, Sulfur Dioxide) sometimes circulate around the hobby. These scare stories are just that. Rarely are they actually experienced by hobbyists. I myself have experienced the problem in a freshwater tank when I completely removed a stagnant sand bed.
I can confidently assure hobbyists that a properly maintained sand bed with plenty of worms and bugs, with good flow over it and with frequent stirring to feed detritus to the tank, will never experience any hint of a problem.
