The skippy is one good option. This type of filter has been in use for a long time and obviously works. However, the skippy site skips over the need for a good mechanical filter before the biofilter, making it sound as if you can let fish waste and leaves rot away in the filter indefinitely. It's fairly obvious when you think about it that the less material you have decomposing in your pond, the better. Here are couple articles you should read before you build one of these:
http://www.akca.org/library/filter2.htm
http://www.wernersponds.com/biofiltermedia.htm
Submerged upflow filters like the skippy are a bit old-school, though. Before you make up your mind, you should also read about:
Fluid Beds Here, the media is kept churning so that it cannot clog and the bacteria get plenty of oxygen.
http://www.koiphen.com/forums/showthread.php?t=69028
http://www.fishkeepersforum.org/hardware/fluidbedfilter/FluidBedFilter.htm
http://www.sydneycichlid.com/fluidised-bed-filter.htm
http://koishack.com/forums/index.php?showtopic=12055
Wet/Dry, Shower or Trickle Filters In these filters, the media is above the waterline, and the water rains down over it. This provides maximum oxygen for the nitrifying bacteria, and helps keep O2 levels up in the pond. It's said that in the most violent shower filters, ammonia is gassed off before it can enter the nitrogen cycle. The downside is that, under the right conditions, this type of filter might extract nitrogen from the air and add it to your pond as nitrate. I'm honestly not sure how big a concern this is, the issue is part of a debate in which both sides tend to spin the truth to advance their own preferred filtration method.
http://www.koi-bito.com/forum/main-forum/2500-trickle-tower-bakki-shower-theory.html
http://www.koiphen.com/forums/showthread.php?t=97511
The Anoxic Filter This one is usually a bit on the large side, but it's by far the best looking filter, taking the form of a small pond, usually with plants. Like the skippy filter, the bacteria grow in a static, submerged substrate. But instead of pot or floor scrubbers, the substrate is cat litter with a bit of laterite, a volcanic clay that is a good source of iron. This is kept in mesh planting baskets, which may be planted for appearance and extra performance, but it is primarily a bacterial filter rather than a plant bog, and will work without any plants at all.
The name "anoxic" can be a bit misleading. What it means is that there is little oxygen in the center of the baskets. This means that it can host not only nitrifying bacteria but denitrifying bacteria, and hence instead of nitrate, the end product will be dinitrogen gas, which will leave the pond at the surface or in a waterfall. However, this is
not a fully anaerobic filter. There is always some oxygen, and hence hydrogen sulfide and and other nasty things associated with anaerobic environments will not be produced. Despite the naysayer's claims of anaerobic sludge, the low oxygen environment in the clay presents no danger to fish (unless you bury your fish in it). Theoretically, the "anoxic" filter actually results in higher dissolved oxygen than most aerobic filters because the bacteria use very little oxygen. It eliminates pollutants well enough that, with reasonalbe stocking levels, it will work with as little as two water changes a year, and even then, the main reason to change the water is to replenish certain nutrients rather than to remove nitrate. The anoxic filter is also said to prevent the buildup of ions, which according to the inventor, is the main reason frequent water changes are necessary with other types of filters.
Unfortunately, this is the filter that is on the other side of the debate with the shower, and it's often hard to tell when the pundits on both sides are stretching the truth or simply don't understand each other's systems. The crowd that favors the highly aerobic shower incorrectly associates the anoxic filter with anaerobic sludge in septic tanks. And from the other side come dire warnings about aerobic filters which are equally unsupported, for instance, that they will clog and become ineffective within a month.
But after wading through the hyperbole, disinformation, speculation, misunderstanding, and genuine science for a few months, I'm convinced that the anoxic filter offers significant advantages, and I'll be incorporating it into my upcoming rebuild. I was able find a few people in the forums who had tried it and reported their results with no apparent agenda other than to help other pond keepers, and every one of them reported success. There were no posts at all from people who had built an anoxic filter and had problems. Hence, though the sample size is small, it seems that this filter is one of the most robust and fool-proof options.
http://www.koiphen.com/forums/showthread.php?t=94835
http://www.mankysanke.co.uk/html/anoxic_filtration.html
Sand & Gravel Finally, the sand and gravel filter combines both biological and very fine mechanical filtration. Many people use them as a "polishing filter" to remove fine debris before sending the water to another biofilter or back to the pond. Fines collect in the sand, and once a week or so you blow air through the sand to release the debris and drain it away.
http://www.koiphen.com/forums/showthread.php?t=68132
