"Oh no! I've got "high" nitrates! What to do?"... First and foremost, understanding what nitrates are, their sources, alleviation and relative placement in a/the grand scheme of "water quality" is absolutely necessary for aquarists. In reality, nitrates by themselves are not "that" deleterious. What accumulating nitrogenous wastes in the form of nitrate compounds indicates may be symptomatic however, and signal a need to react (slowly) in changing ones management/maintenance of captive marine systems.

    Here's what I deem an overall picture of nitrates should be to most saltwater hobbyists; their sources, importance, and means of  keeping in check.

Sources: The Greatest Pet-Fish Story Ever Told!

    Oh for the good old days when real pet-fish-ionados could recite the bacterial nitrification mantra by rout... "Ammonia is converted to nitrite by Nitrosomonas bacteria species and nitrite to nitrate by Nitrobacter species..." Then came friend/college chum Tim Hovanec... and the doubt of which species/groups are actually involved. Well some groups of microbes do this "forward reaction" of conversion: 

Ammonia to Nitrite to Nitrate

NH3 (or NH4OH)<=> NO2 <=> NO3 

with the equal signs representing bacterial involvement. There are a bunch of salient and maybe interesting points that could be made re this reaction statement. The "forward" (reading from left to right) steps are called nitrification... Nitrification is an oxidative process (the reactants on the left "losing electrons" in the process), and results in lowered pH, loss of alkalinity, and is manifest of such energetics that aerobic (oxygen using/needing) organisms product its effects. Hence all the practical things you know or will know re Nitrification in captive systems. For instance, by "driving" nitrification by way of enhanced aerobic/biological filtration (Wet-Dry filters, undergravel filters, fluidized bed filters...) you can produce an abundance of accumulated nitrates (drive the reaction to the right) given sufficient nitrogenous foods, a paucity of organisms that "use up" (export) the nitrates, water changes to dilute them, chemical filtrants to absorb them, efficient skimming to extract them... and/or some version of the reverse reaction series termed denitrification: This occurs in two half-reactions

Nitrate to Nitrite to Nitrogen

    NO3 + 2 electrons + H+ <=>  NO2 + H2O 

NO2 + 3 electrons + 4 H+ 1/2 N2 (nitrogen gas) + 2 H2O

or, taken together:

NO3 + 5 electrons + 6 H+ <=> 1/2 N2 + 3 H2O

Where some anaerobic microbes (lack of oxygen) convert nitrates into transitory nitrite and ultimately ammonia molecules. Note, with the use/incorporation of H+/protons pH is elevated. Note further that this is where the reduction of carbon-based materials comes into play in "feeding" purposeful denitrators, providing the electrons and hydrogen ions, balancing these RedOx equations.

Importance:

    Most fish groups are remarkably tolerant of  "practical" levels/concentrations of nitrates (30-40 ppm. let's say). A few tens of parts per million in their water, changes in same over days time is not life-threatening or stressful compared with fluctuations of temperature, varying light/dark cycles, measurable ammonia or nitrite, or hobbyist hands coming into their space for instance. For fish-only or FOWLR (fish only with live rock) systems, nitrates by themselves are rarely a real worry.

    Some invertebrate groups are notably touchy to too much or sudden increases of nitrate content in their water. Such material may significantly interfere with respiration and other vital metabolic functions. Fifteen to twenty parts per million are upper limits for most non-vertebrate marine livestock, with many corals being mal-affected at lower levels.

Methods of Control:

   Ammonia (ionized or not) is the most common decomposition product of proteins, and excretory product of fishes and invertebrates... and toxic to them at sufficient concentration. It must be dealt with. The easiest, simplest, surest, most fail-safe method is through biological conversion. Hence, the great hub-bub re "cycling" (establishment of beneficial microbial populations) in new captive systems and their protection from population and metabolic stops (e.g. many "medications", abrupt environmental changes...). 

    Other techniques, gear, approaches can be employed for reducing nitrate, its expedient removal and even prevention.

Prevention:

    Foods/Feeding, Livestock Loads. Nitrates are resultant from "too much proteinaceous food input, not enough output"... One easy approach to their limitation is to simply stock and feed your system lightly... particularly with foods of high protein content. Catabolism of protein (made up of amino acids, in turn the source of ammonia) from foods, wastes, decomposition is the root source of the precursors to nitrate: GIGO, garbage in, garbage out.

    Some supplements and salt-mixes have nitrates of appreciable concentration in them. Read labels and if in doubt, test these by diluting into water, and using a test kit.

    Detritus in the substrate can be a reservoir, manufacturer of nitrate. Vacuuming a portion of it along with your water changes can reduce this source. Similarly, mechanical filtration media that traps proteinaceous foods and wastes needs to be cleaned and/or replaced on a regular basis to limit these materials release, cycling in the system. 

    Wet-dry media in same-named filters are a huge source of nitrates. The aerobic bacteria cultured on their vast surface area readily produce nitrates. By and large, aquarists and their aquatic charges are best served by removing such media, converting the same space to live rock, macro-algae culture, either on an alternating light cycle with their main system, or leaving the lights on continuously.    

Dilution:

    Here comes the usual pitch for frequent partial water changes. Obviously switching out ten-twenty percent of your water with that of zero nitrates reduces the composition percentage of nitrates (and other metabolites) by the same amount. The dilution-solution can be a significant source of nitrate reduction in other ways as well. There are ancillary skimming improvements, bacterial et al. microbial involvement and outright chemical interactions with nitrates with such changes.

Bio-mediation/Bacterial Denitrification:

    The so-called "reverse reaction" of nitrification: denitrification is a largely anaerobic set of reactions that serve to convert nitrate ultimately into nitrogen and oxygen.

    Live plant use: Both from micro-organisms growing on the surface of live plants as well as the plants uptake of ammonia, other nitrogenous compounds.

    NNR: Natural nitrate reduction systems include such propositions as Plenums (Jaubert et al.), DSBs (Deep Sand Beds), and various contraptions that are anaerobic to hypoxic containers (boxes, coils, trays) for culturing and feeding denitrifying microbes. All makes, models of the latter have proven fickle. Tricky to slowly drip system water into, to provide "feeder stock" (typically sugars, alcohols, even sulfur).  Plenums and DSBs can be great expedients to reducing nitrate accumulation, but need be large and are often difficult to manage/manipulate when employed in the main/display system. Aquarists are encouraged to build, use these tools in separate sump/refugia where their utilization is more facile and alteration not a matter of tear-down and disruption of the principal tank/s they service. 

    ASD: Autotrophic sulfur denitrification, a type of anaerobic denitrator utilizing elemental Sulfur as a chemical feed source for reducing nitrates, has been advanced and used in places. The reaction series (4 NO3 + 3 S = 2 N2 + 3 SO4) involved is acidic, can be best tied-in with melting down a source of carbonate, does result in excess sulfates, but these don't appear to be problematical (natural seawater contains about 2,700 ppm of Sulfate.  

Chemical Filtration:

    Entails many possible areas for discussion. Of most practical importance is the provision by aquarists of sufficient alkalinity (a notion of balance twixt carbon dioxide/carbonic acid in solution, carbonates and bicarbonates (mainly of sodium) to "resist" the almost assured pressure of driven nitrification (too much livestock, feeding, insufficient microbial denitrification. Whether you boost/maintain your alkalinity with simple baking soda (sodium bicarbonate), calcareous gravels of high solubility, other chemicals or commercial preparations, do keep an eye on your alkalinity (about 2-5 dKH is about right, no higher than 12, and you can multiply these units times 2.8 to get the approximate values in milliequivalents per liter) to assure ready metabolism of your "biological helpers" in the way of nitrifying and denitrifying bacteria.

    There are "nitrate chemical filter materials" sold in the hobby/trade. They don't work in many sets of circumstances, and never directly. There are no practical chemical means of removing nitrate from biological systems. 

Biological Uptake/Export:

    In the hey-days of reef-keeping in the mid- to late-eighties of last century, vigorous algal growth was touted as a/the sure-fire means of reducing nitrate accumulation. In actuality, a good deal of nitrate is taken up by a few processes by many groups of algae, microbes and photosynthetic life forms.

    Both micro- and macro-algae absorb nitrate during photosynthesis. With regular harvesting, removal of biomass, considerable nitrate can be exported from a system by leaving it a little "dirty" with algae growth. 

Other Methods:

    These are numerous and for the most part, unrealistic or otherwise impractical for aquarium hobbyist use. Some have involved electrical current (yes, with saltwater aquariums), ion-exchange resins for removal of nitrate from tapwater (where there is generally little to start with). Folks would/will be better off utilizing an inexpensive reverse osmosis filtration rig for their personal and pet-fish use if they have appreciable nitrate in their source water. Most R.O. units remove 95% plus. 

About Testing Methods/Units of Measure:

    A brief mention that there are two common ways of expressing nitrate concentration by various assay methods. Nitrates as ppm, and nitrogen as nitrate in ppm. Due to the latter's consideration of the three oxygen atoms atomic weight per molecule, measuring nitrate ion by itself results in 4.4 times as much. Do check your test kit though almost all are nitrate ion types on the market nowadays...

Cloze:

    Any amount of investigation into "dreaded nitrates" in hobby literature, the Net, conversations with others... will reveal an enormous amount of differing opinions on the importance/significance of their presence/concentration. Are nitrates important as sources of poisoning of marine livestock? By and large for most species of life, no. Can nitrate measure be useful as an overall indicator of system health, trends in water quality changes, wake-up calls for altering, enhancing methods of overall system improvement? Sure.

    Without implementing some of the above methods of control, nitrates will/do accumulate in freshwater aquariums. Your role as the "creator", and on-going manager of said systems is to devise and impose balance between inputs of nitrate sources and the practical continuous elimination of their accrual.

    Ideally you want your nitrates to be as low as possible. For invertebrate-containing marine systems, a few ppm (up to five as a rule of thumb) is likely acceptable, and a few tens of ppm for fish-only systems. Just having "some" nitrate concentration present in the grand scheme of things that contribute to livestock health is not a real menace. The real value in nitrate measure and maintenance is as a "window" on overall cleanliness, viability of your water.

Bibliography:

Anderson, Frank. 1992. NTS. New tank syndrome may be on its way to oblivion. FAMA 4/92

Booth, George. Nitrate reduction. The next step in water quality. AFM 7/96.

Frakes, Thomas A. 1993. Nitrate menace? SeaScope v.10, Winter 93.

Goemans, Bob. 1997. Plenum. FAMA 10/97

Klostermann, A.F. 1992. Biological denitrification. FAMA 4/92.

Patel, Shilpin. 2000. ASD: Autotrophic sulfur denitrification. FAMA 12/00.