If you buy a lawnmower and leave the blade at the factory setting, you may not cut your grass to the length you want.

The same concept applies to the National Research Council's new modeling program. NRC's 1998 "Nutrient Requirements for Swine" comes with a CD-ROM program providing new estimates of nutrient requirements that you can use to set up a feeding program for your pigs.

The catch is that any computer model will spit out good data if you put good data in. But if your input data is questionable, that's what the model will return to you. That's why it's important to know which modeling settings will help you figure how best to feed your pigs.

But the modeling program isn't the only aspect of NRC's new recommendations. The committee offers a 190-page book with lots of useful information on energy, protein, vitamins, minerals, water and other inputs. Word among swine nutritionists is that the book's advice appears reasonably uncontroversial.

The NRC committee increased protein (and amino acid) levels for high lean-gain genotype hogs. The committee also raised sodium chloride (salt) levels for young pigs and increased manganese, vitamin E and folic acid levels for breeding animals.

Those changes haven't spurred much dispute from most swine nutritionists.

The new toy is the modeling program. It's what everyone is talking about and what some people are questioning.

Few suggest NRC shouldn't have adopted the modeling approach. It's the wave of the future. The consensus: This attempt is a good first step.

NRC designed the models to be easy to use. And in most cases, they are. The three models allow you to look at nutrient requirements for grow/finish pigs (boars are included here), lactating sows and gestating sows.

In the grow/finish program, you can track animals from 3 to 130 kilograms (6.6 to 287 pounds). But the lean-growth curves don't kick in until you hit 20 kilograms (44 pounds).

One problem for the average U.S. user: All calculations are in metric measurements. For example, weights are in kilograms and pen space is in square meters. A conversion button lets you temporarily translate to U.S. measurements, but the program automatically reverts back to the metric system.

NRC, like other scientific organizations, must provide all data in the metric system, reports Gary Cromwell, chairman of this NRC swine nutrition panel. He says putting the model in metric form was a natural result of that directive.

He agrees it would be nice if you could switch to the system you like. But the committee couldn't make all the changes every user might want.

Committee members warn that you must be able to provide accurate and detailed data before the models can provide the appropriate information to develop a sound nutrition program.

The swine nutritionists who spoke to PORK'98 voiced the same warning. And, they added another concern: The model's defaults are a potential danger. If you assume the defaults match your animals and make diet decisions based on that, you're asking for trouble.

John Goihl and Dean Koehler, swine nutritionists with Agri-Nutrition Services in Shakopee, Minn., say the defaults may not apply to many of today's hogs. A pig with average lean potential must gain 1.8 pounds a day from 50 to 250 pounds body weight to gain the 300 grams a day of lean tissue the model suggests as a "medium lean-growth rate."

Cromwell says the committee built the trend of hogs gaining more lean per day into the model so it would be applicable in five years as well as today.

Cromwell agrees the lean gain default values shouldn't be taken at face value for all pigs. But, he adds that the committee felt the range of 300 to 350 grams of fat-free lean gain presented in the NRC tables is representative of pigs from average to high lean-gain capacity.

Those values represent fat-free lean gains of 0.66 to 0.77 pounds per day, or 0.69 to 0.81 pounds per day of lean tissue containing 5 percent fat, Cromwell says.

"We recognize there are fatter, slower-gaining pigs," he explains. "In those cases, the user needs to insert lower lean-gain values."

You can use the models most effectively if you use lean-gain figures that represent your pigs. Those are best obtained from your pigs' closeout data.

"The committee correctly warns you to change those values, but when you hit the reset they come back up," Koehler says. "You know some people will use those default settings."

"It boils down to not using the default numbers," Goihl says. While he and Koehler laud the model as a good idea, Goihl warns it's essential to understand the limitations before you use it.

The NRC model offers the same default lean-growth curve for gilts and barrows. Yet barrows and gilts gain at different rates. The committee contends carcass lean-growth rates are probably not very different for barrows and gilts. The single curve lets you calculate feed for a mixed-sex pen.

The model allows up to 99 individual lean-growth curves. You can measure different groups or run what-if scenarios. Or you can use the NRC default curve, which is the best middle-of-the-road option the committee could construct.

Keep in mind, however, NRC is not the last word in swine nutrition, nor does it claim to be. The committee simply offers the model as a tool to help you get started. With it, you can make changes that can profitably fine-tune pig diets.

Viewed in that vein, it can be a very effective device.

"The new NRC is a tremendous improvement over previous editions," says Jim Nelssen, Kansas State University swine nutritionist. "We agree with the modeling concept and think it can be used for benchmarking."

Nelssen sees the model as a valuable tool, especially if you take the next step and use it in conjunction with real-time ultrasound to create farm-specific lean- growth curves. "It's how you apply it that will tell its value. It's a great conservative model."

Nelssen's biggest concern is the true ideal digestibility values for animal requirements and ingredients. He sees many assumptions that may have been worth probing further.

Koehler agrees the NRC model is a positive move. But, as with all models, be aware of the limitations and have confidence in the input data, he says.

The models may be adjusted. NRC agrees future committees may have to tackle the task of upgrading and improving the models by delving into areas such as vitamin and mineral levels.

In the meantime, if you get a copy of the NRC book and computer disk, consider reviewing it with a swine nutritionist. You can avoid potential mistakes and take steps to optimize your feeding programs. In the end, you'll more accurately hit your pigs' nutritional targets.

Remember, though, pigs present moving targets. A good model is only as good as the latest data you load into it.

What's in an Average?

The National Research Council's "Nutrient Requirements for Swine" contains a series of tables on nutrient requirements and feed ingredient compositions.

Great effort went into creating the feed ingredient composition tables, and you can use them knowing that they are the best available – to a point.

The problem is feed composition tables provide the average values for feed sources. And averages are the middle ground in a range. The feed you have in a bin may or may not be average and may or may not relate to values in the table.

It may not be cost-effective for you to test the composition of every load of feed. But you shouldn't assume the nutritional value is equal to that in any given table.

So if you use the NCR table, remember some variation in corn, wheat, oats and other ingredients may alter the feed value. Formulating to slight nutrient overages may provide a safety net your pigs can't afford to do without.

Where Do I Get a Copy?

If you want a copy of the 1998 National Research Council's "Nutrient Requirements of Swine," contact National Academy Press, 2101 Constitution Ave. NW, Lockbox 285, Washington, D.C. 20055.

You can call them at (800) 624-6242 or (202) 334-3313. Or you can reach them via the Internet at www.nap.edu.

For copies of the informational slides that the NRC committee presented via a satellite introduction program, access www.nas.edu on the Internet.

If you get the CD-ROM version of the model and have trouble installing or using the program, help awaits you at the following e-mail address: ckirk@nas.edu.