Ron Berger spots phony bolognas for a living. The U.S. Department of Agriculture microbiologist can also detect if a hot dog labeled as all-beef has been spiked with lower cost poultry, or whether a chicken pot pie bound for Saudi Arabia inadvertently contains pork.

Berger's breakthrough research has led over the past year to almost 30 voluntary recalls of processed meats after USDA said they were adulterated and misbranded, and has spurred the agency to implement an extensive random monitoring program of the nation's processing plants.

"Substitution of species has always been a problem," said Donald Houston, administrator of the USDA's Food Safety and Inspection Service. Houston said it is "simple economic fraud" when a company intentionally replaces beef with less-expensive, mechanically deboned chicken. According to the USDA, chicken can cost processors anywhere from 30 to 60 cents less per pound than beef or pork.

Until now, however, the agency has had the capability to rapidly detect undeclared species only in raw meat, using a test developed after a 1981 incident in which horse and kangeroo meat was being passed off as Australian beef. Now, for the first time, with Berger's test, the agency has the ability to patrol cooked meat products as well. And according to Carol Seymour, assistant deputy administrator for regional operations in meat and poultry inspections, the agency didn't anticipate finding as many violations as it did.

Where it all started was in Berger's lab, one among many test tube-decorated facilities at the USDA's agricultural research station in Beltsville.

The microbiologist's test, which took him about three years to develop, is essentially based on antigen-antibody responses. Simply put, this is how it works: Animals, including humans, all produce species-distinct proteins -- or antigens. When an antigen from one live animal enters another animal, the invaded animal produces antibodies to fight it off. These antibodies bind to the antigen like a lock and key; in a laboratory host, a very specific antibody is produced for a chicken antigen, another for pork, and so on.

So when Berger takes an emulsified sample of a "pork" sausage, for instance, and inserts it between two layers of lab-produced antibodies specific to chicken, they should form a linkage if a chicken antigen is present. Berger will know this bonding has occurred if the clear liquid in the eye-drop size "reaction vessels" turns emerald green.

The breakthrough in Berger's test came when he was able to isolate an antigen that remains stable when heated. This was crucial since the antigens in processed meat are obviously cooked.

Lining one compartment in a top freezer in Berger's lab are little baggies of bologna, hot dogs, turkey ham and every other imaginable luncheon meat, tagged and labeled and awaiting analysis. In another freezer are assorted-sized plastic bags filled with unfamiliar brown blobs -- meat samples that the agency has used for its raw meat detection tests. The frozen solid collection includes elk, sheep and some kangeroo from the infamous 1981 Australian incident. Another freezer, separated into frost-covered file drawers, holds little vials for the current test -- chicken, pork and beef antibodies derived from rabbit serum.

Aside from the challenge of isolating a heat-resistant antigen, Berger had to demonstrate to skeptical industry representatives that the test worked.

In fact, representatives from the Western States Meat Association produced their own batch of meat products to challenge Berger's test; he was able to accurately identify each species. The association, which was representing a company found to have a mislabeled product, then began discussing appropriate recall procedures, according to Berger.

Some of the companies that have voluntarily recalled products over the past year are currently under investigation by USDA. In several cases, there has been criminal prosecution with indictments handed down, according to Houston. Only one prosecution has been completed.

Ol' Virginia Packing, Inc., of Los Angeles, was fined $10,000 after pleading no contest to charges of selling and transporting "beef" weiners and bologna that contained poultry, according to USDA. During the course of the investigation, the agency found company records indicating that the firm was buying more poultry than the agency believed necessary for those products said to contain poultry, according to Jack Venus, USDA's deputy director of the evaluation and enforcement division in FSIS's compliance program.

In most instances, USDA believes that the problems have been unintentional. Three companies were issued warning letters but were not requested to recall products.

One of them was Safeway, which apparently sold 80,167 pounds of various house brand beef franks and bologna that contained poultry in California, Idaho and the state of Washington, according to the USDA's incident report. Out of 23 samples taken, USDA found 19 to contain poultry.

Safeway did not concede that any product it manufactured contained poultry, nor did it acknowledge the accuracy of the tests. The company voluntarily withdrew affected products from the market and changed production schedules so that beef products were manufactured first each day.

During several investigations, USDA uncovered an unexpected finding. A number of mislabled products were traced to sloppy cleanup between production runs, permitting remnants of chicken hot dogs, for example, to remain in tubes, tubs or grinders while beef franks were being processed. These instances point to companies not operating under good manufacturing practices, although such species crossovers are not unsanitary, Houston said.

ConAgra, the company that makes Banquet frozen foods, decided to recall beef pot pies last year after USDA test samples showed they contained poultry. A ConAgra spokesman said that traces of poultry gravy left in the pipes got into the beef gravy that was subsequently run in the same pipes. The spokesman said that the company no longer processes meat and poultry products on the same day. According to the USDA's synopsis of the matter, it costs the company twice as much to make poultry gravy as to make beef gravy due to the high cost of poultry broth.

The agency initiated tests after a 21-year-old Menominee Falls, Wis., man who said he is allergic to poultry ate two 8-ounce Banquet Beef Pies. According to the USDA's incident report -- which detailed an interview with the man's mother who was present as her son was eating the pies -- within seconds after consuming them, her son's bottom lip began to swell and itch and within five minutes, both his eyelids and lips were largely swollen and itching. The man was taken to the hospital, where he immediately received an IV injection, the report stated.

ConAgra's spokesman said that the source of the man's allergy was never medically confirmed. Norene Salamone, the man's mother, said in a recent interview that her son has been allergic to poultry since the age of 7 and that she has never seen him have such a strong allergic reaction. The family did not hire a lawyer, Salamone said, but was paid about $2,500 by the company as a settlement. The ConAgra spokesman said the payment was partly for medical expenses.

The USDA's Houston said the agency has gotten two or three leads on adulterated products because of consumer complaints of allergic reactions. According to Dr. Dean Metcalfe, senior investigator of the allergy branch at the National Institute of Allergies and Infectious Diseases, meat allergies are extremely rare in adults, although 1 out of 150 children could be allergic to beef, chicken or pork. (Berger has also sampled baby foods that contain meat, but has not found any violations.)

It is very difficult for the agency to prove intent to adulterate, according to Venus. It has been made more difficult by the fact that Berger's test does not identify the quantity of undeclared species in a product. (While Berger said that "anything we call positive is a significant amount," there are no plans to pursue the quantitative angle, due to its complexity.) This aspect of the test potentially gives companies an out, Venus said. Some companies may claim that only minute amounts exist, due to "rework," or crossover in processing.

On the other hand, the non-quantitative aspect leaves room for intentional violators -- who may be substituting substantial amounts of a species for a less expensive one -- to get clumped together with those who aren't. There's a distinction between the few companies who are cheating and those who have been processing meat "for 15 years a certain way in front of the inspector" and have never been reprimanded for faulty cleaning practices, said Bob Hibbert, vice president and general counsel of the American Meat Institute. Hibbert said that USDA should work with plants to correct processing procedures "instead of dropping the regulatory hammer."

Seymour of USDA said that while there may be a difference between intentional and unintentional violations, "that doesn't mean we can ignore the latter case."

As for Berger, the microbiologist's future plans include developing an antigen for sheep and doublechecking work done by two commercial labs, which are planning to develop the test for industry. (Thus far, only the Beltsville lab and three USDA field offices have the antigens necessary to do the tests.)

After all this, what kind of hot dogs does Berger eat? Hebrew National. "They have to answer to a higher authority than us," he said.