For decades, generations of dogs have been bred, raised, and trained as service animals for people with disabilities at a center in England: Labrador retrievers, golden retrievers, curly coat retrievers, border collies, and German shepherds. Scientists at the University of Nottingham realized that they had an ideal opportunity to study dog fertility — five types of purebreds, uniform conditions, one location, systematic record-keeping. So in 1988, they started annually testing the sires’ sperm.

In a study published Tuesday in Scientific Reports, they found declining sperm quality and other effects that they believe could be related to environmental causes. Over 26 years, motility, the progressive forward movement of sperm, dropped 30 percent in all five breeds. Although it has not reached a critical point — the dogs are still successfully impregnating — further decline in motility could eventually harm their ability to reproduce.

“The dogs who share our homes are exposed to similar contaminants as we are,’’ said Richard G. Lea, an associate professor of reproductive biology at the School of Veterinary Medicine and Science at the University of Nottingham. “So the dog is a sentinel for human exposure,’’ added Lea, the study’s principal investigator.

From 1988 to 2014, researchers studied between 42 and 97 stud dogs annually. Between 1994 and 2014, they also noticed that the mortality rate of the female puppies, although small, showed a threefold increase. And the incidence of undescended testicles in male puppies, also small, had a tenfold increase, to 1 percent from 0.1.

When the researchers tested testicular tissue for chemical content (in dogs retired from breeding or neutered for other reasons), they found concentrations of chemicals that had been common in electrical transformers and paint, and others still used in plastics. In additional analyses done in the last three years, researchers found concentrations of the same chemicals in the dogs’ semen. The chemicals include polychlorinated bisphenol (PCB), which, though banned, has a long half-life, and diethylhexyl phthalate (DEHP).

The researchers also found traces of the same chemicals in the food that handlers gave the dogs. The brands were not named in the study (nor was the training center), but Lea said the food, both wet and dry, is sold worldwide — in other words, quite possibly what you serve your own dear beast.

The scientists cannot determine how the chemicals were introduced into the food supply; these are not additives. But Lea and his colleagues speculate that they could be in the packaging as well as in water that came into contact with any ingredients.

They cannot even conclusively say that the dog food is a direct or even the only source of the chemicals found in the testes. But Lea said that it is likely a major one.

The study, the authors said, may support other research findings about environmental exposure and human semen. For at least 70 years, studies have noted a decline in human sperm quality, and a gradual increase in rates of testicular cancer and genital tract abnormalities, like undescended testes. But there has been a debate in the scientific community about whether this trend is directly associated with industrial chemicals discharged in the environment. Critics have noted that lab conditions and standards for the many studies have varied widely. Establishing a cause-effect link over time, they say, is not reliable.

Peter J. Hansen, a professor of reproductive biology at the University of Florida, describes himself generally as a skeptic of many studies linking chemical exposure to declining sperm quality. Much research on the effect of environmental hazards on humans is typically done by administering doses of hazards to research animals in much greater concentrations than is typically found in water supplies, he said.

But the Nottingham study, he noted, detected the chemicals in the dogs’ tissue and also in the dogs’ food. And researchers did so over decades, tracking a concurrent decline in reproductive markers.

“I think it was very rigorous,’’ he said. “It’s much more clear from their data that there was a decline over time, which agrees with the human data but doesn’t suffer from the same research problems.’’