Melissa Kaplan's
Herp Care Collection
Last updated January 1, 2014

Salmonella: Then and Now

©1996 Melissa Kaplan. The California Herper 1(8), California Herpetoculture Industry Advisory Council

Then...
In 1970-1971, the Centers for Disease Control (CDC) conducted three retrospective surveys of salmonellosis to determine the frequency of salmonellosis in households owning turtles. These studies were conducted in Connecticut and Utah, and in Atlanta Georgia. There was also an ongoing surveillance program being conducted in Santa Clara County, California and in the Seattle, Washington area.

In Connecticut, 25 of the 117 families (36 of the 150 patients, or 24 percent of all patients) had pet turtles at the time of the illness. In the control group, only one of the 50 patients (2 percent) had a turtle at onset, for an 22 percent average of all cases being related to turtles. Seventy-two percent of the patients were under five years old. The association of the turtles to the patients ranged from just a few days to several months after acquisition, with more than one-half of all turtle-related infections occurring within a month of acquisition. Hospital stays averaged 11 days for the eight patients hospitalized for treatment for turtle-related salmonellosis.

Salmonellae were isolated from the water housing the turtles in 11 of the 12 pet, variety and department stores from which the turtles were bought by the patients, friends or family members. In addition, Salmonellae were found in all turtle tanks in the 18 pet stores selling turtles in the cities of Bridgeport, Bristol and Meriden. Of the 133 samples analyzed, just over half tested positive for Salmonellae, representing 26 serotypes.

In Utah, 10 of the 64 patients (15 percent) had turtles at onset. The average age of the patients was 6.7 years. Average length of hospital stay was 2.8 days for those with turtle-associated salmonellosis.

In Atlanta, Georgia, 10.9 percent (six of 55 persons with salmonellosis) had turtles at the time of onset.

In Santa Clara County, California, 48 (18 percent) of the 267 confirmed cases of salmonellosis were in families that owned turtles at time of onset.

In Seattle, in the three year period between the years of 1965-1967, three to five years before the one-year studies noted above, there were 619 cases of salmonellosis, an average of 206 cases a year. Seventy-two of the cases (11.6 percent) were associated with families owning turtles.

In 1968, the Washington State Board of Health put into effect a regulation requiring that only those turtles certified as Salmonella-free could be sold within the state. In the subsequent years, there was a marked decrease in the number of turtle-related salmonellosis cases overall, especially in children between the ages of 0-9 years. During this time, the incidence of turtle-related salmonellosis in the same age group increased throughout the rest of the United States.

The results of these studies were combined and averaged with similar studies in New Jersey and Minneapolis-St. Paul, Minnesota, yielding a turtle-associated salmonellosis rate of 18.2 percent. Extrapolating this across the United States, it was determined that 14 percent (280,000) of the 2,000,000 cases of salmonellosis each year were associated with turtles.

Studies conducted during that time found that the breeding farms were the source of contamination. Salmonellae were isolated from the turtle feed (often offal and refuse from rendering plants), in the nesting soil, pond water, ovaries of adult turtles, and the turtle eggs.

One hundred turtles who were refrigerated immediately upon hatching (without first being fed or watered) by the breeder were taken for study. Divided into ten groups, they were placed in water in pans that were sterilized weekly, with water changes done every other day. The water was tested weekly for nine months. Despite these precautions, more than 1100 Salmonella per 100 ml of water was found when testing many of the samples.

The researchers determined that contaminated food and surroundings, the overcrowding and fecal contamination in the ponds, transovarian transmission of the Salmonellae and their ability to penetrate into turtle eggs, and their ability to survive in the intestines of turtles without causing signs of illness were all factors in the contamination farmed turtles. Attempts to eradicate Salmonella with antibiotics and other chemicals failed.

...and now.
It is currently estimated that somewhere between 1-3 percent of all households in the United States own reptiles of some sort. The current cases of reptile-associated salmonellosis are approximately five percent, according to a recent interview with Dr. Jono Mermin of the CDC. While this rate of infection seems minimal, it is still not too far away from the 1970-1971 figures, especially given the apparent decrease in the number of households with reptiles. According to Dr. Mermin, while the number of reptile-related salmonellosis cases is less than during the early 1970s, they have increased since the mid- to late-1980s.

The CDC decided in early September to begin a new study of salmonellosis, including gathering data on those cases associated with reptile ownership and contact. As in the early 1970s, several communities throughout the United States will be selected for study. According to Dr. Mermin, the San Francisco Bay Area is presently slated to be one of them.

One of the issues that may be looked at and discussed, along with the incidence of Salmonellae in the pet trade and patients infected with Salmonellae, is the impact or importance of the severity of the infection. Is a case of a healthy adult who gets diarrhea for a day the same as an adult or child who is hospitalized for several days? Since eradicating Salmonellae from turtles is not possible, what other avenues may be taken besides outright bans?

One variable not looked at during the studies over twenty years ago was to establish whether there were differing rates of salmonellosis in households knowledgeable about turtles, proper turtle care and who employed basic hygiene practices from those households who had no such knowledge and did not practice reasonable hygiene (hand-washing after handling, restricting direct and indirect turtle contact with the very young, proper cleaning and disinfection procedures, etc.). Looking at this variable may prove an assumption held by many experienced herpers that, with proper education and reasonable precautionary practices, the risks can be greatly, if not completely, negated.

Note: Salmonella in substrate will penetrate through the shells of turtle eggs within one hour of contact with a contaminated surface. Salmonella can be passed from the mother to the baby before birth or laying.

The information in the above tables are taken from Johnson-Delaney, Cathy A. 1996. Reptile zoonoses and threats to public health. In: Mader, Douglas. 1996. Reptile Medicine and Surgery. WB Saunders, Philadelphia PA. pp. 20-33.