Jamilly Vitoria Santos da Silva, center, stands next to her mother, Rebeca Arruda as they wash a dog in a bucket in a favela as members of the Brazilian military along with health care workers to talk to residents about the threat of Zika virus. (Matt McClain/The Washington Post)

Brazilian researchers said Thursday they have found Zika in Culex mosquitoes in the northeastern city of Recife in what could prove to be an important discovery. But they cautioned that more study was needed.

Until now, Zika was believed to be carried mainly by the Aedes aegypti mosquito, which is much less numerous, lives in clean water and is more likely to bite during the day. Aedes aegypti thrives in tropical and subtropical climates — it is found in Southern U.S. states such as Florida, but is absent in large parts of the United States. The virus is also carried by the Aedes albopictus mosquito, which lives in more rural environments.

Culex mosquitoes are much more widespread. They breed in dirty water and bite at night. Public health officials have feared that Culex mosquitoes could be involved in Zika transmission, something that would necessitate new strategies to combat the disease — which is blamed for an outbreak of the birth defect microcephaly. Brazil has confirmed around 1,700 cases of the birth defect, which causes babies to be born with unusually small heads and can cause cognitive and learning difficulties.

“It means that we have a second species of vector involved in transmission. And with this vector having totally different habits from Aedes aegypti, we will have to create new strategies to combat Culex as well,” said Constancia Ayres, a researcher from the Oswaldo Cruz Foundation, a leading government-led research institute in Recife, who led the study.

Researchers from the foundation collected 5,000 mosquitoes from houses in Recife where suspected Zika transmission had taken place. They found 86 percent of these mosquitoes were Culex.

A visual journey through Recife

“Culex is a much more abundant species,” Ayres told a news conference in Rio de Janeiro on Thursday, where she presented the results.

The researchers analyzed 456 female Culex mosquitoes, which they divided into 80 “pools” or sample groups of between one and 10 mosquitoes each. They found Zika-infected insects in three of these pools.

Ayres said the research proved that Culex can transmit Zika and that it could have played a role in the rapid spread of the disease in Brazil.

“It can transmit Zika. What we need to know now is which species is the most important — if Culex is the primary vector or the secondary vector. We need to do more research,” she said.

Paulo Gadelha, president of the Oswaldo Cruz Foundation, expressed some caution during Thursday’s news conference. He said that a study at the institution in Rio by researcher Ricardo Lourenço-de-Oliveira had so far failed to find Zika in around 750 Culex mosquitoes.

“It could be that mosquitoes circulating in the Northeast [of Brazil] have more interaction,” he said, stressing that more research was needed.

Gadilha said the discovery did not change the low risk of Zika during Brazil’s Olympics, which open in Rio on Aug. 5 during the country’s cooler winter season. “The risk you have of Zika during this period is very, very remote,” he said.

Stephen Higgs, director of Kansas State University's Biosecurity Research Institute, explains the anatomy of a mosquito. (Kansas State University)

Peter Hotez, dean of the National School of Tropical Medicine at Baylor College of Medicine in Texas said that if researchers have found Zika in Culex mosquitoes only a handful of times, “it’s hard to know what that means; it may not be a finding of great biological significance.” But if they are detecting the virus in Culex mosquitoes in large numbers and on a consistent basis, “That would be a game changer.”

Hotez was skeptical that Culex mosquitoes are going to start spreading Zika in a significant way. “So far, every place we’ve seen Zika has been a place where you have Aedes aegypti mosquitoes,” he said. “There no reason why one would have to speculate that another mosquito vector is involved.”

Brady Dennis in Washington contributed to this report.