A new study by researchers at the University of Florida Institute of Food and Agricultural Sciences (UF/IFAS) offers agencies a tool to identify pythons with the greatest influence on future population growth and uses machine learning to determine when they are most likely to be detected.

Researchers expect the approach will help agencies make better use of limited resources by allowing python removal efforts to target large reproductive females, the removal of which is likely to have the biggest impact on controlling the population. Considering the age, size, sex and reproductive potential of each snake enables removal efforts to focus on areas with the highest likelihood of detecting and removing this key demographic.

“Our previous work treated all Burmese pythons as contributing equally to the population but removing the biggest female we've ever caught is very different from removing a tiny snake that might not survive next year,” said Alex Romer, a quantitative ecologist with the Croc Docs Wildlife Research Team and lead author on the study at the UF/IFAS Fort Lauderdale Research and Education Center (FLREC).

That distinction is important because Burmese pythons have very different survival and reproductive capabilities at different stages of their lives. Young pythons experience high mortality, while mature pythons are more likely to survive and reproduce.

“Every Burmese python removed is a win for the Everglades, but targeting large reproductive females is likely to reduce the population most effectively,” said co-author Melissa Miller, assistant professor of invasive wildlife ecology at FLREC. “This approach is especially valuable at new invasion sites and along the invasion front, where removal efforts should have the greatest impact.”

The researchers developed the Weighted Removal Index (WRI), a demographic weighting system that assigns an ecological value to removing pythons that are more likely to survive and reproduce.

“The idea is that a big female might be worth 100 arbitrary ecological points, while a little snake might be worth five,” said Romer. “That tells you how many little snakes you would need to catch before you've had the same potential impact as removing a large female.”

That distinction could be particularly important for contracted python removal programs, where the number of snakes captured is an important measure of success.

Rather than treating every capture as equivalent, the researchers say evaluating removals throughout the WRI could eventually help managers structure removal efforts and incentives around the snakes most likely to influence population growth.

Burmese pythons are among the most significant invasive predators in South Florida. Their broad diet includes primarily mammals and birds, including species of conservation concern. Because they are highly cryptic and spend much of their time stationery and often submerged or concealed, their low detectability means that finding individual snakes is challenging requiring substantial time and effort. Mature female Burmese pythons can produce dozens of eggs in a single clutch. That makes prioritization of this demographic critical.

The study also found that just looking at the number of pythons captured during a month can portray a misleading picture of when removal efforts are most valuable.

January and February, for example, did not appear particularly important when researchers looked only at the number of snakes captured. But when the researchers applied demographic weighting, those months became much more significant because the snakes being removed included a greater proportion of mature males and females.

“If you just count how many pythons contractors catch in a month, January and February don't seem that important because you're catching a relatively small number of snakes,” Romer said. “But when you add in the demographic weighting, suddenly those months look a lot more important.”

The finding illustrates why researchers believe management decisions should account for who is being removed, not just how many are removed.

The researchers used data from the South Florida Water Management District's Python Elimination Program to identify conditions linked to more impactful python removals.

They tested three machine-learning models to predict when contractors were most likely to find pythons and when those removals could have the greatest impact.

“Understanding what variables are most relevant when managing invasives is paramount as it allows for a more conscious use of resources,” said Sergio Balaguera-Reina, a co-author and wildlife research scientist at FLREC. “Combining and plugging this knowledge into models that allow us to forecast where the most impactful pythons (reproductive females) could be found can be a game changer that can really make a difference in the battle to combat these invaders.”

One model performed best, identifying conditions such as warmer, more humid weather and certain moon phases that were linked to surveys that resulted in removals of snakes with greater potential to reduce the python population.