Science1 publisher3 min readPublished
University of Florida index ranks large female Burmese pythons as the removals that count most
University of Florida researchers built a Burmese python removal index in which one big female might be worth 100 points and a small snake five. The authors say managers could eventually build removal incentives around those weights, in programs that now judge contractors largely by snakes caught.
The Scientist · Science desk

What happened
- The study, published in Ecological Applications, pairs the index with a machine-learning model that predicts when the high-value pythons are most likely to be detected.
- Judged by count, January and February looked unimportant, but weighting raised their value because catches in those months included more mature males and females.
- Co-author Melissa Miller said the approach is especially valuable at new invasion sites and along the invasion front.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- decision Agencies that pay contractors per snake would have to decide whether a winter month of few but mostly adult catches deserves more credit than a busy month of juveniles.
- cost Search hours are the expensive input because pythons hide so well, and a forecast of when big females are detectable lets agencies spend those hours where one capture equals many small ones.
- constraint Until weighted targeting is tested against count-based removal in the field, a program that switches is acting on a population projection, and the 100-to-5 split Romer cited is an illustration.
Contracted python removal programs lean heavily on one figure, the number of snakes captured [8]. Alex Romer, the quantitative ecologist who led the study at UF/IFAS's Fort Lauderdale Research and Education Center [17], gave a simple illustration of what that figure hides. "The idea is that a big female might be worth 100 arbitrary ecological points, while a little snake might be worth five," Romer said. "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." [7] On his numbers the answer is 20 [1]. He called the points arbitrary, so treat the 20 as an illustration of how the index works.
The weighting rests on ordinary demography. Young Burmese pythons die at high rates, while mature ones are more likely to survive and reproduce [4]. A mature female can produce dozens of eggs in a single clutch [10]. The Weighted Removal Index assigns more value to removing the snakes most likely to survive and breed [6], using each animal's age, size, sex and reproductive potential [14]. "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," Romer said [3].
Weighting also changes the calendar. Counted by number, January and February catches looked minor. Weighted, they rose in importance because a greater share of the snakes removed were mature males and females [11]. "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." [12] The thing this doesn't tell you is why the winter catch skews toward adults. A larger share of adults in the catch is a finding about who was removed. On its own it does not say whether adults are easier to find in winter or whether crews search differently then.
Finding the snakes is the hard part. Burmese pythons are cryptic, spend much of their time stationary and often submerged or concealed, and locating one takes substantial time and effort [9]. The machine-learning component targets that problem. It estimates when the high-value snakes are most likely to be detected [1], so effort can go where the chance of removing large adults is highest [14]. The description of the study does not report how accurate those forecasts are, and it does not include a field comparison of weighted targeting against count-driven removal.
So the population benefit is still a projection. The researchers expect the approach to help agencies spend limited resources better [15], and say the index could eventually shape how managers structure removal work and incentives [8]. I think the weighting is a sound correction to a count-only metric. Whether it shrinks a population faster than catching every snake in sight has to be settled in the field.
"Every Burmese python removed is a win for the Everglades, but targeting large reproductive females is likely to reduce the population most effectively," said Melissa Miller, an assistant professor of invasive wildlife ecology at the Fort Lauderdale center and a co-author [5][18]. "This approach is especially valuable at new invasion sites and along the invasion front, where removal efforts should have the greatest impact." [16]
What to watch
- A field trial in which crews target pythons by Weighted Removal Index score while others remove by count, with population outcomes tracked across breeding seasons.
- Published accuracy figures for the machine-learning detection forecast, tested on seasons it was not trained on.
- Whether any contracted python removal program writes weighted scoring into its payments or performance targets.