https://en.wikipedia.org/wiki/The_Screwfly_Solution
Same author wrote a short story about a scientist who developed a tailored virus to save the earth from ecological destruction:
https://www.lightspeedmagazine.com/fiction/last-flight-docto...
"The tape recorder they put by his bed functioned right on through, but if anybody had been around to replay it they would have found little but babbling. “Gaea Gloriatrix,” he crooned, “Gaea girl, queen . . .” At times, he was grandiose and tormented. “Our life, your death!” he yelled. “Our death would have been your death, too, no need for that, no need.”
At other times, he was accusing. “What did you do about the dinosaurs?” he demanded. “Did they annoy you? How did you fix them? Cold. Queen, you’re too cold! You came close to it this time, my girl,” he raved. And then he wept and caressed the bedclothes and was maudlin."
The problems with DDT were that (1) it could last a long time before breaking down, (2) it spread much more widely in the environment than was intended, and (3) it was not narrowly targeted to only affect pests.
When you are making a pesticide that is not narrowly targeted [1], you really want it to break down fast and to not spread much beyond where you specifically apply it.
Just make a pesticide that like DDT spreads far and wide and persists, but unlike DDT make it so one of the things it kills is nitrogen fixing bacteria. That would wipe out almost all of the base of terrestrial animal food chain.
[1] You make a narrowly targeted pesticide by basing it on the hormones that control the target's life cycle. You find some specific behavior of the target, such as its "mate and die" behavior, that is triggered by a specific hormone, and make your pesticide trigger that. Then all you need to do is apply the pesticide when the insect isn't yet sexually mature or when the weather is too cold for eggs to survive, and the insects mate and die, with no offspring produced.