Well, we're all going to die. A new study that combed through 73 recent scientific probes of insect populations now estimates that roughly 40 percent of all insect species are expected to go extinct within the next few decades.
Key causes of the decline included "habitat loss and conversion to intensive agriculture and urbanization," pollution, particularly from pesticides and fertilizers, as well as biological factors, such as "pathogens and introduced species" and climate change.
If it is not immediately obvious, losing nearly half of all insect species worldwide would be extremely, catastrophically bad. It threatens pollination of the vast majority of world plants, the fertility of soils, and directly threatens birds, fish, and other insect-eaters. Given that human beings have grown accustomed to eating, well, food, all of this could present significant global problems; it suggests entire ecosystems will be wiped out within our lifetimes.
The study's authors pin most of the blame on habitat destruction and widespread chemical pollution, and note that insect populations are not just plummeting on developed land, but within designated reserves meant to provide buffers for native species. This suggests we will not be able to stave off a population collapse merely by designating additional buffer acreage, and that the problem appears to be considerably more systemic.
So to put it in shorthand form, we're all going to die, and we may not even need for Florida to become a shallow sea before it happens. Attempting to reinvent new systems of agriculture while whole ecosystems go bottom-up—all the while feeding an ever-increasing population—will take some fancy doing.
But that's not to say we will be devoid of insects entirely: species that have evolved to take advantage of human activities, like roaches and other urban and agricultural pests, "will probably thrive" as their natural enemies "disappear."
The alternative to this is to curb pollution and climate change immediately. Immediately as in immediately, not immediately as in two decades from now. It would have been less expensive and less disruptive to do it at any prior point of the last few decades, but it will still be less expensive than trying to patch up new systems for acquiring food in a new landscape of planetwide die-offs and plagues. Might want to get on that. Might not want to wait for the nation to not be led by a dementia-addled buffoon before starting.