Here’s a link to an excellent summary in the Washington Post of a very disturbing PNAS article. I’ve just a few clarifications to add.

Very brief summary first: previous studies of insect population decline were mostly done in Europe at temperate latitudes. Causes (as well as those in studies of North American temperate locations) were generally attributed to insecticides or habitat destruction.

The current paper looks at the rainforest in Puerto Rico. It has been protected since the days of the Spanish Kings and declared a National Park by Teddy Roosevelt, so no habitat destruction, no or negligible insecticides.

Temperate insects are adapted to the wide range of temperatures found in mid to high latitudes. Tropical insects, not so much. As they evolved in regions that experienced only a narrow band of temperatures, these insects fare poorly outside of that range.

The researchers compared sites sampled in the 1970s with present day readings. The declines were catastrophic, up to sixty-fold. There was parallel decline in insect-eating animals but not in those occupying other niches (like seed-eating birds).

Loss of ecosystem services from insect biodiversity is a serious problem in and of itself. The article cites an American economist who placed that value for wild insects in the US at $58 billion per year, which presupposes that nature can be valued in dollars, but that’s a philosophical rant for another day—suffice it to say the value is big.

However, some areas of the world are warming faster than others. The Puerto Rican rain forest experienced a two degree Celsius increase in average temperature between the 1970s and now. Ding, ding, ding. Two degrees, Paris Accords, utter catastrophe beckoning worldwide with predicted collapse of every world country—ahem.

Here’s the part I looked into: I had no idea what the hell the Post was calling a new method for measuring the role of heat, developed by a Fordham University economist. Well, I can be forgiven, because new means a couple months old. Here’s a link to the paper.

This level of statistics is beyond my current capacity to expatiate upon with any sort of mastery, but I got the gist of it. Let’s say you have two variables, X and Y, that cannot be manipulated, only observed in data. You model the data with X being a predictor of Y and separately as Y being a predictor of X. If X is a much better predictor of Y, the evidence is consistent with X causing Y.

Yeah, that’s a whole lot less definitive than the Post made it out to be. But the authors of the PNAS paper have a compelling argument: what else could explain this drastic species loss? Nothing they could think of.

Nothing I can think of. Occam’s Razor says this is real-world data, rather than just modeling and projection, showing how disastrous two degrees Celsius in warming will be for the world. Sure, it could be changes in precipitation associated with global warming, rather than heat per se. (They factored all cyclical, seasonal, or one-off weather events out of the data, and it didn’t change the conclusions.) Or it could be some other change correlated with global warming.

But it’s still powerful evidence that it’s long past time for politicians to be acting. And that some regions of the world will be fucked to a greater extent than others. With the US primarily responsible—we historically took more of the world’s carbon budget with our fossil fuel burning than than any other country (not to let China, India, or Europe off the hook)—I think we can expect hundreds of millions of climate refugees to be arriving on our shores.

Build a wall—right. It will more likely take perpetual massacres to stem the human tide. And historically, those guns tend to also be turned inward. We all laughed at Trump’s admiration of North Korea, but that place may be a best-case scenario for our future.

Or we can elect politicians who will prioritize this issue and mitigate the harm to the extent that it is possible. Americans reading this, are you registered to vote?