It's a combination of factors. Insecticides, pollution, loss of habitat, light pollution(especially from daylight LEDs), global warming(throwing off reproduction timing), ubiquitous roadways and vehicles, flat surfaces(beetles get stuck on their back).

Douglas Tallamy, an entomologist at the University of Delaware, makes a convincing case that one factor is the replacement of native plants with (often showier) aliens which local insect populations can't eat. This then reduces the populations of birds which feed on those insects. Site for his books: http://www.bringingnaturehome.net/

I think, compared to twenty years ago, the main factor is the higher color temperature of modern lighting systems. There are numerous studies on cold light attracting insects significantly more than the traditional incandescent light sources.

Whenever I see someone claim a given problem is due to a combination of factors, I get a bit suspicious. First, this seems like a way to say "I don't really know the actual cause." Second, in the real world in situations like these, there is almost always one overwhelming factor that accounts for >60% of total influence. I really do wonder what that is because I like insects!

> in the real world in situations like these, there is almost always one overwhelming factor that accounts for >60% of total influence

That may be true when you're optimising a program, but is very much not true in ecology. The parent comment is generally correct with its list of drivers of biodiversity loss (compare the IPBES global summary: https://ipbes.net/global-assessment).

A key feature of ecology is its extreme complexity and the myriad factors at play in any given situation. Of course, there will be individual situations in which one factor is indeed dominant. Those cases actually often end up as textbook examples precisely because they are so rare, like the snowshoe hare/lynx Lotka-Volterra cycles. In most situations, researchers have to resort to some pretty advanced statistics to try and figure out what the most important factors are. There are almost always several, and getting any one factor to explain 60% of the variance is very rare. (That's why principle component analyses -PCAs- are so popular in ecological research papers.)

I studied ecology for a while, and have published a paper in the field. It works exactly the same way as any complex system. If you can't pinpoint a single factor, you're probably failing somewhere. To blame everything on "it's a complex system" is usually a cop out from lack of knowledge, whether the system is genuinely complex or not.

It's not just kind but application.

It's heartbreaking and happens at least once a year, that someone near conventional agriculture/horticulture is posting a desperate for advice in every forum they know about because some asshole upwind let a cloud of 'cide drift over their horticultural masterclass of a lot, and everything is shriveling.

Recompense can be hard to get. Maybe we need an ACLU meets EFF for biodiversity.

Let's ignore for a moment any agenda to ban monocultures, and look at not just the kind but the density. Pollinator species would be doing better if the monocultures weren't so seemless. We need a mosaic of habitat and to reduce the distances between the ones we have. But it's an uphill battle to maintain them in the face of drift and runoff (really I think we shouldn't be cultivating hilltops, that would be a start, and simply dodges the runoff problem)