Bill Kunin

@DLeonhardt You focus on how presidential party affects economic growth, but causation goes both ways. How does economic performance affect who gets elected?

More on the partisan gap in economic performance here: nytimes.com/2021/02/02/opi… (fin)

A thread: The economy has performed much better under Democratic presidents than Republicans over the past century. nytimes.com/2021/02/02/opi… (1/n)

@ThijsFijen Pantraps DO seem to provide discrimination between landscapes. The results from 96 AgriLand sites were highly correlated between years, both for summed bee abundance (r=. 631) and spp richness (r=.735), despite different weather & surveyors. That's vital for long-term monitoring.

@ThijsFijen ...As predicted by Bev Rathcke (1983) for co-flowering plants. Transects have far stronger flower density effects, always positive (but probably non- linear)... Which oddly don't seem to bother people. Sigh.

@ThijsFijen The original argument was made by Cane et al (2000) J. Kansas Ent Soc 73: 225-231, as possible explanation for their low catch. But subsequently quoted by others as established fact. It may be a weak hump-shaped response: facilitation at low flower dens, competition at very high?

@ThijsFijen Oops. That was "AgriLand" not African. I HATE autocorrect!

@ThijsFijen The strongest argument in Portman et AL is that Pantries catch mostly Halictids. Perhaps that's true in their sites... But not my experience here in the UK. My African samples (5269 bees) are 15.6% Andrenid, 60.5% Apid, 0.8% Colletid, 15.4% Halictid, and 7.7% Megachilid.

@ThijsFijen The problem with assessing taxonomic bias is that we don't know the "true" values. We know different methods show different proportions, but ALL are probably wrong. The important thing about passive traps is that they're repeatable -- which is what's needed for monitoring change.

@ThijsFijen Pan traps won't provide a complete species inventory, but the do a good job of providing a robust and repeatable measure of local pollinators abundance, composition and diversity (at least of the locally dominant spp), and are fairly independent of surveyor skill...

@ThijsFijen There's a lot in the lit saying that flowers compete with pantraps, but in my experience the effects are weak: sometimes positive, sometimes negative. Transect counts are very strongly (positively) linked to floral abundance. Combining the two may be best?

@ThijsFijen Pantraps clearly have different biases than transects, catching more solitary bees and far fewer honeybees. Transect counts vary a lot depending on surveyor skill -- the solitary bees in particular get missed by non expert surveyors as they're small, often inconspicuous, flighty