The preprint from @gabdansereau MSc work is out, and it's a fantastic piece of macroecology. It deals with identifying hotspots of unique biodiversity across increasing spatial scales, and it's amazing. https://ecoevorxiv.org/tvmyg 

Gabriel has collected ALL @Team_eBird checklist data with presence/absence of warblers, and measured species richness AND composition uniqueness, based on both raw data and SDMs (hat tips to @wormmaps for embarcadero).

Using SDMs instead of raw data IS important, because it shows how this corrects for over/under sampling. There is a clear East-West effect, which is very visible in the maps of raw data above.

It's a big deal, because we are using indicators like this to understand and manage biodiversity. Gabriel's work is one more piece of evidence that we should inflate the raw data through predictive methods. Field sampling is good, field sampling + computational work is best.

A very cool result (in fact, the answer to one of the questions that prompted this study in the first place!) is that different regions have different relationships between uniqueness and richness (go check out the preprint and have a look at Figure 5, it's EVEN BETTER). But why?

Very long story cut very short: rare species. Sites that have high richness and high uniqueness tend to have more rare species. This is really important, because rare species can more easily be missed when sampling is spatially sparse.

We (a ragtag group including @gabdansereau and @francisbanv) are going to spend some time with the @GEOBON_org Montreal secretariat this summer taking this to the next levels, in order to convert this approach into forecastable indicators. Can't wait.