Where is most biodiverse?

 iNaturalist can be used to compare biodiversity in different locations.  

To do this, it is important to allow for different search effort in different areas: for example, a search of the middle of the Amazon might not return many species, not because there aren't many, but because people haven't observed them.

For the U.S., we will assume that search effort is at least somewhat comparable.  We will also look at total observations and divide them by total species to get some idea of the effort required to document a new species in that area.

We decided to investigate my home town in Washington State, Alexandra's home town in Ohio, one of our favorite places: Maui, and our current home in Arizona.

Port Orchard, WA

A 100 km search radius includes most of Puget Sound, the Olympic mountains, and Seattle.

Dublin, OH

A 100 km search radius includes most of central Ohio creeks, woodlands, agricultural lands, and all of Columbus OH.

Prescott, AZ

A 100 km search radius includes canyons and mountains between Flagstaff and Phoenix.

Maui, HI

A 100 km search radius includes all of Maui as well as ocean channels and nearby islands.

Summary Table

According to this analysis, Washington is most biodiverse, followed by Ohio, then Arizona, and in last place is the island paradise of Maui.  

Island biogeography teaches that islands, although unique for their level of endemic plants and animals, should be deficient in overall biodiversity if there hasn't been enough time for biodiverse taxa to colonize or evolve on the island.  This seems to be the case with Maui, which is especially depauperate in insects, given its tropical location.

Arizona, despite its diverse environments, is significantly lacking in Fungal diversity, which could be due to the arid climate.

The American Midwest, with its perfect seasonal growing climate, is remarkably biodiverse, and would probably surpass Washington if this search had included a more varied part of the Midwest with more mountains and/or rivers and lakes.

Washington tops the list as most biodiverse, probably because of the presence of diverse environments, with everything from alpine mountain tops to ocean kelp forests, and everything in between.  This is despite the lower number of insect species compared to a lower latitude place like Ohio.

A caveat to this entire analysis is revealed by the Total Observations in each area and the resulting Observations per Species.  While I had assumed that search effort would be equivalent across the US, it is evident that many more people are using iNaturalist around major cities like Seattle and Columbus.  Arizona is more sparsely populated and has many less observations, and Maui, despite the huge numbers of tourists, is even less populated.  This means that is only takes 20 or 27 new observations (on average) to observe a new species in Maui and Arizona, respectively, while it takes 47 or 72 new observations in Ohio and Washington, respectively, to observe a new species.  So even though Washington has the most observed biodiversity in iNaturalist, it may be easier to find new, previously-unobserved biodiversity in Maui or Arizona.

Mapping Species Habitat with Appropriate-Sized Buffers

 Previously, I wrote that this Story Map shows small polygons of habitat as buffers around representative observations.  However, the actual locations are not accurate because the underlying observation data has been randomized to protect populations of rare species. 

The first map ("Preliminary Conservation Zones" and "Potential Dispersal Zones" for the American, Rusty-patched, Suckley's, and Western bumble bees) shows the correct kind of critical habitat (buffered observations) USFWS has designated for rusty patch and would likely designate for other proposed species, but the locations are incorrect.  For example, the mapped locations of Rusty patch on that map do not line up to the USFWS GIS for rusty patch critical habitat. 

 

Some of the other species may be are incorrect as well, depending on whether the data source (GBIF) considers the species endangered and so randomized the locations within a 0.2 degree lat/long box.  That seems to be the case for the Western Bumble bee, but not the American bumble bee. 

 

The map shows a mix of accurate and inaccurate, specific habitat points. This is confusing and potentially misleading, if the intent is to facilitate conservation planning.  For example, when I zoom to an area of interest, I might think there is no mapped habitat there. But if there is some nearby, I can't tell from if that habitat is or isn’t within my area of interest.

 

The easiest fix would be to increase the size of the buffers so that they include the entire randomized area (0.2 degree, lat/long) that each point comes from.  A note could say that critical habitat would likely be designated in a subset of those larger polygons based on the buffer size USFWS decides.

Rusty Patched Bumble Bee Critical Habitat

Rusty patched bumble bee range map.

USFWS listed the rusty patched bumble bee (Bombus affinis) as Endangered in 2017 due to a marked decrease in the range and size of populations across the Eastern U.S.  

As one of the first insect species to be listed under the Endangered Species Act, it offers an interesting case study for the way USFWS may approach other insects proposed for listing, including the Monarch butterfly, and numerous other bumble bee species.

From Xerces Society Listing Petition, 2016.

The listing petition states that "the rusty patched bumble bee probably needs floral resources to be located in relative close proximity to its nest sites, as studies of other bumble bee species indicate that they routinely forage within less than one kilometer from their nests ... although in some cases nearly two kilometers ... [It] is likely dependent upon woodland spring ephemeral flowers, since this bumble bee emerges early in the year and is associated with woodland habitats.....Rusty patched bumble bee queens are one of the earliest species to emerge, with observations as early as March and April."

Interestingly, rather than designate critical habitat based on the habitat needs of the species, USFWS chose to designate "High Potential" zones (e.g. critical habitat) as 1 mile buffers, and "Low Potential" zones as 4 mile buffers, around known (since 2006) sightings of the rusty patched bumble bee:

USFWS Map showing "High Potential" and "Low Potential" zones.  

Detail showing example 2x2 mile rusty patched bumble bee "High Priority" habitat in DeKalb, IL from USFWS map.  The buffered area seems to be based on a sighting at Prairie Park, and includes residential and industrial developments.  The only habitat in the area is within Prairie Park.  

USFWS has issued the guidance on whether consultation is required.  For vegetation management activities within the High Potential zones, the guidance provides the following test questions:

• Is there habitat for nesting, foraging, and/or overwintering for the rusty patched bumble bee in the action area or will the proposed action restore habitat for the species in the action area? 

• Will the action cause effects to vegetation in rusty patched bumble bee habitat in the High Potential Zone during the nesting period? Effects could occur as a result of mowing, cutting, grazing, prescribed fire, tree removal, spot-application of herbicide, tree clearing, and/or other activities. 

Based on this case example, it seems likely that USFWS will take a similar approach when listing other bumble bee species.  Specifically, it seems likely USFWS will only designate habitat immediately surrounding recently documented sightings, as opposed to using a general habitat model across the species' range.  Then, Section 7 consultation will be required for any activities that disturb habitat during the nesting period (i.e. growing season).

This seems to be the assumption underlying this Story Map, which shows small polygons of habitat as buffers around representative observations.  Note that the actual locations in this map are not accurate because the underlying observation data has been randomized to protect populations of rare species. 

This map of rusty patched bumble bee habitat around DeKalb, IL schematically shows the kind of habitat USFWS designated (i.e. buffered polygons around point observations) but does not show the accurate locations of the habitat because the data used for the map (GBIF) is randomized within 0.2 by 0.2 latitude/longitude rectangles.  

Mitigation Banking Could Transform the Endangered Species Act

 The Clean Water Act (CWA) --despite its ambiguities-- has the important provision of acre-for-acre wetland mitigation. In other words, the CWA ensures No Net Loss of protected wetlands.

The Endangered Species Act (ESA) --despite controversies over Critical Habitat-- has no automatic provision of no net loss of protected species habitats. Instead, it relies on bespoke mitigations on a project-by-project basis. Most projects are approved with incompletely mitigated impacts to species and their habitats. The result is continual loss of habitat.

Current proposed changes to habitat mitigation could help make ESA more like CWA, moving the ESA toward No Net Loss of habitat. The result would be improved regulatory certainty for projects, mitigation banking opportunities for conservation investors, and better outcomes for listed species.

Environmental Policy Innovation Center's Becca Madsen has more excellent & detailed analysis.

Thistle Misidentification

Thistles are amazing flowers, but there are both native and nonnative invasive species in the Western US.  Some people who try to do good by removing nonnative thistles accidentally kill the native species as well. 

NM Thistle, Cirsium neomexicanum. My photo on iNaturalist. Link: https://www.inaturalist.org/observations/15024423

I worked with an AmeriCorps crew in the Sacramento Mountains (Lincoln NF) that had been removing musk thistles.  But I found out they didn’t know how to tell the difference between native and invasive thistles. 

Photo of musk thistle Carduus nutans from iNaturalist user c-guinan. Link: https://www.inaturalist.org/observations/132367683

Even worse, they didn’t even know that there is an Endangered species of native thistle in those mountains!  We investigated and it looks like they at least didn’t kill any endangered thistles, but it could have been really bad. 

 

Photo of Endangered Sacremento Mountains Thistle Cirsium vinaceum from iNaturalist user ck2az.  Link: https://www.inaturalist.org/observations/14433149

I’m currently trying to find the volunteers on the Prescott national forest who are pulling up all of the Arizona thistle along popular trails.  Its especially troubling to me when people who are trying to do good by removing invasive species end up destroying native species.

Arizona Thistle Cirsium arizonicum.  My photo on iNaturalist.  Link:  https://www.inaturalist.org/observations/87880049

Everyone who’s done natural resource work has stories of project missteps.  I don’t know how to prevent all mishaps, but talking about these misidentifications is a good first step.

The NM Native Plant Society has a great Thistle ID book written by Bob Sivinski.