Using iNat maps to illustrate Ensatina as a ring species

There has been some recent discussion on this forum about why some genera have huge numbers of species and what even constitutes a species.

Some iNat users may have come across the concept of a “ring species”. This is the paradoxical situation where each population within a distribution can breed with its neighbors and yet there are some places where “end” populations coexist and cannot interbreed. Generally there has to be some kind of geographic obstacle that restricts population interactions.

One classic example of this in the western United States is the Ensatina (Ensatina eschscholtzii). Ensatinas are a type od salamander that lives in California’s Coast Ranges and Sierra Nevada, stretching north through Oregon and Washington to southern British Columbia, and south into the northern portion of Baja California. Across this range, herpetologists have identified seven subspecies of Ensatina, each of which has distinct coloration.

One area Ensatinas do not live in is California’s Central Valley (sometimes called the Great Valley), as it is too dry and hot. As a result, as you move down the west coast from north to south, there distribution of Ensatinas splits into two arms at the southern end on the Klamath Range. One arm continues down the coast through San Francisco and Los Angeles all the way to Tijuana, Mexico. The other heads south-east to the western slopes of the Sierra Nevada. Each arm of the population represents a continuum of interbreeding populations forming an elongated oval “ring” of something like 19 identifiable populations from 7 subspecies all of which can interbreed with their neighbors.

But at the south end of the Central Valley, Yellow-blotched Ensatinas of ssp. croceater that live in the Transverse Ranges (marked brown in the first map below) cannot interbreed with neighboring Monterey Ensatinas of ssp. eschscholtzii (marked bright blue).

This illustrates how mutual fertility can be problematic as a criterion for defining the species concept. None of the subspecies can be elevated to species rank because each can interbreed with at least one other. And yet some of the subspecies definitely cannot interbreed even though they live right next to each other.

This paradox was first uncovered by Robert Stebbins in the late 1940s and has been the subject of extensive study since then. There’s a good overview at UC Berkeley’s Understanding Evolution site. Richard Dawkins gave it some prominence in his 2004 book The Ancestor’s Tale.

This made me wonder what this ring pattern looks like among iNat observations. iNat actually has two ways to map out this type of distribution and both of them illustrate the ring species concept pretty well.

One option is an iNat Taxa Map. There isn’t a user interface for this, so you need to build up the URL yourself, but it’s not too complex. You just add the various taxon IDs you want to map into a URL such as the following: https://www.inaturalist.org/taxa/map?taxa=120135,144104,154880,123959,123689,142202,123168,518535#5/41/-120 After your list of taxon IDs you can specify a magnification level (5), and a latitude and longitude (41/-120 to indicate that the map will be centered at 41 degrees north and 120 degrees west). This map shows observations of the seven subspecies of Ensatina plus one recognized hybrid form.

The second mapping option uses the experimental Compare tool. This provides an interface for iNat users to build maps that use different colored markers for each of several different search criteria (typically taxon IDs). Here’s the Compare map for the same seven Ensatina subspecies and the hybrid.

Advantages of the Compare tool include:

  • An actual UI to build your map choices
  • Supports other search criteria, not just taxa
  • Supports reordering of groups, which also swaps marker colors
  • Allows you to label each group of markers

Advantages of the Taxa Map include:

  • Has an intelligible URL
  • Not deemed “experimental”
  • Can default the center point and zoom level
  • Users can toggle on and off markers for GBIF observations

Either way, if you build something nice with one of these tools I recommend you keep a note of the URL (maybe share it here).

[Edited 12/4/2020 to better describe the sequence of radiation proposed by Stebbins.]

30 Likes

Nice job Rupert! I’ve still only seen four of these subspecies. Someday…

3 Likes

This is awesome! Thanks so much for sharing.

2 Likes

I was happy to see the final subspecies in 2019. I haven’t seen the recognized hybrid form yet though.

2 Likes

None of the subspecies can be elevated to species rank because each can interbreed with at least one other.

If only it were so simple. :)

Ring species are sometimes defined as separate species, as far as I understand: see the genus Larus for example. Also, we have examples of successfully interbreeding different genera. Bison (genus Bison) and cattle (genus Bos) can and do interbreed, so much that a small percentage of bovine DNA is today common in bison.

3 Likes

Fascinating. I assume that a small percentage of bison DNA is also found in cattle?

2 Likes

Is it possible to find ring species that were previously undiscovered via mapping, or would ground observations have to also have to collected to verify a species as such?

1 Like

perhaps not since during the time of interbreeding there was presumably a way lower bison than cow population

1 Like

That’s an interesting idea. I think iNat observations and mapping could play a part, but you would need a lot more info to really demonstrate the validity of a ring species. You’re going to need to be able to distinguish different populations visually, which means someone has probably already identified multiple subspecies or a species complex. And you’re probably going to need to know which of those subspecies interbreed and which don’t.

That article on UC Berkeley’s Understanding Evolution site gives a good summary of how evidence for Ensatina as a ring species was built up over time.

The article (which may now be 10+ years old) also hints at some ways that iNat data could contribute to this type of research. At that time, grad student Tom Devitt was looking at how the hybrid Ensatina eschscholtzii ssp. eschscholtzii × klauberi came to be. Part of his research focused on whether the two parent subspecies encountered each other frequently enough for hybridization even to become likely. And that does seem like an area where lots of iNat observations could add valuable data.

2 Likes

@haemocyanin11 used the Taxa Map tool to visualize iNat observations of the 10 infrataxa of the Peppered Moth (Biston betularia): https://www.inaturalist.org/projects/peppered-moths-of-the-world/journal/44532-a-map-showing-the-different-subspecies

Currently, only 120 of the 2,800 Peppered Moth observations on iNat are ID’ed to subspecies- or form-level. That suggests that there’s a lot of scope for identifiers to uncover useful info about the distribution of different infrataxa for this moth, which is common in Europe and North America with scattered populations across Asia, and has long been the subject of evolutionary research.

1 Like

Your described method also works to find interesting outliers, which may be valid waifs, or (usually) a good candidate for quality control on observations–note the pink boxes in the Sierra Nevadas!

3 Likes

Yes. These types of maps can be a good way to spot outliers.

In the case of the pink spots in the Sierras, I believe these are a recognized population of Ensatina eschscholtzii ssp. xanthoptica (Yellow-eyed Ensatina). It is thought that in cooler, wetter conditions, the range of this subspecies stretched across the Central Valley, but that it has split into separate populations as the climate got warmer and drier: http://www.californiaherps.com/salamanders/pages/e.e.xanthoptica.html#description

3 Likes

Very interesting! I wonder if this points to a more recent continous wetland habitat that only recently closed up between SF Bay and that region.

Tulare Lake springs to mind, the northern border may be about right for this, it’d be interesting to do an overlay if possible.

2 Likes

I remember finding a member of that population - totally thought it would have been a different subspecies as xanthoptica is the subspecies in my county. Definitely interesting and cool they are found on both sides of the valley.

1 Like

There were some interesting conclusions from mitochondrial DNA research on Bay Area Ensatinas published in 2009 by Sean Kuchta and Duncan Parks in David Wake’s lab at Berkeley (reference below). They showed that that there was a surprisingly high level of genetic diversity and the evolutionary relationships are a lot more complex than was originally suspected.

The overall picture is a patchwork of many distinct lineages that rarely overlap:

The xanthoptica and eschscholtzii coastal clade

  • The ssp. xanthoptica and ssp. eschscholtzii populations form a “coastal clade” that is separate from the other Ensatina subspecies, which are all more closely related to each other than to the coastal clade.

  • Within xanthoptica there are actually two separate clades (identified as xanthoptica [1] and xanthoptica [2]).

  • The xanthoptica [1] clade has “a dark brown back, a vibrant orange belly, orange proximal limb
    segments, and a striking yellow eye patch [and] is thought to be Batesian mimetic of Pacific newts, genus Taricha … which are highly toxic and possess a similar aposematic coloration.” This clade occurs both in the Sierra Nevada and north and east of San Francisco Bay.

  • mtDNA analysis of xanthoptica [1] suggests there are three different basal lineages in the Sierra Nevada, whereas the East Bay xanthoptica [1] populations form a monophyletic clade. This is consistent with a scenario where xanthoptica [1] originated in the Sierra Nevada and later colonized the Coast Ranges. However, this determination did not exceed the 95% confidence threshold for statistical validation and it’s still possible that xanthoptica [1] originated on the coast.

  • The relatively weak level of divergence between Sierra Nevada and East Bay populations of xanthoptica [1] suggests that this “transvalley leak” was likely open during the Pleistocene (i.e. up until 11,700 years ago). Similar patterns of colonization from the Sierra to the Coast Range have been seen in other California salamanders and snakes.

  • The East Bay xanthoptica [1] lineage was apparently the source for the few xanthoptica [1] populations on the eastern edges of the San Francisco Peninsula.

  • The distinct xanthoptica [2] clade “has a less conspicuous coloration than xanthoptica [1], and was considered by Stebbins (1949) to be xanthoptica x oregonensis intergrades”. This clade occurs in the mountains around Santa Cruz in the southern part of the San Francisco Peninsula.

  • Two different lineages of ssp. eschscholtzii were identified, with the northern lineage stretching as far north as Monterey Bay where it comes within 7.8 km of the southernmost xanthoptica [2] population. Due to extensive farming of this area, it’s unlikely there remains much opportunity for these populations to hybridize.

Separate histories for two oregonensis clades in the Bay Area

Kuchta, Parks and Wade also found interesting results about the ssp. oregonensis populations:

  • There are two different ssp. oregonensis clades in the Bay Area (identified as oregonensis [1] and oregonensis [2]). “Both lineages possess a generalized camouflaged color pattern with a light brown back and a pale belly, and are very similar in appearance. In general, oregonensis [1] is particularly drab in color, while the limbs, tail, and torso of oregonensis [2] possess subdued yellow and orange elements, which is why Stebbins (1949) interpreted them as intergrades between oregonensis and xanthoptica

  • Despite their visual similarity, the two clades are genetically quite distant, with oregonensis [2] apparently more closely related to the picta, platensis, croeater and klauberi subspecies, plus various other ssp. oregonensis clades, than it is to the more basal oregonensis [1].)

  • The oregonensis [1] clade lives close to the coast. Kuchta, Parks and Wake mapped three lineages, one in Sonoma and Mendocino counties, a second in the southern part of the Point Reyes Peninsula and the third, south of San Francisco, along the coast of San Mateo and Santa Cruz counties. They hypothesize that during the Pleistocene, when sea levels were lower and the coastline was further west, these three lineages were part of a continuous population. As the sea level rose at the end of the Pleistocene, the three lineages became isolated from each other. Based on its current distribution the authors suggest that oregonensis [1] may be best suited to “low elevation terraces with a relatively cool, moist climate.”

  • The oregonensis [2] clade generally lives a little further inland. Kuchta, Parks and Wake mapped four Bay Area lineages, one on the San Francisco Peninsula (north of the oregonensis [1] and xanthoptica [2] lineages), a second in Sonoma and Napa counties (north of the North Bay xanthoptica [1] lineage), a third around the Russian River, west of Santa Rosa, and a fourth in southwest Sonoma County and Marin County (surrounding the Point Reyes Peninsula oregonensis [1] population). They noted a fifth oregonensis [2] lineage further north, outside their study area.

Bringing this back to mapping iNat observations, there’s both an opportunity and a challenge here. By adding a substantial number of geolocated Ensatina observations, iNat users can contribute to the knowledge of distribution for these different subspecies. But the cryptic nature of the lineages involved will still limit the conclusions that can be safely drawn. In particular, it seems unlikely that oregonensis [1] can be reliably distinguished from oregonensis [2] without DNA sequencing, so it would be difficult to draw conclusions from iNat observations about interbreeding between these clades.

As is the conclusion of many a research paper: “More study is required.”

Source: Shawn R. Kuchta, Duncan S. Parks, David B. Wake,
Pronounced phylogeographic structure on a small spatial scale: Geomorphological evolution and lineage history in the salamander ring species Ensatina eschscholtzii in central coastal California,
Molecular Phylogenetics and Evolution, Volume 50, Issue 2, 2009, Pages 240–255, ISSN 1055-7903, https://doi.org/10.1016/j.ympev.2008.10.019

5 Likes

Thank you, @rupertclayton . If you don’t mind, a couple of questions that are tangential to this interesting discussion:

  1. Are there instructions on how to use these tools? I’ve not come across that.
  2. looking at your Taxa Map, is there a way to put a calendar boundary on the observations?

Thank you.

1 Like

Very interesting! Both content and the used techniques. Many thanks!

1 Like

This is amazing, and a wonderful way to use these tools! Very cool and informative posts retracing one of the basics of amphibian genetics in iNat, I love it!

2 Likes

Hi @charlescrussell. I don’t believe there’s any official documentation for these mapping tools. I think both of them were mentioned by iNat staff in passing in various discussions on this forum. Here are some threads with info on the Taxa Map and/or the Compare tool:

Maybe there’s something more formal, but I’m not aware of it. Basically, for the Taxa Map, you can take the URL I listed above and substitute your own taxon IDs. For the compare tool there’s an actual interface, but you still need to add in your own search terms, which use the same syntax you’d see on the Explore and Identify pages.

As far as I know, it is possible to do this with the Compare tool but not the Taxa Map. You can use all the same search parameters that are available for Explore and Identify searches. For example, this first string below would map all the Ensatina eschscholtzii ssp. orgeonensis observations between April 1 and June 30 this year, and the second would map all observations in April, May or June of any year:

taxon_id=123959&d1=2020-04-01&d2=2020-06-30

taxon_id=123959&month=4,5,6

Are you looking for something like that?

2 Likes

Thank you - I will explore this further

1 Like