A Hybrid (Stabilized Ancient?) Population of D. arizonica and D. saxosa
Table of Contents
Introduction
How to Distinguish the Species and Their Hybrid
Geographic Distribution
Appendix 1: Detailed Differences between the two Species and Their Hybrid
Appendix 2: Pictures of Every Analyzed Specimen Shown In Order of Their PC1 values
Appendix 3: Detail on the Principal Component Analysis.
Introduction
A typical specimen of D. arizonica (PC1 = -2.91).Photo by Matt Berger (sheriff_woody_pct @iNat) | 
A typical specimen of D. saxosa (PC1 = 2.31).Photo by Grigory Heaton (gheaton @iNat) |
A typical hybrid of D. arizonica and D. saxosa (PC1 = -0.74).Photo by Don Rideout (lagoondon @iNat) | 
A typical hybrid of D. arizonica and D. saxosa (PC1 = -0.53)Photo by Jim Roberts (jimirob1 @iNat) |
| Fig. 1. Typical plants (as found in a Principal Components Analysis discussed below) of D. arizonica (top left) and D. saxosa (top right), and typical plants of a hybrid population between those two species (bottom row). The values of the first principal component (PC1) are given for each specimen.
Click on the pictures to go to the iNat observations for each. | |
In the Borrego Desert, most of the time it is quite easy to identify D. arizonica and D. saxosa at a glance; see the top row of Fig. 1 to see how different in appearance typical plants of these species are.
However, the plants in the Smuggler Canyon area and nearby have been difficult to reliably determine. When I first saw the plants in Smuggler, I thought they were D. arizonica, but the shape of the leaves puzzled me since they weren't what I thought as typical leaves of that species.
Furthermore, in the immediately-surrounding area outside of Smuggler Canyon, I've also puzzled over specimens. I speculated at the time that they might be hybrids since I couldn't clearly determine them between those two species, but over time started thinking they were simply a different expression of D. arizonica.
An observation posted at iNat from Smuggler Canyon generated considerable recent discussion, with, at one time or another, three determinations as D. saxosa and four determinations as D. arizonica, all by people who were pretty familiar with both species!
The discussion of this iNat specimen finally stimulated me to measure some specimens in iNat photographs, and do a PCA on the measurements to try to understand what was going on.
I got the answer quickly, after measuring just 23 plants. The PCA clearly showed that the Smuggler Canyon plants, and some plants from the vicinity of Torote Canyon, are hybrids between the two species. There are probably a number of other hybrid plants that await discovery.
Since the hybrid plants do not have any nearby plants of D. saxosa, it seems likely that this is a stabilized ancient hybrid between the two species. However, much more work is needed to understand what is going on here.
Finding hybrids between these two species was not unexpected. Reid Moran reports, in the Flora of North America Dudleya treatment:
D. Verity (pers. comm.) was able to cross Dudleya species in every combination he tried, regardless of morphology and level of ploidy, including such improbable hybrids as D. blochmaniae × pulverulenta, crossing two of the most unlike. Also, natural hybrids are known between most diploids that grow together; the hybrids are mostly rare (Moran 1951; Moran and Uhl 1952; P. H. Thomson 1993). All diploid hybrids that Uhl studied showed very good chromosome pairing, with no detectable abnormalities at meiosis.
Fig. 1 shows a comparison of the two species and their hybrids. The two species, and their hybrid, are represented there by specimens whose PC1 values were in the middle of their respective distributions.
This page gives a qualitative discussion of how one can distinguish the hybrids from the parents, and shows a map of the distribution of both species in the Borrego Desert.
The detailed analysis is in a separate webpage.
See also Pictures of Every Analyzed Specimen Shown In Order of Their PC1 values.
It will, of course, be very interesting to see what the blooms of these hybrids look like! Unfortunately, very few of the iNat observations determined as D. arizonica, which includes these hybrids, have flowers.
How to Distinguish the Species and Their Hybrid
Distinguishing the Two Species
In order to be able to reliably recognize the hybrid, one has to first know in detail the differences between the two species. Given below are the major differences between the two species, from analysis of the measurements I did on photographs of the largest leaf on each plant that I could measure. Because the leaves within a rosette can be variable, other leaves might have different characteristics. In particular, there is considerable variability in the shapes of the leaves in some plants of D. arizonica, whereas the leaves of D. saxosa tend to be more uniform.
The best discriminants are the width of the leaf, and the location along the leaf where the leaf is widest. D. arizonica has much wider leaves than D. saxosa relative to its leaf length, and its widest portion is typically near the middle of the length of the leaf, whereas the widest portion of the leaf of D. saxosa is at or near its base.
Specifically, the width of the leaf divided by the length of the leaf ranges from 0.3 to 0.75 for D. arizonica, with a median of 0.50, compared to a range of 0.10 to 0.30 for D. saxosa, with a median of 0.17. Those measurements only overlap at their very extreme values of 0.30.
The angle of the leaf tip and its shape also distinguish the two species fairly well. The leaf tip of D. arizonica is typically acuminate (slender, long-tapering to a point, with a different shape than the rest of the leaf), whereas the leaf tip of D. saxosa is just a continuation of the general tapering of its leaf. Surprisingly, though, the actual angle of the leaf tip is somewhat wider for D. arizonica in general.
These four characteristics are probably the main ones that allow most of us to recognize these two species "at a glance".
Distinguishing the Hybrid
As expected, the hybrid of the two species has intermediate values for some characteristics, with other characteristics more similar to one of its parents. This is what made it difficult for us to recognize the hybrid in the field, other than to be confused about which species the hybrids were. In fact, such confusion might be the first clue that one is seeing a hybrid.
The differences detailed below may help one to recognize some hybrids in the field, but I don't expect they can be used to reliably identify all such hybrids. As is often the case, it sometimes take detailed measurements, along with a Principal Components Analysis (PCA) of those measurements, to confirm a given specimen is a hybrid, and not just a specimen at the edge of the range of characteristics of the pure species. Even with a PCA, if the hybrids are not first generation (F1) hybrids and result from continued back-crossing with one of the parents, some specimens may be nearly impossible to distinguish from their parents.
The width to length of the hybrid leaves is in-between those of its parents, with a median ratio of 0.35, compared to a median ratio of 0.17 for D. saxosa and 0.50 for D. arizonica. The location along the leaf where the leaf is widest is also in-between for the hybrid, a median distance of 0.35 from the base of the leaf towards its tip for the hybrid, compared to a median of 0.00 for D. saxosa and 0.50 for D. arizonica.
The shape of the leaf tip tends to be more like that of D. arizonica, with the angle of the leaf tip tending more like that of D. saxosa. The leaves of the hybrid often demonstrate remarkable variability; see the bottom left plant in Fig. 1 and this iNat observation.
Geographic Distribution
The geographic distribution of D. arizonica and D. saxosa from iNat observations is shown in Fig. 2. I have not reviewed the vast majority of the iNat determinations, but a number of people quite familiar with these two Dudleya species have reviewed a lot of them, and hence most of the determinations are undoubtedly correct. I have tossed all observations with poor locations, either obscured or with an uncertainty radius larger than 100 m.
Also marked in Fig. 2, with much larger symbols, are the iNat observations that I used for my analysis, marked with my determination from the PCA.
Fig. 2. Geographic distribution of D. arizonica and D. saxosa from iNat observations (small symbols), with the determinations from my PCA analysis for 23 plants (much larger symbols). Note the strikingly-sharp boundary between the two species at about 33.05° north latitude. So far hybrids have only been found inside the main D. arizonica distribution in the southern part of this map. The hybrids near 33.0° north latitude are at Smuggler Canyon. The hybrids near 32.87° north latitude are near Torote Canyon, and are found near pure plants of D. arizonica.
There are at least two interesting things in this map.
First, there is a strikingly-sharp boundary between the two species at about 33.05° north latitude. Fig. 3 shows that boundary on a map that has physical features identified.
Fig. 3. The sharp boundary between D. arizonica and D. saxosa a bit south of Highway 78. The points correspond to both species, since I was unable to make a map like this with two different colors.
Such a sharp boundary seems very surprising. D. saxosa is a fairly widespread species, and in other areas it grows in habitats very similar to those south of the boundary line. Is it possible that D. arizonica has outcompeted it in that area? Or perhaps D. saxosa easily form hybrids with D. arizonica, and the greater number of D. arizonica plants have wiped out any pure plants of D. saxosa that were originally present there? Or are these plants ancient hybrids of D. arizonica and D. saxosa, and they have replaced pure plants of D. saxosa in this area?
It is also interesting that hybrids so far have only been found south of that dividing line, within the population of D. arizonica.
Second, there is a population of D. arizonica in the northern part of Fig. 2, the Coyote Creek area, where it appears there are roughly equal numbers of the two species present. Yet there are no plants of D. arizonica in the large D. saxosa population to the south of that zone.
In my analysis so far, I chose iNat observations to study based on latitude sampling, after seeing a preliminary version of the plot in Fig. 4. The southernmost hybrids were found when I measured samples near a latitude of 32.88°.
Fig. 4. Plot of latitude vs. PC1 (the first component from the Principal Components Analysis).
Although at first glance there appears to be a pretty good straight-line relationship between latitude and PC1, that is just an artifact of the north-south separation of the two species. There is no significant straight-line relationship within each species separately.
PC1 values between -5 and -1.5 correspond to D. arizonica. PC1 values between 0.7 and 4 correspond to D. saxosa. The hybrids have intermediate PC1 values between -1.1 and 0.5.
More work is needed to fill in this plot, especially near the boundary separating the two species near 33.05° north latitude.
Appendix 1: Detailed Differences between the two Species and Their Hybrid
These are preliminary results, since I have only measured so far seven plants of D. arizonica; 11 plants of D. saxosa; and 5 hybrids. With such few measurements, the range in reported values will undoubtedly be at least somewhat larger after I make more measurements.
The measurements are reported in declining order of how well they separate the species.
Width of leaf divided by length of leaf. D. arizonica has much wider leaves than D. saxosa. The width divided by length ranges from 0.3 to 0.75 for D. arizonica, with a median of 0.50, compared to a range of 0.10 to 0.30 for D. saxosa, with a median of 0.17. The hybrid has a range of 0.19 to 0.38, with a median of 0.35.
Measurements were also made of the width of the leaf 1/4, 1/2, and 3/4 of the leaf length, measured from above the base of the leaf, which similarly differentiated the species as the width divided by the length.
Location on the leaf where it is widest. Nearly all leaves of D. arizonica have their widest portion in the middle of the length of the leaf. Nearly all leaves of D. saxosa have their widest portion at the base of the leaf.
The widest portion of the leaf of D. arizonica occurs 0.33 to 0.70 times the length of the leaf above its base, with a median of 0.50 times. The widest portion of the leaf of D. saxosa is at the base for eight of the 11 measured specimens, with the others at 0.23 to 0.32 times the length of the leaf above its base. The hybrid has a range of 0.29 to 0.45, with a median of 0.35.
Shape of the leaf tip. The tips of the leaves of D. arizonica are essentially all acuminate, as are the tips of the leaves of the hybrid. The tips of the leaves of D. saxosa range from being not mucronate and not acuminate, to being mucronate (having an abruptly smaller short tip).
Angle of leaf tip. The leaf tip angle of D. arizonica ranges from 20 to 60°, with a median of 35°. The leaf tip angle of D. saxosa ranges from 15 to 35°, with a median of 20°. The hybrid ranges from 10 to 25°, with a median of 15°.
Number of rosettes per plant. All measured plants of D. arizonica had only a single rosette. The number of rosettes for D. saxosa ranged from one to 14. The hybrid ranged from one to two rosettes.
Appendix 2: Pictures of Every Analyzed Specimen Shown In Order of Their PC1 values (This Appendix is a separate webpage)
Appendix 3: Detail on the Principal Component Analysis. (This Appendix is a separate webpage)