Plant Species of San Jacinto Mountain: Arabis = Boechera species with non-erect fruit
Introduction
List of Species and Abundances at SnJt
Key to Separate These Species
Principal Component Analysis of These Species
Plots of Characteristics Distinguishing These Species
Pictures of Each Species
Introduction This page currently is a place-holder to give the results of the Principal Component Analysis and show pictures of these species.
For background information on these species, including the characteristics of these species in the floras, voucher information, and recent changes, see the older page, done prior to the analysis reported in this page, Arabis = Boechera species with non-erect fruit: Background Information.
In particular, the linked page explains why our most common high-elevation species now must be referred to as the "Boechera species formerly known as Arabis holboellii"; or "Boechera retrofracta X unknown parent", which we shorten to just Boechera retrofracta X. Briefly, the name A. holboellii was misapplied to these plants, which, like many Arabis / Boechera species, are strictly self-fertile stabilized hybrids from the distant past. Many of these widespread species, originating from different parental crosses, turn out to have similar characteristics, and hence have been lumped under the name A. holboellii.
Unfortunately, the current Arabis / Boechera experts have declined to provide these plants with a Boechera name, preferring that each of these widespread species be called by their hybrid name. Even more unfortunately, no one knows currently who the parents are of the San Jacinto plants formerly known as A. holboellii, so they would have to be referred to as B. unknown parent 1 X B. unknown parent 2, although one voucher has been annotated as B. retrofracta X B. unknown parent. Since that is incredibly unwieldy, and also implies to most people that these are rare hybrids instead of widespread species, we will refer to these plants as "A. holboellii", including the quotes, or "B. retrofracta X", until they get a proper binomial name under Boechera.
These hybrids are for the most part not even listed in the 2011 second edition Jepson Manual. The online edition has this comment: Hybrids with spheric pollen and morphological intermediacy gen more common than sexual diploids (with ellipsoid pollen).
There is nothing special about "A. holboellii" in this regard; A. sparsiflora var. californica is a similar hybrid. If this taxon hadn't been so widespread in southern California, the 2011 Jepson Manual authors would have declined to give it a Boechera name as well. Their Flora of North America treatment states:
Morphological evidence suggests that Boechera californica is an apomictic species that arose through hybridization between B. arcuata and B. pulchra (see M. D. Windham and I. A. Al-Shehbaz 2007 for detailed comparison). In the flora area, B. californica is known only from Los Angeles, Riverside, San Bernardino, and San Diego counties.Arabis pulchra var. gracilis is also said to be a hybrid between A. pulchra var. pulchra and A. perennans, but it was given the Boechera name of B. xylopoda.
List of Species and Abundances at SnJt Table 1 gives a list of the Arabis species with non-erect fruit found in our SnJt surveys at elevations greater than ~4000 feet, with both their 1993 Jepson Manual name under Arabis (JM1 column) and their 2011 Jepson Manual name under Boechera, where present. They are given in descending order of the number of survey areas. Note that species with erect fruit are not listed in this table, nor are discussed in this page!
Table 1. Arabis species found in our high-elevation SnJt Surveys.
JM1 Species Name JM2 Species Name # of Survey Areas # of Vouchers "A. holboellii" none; We call it B. retrofracta X 23 5 A. sparsiflora var. californica B. californica 13 43 A. repanda B. repanda 4 1 A. perennans B. perennans 2 3 Clearly, voucher counts do not reflect the actual abundance of these taxa in our surveys!
This is primarily an elevation effect, in that "A. holboellii" is the only species found above about 6000 feet elevation, where most of our surveys have taken place, but where few vouchers are taken since that area is mostly without roads. Also, this species is most likely to be misdetermined as another species.
A. sparsiflora var. californica is the most common taxon at mid-elevations in the Idyllwild area, so it is not surprising it has the most vouchers.
A. perennans is mostly a lower elevation species that barely enters our survey areas.
There are five vouchers of A. pulchra from higher elevations that are probably actually misdetermined vouchers of "A. holboellii" or A. sparsiflora var. californica.
In addition, A. pulchra vars. gracilis and pulchra probably also occur at SnJt at lower elevations. We include A. pulchra in this page since some of the above species have been previously misidentified as that taxon, including by ourselves. Although A. pulchra almost certainly exists at SnJt at lower elevation, we have never seen either variety of it in flower or fruit here. We have only seen var. pulchra in flower and fruit from the Sierra Nevada, and so include that observation below for comparison. We had not seen var. gracilis when the analysis below was done, so it could not be included. We finally saw var. gracilis at Whale Peak in San Diego County in 2012,and added it to the pictures presented below.
A. sparsiflora var. arcuata apparently does not occur at SnJt. The second edition Jepson Manual treatment does not have that taxon south of the Transverse Range, and vouchers formerly determined as that taxon have been redetermined by the Jepson Manual authors as var. californica. There is no evidence in our Principal Component Analysis that there is more than one taxon within the A. sparsiflora plants here.
Key to Separate These Species The following key has been adapted from the 2011 second edition Jepson Manual Boechera key (numbered key elements), with the addition of "A. holboellii" from SnJt (lettered key elements). There is no guarantee that plants called "A. holboellii" from any other location in southern California are the same as the high elevation SnJt "A. holboellii" plants, since they may be different species.
This key includes B. arcuata, which is probably not present at SnJt, and B. xylopoda, which might not be present at SnJt.
1. Fr, gen ovary hairy 3. Fr pedicel base gen abruptly recurved; fr gen appressed to infl axis .... B. pulchra 3' Fr pedicel base not abruptly recurved; fr gen not appressed to infl axis 5. Proximal sts with simple and 2-rayed hairs only ...................... B. arcuata(2) 5' Proximal sts with at least some 3-12-rayed hairs 13. Fr gen > 7 cm, sparsely hairy throughout; seeds 1.7-2 mm, gen in 1 row ....................... B. californica(2) 13' Fr gen < 7 cm, hairy on distal 2/3, glabrous near base; seeds 2-2.5 mm, gen irregularly in 2 rows........... B. xylopoda 1' Fr, ovary glabrous 50. Fr descending, pendent, or strongly reflexed 55. Fr pedicel hairy ................................................. B. californica(2) 55' Fr pedicel glabrous (sparsely hairy) A. Fr base ~horizontal ............................................. B. perennans A'. Fr base ~vertical ............................................... "B. retrofracta X" 50' Fr ascending to horizontal ............................................ B. arcuata(2)
Principal Component Analysis of These Species The analysis done here is very unusual, in that it has been solely done from pictures. Because we have been immensely confused about these Arabis species at SnJt, we have taken detailed pictures of many different specimens. These pictures allow us to measure 27 different quantities, including the stem, petiole and fruit hair densities; basal and cauline leaf shapes and numbers; number of flowers, their color, and ratios of petiole and calyx to petal lengths; fruit pedicel and body angles to the stem, and the ratio of pedicel in fruit to the fruit body. These quantities are more than enough to reliably distinguish how many taxa are present in our surveys, independent of whether we thought any given specimen was a different species or not, and independent of any floral key or description.
This kind of analysis has the benefit that every measured characteristic is documented with a photograph, and that it is easy to add additional specimens to the analysis from similar detailed photos taken by anyone else in the future.
We measured 21 different specimens, not all of which had both flowers and mature fruits. 14 of those specimens had information on each of 15 different characteristics that were analyzed with a Principal Component Analysis. We included a specimen of A. perennans from Mine Canyon in Anza-Borrego Desert State Park (ABDSP), from 3000 feet elevation, to see how consistent the single specimen of A. perennans from SnJt was with a specimen found in a known location for A. perennans.
Table 2 gives the locations of each specimen that went into the PCA, along with the date each specimen was photographed.
Table 2. Arabis Specimens Analyzed in the Principal Component Analysis.
Taxon Location Date "A. holboellii" Ernie Maxwell Trail 5/22/2009 "A. holboellii" Hidden Lake 6/25/2008 "A. holboellii" Round Valley 7/15/2008 "A. holboellii" Tahquitz Peak 5/27/2009 "A. holboellii" Tahquitz Peak 5/11/2007 A. perennans Mine Canyon, ABDSP 3/4/2011 A. perennans PCT North of SR74 4/19/2009 A. pulchra Walker Pass, Sierra Nevada 5/12/2011 A. sparsiflora var. californica Black Mtn Trail - on trail 5/25/2011 A. sparsiflora var. californica Black Mtn Trail - trailhead 5/25/2011 A. sparsiflora var. californica Ernie Maxwell Trail - 1 5/22/2009 A. sparsiflora var. californica Ernie Maxwell Trail - 2 5/22/2009 A. sparsiflora var. californica Ernie Maxwell Trail 6/17/2008 A. sparsiflora var. californica South Ridge Trail 5/27/2009 Specimens of A. repanda were not measured since they are very distinctive in the field and not likely to be confused with one of the other taxa.
The analysis showed clearly that those 14 specimens comprised four separate taxa, which were then identified to species using the 2011 second edition Jepson Manual key except for "A. holboellii", taken from the 1974 Munz Flora of southern California.
Figure 1 shows the results of the PCA, with the taxa labeled:
Fig. 1. Principal Component Analysis of 21 Arabis Specimens
Species separations hardly get any better than shown in Fig. 1. Each species is cleanly separated from the other species, and specimens of a single species are spread over a very tight range of only ± one standard deviation or less.
As advertised on the Background Information page, the PCA shows that A. pulchra is a very distinctive species, separated from all the others here by eight standard deviations, reflecting the large number of characteristics for which it is distinctive.
Of course, a plot like the above is spectacularly uninformative about exactly what is it that differentiates the species. The following section gives links to plot of the characteristics found to be significantly different, plotted two at a time. The power of the PCA is that it in essence combines all of the following plots into a single one.
Plots of Characteristics Distinguishing These Species These plots include all 21 measured specimens, not just the 14 that went into the PCA. Not every plot includes every specimen, since some specimens didn't have some of the features needed for every measurement, such as mature fruit, and some sets of photographs were incomplete.
Because A. pulchra is so distinctive, and thus compresses the range shown in each plot for the other specimens, it is included only in the first plot.
Explanations of what quantities are measured will be provided in the future.
Number of Cauline Leaves vs. Number of Basal Leaves
Cauline Leaf Shape vs. Basal Leaf Shape
Petal Color vs. Number of Flowers per Stem
Pedicel vs. Calyx Length, both scaled by Petal Length
Angle of Pedicel Tip (Fruit Body Base) from Erect Vs. Angle of Pedicel Base (maximum values for each specimen)
Pictures of Each Species Fig. 2 gives pictures of the three taxa with fruit that are ± spreading at their base and recurved toward their tips: A. sparsiflora var. californica, A. perennans and A. repanda. Fig. 3 gives pictures of the three taxa with fruit that are mostly hanging downward: "A. holboellii", A. pulchra var. gracilis and A. pulchra var. pulchra.
All pictures are from San Jacinto Mountain except for all pictures of A. pulchra var. gracilis (which are from Whale Peak, Anza-Borrego Desert State Park in San Diego County); A. pulchra var. pulchra (which are from Walker Pass in the southernmost Sierra Nevada in Kern County); and the picture of the inflorescence in fruit for A. perennans, which is from Mine Canyon, Anza-Borrego Desert State Park in San Diego County).
Because Arabis pulchra var. gracilis is said to be a hybrid between A. pulchra var. pulchra and A. perennans, we put the pictures of those three taxa side by side in Fig. 4 on a separate page.
Click on each picture to get a larger version.
Fig. 2. Pictures of the three taxa with fruit that are ± spreading at their base and recurved toward their tips: A. sparsiflora var. californica, A. perennans and A. repanda
Fig. 3. Pictures of the three taxa with fruit that are mostly hanging downward: "A. holboellii", A. pulchra var. gracilis, and A. pulchra var. pulchra
Go to:
Copyright © 2011-2018 by Tom Chester and Dave Stith.
Commercial rights reserved. Permission is granted to reproduce any or all of this page for individual or non-profit institutional internal use as long as credit is given to us at this source:
http://tchester.org/sj/species/arabis_non_erect_fr/index.html
Comments and feedback: Tom Chester
Updated 15 July 2018.