Field Trips Log
September 2004

Thursday, 2 September 2004 (Eastern end of Angeles National Forest)

Continuing my work on Tom Chester's rabbitbrush survey, I drove up I-15 to Cajon Junction and the turnoff onto SR138 and headed toward Lone Pine Canyon Road. Tom had GPS'd quite a few locations of rabbit-
brushes in this area on a trip in 2002 and had concluded at the time that they were all ssp. mohavensis. Since this was the subspecies that I felt the least familiarity with, I was eager to see a lot of it and compare it with what I thought had been mohavensis on the Angeles Forest Highway.

When I stopped at the first GPS point at the entrance to Mormon Rocks Fire Station and saw a bunch of obvious rabbitbrushes, my initial impression was, "These plants sure look a lot like the "green forms" of what I had decided must be hololeucus on the Angeles Forest Highway." They were about at the same stage, about the same overall size, with long gray-green stems and little clusters of new greenish buds. But since I didn't know what mohavensis was supposed to look like, I couldn't immediately rule it out as a possibility. Although the stems were just as leafy as the "green forms" on AFH (whereas mohavensis is supposed to be ± leafless at bloom time, Tom had said that the mohavensis he had seen here in 2002 were leafy at this same stage.

I next stopped at the junction of SR138 and Lone Pine Canyon Road and observed the same-looking rabbit-
brushes there. I continued up Lone Pine Canyon Road, stopping every half-mile or so, and seeing nothing but the greenish looking rabbitbrushes, which really had me wondering. As I neared the highest point of the road just south of Wrightwood, I noticed a tall blooming "rabbitbrush" off to the side of the road, so I stopped to take a look at it. It was about 7-1/2' tall, much taller than a rabbitbrush should be. It had yellow blooms that looked very similar to rabbitbrush blooms, with spreading corolla lobes, but the phyllaries were not at all like typical rabbitbrush phyllaries. They were wide, almost flat, with rounded tips, and were not lined up in vertical rows. I took pictures of it and was glad I did, because when Tom saw them, he was able to immediately identify this species as Nevada broomsage (Lepidospartum latisquamum), which has a very limited range in Southern California and which was a new species for me.

Not far beyond this, I came to a place I had stopped at last year and taken pictures of what we had identified as mohavensis in bloom. When I went back later and looked at my photos of this plant, it was clear to me from its densely white-tomentose stems and leaves that it was ssp. hololeucus. At this location there were many of the "green form" rabbitbrushes (not blooming) and a rabbitbrush that looked very much like what I now thought I recognized as ssp. consimilis, which was blooming.

Continuing on through the community of Wrightwood to the west along SR2, I saw only a few more of the "green form" rabbitbrushes, but now the blooming "consimilis" was lining the road on both sides. I stopped half a dozen times to photograph and sample different plants as I approached then passed through Big Pines. Now I was looking for ssp. bernardinus because I knew that along this eastern end of the Angeles Crest Highway both consimilis and bernardinus could be expected. I didn't see the first bernardinus until I was heading up toward Blue Ridge Summit, and there was one on the side of the road with a consimilis conveniently located right next to it. The different between these two subspecies is quite striking, and once you have seen a lot of them there's no way they could be confused. The fact that the consimilis was overlooked in the original surveys illustrates the very important point that you always have to be on guard against making conclusions based on preconceptions. Consimilis was not expected to be here, and there is supposed to be a greener form of ssp. bernardinus, just as there is a green form of hololeucus, so when this was originally observed, it was believed to be a part of the bernardinus complex.

Arriving at Vincent Gap, I headed down Big Rock Creek Road toward the desert. Vincent Gap is also the trailhead for the Manzanita Trail going north, and when I looked at some photos I had taken of what I had down as bernardinus (clearly taken at a time when I didn't know any of the sspp) on the very top part of that trail, I was surprised to see that it was an obvious hololeucus. So Vincent Gap is one of the places where three sspp. are situated within a very close proximity to each other.

Big Rock Creek is a dirt road but is not too bad to drive with a normal vehicle. As I descended, I saw both scattered individuals of both bernardinus and what I thought was consimilis. The road becomes paved just about where there is an LA County prison, and then a few miles further on I arrived at Tom's first GPS point for hololeucus. There was a large "white ghost" surrounded, as all the ones on Angeles Forest Highway had been, by "green forms." And looking at the "green forms" again closely, I was beginning to suspect that all the greenish looking rabbitbrushes down on Lone Pine Canyon Road were in fact hololeucus and not mohavensis. This was good because it related the hololeucus on AFH with the hololeucus in Lone Pine Canyon and Big Rock Creek, but it now raised the question of where was all the mohavensis that was supposed to be here. I continued to the junction of Big Rock Creek Road and Valyermo/Big Pines Highway and saw masses of white ghosts," more than I had sen anywhere else. I also saw several individuals of a somewhat different-looking rabbitbrush with slender, green, leafless stems, and I thought at the time that this might be the mohavensis, but the difficulty was that it didn't look like the plant I had seen and thought was mohavensis on AFH. Tom looked at these plants later and concluded that they were probably just consimilis that were in a drier location and had lost their leaves.

I drove back up Rock Creek Canyon Road and returned home via the Angeles Crest, stopping at a few places to confirm again that both bernardinus and consimilis mixed together along that highway. At a turnout near Three Points, Tom had GPS'd a location where there were individuals of all three sspp. within about 30' of each other, so I checked that out and then went home. I felt fairly confident now in both of the forms of hololeucus (although I plan to go back in a month or so and see the "green form" blooming) and both bernardinus and consimilis. Whether at this point I have seen mohavensis or not, I am not sure, but that will have to wait for another day. My next excursion will be to survey SR39 above Azusa which should have consimilis just like Mt. Baldy Road, but possibly also bernardinus near the top. Steve Boyd's Liebre Mts flora suggests that the core of the mohavensis population may be in that area, so that will be yet another trip.

Friday, 3 September 2004 (SR39, San Gabriel Mts)

My goal for today was to survey for rabbitbrushes along SR39 above Azuza. I had no anticipations of surprises on this road since it, like Mt. Baldy Road, would be expected to have ssp. consimilis, except near the very top where ssp. bernardinus might put in an appearance. It would seem unlikely for mohavensis to have come over onto the coastal side, although Tom's assessment of the consimilis/mohavensis on the south side of Mill Creek Summit is that it was in fact mohavensis, so I needed to be prepared for anything.

SR39 winds up through San Gabriel Canyon, past Morris and San Gabriel Reservoirs, Rincon Fire Station, Coldbrook Ranger Station and campground, the turnoff to Crystal Lake, and ends at a gate several miles from the Angeles Crest Highway. It originally connected with the Angeles Crest, but due to unstable terrain and repeated rockslides was closed several times, and in 1978, closed indefinitely, a closure which has continued to the present. I hear occasional rumors of Caltrans plans to reopen it, but frankly think it better for it to be left alone.

On the way up 39 I saw in bloom quite a few shrubby butterweeds (Senecio flaccidus), erect goldenasters (Heterotheca sessiliflora), twiggy wreathplants (Stephanomeria virgata), feltleaf or white everlastings (Gnaphalium canescens ssp. microcephalum), and some species of Oenothera. Approximately three miles above Coldbrook Campground, I passed an area where there was water actually cascading down some rocks right next to the road, so I stopped to take a look. There was a lot of beautiful California goldenrods (Solidago californica), Hooker's evening primrose (Oenothera elata ssp. hookeri), identifiable by its obvious red blister bases on the long stem hairs, and columbines (Aquilegia formosa). I was momentarily excited because I thought the maidenhair fern was Adiantum jordanii, which I have yet to knowingly see (it turned out just to be the usual capillus-veneris), but then caught a flash of blue among some grasses. I bent down and peered more closely, and was immensely pleased to have found at last the lovely Dunn's lobelia (Lobelia dunnii var. serrata) which I had wanted for years to see. Now if I could just find the red one, Lobelia cardinalis, I'd be even happier.

After having had to look at hundreds of yellow-blooming scalebrooms (Lepidospartum squamatum), I finally saw the first rabbitbrush at an elevation of about 4800', and it looked like consimilis as expected. I thereafter stopped every half-mile or so to check the identity of the rabbitbrushes which were all blooming and all looked like consimilis to me. The road to Crystal Lake was closed, but I continued on the top of SR39, which is at an elevation of about 5500'. I saw nothing but what appeared to me to be consimilis. The views from the top of the road are spectacular, back down San Gabriel Canyon, and north toward Mt. Williamson and the other peaks of the higher San Gabriels. From the closed gate, I set out on foot and continued another mile and half, but it appeared that it was at least another couple of miles to the Angeles Crest, and the heat persuaded me to turn back, not having found anything different. Having accomplished my mission for today, I set off for home.

Addendum 9/4/04: Tom's analysis of his rabbitbrush samples from the Angeles Forest Highway and Angeles Crest Highway (as compared with samples of ssp. consimilis from Mt. Baldy Road) has introduced a fly into the ointment, as they say. I will not attempt to summarize the specifics of his analysis, which at some point will be linked from this page, but suffice it to say that it is his at least tentative conclusion not only that the plants I saw on AFH were mohavensis, but likely also the ones I saw on Upper Big Tujunga Canyon Road (which he did not see), and the plants which he and Jane surveyed along the western part of the Angeles Crest Highway a few days ago (including the ones I saw near the junction of ACH and Upper Big Tujunga, referred to in the 30-31 August report). Apparently, consimilis and mohavensis are much more similar than we had thought, and this brings into question many of the consimilis identifications made thus far. He is currently at work trying to develop characters which will more easily separate these two sspp.

However, it's beginning to look like mohavensis is coming into the range from the NW (which makes sense if that's where Steve Boyd has it as its main range), coming up Upper Big Tujunga and spreading along that part of the ACH. Hololeucus is on the desert margins of the range and starting a bit up the slopes, east, north and AFH (and the NW part of ANF), but not (except for a few exceptions) making it up to the ACH. Bernardinus is in the 5500-9000' range of the eastern end of the ACH (I saw it on top of Throop Peak) and doesn't seem to spread far from there. And consimilis is primarily in the eastern half of the ACH, dropping down in places on the north side (Big Rock Creek), and definitely on the coastal side (Mt. Baldy Rd., SR39).

But what this means is that I need to confirm SR39 specimens as consimilis and make sure there are no bernardinus or mohavensis mixed in there, which if Tom is correct that consimilis has longer leaves and shorter involucres, while mohavensis has shorter leaves and longer involucres, may not be terribly difficult, especially if numerous individuals are surveyed. Then I am going to survey Mt. Gleason Road and the road from Mill Creek Summit to Three Points on the ACH.

Tuesday, 7 September 2004 (SR39, San Gabriel Mts)

When Tom and Jane began their survey of rabbitbrushes in the San Gabriels, they initially felt relatively confident about identifying three of the sspp. of Chrysothamnus nauseosus, specifically bernardinus, hololeucus and mohavensis. Then consimilis was found in the Mt. Baldy area by Tom and it was realized that much of the rabbitbrush along the Angeles Crest Highway was that ssp. also. It turned out to be fairly easy to differentiate consimilis and bernardinus, two sspp. that are often interspersed, but it meant we had to go back and look at all those again. Then we realized that we had been confusing the green form of hololeucus with mohavensis in the Lone Pine Canyon area, but because of the color of the leaves and the hairiness of the involucres, plus the fact that in bloom mohavensis has spreading corolla lobes while hololeucus has erect corolla lobes, and the fact that there are usually white forms of hololeucus present with the green forms, it is possible to differentiate these sspp. quite easily.

Now the biggest problem is to differentiate consimilis and mohavensis, which are very similar looking. They are only separated in the Jepson Manual by the following characters: consimilis, involucres 6-10 mm, stems generally leafy; and mohavensis, involucres 8.5-12 mm, stems often leafless. Since many of the involucres on both ssp. are in the range of 8-9 mm, this is not a helpful discriminant, and the use of "generally" and "often" in describing the leafiness of the stems makes this a not very useful character either. However, in the species description, it gives 2-6 cm as the length of consimilis leaves and 1-3 cm as that for mohavensis, if indeed it has any leaves. Munz's Flora of Southern California also gives the following discriminants: consimilis, involucres not sharply angled, phyllaries slightly if at all keeled; and mohavensis, involucres sharply angled, phyllaries strongly keeled. The length of the leaves in particular seems to be something that can be easily seen, and individuals with leaves of 4-6 cm fall nicely into the consimilis ssp. The percentage of live as opposed to dead or dried leaves also is a clue, thus an individual with 50-100% of lives leaves would seem to be consimilis. A further characteristic which has recently been noted but needs to be further investigated is that the phyllaries of consimilis seem to be two-tone with a narrow yellow-green midrib with strips of white on both sides, whereas those of mohavensis seem to be a solid one-tone yellow. Thus it is beginning to look as though if one finds a specimen with lots of live leaves at least some of which are more than 4 cm long, and the involucre has two-tone basically unkeeled phyllaries, and the involucres are not more than 9 or 10 mm long, it is probably consimilis, and if one finds a specimen that has few or no live leaves (or no leaves at all), all of which are 3 cm or less in length, and the phyllaries are solid yellow, sharply angled and keeled, and not less than about 8 mm, it is probably mohavensis.

Today I drove up San Gabriel Canyon Road (SR39) again to look at the rabbitbrushes with an eye to checking out our new discriminants. All of the individuals I looked at fit into the consimilis category in all respects, which is what I expected (although one has to guard against only seeing what one expects to see). Due to the extreme heat, I was not able to go any further than I did on Friday, so will have to examine the plants on the upper part of the closed section of SR39 near Islip Saddle another day. I know there is both bernardinus and consimilis there, and I want to verify that those are the only sspp. there or to find out if mohavensis is there too.

On my way home, I decided to drive over Glendora Ridge Road to Baldy Village and survey that route for rabbitbrushes. Again, I would have been extremely surprised to find mohavensis there, and in fact all the specimens which I checked seemed to fall into the consimilis ssp. in each of the above discussed characters. However the length of the involucres is not something that can be eyeballed, and when I got my samples home and measured them, there appeared to be a significant possibility that they were either a mixed population or that they were all mohavensis.

Tomorrow I plan to go back up to the western part of the Angeles Crest Highway, Upper Big Tujunga Canyon Road, and the Angeles Forest Highway to look at some of what Tom now believes to be mohavensis, to compare it with the consimilis according to our new means of differentiating the two ssp., and then survey Mt. Gleason Road.

Wednesday, 8 September 2004 (Mt. Gleason Rd., San Gabriel Mts)

For me this rabbitbrush project has been a process of repeatedly thinking that we were making progress only to have more confusion ensue, that is, two steps forward and one back. And today was no exception. I set off this morning with the intention of reexamining some of the plants along the western section of the Angeles Crest Highway, Upper Big Tujunga Canyon Road, and the Angeles Forest Highway, and to do a brief survey along Mt. Gleason Road, which has not been looked at so far. Remember that Tom Chester had concluded that all of these specimens which we had thought were ssp. consimilis at one time are actually ssp. mohavensis. My plan was to look at plants in these areas to see if the disciminants I discussed in yesterday's report are ones that can be used to fairly easily separate these two taxa.

The very first individuals I looked were leafy (i.e. had a fair number of leaves), had 50-60% live leaves, at least some leaves more than 5 cm long (although most were less than 3 cm), and those two-tone involucres we had noticed in the Mt. Baldy plants. Tom had said these were mohavensis, but they looked like consimilis to me. As I stopped to see several plants on the Angles Crest Highway, then drove down Upper Big Tujunga Canyon Road and up the Angles Forest Highway, I tried to compare these specimens with the consimilis farther east, but they all seemed to look pretty much alike, that is, there did not appear to be much correlation between the characters I was looking at and the two taxa; there just was a lot of variability across all the plants I was observing. I did find one individual of a hololeucus just beginning to bloom on Upper Big Tujunga Canyon, and then I drove up to Mill Creek summit and turned onto Mt. Gleason Rd. It was dry and rabbitbrush-free for the first few miles, then they started to appear in large numbers, and I stopped several times to look at them.

I knew from the floras that mohavensis is supposed to have a somewhat longer involucre (Munz says 6-9 mm for consimilis and 9-10 mm for mohavensis, while the JM gives 6-10 mm and 8.5-12 mm). When I began to measure involucres in excess of 11 mm, I was pretty sure I was dealing with the Mojave rubber rabbitbrush. (Previously, I had measured some involucres on the Angeles Crest Highway and this was what had confused me because it was pretty much the only characteristic that pointed to mohavensis when everything else pointed to consimilis.) But except for finding what seemed to be a large group of Mojave rubber rabbitbrushes on Mt. Gleason Road, I didn't feel like I had accomplished much today to advance our cause.

Thursday, 9 September 2004 (Upper section of SR39, San Gabriel Mts)

Since I didn't get as far up the closed section of SR39 on 3 September, I decided to drive up to Islip Saddle and walk down as early in the morning as possible to beat the current heat wave we are experiencing. Thunder-
storms were predicted for today but it was pleasantly cool when I arrived there about 9:30 and started down the road. I had two goals, first to establish if all the non-bernardinus plants on this upper section of SR39 are actually ssp. consimilis, and second to see how far down the road I might find examples of ssp. bernardinus, since it is to be found at Islip Saddle and could be expected to extend down the road.

I stopped every quarter mile and collected a sample, recording a GPS position for each. I also noted what else was blooming along the road, and that included rock goldenbush (Ericameria cuneata), several species of buckwheats (Eriogonum wrightii, E. parishii, E. saxatile, E. fasciculatum var. foliolosum, and E. nudum var. pauciflorum), Parish's oxytheca (Oxytheca parishii), California brickellbush (Brickellia californica), twiggy wreathplant (Stephanomeria virgata), prickly poppy (Argemone munita), blazing stars (Mentzelia laevicaulis), California fuchsia (Epilobium canum ssp. latifolium), and a few California-asters (Lessingia filaginifolia). There were also a couple of seepy areas with California goldenrod (Solidago californica) and lots of scarlet monkeyflower (Mimulus cardinalis).

I recorded locations for the several San Bernardino rubber rabbitbrushes I saw, the last of which was 1.3 miles down the road. By this time it had clouded over significantly, and when I was about 2-3/4 miles down the road, it began to rain. I guessed that it was another 3/4 of a mile to a mile further to the point I had reached on Friday and I decided that I had a sufficient number of samples to measure. The wind began to rise, it thundered over-
head, and I put my raincoat on and packed away my GPS and notebook. The rain never came down hard, but it did continue until I got back to the car. At least it was not 103° which was predicted for the valleys.

Later, when I measured my samples, I found that just by going on involucre length, I had probably verified the presence of both consimilis and mohavensis, some involucres being 10-11 mm and some being 7-8. At Shortcut Saddle I met Tom, who had come up to do some more sampling along the section of the Angeles Crest from there to Chilao. Before going to bed, I got an e-mail from him stating his conviction based on his latest analyses that the majority of non-bernardinus rabbitbrushes along the Crest are ssp. mohavensis.  So this is our second major change of identity for these plants, first from bernardinus to consimilis, and now from consimilis to mohavensis. But this explains a lot, and it explains largely why we have been so frustrated trying to separate out these subspecies, because we have been dealing for the most part with only a single taxon. And we can console ourselves with the knowledge that we now have a pretty good idea of where these plants live.

The key here is that none of the subspecies have exclusive ranges, and very few localities only have a single taxon within it. The usual situation is for two or more subspecies to be present in a given area, or at least close by. And given the similarity between some of these taxa (even the Jepson Manual comments on how they are often confused with each other), the range of variability within each taxon and the fact that there is more than one form for most of the subspecies, and the further fact that they can intergrade between them, the task of identifying them can become a daunting one. Now that most of what we have heretofore been thinking of as ssp. consimilis has been reassigned to ssp. mohavensis, I really want to go back and see ssp. consimilis again, and the one locality where I am sure I can do that is the Mt. Baldy area, so before leaving this project I will return there tomorrow.

Friday, 10 September 2004 (Mt. Baldy Road, San Gabriel Mts)

I think we are nearing the end of our rabbitbrush survey at least in the San Gabriels. It has been an epic project, but we now have a pretty good sense of where things are. One important realization we have made as noted in yesterday's report is that none of the subspecies have unique or exclusive ranges, and there are few localities that do not have two or more of the subspecies in residence or at least nearby. There appears to be a boundary of sorts near Three Points/Santa Clara Divide Road, with no consimilis and only mohavensis to the west, and both subspecies continuing east. Mohavensis is also present in Upper Big Tujunga Canyon, around Mill Creek Summit, and along Mt. Gleason Road. Both of the taxa are present on SR39, but only consimilis in the Mt. Baldy area. Hololeucus is present in Lone Pine Canyon, along Big Rock Creek and Angeles Forest Highway, and bernardinus seems to share most of the range of consimilis. There are still areas left to be surveyed, such as the Liebre Mts and Lytle Creek, and the road from Mill Creek to Three Points. Also the area east of Dawson Saddle down through Wrightwood needs to be looked at again because we have mostly consimilis there and there may in fact be mohavensis.

I should state at this point that despite Tom's valiant efforts to document characters which would show that mohavensis and consimilis are two separate taxa, I continue to feel uncomfortable with this position. I am not the one who is doing the rigorous analysis that Tom is doing, but it has always been my opinion that whenever you have two subspecies that are so difficult to separate, each of which is possessed of so many characteristics that either through overlap or variability make them only marginally different, any attempt to say that they are unique taxa strikes me as an arbitrary or artificial one. In situations like this, I believe that botanists have found populations that exhibit one extreme end of a spectrum in one place and populations that exhibit the other end of the spectrum in another place, and on this basis have concluded that they were dealing with two separate taxa. If they had conducted a more far-ranging and inclusive study of the entire range of both subspecies, they would perhaps have realized that what they had was a single taxon with a greater range of variability. But again, I am not a professional botanist, and do not feel qualified to state this as a certainty. It is merely my impression, and one which Tom may yet be able to demonstrate as false.

Having only a short time available this morning, I drove up to the Mt. Baldy ski-lift parking area, where I collected samples from a half-dozen plants in order to measure the involucres. Because of our confusion about mohavensis and consimilis along the Angeles Crest, we were in effect struggling to separate some groups of mohavensis from others, and so it was reassuring to find once again the true consimilis. They are indeed leafy, with typical leaves in the 3-4 cm range, but as we have discovered mohavensis can be quite leafy as well, so that is not a particularly useful discriminant. This is an example of two taxa which really can't be told apart just be looking at them, and can only be separated based on the length of the involucre, and you have to have a sufficient number of measurements to establish a range.

Later I took 110 measurements from five plants under the microscope, and the median involucre lengths were 7.0, 7.35, 6.19, 6.77, and 7.2 (in mm). This is a substantial difference from the involucre lengths of 9 mm or more that are typical of mohavensis. It appears that mohavensis is by far the more common subspecies, with consimilis being occasional. Next week I will survey some of our remaining areas, and then we need to take a look at the San Bernardinos, not only for C. nauseosus, but also for some of the other rabbitbrush species like parryi, viscidiflorus and teretifolius.

Thursday, 16 September 2004 (Lone Pine Canyon, Big Rock Creek)

I had a couple of goals for today's driving excursion around the eastern end and the back side of the San Gabriels. I wanted to go up into the Lytle Creek area which I had never been into before, I wanted to see if the Chrysothamnus nauseosus ssp. hololeucus was blooming yet along Lone Pine Canyon Road, and I wanted to photograph the southern goldenrod (Solidago confinis) that Tom had told me about at a location along Big Rock Creek.

I drove up the I-15 and turned onto Lytle Creek Road following the Lytle Creek drainage up the canyon, and soon passed the ranger station and a little community that reminded me of Mt. Baldy Village. The terrain and habitats that I was traversing did not seem to me to be rabbitbrush territory and in fact I didn't see any. After the road became dirt-surfaced, I was confronted by a sign that said "road construction ahead, possible delays of 1 hour," so I turned around. A map I had picked up at the ranger station showed a road that might have taken me over to Lone Pine Canyon Road, but I decided to play it safe and return to the interstate. And even though I hadn't seen any rabbitbrush, I was glad to have done this because I love exploring new areas. Now when someone says Lytle Creek, I'll have some idea of where it's located and what kind of place it is.

I was soon heading along Lone Pine Canyon Road wondering whether two weeks would make any difference in the blooming status of the hololeucus. At first all I saw was greenish shrubs still in bud, and a lot of yellow sticky lessingia (Lessingia glandulifera) almost covering the ground in places, but then there was a flash of yellow ahead and when I stopped to examine it, it turned out to be a blooming ghostly rubber rabbitbrush. I stopped a half dozen times before reaching Wrightwood, finding a few more hololeucus in bloom and a single mohavensis also blooming farther down Lone Pine Canyon than I had seen any previously. Since a lot of the mohavensis or consimilis (whichever you want to call it) near Wrightwood was past its prime bloom, and the hololeucus was only just beginning, I formed the opinion that there is only a small amount of overlap in the blooming periods of the two subspecies.

I drove through Wrightwood and on down the Big Pines Highway, turning left onto Big Rock Creek Road and proceeding along the creek. It's irritating that all along there, there are no parking signs, so that it's really difficult to find a place to stop. Fortunately I had Tom's GPS location of the Solidago, so I was able to pull over just long enough to get down near the creek where it was growing. It gets a lot taller than the goldenrod I'm used to seeing. Having finished what I wanted to do, I called it a day and headed home via Valyermo Road and the 138.

Addendum 9/29/04: The upper portions of the Lytle Creek drainage are probably closed now due to the ANF closure that was effective Monday, 9/27, so I probably won't get back up in there this year. I have a feeling that a 4-wheel-drive vehicle would be useful in that area anyway.

Monday, 20 September 2004 (Lower Hondo Canyon, Santa Monica Mountains)

I have been eager to identify a species of mint which I have seen a number of times along a seasonal creekbed near the trailhead for the Hondo Canyon section of the Backbone Trail, which goes from Old Topanga Canyon Road up to Stunt Road. I was with Tom Chester in the spring when we saw it and since it was not blooming, we were unable to put a name on it. We thought at the time, and I was very much hoping that this would be the case, that it was field mint (Mentha arvensis), our only native Mentha, which I have yet to observe. When we saw the plant in May, the leaves were clearly pubescent which seemed to rule out M. spicata or spearmint, but when I looked at them today, they had only a few small hairs on the veins of the underside. But again, spicata was ruled out by the fairly long petioles.

Perhaps later in the season more inflorescences will develop (in fact, I probably should go back in a few weeks to check this), but as of this time each flowering stem only had a single, small, ball-like inflorescence, so the key relating to the internodes between upper inflorescences was not possible to use. However, the stamens were included which would seem to rule out both M. arvensis and M. pulegium (which is also ruled out by the subsessile cauline leaves), thus I was left with the choice of Mentha Xpiperita (peppermint), which is the only one with leaves with substantial petioles. Jane Strong looked at my pictures and suggested Mentha citrata (bergamot mint) which she has in her yard and which look very much like this. The Jepson Manual has citrata as a synonym for Xpiperita, so I am unclear whether it is in fact a distinct species. L.H. Bailey's Manual of Cultivated Plants lists both piperita and citrata as pure species, with the differentiation being on the shape of the leaves, piperita having lanceolate or ovate-lanceolate acute-tipped leaves and citrata having mostly ovate or elliptic obtuse-tipped leaves. This is an old source (1949), and on this basis I would have to go with citrata, but citrata may in fact not be distinct from Xpiperita. You can see my pictures of it here.

Addendum 10/6/04: Thanks to discussions with Tom Chester regarding the meaning of 'inflorescence' as it relates to this species, I now understand better how to key this species. What appeared to me to be a single ball-like flowering head or inflorescence is actually a series of several inflorescences at nodes that are only separated by a few mm. That definitely keys to 1' in the Jepson Mentha key, which says "internodes between upper inflorescences gen inconspicuous, gen < 6 mm."  But then M. citrata is not differentiated from Xpiperita in the Jepson key, and online pictures of peppermint do not look much like this. So I have decided to go with citrata as a variety or subspecies of Xpiperita with the name bergamot mint. Clearly they are very closely related.  Jay Sullivan has just told me that he has seen a similar plant at several stream crossings in Escondido Canyon in the Santa Monica Mountains, and I intend to go there to examine those plants to see if they are the same. This species is not listed in Raven, Thompson and Prigge's Flora of the Santa Monicas, or in Wishner's more recent update, possibly because it is a recent introduction.

Tuesday, 21 September 2004 (Blue Sky Meadows, San Bernardino Mts)

Tom Chester and Jane Strong recently visited RT Hawke, owner of a property at about 8000' elevation south of Onyx Summit in the San Bernardinos, and Tom reported to me that they had seen sticky-leaved rabbitbrush and autumn dwarf gentian there recently, so I decided to drive up there this morning and meet RT. Although the temperatures down below were predicted to be in the 90's today, it had been 42° this morning up on the hill, and was still quite chilly when I arrived in my shorts.

I had stopped in the Barton Flats area on my way up to RT's place to see again the diminutive Parry's rabbit-
brush (Chrysothamnus parryi ssp. asper) which I had photographed on a quick visit there a few days ago. This is a species that could be easily overlooked, especially since it was in amongst large Chrysothamnus shrubs scattered under the pines of the Flats, and even if seen might be taken at first for baby rabbitbrushes. However, on closer inspection, the stems are white-tomentose, and the involucres are much longer than those on the consimilis/mohavensis complex (I measured one at 15 mm!). The phyllaries are somewhat membran-
ous with green tips, and are only weakly ranked. Also the lower ones are often spreading. Once you gain a mental picture for what this species looks like, you suddenly realize that they are all around and you just hadn't noticed them. In the same area were quite a few Parish's buckwheats (Eriogonum parishii) with their delicate slender stems and tiny flowers.

It had become a beautiful cool and sunny day when I arrived at the Blue Sky Meadow Science Institute, RT's mountain home otherwise known as the Los Angeles County Outdoor Science School, where he has been teaching for the past 17 years. Actually he's only been at this location for a couple of years, before which he was at a location in Wrightwood in the San Gabriels. I had met RT on two previous botanical field trips in the East Mojave and the Big Bear area, and we had spent an enjoyable day in the burn area north of Lake Arrow-
head in June, and I had developed a considerable respect for his knowledge of local flora. We set out and he soon led me to an area where there were some obvious blooming rabbitbrushes. It took only a quick glance to confirm that these were in fact sticky-leaved rabbitbrushes (Chrysothamnus viscidiflorus) because their leaves are so distinctive, each looking as though someone had taken them by the tip and given them a twist. To someone who has seen as many rabbitbrushes over the past month or so as I have, these clearly looked very different. The glabrous stems revealed their subspecies as viscidiflorus, in contrast to the other subspecies in the San Bernardinos, puberulus, which has as its name implies puberulent stems.

Our second destination was a little seasonal pond ringed with cat-tails and the seepy area above it where RT had seen the Gentianella and Sisyrinchium idahoense, a wet meadow montane species that I was unfamiliar with. Autumn dwarf gentian (Gentianella amarella ssp. acuta) is a tiny plant and another one that is easily overlooked. We did find at least a dozen of them, but the flowers on most were closed. The best I could do was to photograph a couple that had flowers partly open. Normally the petals spread well apart, but at least I was able to see the hairy fringe on the insides of the lobes. The Sisyrinchium had apparently disappeared as we were unable to find any of that, not surprising since it is only supposed to bloom through August.

RT mentioned that there was some dwarf four o'clock (Mirabilis pumila) on a high ridge above his meadow, and I just had barely enough time to make it up there before getting back to Pasadena to pick up my daughter from school, so we headed off through a pinyon-juniper woodland, passing some blooming sapphire woolstar (Eriastrum sapphirinum) and strangely a single woolly marigold (Baileya pleniradiata) well above its normal range, and up a rocky wash to an open clearing on top of the ridge, somewhat over 9000' in elevation. RT soon found the Mirabilis and I was able to photograph a few open blooms. Since I didn't take a sample of it, I wasn't able to appreciate what separated this from four o'clocks that I had seen before, but according to the Jepson Manual the involucre of pumila in fruit is brown, papery and enlarged, whereas in californica and multiflora it is basically unchanged. RT pointed out several dark green clumps of ferns scattered among the rocks and we agreed that it looked like a Pellaea similar to mucronata because it did have tiny mucronate tips on the ends of the leaf segments, but not being sure which species of that genus inhabits the upper elevations of the San Bernardinos, I could not put a species on it. Later, I looked in my San Bernardino flora and eliminated andromedifolia based on altitude and breweri based on its being 1-pinnate, leaving only the two variants of mucronata, of which var. californica fit better due to the elevation and not having any ternate leaf divisions. Its common name is California cliff brake.

All in all, it was a lovely though fairly rushed outing, and I just made it back to Pasadena with a few minutes to spare, and I thank RT Hawke for his time and for introducing me to a few new species.

Friday, 24 September 2004 (Cheeseboro Canyon, Simi Hills)

Today's excursion to the Cheeseboro Canyon area was mainly in search of a shrub I thought I had seen there with my friend Richard Sapiro in May of 2000. We hiked the Cheeseboro Canyon/Palo Comado loop during the spring after a burn had swept through the canyon, pretty well denuding the area of most of its vegetation. We saw a grayish-white tomentose shrub that was not in bloom and that we identified at the time as Tetradymia comosa. The Tetradymias are alternately called horsebrushes or cottonthorns, and I dubbed comosa tufted horsebrush because I couldn't find any other common name for it and 'comosa' means tufted. Since then I have begun calling it cottonthorn, although that is a name also applied to Tetradymia spinosa which I call spiny horsebrush. I have not seen T. comosa anywhere else since then and consequently was not at all familiar with its appearance. Something made me think of it recently, and I read in Munz's Flora of Southern California that its blooming period was August to September, so since I had no photographs of its flowers, I decided to go out there and see if I could get lucky and find it again.

Cheeseboro Canyon (often misspelled as Chesebro because the road leading to it for some reason is spelled that way) is a very dry canyon in the Simi Hills just north of the 101 freeway and a few miles west of Las Virgenes/Malibu Canyon Rd. It is a favorite destination for bikers who ride up to Shephards' Flat and Simi Peak. The first mile or so is basically a flat walk through an area that has been pretty much taken over by non-native grasses and other introduced plants. I was pleased to see this time that a large section had been cleared by burning and was being used as an experimental plot to test the feasability of reseeding with native needlegrasses. There was some blooming sawtooth goldenbush and California aster, and lots of doveweed, western jimsonweed (Datura wrightii), and mediterranean mustard, and the hillsides were sprinkled with valley oaks. I have also seen lots of black mustard in there, but that did not seem to be in evidence today. In any year that sees normal rainfall, Cheeseboro is a beautiful place, with blooming gilias, fiddlenecks, snapdragons, poppies, suncups, redmaids, goldenstars, owlsclover, pincushions, clarkias, chinese houses, popcorn flowers, bluedicks, larkspurs and many other things. Richard and I counted 166 species on our 2000 hike with 111 of them in bloom!

After about a mile, you start going through intermittent patches of oak woodlands which was welcome as it was a fairly hot day. Another mile (about 3.3 from the start) brings you to what is called Sulpher Springs where the smell of rotten eggs can often be detected from some seeps in a dry streambed. If the smell isn't too bad, and it's usually not, this is a good place to stop for some lunch, depending on how far you're going. Just beyond this point, the trail enters an open valley strewn with boulders and now covered with chamise, most of which had been burned off the last time I was there. Around 4 miles from the start is where I had seen the Tetradymia, and my only picture of it from 2000 shows an ashy-white shrub. I was also picturing a fairly large shrub, so when I passed by the area where I was pretty sure I had seen it and kept going for another 1/2 mile without finding it, I had about decided that it wasn't there anymore. The chamise had grown to almost head height in many places and could easily have obscured smaller shrubs.

I turned around and headed back more slowly, looking carefully on the side of the trail where I had seen it. About thirty feet from the trail I spotted a small darkish-colored shrub that I couldn't immediately recognize, and I tramped over to it. It was either dead or completely dried up, but for some reason it looked a bit like what I remembered of the horsebrush. I looked around and saw a few others. It was not spiny, but the dried leaves looked like they might have been almost spine-tipped once. One of these shrubs had some new growth starting at the base and it was the color of and had the felty tomentum of Tetradymia comosa. Still, it bothered me that no spines were in evidence, and I returned to the car unconvinced that I had ever seen T. comosa at all. And on the way back, I saw many more of these apparently dead shrubs that I had passed without noticing before.

Also on the way back I recorded blooming Stephanomeria virgata, Asclepias fascicularis, Eriogonum both elongatum and fasciculatum, Salsola tragus in great abundance, Malacothrix saxatilis, Malcothamnus fasciculatus, Ericameria pinifolia, Baccharis pilularis and B. salicifolia, and some viciously-fruited Tribulus terrestris. When I got home I noted in the Jepson Manual that Tetradymia comosa is "± spiny or not," so that made me feel a bit better about my identification, but still I wondered why it was almost all dead or apparently so. I decided to consult my all-knowing friends Tom and Jane. Tom said he had never seen this species so couldn't help, and Jane told me of a location near JPL where she had seen it with Bob Muns and a Sierra Club group this past spring, so I will try to go there sometime this weekend to see if it's still blooming there, and to compare it to the ones I saw in Cheeseboro Canyon. Most likely I will have to wait until next year for its flowers.

Thanks to Jay Sullivan for straightening out my spelling of Chesebro and Cheeseboro!

Sunday, 26 September 2004 (Hahamongna Watershed Park, Altadena)

I had some time this afternoon so I slipped over to Hahamongna Watershed Park which is adjacent to JPL on the western side of Altadena. This is where Jane had told me she saw many Tetradymia comosa plants this past June with Bob Muns. Since I was pretty much unsure of the identity of the Tetradymia I had seen on Friday in Chesebro Canyon, I welcomed the opportunity to see some whose identity was fairly certain. Jane gave me very specific directions to the Tetradymia and I found it within minutes after entering the park. I was pleased to see that it looked very much like my memory of the one at Chesebro, at least the one I had seen in 2000. Unfortunately it was not blooming, but was beautifully in seed so I was able to photograph that. And since the Calphotos website didn't have any pictures of this taxon, I was able to at least contribute these.

I will have to return next year to photograph it during its bloom, but now I now a convenient location that is less than a half-hour from my house, and this will be one of my top priorities for next year. Meanwhile, you can see a picture I took of it here.

Monday-Tuesday, 27-28 September 2004 (Western Angeles National Forest)

The Liebre (pronounced Lee-bray) Mts area at the western end of the San Gabriels remained the last part of the Angeles National Forest that had not been covered in our rabbitbrush survey, so I was determined to get an idea of what ssp. of Chrysothamnus nauseosus was prevalent there. I had recently noted the inclusion of C.n. ssp. mohavensis in Carl Wishner's Flora of the Santa Monica Mountains: A Synonymized Checklist, and I was curious as to whether it was considered native there or perhaps was a recent introduction. It did seem unlikely to me that ssp. mohavensis would be anywhere near the Santa Monicas, so I contacted Carl and he very kindly enlisted the assistance of Barry Prigge, one of the authors of Peter Raven's original Flora of the Santa Monica Mountains (1986), in going back to look again at the two UCLA vouchered specimens that this listing was apparently based on. Barry got back to us with his opinion that one of the specimens was definitely ssp. hololeucus, and it appeared that the other was as well.

In communicating about this and in reference to reports of old sightings of Chrysothamnus in areas that are now likely to have been developed, Carl mentioned the Simi Hills and Woolsey Canyon as possible locations. He made several trips to the Gavin Canyon area near the Calgrove exit of I-5 and reported the existence of one large shrub the taxon of which he was uncertain. Subsequently he notified me of what he referred to as the "mother lode" of Chrysothamnus in that area along the 14 freeway and Sierra Highway leading up to the Placerita Canyon vicinity. So I decided to survey both of these areas, and spent part of Monday doing that.

First I checked on the one large shrub Carl had observed on a piece of private property off The Old Road, which parallels I-5 west of the Sierra Highway.  Having seen as much Chrysothamnus as I have recently, I only had to glance at it to be certain in my own mind that it was ssp. hololeucus, in particular what Tom and I have come to refer to as the green form. It was still only in bud with anthesis a few weeks away, but the buds themselves were slightly tomentose and very small, and of course the consimilis/mohavensis plants are already well into their blooming period everywhere we have seen them. I then drove up the Sierra Highway toward Placerita, and in many places both sides of the highway were lined with the same grayish-green non-blooming shrubs that I believe to have been ssp. hololeucus. I will return to this area in a couple of weeks and try to confirm this id by observing the short erect corolla lobes, but in the meantime Carl's use of the term "mother lode" could not have been more apt.

My next goal was the Liebre Mts area. We had seen both mohavensis and hololeucus listed as common species in Steve Boyd's Vascular Flora of the Liebre Mts (1999), and so I did not expect to find any real surprises there. I drove up to Quail Lake, followed Rt 138 toward Lancaster, and turned off on the Old Ridge Route that plunges south through the ANF to its southern terminus at Castaic. This is a seldom driven-on road that is not maintained by the County and therefore in poor condition, but it is both scenic and historic, so I was surprised not to encounter a single other car. I had seen a lot of ssp. hololeucus as I drove up the interstate but as I approached the Gorman area I began seeing quantities of blooming rabbitbrushes, and I suspected that they were mohavensis for the most part. Turning off onto Quail Lake Road, I stopped at the first turnout and immediately found both mohavensis and hololeucus blooming right next to each other, so it was an easy thing to compare them.

From that point on along the 138 and then down the Old Ridge Route, I stopped every half mile or so and sampled the rabbitbrush population, doing some photography, collecting some specimens, and taking GPS readings of the locations of individual shrubs. Even though there were some areas where it seemed that the pre-
dominant ssp. was mohavensis, overall I would have to say that ssp. hololeucus was defnitely the most prevalent. Although there were a very few individuals of hololeucus in bloom, the overwhelming majority of bloomers were ssp. mohavensis.

On Tuesday I returned to this area to discover that the recent Angeles National Forest closure, effective yesterday, had been extended to the Old Ridge Route, and it too was closed. I breathed a sigh of relief that I had managed to survey it on its last day of being open for the indefinite future, or at least until we get some significant rainfall. I turned left on Pine Canyon Road and headed toward Lake Highes Road and Bouquet Canyon, sampling rabbitbrushes along there. There were lots of both hololeucus and mohavensis, and as I had come to suspect, it was the mohavensis that was blooming. After a few miles I entered a burn area and there was no more live rabbitbrush or anything else.  I pulled up into the parking area for the Liebre Mt. trail and sat there gazing around and thinking about the beautiful black oak and buckeye forest that used to clothe that side of the mountain ridge. It was now a scene of utter devastation, caused by the Pine fire this past July, but I knew that one day its natural vegetation would return.

At some point along Pine Canyon Road after I had left the burn area, I stopped to look at a particularly robust and thickly-blooming rabbitbrush, and I saw that although it had spreading corolla lobes, its involucres were noticeably shorter than any of the other ones I had seen. I had run across one of these the day before, and there had been something different about it, but I couldn't put my finger on it, and stupidly did not even notice the thing that set it so characteristically apart. When I got home and looked at it under my microscope, I saw what I had failed to see earlier, and when I described it to Jane Strong that night, back came the immediate response with an identification that fit perfectly with all its physical features. It was sharp-bracted.rubber rabbitbrush (Chrysothamnus nauseosus ssp. ceruminosus), and was distinguished most particularly by its short involucres and by its spreading and often recurved phyllaries. As I drove along Pine Canyon Road beyond the burn area, most particular from White Oaks Ranch to Lake Hughes Road, I stopped frequently and found many of these rabbitbrushes, which I had never seen before, and which for some reason Steve Boyd did not have listed in his Liebre Mts flora. I have contacted him and will be interested to hear his response.

I was particularly excited to make this find because this is the fifth subspecies of Chrysothamnus nauseosus that we have now in our survey of the Angeles National Forest (assuming that mohavensis and consimilis are to be considered as separate sspp.), and we can now add this to our distribution plot.