Van Reenens Crest

Posted on October 22, 2011 by Bruce Bayer

Photos from a field trip to the farm Van Reenes Crest
(half way between Swellendam & Heidelberg along the N2).
See an earlier article in Haworthia Updates Vol. 7 Chapter 6.

Trevenen & Hesphia Barry
Honey Guide Cabin
P.O. Box 8, Buffeljagsrivier. 6742
084 516 3596

Continue reading →

The Haworthia pollinator

Posted on March 25, 2012 by Bruce Bayer

While I have seen and recorded a Solitary Bee species pollinating the flowers of Haworthia, I have never succeeded in photographing this pollinator. Slightly larger than the ordinary Honey Bee, it is a very rapid and busy flyer and does not spend time at any one flower. Here in Cape Town it is relatively rare and there has been an odd season where I have not seen it at all. It makes a nest consisting of a short tunnel dug into the ground where it makes a series of nectar and pollen filled cells in which the eggs are laid. At Worcester these insects were very common and on one occasion I came across them nesting in large numbers in a small patch of bare gravelly clay. It is remarkable how they were obviously able to recognize and home in on their own small tunnels. It does come into relatively insect proof enclosures and can make isolation of plants for pollination difficult. It is unlikely that flying distance has been measured and it seems very unlikely that it will match the observed maximum of the Honey Bee at 13km (8miles). However, I do not know the length of the life cycle and the relation of the feeding/foraging activity to its nesting behavior. It may be possible that feeding but non-nesting bees disperse over greater distances.

The chance to take the two photographs presented here came quite by chance. I was in my plant house that is partly accessible to insects, and saw the bee on a Haworthia flower stalk. Its jaws were clamped on the peduncle and it was totally stationary. I took a few pictures and then attempted one from another angle. In doing so I had to dislodge a second peduncle from nearly under the bee. The slight difficulty in doing so did not register properly. I took a picture and then suddenly the bee came to life and flew off to resume foraging in the house. But it soon settled on another inflorescence (on a plant at another table) again on the peduncle with its jaws clamped on the talk as before. I removed some pollen coated threads hanging from its back legs and left. While downloading the pictures I saw that I had photographed a spider on the peduncle under the bee in my second angle. So went back to see. The bee was still stationary on the flower and then I saw what I thought was the same spider crawling over the bee. Somehow I disturbed the bee again and it flew off up into the shade cloth in seemingly healthy fashion. The next day I thought I would check the original peduncle to see if there was any web. Much to my surprise I saw the original spider and no web. I have seen these small yellowish spiders on flowers before and also noted small amounts of web. Very occasionally there have been dead flies sitting attached to the flower stalks but not with any noticeable web.

So it is still a bit of a mystery if this small spider, or spider pair, could actually capture and immobilize such a bigger potential prey. Certainly the bee I observed had encountered web and may have grasped the stalk to free its hind legs to work the web off. I did not observe anything more than minimal movement to do so.

Regarding pollination. I was not doing any controlled pollination during which I would have been checking for the presence and exclusion of the bee. But what I did observe was seed set on H. limifolia clones that I could not achieve although in desperation I have transferred pollen between different collections. The bee had an advantage in also bringing pollen from unrelated species. This was on H. limifolia ‘gigantea’ and on H. limifolia ‘glaucophylla’. While I have set a few capsules on the former, I have achieved none on the latter. What is interesting is that no seed was set on field collected clones of H. limifolia presumably ‘keithii’ from Isiteki.

During feeding/foraging, the bee holds onto the lower flower limbs and very briefly inserts its ‘tongue’ into the flower and buzzes away. I could not see any deliberate hovering and pollen transfer to the back legs as does the Honey Bee. But pollen is collected and carried in approximately the same way. When I hand-pollinate I adopt the simple approach of approximating the insertion of hair (bees ‘tongue’) into one flower to obtain pollen and then into another to deposit it. Better results are claimed for physically exposing the stigma and transferring pollen using a brush.

A recent trip to photograph flowers resulted in us finding H. floribunda southeast of the Bontebok Park at Swellendam. I have reported elsewhere that I had seen this species within the park but repeated visits had turned up nothing but similar looking H. mirabilis. On this occasion we were outside the park where we have recorded H. mutica, H. marginata, H. minima, and also similar small forms of H. mirabilis as occur inside the park. We explored a small area we had not covered before and were delighted to find H. floribunda, now easy to see because they were in flower. However, we proceeded to the H. mirabilis locality and on the way again found H. floribunda that we had missed on the previous visit. We also found more H. mirabilis about 60m further along and still about 100m away from the original H. mirabilis locality. These four species were along a stretch of about 5-600m and not occupying shared habitat. The vegetation was fynbos on shallow alluvium, quite stony and well-drained.

It was while photographing the flowers of H. floribunda that a furry fly appeared and ignoring us and the camera, attended to the flowers. While I was trying to get a picture of this fairly fast moving fly, the Anthophorid bee pollinator also appeared but flyng too fast and haphazardly to be pictured. We had seen the same fly on H. mutica at the Buffeljags habitat. The species is Australoechis hirtus known to be a pollinator and visitor to many flower kinds. The flies are nectar feeders and parasitize other insects in their reproductive process.

While we do not know what the pollinating effectiveness of these insects over distance is, it is quite obvious that in this situation there must be substantial transfer of pollen across all the species given the short distances involved. Three of the species were in flower. But H. marginata, curiously, at this locality flowers nearly 5 months earlier than at other known places. Despite that, there are hybrids with H. minima. There did not seem to be hybrids among the H. floribunda plants, but we did think that some of the plants of H. mirabilis could have been hybrid. ♦

Australoechus hirtus on H. mutica. Rotterdam.

Haworthia flowers – some comments as a character source, part 1

Posted on May 12, 2012 by Bruce Bayer

M B Bayer, Kuilsriver, RSA

Introduction

The object of this essay is to discuss where we are now with respect to classification of Haworthia. Despite my comments and observations stretching over 50 years, there are still taxonomists writing and arguing on the basis of method and practice that generated the anarchy of names that existed at the start of my involvement. This method is what is probably referred to as “typological” i.e. there is a single herbarium specimen and anything that departs from this in some mind catching way is a different and thus needs a new name. At the generic level, recent DNA studies show that Haworthia is indeed three separate entities (the subgenera), and that these cannot be rationally separated from the aloid genera. Formal classification requires that Haworthia thus be subsumed in Aloe (see Treutlein et al, Rhamdani et al and Daru et al). This is both incomprehensible and anathema to writers and collectors locked into method that does not rest on any insight to what the problem of species actually is, let alone take proper cognizance of the problems that exists at generic level.

In response to some of the expressed anxiety about the proposed changes to Aloe, Dr John Manning, in a personal communication, wrote…”Most folk follow the biological species concept, which in its simplest form proposes that species are groups of interbreeding populations that are more or less reproductively isolated from other species. The existence of infrequent hybrids does not invalidate it. HOWEVER this is just about always a purely theoretical construct because there is no experimental data in most species to back it up. In practice, therefore, folk use morphology as a surrogate for reproductive isolation on the argument that discontinuities in morphology will coincide with these reproductive boundaries. Naturally there are all sorts of natural variabilities within species and so it becomes tricky to decide which of them fall within the boundary of the species and which fall outside of it”.

Two interesting things arise from this. Firstly, a “species” is itself a biological entity and there are not (or should not) be species that are not biological. Secondly there is a difference between a species concept and a definition. The concept is a general idea, while a definition lays down requirements that must be met. It might be better to say…”species determined by their interbreeding potential”, than use the term “biological species”. There is a huge amount of real experience and observation of the interbreeding capabilities of Haworthia both in the field and in cultivation. This suggests that if we really followed this criterion for determining in the genus then most of the Latin species names are superfluous, as I indeed consider them to be. Therefore, while there is a problem with the formal recognition of genera, the seems to be a total vacuum in respect of considerations around the constitution of species and how they can be, or are to, be circumscribed.

I have of course based what I regard as Haworthia species on a view of species as dynamic systems in a geographic framework, because history and my own long experience suggest that there are no morphological (or biological) criteria that produce the kind of classification that writers have produced or that collectors may need or expect. I proved by my studies in Oxalis, that morphological criteria can be extremely misleading, and that characters (or even character sets) thought to be useful in delimiting species, may in fact vary more within species than between species. I observe that this is true in many other genera. Flowers have recently been touted again by aspiring taxonomists as characters that will solve problems and it has even been said that I have ignored them. This is not true. I do not deny their value as characters. My experience simply suggests that variation and similarities in floral characters, mirror the situation with regard to vegetative morphology and may actually provide evidence for further amalgamation of species and reduction of names. This essay thus explores the situation in a data set, fortuitously available by virtue of modern digital photography and opportunity.

The elemental issue underlying all of this is dependent on the species definition. If some floral character is, or characters are, identified and specified as species characters then of course they would acquire significance in that context. What I can only show is that I observe nothing of the kind to relate to such a definition and insist on as species as systems in a geographic framework. Can they be rationally organized in any other system?

The hypothesis this essay addresses is that the floral morphology can be used to identify and circumscribe taxonomic ranks in Haworthia.

Overview

Originally I very laboriously drew flowers – often only one per collection. But one of my projects was trying to find more consequential objective evidence than just subjective vegetative differences to distinguish between H. reticulata and H. herbacea. I used a range of floral measurements but found the variation so great that to get a statistically significant result would have entailed many more samples than the number I used. Doing this would have been both very labour intensive and very destructive to the populations. I must point out that I am confident I could separate H. arachnoidea, H. mirabilis, H. pubescens as well as H. herbacea and H. reticulata (to some degree), all in the Worcester area, on the basis of their flowers. But this is not the level at which any other solution is needed beyond the appearance of the plants. There are only general differences by which one can separate H. herbacea and H. reticulata in respect of both vegetative and floral characters. H. wittebergensis could probably be identified just on the basis of its small white and more recurved flower, but its vegetatitve characters are quite adequate for the purpose.

Photography in those early days was not an option for recording flower detail. The advent of digital photography changes that quite dramatically and this essay is an exercise using this technology to attempt to document field observations. I photographed flowers as available in approximately 30 populations during February 2012, and these are used now in a posteriori manner to test the hypothesis regarding flowers. These images cover three species as I recognize them as systems in a geographic context viz. H. mirabilis (20 populations from three geographic zones), a summer flowering set of H. mutica (3 populations), H. floribunda (5 populations). Several outlier populations are added although sometime by single clones (see Table 1 listing the accessions. Map 1 (at end) indicates relative geographic position). These photographs provide a pictorial record of variation and difference within populations, between populations of the same species and between populations of different species.

Data

1. The plants. Figure Set 1 is a set of to 4 each, of the plants as they are in the populations covered. I have done this as a reference to how the names were derived and as I use them. The plants are far more variable than is generally understood and four images per population is really grossly inadequate as a sample to base a circumscription on. In the case of 7163 H. mirabilis, Frehse Reserve, S Riversdale, I have included pictures of plants (no flowers) because this is the type locality for the name H. magnifica (=H. mirabilis) that is widely used without any consideration of the range of variants there and how they may relate to other populations of H. mirabilis that are equally variable. Curiously there were no flowers and it appeared that the few plants that had flowered had done so earlier than expected. The same applies to a very similar population 6651 from a little to the southwest where there was only a single open flower. But two other populations viz. 7778 and 7818 in the area had many flowers at the time. It was rather similar at Van Reenens Crest where the various populations had not flowered with equal vigour and at some it was not possible to get a respectable number of flowers. It should be noted that in a good season with some summer rain, the plants will flower more successful with more plants flowering and some plants producing successive inflorescences. Thus the flower season could be extended be several weeks. Note that in the case of 7976 H. floribunda I did not take plant pictures for this second population.

Table 1. List of populations and identifications.

1. Haworthia mirabilis – eastern zone, Riversdale area (cf H. magnifica).
6651 H. mirabilis, E Riversdale.
7163 H. mirabilis, S Riversdale.
7778 H. mirabilis, Komserante, E Riversdale.
7818 H. mirabilis, Windsor, SE Riversdale.
7809 H. mirabilis, Koeisekop, Heidelberg.

2. H. mirabilis – central zone, Swellendam area.
7887 H. mirabilis, Rotterdam.
7912 H. mirabilis, Rietkuil.
7913 H. mirabilis, Rietkuil.
7916 H. mirabilis, Van Reenens Crest.
7917 H. mirabilis, NE Dam VR Crest.
7918 H. mirabilis, Van Reenens Crest, SE Dam.
7919 H. mirabilis, Van Reenens Crest, Game camp.
7955 H. mirabilis, Van Reenens Crest.
7959 H. mirabilis, W Van Reenens Crest.
7960 H. mirabilis, Kruiskloof.

3. H. mirabilis – western zone, Greyton area (cf. vars beukmannii, rubrodentata).
7262 H. mirabilis, S Greyton.
7978 H. mirabilis, Schuitsberg.
7979 H. mirabilis, Nethercourt.
7980 H. mirabilis, E Ouplaas.
7981 H. mirabilis, Schuitsberg N.

4. H. mutica – north-eastern zone, Buffeljags (Swellendam area, cf. H. groenewaldii).
7801 H. mutica, Buffeljags.
7888 H. mutica, Rotterdam 1.
7889 H. mutica, Rotterdam 2.

5. H. floribunda – central zone, Swellendam area.
7774 H. floribunda, S Swellendam.
7910 H. floribunda, Rietkuil.
7963 H. floribunda, Niekerkshek.
7975 H. floribunda, Rotterdam 1.
7976 H. floribunda, Rotterdam.

6. H. mirabilis – outliers and ‘outgroups’.
7248 H. mirabilis, Ballyfar, Infanta.
7513 H. mirabilis, Klipfontein N Bromberg.
7612 H. mirabilis, Diamant W.
7780 H. retusa ‘geraldii’, Komserante.
7781 H. retusa ‘foucheii’, Komserante.
7920 H. retusa ‘ nigra’, Van Reenens Crest.
7974 H. mirabilis, Klipfontein, Potberg.
7982 H. mirabilis, Melkhoutkraal, Goukou.
JDV87-132 H. parksiana, Mossel Bay.