EVOLUTION OF THE CACTACEAE

As in nearly all fields of science, it seems we DON'T know more than we DO know, but if you will forgive the brevity of discussion for each of these enormous topics, I'll try to give a summary of what we do know about cactus evolution in general. Readers are advised to examine pertinent references such as Gibson and Nobel's (1986) ^Cactus Primer^ and Benson's (1982) Cacti of the United States and Canada, as well as other resources such as Cullmann, Goetz & Groener, "The Encyclopedia of Cacti", all of which have some discussion of cactus evolution. None of these sources are complete with respect to a "total assessment" of current data, but Gibson & Nobel do an admirable job of covering a very complicated topic. Given the above caveat, I'll try to give some "highlights" of cactus evolution, with respect to the systematic (phylogenetic) study of the family, and relevant biology when appropriate.

FOSSIL CACTI - Since the major processes of fossilization generally require sedimentation of mineral materials over vegetative/floral materials, you can realize very quickly that with even a hypothetical [xerophytic] ancestor to the cacti (perhaps Pereskia-like in appearance), the environmental factors that favor fossil preservation (save for amber entombment) would have been an unlikely combination of cacti and ample water for sedimentation. Thus the fossil record for cacti is poor, or in fact absent. In 1944, a presumed fossil cactus ^Eopuntia douglasii^ was described from Eocene deposits in Utah. There was much controversy over whether this compression fossil was indeed a cactus, and after significant bantering about the literature, most people do not believe this "fossil" is indeed a cactus. For literature citations see Benson (1982), pages 76-79. Except for relatively recent pack rat middens (containing a variety of cactus seed) our knowledge of cacti from any fossil remains continues to be depauperate at best. Scientists today must rely on the morphology of present day cacti, or other sources of data (biochemical, chromosomal, or in my case nucleic acids [DNA]). Through the study of these sources of evolutionary information, we can make some general conclusions about cactus evolution.

CACTUS RELATIVES:

From some~ very recent studies of DNA variation and from vascular anatomy, the closest angiosperm family to the cacti is the Portulacaceae, with the genera Portulaca, Talinum, and Anacampseros the likely "cousins" (among others). The long standing previous assumptions that the Aizoaceae (including the mesembs) was the sister family to the cacti (due to the floral hypanthium) was proved fallacious in independent tests of phylogeny for the order (Caryophyllales). We are ~continuing to amass various sorts of data which continue to corroborate this hypothesis.

EVOLUTION WITHIN THE CACTI

ASSUMING that from some common ancestor between the Portulacaceae and the "proto-cacti", a xerophytic lineage arose which was capable of radiating and speciating in the "New World", most likely after the split up of Gondwanaland. The "exact" center of origin for the cacti has been disputed for many years. two prevailing "centers" of likely origin persist: The first is within the Caribbean Islands (proposed by Buxbaum); and the other is in north [western] South America; I tend to support the latter hypothesis, and hopefully my studies of chloroplast DNA variation will help provide data to answer this question. Regardless of the geographic center of origin, the result we see today is a very diverse and widespread New World family of angiosperms, having approximately 1,500-2,200 species and about 100 genera [conservatively]. It is the second largest family of angiosperms restricted to the New World (the first being the Bromeliaceae), and has geographic limits extending from central Canada to Patagonia, and virtually every habitat in-between.

Divergence of the Major Groups

Students of of the cacti can generally recognize three major lineages within the family, based upon gross morphology, flower and fruit structures, as well as other characters. These three groups, treated previously as tribes, are now considered to be subfamilies, and have been shown to be monophyletic (of one common ancestral lineage). The three subfamilies are briefly discussed below:

SUBFAMILY PRERESKIOIDEAE

The smallest subfamily of the Cactaceae has 18 species; all having persistent leaves and large shiny black seeds. The two genera, ^Pereskia^ (16 spp) and ^Maihuenia^ (2 spp.) are found predominantly in South America. Pereskia plants are usually shrubby, tree-like, or in one species (P. aculeata) form vines. The flowers can be very showy (i.e P. grandifolia, P. bleo) or can be diminutive (i.e. P. humboldtii, P. weberiana, P. diaz-romeroana). The genus Maihuenia is found only in relatively high elevation habitats of Andean Argentina and Chile; these are low growing mat or cushion-forming plants, and are strikingly different than their sister genus Pereskia in vegetative form. Maihuenia was originally placed next to (or in some cases within) the genus Opuntia. Both of these genera of the subfamily Pereskioideae are on an independent evolutionary lineage from the other cacti, and have not been as "successful" as the other two subfamilies in terms of speciation and geographic range. They are likely a relictual group, PERHAPS similar (but not the same) as the "proto-Cactaceae" ancestor, and have differentiated into at least four or five geographically isolated groups within South America and the Caribbean.

SUBFAMILY OPUNTIOIDEAE

The monophyly of this subfamily can be clearly (and painfully) demonstrated with the uniquely derived feature of the opuntioid areole... the presence of small bristles called glochids are unique derived feature of the opuntioids, obvious to anyone~who~has~touched one of these plants. Additionally, the seeds of members of this family (originally described somewhat inaccurately as arillate) have a funicular envelope and accessory tissues (which become stone-like in many taxa) which are shared by virtually all members of the subfamily. There are about 300 species or so (more or less) within this subfamily. Conservatively, there are five genera [Opuntia with >200 species; Pereskiopsis with ca. 9 spp; Quiabentia with 2-4 spp.; and Tacinga with ca. 2 spp.] The taxonomy of this subfamily is perhaps the worst in the family, and is in desperate need of study. "Splitters" will fragment the genus Opuntia into as many as 20 genera, while "lumpers" will take a broader concept of the genus and transfer the splitting to an infrageneric level. Regardless of its~ generic level classification, this subfamily is clearly the winner when it comes to occupying the most geographic range for may major group of cacti----> coast to coast (E-W) in both N. America and S. America, and from central Canada to Patagonia... It's also successfully invaded (with the help of.....???) southern Africa and Australia where it's become a noxious plant pest, and it's been cultivated in the Mediterranean region for perhaps nearly 500 years (probably brought back by good ol' Chris Columbus). Except for horticultural interest in the other cacti, this subfamily represents a major economic use for the cacti, especially the "prickly pear" group of Opuntia, which provides Nopales and Tunas for human consumption. The major growth forms are the cholla-type, cylindrical stemmed forms (chollas with short-lived ephemeral leaves; Pereskiopsis and Quiabentia with cylindrical stems and persistent succulent leaves; Tacinga with cylindrical stems and hummingbird flowers (!), and Pterocactus (9 spp) from Argentina with unusually winged seeds and a geophyte growth habit. Some very interesting research on Opuntioid evolution is being done by Wolfgang Stuppy (University of Kaiserslautern, Germany) a Ph.D. student (who's actually writing his thesis on Euphorbiaceae!) nearly completed, using the seed internal anatomy as a useful source of phylogenetic morphological information. We need as much information as we can get about this group......Any prospective students want to go to grad school???

SUBFAMILY CACTOIDEAE

Wow! About 86% of the species diversity of the family is found in this subfamily..Probably 99% or more of the cactus cultivated in hobbyists collections are from this subfamily as well. With over 1,000 species, showing extremes of morphological diversity where do I start? Basically, this group shows many of the really confusing problems of parallel evolution within the cacti. Witness, for example, the morphological similarities of the north American "ball" cacti (Mammillaria, Coryphantha, etc. of Tribe Cacteae) and the similarly structured members of the Notocacteae (Parodia, Notocactus s.str., Frailea) or the Tall, columnar members of the Tribe Pachycereeae (saguaro, Lophocereus, Stenocereus, Pachycereus, Myrtillocactus, Polaskia) from Mexico/Central America/USA and the Andean members of Tribes Browngieae and Trichocereeae which have virtually the same vegetative structures. It is no wonder that Britton and Rose, using a purely "pigeon-hole" concept of taxonomy grouped evolutionarily unrelated cacti based upon superficial resemblance. Today most cactologists recognize about eight to ten (presumably) evolutionarily independent lineages within this subfamily which are called TRIBES. It would take many many pages of text to explain the presumed hypotheses of relationships between the tribes of this subfamily, and for many of these the relationships are nothing more than educated speculation. I am attempting to clarify some of these relationships using DNA techniques, but we are several months (years?) away from resolving the problems conclusively. There also is the question of which of certain controversial genera should be placed within what tribe. There are countless unanswered questions still remaining about this subfamily as well. Many of these would warrant their own discussions!!! Basically there are about 4 tribes which we feel evolved in North America (Tribes Cacteae [largest], Pachycereeae, Echinocereeae [incl. Leptocereeae], and Hylocereeae). These have morphological "counterparts" in South America: Tribes Notocacteae, Trichocereeae, Browningieae, Cereeae, and Rhipsalideae. There~are~parallel evolution scenarios hypothesized for barrel cacti, columnar cacti, and epiphytic cacti, each occurring in North and South America. This creates many more phylogenetic and biogeographic hypotheses which are in need of study. The degrees of specialization, biogeographic affinities, and changes in floral/pollinator syndromes preclude further elaboration here; it would take at least a seminar or two to get through the basics of cactus evolution, let alone discuss adequately, how recent data is answering many of these questions.

SUMMARY

I hope I've given a digestible synopsis of major trends in cactus evolution, -- if you're still awake, I'll refer you the previously mentioned references for more reading. Please feel free to discuss various issues - I'll TRY to jump in when I can. The preceding summary was done as part of my displacement behavior which is keeping me from grading ca. 60 term papers, each about 20 pages. (Arrrgh!) Cactus evolution is fascinating, and its result is the manifestation of natural selection and speciation processes which has given us a marvelous array of xerophytes which we cherish and admire in our collections. Please respect the evolutionary process and practice conservation whenever possible; don't buy field collected plants, and propagate as much as possible.

I'd appreciate knowing if this was digestible and/or useful (I don't want to babble-on incessantly if no one can use the information!).

All Best Regards,

(Dr.) Robert S . Wallace Internet: rwallace@iastate.edu Department of Botany Iowa State University Ames, Iowa 50011 USA