Cactus & Succulent Society
of New Zealand (CSSNZ)

The Evolution of Cacti

In the beginning ...

No fossils of early cacti have ever been found, because for a fossil to form there needs to be ample water for sedimentation or laying down of soil over the plant like a plaster cast - an unlikely occurrence in the case of Cacti in a water scarce environment of the deserts. Also entombment in Ice or Amber is unlikely.

However DNA, molecular data and bio-geography (how geology e.g.. plate tectonics impacts on biology) has given us a few clues.

Relatives of Cacti

From DNA studies it has been found that the closest family to the cacti family is the Portulacaceae ( e.g.. genera Portulaca, Talinum, and Anacampseros) and Didiereacea families which places them in the Caryophyllales, which originated in Africa. Portulacaceae now occur in South Africa, Kerguelen Island and New Zealand - all part of the original Gondwanaland. Didiereaceae live on Madagascar - another part of Gondwanaland.
Which all makes sense as cacti are in the Americas, starting out in South America - another part of Gondwanaland.

Evolution of cacti

In the beginning there was one land mass - Pangaea - which in turn split into Laurisa (north - North America, Europe, and Asia) and Gondwanaland (south - Africa, South America, India, Madagascar, Antarctica, and Australia) about 200-165 MYA (million years ago). About 130 MYA Gondwanaland separated into Africa, Madagascar and South America and these started moving apart. Caryophyllales were in an area of Gondwanaland that split into Africa and South America.

By 65 MYA there was about 800km between the two continents of Africa and South America, which is sufficient to ensure there be no interchange of organisms between the two land masses e.g. migratory birds or floating rafts of flotsam, so evolution of the species on each land mass occurred in different and independent directions.

Because cacti are only found in the new world it is assumed that they only started to appear after the separation of South America from the other continents in Gondwanaland, and after the 65 MYA mark and then moved into North America when these two land masses connected, then evolved separately when the two landmasses of North and South America became separate again when the land bridge drifted away, only to be rejoined with the central America land bridge. In fact from studies it is thought cacti first appeared/evolved 30 MYA.

Caryopyllales were now split into 2 continents (really the Cactaceae-Portulacaceae-Didiereaceae complex were stranded and then diverged on five separate land masses from the original Gondwanaland as mentioned in "relatives of cacti" above) and the south American ones evolved into cacti about 30 MYA.
The "exact" starting point in terms of area for the origin of the first cacti has been disputed for many years, possibles are the Caribbean, southern Mexico, northern South America or along the Andes rain-shadow. When South American connected with North America (twice), cacti moved northwards (and south) so that today cacti extend from central Canada to near the bottom of South America and every place in between.

Between 30 MYA and today the climate changed, glaciers retreated, ocean currents changed - all factors leaving behind arid deserts especially at a belt 30 degrees latitude worldwide. During this time Cactaceae had evolved in a similar way to the African succulents with CAM/C4 metabolism, water storage cells, reduction of leaves and surface area etc., but with a few extra adaptations like areoles. They expanded into this increasingly arid territory, evolving into new species. Isolation e.g. the Baja California peninsula (which was at one stage a separate landmass) and its offshore islands led to new species and varieties evolving in their "own little world".

The Rhipsalis Puzzle

Rhipsalis is the only cacti found naturally outside of the Americas, others e.g. Opuntia have been artificially introduced by man into Africa, Mediterranean, Australia etc in fairly modern times. Rhipsalis grows on the Africa continent (in an area 3500km across in the tropics) as well as Madagascar, Seychelles and Sri Lanka plus neighbouring islands.
It was not introduced there by man, unless some ancient African peoples sailed to America and back, or conversely some ancient American people sailed to Africa AND went to the trouble of bringing some obscure cacti that grows in the trees that neither has a nice flower or any ethnobotanical uses, with them (and didn't bother to record it in their history).
The distance between America and Africa also means it unlikely that it was carried there by migratory birds, also birds tend to migrate North/South according to seasons not in the East/West direction, also introduction by rafting or floating on flotsam in ocean currents from America to Africa is an unlikely scenario.

The other possibility is that cacti had appeared/evolved up to rhipsalis stage BEFORE the splitting up of Gondwanaland and the Rhipsalis that were left stranded in Africa did not evolve further, while it cousins carried away on the South American landmass evolved further in response to changing climates, so the Rhipsalis genus is the ancestor of all cacti.

This scenario put forward by Phil Maxwell is that Pereskia evolves in West Gondwanaland (area of future South America) from the Portulacaceae(?) family, Opuntia form a evolutionary branch (or may even be a separate independent evolution of 'cactus') at an early stage, the Pereskia evolves into a epiphyte - a Rhipsalis - with CAM and no leaves which spreads rapidly throughout West Gondwanaland, the American and the African populations become genetically isolated due to plate movements/oceans and begin to diverge. The African Rhipsalis reach an evolutionary 'cul-de-sac' while the American ones go on to form the cacti genus we know today. Also another evolutionary branch that diverged from Pereskia early on is the Hylocereeae e.g.. Hylocereus, Epiphyllum and Selenicereus according to Phil.

Cacti evolve into 3 subfamilies

Three subfamilies, which had one common ancestor but evolved about 20 MYA are now recognised by (most) botanists as below:(see above "Rhipsalis" above for an alternative evolutionary path).

PERESKIOIDEAE
The smallest (18 species) and most primitive subfamily of the Cactaceae are found predominantly in South America/Mexico. Pereskia plants are usually shrubby, tree-like with leaves.
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 specialisation and distribution.
OPUNTIOIDEAE
This subfamily (300 species) has small bristles called glochids which are unique to opuntioids. It is the most widely distributed from coast to coast and Canada to southern south America, plus with the help of humans now the Mediterranean, Africa and even Australia (where it is considered a pest).
Economically it is important as the "prickly pear" for human consumption, or as a host for the red dye cochineal insect in places from Sicily to Egypt to Mexico to South America.
CACTOIDEAE
more than 99% of the cactus cultivated by hobbyists are from this subfamily, with 1,000 to 2,000 species of all shapes and forms (probably 1600).
There are eight to ten, possibly independent, tribes within this subfamily.
Four tribes which evolved in North America (Cactacae, Pachycereeae, Echinocereeae, and Hylocereeae), and their parallels in South America (Notocacteae, Trichocereeae, Browningieae, Cereeae, and Rhipsalideae).
Barrel cacti, columnar cacti, and epiphytic cacti have parallel evolved in both North and South America after spreading from point of origin then being separated when the central American land bridge disappeared for a while.