Root Succulence


Apical Meristems







Published papers



Modifications to the Cactus Cortex

    The changes that occurred in the shoot cortex during cactus evolution involved more modifications than just adding more water-storage capacity by adding more parenchyma cells. The evolutionary increase in width of cactus stems due to having an increasingly wider cortex affected the selection pressures acting on numerous aspects of cactus biology, including:

    1) the need to translocate water and sugar within the cortex itself;

    2) the ability to change volume without changing surface area,

    3) the mechanical tissues needed to support the increased weight of stems, and

    4) the number of shoot apical meristem cells needed to produce the numerous cells of the voluminous body.


Cortical bundles: The translocation of water and sugar through the voluminous cortex

One consequence of the increased cortex width is that the photosynthetic outer cortex is pushed much farther away from the vascular bundles of the stem (the cauline bundles) as the cortex differentiates. Most cacti still have a relatively narrow pith, so in broad, globose barrel cacti and those with wide columnar shoots, the epidermis and chlorenchyma are often several centimeters – and as much as 30 cm in Echinocactus – from the xylem and phloem of the shoot’s vascular bundles. Over such distances, diffusion of water is extremely slow, diffusion of glucose or sucrose is even slower. Even with an extremely thick covering of cuticle and wax, any epidermis will lose some water to dry desert air, and that water must be replaced rapidly if the epidermis and outer cortex tissues are to remain alive. Similarly, as the outer chlorophyll-bearing cortex cells produce sugars and starch, those carbohydrates must be moved away from the plant’s surface if they are to be of any use to the rest of the plant. Rapid translocation throughout the voluminous cortex of many cacti is made possible by the presence of a system of collateral cortical bundles.

Cortical bundles branch off from stele bundles and from leaf/bud traces, then run outward, tangentially and upward through the cortex, branching extensively and vascularizing all regions of the cortex except the outermost layers of palisade cortex. Cortical bundles always end near or in the basal layers of the palisade cortex, they never extend to the hypodermis or epidermis. As with leaf venation, the most distal portions of cortical bundles are small, narrow bundles with few conducting cells, so that they look like leaf minor veins. In a large number of species, cortical bundles end in clusters of cells that resemble terminal tracheids in leaf veins: the cells are large and irregularly isodiametric but they have helical or reticulate secondary walls reinforcing their primary walls. These “terminal tracheid-like” cells usually occur in small clusters of about ten to twenty cells.

   This plant of Echinocactus platyacanthus has a very wide cortex. Many non-cactus succulents have very broad shoots like this, but those are invariably due to a large amount of parenchymatous wood.

Click on any photo for a larger image.

  Slice of stem of Echinopsis hammerschmidii held up to the sun, showing ring of ordinary vascular bundles (cauline bundles), radiating leaf/bud traces, and numerous dots and threads, which are cortical bundles and medullary bundles.  Transverse section of cortical bundle in Pediocactus simpsonii. There is a small amount of phloem at top, a larger amount of xylem below. Note large intercellular spaces in the cortex.

Because cactus cortex is long-lived, cortical bundles must also function for many years. In all species, cortical bundles produce secondary phloem, and in very old bundles at the base of very old shoots, the amount of collapsed secondary phloem can be many millimeters thick. Many species also produce secondary xylem in their cortical bundles, but typically it is only a small amount even in very old bundles in very old cortex. An unusual feature that needs more study is that in a few species, cortical bundle secondary xylem contains a matrix of xylary fibers, not xylary parenchyma.

  The apical regions of this Pilosocereus are less than one year old but the trunk regions are several decades old. Cortical bundles supplying the trunk cortex cells must function much longer than leaf veins.

Click on any photo for a larger image.

  This cortical bundle is in apical cortex and is thus young and very small, with only primary xylem and just a bit of secondary phloem.   This cortical bundle is in the trunk cortex and is thus very old. The dark streak running vertically is collapsed secondary phloem: the bundle's cambium had to produce huge amounts of secondary phloem for it to still be so prominent despite collapsing.


    Because cortex of most cacti is vascularized by cortical bundles, the epidermis and palisade chlorenchyma can be kept hydrated and healthy no matter how much water they lose to the dry desert air, as long as the plant itself has enough water of course. The presence of cortical bundles may be a key innovation in the evolution of cacti because it has permitted cactus cortex to become extremely broad as mentioned above. It is hard to over-emphasize the importance of the evolution of cortical bundles in cacti: whereas we animals vascularize any additional tissues we produce (such as tumors), plants do not do this automatically. In contrast, a plant’s ability to produce extra tissues is extremely limited by its inability to vascularize them. Consequently other stem-photosynthetic plants such as euphorbias, pachypodiums, stapeliads never become as broad as cacti do. It is certainly popular to point out the convergent evolution between euphorbias and cacti, but no euphorbia ever becomes as broad as a barrel cactus, or even as broad as many smaller cacti. The reason is simple: none of the non-cactus stem-photosynthetic succulents have cortical bundles in their stems. They are free to evolve to have a wider, more voluminous cortex only up to the point at which diffusion from the stele becomes too slow to keep the epidermis healthy. One of the broadest euphorbias, E. obesa, achieves its width by widening its stele: it has a broad pith that spreads its set of vascular bundles into a ring with a wide diameter, which is in turn surrounded by a cortex that is only moderately thick. But storing water in pith is more difficult than storing it in the cortex because a pith can absorb a large amount of water and swell only if the wood is flexible, otherwise either the wood would be torn apart or the pith would be prevented from expanding enough to absorb much water.

    Cortical bundles are present in almost all members of subfamily Cactoideae (only Blossfeldia liliputana lacks them) but are absent from subfamilies Pereskioideae and Opuntioideae.

  This is a full-sized plant of Euphorbia obesa -- it has the same shape as a globose cactus such as the Echinocactus platyacanthus above, but it never becomes as large -- without cortical bundles, its cortex cannot become truly thick and voluminous.

Click on any photo for a larger image.

  This cladode ("pad") of Platyopuntia engelmannii may appear to be big, fat and succulent, but look at it from the side (the exact same pad is shown in the photo on the right). The green color of the pad indicates the epidermis is still present.   Side view of the same pad as in the photo to the right. Without an extensive system of cortical bundles, the cortex of these stems is not free to evolve to become very thick.

Mauseth, J. D. 2004. The structure of photosynthetic, succulent stems in plants other than cacti. The International Journal of Plant Sciences 165: 1-9.

Mauseth, J. D. 2004. Cacti and other succulents: stem anatomy of “other succulents” has little in common with that of cacti. Bradleya 22: 131-140.

Sajeva, M., and J. D. Mauseth. 1991. Leaflike structure in the photosynthetic, succulent stems of cacti. Annals of Botany 68: 405 – 411.

Mauseth, J. D., and M. Sajeva. 1992. Cortical bundles in the persistent, photosynthetic stems of cacti. Annals of Botany 70: 317 – 324.

Mauseth, J. D. 1993. Medullary bundles and the evolution of cacti. American Journal of Botany 80: 928 – 932.


Collapsible water-storage cells.

    Cactus cortex is divided into an inner water-storing region and an outer photosynthetic layer. More than cortical bundles is needed to move water from the storage region to the chlorophyllous region: water-storing cells should give up water more easily than do other cells. The selective advantage of storing water is not just that it keeps water-storage cells alive but rather that the water can be made available to cells of the photosynthetic palisade cortex, the apical and axillary buds, any flower buds or developing fruits and so on. Consequently, water-storage cells should have thin, flexible walls that can contract or shrink readily such that the cell’s volume diminishes as water is transferred out. On the other hand, the cells that need the water should be more resistant to shrinkage: if all cell walls were equally flexible, all parts of the plant would suffer water stress equally, but that is not adaptive. Instead, water-storage tissues should give up water so easily that the more active cells do not experience water stress unless drought is extremely prolonged. In all cacti, cell walls of the inner cortex are especially thin and flexible, but in many cacti there is an additional modification: the walls are folded or undulate, even when young and recently produced by the shoot apical meristem. Because the walls already have folds in them, the cells are presumably set to shrink very easily. These cells have been called collapsible cortex cells. In contrast, wall of palisade cortex cells – the photosynthetic cells – are thicker and lack folds, so these cannot shrink as easily. Consequently, as the volume of water within a cactus shoot diminishes, cells of the inner cortex give up water preferentially, transferring it to other cells, minimizing water stress and thus allowing photosynthesis to continue.

  Turgid water-storage parenchyma cell in Brasilicereus phaeacanthus inner cortex. The cells are turgid and have smooth, thin walls. No organelles are visible because the cell consists almost entirely of central vacuole.

Click on any photo for a larger image.

  Slightly collapsed water-storage parenchyma cells in the inner cortex of a Brasilicereus plant that had not been watered for several weeks.   Very collapsed water-storage parenchyma cells in the inner cortex of Jasminocereus thouarsii. This was collected from a plant in the Galapagos Islands during a drought so severe that the cactus was extremely wilted and flaccid but alive and capable of recovering. If given water, these cells would swell and quickly resemble those in the turgid Brasilicereus on the left.

Mauseth, J. D. 1995a. Collapsible water-storage cells in cacti. Bulletin of the Torrey Botanical Club 122: 145 – 151.


Mucilage cells.

A perplexing aspect of water storage in cacti and many other plants is the role of mucilage cells. These are present in many cacti as sparsely dispersed individual cells in some species, as such abundant cells in other species that the plants feel slimy when cut. Spectacular concentrations of mucilage cells occur in prickly pears (members of Opuntia with flat pads) where they form canals many micrometers wide running just exterior to the phloem of the vascular bundles. Plants of living stones (Ariocarpus fissuratus) also have a giant mucilage cavity filling the center of their large taproot. Mucilage is predominately a carbohydrate capable of binding and holding water very firmly. For example, when samples of Opuntia are being prepared for microscopy, they are first dissected, preserved in formaldehyde, then dehydrated by placing them is a series of alcohol solutions, each solution being a stronger solution with less water than the previous one (remember, beer is 5% alcohol, 95% water; wine is 14% alcohol, 86% water). When the formaldehyde is poured off so that the 25% alcohol solution can be poured onto the samples, the mucilage has turned the formaldehyde solution into a slimy mess that – to be really vivid – resembles a runny nose. The mucilage is often so slimy/sticky that the tissue pieces have to be pulled out of it with forceps, it simply will not pour off. This problem continues until the samples have been soaked in 85% alcohol, at which point the slimy/stickiness disappears completely. Even when the samples are exposed to 70% alcohol (140 proof), the mucilage has such a high affinity for water that it can hold the water in the cactus tissue, preventing its loss to the alcohol. You can imagine that the mucilage will be just as effective in a tug-of-war with dry desert air. Undoubtedly, mucilage does help cacti retain water.

    But with water held tightly in mucilage cells, how does that help all the other cells of the plants? As described with collapsible cortex above, water-storage cells must provide water to the other cells of the plant, not keep it for themselves while the rest of the plant desiccates. One hypothesis is that as cacti experience a prolonged drought, they should be able to digest their mucilage and thus release the water to other cells. If so, the mucilage content of well hydrated plants should differ from that of plants suffering from drought. However, this has not been tested.

  This cortex cell has filled itself with mucilage (generated by dictyosomes then secreted between the cell wall and the plasma membrane, causing the protoplast to shrink and finally die -- programmed cell death?). Often the mucilage appears layered, but this might be a fixation artifact. This is Corryocactus brachypetalus, but many cacti have mucilage cells like this.

Click on any photo for a larger image.

  Almost all opuntias with flattened pads (cladodes) have canals of mucilage cells aligned parallel to the vascular bundles and located just exterior to the phloem. When cut, the mucilage cells ooze out like toothpaste. This is a pad of Nopalea.
  The center of the roots of Ariocarpus fissuratus are occupied by large canals filled with mucilage cells.   Shoots of Uebelmannia gummifera have canals of mucilage cells that run just at the bases of the ribs.

Mauseth, J. D. 1980. Release of whole cells of Nopalea (Cactaceae) into secretory canals. Botanical Gazette 141: 15 – 18.

Mauseth, J. D. 1980. A stereological morphometric study of the ultrastructure of mucilage cells in Opuntia polyacantha (Cactaceae). Botanical Gazette 141: 373 – 378.


Volume of stem must increase but surface area of epidermis must remain constant.

    The evolutionary expansion of water-storage tissues in the cortex created problems for the epidermis. Whereas we animals produce new epidermis cells on just about any part of our bodies all through our lives, plants produce new epidermis cells only at shoot and root tips or in developing leaves and flowers.

  The shoot apex of this Mammillaria magnimamma is the only place on the shoot where new epidermis cells are being made.

Click on any photo for a larger image.

  The shoot apex of Cereus forbesii, the only source of new epidermis cells. Just a few centimeters below the apex, all three ribs have achieved their full size and full complement of epidermis cells; if the ribs were to expand now, the epidermis would be torn apart.    This shoot of Cleistocactus was cut off, causing three axillary buds to be released from apical dominance. The shoot apex of each bud is making new epidermis cells.

    Consequently, as a succulent stem swells to an increased volume after a rain, it can not increase its surface area correspondingly by producing new epidermis cells. If it were limited by the amount of stretching the epidermis could undergo without splitting open, then the very limited ability to increase surface area would result in a very limited absorption of water. But most cacti – and also many other stem-succulent taxa – have a pleated surface composed of either ribs (vertical pleats only) or tubercles (both vertical and transverse pleats).

  Three ribs on a stem of Coleocephalocereus goebelianus. If the plant were desiccated, the bases of the ribs would be contracted tangentially, and four or five ribs would be visible.

Click on any photo for a larger image.


  Very short cacti like this Mammillaria magnimamma usually have their surface folded into tubercles rather than ribs. The tubercles allow the plant to shrink as it loses water during a drought, or to re-expand vertically as it absorbs water after a rain.   Ribs are visible in this stem cross section of this Cereus repandus. The outermost layer of cortex -- about 2 mm thick just under the epidermis -- is dark green chlorenchyma whereas the inner portions of the cortex are light green and mostly just store water rather than carry out photosynthesis.

    It is easiest to think of a cactus cortex as having two parts, an outer cortex consisting of ribs or tubercles and an inner part between the vascular bundles and the bases of the tubercles (see the photo above, right). After the plant has lost water during a prolonged drought, its ribs will have less volume and the two sides of each rib will have pulled together, thus the base of each rib will be narrow. The outer circumference of the inner cortex is the sum of the bases of all the ribs, so with the ribs shrunken, the outer circumference of the inner cortex is smaller, which means that the inner cortex has become thinner through dehydration. When rains return and the plant absorbs water, the inner cortex can swell, becoming thicker and having a larger outer circumference. This increase in outer circumference is accommodated by an expansion of the base of each rib: the volume of both the inner cortex and of the ribs increases whereas the surface area of the shoot remains unchanged. With tubercles instead of ribs, the shoots can cycle between being shorter and taller as well as thinner and thicker. Stem-succulents with pleated surfaces can undergo great changes in volume without any risk of their epidermis being torn open.

  Plants of Ariocarpus fissuratus are never tall, but because their surface is covered with tubercles, they can actually contract below the soil surface if they become extremely desiccated.

Click on any photo for a larger image.

  Many species of Mammillaria, like this M. meiacantha, grow as short, disk-like plants. Their tubercles permit the shoot surface to become both shorter and narrower as the plants lose water during dry periods.

Of course, having a pleated surface keeps the shoot from having a minimum S/V ratio, so transpirational surface area is not as low as possible. But the surface area is also important for light absorption, and having a pleated surface increases the amount of light that can be absorbed as well as gives the shoot an ability to contract and expand. As you might guess, the number and height of ribs are important factors: many tall ribs can expand more than a few short ribs. A curious aspect of ribs that is not immediately obvious is that whereas the total surface area of a set of ribs is affected by both the number and height of the ribs, the total volume of water-storage tissues located within the set of ribs is determined only by their height, not by their number. If the number of ribs on a shoot is doubled, then surface area is doubled but the width of the base of each must be reduced to half, so each rib will have half the volume of an original rib. Thus, having numerous ribs of a certain height allows more light absorption (and water loss) than does having fewer ribs of the same height. It might be that stem-succulents of drier habitats would have fewer ribs whereas those of more mesic ones would have more ribs, but this still needs to be examined.

Mauseth, J. D. 2000. Theoretical aspects of surface-to-volume ratios and water-storage capacities of succulent shoots. American Journal of Botany 88: 1107 – 1115.

Mauseth, J. D. & R. Ross. 1988. Systematic anatomy of the primitive cereoid cactus Leptocereus quadricostatus. Bradleya. 6: 49 – 64.

Mauseth, J. D. 1996. Comparative anatomy of tribes Cereeae and Browningieae (Cactaceae). Bradleya 14: 66 – 81.

Mauseth, J. D., and J. V. Landrum. 1997. Relictual vegetative anatomical characters in Cactaceae: the genus Pereskia. Journal of Plant Research 110: 55 – 64.

Mauseth, J. D., T. Terrazas, and S. Loza-Cornejo. 1998. Anatomy of Relictual Members of Subfamily Cactoideae, IOS Group 1a (Cactaceae). Bradleya 16: 31 – 43.

[End cortex page]