American Journal of Botany 87(8): 11071115. 2000.

THEORETICAL

ASPECTS OF SURFACE-TO-VOLUME

RATIOS AND WATER-STORAGE CAPACITIES OF

SUCCULENT SHOOTS1

JAMES D. MAUSETH2
Section of Integrative Biology, BIO LABS 311, The University of Texas, Austin, Texas 78712 USA
Surface-to-volume (S/V) ratios of drought-adapted plants affect transpiration, photosynthesis, and water-storage capacity. The S/V
ratio of cladodes and flat leaves is S/V 5 2/T, where T is thickness: even slight thickening greatly reduces S/V. During rain/drought
cycles succulent stems swell and shrink without tearing by having flexible ribs, but ribs increase S/V above that of a smooth cylindrical
stem with equal volume: the increased surface area is Sribbed/Scylindrical 5 N[x2 1 (p/N)2] /p(1 1 x ), where N is number of ribs and
x is rib height relative to the radius of the inner stem. Numerous low ribs provide moderate expandability (storage volume) with little
increase in S/V and are adaptive where droughts are short. Tall ribs provide greater expandability but greatly increase S/V and probably
are adaptive only in mesic habitats. Having ;815 ribs, each about as tall as the inner stem radius, provides large storage capacity
and intermediate increase in S/V. By increasing absolute size, S/V is reduced so greatly that even large ribs can have an S/V smaller
than that of a narrow cylindrical or spherical stem with less volume.
Key words:

adaptation; Cactaceae; cactus; cladode; desert; evolution; succulent; surface-to-volume ratio; xeric.

Many plant species have adapted to xeric conditions by becoming succulent, and during their evolution, several problems
had to be solved. First, the transpirational surface area could
be reduced either temporarily by leaf abscission or permanently by evolutionary reduction of leaves. Second, sufficient
water storage capacity had to be available to allow persistent
organs such as buds, roots, and the stem axis to survive
droughts. Third, seasonal rain/drought cycles caused the
plants volume to increase and decrease cyclically. These three
factors affect a plants surface-to-volume (S/V) ratio.
Simply reducing a plants S/V ratio to a minimum may not
be an optimal survival strategy. Because the transpirational
surface is typically also a photosynthetic surface, reducing surface area reduces photosynthesis (Fig. 1). If the habitat has
high atmospheric humidity and only short periods without
rain, then maintaining extra surface area (a non-minimum S/V
ratio) may be advantageous (Figs. 26). Conversely, if a plant
will experience droughts that last a year or longer, large
amounts of succulent tissue and a low S/V ratio may be necessary (Figs. 7, 8). Several species of cactus survive up to
three years without water (Szarek and Ting, 1975; Smith and
Madhaven, 1982).
As the amount of succulent tissue increases in a stem, so
does the potential for large changes in volume: the plant will
swell greatly after a rain and shrink during drought (Nobel,
1981; Barcikowski and Nobel, 1984; Mauseth, 1995). The epidermis and hypodermis must accommodate this, but whereas
young, growing dermal tissues are extremely extensible, mature ones are not: the total surface area of a region of mature
stem tends to be constant. Many succulent stems have contiguous ribs that can widen or shrink at the base whenever the
stem swells or contracts (Fig. 9; Porembski, Martens-Aly, and
Barthlott, 1991; Felger and Henrickson, 1997). When dry, the

stem has lost volume and the ribs are narrow; when hydrated,
the stem is swollen and its ribs are broad. Thus, volume cycles,
while surface area remains constant. Ribbed stems occur in
Asclepiadaceae, Cactaceae, Euphorbiaceae, and Vitaceae as
well as other families.
Certain aspects of the effects of ribs on the S/V ratio are
intuitive, but others are not. If a stem has a certain size, shape,
and number of ribs, a mutation that causes it to have more
ribs will also cause it to have more surface area (but how much
more?), a higher S/V ratio (is the change significant?) but a
greater capacity to expand (how much more capacity?). If a
mutation causes the ribs to be taller, the mutant stem will have
more surface area, but how is the S/V ratio affected (ribs add
both surface and volume) and how is expandability affected?
Would a particular number, size, and shape of ribs minimize
the S/V ratio but maximize capacity to expand and store water?
Are particular rib geometries adaptive in moderately xeric habitats, whereas others are more functional in extremely dry regions? In an attempt to answer these questions, the geometry
of leaves, stems, and ribs was analyzed to study the effect of
these factors on the S/V ratio of plants. Actual data from several sample plants were used to calculate real values.
MATERIALS AND METHODS
The values listed in Table 1 and the plants illustrated in Figs. 18 are part
of a large data set of almost 200 species that were studied in the field, in
botanical gardens or obtained from nurseries (Abbey Garden Nursery, P. O.
Box 2249, La Habra, California 90632-2249 USA, phone: 562-905-3520;
Mesa Garden Nursery, P. O. Box 72, Belen, New Mexico 87002 USA, phone:
505-864-3131; Miles to Go Nursery, P. O. Box 6, Cortaro, Arizona 85652
USA, phone: 520-682-7272; also see http://www.cactus-mall.com). All measurements were taken on mature, living plants before fixation and dehydration.

RESULTS
Manuscript received 8 June 1999; revision accepted 2 November 1999.
This research was funded in part by grants from the Mellon Foundation
through the Institute of Latin American Studies at the University of Texas and
from the Research Committee of the Cactus and Succulent Society of America.
2
E-mail: j.mauseth@mail.utexas.edu.
1

LeavesFor ordinary flat, non-succulent leaves, the surface

area S of the upper surface is approximately equal to that of
the lower surface, and the total leaf surface 5 2S. Leaf volume
V can be computed by multiplying leaf thickness T by either
the upper or lower surface: Vleaf 5 ST. Therefore:

1107

1108

AMERICAN JOURNAL

OF

BOTANY

[Vol. 87

Figs. 18. Variations in ribs. 1. Rhipsalis heteroclada is a rainforest epiphyte with narrow stems only ;3 mm in diameter. It is ribless, undergoing little
swelling or shrinking in its almost continuously moist habitat. Because it has only tiny, highly reduced leaves, it has little photosynthetic surface. 2. Notocactus
succineus is a small cactus (;25 mm radius) that occurs in grasslands. Droughts are not severe or prolonged, and the numerous low ribs provide enough
expansion capacity that stems swell enough to store moderate amounts of water. Due to the large number of ribs, no rib has to expand very much after a rain.
3. Trichocereus terscheckii is a large columnar cactus (stem radius ;100 mm) of very dry areas (surrounding plants are shrubby legumes). Although its ribs
are much larger than those of N. succineus, they are about the same size relative to the radius of the inner part of the stem. Due to its much larger size, T.
terscheckii has a much lower S/V ratio than N. succineus. 4. Deamia testudo has five very thin ribs (8 mm) that are tall (10 mm) relative to the inner stem
radius (3 mm), and its S/V ratio is extremely high. It survives only in rainforest habitats, and in cultivation it requires high humidity and frequent watering. 5.
Monvillea spegazzinii, with only four low ribs, has little capacity to swell or shrink during rainy/dry cycles. It occurs in only moderately dry areas, growing
among leafy shrubs and small trees. 6. Frailea chiquitana has numerous low ribs but its very small size (16 mm from rib tip to rib tip) gives it a high S/V
ratio, despite its almost spherical shape. It occurs only in shady, forested areas where soil retains moisture during short droughts. Notice the Selaginella, which
is larger than the cactus. 7. Neoraimondia gigantea has only 5 to 7 very tall ribs (55 mm), which would tend to give it a high S/V ratio (somewhat like D,
testudo in Fig. 4), but its stems are very thick (240 mm from rib tip to rib tip) and its large absolute size decreases its S/V ratio. N. gigantea occurs only in
extremely xeric coastal deserts of Peru, often with no other vegetation able to survive in the area. 8. Platyopuntias are the prickly pears, the opuntias with
flattened cladodes. This Opuntia phaeacantha stem is ;10 mm thick or less in dry conditions but can swell to 15 or 20 mm after a rain.

S/V ratio of a thin, flat leaf 5 2S/ST 5 2/T.

(1)

The S/V ratio is not related to the leafs length, width, or shape,
but only to its thickness (the surface along the leaf edge is
inconsequential unless the leaf is unusually thick relative to
the leafs length or width, or unless the leaf is highly dissected). Because the leafs S/V ratio is not directly related to
length or width, small leaves and large ones have the same S/
V ratio if they have equal thickness, and evolutionary reduction of leaf length or width would decrease total leaf surface
area but not the S/V ratio unless the leaf becomes so small
that length or width approaches T (Fig. 10). In contrast, an
evolutionary increase in leaf thickness would cause an increase
in leaf volume and a decrease in the S/V ratio (Fig. 15). A leaf
of ordinary thickness (for example, T 5 0.5 mm) has S/V 5
2/0.5 mm 5 4.0 mm2/mm3 and a thicker, markedly succulent
one (T 5 1.0 mm) has S/V 5 2.0 mm2/mm3. The same relationship would apply to flattened leaflike cladodes of plants
such as some opuntias (prickly pears; Figs. 8, 17).

Leafless, ribless stemsIf an ordinary cylindrical internode

is viewed in transverse section (Fig. 11), its perimeter P is
circular and equals the circumference C 5 2pr (r is the radius), whereas the cross-sectional area Ac-s 5 pr2. For a given
length L of stem, the surface area S 5 PL and volume V 5
Ac-sL, so the ratio of P/Ac-s is identical to the ratio of S/V: S/V
5 PL/Ac-sL 5 P/Ac-s. The ratio of perimeter to cross-sectional
area (and thus S/V) is
P/Ac-s of a cylindrical stem 5 2pr/pr2 5 2/r.

(2)

Thus for a stem that is smooth, cylindrical and has neither

leaves nor ribs, S/Vstem is related only to the stems radius (see
also Felger and Henrickson, 1997). This equation is basically
identical to S/Vleaf 5 2/T: a cylindrical leafless, ribless stem
has the same S/V as a flat thin leaf if r 5 T. Of course, most
leaves are thinner (;0.5 mm, S/V 5 4.0 mm2/mm3) than most
stems (radius of at least 2 mm, S/V of at most 1.0 mm2/mm3),
and consequently leaves have higher surface-to-volume ratios.

August 2000]

MAUSETHSURFACE-TO-VOLUME

RATIOS IN SUCCULENT STEMS WITH RIBS

1109

Fig. 10. The S/V ratio of flat leaves and cladodes is S/V 5 2/T, so the
leaves in (a) and (b) have the same S/V ratio because they have the same
thickness. The S/V ratio of leaf (c) is much lower because it is three times
thicker (see Fig. 15).
Fig. 9. While dehydrated during a drought, a ribbed stem has a small
volume (a), but when swollen after a rain its volume is greater (b), although
the surface area is unchanged. Consequently, the S/V ratio is high when dehydrated, low when filled with water. Ribs expand laterally but the collenchymatous or sclerenchymatous hypodermis prevents them from becoming
taller. Typically, ribs touch each other at their base and the stem axis has no
surface other than rib surface.

Most plants could reduce their S/V ratio by merely abscising

their leaves or losing them evolutionarily. But if a plant has
succulent, thick leaves on slender stems (e.g., T 5 1.0 mm
and r 5 1.0 mm, as occurs in many Mesembryanthemaceae
and Sansevieria species; Koller and Rost, 1986; Sajeva and
Costanzo, 1994), then S/Vleaf 5 S/Vstem and the plant could
reduce its total surface area but not its S/V ratio by becoming
leafless. Furthermore, such a plant could reduce its S/V ratio
by one-half by either doubling leaf thickness or stem radius
doubling leaf thickness would double leaf volume and have
only a small effect on surface, whereas doubling stem radius
would double the stem surface and quadruple its volume. Decreasing the S/V ratio by increasing stem radius actually in-

creases total surface area, unless the stem simultaneously becomes shorter.
A consequence of the relationship S/Vstem 5 2/r is that
among cylindrical leafless, ribless stems, broader stems have
lower S/V ratios than do narrower ones, although all have circular cross sections (Fig. 15). Without changing shape at all,
merely becoming wider, having a larger r (either as an individual plant grows or as a taxon evolves) will cause the S/V
ratio to decrease. In contrast, if radius remains constant, increasing a stems length does not change S/V; this is important
for succulents with stems that are extremely long (upright in

TABLE 1. Minimum, maximum, and mean values for several parameters, taken from a sample of almost 200 species of cacti. All dimensions are in mm or mm2.
Dimension

Pith radius
Wood thickness
Cortex thickness
Inner stem radius (r)
Rib number (N)
Rib height (H)
Rib base (B)
xa
Total stem crosssectional area (CSA)
Perimeter (P)
S/V
a

Minimum

0.2
0.1
1.0
1.3
2.0
0.8
1.0
0.03
20.3
18.9
0.012

Maximum

72.5
14
140
184
38
55
75
10.0
116 802
1509
2.958

Mean

10.7
1.9
19.9
32.5
13.6
11.2
15.0
0.59
7299
347
0.169

This is height of ribs as a factor of inner stem radius: H 5 xr.

Fig. 11. A cross section of a cylindrical stem has a circular outline. The
formulas for perimeter P, cross-sectional area Ac-s, surface S, and volume V
show that P/Ac-s 5 S/V.

1110

AMERICAN JOURNAL

OF

BOTANY

[Vol. 87

radius. Also, as rib height increases, rib cross-sectional area

and volume increase proportionally.
In a stem with ribs, total stem cross-sectional area (CSA)
and thus volume occur in both inner stem and in ribs:
total stem CSA 5 CSA of inner stem 1 CSA of all N ribs
5 pr 2 1 prH 5 pr(r 1 H).

Fig. 12. A rib can be considered as two right triangles back to back, with
the rib perimeter forming the hypotenuses. The formula Brib 5 2pr/N slightly
overestimates the length of the base of the rib and thus its cross-sectional area
and volume, but this is significant only for very low, wide ribs.

(6)

This shows that total CSA and volume of a ribbed stem depend only on inner stem radius and rib height, not on rib
number.
The photosynthetic cortex of succulent stems occurs as a
layer of cells just below the rib epidermis or hypodermis, so
water stored within ribs themselves is closer to the cells that
must be kept hydrated. From a water distribution standpoint,
it may be more advantageous to have a large fraction of the
water storage capacity located within the ribs rather than the
inner stem. This fraction can be calculated as
fraction of total stem CSA in ribs

some, but climbing, scrambling, or prostrate in others) but

which have parallel sides, being as broad near the shoot apex
as at the base (Figs. 3, 5, 7).
Stems with ribsIn a transverse section of a stem with ribs,
each rib appears as a triangle with its base Brib located on the
surface of a cylindrical stem (which will be called the inner
stem) and its height H projecting radially away from the inner
stem (Fig. 12). The triangular transverse section can be considered as two right triangles back to back with their hypotenuses constituting the rib perimeter. Ribs are usually symmetrical, so the dimensions of one half of the rib are identical to
those of the other. As discussed above, rib perimeter in cross
section multiplied by length results in surface area of the rib,
and cross-sectional area multiplied by length results in rib volume, therefore P/Ac-s 5 S/V for ribbed stems just as for ribless
ones. Typically each rib contacts two others along its base
(Figs. 27), and the surface of the stem consists of just rib
surface. In some species with only two or three thin ribs, there
is a strip of ordinary stem cortex and epidermis separating
each rib from its neighboring ribs.
Rib cross-sectional area is 1/2BribH. Because the entire circumference of the inner stem (with radius r) is covered with
ribs, if there are N ribs, then the base Brib of each rib is given
approximately by Eq. 3:
base of each of N ribs 5 2pr/N.

(3)

This is a slight overestimate because the arc calculated by 2pr/

N is a bit longer than Brib, which is the cord of the arc. With
this value for the base, the cross-sectional area of ribs can be
calculated as
Ac2s of one rib 5 (1/2)Brib H 5 (1/2)(2pr/N)(H)
5 prH/N.
Ac2s of all N ribs 5 N(prH/N) 5 prH.

(4)
(5)

Equation 5 shows that the cross-sectional area and thus the

volume of all the ribs in a stem depend only on the height of
the ribs and the radius of the inner stem that the ribs sit on.
Stems with many thin ribs (N is large) or just a few thick ones
(N is small) have equal amounts of volume in their ribs if the
ribs are the same height and sit on inner stems of the same

5 CSA ribs /CSA total stem 5 prH/pr(r 1 H)

5 H/(r 1 H).

(7)

If H is expressed as a fraction of r (for example, H 5 xr),

then Eq. 7 becomes
fraction of total stem CSA in ribs
5 xr/(r 1 xr) 5 x/(1 1 x).

(8)

Equation 8 shows that regardless of absolute size, the volume

of ribs relative to inner stem is related only to the height of
ribs relative to the inner stem radius. As H increases relative
to r, the fraction of stem CSA and volume that occurs as rib
tissue increases linearly, and once H 5 r (that is, x 5 1.0),
the volume of ribs equals that of the inner stem. If H . r (that
is, x . 1.0), more water is being stored in the ribs than in the
inner stem.
Rib perimeter and surface are calculated as the hypotenuses
of two right triangles constituting the rib (Fig. 12), with the
height H of each triangle equaling Hrib and base of each triangle Btriangle equaling half of Brib 5 (1/2)(2pr/N) 5 pr/N. Each
hypotenuse is [H2 1 (pr/N)2], and this must be doubled to
get the two hypotenuses of one rib:
rib perimeter 5 2[H2 1 (pr/N)2].

(9)

Note that this gives the rib perimeter that is covered by epidermis, not the total perimeter of a triangleit does not include the rib base, which does not transpire or photosynthesize. To calculate the total perimeter of the stem, Eq. 9 would
be multiplied by the number N of ribs:
total perimeter of all N ribs 5 2N[H2 1 (pr/N)2]. (10)
The CSA of each rib is prH/N (Eq. 4), and the P/Ac-s ratio of
one rib is obtained by dividing this into the perimeter (Eq. 9):
P/Ac2s of one rib 5 2[H 2 1 (pr/N) 2 ]/(prH/N)
5 2N[H 2 1 (pr/N) 2 ]/prH.

(11)

The P/Ac-s ratio (and thus the S/V ratio) of all the ribs is the
same as that of a single rib, which can be shown by dividing
Eq. 10 by the CSA of all N ribs (5 prH; Eq. 5):
P/Ac-s of all N ribs 5 2N[H2 1 (pr/N)2]/prH
which is identical to Eq. 11.

(12)

August 2000]

MAUSETHSURFACE-TO-VOLUME

RATIOS IN SUCCULENT STEMS WITH RIBS

1111

several right triangles have the same H/Btriangle ratio, larger ones
have smaller P/Ac-s and S/V ratios, just as broader cylindrical
stems have smaller S/V ratios than narrower ones. Tall, thin
ribs can have a smaller S/V ratio than those with H 1/2Brib
if their absolute size is sufficiently larger. Similarly, if two
stems have the same number and height of ribs but one stem
is broader than the other (and thus has broader ribs), the broader stem has a lower S/V ratio due to absolute size. If a stem
must have ribs, minimum S/V ratio is achieved by having the
largest possible ribs with H 1/2Brib.

Fig. 13. If the base of a low triangle or rib is held constant but the height
is increased, the S/V ratio decreases until it is an isosceles triangle with H 5
1/2B; then it increases again. Absolute size must be considered: a tall, thin
rib can have a lower S/V ratio than an isosceles rib if it is much larger.

Equations 912 demonstrate several important points. Rib

number affects P/Ac-s and S/V ratios: if two stems are identical
in r and H but one has more ribs (and thus Brib 5 2pr/N of
each rib is less), total CSA of all the ribs is the same in the
two stems (Eq. 5), but total perimeter is greater in the stem
with more ribs (Eq. 10), so its S/V ratio is greater. Consequently, when considering plants with similar inner stem radius and similar rib height, plants with more ribs have more
transpirational and photosynthetic surface area relative to volume and might be expected in more mesic habitats.
Rib shape affects P/Ac-s and S/V ratios. Being composed of
two right triangles, the hypotenuse is at a minimum relative
to CSA when Htriangle 5 Btriangle, that is, when each half of the
rib is a right isosceles triangle and the whole rib itself has Hrib
5 1/2Brib (Fig. 13). For any rib that is taller relative to its base
(H k B) or shorter (H K B), the P/Ac-s and S/V ratios are
higher: plants with tall thin ribs or low flat ribs have higher
S/V ratios than those in which rib height is half that of width,
if their cross-sectional areas are equal. Ribs provide flexibility
to the stem surface but automatically increase the S/V ratio
above that of a ribless cylindrical stem (P/Ac-s 5 2/r; Eq. 2),
but ribs with H 1/2Brib cause minimum increase in stem S/
V ratio. It might be expected that species adapted to arid conditions would have ribs with H 1/2Brib, whereas those in
more mesic environments could have tall, thin ribs or short,
broad ones, both of which would increase the S/V ratio, increasing photosynthetic surface relative to water storage volume.
Absolute size of ribs affects P/Ac-s and S/V ratios: even if

Cyclical changes due to water absorption and lossFor

any particular mature stem, N is constant but r, H, Brib and the
S/V ratio vary as the stem absorbs water after a rain and loses
it during drought (Fig. 9). Rib shape also changes: although it
is theoretically possible for H and Brib to expand or shrink
symmetrically such that rib shape remains constant, ribs on
real plants are significantly narrower when desiccated than
when hydrated, but not significantly shorter. As tissues of the
inner stem dehydrate and shrink, stem radius r decreases as
do inner stem circumference (2pr) and thus rib bases (2pr/
N). Rib height could decrease as well, but that would require
that the perimeter (the two hypotenuses) also decrease, but the
epidermis and hypodermis tend to have thick walls in ribbed
succulents, and probably can neither stretch nor shrink significantly (Mauseth, 1996; Mauseth, Terrazas, and Loza-Cornejo,
1998).
As a stem absorbs water and swells, ribs expand and their
bases become broader. As the inner stem expands, its original
radius r1 increases to r2 and its original circumference C1 increases to C2; the base of each rib 2pr1/N increases to 2pr2/
N. The total perimeter of all N ribs is constant, as given by
Eq. 10, which shows that the perimeter is the same whether
the stem is dehydrated and the ribs are thin and sitting on a
shrunken inner stem or whether the stem is fully hydrated and
the ribs are expanded and broad. Under ideal conditions, ribs
would be capable of expanding so much that the stem would
become circular in transverse section. But for this to happen,
the ribs would have to be so flexible that they could swell
until they become flat and the base of each rib is as long as
the two hypotenuses (Fig. 14); rib height would have to either
shrink or be very much smaller than the radius of the inner
cortex (x of H 5 xr would have to be small even when the
stem is dehydrated). If the stem could swell to a new circular
perimeter, then its new radius r2 would be the radius of a circle
whose circumference equals that of the perimeter of all the
original ribs. It is possible to express the new radius r2 as a
factor of r1 (for example, r2 5 yr1), and the circumference of
yr1 equals the perimeter of all N ribs (given by Eq. 10):

Fig. 14. If the cells in rib (a) were perfectly elastic (but the hypodermis in the hypotenuses was not), the rib could theoretically expand until the surface of
the plant was cylindrical and the base of the rib was stretched so much that its length equaled that of the two hypotenuses (b). Numerous low ribs may be able
to approximate this, but a few tall, thin ribs probably cannotcell expansion would have to be unrealistic.

1112

AMERICAN JOURNAL

OF

BOTANY

[Vol. 87

2pyr1 5 2N[H 2 1 (pr1 /N) 2 ]

y 5 (N[H 2 1 (pr1 /N) 2 ])/pr1.

(13)

Expressing rib height H as a fraction of inner stem radius r1,

again letting H 5 xr1 as was done for Eq. 8, we get
y 5 (N[(xr1 ) 2 1 (pr1 /N) 2 ])/pr1
y 5 (N[(x 2 1 (p/N) 2 ])/p.

(14)

The CSA in the fully expanded condition would be

CSA hydrated 5 p(yr1 ) 2 5 (N 2 r1 2 [x 2 1 (p/N) 2 ])/p.

(15)

The hydrated volume (Eq. 15) relative to the desiccated, ribbed

volume [pr21(1 1 x), which is obtained by letting H 5 xr1 in
Eq. 6: pr1(r1 1 H)] would be
CSA hydrated /CSA ribbed 5 {(N 2 r1 2 [x 2 1 (p/N) 2 ])/p}/pr1 2 (1 1 x)
5 (N 2 x 2 1 p 2 )/p 2 (1 1 x).

(16)

Equation 16 shows the amount of volume change that can be

accommodated by the stem as either the number N or height
x of its ribs increases. Although both N and x appear to contribute equally, it is probably more feasible to increase the
number of ribs such that each rib has to expand only a small
amount. With a small number of tall ribs, each rib might be
required to undergo an unrealistic change in shape, having
their cells and tissues expand in ways that do not seem structurally realistic. For example, even if all cortex cells could
swell greatly, it is difficult to imagine that ribs like those of
Figs. 3 and 4 could swell to form a round stem.
Considering all factors simultaneouslyThe effects of inner stem radius, rib number, and rib height can be considered
simultaneously. The total perimeter of a ribbed stem is given
by Eq. 10, and if rib height H is expressed as a factor of r
(again, H 5 xr) as was done to derive Eq. 8, we get
total perimeter of a ribbed stem
5 2N[(xr) 2 1 (pr/N) 2 ] 5 2Nr[x 2 1 (p/N) 2 ].

(17)

The total cross-sectional area of this stem is given by Eq. 6

as pr(r 1 H), and if H is expressed as xr, we get
total CSA of a ribbed stem
5 pr(r 1 xr) 5 pr 2 (1 1 x).

(18)

With these formulas, it is possible to determine the radius,

perimeter (surface), and S/V ratio of a cylindrical stem with
the same cross-sectional area (the same volume) as a ribbed
stem, that is, to evaluate the effects of various types of ribs
while holding CSA and volume constant. The radius of the
cylindrical stem can be expressed as a factor y of the radius
of the ribbed stem: yr. Because the cylindrical stem with radius
yr has the same total cross-sectional area [pr2(1 1 x); Eq. 18]
as the ribbed stem:
total stem CSA 5 p(yr)
pr 2 (1 1 x) 5 py 2 r 2

(19)

(1 1 x) 5 y.
Using this expression of y, it is possible to calculate the parameters of a cylindrical stem with the same CSA as a ribbed
stem:

Fig. 15. Graph of S/V 5 2/r (and simultaneously a graph of S/V 5 2/T).
Units for radius are in mm; those for S/V are in mm2/mm3. Narrow stems and
thin leaves or cladodes (r or T 5 1 or 2 mm) have a high S/V ratio, but
increasing thickness causes a very rapid decrease in the S/V ratio.

radius 5 ry 5 r(1 1 x).

(20)

circumference (perimeter) 5 2pry 5 2pr(1 1 x).

(21)

S/V 5 P/Ac2s 5 [2pr(1 1 x)]/pr (1 1 x)

2

5[2(1 1 x)]/r(1 1 x).

(22)

When comparing two stems with the same CSA, it may

seem logical to compare the S/V ratios of the two, but because
the two have the same CSA and volume, the denominators in
any comparison cancel, and by simply considering only the
difference in surface we automatically consider the difference
in S/V ratio.
Perimeter of ribbed stem (Eq. 17)
5 2Nr[x 2 1 (p/N) 2 ].
Perimeter of cylindrical stem (Eq. 21)
5 2pr(1 1 x).
Pribbed /Pcylindrical 5 Sribbed /Scylindrical 5 (S/Vribbed )/(S/Vcylindrical )
5 N[x 2 1 (p/N) 2 ]/p(1 1x).

(23)

Notice that the radius has been eliminated and the only variables that determine the extent of surface increase are the number N of ribs and their height as a factor x of the inner stem
radius. Equation 23 has been plotted for several values of rib
number and height that include the known range of values in
cacti (Tables 1, 2; Fig. 16). The mean number of ribs in available samples in Table 1 is ;14 and the mean height as a factor
x of inner stem radius is ;0.6. A stem with ribs like this would
have a surface area and an S/V ratio 2.25 times larger than a

August 2000]

MAUSETHSURFACE-TO-VOLUME

1113

RATIOS IN SUCCULENT STEMS WITH RIBS

Fig. 17. (a) In transverse section, a dehydrated cladode of a platyopuntia

(as in Fig. 8) has parallel sides and its ring of vascular bundles is also flattened. (b) When hydrated, the pith becomes swollen and the ring of bundles
is forced to change shape. As the cladode ages and has more wood, it will
become stiffer and such shape changes become more difficult.

Fig. 16. Graph of Eq. 23: Sribbed/Scylindrical 5 N[x2 1 (p/N)2]/p(1 1 x)

for values of rib number N up to 50 and values of rib height x as a factor of
inner stem radius (H 5 xr) between x 5 0.1 (very low ribs) and x 5 2.0 (tall
ribs).

cylindrical stem with equal CSA (Table 2) and would be able

to expand to a new volume 4.73 times larger than the ribbed
condition (if it could expand until it the new perimeter were
circular).
DISCUSSION
Low ribs provide flexibility without greatly increasing the
amount of surface or the S/V ratio. All values in the x 5 0.1
column of Table 2 are ,2.0, indicating that even if a stem had
50 very low ribs, its surface and S/V ratio would not even be
double that of a ribless cylindrical stem with the same CSA
and volume. Thus, having these numerous low ribs would not
greatly increase the risk of excessive transpiration in xeric habitats. Conversely, low ribs, even when there are many of them,
do not permit much expansion of the stem, so if an abundance
of water would be available after a heavy rain, the stem would
not have much expandability to store it. Both tendencies are
more extreme if there are just a few low ribs: the S/V ratio is
almost the same as that of a cylindrical stem but expandability
is almost zero. Plants with this type of ribbing might be expected to occur in habitats that are extremely dry but which
receive small amounts of rain periodically throughout the year:
low S/V ratio and considerable water storage capacity are advantageous, but the plant does not need to absorb a years
worth of water in a brief rainy season (Figs. 2, 5, 6).
Ribs of moderate height (x 5 0.6) have little effect on the
surface area and the S/V ratio unless they are quite numerous
(N . 25 or so; Table 2). A stem with ten ribs of height 60%
of the radius of the inner stem (x 5 0.6) has a S/V ratio that
is only 1.7 times larger than a cylindrical stem with the same
CSA and volume, and the increase does not double until a
plant has 12 or 13 ribs like this. These plants have low S/V
ratios, moderate expandability, and some extra photosynthetic
surface area. This might be adaptive in habitats that are not
so xeric that the S/V ratio is an overriding factor and in which
rainfall is strongly episodic yet reasonably reliablerains may
occur twice a year, and completely dry years are rare. In such
a habitat, plants would need to store sufficient water for four
or five very dry months but would rarely draw down their
stored water during a full year or two without rain. Felger and

Lowe (1967) found that in the columnar cactus Lophocereus

schottii, populations in drier habitats had fewer ribs and thus
lower S/V ratios than populations in more mesic regions.
If ribs of moderate height (x 5 0.6) are numerous (N .
25), then the S/V ratio becomes high, as much as 47.5 times
higher than that of a cylindrical stem with the same volume.
Such high S/V ratios might be advantageous in habitats such
as dry forests of Mexico and Brazil, habitats with enough
moisture to support the growth of tall trees but which have a
very dry, hot season in which leaves are abscised. During the
rainy season, the surrounding trees are leafy and shade the
succulent plants but during the dry, leafless season, the succulents receive full sunlight and are healthy and metabolically
active, relying on their stored water. In such habitats the increased surface area of the ribs may be advantageous in facilitating greater photosynthesis in the brief sunny period, and
the increased transpiration is tolerable because the dry season
is not prolonged. Succulents in these regions may actually not
conserve water very well, absorbing and losing large amounts
each year, and thus needing the increased expandability that
the numerous moderate-sized ribs provide. Coryphantha viviTABLE 2. The increase in surface area and S/V ratio caused by various
numbers of ribs of various heights x (with H 5 xr) in stems with
the same cross-sectional area. Values are calculated from Eq. 23:
Sribbed/Scylindrical 5 Nx2 1 (p/N)2)/p(1 1 x). The row for 14 ribs
and the column for x 5 0.6 are boldface because these are the
mean values for rib number and rib height (expressed as x) in
Table 1.
Number
of ribs

x 5 0.1

x 5 0.3

x 5 0.6

x 5 1.0

x 5 2.0

3
4
5
6
7
8
9
10
14
15
20
25
30
35
40
45
50

0.958
0.961
0.965
0.971
0.977
0.984
0.992
1.001
1.044
1.057
1.130
1.219
1.318
1.427
1.544
1.666
1.792

0.912
0.939
0.972
1.011
1.055
1.104
1.156
1.213
1.464
1.532
1.891
2.270
2.661
3.060
3.463
3.870
4.279

0.911
0.995
1.093
1.202
1.320
1.444
1.572
1.704
2.257
2.399
3.122
3.857
4.598
5.343
6.091
6.840
7.591

0.978
1.145
1.329
1.524
1.727
1.934
2.146
2.359
3.229
3.449
4.557
5.671
6.789
7.909
9.031
10.153
11.276

1.245
1.580
1.926
2.280
2.637
2.997
3.358
3.721
5.178
5.543
7.374
9.207
11.042
12.877
14.713
16.550
18.387

1114

AMERICAN JOURNAL

para undergoes cycles in which it survives losing as much as

91% of its water (Nobel, 1981).
Tall ribs greatly increase the S/V ratio, even if there are few
ribs. Ribs with x 5 1.0 double the S/V ratio of a stem even if
there are as few as eight or nine ribs, and as few as five ribs
with x 5 2.0 doubles the S/V ratio over that of a cylindrical
stem with the same CSA. Fifty tall ribs would increase the S/
V ratio by as much as 1118 times. Plants with a few tall ribs
would be expected to be restricted to extremely mesic habitats
that experience either brief or no drought. Very tall ribs (x .
2.0) essentially would be acting like thick succulent leaves
(ribs are never as thin as ordinary leaves: minimum Brib 5 1.0
mm in Table 1), with the significant difference that ribs cannot
be abscised during drought, whereas leaves can be. Certainly
the three or four very tall ribs of Deamia (Fig. 4) could never
expand fully to make the plant cylindrical, and such a huge
water storage capacity would be completely unnecessary in
their mesic habitat in Central America (Backeberg, 1977). Instead, the large surface area provided by these ribs undoubtedly increases the photosynthetic surface area. Plants with
many tall ribs (N . 25, x 5 2.0) might not be expected to
occur at all: with their extremely high S/V ratio they would be
restricted to very humid habitats, and whereas a few tall ribs
might be effective at photosynthesis, large numbers of tall
ribscrowded on a relatively narrow stemwould shade each
other (Nobel, 1980). It is difficult to conceive of a habitat in
which these features would be selectively advantageous. Very
tall ribs do occur (Table 1), with a maximum of x 5 10.0 in
Rhipsalis (N 5 2, H 5 10 mm) and H 5 55 mm in Dendrocereus (N 5 3 or 4; x 5 7.86). As might be expected, these
are both plants of very humid habitats where droughts
would last at most a week or twoRhipsalis is a rainforest
epiphyte and Dendrocereus occurs in rainy areas of Hispaniola
and Cuba. However, it is not yet known if these two species
are exceptional, and more actual species must be examined.
The effect of absolute size may be the most significant factor in maximizing water storage and minimizing the S/V ratio.
Increasing the radius r of the inner stem dramatically decreases
the S/V ratio (Fig. 15), and similarly, increasing the absolute
size of a rib while holding its shape constant causes its S/V
ratio to decrease (Fig. 7). In both cases, as S/V is decreasing,
the actual surface area is increasing, so photosynthetic capacity
is increasing. Consequently, maximum water storage capacity
and minimum S/V ratio can be achieved primarily by having
stems become very large. Some cacti (Soehrensia) have enormous stems (CSA 5 116 802 cm2; Table 1). If a large number
of ribs of moderate height is added to a large body, expandability is achieved without increasing the S/V ratio too severely, again because of the large absolute size of the ribs. In
contrast, stems with the same relative dimensions (the same N
and x) would have a much higher S/V ratio if they are narrow
in absolute size (r is small). The larger stems could survive in
very xeric habitats, whereas the narrow stemswith the same
shapewould be restricted to mesic ones (Figs. 2, 6).
The presence of ribs also affects the stems interaction with
wind and its strength. Air moves smoothly across a large flat
surface, creating a thick boundary layer, but ribs create turbulent air flow and a thinner boundary layer (Nobel, 1988). It
may be that all ribbed stems have so much more turbulence
than smooth cylindrical stems that number and size of ribs are
inconsequential. Turbulent air flow carries away excess heat
more rapidly than laminar air flow, but it also removes transpired water vapor more rapidly. Presence of ribs increases the

OF

BOTANY

[Vol. 87

strength and rigidity of stems as well, in some cases providing

as much support as the wood (Cornejo and Simpson, 1997;
Niklas, Molina-Freaner, and Tinoco-Ojanguren, 1999). The
mechanical aspects of ribbed stems should be an important
phenomenon for further studies, especially in species with
very tall ribs.
Unlike ribbed stems, flattened cladodes can swell and shrink
without tearing or wrinkling. By being broad and thick, their
S/V ratio 5 2/T and as Fig. 15 shows, even moderately thick
cladodes have low S/V ratios, although not nearly as low as
cylindrical stems with the same CSA. Benson (1982) mentions
several platyopuntias (O. chlorotica, O. ficus-indica, O. lindheimeri, O. littoralis) that have cladodes 2025 mm thick and
that would have an S/V ratio of 0.080.1 mm2/mm3. In cladodes of platyopuntias (Figs. 8, 17), not only is the stem flattened, but so is the ring of vascular bundles and pith. Water
is stored in both cortex and pith, and volume change requires
flexion of the wood; this is possible while the cladodes are
young, but as they age and develop more wood, they must
become rigid. Ribbed cylindrical stems provide flexibility despite the woodiness of the stem.
During the course of their evolutionary history, succulent
plants of various families have become adapted to an extremely wide range of habitats that differ in their rainfall patterns.
It is to be expected that a variety of rib shapes and numbers
and of plant sizes will have evolved, with certain combinations
of characters being adaptive in certain environments and other
combinations adaptive elsewhere. A sample of almost 200 species of cacti is being studied to determine whether actual species correspond to the principles outlined here.
LITERATURE CITED
BACKEBERG, C. 1977. Cactus lexicon. Blandford Press, Dorset, UK.
BARCIKOWSKI, W., AND P. S. NOBEL. 1984. Water relations of cacti during
desiccation: distribution of water in tissues. Botanical Gazette 145: 110
115.
BENSON, L. 1982. The cacti of the United States and Canada. Stanford University Press, Stanford, California, USA.
CORNEJO, D. O., AND B. B. SIMPSON. 1997. Analysis of form and function
in North American columnar cacti (Tribe Pachycereeae). American Journal of Botany 84: 14821501.
FELGER, R., AND J. HENRICKSON. 1997. Convergent adaptive morphology of
a Sonoran desert cactus (Peniocereus striatus) and an African spurge
(Euphorbia cryptospinosa). Haseltonia 5: 7785.
, AND C. H. LOWE. 1967. Clinal variation in the surface-volume relationships of the columnar cactus Lophocereus schottii in northwestern
Mexico. Ecology 48: 530536.
KOLLER, A. L., AND T. L. ROST. 1986. The microscopic anatomy of Sansevieria leaves. Cactus and Succulent Journal (U. S.) 58: 3033.
MAUSETH, J. D. 1995. Collapsible water-storage cells in cacti. Bulletin of the
Torrey Botanical Club 122: 145151.
. 1996. Comparative anatomy of Tribes Cereeae and Browningieae
(Cactaceae). Bradleya 14: 6681.
, T. TERRAZAS, AND S. LOZA-CORNEJO. 1998. Anatomy of relictual
members of Subfamily Cactoideae, IOS Group 1a (Cactaceae). Bradleya
16: 3143.
NIKLAS, K. J., F. MOLINA-FREANER, AND C. TINOCO-OJANGUREN. 1999. Biomechanics of the columnar cactus Pachycereus pringlei. American Journal of Botany 86: 767775.
NOBEL, P. S. 1980. Interception of photosynthetically active radiation by cacti
of different morphology. Oecologia 45: 160166.
. 1981. Influence of freezing temperatures on a cactus, Coryphantha
vivipara. Oecologia 48: 194198.
. 1988. Environmental biology of agaves and cacti. Cambridge University Press, Cambridge, UK.
POREMBSKI, S., B. MARTENS-ALY, AND W. BARTHLOTT. 1991. Surface/vol-

August 2000]

MAUSETHSURFACE-TO-VOLUME

RATIOS IN SUCCULENT STEMS WITH RIBS

ume-ratios of plants with special consideration of succulents. Beitrage

zur Biologie der Pflanzen 66: 189209.
SAJEVA, M., AND M. COSTANZO. 1994. Succulents, the illustrated dictionary.
Cassell plc, Villiers House, London, UK.
SMITH, B. N., AND S. MADHAVEN. 1982. Carbon isotope ratios in obligate
and facultative CAM plants. In I. P. Ting and M. Gibbs [eds.], Crassu-

1115

lacean acid metabolism, 231243. American Society of Plant Physiologists, Rockville, Maryland, USA.
SZAREK, S. R., AND I. P. TING. 1975. Photosynthetic efficiency of CAM pants
in relation to C3 and C4 plants. In R. Marcelle [ed.], Environmental and
biological control of photosynthesis, 289297. Dr. W. Junk, The Hague,
The Netherlands.