Content uploaded by Byron B. Lamont
Author content
All content in this area was uploaded by Byron B. Lamont on Mar 25, 2014
Content may be subject to copyright.
Relative Performance of Clumped vs. Experimentally Isolated Plants in a South African
Winter-Rainfall Desert Community
Author(s): Neil Eccles, Byron Lamont, Karen Esler and Heather Lamont
Source:
Plant Ecology,
Vol. 155, No. 2 (Aug., 2001), pp. 219-227
Published by: Springer
Stable URL: http://www.jstor.org/stable/20051124 .
Accessed: 25/03/2014 01:55
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
.
Springer is collaborating with JSTOR to digitize, preserve and extend access to Plant Ecology.
http://www.jstor.org
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
^L Plant Ecology 155: 219-227, 2001. 219
? 2001 Kluwer Academic Publishers. Printed in the Netherlands.
Relative performance of clumped vs. experimentally isolated plants in a
South African winter-rainfall desert community
Neil Eccles1, Byron Lamont2, Karen Esler3* and Heather Lamont2
'Institute for Plant Conservation, University of Cape Town, Rondebosch, 7701, South Africa; 2Department of
Environmental Biology, Curtin University of Technology, P.
O Box, U1987, Perth, 6845, WA, Australia;
3Department of Botany, University of Stellenbosch, Private Bag XI, Matieland, 7602, South Africa; *Author
for correspondence (e-mail: kje@land.sun.acza; fax: +27 -21 808 3607)
Received 1 February 2000; accepted in revised form 28 January 2001
Key words: Branch growth, Competition, Positive interactions, Risk, Shrubs, Strandveld, Vegetation clumps,
Water relations
Abstract
We conducted a factorial removal experiment in the arid strandveld on the west coast of South Africa to test the
hypothesis that perennial species with a "preference" for occurring in
multi-species clumps should derive ben
efits from their association into clumps. Contrary to our hypothesis, we obtained evidence of competition for
water in the clumped non-succulent species studied in the form of depressed water potentials. We were not able
to detect any effect on leaf water contents associated with isolation, suggesting that clumped plants are able to
compensate physiologically in response to competitive stress. We also found that isolating individuals had no
effect (positive or negative) on branch growth. Finally, we showed that isolating individuals exposed them to a
far greater risk of damage by wind or animals. In light of these results we conclude that the spatial arrangement
of plants in this community does suggest a situation where the benefits associated with occurring in clumps ex
ceed any competitive costs.
Introduction
One of the most striking features of plant communi
ties in the arid west coast strandveld (part of the win
ter rainfall Succulent Karoo biome) of South Africa
is that most perennial plants are concentrated into
multi-species clumps (Eccles et al. 1999). This in it
self is not particularly unusual. There are several ex
amples in the literature, many of them in desert sys
tems, which suggest that clumped vegetation patterns
are in fact very common, if not predominant. For in
stance, the majority of spatial pattern studies re
viewed by Fowler (1986) suggested that clumped and
random patterns were common in arid and semi-arid
plant communities, while regular patterns were rela
tively infrequent. The most popular interpretation of
these patterns has been the "maturation thinning"
model which is nested in competition theory (Phillips
and MacMahon 1981; Prentice and Werger 1985;
Miriti et al. 1998). This model involves a combina
tion of seed dispersal processes that produce aggre
gated patterns in juveniles, and competition that will
convert these into random, and ultimately into regu
lar ones (Phillips and MacMahon 1981). The overall
clumped pattern is then attributed to the fact that there
tend to be greater numbers of juveniles in popula
tions.
There are, however, two weaknesses with this
competition-based model. Firstly, empirical support
for the model is often not convincing. The possibility
that
maturation thinning is occurring has usually been
explored by dividing populations into size classes and
characterizing the spatial patterns of these size classes
separately (Phillips and MacMahon 1981; Prentice
and Werger 1985; Wright and Howe 1987). In these
studies, complete sequences of patterns (i.e. clumped
to random to regular) are seldom resolved. The most
common trend is clumped to random. It is possible to
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
220
argue that if individuals are clumped to start with, and
if some mortality is inevitable (as it
must be even in
the absence of competition) then a tendency towards
randomness may also be inevitable. In addition to
this, the failure to identify individuals growing close
together may often bias the interpretation of patterns
(Ebert and McMaster 1981; Cox 1987). For instance,
Prentice and
Werger (1985) made this error explicitly
by exploring the spatial arrangement of clumps rather
than the spatial arrangement of individuals. The sec
ond weakness with the model is that it neglects the
likely possibility that seed dispersal is subject to se
lection (Dieckmann et al. 1999). Evolutionary theory
suggests that if there is a net disadvantage associated
with living in clumps (e.g. reduced fecundity due to
competition) then dispersal strategies that ensure ju
veniles do not aggregate should evolve.
So, while few would deny that some competition
between plants is axiomatic, several authors have be
gun to question whether competition on its own is a
suitable framework within which to interpret clumped
patterns (e.g. Wright and Howe (1987); Schlesinger
et al. (1990); Couteron and Kokou (1997); Evans and
Belnap (1999); Eccles et al. (1999); Reynolds et al.
(1999). One alternative interpretation of clumped
vegetation patterns is the notion of "pseudointerac
tion" (Garrett and Dixon 1997). These authors argue
that environmental heterogeneity may produce spatial
characteristics that suggest interactions even in the
absence of actual interactions. This idea is compli
cated by the fact that
most biotic interactions are me
diated by changes in the environment (Jones et al.
1997; Schlesinger et al. 1990; Reynolds et al. 1999).
Another more tangible framework for interpreting
clumped patterns is the balance between positive and
negative interactions (Callaway and Walker 1997;
Holmgren et al. 1997; Feldman et al. 1999; Herr? et
al. 1999). According to this notion, plants will have a
preference for clumping if there is some net benefit
associated with occurring in clumps (i.e. if the ben
efits accruing through association with neighbors are
greater than the inevitable competitive costs).
Armed with this basic prediction that plants with a
preference for clumping should derive a net benefit
through association into clumps, and with the ob
served strong tendency towards clumping in strand
veld communities, we set out to establish whether we
could detect advantages associated with living in
clumps. In this study we focused on mature individu
als rather than seedlings and recent recruits. Our ex
perimental approach was to isolate individuals from
their clumps and to compare the performance of these
individuals against matched controls over a year. In
addition to assessing growth (of tagged branches), we
also monitored plant water relations as well as physi
cal damage by herbivory and wind.
Materials and Methods
Study site
The study area is located near the mouth of the Groen
River on the west coast of South Africa (30?51'S
17?34'E) and is part of the Succulent Karoo biome.
Annual average rainfall is in the region of 140 mm,
most of which occurs in
winter. Records of rainfall at
Groen River [station number 01571113, Climate In
formation, Directorate Climatology, South African
Weather Bureau] only began in 1999, indicating an
extremely dry year with a total of 35 mm falling be
tween January and December. Records from an equiv
alent, nearby station [Sarrisam, 01570357] show that
the previous year, 1998, was also drier than average,
with a total of 79 mm falling. Using a composite of
rainfall records from the two stations, the rainfall dur
ing this study period (August 1998 to August 1999)
was about 90 mm. Fog and dew are a regular occur
rence and may contribute significantly to the total
precipitation. Temperatures in the area are generally
moderate. The average maximum temperature in Jan
uary (the hottest month) is about 20 ?C, while the av
erage minimum temperature in June (the coolest
month) is about 9 ?C. Frosts do not occur.
Wind is a very important feature of the climatic
regime (Desmet and Cowling 1999). Wind results in
high evaporative demands (particularly during hot
f?hn-like berg winds); it
may result in physical dam
age to plants directly or through sand blasting (J?r
gens 1996); and may have profound effects on seed
dispersal. In addition to this source of disturbance, the
area is also subject to grazing. Currently this is
mainly in the form of grazing by sheep. Prior to
wide
spread human settlement though, indigenous ante
lope, such as springbok, oryx and red hartebeest,
would have had much the same impact (Cowling and
Pierce 1999). The soils are red aoelian sands overly
ing an impenetrable hardpan of impervious silcrete
about 10 cm thick at about 2 m. The vegetation in the
area is known as the medium strandveld. It is domi
nated by medium-lived shrubs that
may reach 2 m tall
in places. The vegetation clumps are irregularly (or
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
221
"leopard", Aguiar and Sala (1999)) shaped and are
generally separated by between 1 and 2 m of bare
sand. An average clump contains between 10-25 pe
rennial plants 7-11 species and there is a large ele
ment of chance involved in the species composition
of these clumps (Eccles 2000). The most important
families are the Asteraceae and the Mesembryanthe
maceae.
Experimental design and analysis
In order to test the hypothesis that
mature individuals
(as opposed to seedlings and recent recruits) benefit
by living in
multi-species vegetation clumps, we con
ducted a factorial removal experiment. We selected
pairs of individuals matched as closely as possible in
terms of size and clump characteristics (Figure 1)
from eight species. As previously mentioned, a wide
array of simultaneous species interactions are possi
ble within clumps, making individual species interac
tions as they might vary between clumps irrelevant to
explaining our results. Besides the eight species be
ing among the most abundant in this community,
these species exhibited a wide range of physiological
traits (Table 1).
We then experimentally isolated one
of the paired individuals from its clump and left the
other with its clump intact (control). Plants were re
moved by cutting to ground level and lignotubers and
rhizomes were removed by cutting off roots at their
bases with minimal disturbance of the soil. Litter was
raked off. Any resprouts of non-target plants were cut
off during each visit. We replicated these 16 treat
ments (8 species x 2 removal treatments) in three
blocks. Replication was intended to even out any dif
ferences in composition between the controls. The
treatments were imposed in late winter (August
1998).
Growth of five randomly selected marked branches
per plant was carried out between April 1999 and July
1999. Stem elongation was used as the index of
growth. We justified measuring plants during this pe
riod as an analysis of the seasonal frequencies of
growth of another Succulent Karoo site (Goegap)
showed that growth occurs from the onset of rains in
autumn through to late winter. Moderate winter low
temperatures allow winter growth to go on until the
beginning of spring when mean monthly precipitation
again drops (Esler and Rundel 1999).
Assessments of plant water status were carried out
in October and November 1998 and January, April
and late July 1999. At each assessment date we mea
7.5
i 5.0
2
?D
?
2 2.5
o.o
I
I 1
I
Po Sl Ea Su Rf 2m Oc Ss
Species
7.5
a
2
OX)
fi
03
u
CD
5.0
2.5H
0.0 1
x
i
Po Sl Ea Su Rf Zm Oc Ss
Species
Figure 1. Attributes of clumps and individuals that were isolated
(0 plotted against their paired controls (c). a) Clump cover, b)
Clump height, c) number of plants per clump and d) Plant height.
Points falling on the diagonal line would correspond exactly. Bars
represent standard errors. Po = Pteronia, Sl = Salvia, Ea = Erio
cephalus, Su = Stoeberia, Rf = Ruschia, Zm = Zygophyllum, Oc =
Othonna, Ss = Senecio.
sured pre-dawn and mid-day water potentials using a
digital pressure chamber (PMS Instruments, Corval
lis, Oregon, USA, model 1003) and leaf water con
tents [(wet weight - dry weight)/wet weight]. For the
wet weights, leaves were weighed on site or wrapped
in plastic film, kept cool and in the dark, and weighed
within 2 hours of collection. They were then frozen
at -16 ?C to hasten drying, dried at 70 ?C for 48 hours
and then re-weighed for dry weight. A paired i-test
revealed that overall pre-dawn and mid-day leaf wa
ter contents did not differ and we therefore only
present the pre-dawn results. Finally, we noted any
damage to the tagged branches resulting from either
animal activity or wind.
We analyzed the branch growth data by conven
tional ANOVA, treating each plant as the experimen
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
222
Table 1. List of the eight species used in this study along with their families, photosynthetic mode and an index of their succulence (turgid
mass/leaf area ratio).
Species Family Photosynthetic pathway Turgid/area ratio
Salvia lanceolata Lam.
Eriocephalus africanus L.
Pteronia onobromoides DC.
Zygophyllum morgsana L.
Othonna cylindrica (Lam.) DC
Senecio sarcoides (DC.) Sch.Bip.
Stoeberia utilis (L. Bolus) Van Jaarsv.
Ruschia fugitans L. Bolus
Lamiaceae
Asteraceae
Asteraceae
Zygophyllaceae
Asteraceae
Asteraceae
Mesembryanthemaceae
Mesembryanthemaceae
C3
C3
C3
C3
C3
C3
CAM
CAM
0.5484
0.9715
0.9961
1.0111
2.6286
4.1825
4.2853
5.5009
tal unit and using average branch growth per plant
(excluding damaged branches) as the var?ate. In this
way, the data met the assumptions of normality and
homogeneity of variance necessary for a valid
ANOVA. We analyzed the water relations variables
using the repeated measures analysis of variance pro
cedure in the Genstat procedure library (Genstat 5
Committee 1993). This procedure uses maximum
likelihood to estimate an adjustment factor for de
grees of freedom in order to account for non-spheric
ity of data in the "subject time" stratum of the
ANOVA (Greenhouse and Geisser 1959; Everitt
1995). Following on from this analysis, we used nor
mal order plots (Perry 1986; Cousens 1988) to ex
plore groupings of treatments at the "species.treat
ment" level, as an alternative to the usual (but often
misleading) multiple range tests. We opted to use this
approach because we could find no evidence to sup
port our basic hypothesis that plants should benefit
from living in clumps. As such, this exploration was
unplanned. Finally for branch damage, we conducted
an angular transformation on the percentage of dam
aged branches per plant and analyzed this using a
conventional ANOVA. It should be noted that despite
the transformation of this var?ate, the conditions of
normality and of homogeneity of variance were not
met and as such the statistical interpretation is tenta
tive and only presented to reinforce what is clear from
the actual data.
Results
Plant water relations
In terms of water relations our experiment failed to
confirm the hypothesis that plants should benefit by
living in clumps. In the case of leaf water contents
there was no evidence that the isolation of individu
als had any significant effect (negative or positive)
(Table 2). Different species certainly showed clear
differences over time (significant species.time inter
action) (Figure 2). As expected, the water contents in
the more succulent species (Senecio, Othonna, Stoe
beria and Ruschia) were generally higher and gener
ally changed less over the season than the non-succu
lent species (Pteronia, Eriocephalus and Salvia). Al
though Zygophyllum is summer deciduous, prior to
losing its leaves its performance was much the same
as the succulent group. The pre-dawn and mid-day
water potentials were slightly more difficult to inter
pret. Once again there was a significant species.time
interaction (Table 2). Zygophyllum was unusual in
that it behaved more like the non-succulents than the
succulents, which is in contrast to the leaf water con
tents.
In addition to these species.time interactions, the
species.treatment interactions were also significant in
both pre-dawn and mid-day water potentials, and
there was a significant treatment.time interaction in
the case of the mid-day water potentials (Table 2).
The treatment.time interaction reflects the fact that the
treatment effect was more pronounced in mid-sum
mer than in winter. In the case of the species.treat
ment interactions, once again, the general distinction
between the succulent species and the non-succulent
species was apparent. Clumped and isolated individ
uals of the succulent species showed no evidence of
performing differently while the non-succulent ones
did exhibit a treatment effect (Figure 3). Salvia was
an anomaly since its lack of response resembled the
behavior of succulents. Zygophyllum behaved in the
same way as the non-succulents. Where treatment ef
fects were evident, they indicated that isolated indi
viduals were less stressed (higher water potentials)
than their counterparts in clumps. Finally, it would
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
223
t?
53
ti
111
t?
o 70
u
50
40
0
s
-5.0
-7.5 -|-1-1-1-1-1-1-1
0
-2.5
-5.04
* -7.5
-Po
-SI
-Ea
-Su
-Rf
Zm
-Oc
-Ss
jan apr jul
Month
Figure 2. Time series showing a) rainfall and b) the development
of the significant species time interactions for pre-dawn leaf water
contents and c) pre-dawn and d) mid-day water potentials. The
species codes are as in (Figure 1). In the case of Zygophyllum, this
species was summer deciduous and there was therefore no data re
corded in
April.
appear that within the succulent group (Senecio, Oth
onna, Stoeberia and Ruschia), the CAM Mesembry
anthemaceae species (Stoeberia and Ruschia) gener
ally had higher water potentials than the C3 Aster
aceae (Senecio and Othonna).
Plant growth and damage
While there was once again a significant species ef
fect on branch growth, there was no significant treat
ment main effect or species.treatment interaction (Ta
ble 3). The two Mesembryanthemaceae species (Stoe
beria and Ruschia) responded most to the light rains
that fell during this growth period (Figure 4). At the
other end of the spectrum, Pteronia hardly showed
any response to the rains. In terms of plant damage, it
is clear that isolating plants exposes them to a much
greater risk of being damaged (p < 0.001, (Table 4)).
Senecio individuals in particular were highly suscep
tible to browsing damage when isolated from their
clumps. The damage in
Eriocephalus, and Salvia iso
lated plants appeared to have been caused by either
sheep walking through the plants or wind damage,
while the damage in Stoeberia and Zygophyllum re
sulted from sheep using the plants as a scratching
post. The damage in isolated Pteronia and Othonna
plants appeared to be the result of animals brushing
past or wind.
Discussion
Our working premise was that clumped vegetation
patterns, such as those present in the strandveld, most
likely represent situations where the benefits accrued
by plants through their association with neighbours
are greater than the inevitable competitive costs. In
line with this notion of net effect, the vari?tes that we
measured did indeed reflect the full spectrum of in
teractions (negative in the case of water potentials in
some species; neutral in the case of leaf water con
tents and growth; and positive in the case of physical
damage). In support of the findings of Briones et al.
(1998) in a Chihuahuan system, occurring in clumps
had little or no effect on plant water relations in the
more succulent species studied (Senecio, Othonna,
Stoeberia and Ruschia). In contrast, the less succulent
species (Pteronia, Eriocephalus, Zygophyllum but not
Salvia), showed some evidence of competition for
water in the form of lower (more negative) water po
tentials in clumped individuals. Lamont and Lamont
(2000) recently showed that these two groups of spe
cies were fundamentally different in their water up
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
224
Table 2. Repeated measures ANOVA results for pre-dawn and mid-day water potentials and pre-dawn leaf water contents. Values are the
variance ratios. When assessing the significance of these, the degrees of freedom under the site.subject.time stratum were multiplied by the
Greenhouse-Geisser (G-G) epsilon value.
Source of variation df Pre-dawn water potential/P Mid-day water potential/P Pre-dawn water content/P
Site stratum 2
Site.Subject stratum
Species 7
Treatment 1
Species.treatment 7
Residual 30
Site.Subject.time stratum
Time 4
Time.species 28
Time.treatment 4
Time.species.treatment 28
Residual 128
Total 239
G-G epsilon
0.27/> 0.05
186.6/< 0.0001
16.91/< 0.0001
5.64/< 0.001
1.31
218.5/< 0.0001
23.10/< 0.0001
1.91/>0.05
1.04/>0.05
0.7814
5.00/> 0.05
189.9/< 0.0001
19.29/< 0.0001
4.93/< 0.001
1.88
148.3/< 0.0001
13.63/< 0.0001
4.87/< 0.025
1.39/>0.05
0.6396
0.34/> 0.05
170.8/< 0.0001
0.4/> 0.05
0.7/> 0.05
2.33
411.7/< 0.0001
57.65/< 0.0001
0.76/> 0.05
0.82/> 0.05
0.7594
take and holding properties. There were, however, no
differences in leaf water contents between clumped
and isolated individuals in any species. If we assume
that leaf water content is a better indicator of plant
water status than water potentials (Turner 1981; La
mont 1999), it would appear that the non-succulent
species are able to minimize the effects of negative
interactions by means of some form of osmotic ad
justment. Osmotic adjustment is a response whereby
tissues at lower water potentials are able to attract
more water to them, maintaining a higher water con
tent and avoiding detrimental water losses. Compen
satory responses to stress at similar time scales (an
entire season) have recently been reported in other
desert systems (Reynolds et al. 1999). In addition,
while one would assume that such compensatory re
sponses are associated with metabolic costs, these are
not manifest in any significant growth response. This
suggests that the costs are either insignificant, or that
some other growth related benefit balances them out.
This still leaves the question of why occur in
clumps and it would appear that our least rigorously
assessed var?ate, namely damage, may hold the key.
Prior to human colonization, large herds of ungulates
were widespread in the succulent and adjacent Nama
Karoo region, migrating to and from areas of high and
low productivity (e.g. Springbok, Skinner 1993).
These herds would have heavily impacted local areas
on an intermittent and unpredictable basis (Cowling
and Pierce 1999). With the advent of perennial pas
toralism in the eighteenth and nineteenth centuries,
the impact of domestic ungulates is believed to have
contributed to the decline in productivity of the re
gion. In addition, domestic herbivores are seen as be
ing one of the major threats to biodiversity, changing
the relative abundance of species through selective
herbivory (Joubert 1971; Todd and Hoffman 1999).
These herbivores may certainly have influenced the
structural characteristics of the vegetation, such as
clumping which may confer benefits in terms of pro
tection. Of the 24 isolated plants in our experiment,
no less than 13 (or 54%) were damaged to some ex
tent by the end of the experiment. This is in contrast
to about 4 (or 16%) in the clumped plants. The de
gree and type of damage ranged from broken
branches and isolated browsing to wind throw of
whole plants and severe browsing. Although flower
ing and fruiting were not measured, the loss of
branchlets in isolated plants is likely to have an ad
ditional impact on the reproductive potential of all
species. The exception is
Ruschia, which is accessible
to herbivores even when it is clumped, as it occurs in
low clumps or at their edges. It can be argued that by
isolating mature individuals that had benefited from
associational defenses for most of their lives, our ex
perimental design may have exaggerated this effect.
These plants would not have had the need to invest
in structural growth or defense prior to treatment and
reallocation of resources following treatment would
obviously have been impossible over the time frame
of a single season. However, this does not weaken our
main contention that these plants are faced with mak
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
225
CU ?
won
.s s
fi
c?
"S
5-,
4
3
2
1
Poc
Zmi
Poi
Eac
Oci Ssi Ssc
2^H
.2
S 3
Cl
"3 2H
1
0
Zmi"
Eac
Zmc
Poi
Oci &i
-1 1
Normal order score
Figure 3. Means of the 8 x 2 treatments positioned along the x
axis according to their normal var?ate scores for n = 16, and along
the y-axis according to the actual values for a) pre-dawn and b)
mid-day balancing pressures (-water potentials). The values are
means for all 5 sampling times. The solid line has a slope equiva
lent to the standard error of the mean and intersects the y-axis at
the overall mean. The dotted lines are simply parallel lines (i.e.
same slope). Points falling along any one of these lines form natu
ral groups that behave similarly. The points are coded with the
species codes in Figure 1 together with an "i" or a "c" indicating
the treatment (isolated or clumped).
ing trade-offs between the costs and benefits of living
in clumps as opposed to living in the open. On the
costs side, competition for water is likely to present a
problem for non-succulent species, although we have
shown that these plants are able to compensate physi
ologically for this even during a relatively dry year.
In even drier years, clumping might be beneficial due
to a) mutual shading by overlapping pants (2-5 over
lapping plants per target plant in our study, Authors
unpublished data), b) soil cooling by the extra litter
and/or c) extra capture of dew and fog by the greater
concentration of foliage, stems and litter. On the ben
efit side, plants are exposed to much lower risks of
a
C3
7.5
5.0
2.5
0.0
i
I i
i
Po Sl Ea Su Rf Zm Oc Ss
Species
7.5
a
2
W)
?
fi
5.0
2.5
0.0 x
x
i
Po Sl Ea Su Rf Zm Oc Ss
Species
Figure 4. Box and whisker plots of branch growth for the eight
species in a) clumped and b) isolated situations. Once again spe
cies codes are as in Figure 1.
Table 3. Results of an analysis of variance (two-way ANOVA) for
branch growth data, treating each plant as the experimental unit and
using average branch growth per plant (excluding damaged
branches) as the variate.
Source of variation df P-value
Species
Treatment
Species treatment
0.001
0.564
0.332
damage and are free to allocate resources that might
otherwise have gone to structural or defense functions
into other activities. In wetter years, water would not
be as limiting, so that mutual protection from her
bivory might become the more important function,
especially as more palatable foliage would then be
available.
Whether or not this simple 'balance sheet' is
enough to conclude that plants experience a net ben
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
226
Table 4. Percentage of tagged branches (out of a total of 15) that
were damaged between April and late July (14 weeks).
Species Isolated Clumped
Pteronia 13.3 0.0
Salvia 26.7 0.0
Eriocephalus 26.7 0.0
Stoeberia 33.3 0.0
Ruschia 0.0 6.6
Zygophyllum 13.3 0.0
Othonna 13.3 0.0
Senecio 66.7 26.7
efit through association with other plants though is
still open to debate. It is possible that the net interac
tion in adults is neutral and that stronger positive in
teractions during other life stages (e.g. seedlings)
have, over evolutionary time, resulted in the observed
preference to clump. Several authors have presented
data to suggest that establishment in
many desert spe
cies is facilitated by the presence of existing vegeta
tion (Yeaton and Manzanares 1986; Yeaton and Esler
1990; Valiente-Banuet and Ezcurra 1991). However,
very often these facilitative relationships become
competitive during adult stages and a cyclical succes
sion phenomenon emerges. While there is little doubt
that facilitation during establishment is a feature of
the strandveld (around 80% of recruitment occurs un
der clumps, Authors unpublished data), it does not
appear that cyclical succession occurs.
To conclude, our results provide evidence that
adult plants in this arid system simultaneously expe
rience both positive (reduced herbivory) and negative
(lower water potentials) interactions with other plants.
While our evidence does not constitute proof that
plants experience a net overall benefit (although it
points in that direction), we are confident that at worst
the interactions between adults are neutral. During the
experimental period, rainfall was lower than average;
thus any detrimental effects of clumping on water
content or growth rates should have been highlighted
when in fact they remained negligible or non-existent.
In addition, we feel justified in advancing the circular
argument that if, over the entire life span of individu
als, the net interactions were not usually positive, then
plants would not occur in clumps because there would
have been strong selection against this.
Acknowledgements
This work was funded by the National Research
Foundation (South Africa) in the form of a grant to
KJE, a fellowship to BL, and scholarship to NE. It
was made possible by access to a Mazda Wildlife
Fund courtesy vehicle. Thanks to Richard Cowling
for his comments on a draft. NE is an adjunct research
fellow at Curtin University of Technology.
References
Aguiar M.R. and Sala O.E. 1999. Patch structure, dynamics and
implications for the functioning of arid ecosystems. TREE 14:
273-277.
Briones O., Monta?a C. and Ezcurra E. 1998. Competition inten
sity as a function of resource availability in a semiarid ecosys
tem. Oecologia 116: 365-372.
Callaway R.M. and Walker L.R. 1997. Competition and facilita
tion: a synthetic approach to interactions in plant communities.
Ecology 78: 1958-1965.
Couteron P. and Kokou K. 1997. Woody vegetation spatial patterns
in a semi-arid savanna of Burkina Faso, West Africa. Plant
Ecol. 132: 211-227.
Cowling R.M. and Pierce S. 1999. Namaqualand. A Succulent
Desert. Fernwood Press, Cape Town.
Cousens R. 1988. Misinterpretation of results in weed research
through inappropriate use of statisitcs. Weed Research 28: 281?
289.
Cox G.W. 1987. Nearest-neighbour relationships of overlapping
circles and the dispersion pattern of desert shrubs. J. Ecol. 75:
193-199.
Desmet P.G. and Cowling R.M. 1999. The Climate of the Karoo
A functional approach. In: Dean W.R.J. and Milton S.J. (eds),
The Karoo: ecological patterns and processes. Cambridge Uni
versity press, Cambridge, pp. 3-16.
Dieckmann U., O'Hara B. and Weisser W. 1999. The evolutionary
ecology of dispersal. TREE 14: 88-91.
Ebert T.A. and McMaster G.S. 1981. Regular pattern of desert
shrubs: A sampling artifact? J. Ecol. 69: 559-564.
Eccles N.S. 2000. Plant ecology of the Namaqualand Strandveld:
community structure and dynamics in a winter-rainfall desert.
Eccles N.S., Esler K.J. and Cowling R.M. 1999. Spatial pattern
analysis in Namaqualand desert plant communities: evidence
for general positive interactions. Plant Ecol. 142: 71-85.
Esler K.J. and Rundel P.W. 1999. Comparative patterns of phenol
ogy and growth form diversity in two winter rainfall deserts:
the Succulent Karoo and Mojave desert ecosystems. Plant Ecol.
142: 97-104.
Evans R.D. and Belnap J. 1999. Long-term consequences of dis
turbance on nitrogen dynamics in an arid ecosystem. Ecology
80: 150-160.
Everitt B.S. 1995. The analysis of repeated measures: a practical
review with examples. The Statistician 44: 113-135.
Feldman R., Tomback D.F. and Koehler J. 1999. Cost of mutual
ism: Competition, tree morphology, and pollen production in
limber pine clusters. Ecology 80: 324-329.
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
227
Fowler N. 1986. The role of competition in plant communities in
arid and semiarid systems. Ann. Rev. Ecol. Syst. 17: 89-110.
Fuentes E.R., Otaiza R.D., Alliende M.C., Hoffman A. and Poiani
A. 1984. Shrub clumps of the Chilean matorral vegetation:
structure and possible maintenance mechanisms. Oecologia 62:
405-411.
Garrett K.A. and Dixon P.M. 1997. Environmental pseudointerac
tion: The effects of ignoring the scale of environmental hetero
geneity in competition studies. Theoret. Pop. Biol. 51: 37-47.
Genstat 5 Committee 1993. Genstat 5 Release 3 Reference Manual.
Oxford Science Publications, Oxford.
Greenhouse S.W. and Geisser S. 1959. On methods in the analysis
of profile data. Psychometrika 24: 95-112.
Herr? E.A., Knowlton N., Mueller U.G. and Rehner S.A. 1999. The
evolution of mutualisms: exploring the paths between conflict
and cooperation. TREE 14: 49-53.
Holmgren M., Scheffer M. and Huston M.A. 1997. The interplay
of facilitation and competition in plant communities. Ecology
78: 1966-1975.
Jones CG., Lawton J.H. and Shachak M. 1997. Positive and nega
tive effects of organisms as physical ecosystem engineers. Ecol
ogy 78: 1946-1957.
Joubert J.G.V 1971. 'N Voorlopige vergelyking tussen twee be
weidingstelsels op weskus strandveld. Hand. Weidingsveren. S.
Afr. 6: 78-81.
J?rgens N. 1996. Psammophorous plants and other adaptations to
desert ecoystems with high incidence of sand storms. Feddes
Repertorium 107.
Lamont B. 1999. Pressure volume curves. In: Atwell B.J., Kriede
mann P.E. and Turnbull C.G.N. (eds), Plants in Action. Mac
Millan Australia, South Yarra, Vic, pp. 465-467.
Lamont B.B. and Lamont H.C. 2000. Utilizable water in leaves of
eight arid species as derived from pressure-volume curves and
chlorophyll fluorescence. Physiologia Plantarum 110: 64-71.
Miriti M.N., Howe H.F. and Wright S.J. 1998. Spatial patterns of
mortality in a Colorado desert plant community. Plant Ecol.
136: 41-51.
Perry J.N. 1986. Multiple-comparison procedures: a dissenting
view. J. Econ. Ent. 79: 1149-1155.
Phillips D.L. and MacMahon J.A. 1981. Competition and spacing
patterns in desert shrubs. J. Ecol. 69: 97-115.
Prentice I.C. and Werger M.J.A. 1985. Clump spacing in a dwarf
desert shrub community. Vegetatio 63: 133-139.
Reynolds J.F, Virginia R.A., Kemp PR., de Soyza A.G. and Trem
mel D.C. 1999. Impact of drought on desert shrubs: effects of
seasonality and degree of resource island development. Ecol.
Monographs 69: 69-106.
Schlesinger W.H., Reynolds J.F., Cunningham G.L., Huenneke
L.F, Jarrell W.M., Virginia R.A. et al. 1990. Biological feed
backs in global desertification. Science 247: 1043-1048.
Skinner J.D. 1993. Springbok (Antidorcas marsupialis) treks.
Trans. Roy. Soc. S. Afr. 48: 291-305.
Todd S.W. and Hoffman M.T. 1999. A fence-line contrast reveals
effects of heavy grazing on plant diversity and community com
position in Namaqualand, South Africa. Plant Ecol. 142: 169?
178.
Turner N.C. 1981. Techniques and experimental approaches for the
assessment of plant water status. Plant Soil 58: 339-366.
Valiente-Banuet A. and Ezcurra E. 1991. Shade as a cause of the
association between the cactus Neobuxbaumia tetetzo and the
nurse plant Mimosa luisana in the Tehuac?n valley, Mexico. J.
Ecol. 79: 961-971.
Wright S.J. and Howe H.F. 1987. Pattern and mortality in Colo
rado Desert plants. Oecologia 73: 543-552.
Yeaton R.I. and Esler K.J. 1990. The dynamics of a succulent ka
roo vegetation. A study of species association and recruitment.
Vegetatio 88: 103-113.
Yeaton R.I. and Manzanares A.R. 1986. Organization of vegetation
mosaics in the Acacia schaffneri ? Opuntia streptacantha as
sociation, southern Chihuanhuan Desert, Mexico. J. Ecol. 74:
211-217.
This content downloaded from 134.7.248.132 on Tue, 25 Mar 2014 01:55:25 AM
All use subject to JSTOR Terms and Conditions
