Nc7c I'hytol. (148')), i n, 2.1.

^ 243
Ecophysiology of xerophytic and halophytic
vegetation of a coastal alluvial plain in
northern Venezuela
I. Site description and plant communities
BY E. MEDINA' , W, J. CRAM", H, S. J. LEE'^^ U. LUTTGE' ' *, M. POPP' ',
J. A, C. SMITH' ' AND M. DIAZ'
' Centro de Ecologia y Ciencias Ambientales, Instituto Venezolano de Investigaciones
Cientificas, Caracas 1020-A, Venezuela
'^Department of Biology, The University, Newcastle upon Tyne, NEl 7RU, UK
'^ Institut fur Botanik, Technische Hochschide Darmstadt, D-6100 Darmstadt, FRG
^ Institut fiir Pflanzenpliysiologie der Universitdt, A-1091 Wien, Austria
^ Institut fiir Angewandte Botanik, Westfdlische Wilhelms- Universitdt,
D-4400 Munster, FRG
*' Department of Botany, University of Edinburgh, Edinburgh, EH9 3JH, UK
' Centro de Investigaciones en Ecologia y Zonas Aridas, Universidad Naciottal
Experimental Erancisco de Miranda, Coro, Venezuela
{Received 8 March 1988; accepted 22 July 1988)
SUMMARY
This paper describes the ecology of a coastal alluvial plain at Chichiriviche in northern Venezuela. The area
supports a great diversity of plant communities, ranging from mangroves on the seaward edge of the plain to non-
halophytic, fresh-water communities on the landward side. Small differences on topography result in a mosaic of
saline and less-saline environments. Rainfall is strongly seasonal, causing superficial flooding of the alluvial plain
in the rainy season and the creation of a hypersaline substratum during the dry season. As a consequence, much
of the plain is devoid oi vegetation. Towards the landward side of the plain there are numerous small 'vegetation
islands', fringed by halophilic succulent herbs, and made up of deciduous and semi-deciduous shrubs and trees
together with non-halophytic CAMplants such as cacti and bromeliads. In subsequent papers the results of
ecophysiological studies of these diverse plant species are presented.
Key words: CAM plants; epiphytes; halophytes, mangroves; salt flat; Venezuela,
INTRODUCTION habitat patchiness. Deciduous and semi-deciduous
non-halophytic shrubs and trees coexist with man-
Salt flats associated with estuarine mangroves under groves and shrubby and herbaceous true halophytes,
strong seasonal rainfall in the tropics constitute an Within this heterogeneous vegetation terrestrial and
interface between terrestrial halophytic and non- epiphytic CAMplants may constitute a significant
halophytic vegetation. Proximity to the coast and percentage of the total biomass, as is the case of the
seasonal distribution of rainfall result in pronounced columnar cacti and opuntias, and Bromehaceae of
seasonal variation in soil sahnity and availabihty of the genera Tillandsia (epiphytic) and Bromelia
superficial fresh water. Under these conditions an (terrestrial). Coexistence of species such as succulent
array of higher plants with contrasting ecophy- halophytes, with high chloride concentrations in
siological properties coexist due to microtopo- their leaf sap, and succulent CAMplants, with
graphical heterogeneity resulting in a high degree of negligible chloride content, has been explained on
* To whomall correspondence should be addres,sed, the basis of microhabitat differentiation and dif-
234 E. Medina and others
1 CHI CHI RI VI CHE
2 TOCUYO DE LA COSTA
3 CURAMI CHATE
* STODY SI TE
Figure 1. Location of the study site on the northern coast ot Venezuela
ferential distribution of active absorbing roots
(Walter, 1973). Non-halophytic cacti and terrestrial
bromeliads have supeflcial root systems which are
active only during the rainy season, when superficial
soil salinity is washed away by free running water.
Shrubby and herbaceous halophytes on the contrary,
are able to obtain water and nutrients from deeper
soil layers, with higher chloride content throughout
the year. The present project aims to document the
variety of ecophysiological types co-occurring in such
an environment, and characterise differentiation of
microhabitats. The field work was undertaken
during the months of November-December 1985
and March-April 1986.
create a mosaic of saline and less or non-saline soil
environments. Areas in lower geomorphological
levels are heavily flooded during the rainy season but
become brackish and salty as the dry season
progresses. These areas may be completely devoid or
vegetation, probably as a result of the strong seasonal
changes in salinity and flooding (Walter, 1973;
Walter & Breckle, 1984). Towards the Chichiriviche
bay the influence of sea water becomes more
pronounced and flooding tends to be permanent.
There, a typical estuarinc mangrove vegetation
dominates the landscape. At the other extreme,
deciduous forest vegetation surrounds the areas
flooded seasonally by the Tocuyo river.
STl' DV ARK A
The study area, locally known as the Cienega el
Ostional, is located about 6 km west of the town of
Chichiriviche in Estado Falcon in north-western
Venezuela ( 1O55' N, 68 21' W) (Fig. 1). The
whole area is made up of alluvial sand plains and
harrier beaches created by the Tocuyo and Tucurerc
rivers (Goddard & Piccard, 1972). Beaches extend
for about 30 km from Chichiriviche in the direction
of Curamichate. These alluvial plains are period-
ically flooded by the Tocuyo river, exerting a
superfical salt washing effect. Saline influence arises
from upward percolation of sea water from the
Chichiriviche bay. Small differences in topography
CLIMATE
Long-term climatoiogical data are available from the
town of Tocuyo de la Costa, approximately 16 km
north-west of the study site (Fig. 1) for the period
1963-86. Annual rainfall averages 1029 mm. The
annual distribution of rainfall (Fig. 2) shows two
clear peaks, a small one in April and a very
pronounced one in November. There is no rainless
month, even in the driest years, and only the months
of February and March present average rainfall
helow 40 mm. This climate type is very favourable
for succulents (Walter, 1973), including epiphytes,
which have the possibility of recharging their \yater
reserves shortly after significant rains. Potential
Ecophysiologv of xerophvtic and halophvtic vegetation in Venezuela. I 235
E
d 200
o
Q.
a
o 100
o
o
CC
TOCUYO DE LA COSTA (20m)
(1963-1985) o Rai nf al l av 1020 mm
Evaporation av 2325 mm
A Temperature av 26 4 "C
O
o
30 S.
E
cu
2 0 ' "
10
J F M A M J J A S O N D
Figure 2. Axera^c tnonllily \alue.s of raitil'all, cxaporatioti
and temperature tiir a 22 years period as measured in the
meteorological station of the Ministerio del .Ambiente in
Tocuyo dc la Costa, Estado I'alcon, Venezuela.
evaporation is very high throughout the year and is
below the rainfall curve only during November and
December. The evaporation curve shows two small
peaks, the first in March and the second in August,
in both cases just before the onset of the respective
rainy periods, associated with the apparent path of
the sun towards and from higher latitudes during the
summer in the northern hemisphere. Lowest evap-
oration rates are registered in the period November
to February, when they reach 5-5+ 0-3 mm d^';
during the rest of the year daily evaporation is signifi-
cantly higher reaching 6-4 + 0-3 mmd"'. Average
monthly temperature varies very little during the
year; daily temperature differences between mean
maxima and mean minima (7-1 C on average) are
larger than the differences between mean tempera-
tures of the hottest and the coldest month (August
and January respectively), which amounts to 2-1 C.
The period analysed shows a pronounced rainfall
Variation between successive years (Fig. 3). This
variation is of ecological importance because it can
regulate the expansion or contraction of the halo-
phobic vegetation through the degree of surface-soil
Washing occurring every year. Besides, occurrence of
humid years probably causes erosive impacts which
rnay reduce the soil level in the vegetation covered
islands, thereby decreasing survival of non-halo-
phytic elements. The 23-year observation period can
be roughly diflerentiated into two phases. The
period between 1964 and 1975 was rather humid
vv'ith a rainfall average of 1170 348 mmy"'. The
period from 1976 to 1986 was much drier with a
rainfall average of 856 + 357 mm y"\ Out of 23
years, 7 years had rainfall above average plus twice
the standard error of the mean, and of those, 5
occurred before 1975. In contrast, 7 out of 9 years
with rainfall below average minus twice the standard
2000-
TOCUYO DE LA COSTA {20m)
64 68 72 76
Ye a r s
80 84
Figure 3. .\tintial variations in total rainfall registered
the last 2.^ \ears mthe TocuNO df la Costa station.
ISO
150-
70-
60-
50-
40-
_ 30
E
l 2 0
10
AL
15 20 25 30
October
5 10 15 20 25 30 5 10 15
November December 1985
15 20 25 28 5 10 15 20 25 30 5 10 15
February March April 1986
Figure 4. Daily rainfall distributioti during the study
period of the rainy seasoti ot 1985 and the dry season ot
1986.
error occurred after 1975. However, variability is
high. In the month of November 1985, during the
period when the first series of field measurements
were undertaken, rainfall was nearly twice the long-
term average recorded in Fig. 3. Rainfall figures
registered durmg the periods of fieldwork charac-
terize the strong seasonal distribution of rain (Fig. 4).
During November 1985, 8d had rainfall x'alues
above 25 mm, among them 1 d with more than
150 mm.
V II G K T A T I O N
The vegetation in the Cienega el Ostional is
diflerentiated between the extremes of mangroves
and deciduous forest. Between these two extremes a
mixture of halophylic and halophobic vegetation is
236 E. Medina and others
Figure 5. General vegetation map of the study area
developed from aerial photographs provided by the
Ministerio del Ambiente, Caracas. L, lagoon with brackish
water; M, tall mangrove vegetation; B, deciduous forest,
G, secondary grassland; I.V., island vegetation complex;
vf, mainly vegetation-free, flooding-prone areas; C, long
term fresh-water flooded areas dominated by sedges; S,
the broken line indicates the border of areas of pre-
dominantly saline soils in the east from higher non-flooded
areas in the west.
found (Fig. 5). The lowest geomorphological level is
occupied by a lagoon (L) with brackish water and is
completely surrounded by tall mangrove vegetation
(M) [mainly Rhizophora mangle L. and Avicennia
germinans (L.) Stearn and scattered Laguncularia
racemosa CJaertn.]. The highest geomorphological
level is occupied by several areas of deciduous forest
(B) and grassland (G), probably of secondary origin,
dominated by the grass, Paspalum plicatulum Michx.
The area between the mangroves and the deciduous
forests contains a mosaic of vegetation units desig-
nated by us as an island-vegetation complex (I.V.)
where very slight changes in the soil level, in the
order of 5-20 cm, determine changes in salinity and
flooding and the coexistence of halophylic and
halophobic plant species. Between the mangrove
area and the island-vegetation units extend areas
completely devoid of vegetation or occupied by
sparse pure stands of Batis maritima L. and/or
Sesuvium portulacastrum L. (vf). In some areas fresh
water accumulates probably as a result of impeded
drainage by the elevated road. These areas are
occupied by almost pure stands of a few species of
Cyperaceae (C).
Vegetation units emerging from the plain, ex-
cluding pure mangrove associations, can be dif-
ferentiated as follows:
(1) Flooded areas with fresh water accumulated
during the rainy season. These areas are dominated
by species of Cyperaceae such as Eleocharis geniculala
R. Br. Prod., E. nutans R. Br. Prod., Fymbristylis
cymosa Vahl and F. spadicea Vahl. Some species
such as Nyniphaea sp., with floating leaves in open
water, are ephemeral and disappear during the dry
season (Fig. be).
(2) A Batis maritima-Sesuvium protulacastrinn
unit, bordering the vegetation-free salt fiats, which
can be temporarily flooded after heavy rains [Fig.
6(a) and (6)].
(3) A grassland elevated between 5-10 cm from
the salt flat dominated by Sporobolus virginicus
Kunth, intermixed with Oxycarpha suaedifolia Blake
and isolated individuals of the cacti, Opuntia zven-
tiana Britton & Rose and Acanthocereus tetragonus
(L.) Humik. (Fig. 6a).
(4) Small islands, 3-10 m in diameter, where the
soil surface can be 10-40 cmhigher than the salt-fiat
[Fig. 6(c) and {d)~\. The vegetation on top of these
islands is frequently dominated by a mangrove
associated species, Conocarpus erectus L. (Tom-
linson, 1986), reaching approximately 2 min stature,
frequently accompanied by a columnar cactus Sub-
pilosocereus ottonis Backeb., and a broad-joined
Opuntia species (unit 4a in Table 1). On isiands
without C. erectus a number of isoiated non-
halophylic tree species such as Prosopis fuliflora
D.C., Capparis hastata Jacq. and Maytenus karstenit
Reiss. may be found. Another occasional species is
the cactus, Pereskia guamacho Weber, a small
deciduous tree or shrub covered with succulent
fleshy ieaves ciuring the rainy season (unit 4/; in
Tabie 1). As an epiphyte on the shrubs and cacti, the
bromeliad Tillandsia flexuosa Sw. is abundant. In
islands nearby and in deciduous lorest relicts, tiie
orchid Schomburgkia humboldtiana Reichb. is also
frequently found. The floor of these islands may be
covered by dense thickets of Bromelia humilis Jacq-
or scattered tufts of Sporobolus virginicus.
The contact between the islands and the sand
plains on lower ground is occupied by the Balis-'
Sesuvium community. During the rainy season the
reddish-flowered Portulaca riibricaulis H.B.K., with
well developed underground tubercules, is quite
often seen.
(5) The deciduous woodland. These areas are
represented hy deciduous forest remnants of variable
size which have been clearly isolated from the main
forest core by erosive influences of running water
during humid years, but also to some extent by
human infiuence (Fig. bf). The erosion reduces the
soil level thereby increasing salinity percolation from
underground brackish water in those areas con-
stituting the transition zone between saline and non-
Ecophysiology of xerophytic and halophytic vegetation in Venezuela. I 237
Figure 6. Vegetation units within the study area, {a) General view of the alluvial plains showing Sporobolus
virginicus and Acanthocereus tetragonus (foreground), Batis maritifna and Sesuvium portulacastrum (middle) and
Conocarpus erectus. Background shows the main core of deciduous forest towards the north-east border of the
study area, (6) Edge of the vegetation zone with Sesuvium potulacastrum leading to higher ground with mixed
grass species, (c) View of the island vegetation in the rainy season, (d) View of a typical island during the dry
season, (e) Example of fresh water vegetation with Nymphaea sp, and Eleocharis geniculata. (f) Remnants of
deciduous forest vegetation.
saline soils. The borders of these forested areas
^re characterised by the presence of Capparis
odoratissima Jacq,, C. hastata, Capparis cf. pachaca
Jacq. (broad, round leaves), Caesalpinia coriaria
Willd, Prosopis juliflora, Jacquinia revoluta Jacq,,
^aytenus karenii, Erythroxylon cumanense H, B, K,
3nd shrubs of Croton sp, and Pereskia guamacho.
Occasionally there are dense thickets of the terrestrial
Dromeliads B. humilis and B, chrysantha Jacq. As
Epiphytes the bromeliads Tillandsia flexuosa and T.
''ecurvata L,, and the orchids Schomburgkia hum-
"oldtiana and Brassovala nodosa Hook, are commonly
found.
A profile through the study area which includes all
the vegetation units described above and a number
of the most representative species shows the species
distribution pattern associated with topographical
position and the fluctuation of the water table
(Fig, 7),
SOI LS
It is clear that distribution of vegetation units is
determined by superficial soil salinity resulting from
underground water percolation, which varies sea-
sonally because of rain washing. During the rainy
238 E. Medina and others
metres 20 AO 60 80 100 120 UO 160 180 200 220 240 260 280
Bare soil
Conocarpus erectus
Cupparis flexuosa
Prosopis juliflora
Eugenia sp.
Croton heliaster
Capparis odoratissima
Mimosa oligocantha
Capparis hastata
Maytenus karstenii
Trichilia trifoliata
Guapira sp.
Capparis Iinearis
Fimbristylis spadicea
Fimbristylis cymosa
Carex sp.
Desmanthus virgatus
Paspalum pUcatulum
Leptochioa sp.
Sporobolus virginicus
Acanthocereus tetragonus
Opuntia wentiana
Subpiiosocereus ottonis
Bastardia viscosa
Cissus sicyoides
Anthurium crassinervium
Philodendron sp.
Evolvulus tenuis
Sesuvium portulacastrum
Batis maritima
Oxycarpha suaedifolia
Bromeiia humilis
Tillandsia flexuosa
Phoradendron mucronatur
Tillandsia recurvata
cm
150
100
50
island
deciduous forest
halophytes grassland remnant grassland halophytes
- soil pit
water table - rainy season
soil pit
water table -dry season
Figure 7. Transect indicating location of the three soil pits excavated for measurement of ground-water depth
and vertical variations in soil salinity. The transect reveals distribution of most frfqiicnt plant species as related
to topographical variations.
season differences in soil salinity (expressed as
chloride content) among vegetation units are re-
latively small but it is clear that salinity decreases
from the salt fiats towards the deciduous forest
remnants (Table 1). Differences among vegetation
units increase markedly during the dry season. In
April 1986, in the top 0-10 msoil, salinity decreases
from a high of 340 mequiv. Cr kg"' soil in the salt
flats to a low of 18 mequiv. Cr kg"' in tbe deciduous
forest remnants. The sodium/cbloride equivalent
ratio was O-88 + O-O3, indicating tbat most of tbe salt
in these soils is NaCI. Seasonal variation in salinity is
quite large in tbe salt flats, particulary in the
Batis-Sesuvium zone (unit 2) and the Sporobotus
zone (unit 3). Those are the typical halophylic plant
communities along the Caribbean coast in Ven-
ezuela. Least variation in salinity is observed in the
remnants of the deciduous forests (unit 5). In-
dividuals of Subpitosocereiis ottonis are found only in
tbese areas vvbcre the superficial soil does not
increase markedly in chloride concentration during
tbe dry season. However, Opuntia sp. often occur
within Sporobolus grasslands, therefore their roots
reach into the soil layer of high salt content during
the dry season. Individuals oi B. humilis are found in
a wider range of superficial soil salinity. For this
species however, soil chloride content is of no
consequence because functional roots are not in tbe
soil but grow between the leaf sheaths into the tank
constituted by tbe leaf bases (Pittendrigh, 1948).
Soil profiles were excavated during the dry season
(in March and April 1986 and 1987) to determine the
vertical variations in nutrients and salinity and to
measure tbe depth of the ground water (Fig. 8).
Sodium and chloride were by far tbe dominant ion
species in tbe profile. The ground water from the
tbree vegetation sites (units 1, 4a and 4-b) had
significantly higher concentrations than sea water
collected nearby (Fig. 9). In all cases Cl con-
centration increased witb soil depth (Fig. 10).
Moreover, there vvere considerable differences in the
profiles depending on the location in relation to the
vegetation units. Down to 0-50 m in a profile dug
near a 6-5 mtall Subpiiosocereus ottonis at the border
of a deciduous forest remnant (Soil pit II, Fig. 9),
salinity was about half of that measured in a profile
in the salt plain near a Conocarpus erectiis-covercd
island (Soil pit I). At lower levels, nearer the lagoon
Ecophysiology of xerophytic and halop/iytic vegetation in Venezuela. I
239
Table 1. Seasonal changes in superficial soil chloride contents. Cienega el Ostional Chichiriviche, Stado Falcon
(see text for description of vegetation units)
(1) Vegetation-free salt flats
(2) Batis + Scsuiiuni zone
(3) Spnrobolus + Oxxcarpha zone
4() Conocarpus island with
Bromelia humilis
4(6) Non-halophilic trees and
Bromelia sp. underneath, moss
covered soil surface
(5) Deciduous forest with
Bromelia sp. beneath
Chloride content
Depth (cni)
0-10
10-20
0-10
0-10
10 20
0-10
10-20
0-10
10-20
0-10
10-20
in nict|uiv. kj
Dry season
340+ 133
219 + 49
203 + 107
110 + 87
70 + 26
46 + 34
114+149
28 + 20
44 + 9
18 + 21
21+27
X ' dry soil
Rainy season
65
107
37
17

11-36

11

16

Na/Cl equivalent ratio = 0-88 + 0-03

Figure 8. Depth of ground water in soil pit III of Figure
7 excavated during the dry season of 1986. The whole
profile shows a predominantly sandy fine clay con-iposition.
Red colour is an indication of the presence of ferric oxide
and grey of ferrous oxides which indicates pronounced
seasonal oscillations of the ground water.
(Soil pit III) salinities were even higher. Ground
water was reached at a depth of 1-05 min the profile
in the S. ottonis site, at 1-10 mat the ishmd and at
0-70 mat the vegetation-free plain (Soil pits II, I and
III respectively. Fig. 9). Osmolalities of the ground
2iOO
2000
1500
1200
n
in
I - s o i l pi t I
D - - II
sea water
K'Co'Mg"
Figure 9. Ionic composition of ground water in the dry
season frompits excavated near a 6-5 mtall Subpilosocereus
ottonis (Soil pit II, Fig. 7) at the border of a small island
dominated by a Conocarpus erectus tree (Soil pit I, Fig. 7),
on the alluvial plain (Soil pit III) and sea water from the
Chichiriviche bay.
water were 1-73, 1-75 and 2-30 osmol kg"' respec-
tively, corresponding to an osmotic pressure of
more than 4-5 MPa, i.e. about 1-8 times that of sea
water. At a depth of 0-50 m the profile near the
lagoon (Soil pit III) had a much higher salinity,
approaching 600 mequiv. kg ' soil. Ground water
was already struck at 0-70 m as stated above.
Differences it-i ground water level indicate that the
vegetation islands are located about 40-45 cmabo\-e
the water table of the lagoon (see Fig. 7). The
vegetation-free plains are salt fiats during the dry
seasoi-i and mud flats in the rainv season. Thev are
240 E. Medina and others
100
100 200 300
meq Cl " / kg soil
400
Figure 10. Variations of chloride concentrations with
depth in soil pits. I, Conocarpus profile (A); H, Cactus
profile ( ) ; HI, Batis profile (O) (cf. also Fig. 7).
flooded after heavy rain [Fig. 6(c) and (e)]. In the dry
season of 1983 most of the vegetation-free plain was
covered by a continuous salt crust (Fig. 11 a). This
was the dry season following the driest year in the 22
year period shown in Fig. 3 and was exceptionally
dry in the whole Caribbean region (Smith, Griffiths
& Luttge, 1986). During the period of the present
investigation, in the dry season of March to April
1986, a continuous salt crust was not observed,
although crystallized salt was frequently found on
top and in the upper layers of the soil (data in Table
1). Extensive formation of salt crusts was observed
again in the dry season of 1987 (Fig. 11 b), following a
year with below average rainfall (see Fig. 3).
ORIGIN OF THE ISLAND VEGETATION AND
CO- OCCURRENCE OE HALOP 111L I C AND
HALOPHOB IC VEGETATION
The distribution of vegetation units, as related to
topographical positions and soil salinity, clearly
indicates that soil salinity is very heterogeneous,
contrasting soils occurring side by side. In most
cases however, non-saline patches are slightly ele-
vated above the salt flat level. This level difference
ranges from 5 to 40 cm. Moreover, salinity in the
upper soil layers changes dramatically throughout
the year. During the rainy season superficial salt is
washed away, while it is concentrated through
evaporation during the dry season. This last situation
was considered by Walter (1973) when explaining
the co-occurrence of halo-intolerant succulent cacti
with succulent halophytes rooting in the same soil.
We have observed similar behaviour in Opuntia
wentiana, a salt-intolerant succulent cactus and the
halophytes Sporobolus virginicus and Oxycarpha
suaedifolia (unpublished data). The cacti grow and
recharge their water reserves only during the rainy
season when salinity is low (Table 1).
One of the main objectives of the Chichiriviche
project is to explain the vegetation dynamics in the
salt flats and to clarify the origin of island vegetation.
Are those islands in the process of formation, gaining
land from the inhospitable salt flats, or do they
represent phases of disintegration of vegetation units
originally growing on salt-free areas? Three hypo-
theses are worth considering.
(1) Walter (1973, pp. 411-12) assumes that the
succession leading to the formation of the vegetation
islands starts with individual rosettes of B. humilis
Figure 11. Extensive salt crust development ohserved in the study area during the dry seasons of 1983 (a) nnd
1987 (b).
Ecophysiology of xerophytic and halophvtic vei^etatioit in Jetteziiela. I
241
" pWtej^vsj,
Figure 12. A.spects of island vegetation witliin tlie alluvial plains, (a) Typical small island with Sidtfyi/osoccrcus
ottonis and Capparis odoratissima with Boniclici huniilis utiderneath. In the backgrotitid the core of the deciduous
forest with which these types of v'egetation islands seem to be originally associated, (h) .A.spect of plants growing
on slightly elevated ground within the salt flat {Sporohoius virffiniciis and Sesuvium portulacastrum). (c), (d)
Aspect of disintegrating island during the rainy and the following dry season respectively. Notice that two
Bromeiia humilis specimens atid the small Subpiiosocereus ottonis died in the intervening period. The tree
remnants behind also itidicate that this partictilar island is in the process ot decay.
transported to the sand | Jain by gu.sty wind.s. 'Iliis is
possible because the tank torminj^ rosettes of B.
humilis do not need to root in the substratum and
only loosely He on it for tiiccbanical support.
According to Walter's hypothesis tbese rosettes
would then form large colonies by \-egetati\e pro-
pagation. The wind could accumulate sand around
these colonies so tbat otber plants, including cacti,
may establish themselves. However, from pro.,
ductivity measurements performed in tbe present
project, it appears improbable that a single rosette of
B. humilis loosely laid down on tbe sand could give
rise to a community of rosettes rapidly enough to
form a new island,
(2) Vegetation-free areas may be occupied by salt-
toleratit herbs, sbrubs and/or trees (such as Spo-
robolus virginicus, Batis maritima atid Conocarpus
erectus) wbicb, by their presence, cause accumulation
of soil in their surroundings so ele\'ating tbe soil level
above the salt Hat itself. Tbis cle\ation of the soil
level allows leaching of salts during tbe rainy season
tbereby creating favourable conditions for tbe estab-
lishment of salt-intolerant plant species.
(3) Salinity variations due to topographical pos-
itions are associated witb tbe erosive effects brought
about through flooding by the river and after beavy
rainfalls. It is proposed that tbe deciduous forest soil
is being eroded away leading to tbe formation of
deciduous forest patcbes surrounded by soils prone
to salinizatit)n from percolation of sea water. Tbis
may explain the occurrence of salt-intolerant trees
and shrubs in small islands surrounded by balopbytic
vegetation.
Field observations suggest tbat both processes of
aggregation and disintegration of \-egetation islands
as stated in bypotheses (2) and (3) are co-occurring
in tbe area. Small islands, constituted by salt-
intolerant species, appear to be separated from tbe
main core of deciduous forest as a result of erosive
processes [bypotbesis (3), Fig, Ma]. Conversely,
slightly elevated areas arising from changes in
bydrological patterns may be invaded by salt tolerant
species. This is an alternative mode of island
formation [bypotbesis (2), Fig. \2h\.
While hypotheses (2) and (3) and Fig 12/; and a are
concerned witb tbe formation of vegetation islands.
Figure 12(c) and {d) show island disintegration
wbicb may be due to continuation of erosion as in
. \ M' 11 I
242 E. Medina and others
hypothesis (3) of island formation. Conversely,
degradation may occur with juvenile islands formed
according to hypothesis (2), i.e. for such cases young
islands are unstable.
STUDIES AND AIMS OF THE CHt CHl Rl VI CHE
PROJECT
The studies of the Chichiriviche project presented in
this series of articles were primarily devoted to
comparative ecophysiology of plants in the rainy and
the dry season and to the physiological changes
brought about hy the transition between seasons.
Several species were studied. They represent an
array of life-forms such as halophyte, mangrove,
deciduous and evergreen shrubs, tank-forming bro-
meliads, leaf and stem succulents and epiphytes.
They comprise species with C.|-photosynthesis and
crassulacean acid metabolism (CAM). The com-
parative ecophysiology led to synecological con-
siderations particularly by the work on the vegetation
islands (units 4a and b). Each island covered a
relatively small area on the plain and possessed a
limited number of species. Thus, the species selected
for measurements were representative and allowed a
comparative evaluation of adaptive strategies of co-
occurrence. Terrestrial species of the island and the
Batis-Sesuvium zone (unit 2) bordering them, which
are all affected by seasonal changes in salinity of the
substratum, are described in separate papers: II, III,
V and VI.
The dominant shrub of the islands, Conocarpus
erectus, was investigated in greater detail. It also
occurs in sand and sand dunes close to the sea shore
and these were included in the study. At such
locations, fundamentally different from the plains of
the Cienega et Ostional, C. erectus also behaves in a
different way and has been compared with a salt
excreting mangrove growing in the same area,
Avicennia germinans (Paper VI).
Another monographic treatment is devoted to B.
humilis (Paper III). It is one of the most typical and
abundant terrestrial plants of the islands but it also
occurs within the fringe of the deciduous forest
remnants (unit 5). This tank-forming hromeliad is
not directly exposed to the substratum since mostly
it has few or no functional roots in the soil. It only
uses the substratum as a support, occasionally even
becoming semi-epiphytic, resting on other vege-
tation from a few decimeters up to a 1 mabove the
ground. The species occurs in habitats with con-
trasting light climates, frompartially shaded habitats
under tree groves, where the plants are larger and
greener, to fully exposed areas, where plants are
smaller and yellowish in colour, as described by
Medina, Olivares & Diaz (1986).
^^he true epiphytes studied (Paper IV), the bro-
meliad Tillandsia flexuosa and the orchid Schom-
burgkia humboldtiana, are both CAMplants. For
these epiphytes, frequency of rainfall is more
important than amount. This factor is particularly
critical for S. humboldtiana, which has to absorb
water when it is available through the root velamen
and accumulate it in succulent bulbs and leaves. T.
flexuosa forms a small tank with its leaf sheaths
allowing the retention of rain water for longer
periods. T. flexuosa leaves are heavily coated with
trichomes which are highly effective in absorbing
rain water.
Finally, investigations of mangroves were not
limited to the plants oi A. germinans occurring on the
plains of the Cienega el Ostional. Fxtensive sampling
of mangroves including A. germinans, R. mangle and
L. racemosa along the lagoon were performed during
the rainy season in November/December 1985, and
during the dry season in March/April 1986 (Paper
VII, in preparation).
ACKNOWLEDGEMENTS
This project was a multinational endeavour with the
participation of the following institutions: (1) Institut fCir
Botanik der Technischen Hochschule, Darmstadt, FRG;
(2) Centro de Ecologia, Instituto Venezolano de In-
vestigaciones Cientificas, I VIC, Caracas, Venezuela; (3)
Institut fur Pflanzenphysiologie der Universitat Wien,
Austria; (4) The Department of Plant Biology, The
University of Newcastle upon Tyne, UK; (5) Centro de
Investigaciones de Zonas Aridas, Universidad Fransisco
de Miranda, Coro, Venezuela. The project was supported
mainly by grants from the Deutsche Forschungsgemein-
schaft, Bonn-Bad Godesberg, FRG (U. L. and J. A. C. S. )
and Fonds zur Foderung der wissenschaftlichen Forschung
(project no. 5784), Austria (M.P.). Additional funds were
received from The Royal Society of London (W. J. C. and
H. S. J. L. ), the Dax Copp Travelling Fellowship of the
Institute of Biology (H. S.J. L.) and the Studienstiftung des
deutsclien Volkes (C.S.). Local laboratory facilities and
supplies were provided by the Centro de Ecologia of IVIC.
Dr R. Wingfield and Lies G. Carnevali and G. Colonello
assisted in the identification of plant samples. Meteoro-
logical data was supp'ied by the central Falcon office of
the Ministerio del Ambiente y Recursos Naturales. Rosel
Heger prepared the half-tone pictures and Doris Schafer
and Lelis Oehoa the india ink illustrations. Technical
assistance was supplied by Frika Ball, Eileen Goundry,
Sabas Perez and Elisa Martinez. Thanks also to Dr J.
Gorham (University College of North Wales) for ion
exchange chromatography of soil and water samples.
REEERENCES
GODDARD, D. & PICCARD, X. (1972). Geomorfologia y sed-
imentacion en la costa del Estado F'alcon, Cabo San Roman a
Chichiriviche. Memoria Segundo Congreso Latinoamericano
do Geologia. Tomo II. Boletin de Geologiii. Publ. Esp. No. 7,
1157-1179.
MEDINA, E. , OLIVARES, E. & DIAZ, M. (1986). Water stress and
light intensity effects on growth and nocturnal acid ac-
cumulation in a terrestrial CAMbromeliad {Brometia humilis
Jacq.) under natural conditions. Oecologia 70, 441-446.
PnTENDRiOH, C. S. (1948). The Bromeliad-Anopheles-Malaria
Eeophvsiologv of .xerophvtic and halophvtic vegetation in Venezuela. I 243
complex in Trinidad. 1. Tho bronieliad Hora, Evolution 2, WALTER, ]]. {\'-)73). Die l'i%'ct(ition der Ertie in okopliysiologisclier
58-89, Betrrichtuiif;. Band 1 : Die tropischeii iiiul suhtropmhen Zonen.
SMITH, J, A, C, GRIFFITHS, H, & LITTGF, U, (1486), Com- VF.B Gustav Fischer Verlag, Jena,
parative ecophysiology of ChM and C., bromeliads, I, The W.M.TFR, 11, & HRFCKLE, S,-\V, (1984), OWoi'/f rfc) AV(A', Band 2:
ecology of the Iirotncliaceae in Trinidad, Plant, Cell and Spezielle Okologie der tropischen und siibtropischen Zonen, pp.
Environment 9, 359 376, 180 182, Gimta\' iM.scher, Stuttgart,
ToMLiNSON, P, B, (1986), The Botany of Mangroves. Cambridge
University Press, Cambridge, L ,K,