Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

Plant Ecology (2005) 181:127–137  Springer 2005

DOI 10.1007/s11258-005-5321-x
0

Gypsum physical soil crusts and the existence of gypsophytes in semi-arid


central Spain

Roberto L. Romão1 and Adrián Escudero2,*


1
Universidade Estadual de Feira de Santana. Laboratório de Ecologia Evolutiva da, Av. Universitária, s/n.
44031-460 Feira de Santana – BA, Brasil; 2Biodiversity and Conservation Group. E.S.C.E.T. Universidad
Rey Juan Carlos. C/. Tulipán s/n. Móstoles E-28933. Spain; *Author for correspondence (e-mail: adrian.
escudero@urjc.es; phone: +34-91-4887070)

Received 22 January 2004; accepted in revised form 7 April 2005

Key words: Edaphic specialist, Emergence, Physical soil crusts, Seedling growth, Seedling survival

Abstract

Probably gypsophytes are the most interesting set of edaphic specialists of arid and semiarid climates.
Despite they conform a global biodiversity priority, there are almost no information about those adaptive
traits that confer such a specialised behaviour. Our broad hypothesis is that gypsophytes are ‘‘refuge-
endemics’’ that are able to grow on gypsum soils due to their ability to surpass extremely hard gypsum soil
physical crust during emergence. With this in mind we have conducted an experimental approach com-
bining field and greenhouse assays. Seeds from two gypsophytes, genuine and widely distributed in the
Iberian Peninsula gypsophytes (Helianthemum squamatum and Lepidium subulatum) and one gypsovag
(Teucrium capitatum), a generalist plant that can also grow on gypsum soils were used in our experiments.
Two complementary experimental approaches were conducted. The first involved a field experiment in
which the presence or absence of the physical crust together with the sowing date were manipulated and a
greenhouse experiment in which the irrigation amount and the types of soil were controlled. Variables of
interest were the percentage of germination, growth and survival. In the field experiment we found a
significant decrease in the final germination of the gypsovag in the plots with intact crusts. On the other
hand, H. squamatum is able to grow in the three tested soils, despite higher survival and growth on genuine
gypsum soils. Our results confirm the hypothesis that gypsum edaphic specialists base their behaviour to a
great extent on the ability to surpass extremely hard gypsum surface crusts, although this seems a marginal
adaptive trait as shown by the capability to grow on a complete array of soils and the negative effect of the
crust along the earlier development life stages of gypsophytes. Furthermore, a gypsovag such as Teucrium
capitatum presents extreme difficulties to emerge on non-disturbed gypsum physical crusts but once sur-
passed its growth and survival is not limited.

Introduction Garcı́a-Moya 1989; Mota et al. 2003). These


plants exclusively grow in gypsum (hydrous cal-
Gypsophytes probably represent the most con- cium sulfate) soils under arid or semiarid climates.
spicuous and diversified set of arid and semiarid Surprisingly, such type of specialists have received
endemic plants (Johnston 1941; Parsons 1976; remarkably little study (Meyer 1986; Meyer et al.
Powell and Turner 1977; Meyer 1986; Meyer and 1992; Escudero et al. 1999, 2000) in spite of
128

gypsisols which are characterised by a gypsum where they are competitively excluded under
content over 5%, extending over one million km2 natural conditions. In order to test these two
in the world (Verheye and Boyadgiev 1997). In complementary hypotheses, we have conducted a
addition to this, most of these plants conforms a two-fold experimental approach combining field
global biodiversity priority because most of them and greenhouse experiments. If our gypsum
are narrowly distributed and seriously threatened physical crust hypothesis is correct, gypsophytes
(Meyer 1986). In these sense, Mediterranean gyp- should germinate, grow and survive on crusted
sum ecosystems are considered one of the most soils whereas gypsovags, plants that marginally
threatened habitats in the Mediterranean Basin can grow on gypsum soils, should perform much
(Gómez-Campo 1987; European Community better on finely raked gypsum soils (crust experi-
1992; Mota et al. 2003). mentally crushed). On the other hand, if gypso-
Historically, gypsophily has been related to the phytes are not soil specialists and exploit gypsum
ability of gypsophytes to persist under certain environments as marginal habitats where they are
chemical restrictions of gypsum soils (Duvigneaud competitive, they should germinate and recruit
and Denaeyer-De Smet 1966, 1968; Boukhris and individuals not only on gypsum soils but also on
Loissant 1970; Cannon 1971). However several other types of soil.
field studies have shown little evidences of this type
of restriction (Meyer 1986; Meyer and Garcı́a
Moya 1989; Meyer et al. 1992; Verheye and Boy- Methods
adgiev 1997). It has been also suggested that the
specialist behaviour of gypsophytes lies to a great We have selected three conspicuous shrubby plants
extent in their capacity to surpass extremely hard of the gypsum communities of central Spain (Rivas-
gypsum physical surface crusts immediately after Martı́nez and Costa 1970; Rubio et al. 1994). He-
germination (Campbell and Campbell 1938; Meyer lianthemum squamatum (L.) Cours (Cistaceae) and
1986; Escudero et al. 1999, 2000). So, emerging Lepidium subulatum L. (Cruciferae) are small
gypsophytes surpass the gypsum soil crust, shrubs (20 – 60 cm) that exclusively grow on gyp-
whereas non-gysophytes are unable to. Our spe- sum outcrops. These two plants are typical ele-
cific hypothesis is that gypsophytes take selective ments of the so-called Gypsophiletalia sthrutii
advantage due to this capability, whereas gypso- communities, which are endemic but widely dis-
vags are plants that able to grow on gypsum but tributed on gypsum outcrops under semiarid con-
not exclusively, only recruit individuals in the ditions in the Iberian Peninsula (Loidi and
vicinity of standing gypsophytes where gypsum Fernández-González 1994). However, Teucrium
physical crusts are narrower; probably via facili- capitatum L. (Labiatae) which also is a small shrub
tation (Escudero et al. 2000). It is widely known can grow on gypsum soils but preferentially on the
that gypsum crusts are notoriously thicker and widely distributed calcareous soils from the Eastern
harder than those found in other arid and semiarid half of the Iberian Peninsula. This gypsovag has
habitats (Verherye and Boyadgiev 1997). Ad- been selected because it is one of the most out-
ditionally, during the past few years we have standing elements both in gypsum and calcareous
accumulated field evidences of notorious gypso- soils in the vicinity of our field experimental site.
phyte emergences and survival on extremely well
conserved hard-crusted gypsum surfaces (Escu-
dero et al. 1999; 2000; Romão 2003). Experimental approach
Our broad hypotheses are (1) the existence of a
profound link between gypsophily and the pres- Experiment one – The field study site was located
ence of thick non-perturbed gypsum physical in Chinchón, Madrid province, within an excep-
crusts that only can be surpassed by genuine gy- tionally well-conserved remnant of a southern
psophytes at least on massive gypsum outcrops, exposed gypsum habitat (4011¢ N, 335¢ W,
and (2) the fact this behaviour is a marginal 550 m a.s.l.). The climate is upper semiarid with an
adaptive trait, so gypsophytes constitute ‘‘refuge’’ annual average rainfall of 415 mm but almost no
endemics (in the sense of Gankin and Major 1964) rainfall in summer. Mean daily maximum and
and consequently are able to grow on other soils minimum temperatures in January are 9.6 C and
129

0.6 C, respectively, and 32.7 C and 9.6 C in July. Sowing was conducted in January immediately
Soils are Calcic gypsisols developed over gypsum before field emergences under natural conditions
parental rocks (Monturiol and Alcalá del Olmo (Escudero et al. 1997). Seeds were collected during
1990) with gypsum contents above 50%. Vegeta- the previous summer from 20 to 30 plants per
tion is conformed by conspicuous vegetated pat- species randomly selected within the experimental
ches dominated by creeping chamaephytes area and stored at 6 C in darkness in hermetically
interspersed on skeleton crusted surfaces covered sealed containers (see Escudero et al. 1997). Before
by specialised lichens. Cover of perennial plants is sowing, seeds were soaked in distilled water in
20% in the 1000 m2 where the experimental plots order to promote the development of seed muci-
were set. The dwarf scrub community has been lage. The existence of mucilaginous seed coats is a
named Herniario fruticosae-Teucrietum pumili widely extended trait in gypsum environments and
(Rivas-Martı́nez and Costa 1970). The bare zone is these three species have mucilaginous coats
covered by a biological crust (50%) and some (Escudero et al. 1997). Such a type of seed trait
gypsum crystals (5%). Soil biological crust is determinates both a favourable environment for
dominated by Diploschistes diacapsis, Squamarina germination and an efficient mechanism to avoid
cartilaginea together with some gypsum specialised washing, so seeds remain anchored at the sowing
lichens such as Toninia sedifolia, Fulgensia subb- point. In the case of the ‘intact crust’ plots lichens
racteata and Psora decipiens and with almost non were carefully eliminated in the sowing hole when
fix-nitrogen elements such as Collema crispum. necessary to avoid bias due to the nature of the
This biological crust is extremely hard with an lichen species.
average strength in May of 1230 kPa (SE = 350, Field germination was regularly recorded till
n = 20) measured with a portable penetromer October because the most critical period for
(RIMIK CP-20, Australia). During the summer seedling survival corresponds to the first summer
drought the physical crust reaches significantly drought. Second and third germination per hole
higher values (Escudero, personal observation). were noted but seedlings were immediately clipped
Experimental plots consist 30 cm · 30 cm qua- to avoid pernicious seedling competition. Seedling
drates randomly distributed on bare (without survival and growth was also recorded and mea-
perennial cover) and flat (<5%) soil surfaces. sured as the number of leaves (Helianthemum
Plots were randomly assigned to two contrasting squamatum) and total length (Lepidum subulatum
and fixed treatments. In the first, named ‘‘intact and Teucrium pumilum) at each census.
crust’’, the soil surface was unaltered, whereas in As shown in (Escudero et al. 1999) survival is
the ‘‘disturbed crust’’, soil surface was evenly higher for the first Helianthemum squamatum
raked to destroy the gypsum crust in January cohort. In order to evaluate the relationships
2000. Finely divided microphytic rests were not between germination time and crust constraints we
retired from the quadrates. Six plots per treatment conducted a similar experiment with Helianthe-
and plant species were considered. Twenty five mum squamatum but delaying the sowing time to
sowing circles (15 mm in diameter) were regularly the first week of March. The new source of vari-
distributed in a transparent plastic sheet in order ation in the corresponding linear models was the
to conform a 5 cm side-cell grid. The plastic sheet emerging time (two fixed levels: January and
was used for sowing and monitoring purposes but March sowing). During the germination window
not left on soil. Three seeds per cell grid per species (January to May) the strength of the gypsum crust
were sown in middle January within each plastic remains rather constant and relatively low.
hole conforming a set of 75 seeds per quadrate. Experiment two – In a parallel greenhouse
Seeds were separated to avoid interferences during experiment the specialist behaviour of a widely
the emergence. Seeds were sown on the soil surface distributed gypsophyte such as Helianthemum
and not buried. Previous works on gypsophyte squamatum was also surveyed. A two-way factorial
germination responses showed that final germina- experiment was conducted. Soil type was consid-
tion percentage reached more than 50% in the ered the first source of variation with three fixed
three considered species (Escudero et al. 1997), so levels, gypsum soils and calcareous soils which
the number of sown seeds should guarantee at were collected in the field in the vicinity of the
least one seedling per sowing point once emerged. above mentioned locality (soil data in Romão
130

2003), and an enriched commercial substratum probability of one seed yielding an emerged seed-
(COMPO SANA Universal). The second source of ling per pot (‘‘seedling emergence rate’’), proba-
variation was the amount of irrigated water. It bility of survival at June and probability of
must be reminded that the gypsophile behaviour is survival at September were also assessed by fitting
related to the amount of rainfall, being confined to generalised linear models. These variables were
arid and semiarid climate conditions (Rivas- modelled as binomial response due to their nature
Martı́nez and Costa 1970). Two fixed levels were and difficulties to normalise, using logits as the
considered, a control which mimics the amount of link function. As variables were sequentially in-
rainfall during spring and early summer in our cluded, we tested different sequences and nested
field locality (12 ml per pot) and a second treat- models (Mc Cullagh and Nelder 1989) and sub-
ment which double this last amount (one irrigation sequently compared. Models were adjusted using
per week). Twenty-five pots of 10 · 10 · 10 cm S-Plus 2000.
were filled per treatment combination. Five seeds
were sown per pot in January. This density is
similar to that found under natural conditions in
favourable microsites (Escudero et al. 1999). Results
Provenance and conservation of seeds were similar
to the above mentioned experiment. Germination, Experiment one
survival and growth were also weekly recorded
during 6 months. As shown in Figure 1 and Table 1, Lepidium
subulatum germination did not show differences
between treatments (average final germination =
Statistical analysis 43.38% ± 2.82), nor survival was affected (sur-
vival in May = 72.39% ± 8.09; in June =
Experiment one – Normality of the variables – 35.31% ± 8.17; and in September = 7.17% ±
percentage of germination, survival at June and 4.35). On the other hand Teucrium capitatum seeds
survival at September per plot was studied emerged significantly better without crust (final
through the Kolgomorov – Smirnov test germination = 29.11% ± 4.11) than on crusted
(p < 0.1). Appropriate transformations were soils (5.11% ± 2.31) (Figure 1 and Table 1).
applied to improve normality when necessary. However survival was rather similar between
Linear models were independently conducted for treatments (survival in May = 68.85% ± 7.37; in
each species with two fixed treatment, crust (1 June = 48.40% ± 8.84, and in September =
d.f.) and germination date (1 d.f.). When nor- 8.98% ± 3.16). Germination of Helianthemum
mality was not reached, non-parametric tests squamatum seeds only showed significant differ-
were conducted. ences for the sowing treatment (Figure 1 and
Experiment two – Normality of variables were Table 2), being higher the final germination for the
also checked – average plant size per pot in June, first cohort (final germination = 53.00% ± 3.29
size in September, average number of leaves per for the January cohort and 33.77% ± 2.37 for the
plant and pot in June and number of leaves per March cohort). After germination, survival did
plant in June – and two-way anovas was con- not present significant differences neither for the
ducted on log-transformed data. To account crusted treatment nor for the sowing moment
statistically for possible differences among pots in (survival in May = 71.14% ± 3.64; in
their response to experimental treatments due to June = 54.73% ± 3.96; and in September =
competition, the number of seedlings per pot was 16.06% ± 3.22).
included in the model as a covariate (substra- Size in June was extremely different in the case
tum*water*number of seedlings per pot). As this of L. subulatum (U = 1633, p < 0.0001) being
study focuses on the consequences of experimental higher in the disturbed and crushed crust treat-
manipulations, significance tests associated with ment, marginally different in H. squamatum
covariate in models are not considered. (U = 1089, p = 0.036) following a similar trend,
The significance of the effects of sowing sub- and rather similar in the case of T. capitatum
stratum (substratum) and irrigation (water) on the (U = 145, p = 0.487)
131

Figure 1. Mean value and standard error for the original data of germination in disturbed and intact crusts and taking into account the
sowing date for H. squamatum (field experiment).

Table 1. Results from the experiment I (field conditions) for Lepidium subulatum (gypsophyte) and Teucrium capitatum (gypsovag).

Plant Species Variable df F p-value

Lepidium subulatum Germination percentage 1,10 0.685 0.427


May Survival 1,10 0.268 0.618
June Survival 1,10 3.09 0.108
September Survival v 2 = 3.578 0.058
Teucrium capitatum Germination percentage 1,10 13.5 0.000
May Survival 1,10 2.10 0.177
June Survival 1,10 2.06 0.181
September Survival v 2 = 0.117 0.731

Survival in September could not be properly transformed, so a non-parametric test was conducted.
132

Table 2. Results for H.squamatum in the experiment I (field conditions). Models included crust treatment (intact and disturbed crust),
and sowing cohort (January and March).

Variable Source of variation df F p-value

Germination percentage Crust treatment 1, 20 1.12 0.301


Sowing cohort 1, 20 14.62 0.000
Interaction 1, 20 3.06 0.095
May Survival Crust treatment 1, 20 3.76 0.066
Sowing cohort 1, 20 0.61 0.000
Interaction 1, 20 1.69 0.207
June Survival Crust treatment 1, 20 0.000 0.992
Sowing cohort 1, 20 0.865 0.363
Interaction 1, 20 1.007 0.327
September Survival Crust treatment 1, 20 0.012 0.914
Sowing cohort 1, 20 2.275 0.147
Interaction 1, 20 0.593 0.450

Experiment two 1.00 ± 0.40; log transformed), but marginally


higher in the enriched soil also in September (Ta-
Treatments showed no differences in the Helian- ble 3).
themum squamatum average size per pot in June Maximum likelihood analysis of variance of
(average size per pot – log transformed = 0.71 ± dependent variables for the effects of sowing sub-
0.17) and only for substratum in size in September stratum (Substratum) and irrigation (Water) were
(Table 3). Pairwise comparisons between substra- obtained adjusting generalised models as binomial
tum levels clearly show that size in enriched com- responses (Table 4). Seedling emergence rate
mercial soils becomes significantly larger showed no differences among treatments. How-
(1.57 ± 0.64; log-transformed p < 0.000) than in ever, survival in June and survival in September
gypsum soils (log-transformed size = 0.88 ± 0.17) showed highly significant differences among sub-
or calcareous soils (0.87 ± 0.15). These last two strata (Figure 2). Survival in gypsum soils is
levels do not exert significant differences. We have always significantly higher, mainly in relation to
also no detected differences in the average number enriched soils. A significant interaction between
of leaves per plant and pot in June (mean = irrigation and substratum was detected in both

Table 3. Results of the experiment II (greenhouse conditions).

Variable Source of variation df F p-value

Size in June Substratum 2,181 0.424 0.655


Water 1,181 0.525 0.470
Substratum*Water 2,181 2.117 0.124
Size in September Substratum 2,181 4.54 0.012
Water 1,181 0.616 0.433
Substratum*Water 2,181 0.397 0.673
Leaves in June Substratum 2,181 0.092 0.912
Water 1,181 1.623 0.203
Substratum*Water 2,181 2.363 0.097
Leaves in September Substratum 2,181 4.011 0.020
Water 1,181 4.392 0.047
Substratum*Water 2,181 2.560 0.080

The treatments included were the different soil substrata (gypsum, calcareous soils and enriched commercial substratum) and water
(two levels of irrigation). All variables were log transformed to fulfil normality criteria. Average values per pot (n = 50 per treatment
combination).
133

Table 4. Maximum likelihood analysis of variance of dependent variables from experiment II for the effects of sowing substratum
(Substratum) and irrigation (Water) were obtained adjusting generalised models as binomial responses (logit as link function).

Variable Source of variation df Chi-square p-value

Seedling emergence rate Substratum 2 1.269 0.530


Water 1 0.227 0.633
Substratum*Water 2 1.59 0.450
Survival at June Substratum 2 42.13 0.0000
Water 1 0.032 0.856
Substratum*Water 2 15.07 0.0005
Survival at September Substratum 2 39.29 0.0000
Water 1 1.29 0.254
Substratum*Water 2 8.58 0.013

cases, mainly because under double irrigation about the origin of gypsophily (see Meyer 1986;
treatment, survival in calcareous soils is signifi- Escudero et al. 1999). Field evidences had previ-
cantly lower (Figure 2). In the case of the survival ously shown that the two studied gypsophytes,
in September the reduced model (without the Lepidium subulatum and Helianthemum squamatum
interaction) was not a good reduction of the sat- are able to recruit in extremely hard gypsum sur-
urated model (v2=15.07, p = 0.0005). faces (>1500 kPa in gypsum crests) under a wide
array of climatic conditions (Escudero et al. 1999;
2000). Such observational data seems to support
Discussion for our experimental results. They confirm that
emerging radicles of gypsophytes can surpass the
Many authors have highlighted the relevance of gypsum physical crust without significant difficul-
biological crusts in arid systems (Watson 1979; ties. However, gypsovag Teucrium capitatum has
Belnap et al. 2001). They redistribute water (Eld- enormous difficulties to emerge (only 5% in aver-
ridge et al. 2000; Maestre et al. 2002), enhance soil age) under crusted conditions whereas not on un-
stability (Belnap and Gillette 1998) and the activity crusted soils. Although Bridges and Burnham
of microphytes improves soil resource levels (Bel- (1980) pointed out that roots of generalists plants
nap and Harper 1995; Zaady et al. 1998) deter- were not able to penetrate in soils with more than
mining ecosystem function and dynamics (West 25% of gypsum (gypsisols) because the soil mate-
1990), and resilience, facilitating ecosystem recov- rials lack plasticity, cohesion and aggregation, and
ery after severe droughts (Aranibar et al. 2003). is completely mechanically unstable, we consider
However several sowing experiments have shown that the most relevant fact is the inability of
an important reduction in germination on undis- emerging seedlings to surpass the gypsum crusts
turbed crusts (Zaady et al. 1997). It is also known and not the lately seedling or developed plant. It is
that the development of desert physical crusts is one necessary to note that the tested effect is only re-
of the most limiting factors for the emergence of lated to the extremely hard physical constraint of
new seedlings under field conditions (Hanks and the gypsum crust, because the filtering effect of the
Thorp 1956; Harper et al. 1965; Cary and Evans biological components of the crust on any life stage
1974). Additionally, Sylla (1987) and Johansen of vascular plants remain unknown. In any case this
(1993) have suggested that the inhibition mecha- non-explored effect should be additive because well
nism in microphytic crust is related to the difficul- conserved biological crusts always present extreme
ties of the penetration of radicles and seeds into the strength values. Lichen crust cover appears high
soil below the crust layer. Prasse and Bornkann and positively correlated with the strength of the
(2000) pointed out that the effect of crusts on plant gypsum crust (Tarazona et al. 1980).
is extremely diverse and probably include a trade- Other two evidences reinforce the physical
off between negative effects at the germination stage hypothesis: first, Merlo et al. (1997) and Romão
and the above mentioned benefitial features. Our (2003) have not found significant differences in
results clearly support the physical hypothesis laboratory germination under a complete set of
134

Figure 2. Mean value and standard error for the original data (not our binomial model) of survival in June and September in
experiment two (greenhouse experiment).

gypsum dilutions, showing that the presence of hand, Teucrium capitatumis relatively abundant in
gypsum and or calcium content do not alter the the gypsum outcrops of the studied area, but
germinability in these two gypsophytes, and sec- confined to vegetated patches where genuine
ond, gypsophytes are confined to arid or semi-arid gypsophytes are dominant (see Romão 2003).
climates where gypsum crusts exclusively appear There are no multi-species patches dominated by
(Rivas-Martı́nez and Costa 1970; Meyer 1986). T. capitatum nor isolated individuals in non-dis-
Under more mesic climates, gypsum physical turbed community (Romão 2003).
crusts do not appear and the gypsum outcrops, As previously reported (Escudero et al. 1999),
sometimes with percentages of sulfates in soil far timing of emergence is a key factor for growth and
above 50%, are covered with calcicole vegetation consequently for survival in many desert plants
(Rivas-Martı́nez and Costa 1970). On the other (see Fowler 1988; Eldridge and Westoby 1991), but
135

as shown here, also for the germination of primary refuges on this edaphic islands because its
H. squamatum. Germination is significantly higher capability to surpass gypsum crusts at the emer-
for the first sowing cohort. Factors underlining gence stage but also has evolved adaptive strate-
this behaviour might be related to a temperature gies to perform better on such soils, as it occurs in
control because most of the Iberian gypsophytes the case of calcicole plants. Note that under dou-
present optim temperature for germination rang- ble irrigation, recruitment does not appear to be
ing between 5 C and 15 C (Escudero et al. 1997) improved. Even more, seedlings do present sig-
and soil’s water contents remain rather similar nificantly lower survival and growth in calcareous
along all the autumn-early spring period. This soils under double irrigation (see the significant
widely extended trait in Mediterranean plants water*soil substratum interaction).
(Thanos et al. 1992; Bell et al. 1993) must guar- In conclusion our results confirm our hypothesis
antee an optimum seedling survival. that gypsum edaphic specialists base their behav-
On the other hand, no significant differences in iour in a great extent on the ability to surpass
survival (in May, in June and even in September) extremely hard gypsum surface crusts (see Meyer
appear in any of the three studied plant species, 1986), although this seems a marginal trait as
two genuine gypsophytes and a gypsovag. This shown by the capability to grow on a complete
fact suggests that once the crust is surpassed the array of soils and the negative effect of the crust
chance for survival is maintained both for gypso- along the earlier development life stages of the two
phytes and gypsovags in the two experimental studied gypsophytes.
scenarios. As expected the most limiting factor for
survival is the summer drought as previously
reported under non-experimental field conditions Acknowledgements
(Escudero et al. 1999; 2000), even for other Iberian
gypsophytes such as Gypsophila struthium subsp. We thank Luis G. Benavides and Dr. Albert
struthium (Solá, personal communication). (‘Pizca’)for collaboration during fieldwork and
Although not included, logistic models for survival Francisco Moreno for providing valuable data.
in July (mortality peak) for the three studied spe- Dr. Albert kindly read a draft version of the paper.
cies were highly significant and included the seed- Financial support was provided by a MCYT
ling size as predictor with positive coefficients (see research project to A.E (REN2000-0254-P4-03).
also Escudero et al. 2000). Anyway, seedlings
better grow on disturbed crust in the case of L.
subulatum but also marginally in H. squamatum.
References
This fact points out the existence of a negative
effect of physical gypsum crust during the earlier Aranibar J.N., Anderson I.C., Rigrose S. and Macko S.A.
growth stages on both gypsophytes, suggesting 2003. Importance of nitrogen fixation in soil crusts of
that the population mediator role of the physical southern African arid ecosystems: acetylene reduction and
gypsum crust is not only confined to emergence stable isotope studies. J. Arid Environ. 54: 345 – 358.
Bell D.T., Plummer J.A. and Taylos S.K. 1993. Seed germina-
(see Bridges and Burnham 1980).
tion ecology in southwestern, Australia. Bot. Rev. 59: 24 – 73.
H. squamatum seems to conform a ‘‘refuge- Belnap J. and Gillette D.A. 1998. Vulnerability of desert soil
model’’ endemic (sensu Gankin and Major 1964; surfaces to wind erosion: impacts of soil texture and distur-
Proctor and Woodell 1975) because seeds of bance. J. Arid Environ. 39: 133 – 142.
H.squamatum germinate without significant dif- Belnap J. and Harper K.T. 1995. Influence of cryptobiotic soil
crusts on elemental content of tissue in two desert seed plants.
ferences in the three tested soils and can efficiently
Arid Soil Res. Rehabil. 9: 107 – 115.
grow on non-gypsum soils. This result would be in Belnap J., Rosentreter R., Leonard S., Kaltenecker J.H.,
agreement with Boukhris and Loissant (1970), Williams J. and Eldridge D. 2001. Biological Soil Crusts:
who pointed out the capability of gypsophytes to Ecology and Management. In: Peterson P. (ed.), U.S.
grow on several types of soil. However, we have Departament of the Interior T.R. 1730 – 2. pp.110.
Boukhris M. and Lossaint P. 1970. Sur la teneur en soufre de
found significant differences in survival and
quelques plantes gypsophiles de Tunisie. Oecologia Planta-
growth, being always higher on genuine gypsum rum 5: 345 – 354.
soils (Figure 2). This suggest that at least for this Bridges E.M. and Burnham C.P. 1980. Soils of the state of
gypsophyte, there is an intermediate strategy: it Bahrain. J. Soil Sci. 31: 689 – 707.
136

Campbell R.S. and Campbell I.F. 1938. Vegetation on gypsum crust composition in contrasted microsites within a mediter-
soils of the Jornada Plain, New Mexico. Ecology 19: 572 – rean semi-arid steppe. Soil Biol. Biochem. 34: 895 – 898.
577. McCullagh P. and Nelder J.A. 1989. Generalized Linear
Cannon H.L. 1971. The use of plant indicators in ground water Models. Chapman and Hall, New York.
surveys, geologic mapping, and mineral prospecting. Taxon Merlo M.E., Cabello J., Márquez M.M. and Alemán M.M.
20(2/3): 227 – 256. 1997. On the germination of plants of gypseous soils in
Cary J.W. and Evans D.D. 1974. Soil crusts. Technical Bulletin relation to the medium calcium content. 36th IAVS sym-
N. 214. University of Arizona, Tucson, Arizona. posium. Island and High Mountain Vegetation: Biodiver-
Duvigneaud P. and Denaeyer-De Smet S. 1966. Accumulation sity, Bioclimate and Conservation. Serie Informes 40: 197 –
du soufre dans quelques espèces gypsophiles d’Espagne. 206.
Bulletin Societe Royale Botanique du Belgique 99: 263 – 269. Meyer S.E. 1986. The ecology of gypsophile endemism in the
Duvigneaud P. and Denaeyer-De Smet S. 1968. Essai de clas- eastern Mojave desert. Ecology 67: 1303 – 1313.
sification chimique (éléments minéraux) des plantes gypsic- Meyer S.E. and Garcı́a-Moya E. 1989. Plant community pat-
oles du bassin de l’Ebre. Bulletin Societe Royale Botanique terns and soil moisture regime in gypsum grasslands of north
du Belgique 101: 279 – 291. central Mexico. J. Arid Environ. 16: 147 – 155.
Eldridge D.J. and Westoby M. 1991. Recruitment and survival Meyer S.E., Garcı́a-Moya E. and Lagunes-Espinoza L.C. 1992.
in Atriplex visicaria populations in semi-arid Western New Topographic and soil surface effects on gypsophile plant
South Wales, 1977 – 1987. Aust. J. Ecol. 16: 309 – 314. community patterns in central Mexico. J. Veg. Sci. 3: 429 – 438.
Eldridge D.J., Zaady E. and Schachack M. 2000. Infiltration Monturiol F. and del Olmo L. 1990. Mapa de asociaciones de
through three contrasting biological soil crusts in patterned suelos de la Comunidad de Madrid. Escala 1:200.000. Escala
landscapes in the contrasting biological soil crusts in pat- 1:200.000. Consejo Superior de Investigaciones Cientı́ficas,
terned landscapes in the Negev, Israel. Catena 40: 323 – 336. Madrid.
Escudero A., Carnes L. and Pérez-Garcı́a F. 1997. Germination Mota J.F., Solá A.J., Dana E. and Jiménez-Sánchez M.L. 2003.
of gypsophytes and gypsovags in semiarid central Spain. Plant succession in abandoned quarries in southeast Spain.
J. Arid Environ. 36: 487 – 497. Phytocenologia 33: 13 – 28.
Escudero A., Iriondo J.M., Olano J.M., Rubio A. and Somo- Parsons P.F. 1976. Gypsophily in plants a review. Am. Midland
linos R. 2000. Factors affecting establishment of a gypsoph- Nat. 96: 1 – 20.
ite: the case of Lepidium subulatum. Am. J. Bot. 87: 861 – 871. Powell A.M. and Turner B.L. 1977. Aspects of the plant biol-
Escudero A., Somolinos R.C., Olano J.M. and Rubio A. 1999. ogy of the gypsum outcrops of the Chihuahuan desert. In:
Factors controlling the establishment of Helianthemum Wauer R.H. and Riskind D.H. (eds), Transactions Symp.
squamatum (L.) Dum., an endemic gypsophile of semi-arid Biol. Resources of the Chihuahuan Desert. U.S. Dep. Inte-
Spain. J. Ecol. 87: 290 – 302. rior, Natl. Park Serv., Trans. Proc. Series 3, pp. 315 – 325.
European Community 1992. Council Directive 92/43/EEC of Prasse R. and Bornkann R. 2000. Effect of microbiotic soil
21 May 1992 on the Conservation of Natural Habitats and of surface crusts on the emergence of vascular plants. Plant
Wild Fauna and Flora. European Community, Brussels, Ecol. 150: 65 – 75.
Belgium. Proctor J. and Woodell S.J. 1975. The ecology of serpentine
Fowler N. 1988. What is a safe site? Neighbor, litter, germi- soil. Adv. Ecol. Res. 9: 255 – 284.
nation date, and patch effects Ecology 69: 947 – 961. Rivas-Martı́nez S. and Costa M. 1970. Comunidades gipsı́colas
Gankin R. and Major J. 1964. Arctostaphylos myrtifolia, its del centro de España. Anales del Instituto de Botánica
biology and relationship to the problem of endemism. Ecol- Cavanilles 27: 195 – 223.
ogy 45: 792 – 808. Romão R.L. 2003. Estructura espacial de comunidades de
Gómez-Campo C. 1987. Libro Rojo de especies vegetales gipsófitos: Interacciones bióticas y constricciones abióticas.
amenazadas de España Peninsular e Islas Baleares. M.A.P.A. PhD Thesis, Universidad Politécnica de Madrid.
ICONA, Madrid, Spain. Rubio A., Escudero A., Molina A. and Pajarón S. 1994. Una
Hanks R.J. and Thorp F.C. 1956. Wheat seedling emergence as nueva aproximación metodológica al estudio de los gipsófitos
related to soil moisture content, bulk density, oxygen diffu- ibéricos. Mediterranea 13: 15 – 26.
sion rate, and crust stregth. Soil Sci. Soc. America Proc. Sylla D. 1987.Effect of microphytic crust on emergence of range
20(3): 307 – 310. grasses. MSc- Thesis, School of renewable natural resources,
Harper J.L., Willians J.J. and Sagar G.R. 1965. The behavior of University of Arizona.
seeds in the soil I. The heterogeneity of the soil surface and its Tarazona T., Barreno E., Crespo A. and Dı́az-Llanos J. 1980.
role in determining the establishment of plant from seed. J. Estudio estadı́stica de la vegetación liquénica de los yesos del
Ecol. 53: 273 – 286. centro España. Anales del Instituto Nacional de Investigac-
Johansen J.R. 1993. Cryptogamic crust of semiarid and arid iones Agrarias 4: 139 – 155. (In spanish).
lands of North American. J. Phycol. 29: 140 – 147. Thanos C.A., Georghiou K., Kadius C. and Pantazi C. 1992.
Johnston I.M. 1941. Gypsophyly among Mexican desert plants. Cistaceae: a plant family with hard seeds. Isr. J. Bot. 41:
J. Arnold Arboretum 22: 145 – 170. 251 – 263.
Loidi J. and Fernández-González F. 1994. The gypsophilous Verheye W.H. and Boyadgiev T.G. 1997. Evaluating the land
scrub communities of the Ebro valley (Spain). Phytocoeno- use potential of gypsiferous soils from field pedogenic char-
logia 24: 383 – 399. acteristics. Soil Use Manage. 13: 97 – 103.
Maestre F.T., Huesca M., Zaady E., Bautista S. and Cortina J. Watson A. 1979. Gypsum crusts in deserts. J. Arid Environ. 2:
2002. Infiltration, penetration resistance and microphytic 3 – 20.
137

West N.E. 1990. Structure and function of microphytic soil Zaady E., Gutterman Y. and Boeken B. 1997. The germi-
crusts in wildland ecosystems of arid to semi-arid regions. nation of mucilaginous seeds of Plantago coronopus,
Adv. Ecol. Res. 20: 179 – 223. Reboudia pinnata, and Carrichtera annua on cyanobacterial
Zaady E., Groffman P. and Shachak M. 1998. Nitrogen fixation soil crusts from the Negev Desert. Plant Soil 190: 247 –
in macro- and microphytic patches in the Negev desert. Soil 252.
Biol. Biochem. 30: 449 – 454.

You might also like